diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,85668 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.9999591319628918,
+  "eval_steps": 500,
+  "global_step": 12234,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 10.763204990230978,
+      "learning_rate": 2.7173913043478262e-08,
+      "loss": 1.9124,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 10.252056662172615,
+      "learning_rate": 5.4347826086956524e-08,
+      "loss": 1.2475,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 12.5031240421653,
+      "learning_rate": 8.152173913043479e-08,
+      "loss": 1.8097,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 11.19022777352965,
+      "learning_rate": 1.0869565217391305e-07,
+      "loss": 1.7408,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.74911389200089,
+      "learning_rate": 1.3586956521739132e-07,
+      "loss": 1.3384,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 16.414301597091637,
+      "learning_rate": 1.6304347826086958e-07,
+      "loss": 2.1608,
+      "step": 6
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 9.441165175742352,
+      "learning_rate": 1.9021739130434786e-07,
+      "loss": 1.8799,
+      "step": 7
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 9.77584594887577,
+      "learning_rate": 2.173913043478261e-07,
+      "loss": 1.8447,
+      "step": 8
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 10.111184232798522,
+      "learning_rate": 2.445652173913044e-07,
+      "loss": 1.4988,
+      "step": 9
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 9.567135461407071,
+      "learning_rate": 2.7173913043478264e-07,
+      "loss": 1.5362,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 10.00231116510922,
+      "learning_rate": 2.989130434782609e-07,
+      "loss": 1.2534,
+      "step": 11
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.806732616267708,
+      "learning_rate": 3.2608695652173915e-07,
+      "loss": 1.4756,
+      "step": 12
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 10.617002504828708,
+      "learning_rate": 3.532608695652174e-07,
+      "loss": 1.7337,
+      "step": 13
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 13.54765846286333,
+      "learning_rate": 3.804347826086957e-07,
+      "loss": 2.1608,
+      "step": 14
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 11.123744709657245,
+      "learning_rate": 4.0760869565217393e-07,
+      "loss": 1.8026,
+      "step": 15
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.142452879045726,
+      "learning_rate": 4.347826086956522e-07,
+      "loss": 1.38,
+      "step": 16
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 10.641416240703439,
+      "learning_rate": 4.6195652173913045e-07,
+      "loss": 1.439,
+      "step": 17
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 9.110065588139014,
+      "learning_rate": 4.891304347826088e-07,
+      "loss": 1.8888,
+      "step": 18
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 10.116752917147823,
+      "learning_rate": 5.16304347826087e-07,
+      "loss": 1.4248,
+      "step": 19
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 10.439796920171236,
+      "learning_rate": 5.434782608695653e-07,
+      "loss": 1.5432,
+      "step": 20
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.992746653280529,
+      "learning_rate": 5.706521739130435e-07,
+      "loss": 1.7672,
+      "step": 21
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 11.46654871784486,
+      "learning_rate": 5.978260869565218e-07,
+      "loss": 1.66,
+      "step": 22
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 10.103379665707928,
+      "learning_rate": 6.25e-07,
+      "loss": 1.5059,
+      "step": 23
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.922521662725512,
+      "learning_rate": 6.521739130434783e-07,
+      "loss": 1.3644,
+      "step": 24
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 9.482787698784223,
+      "learning_rate": 6.793478260869566e-07,
+      "loss": 1.7136,
+      "step": 25
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 10.131149963147784,
+      "learning_rate": 7.065217391304348e-07,
+      "loss": 1.5865,
+      "step": 26
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.553296695300296,
+      "learning_rate": 7.336956521739132e-07,
+      "loss": 1.0809,
+      "step": 27
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.447837089466901,
+      "learning_rate": 7.608695652173914e-07,
+      "loss": 1.3542,
+      "step": 28
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.646186450438408,
+      "learning_rate": 7.880434782608697e-07,
+      "loss": 1.3759,
+      "step": 29
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 11.363825771345514,
+      "learning_rate": 8.152173913043479e-07,
+      "loss": 1.3261,
+      "step": 30
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.707159536720378,
+      "learning_rate": 8.423913043478261e-07,
+      "loss": 1.2247,
+      "step": 31
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.5094610599711125,
+      "learning_rate": 8.695652173913044e-07,
+      "loss": 1.4841,
+      "step": 32
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.783281889567137,
+      "learning_rate": 8.967391304347826e-07,
+      "loss": 1.5589,
+      "step": 33
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 10.324049872893895,
+      "learning_rate": 9.239130434782609e-07,
+      "loss": 1.402,
+      "step": 34
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.307774271557626,
+      "learning_rate": 9.510869565217393e-07,
+      "loss": 1.4388,
+      "step": 35
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.48905313529863,
+      "learning_rate": 9.782608695652175e-07,
+      "loss": 1.5702,
+      "step": 36
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.834200227156316,
+      "learning_rate": 1.0054347826086958e-06,
+      "loss": 1.3819,
+      "step": 37
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.867659654983689,
+      "learning_rate": 1.032608695652174e-06,
+      "loss": 1.5146,
+      "step": 38
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.104395892275559,
+      "learning_rate": 1.0597826086956523e-06,
+      "loss": 1.2098,
+      "step": 39
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.485504914038025,
+      "learning_rate": 1.0869565217391306e-06,
+      "loss": 1.3955,
+      "step": 40
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 10.13078856345608,
+      "learning_rate": 1.1141304347826088e-06,
+      "loss": 1.4476,
+      "step": 41
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.66501688643045,
+      "learning_rate": 1.141304347826087e-06,
+      "loss": 1.1798,
+      "step": 42
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.056438781274291,
+      "learning_rate": 1.1684782608695653e-06,
+      "loss": 1.1171,
+      "step": 43
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 9.524989559260273,
+      "learning_rate": 1.1956521739130436e-06,
+      "loss": 1.2799,
+      "step": 44
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.207602666230073,
+      "learning_rate": 1.2228260869565218e-06,
+      "loss": 1.1379,
+      "step": 45
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 10.741407362172229,
+      "learning_rate": 1.25e-06,
+      "loss": 1.5437,
+      "step": 46
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.112828564926632,
+      "learning_rate": 1.2771739130434786e-06,
+      "loss": 0.9981,
+      "step": 47
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.503495238241442,
+      "learning_rate": 1.3043478260869566e-06,
+      "loss": 1.0219,
+      "step": 48
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.465625542087716,
+      "learning_rate": 1.3315217391304349e-06,
+      "loss": 1.4285,
+      "step": 49
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.756922552940579,
+      "learning_rate": 1.3586956521739131e-06,
+      "loss": 1.298,
+      "step": 50
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.191522642983566,
+      "learning_rate": 1.3858695652173914e-06,
+      "loss": 1.2136,
+      "step": 51
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.345904517953232,
+      "learning_rate": 1.4130434782608697e-06,
+      "loss": 1.2073,
+      "step": 52
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.782042045608577,
+      "learning_rate": 1.440217391304348e-06,
+      "loss": 1.197,
+      "step": 53
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.103653842504592,
+      "learning_rate": 1.4673913043478264e-06,
+      "loss": 1.1927,
+      "step": 54
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 9.288565917912903,
+      "learning_rate": 1.4945652173913044e-06,
+      "loss": 1.728,
+      "step": 55
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.496033995601524,
+      "learning_rate": 1.521739130434783e-06,
+      "loss": 1.1772,
+      "step": 56
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.70760468187945,
+      "learning_rate": 1.548913043478261e-06,
+      "loss": 1.0784,
+      "step": 57
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.164233229137625,
+      "learning_rate": 1.5760869565217394e-06,
+      "loss": 0.8166,
+      "step": 58
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.153545718589251,
+      "learning_rate": 1.6032608695652175e-06,
+      "loss": 1.0213,
+      "step": 59
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 9.224145677156148,
+      "learning_rate": 1.6304347826086957e-06,
+      "loss": 1.2201,
+      "step": 60
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.929420117992498,
+      "learning_rate": 1.657608695652174e-06,
+      "loss": 1.0981,
+      "step": 61
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.832090681262447,
+      "learning_rate": 1.6847826086956522e-06,
+      "loss": 1.1829,
+      "step": 62
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.827788334408938,
+      "learning_rate": 1.7119565217391307e-06,
+      "loss": 1.1005,
+      "step": 63
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.5134565802844415,
+      "learning_rate": 1.7391304347826088e-06,
+      "loss": 0.8607,
+      "step": 64
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.975177473268082,
+      "learning_rate": 1.7663043478260872e-06,
+      "loss": 1.0803,
+      "step": 65
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.425166183961684,
+      "learning_rate": 1.7934782608695653e-06,
+      "loss": 1.1596,
+      "step": 66
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.723229692609811,
+      "learning_rate": 1.8206521739130437e-06,
+      "loss": 1.2265,
+      "step": 67
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.7467229371397694,
+      "learning_rate": 1.8478260869565218e-06,
+      "loss": 0.7996,
+      "step": 68
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.146784934298589,
+      "learning_rate": 1.8750000000000003e-06,
+      "loss": 0.3544,
+      "step": 69
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 10.983495052283711,
+      "learning_rate": 1.9021739130434785e-06,
+      "loss": 1.5448,
+      "step": 70
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.528084910725228,
+      "learning_rate": 1.9293478260869568e-06,
+      "loss": 0.9641,
+      "step": 71
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.772154961625803,
+      "learning_rate": 1.956521739130435e-06,
+      "loss": 0.6387,
+      "step": 72
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.372610251132539,
+      "learning_rate": 1.9836956521739133e-06,
+      "loss": 0.7883,
+      "step": 73
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.3183594429603716,
+      "learning_rate": 2.0108695652173916e-06,
+      "loss": 0.6512,
+      "step": 74
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.804934889637244,
+      "learning_rate": 2.03804347826087e-06,
+      "loss": 0.8766,
+      "step": 75
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.42791079405579,
+      "learning_rate": 2.065217391304348e-06,
+      "loss": 0.8556,
+      "step": 76
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.519309928339856,
+      "learning_rate": 2.0923913043478263e-06,
+      "loss": 1.5128,
+      "step": 77
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.595715930891681,
+      "learning_rate": 2.1195652173913046e-06,
+      "loss": 1.3523,
+      "step": 78
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.709132076023693,
+      "learning_rate": 2.146739130434783e-06,
+      "loss": 0.8167,
+      "step": 79
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.721035794036643,
+      "learning_rate": 2.173913043478261e-06,
+      "loss": 0.8613,
+      "step": 80
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.955864530501376,
+      "learning_rate": 2.2010869565217394e-06,
+      "loss": 1.4923,
+      "step": 81
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.226117141588304,
+      "learning_rate": 2.2282608695652176e-06,
+      "loss": 1.0378,
+      "step": 82
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.578554182689265,
+      "learning_rate": 2.255434782608696e-06,
+      "loss": 1.2069,
+      "step": 83
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.854813831700824,
+      "learning_rate": 2.282608695652174e-06,
+      "loss": 0.8649,
+      "step": 84
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.154347006791653,
+      "learning_rate": 2.3097826086956524e-06,
+      "loss": 0.8471,
+      "step": 85
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.017133522659245,
+      "learning_rate": 2.3369565217391307e-06,
+      "loss": 1.1938,
+      "step": 86
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.026983949592643,
+      "learning_rate": 2.364130434782609e-06,
+      "loss": 0.8922,
+      "step": 87
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.405981979152875,
+      "learning_rate": 2.391304347826087e-06,
+      "loss": 1.4834,
+      "step": 88
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.5999817340984075,
+      "learning_rate": 2.4184782608695654e-06,
+      "loss": 1.0044,
+      "step": 89
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.162604562149298,
+      "learning_rate": 2.4456521739130437e-06,
+      "loss": 1.3329,
+      "step": 90
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.074359325975014,
+      "learning_rate": 2.472826086956522e-06,
+      "loss": 1.5377,
+      "step": 91
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.41639145189624,
+      "learning_rate": 2.5e-06,
+      "loss": 1.0232,
+      "step": 92
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.4783028052120395,
+      "learning_rate": 2.5271739130434785e-06,
+      "loss": 0.955,
+      "step": 93
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.739755004143212,
+      "learning_rate": 2.554347826086957e-06,
+      "loss": 0.753,
+      "step": 94
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.632644951561571,
+      "learning_rate": 2.581521739130435e-06,
+      "loss": 1.1089,
+      "step": 95
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.707986315608768,
+      "learning_rate": 2.6086956521739132e-06,
+      "loss": 1.6927,
+      "step": 96
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.928093931905623,
+      "learning_rate": 2.6358695652173915e-06,
+      "loss": 0.7765,
+      "step": 97
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.761801614748697,
+      "learning_rate": 2.6630434782608698e-06,
+      "loss": 1.3454,
+      "step": 98
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.71085682910734,
+      "learning_rate": 2.6902173913043476e-06,
+      "loss": 1.2272,
+      "step": 99
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.743239729296844,
+      "learning_rate": 2.7173913043478263e-06,
+      "loss": 0.8723,
+      "step": 100
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.8842539643647065,
+      "learning_rate": 2.7445652173913045e-06,
+      "loss": 1.1771,
+      "step": 101
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.6027040012838265,
+      "learning_rate": 2.771739130434783e-06,
+      "loss": 1.1343,
+      "step": 102
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.93038824774127,
+      "learning_rate": 2.7989130434782615e-06,
+      "loss": 0.8708,
+      "step": 103
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.992654839808304,
+      "learning_rate": 2.8260869565217393e-06,
+      "loss": 0.7075,
+      "step": 104
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.638844600072942,
+      "learning_rate": 2.8532608695652176e-06,
+      "loss": 0.8676,
+      "step": 105
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.919202754473433,
+      "learning_rate": 2.880434782608696e-06,
+      "loss": 1.1083,
+      "step": 106
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.808777508326425,
+      "learning_rate": 2.9076086956521745e-06,
+      "loss": 0.8974,
+      "step": 107
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.049341085789842,
+      "learning_rate": 2.9347826086956528e-06,
+      "loss": 0.4606,
+      "step": 108
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.734898712066466,
+      "learning_rate": 2.9619565217391306e-06,
+      "loss": 0.7338,
+      "step": 109
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.308385711610393,
+      "learning_rate": 2.989130434782609e-06,
+      "loss": 1.2931,
+      "step": 110
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.37792345474363,
+      "learning_rate": 3.016304347826087e-06,
+      "loss": 1.3357,
+      "step": 111
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.207211713951564,
+      "learning_rate": 3.043478260869566e-06,
+      "loss": 0.4874,
+      "step": 112
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.291411813399872,
+      "learning_rate": 3.0706521739130436e-06,
+      "loss": 1.3978,
+      "step": 113
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.685180426322534,
+      "learning_rate": 3.097826086956522e-06,
+      "loss": 0.8724,
+      "step": 114
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.104442131372095,
+      "learning_rate": 3.125e-06,
+      "loss": 1.0527,
+      "step": 115
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.478605850436171,
+      "learning_rate": 3.152173913043479e-06,
+      "loss": 1.4413,
+      "step": 116
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.531709196936938,
+      "learning_rate": 3.179347826086957e-06,
+      "loss": 1.6845,
+      "step": 117
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.4759958206666175,
+      "learning_rate": 3.206521739130435e-06,
+      "loss": 1.0112,
+      "step": 118
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.049621112602324,
+      "learning_rate": 3.233695652173913e-06,
+      "loss": 1.3741,
+      "step": 119
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.5817184153763173,
+      "learning_rate": 3.2608695652173914e-06,
+      "loss": 0.7584,
+      "step": 120
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.404142910694771,
+      "learning_rate": 3.28804347826087e-06,
+      "loss": 1.1528,
+      "step": 121
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.906408150238798,
+      "learning_rate": 3.315217391304348e-06,
+      "loss": 0.512,
+      "step": 122
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.488336599440538,
+      "learning_rate": 3.3423913043478262e-06,
+      "loss": 1.018,
+      "step": 123
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.981920682664463,
+      "learning_rate": 3.3695652173913045e-06,
+      "loss": 0.8915,
+      "step": 124
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.05588044254937,
+      "learning_rate": 3.396739130434783e-06,
+      "loss": 0.8899,
+      "step": 125
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.854430041120995,
+      "learning_rate": 3.4239130434782614e-06,
+      "loss": 1.0416,
+      "step": 126
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 10.556756679163271,
+      "learning_rate": 3.4510869565217393e-06,
+      "loss": 1.6121,
+      "step": 127
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.0566625292176495,
+      "learning_rate": 3.4782608695652175e-06,
+      "loss": 1.4485,
+      "step": 128
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.883945959657133,
+      "learning_rate": 3.5054347826086958e-06,
+      "loss": 0.6082,
+      "step": 129
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.701849857905996,
+      "learning_rate": 3.5326086956521745e-06,
+      "loss": 1.3073,
+      "step": 130
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.27587282635949,
+      "learning_rate": 3.5597826086956527e-06,
+      "loss": 0.6537,
+      "step": 131
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.3323783890401635,
+      "learning_rate": 3.5869565217391305e-06,
+      "loss": 0.5372,
+      "step": 132
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.843012301008066,
+      "learning_rate": 3.614130434782609e-06,
+      "loss": 1.243,
+      "step": 133
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.200673205505264,
+      "learning_rate": 3.6413043478260875e-06,
+      "loss": 1.2031,
+      "step": 134
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.178859040391603,
+      "learning_rate": 3.6684782608695657e-06,
+      "loss": 0.7575,
+      "step": 135
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.493390921146402,
+      "learning_rate": 3.6956521739130436e-06,
+      "loss": 1.2118,
+      "step": 136
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.517598073215942,
+      "learning_rate": 3.722826086956522e-06,
+      "loss": 1.0804,
+      "step": 137
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.872001208747307,
+      "learning_rate": 3.7500000000000005e-06,
+      "loss": 0.8877,
+      "step": 138
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.751199032083687,
+      "learning_rate": 3.7771739130434788e-06,
+      "loss": 1.0684,
+      "step": 139
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.814160757707059,
+      "learning_rate": 3.804347826086957e-06,
+      "loss": 1.4598,
+      "step": 140
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.7161018974223925,
+      "learning_rate": 3.831521739130435e-06,
+      "loss": 0.2583,
+      "step": 141
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.528220495890977,
+      "learning_rate": 3.8586956521739136e-06,
+      "loss": 0.9408,
+      "step": 142
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.7952554346766822,
+      "learning_rate": 3.885869565217392e-06,
+      "loss": 0.4543,
+      "step": 143
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.48531466674868,
+      "learning_rate": 3.91304347826087e-06,
+      "loss": 0.8372,
+      "step": 144
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.993848127713756,
+      "learning_rate": 3.9402173913043475e-06,
+      "loss": 1.1149,
+      "step": 145
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.60199739787131,
+      "learning_rate": 3.967391304347827e-06,
+      "loss": 1.4652,
+      "step": 146
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.180849053327101,
+      "learning_rate": 3.994565217391305e-06,
+      "loss": 0.3403,
+      "step": 147
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.675303114346352,
+      "learning_rate": 4.021739130434783e-06,
+      "loss": 1.1365,
+      "step": 148
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.618577228665263,
+      "learning_rate": 4.048913043478261e-06,
+      "loss": 1.067,
+      "step": 149
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.078694781710514,
+      "learning_rate": 4.07608695652174e-06,
+      "loss": 1.0104,
+      "step": 150
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.767941978563566,
+      "learning_rate": 4.103260869565218e-06,
+      "loss": 1.3224,
+      "step": 151
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.885594442243772,
+      "learning_rate": 4.130434782608696e-06,
+      "loss": 0.7112,
+      "step": 152
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 10.457674878448332,
+      "learning_rate": 4.157608695652174e-06,
+      "loss": 1.8655,
+      "step": 153
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.904533607873724,
+      "learning_rate": 4.184782608695653e-06,
+      "loss": 1.4322,
+      "step": 154
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.049829569078577,
+      "learning_rate": 4.211956521739131e-06,
+      "loss": 1.1242,
+      "step": 155
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.5836948972226566,
+      "learning_rate": 4.239130434782609e-06,
+      "loss": 0.6328,
+      "step": 156
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.511416011994131,
+      "learning_rate": 4.2663043478260874e-06,
+      "loss": 0.7455,
+      "step": 157
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.415113124345645,
+      "learning_rate": 4.293478260869566e-06,
+      "loss": 0.8246,
+      "step": 158
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.563836391840908,
+      "learning_rate": 4.320652173913044e-06,
+      "loss": 1.2982,
+      "step": 159
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.255558685888354,
+      "learning_rate": 4.347826086956522e-06,
+      "loss": 0.922,
+      "step": 160
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.8007257579161235,
+      "learning_rate": 4.3750000000000005e-06,
+      "loss": 0.6111,
+      "step": 161
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.138614187559548,
+      "learning_rate": 4.402173913043479e-06,
+      "loss": 1.0949,
+      "step": 162
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.9221676977686535,
+      "learning_rate": 4.429347826086957e-06,
+      "loss": 0.909,
+      "step": 163
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.942921539191711,
+      "learning_rate": 4.456521739130435e-06,
+      "loss": 0.7227,
+      "step": 164
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.900295438543103,
+      "learning_rate": 4.4836956521739135e-06,
+      "loss": 0.6749,
+      "step": 165
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.789778017522469,
+      "learning_rate": 4.510869565217392e-06,
+      "loss": 0.9947,
+      "step": 166
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.048409907918103,
+      "learning_rate": 4.53804347826087e-06,
+      "loss": 1.2421,
+      "step": 167
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.367898594382595,
+      "learning_rate": 4.565217391304348e-06,
+      "loss": 0.5863,
+      "step": 168
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.038484443485752,
+      "learning_rate": 4.5923913043478265e-06,
+      "loss": 0.8006,
+      "step": 169
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.980449471454865,
+      "learning_rate": 4.619565217391305e-06,
+      "loss": 1.2171,
+      "step": 170
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.291826025186352,
+      "learning_rate": 4.646739130434783e-06,
+      "loss": 0.969,
+      "step": 171
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.18831539912512,
+      "learning_rate": 4.673913043478261e-06,
+      "loss": 0.6138,
+      "step": 172
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.608150004360503,
+      "learning_rate": 4.7010869565217396e-06,
+      "loss": 0.7879,
+      "step": 173
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.410996394740663,
+      "learning_rate": 4.728260869565218e-06,
+      "loss": 0.4698,
+      "step": 174
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.1740205070337355,
+      "learning_rate": 4.755434782608696e-06,
+      "loss": 1.0853,
+      "step": 175
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.13726770207731,
+      "learning_rate": 4.782608695652174e-06,
+      "loss": 0.8261,
+      "step": 176
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.361680179795122,
+      "learning_rate": 4.809782608695653e-06,
+      "loss": 1.1934,
+      "step": 177
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.67123136273917,
+      "learning_rate": 4.836956521739131e-06,
+      "loss": 1.1739,
+      "step": 178
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.2229598582844945,
+      "learning_rate": 4.864130434782609e-06,
+      "loss": 1.0926,
+      "step": 179
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.7046608872567734,
+      "learning_rate": 4.891304347826087e-06,
+      "loss": 0.8426,
+      "step": 180
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.198449576398156,
+      "learning_rate": 4.918478260869566e-06,
+      "loss": 1.1743,
+      "step": 181
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.525933837362583,
+      "learning_rate": 4.945652173913044e-06,
+      "loss": 0.8965,
+      "step": 182
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.102006691809892,
+      "learning_rate": 4.972826086956522e-06,
+      "loss": 0.4219,
+      "step": 183
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.716964784019454,
+      "learning_rate": 5e-06,
+      "loss": 1.3908,
+      "step": 184
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.812408168037797,
+      "learning_rate": 5.027173913043478e-06,
+      "loss": 0.9765,
+      "step": 185
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.7300425111293345,
+      "learning_rate": 5.054347826086957e-06,
+      "loss": 0.7135,
+      "step": 186
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.29453599820772,
+      "learning_rate": 5.081521739130435e-06,
+      "loss": 0.6838,
+      "step": 187
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.3872345244405215,
+      "learning_rate": 5.108695652173914e-06,
+      "loss": 0.7755,
+      "step": 188
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.350885267174286,
+      "learning_rate": 5.135869565217392e-06,
+      "loss": 0.7441,
+      "step": 189
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.705837768030783,
+      "learning_rate": 5.16304347826087e-06,
+      "loss": 0.9734,
+      "step": 190
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.031565248567177,
+      "learning_rate": 5.190217391304348e-06,
+      "loss": 0.521,
+      "step": 191
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 9.231182683688848,
+      "learning_rate": 5.2173913043478265e-06,
+      "loss": 1.2869,
+      "step": 192
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 8.081212572731191,
+      "learning_rate": 5.244565217391306e-06,
+      "loss": 1.1,
+      "step": 193
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.923947857423653,
+      "learning_rate": 5.271739130434783e-06,
+      "loss": 0.1881,
+      "step": 194
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.654283650048554,
+      "learning_rate": 5.298913043478261e-06,
+      "loss": 0.9192,
+      "step": 195
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.772455621286309,
+      "learning_rate": 5.3260869565217395e-06,
+      "loss": 0.7107,
+      "step": 196
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.8759240797330703,
+      "learning_rate": 5.353260869565218e-06,
+      "loss": 0.3039,
+      "step": 197
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.517060510758465,
+      "learning_rate": 5.380434782608695e-06,
+      "loss": 0.8079,
+      "step": 198
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.702676517665291,
+      "learning_rate": 5.407608695652174e-06,
+      "loss": 1.3456,
+      "step": 199
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.5636098368903895,
+      "learning_rate": 5.4347826086956525e-06,
+      "loss": 0.9768,
+      "step": 200
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.958333903626813,
+      "learning_rate": 5.461956521739132e-06,
+      "loss": 0.7762,
+      "step": 201
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.581754010001346,
+      "learning_rate": 5.489130434782609e-06,
+      "loss": 0.7291,
+      "step": 202
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.923986836029897,
+      "learning_rate": 5.516304347826087e-06,
+      "loss": 0.9687,
+      "step": 203
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.081234676189072,
+      "learning_rate": 5.543478260869566e-06,
+      "loss": 0.8964,
+      "step": 204
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.390859736544384,
+      "learning_rate": 5.570652173913044e-06,
+      "loss": 0.3286,
+      "step": 205
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.417543485088989,
+      "learning_rate": 5.597826086956523e-06,
+      "loss": 1.6375,
+      "step": 206
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.992210187529088,
+      "learning_rate": 5.625e-06,
+      "loss": 0.8719,
+      "step": 207
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.392475147231553,
+      "learning_rate": 5.652173913043479e-06,
+      "loss": 0.9143,
+      "step": 208
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.174327311086821,
+      "learning_rate": 5.679347826086957e-06,
+      "loss": 0.6731,
+      "step": 209
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.743161076840916,
+      "learning_rate": 5.706521739130435e-06,
+      "loss": 0.6002,
+      "step": 210
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.316271470573057,
+      "learning_rate": 5.733695652173914e-06,
+      "loss": 0.9979,
+      "step": 211
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.871275447806914,
+      "learning_rate": 5.760869565217392e-06,
+      "loss": 0.8193,
+      "step": 212
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.000709896186412,
+      "learning_rate": 5.78804347826087e-06,
+      "loss": 0.8606,
+      "step": 213
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.24528787539437,
+      "learning_rate": 5.815217391304349e-06,
+      "loss": 0.7,
+      "step": 214
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.942273062178936,
+      "learning_rate": 5.842391304347826e-06,
+      "loss": 0.9734,
+      "step": 215
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.013164920105653,
+      "learning_rate": 5.8695652173913055e-06,
+      "loss": 0.9056,
+      "step": 216
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.385221534082321,
+      "learning_rate": 5.896739130434783e-06,
+      "loss": 0.6014,
+      "step": 217
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 8.202981310804969,
+      "learning_rate": 5.923913043478261e-06,
+      "loss": 1.2524,
+      "step": 218
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.673247533397,
+      "learning_rate": 5.95108695652174e-06,
+      "loss": 1.2265,
+      "step": 219
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.655756570467079,
+      "learning_rate": 5.978260869565218e-06,
+      "loss": 1.0721,
+      "step": 220
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.234835376562726,
+      "learning_rate": 6.005434782608696e-06,
+      "loss": 1.1082,
+      "step": 221
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.974427392129233,
+      "learning_rate": 6.032608695652174e-06,
+      "loss": 1.0017,
+      "step": 222
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.916925594285953,
+      "learning_rate": 6.0597826086956525e-06,
+      "loss": 1.1169,
+      "step": 223
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.069754222166293,
+      "learning_rate": 6.086956521739132e-06,
+      "loss": 1.1844,
+      "step": 224
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.5925567538816,
+      "learning_rate": 6.114130434782609e-06,
+      "loss": 1.3222,
+      "step": 225
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.406797145497836,
+      "learning_rate": 6.141304347826087e-06,
+      "loss": 0.7072,
+      "step": 226
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.960839328635132,
+      "learning_rate": 6.1684782608695655e-06,
+      "loss": 1.0292,
+      "step": 227
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.2287351299289333,
+      "learning_rate": 6.195652173913044e-06,
+      "loss": 0.4187,
+      "step": 228
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.6174718743187375,
+      "learning_rate": 6.222826086956523e-06,
+      "loss": 0.9114,
+      "step": 229
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.422614534330612,
+      "learning_rate": 6.25e-06,
+      "loss": 0.8558,
+      "step": 230
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.2228528011626585,
+      "learning_rate": 6.2771739130434786e-06,
+      "loss": 0.8636,
+      "step": 231
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.503777571503567,
+      "learning_rate": 6.304347826086958e-06,
+      "loss": 0.8099,
+      "step": 232
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.906673443489322,
+      "learning_rate": 6.331521739130435e-06,
+      "loss": 0.7918,
+      "step": 233
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.857207855313775,
+      "learning_rate": 6.358695652173914e-06,
+      "loss": 1.8075,
+      "step": 234
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 8.148563326223139,
+      "learning_rate": 6.385869565217392e-06,
+      "loss": 1.1465,
+      "step": 235
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.4249316447648654,
+      "learning_rate": 6.41304347826087e-06,
+      "loss": 1.1498,
+      "step": 236
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.644600409673028,
+      "learning_rate": 6.440217391304349e-06,
+      "loss": 0.9854,
+      "step": 237
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.008422155861959,
+      "learning_rate": 6.467391304347826e-06,
+      "loss": 0.4489,
+      "step": 238
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.686369840867863,
+      "learning_rate": 6.4945652173913055e-06,
+      "loss": 1.2196,
+      "step": 239
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.389822453643525,
+      "learning_rate": 6.521739130434783e-06,
+      "loss": 1.5096,
+      "step": 240
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.40452020190104,
+      "learning_rate": 6.548913043478261e-06,
+      "loss": 1.0686,
+      "step": 241
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.8513034466412925,
+      "learning_rate": 6.57608695652174e-06,
+      "loss": 0.5596,
+      "step": 242
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.0985386159169215,
+      "learning_rate": 6.603260869565218e-06,
+      "loss": 0.8382,
+      "step": 243
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.486981558596191,
+      "learning_rate": 6.630434782608696e-06,
+      "loss": 1.1326,
+      "step": 244
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.04189008786626,
+      "learning_rate": 6.657608695652175e-06,
+      "loss": 1.0986,
+      "step": 245
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.826697566070404,
+      "learning_rate": 6.6847826086956524e-06,
+      "loss": 1.224,
+      "step": 246
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.079863511831709,
+      "learning_rate": 6.7119565217391315e-06,
+      "loss": 0.9992,
+      "step": 247
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.103408839897494,
+      "learning_rate": 6.739130434782609e-06,
+      "loss": 0.71,
+      "step": 248
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.1228953020068,
+      "learning_rate": 6.766304347826087e-06,
+      "loss": 0.9402,
+      "step": 249
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 9.976070490177033,
+      "learning_rate": 6.793478260869566e-06,
+      "loss": 1.4335,
+      "step": 250
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.589589798539868,
+      "learning_rate": 6.820652173913044e-06,
+      "loss": 0.7684,
+      "step": 251
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.862103581613901,
+      "learning_rate": 6.847826086956523e-06,
+      "loss": 1.3673,
+      "step": 252
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.107711867291614,
+      "learning_rate": 6.875e-06,
+      "loss": 0.7849,
+      "step": 253
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.781420070769819,
+      "learning_rate": 6.9021739130434785e-06,
+      "loss": 1.1199,
+      "step": 254
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.074452327593561,
+      "learning_rate": 6.929347826086958e-06,
+      "loss": 0.9971,
+      "step": 255
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.882879030652116,
+      "learning_rate": 6.956521739130435e-06,
+      "loss": 0.7556,
+      "step": 256
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.997966863791385,
+      "learning_rate": 6.983695652173914e-06,
+      "loss": 1.1353,
+      "step": 257
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.651101842854983,
+      "learning_rate": 7.0108695652173915e-06,
+      "loss": 1.2194,
+      "step": 258
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.586971534165365,
+      "learning_rate": 7.03804347826087e-06,
+      "loss": 0.8826,
+      "step": 259
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 8.149993046803752,
+      "learning_rate": 7.065217391304349e-06,
+      "loss": 1.2882,
+      "step": 260
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.5061265280242075,
+      "learning_rate": 7.092391304347826e-06,
+      "loss": 0.625,
+      "step": 261
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.478712447315659,
+      "learning_rate": 7.119565217391305e-06,
+      "loss": 0.6542,
+      "step": 262
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.837884654347885,
+      "learning_rate": 7.146739130434784e-06,
+      "loss": 1.0407,
+      "step": 263
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.821659738606539,
+      "learning_rate": 7.173913043478261e-06,
+      "loss": 0.4997,
+      "step": 264
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.876378892024102,
+      "learning_rate": 7.20108695652174e-06,
+      "loss": 1.266,
+      "step": 265
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.167619207924342,
+      "learning_rate": 7.228260869565218e-06,
+      "loss": 0.5682,
+      "step": 266
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.240061974639415,
+      "learning_rate": 7.255434782608696e-06,
+      "loss": 0.9576,
+      "step": 267
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.287843463237732,
+      "learning_rate": 7.282608695652175e-06,
+      "loss": 0.4013,
+      "step": 268
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.593555393743031,
+      "learning_rate": 7.309782608695652e-06,
+      "loss": 0.799,
+      "step": 269
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.661077596444041,
+      "learning_rate": 7.3369565217391315e-06,
+      "loss": 1.6408,
+      "step": 270
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.307588962336696,
+      "learning_rate": 7.364130434782609e-06,
+      "loss": 0.749,
+      "step": 271
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 8.37410791120305,
+      "learning_rate": 7.391304347826087e-06,
+      "loss": 0.8649,
+      "step": 272
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 9.370566236174257,
+      "learning_rate": 7.418478260869566e-06,
+      "loss": 1.5156,
+      "step": 273
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.110290671960013,
+      "learning_rate": 7.445652173913044e-06,
+      "loss": 0.9931,
+      "step": 274
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.699925092422563,
+      "learning_rate": 7.472826086956523e-06,
+      "loss": 1.0836,
+      "step": 275
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.236535462488547,
+      "learning_rate": 7.500000000000001e-06,
+      "loss": 0.6872,
+      "step": 276
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 8.182701055682035,
+      "learning_rate": 7.5271739130434784e-06,
+      "loss": 1.3735,
+      "step": 277
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.376412752164515,
+      "learning_rate": 7.5543478260869576e-06,
+      "loss": 0.8981,
+      "step": 278
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.9740375817891804,
+      "learning_rate": 7.581521739130435e-06,
+      "loss": 0.5276,
+      "step": 279
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.909735996024911,
+      "learning_rate": 7.608695652173914e-06,
+      "loss": 0.8678,
+      "step": 280
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.243795496334634,
+      "learning_rate": 7.635869565217392e-06,
+      "loss": 0.8842,
+      "step": 281
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.985454133185386,
+      "learning_rate": 7.66304347826087e-06,
+      "loss": 0.3469,
+      "step": 282
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.329242088544714,
+      "learning_rate": 7.690217391304349e-06,
+      "loss": 0.8734,
+      "step": 283
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.438245468838116,
+      "learning_rate": 7.717391304347827e-06,
+      "loss": 1.1468,
+      "step": 284
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.965606920686939,
+      "learning_rate": 7.744565217391305e-06,
+      "loss": 0.9502,
+      "step": 285
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.078318625119426,
+      "learning_rate": 7.771739130434784e-06,
+      "loss": 1.2333,
+      "step": 286
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.360334548540601,
+      "learning_rate": 7.798913043478262e-06,
+      "loss": 0.7249,
+      "step": 287
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3335049330687503,
+      "learning_rate": 7.82608695652174e-06,
+      "loss": 0.2457,
+      "step": 288
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.7362915593195667,
+      "learning_rate": 7.853260869565218e-06,
+      "loss": 0.516,
+      "step": 289
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.8781325692758477,
+      "learning_rate": 7.880434782608695e-06,
+      "loss": 0.3987,
+      "step": 290
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.936096915610338,
+      "learning_rate": 7.907608695652175e-06,
+      "loss": 0.6694,
+      "step": 291
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.903385405984172,
+      "learning_rate": 7.934782608695653e-06,
+      "loss": 0.9925,
+      "step": 292
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.4190886985795634,
+      "learning_rate": 7.961956521739131e-06,
+      "loss": 0.2049,
+      "step": 293
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.404488252552992,
+      "learning_rate": 7.98913043478261e-06,
+      "loss": 0.7957,
+      "step": 294
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.097551330154639,
+      "learning_rate": 8.016304347826088e-06,
+      "loss": 0.966,
+      "step": 295
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.976238038990829,
+      "learning_rate": 8.043478260869566e-06,
+      "loss": 1.1818,
+      "step": 296
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 8.421521826437415,
+      "learning_rate": 8.070652173913044e-06,
+      "loss": 1.4973,
+      "step": 297
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.077754124133221,
+      "learning_rate": 8.097826086956523e-06,
+      "loss": 0.5261,
+      "step": 298
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.4070308899500175,
+      "learning_rate": 8.125000000000001e-06,
+      "loss": 1.1101,
+      "step": 299
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.318753575824465,
+      "learning_rate": 8.15217391304348e-06,
+      "loss": 1.104,
+      "step": 300
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.284465274777319,
+      "learning_rate": 8.179347826086957e-06,
+      "loss": 0.7641,
+      "step": 301
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.625176035390117,
+      "learning_rate": 8.206521739130436e-06,
+      "loss": 0.8921,
+      "step": 302
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.339376984014242,
+      "learning_rate": 8.233695652173914e-06,
+      "loss": 1.2197,
+      "step": 303
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.157859884255341,
+      "learning_rate": 8.260869565217392e-06,
+      "loss": 0.9333,
+      "step": 304
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.3629598823237234,
+      "learning_rate": 8.28804347826087e-06,
+      "loss": 0.9097,
+      "step": 305
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.868071909896863,
+      "learning_rate": 8.315217391304349e-06,
+      "loss": 0.9008,
+      "step": 306
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.978604409945964,
+      "learning_rate": 8.342391304347827e-06,
+      "loss": 0.5327,
+      "step": 307
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.7437093095606695,
+      "learning_rate": 8.369565217391305e-06,
+      "loss": 0.8238,
+      "step": 308
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.454985051623668,
+      "learning_rate": 8.396739130434784e-06,
+      "loss": 0.6295,
+      "step": 309
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.614788356863233,
+      "learning_rate": 8.423913043478262e-06,
+      "loss": 0.9974,
+      "step": 310
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.608916913600715,
+      "learning_rate": 8.45108695652174e-06,
+      "loss": 1.1464,
+      "step": 311
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.436932595933803,
+      "learning_rate": 8.478260869565218e-06,
+      "loss": 0.4595,
+      "step": 312
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.57722024795812,
+      "learning_rate": 8.505434782608697e-06,
+      "loss": 0.5938,
+      "step": 313
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.677248597091288,
+      "learning_rate": 8.532608695652175e-06,
+      "loss": 1.1063,
+      "step": 314
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 8.034438340002367,
+      "learning_rate": 8.559782608695653e-06,
+      "loss": 1.6156,
+      "step": 315
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.2062511410316326,
+      "learning_rate": 8.586956521739131e-06,
+      "loss": 0.6021,
+      "step": 316
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.145723079176569,
+      "learning_rate": 8.61413043478261e-06,
+      "loss": 1.6035,
+      "step": 317
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.34202953981997,
+      "learning_rate": 8.641304347826088e-06,
+      "loss": 0.8125,
+      "step": 318
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.337229311201793,
+      "learning_rate": 8.668478260869566e-06,
+      "loss": 1.1929,
+      "step": 319
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.225481053597104,
+      "learning_rate": 8.695652173913044e-06,
+      "loss": 1.1157,
+      "step": 320
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.017710989732211,
+      "learning_rate": 8.722826086956523e-06,
+      "loss": 0.8339,
+      "step": 321
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 7.882373233059433,
+      "learning_rate": 8.750000000000001e-06,
+      "loss": 1.262,
+      "step": 322
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.066556353039955,
+      "learning_rate": 8.77717391304348e-06,
+      "loss": 0.2942,
+      "step": 323
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 7.770860273190021,
+      "learning_rate": 8.804347826086957e-06,
+      "loss": 1.4823,
+      "step": 324
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.7343550772185505,
+      "learning_rate": 8.831521739130436e-06,
+      "loss": 0.7482,
+      "step": 325
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.639168375286053,
+      "learning_rate": 8.858695652173914e-06,
+      "loss": 0.9131,
+      "step": 326
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 7.518462985960601,
+      "learning_rate": 8.885869565217392e-06,
+      "loss": 1.058,
+      "step": 327
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 7.54762146148649,
+      "learning_rate": 8.91304347826087e-06,
+      "loss": 1.2194,
+      "step": 328
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 7.871385299233604,
+      "learning_rate": 8.940217391304349e-06,
+      "loss": 1.0078,
+      "step": 329
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.359117961957575,
+      "learning_rate": 8.967391304347827e-06,
+      "loss": 0.7769,
+      "step": 330
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.0982811956203244,
+      "learning_rate": 8.994565217391305e-06,
+      "loss": 0.7377,
+      "step": 331
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.830234028030018,
+      "learning_rate": 9.021739130434784e-06,
+      "loss": 0.9978,
+      "step": 332
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.8181359163099655,
+      "learning_rate": 9.048913043478262e-06,
+      "loss": 0.5928,
+      "step": 333
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.681560068759461,
+      "learning_rate": 9.07608695652174e-06,
+      "loss": 1.0175,
+      "step": 334
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.80680243793632,
+      "learning_rate": 9.103260869565218e-06,
+      "loss": 1.1993,
+      "step": 335
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.186937644327905,
+      "learning_rate": 9.130434782608697e-06,
+      "loss": 0.6584,
+      "step": 336
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.560489553639137,
+      "learning_rate": 9.157608695652175e-06,
+      "loss": 1.1444,
+      "step": 337
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 8.058229643385323,
+      "learning_rate": 9.184782608695653e-06,
+      "loss": 1.1921,
+      "step": 338
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.668465254356496,
+      "learning_rate": 9.211956521739131e-06,
+      "loss": 0.7071,
+      "step": 339
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.35169558250226,
+      "learning_rate": 9.23913043478261e-06,
+      "loss": 0.9809,
+      "step": 340
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.837242529338085,
+      "learning_rate": 9.266304347826088e-06,
+      "loss": 0.9961,
+      "step": 341
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.198233376887303,
+      "learning_rate": 9.293478260869566e-06,
+      "loss": 1.0778,
+      "step": 342
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.4647916149997653,
+      "learning_rate": 9.320652173913044e-06,
+      "loss": 0.3007,
+      "step": 343
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.261137583211198,
+      "learning_rate": 9.347826086956523e-06,
+      "loss": 1.3224,
+      "step": 344
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.556407019061477,
+      "learning_rate": 9.375000000000001e-06,
+      "loss": 1.2377,
+      "step": 345
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 8.46885744874352,
+      "learning_rate": 9.402173913043479e-06,
+      "loss": 1.5947,
+      "step": 346
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.194630590437151,
+      "learning_rate": 9.429347826086957e-06,
+      "loss": 0.8361,
+      "step": 347
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.769530302655123,
+      "learning_rate": 9.456521739130436e-06,
+      "loss": 1.2915,
+      "step": 348
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.27229215981989,
+      "learning_rate": 9.483695652173914e-06,
+      "loss": 0.4573,
+      "step": 349
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.0281628916208145,
+      "learning_rate": 9.510869565217392e-06,
+      "loss": 0.9685,
+      "step": 350
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 7.614906153133044,
+      "learning_rate": 9.53804347826087e-06,
+      "loss": 2.14,
+      "step": 351
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.132861967845814,
+      "learning_rate": 9.565217391304349e-06,
+      "loss": 0.4094,
+      "step": 352
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 7.769971710510702,
+      "learning_rate": 9.592391304347827e-06,
+      "loss": 1.5559,
+      "step": 353
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.214541771357837,
+      "learning_rate": 9.619565217391305e-06,
+      "loss": 0.6542,
+      "step": 354
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.368108862701037,
+      "learning_rate": 9.646739130434783e-06,
+      "loss": 0.3184,
+      "step": 355
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 9.467183830886526,
+      "learning_rate": 9.673913043478262e-06,
+      "loss": 2.0,
+      "step": 356
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.148717928615864,
+      "learning_rate": 9.70108695652174e-06,
+      "loss": 0.7562,
+      "step": 357
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.699142596546084,
+      "learning_rate": 9.728260869565218e-06,
+      "loss": 0.5602,
+      "step": 358
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.778849120224696,
+      "learning_rate": 9.755434782608696e-06,
+      "loss": 1.285,
+      "step": 359
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.325812744530062,
+      "learning_rate": 9.782608695652175e-06,
+      "loss": 1.035,
+      "step": 360
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.270446437848442,
+      "learning_rate": 9.809782608695653e-06,
+      "loss": 1.1036,
+      "step": 361
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.098595435383664,
+      "learning_rate": 9.836956521739131e-06,
+      "loss": 0.7118,
+      "step": 362
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.120947794599988,
+      "learning_rate": 9.86413043478261e-06,
+      "loss": 1.0194,
+      "step": 363
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.753519153874003,
+      "learning_rate": 9.891304347826088e-06,
+      "loss": 0.8027,
+      "step": 364
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.54491484144688,
+      "learning_rate": 9.918478260869566e-06,
+      "loss": 0.9281,
+      "step": 365
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.88514814649314,
+      "learning_rate": 9.945652173913044e-06,
+      "loss": 0.8455,
+      "step": 366
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.574575608802646,
+      "learning_rate": 9.972826086956523e-06,
+      "loss": 1.2401,
+      "step": 367
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.295452016714822,
+      "learning_rate": 1e-05,
+      "loss": 0.7394,
+      "step": 368
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.827174479039339,
+      "learning_rate": 9.999999824760881e-06,
+      "loss": 0.912,
+      "step": 369
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.053210078293678,
+      "learning_rate": 9.999999299043533e-06,
+      "loss": 0.455,
+      "step": 370
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.67096244105842,
+      "learning_rate": 9.999998422847994e-06,
+      "loss": 0.8681,
+      "step": 371
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.554637130438726,
+      "learning_rate": 9.999997196174327e-06,
+      "loss": 0.7649,
+      "step": 372
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.5853043492657246,
+      "learning_rate": 9.999995619022615e-06,
+      "loss": 0.6937,
+      "step": 373
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.6670712091283875,
+      "learning_rate": 9.999993691392973e-06,
+      "loss": 0.8806,
+      "step": 374
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.168060014150088,
+      "learning_rate": 9.99999141328553e-06,
+      "loss": 0.8426,
+      "step": 375
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 7.843194545607991,
+      "learning_rate": 9.999988784700451e-06,
+      "loss": 1.1695,
+      "step": 376
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.168568665105547,
+      "learning_rate": 9.999985805637917e-06,
+      "loss": 0.7686,
+      "step": 377
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.894165959519665,
+      "learning_rate": 9.999982476098138e-06,
+      "loss": 0.7126,
+      "step": 378
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 9.08781693842811,
+      "learning_rate": 9.99997879608135e-06,
+      "loss": 0.9496,
+      "step": 379
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.7907899593817636,
+      "learning_rate": 9.999974765587805e-06,
+      "loss": 1.0241,
+      "step": 380
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.14564735190624,
+      "learning_rate": 9.99997038461779e-06,
+      "loss": 1.084,
+      "step": 381
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.632103552875158,
+      "learning_rate": 9.99996565317161e-06,
+      "loss": 0.6177,
+      "step": 382
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 7.06828195446152,
+      "learning_rate": 9.999960571249599e-06,
+      "loss": 1.2956,
+      "step": 383
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.651102377791331,
+      "learning_rate": 9.999955138852111e-06,
+      "loss": 1.3542,
+      "step": 384
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.042294337025638,
+      "learning_rate": 9.99994935597953e-06,
+      "loss": 0.6006,
+      "step": 385
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.402994414654537,
+      "learning_rate": 9.999943222632258e-06,
+      "loss": 0.4791,
+      "step": 386
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.965056543386612,
+      "learning_rate": 9.999936738810725e-06,
+      "loss": 1.0879,
+      "step": 387
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.1713896321431605,
+      "learning_rate": 9.999929904515386e-06,
+      "loss": 1.053,
+      "step": 388
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.145030010793328,
+      "learning_rate": 9.99992271974672e-06,
+      "loss": 0.7253,
+      "step": 389
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.336075295397067,
+      "learning_rate": 9.999915184505233e-06,
+      "loss": 0.7442,
+      "step": 390
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.5450996060354,
+      "learning_rate": 9.99990729879145e-06,
+      "loss": 1.277,
+      "step": 391
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 7.636158624112354,
+      "learning_rate": 9.999899062605928e-06,
+      "loss": 1.239,
+      "step": 392
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.477488726410057,
+      "learning_rate": 9.999890475949236e-06,
+      "loss": 0.9772,
+      "step": 393
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.051495747823013,
+      "learning_rate": 9.999881538821985e-06,
+      "loss": 0.5469,
+      "step": 394
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.300752608556988,
+      "learning_rate": 9.999872251224796e-06,
+      "loss": 0.7399,
+      "step": 395
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.5526737909232016,
+      "learning_rate": 9.999862613158323e-06,
+      "loss": 0.513,
+      "step": 396
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.933537207255718,
+      "learning_rate": 9.99985262462324e-06,
+      "loss": 0.8352,
+      "step": 397
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.738769227354006,
+      "learning_rate": 9.999842285620247e-06,
+      "loss": 1.0434,
+      "step": 398
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.762740426870926,
+      "learning_rate": 9.999831596150069e-06,
+      "loss": 0.3968,
+      "step": 399
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.1264125436257375,
+      "learning_rate": 9.999820556213455e-06,
+      "loss": 0.7019,
+      "step": 400
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.500334497583717,
+      "learning_rate": 9.99980916581118e-06,
+      "loss": 0.8607,
+      "step": 401
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.481620748046511,
+      "learning_rate": 9.999797424944041e-06,
+      "loss": 0.8718,
+      "step": 402
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.151238330741133,
+      "learning_rate": 9.999785333612863e-06,
+      "loss": 0.7545,
+      "step": 403
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.976320787775345,
+      "learning_rate": 9.999772891818493e-06,
+      "loss": 1.2427,
+      "step": 404
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.000150498697293,
+      "learning_rate": 9.999760099561802e-06,
+      "loss": 0.9778,
+      "step": 405
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.6502724932396453,
+      "learning_rate": 9.999746956843685e-06,
+      "loss": 0.6976,
+      "step": 406
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.193776525977321,
+      "learning_rate": 9.999733463665067e-06,
+      "loss": 0.8279,
+      "step": 407
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 7.239212491469085,
+      "learning_rate": 9.999719620026891e-06,
+      "loss": 1.3266,
+      "step": 408
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.596325626752294,
+      "learning_rate": 9.99970542593013e-06,
+      "loss": 1.1974,
+      "step": 409
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.706543574475022,
+      "learning_rate": 9.999690881375777e-06,
+      "loss": 1.2101,
+      "step": 410
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.644646579385012,
+      "learning_rate": 9.99967598636485e-06,
+      "loss": 0.9091,
+      "step": 411
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.9175230022782546,
+      "learning_rate": 9.999660740898397e-06,
+      "loss": 0.498,
+      "step": 412
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.173890194845122,
+      "learning_rate": 9.999645144977483e-06,
+      "loss": 0.7663,
+      "step": 413
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.748467260869798,
+      "learning_rate": 9.999629198603205e-06,
+      "loss": 0.7728,
+      "step": 414
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.778886530785687,
+      "learning_rate": 9.999612901776678e-06,
+      "loss": 1.2248,
+      "step": 415
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.935476197785821,
+      "learning_rate": 9.999596254499044e-06,
+      "loss": 0.6129,
+      "step": 416
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.02337258831855,
+      "learning_rate": 9.999579256771473e-06,
+      "loss": 0.6155,
+      "step": 417
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.8062484181428986,
+      "learning_rate": 9.999561908595153e-06,
+      "loss": 0.6679,
+      "step": 418
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.8954362755783265,
+      "learning_rate": 9.999544209971299e-06,
+      "loss": 0.66,
+      "step": 419
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.099290865501664,
+      "learning_rate": 9.999526160901156e-06,
+      "loss": 0.8223,
+      "step": 420
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.863034209422073,
+      "learning_rate": 9.999507761385989e-06,
+      "loss": 0.6422,
+      "step": 421
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.083975787293613,
+      "learning_rate": 9.999489011427084e-06,
+      "loss": 0.7041,
+      "step": 422
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.3178956260126076,
+      "learning_rate": 9.999469911025756e-06,
+      "loss": 0.5422,
+      "step": 423
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.650247498520858,
+      "learning_rate": 9.999450460183347e-06,
+      "loss": 1.0725,
+      "step": 424
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.95259074049211,
+      "learning_rate": 9.999430658901217e-06,
+      "loss": 0.9106,
+      "step": 425
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 7.71544477110711,
+      "learning_rate": 9.999410507180757e-06,
+      "loss": 1.443,
+      "step": 426
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.346969103609945,
+      "learning_rate": 9.999390005023377e-06,
+      "loss": 1.1114,
+      "step": 427
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.434687806214507,
+      "learning_rate": 9.999369152430514e-06,
+      "loss": 1.0029,
+      "step": 428
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.2244364829348955,
+      "learning_rate": 9.999347949403633e-06,
+      "loss": 0.8434,
+      "step": 429
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.439228351059763,
+      "learning_rate": 9.999326395944217e-06,
+      "loss": 1.0175,
+      "step": 430
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.899604465621711,
+      "learning_rate": 9.999304492053777e-06,
+      "loss": 0.9335,
+      "step": 431
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.3167350747899755,
+      "learning_rate": 9.999282237733849e-06,
+      "loss": 1.1625,
+      "step": 432
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.770326297294352,
+      "learning_rate": 9.999259632985996e-06,
+      "loss": 0.896,
+      "step": 433
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.427169696584599,
+      "learning_rate": 9.999236677811796e-06,
+      "loss": 0.8969,
+      "step": 434
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.297151047850789,
+      "learning_rate": 9.999213372212863e-06,
+      "loss": 0.9043,
+      "step": 435
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.20610963603229,
+      "learning_rate": 9.99918971619083e-06,
+      "loss": 1.2859,
+      "step": 436
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.8465537937699406,
+      "learning_rate": 9.999165709747353e-06,
+      "loss": 1.0852,
+      "step": 437
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.657766614343768,
+      "learning_rate": 9.999141352884118e-06,
+      "loss": 1.0622,
+      "step": 438
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.9847198110116775,
+      "learning_rate": 9.999116645602828e-06,
+      "loss": 0.9139,
+      "step": 439
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.4219398024801,
+      "learning_rate": 9.99909158790522e-06,
+      "loss": 0.8118,
+      "step": 440
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.208092284451877,
+      "learning_rate": 9.999066179793047e-06,
+      "loss": 1.0925,
+      "step": 441
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.381473249926374,
+      "learning_rate": 9.99904042126809e-06,
+      "loss": 1.1494,
+      "step": 442
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.171395550274936,
+      "learning_rate": 9.999014312332156e-06,
+      "loss": 1.1686,
+      "step": 443
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.07792792749062,
+      "learning_rate": 9.998987852987074e-06,
+      "loss": 0.9116,
+      "step": 444
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 7.609250844455039,
+      "learning_rate": 9.9989610432347e-06,
+      "loss": 1.7275,
+      "step": 445
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.042050660855812,
+      "learning_rate": 9.998933883076912e-06,
+      "loss": 1.1827,
+      "step": 446
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.387281639470219,
+      "learning_rate": 9.998906372515615e-06,
+      "loss": 0.9077,
+      "step": 447
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.807479350295363,
+      "learning_rate": 9.998878511552734e-06,
+      "loss": 0.9013,
+      "step": 448
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.673383829189851,
+      "learning_rate": 9.998850300190226e-06,
+      "loss": 0.7067,
+      "step": 449
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.482918184763173,
+      "learning_rate": 9.99882173843007e-06,
+      "loss": 1.2614,
+      "step": 450
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.972131746168448,
+      "learning_rate": 9.99879282627426e-06,
+      "loss": 0.911,
+      "step": 451
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 7.2324493611028275,
+      "learning_rate": 9.998763563724831e-06,
+      "loss": 1.1946,
+      "step": 452
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.36145192963037,
+      "learning_rate": 9.99873395078383e-06,
+      "loss": 1.0718,
+      "step": 453
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.139122292070844,
+      "learning_rate": 9.998703987453334e-06,
+      "loss": 0.8382,
+      "step": 454
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.657219504198628,
+      "learning_rate": 9.998673673735442e-06,
+      "loss": 1.268,
+      "step": 455
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.009202914261718,
+      "learning_rate": 9.998643009632281e-06,
+      "loss": 0.7224,
+      "step": 456
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.475889755505141,
+      "learning_rate": 9.998611995145997e-06,
+      "loss": 0.7266,
+      "step": 457
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9204812272815377,
+      "learning_rate": 9.99858063027877e-06,
+      "loss": 0.2579,
+      "step": 458
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.499067475167554,
+      "learning_rate": 9.99854891503279e-06,
+      "loss": 1.1072,
+      "step": 459
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.41932921975513,
+      "learning_rate": 9.998516849410287e-06,
+      "loss": 0.4482,
+      "step": 460
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.714106357668273,
+      "learning_rate": 9.998484433413507e-06,
+      "loss": 0.7622,
+      "step": 461
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.043566688600208,
+      "learning_rate": 9.998451667044721e-06,
+      "loss": 0.6495,
+      "step": 462
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.919820905006889,
+      "learning_rate": 9.998418550306228e-06,
+      "loss": 0.8238,
+      "step": 463
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.294308724711821,
+      "learning_rate": 9.998385083200346e-06,
+      "loss": 0.5751,
+      "step": 464
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.301981205204581,
+      "learning_rate": 9.998351265729423e-06,
+      "loss": 0.4563,
+      "step": 465
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.89628096529308,
+      "learning_rate": 9.99831709789583e-06,
+      "loss": 1.39,
+      "step": 466
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.64203438591841,
+      "learning_rate": 9.99828257970196e-06,
+      "loss": 0.8958,
+      "step": 467
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.151009132010121,
+      "learning_rate": 9.998247711150235e-06,
+      "loss": 0.4676,
+      "step": 468
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.055988336093534,
+      "learning_rate": 9.998212492243099e-06,
+      "loss": 1.0388,
+      "step": 469
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.290773029422374,
+      "learning_rate": 9.998176922983017e-06,
+      "loss": 1.0765,
+      "step": 470
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.447842072761543,
+      "learning_rate": 9.998141003372486e-06,
+      "loss": 1.2072,
+      "step": 471
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.586951220708749,
+      "learning_rate": 9.998104733414022e-06,
+      "loss": 1.0186,
+      "step": 472
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.485052880067488,
+      "learning_rate": 9.998068113110168e-06,
+      "loss": 0.9249,
+      "step": 473
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.381424161053315,
+      "learning_rate": 9.99803114246349e-06,
+      "loss": 1.4598,
+      "step": 474
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.656368392426574,
+      "learning_rate": 9.997993821476583e-06,
+      "loss": 0.7038,
+      "step": 475
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.572194725600882,
+      "learning_rate": 9.99795615015206e-06,
+      "loss": 1.0368,
+      "step": 476
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.633066802629584,
+      "learning_rate": 9.99791812849256e-06,
+      "loss": 1.0351,
+      "step": 477
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.647468770382145,
+      "learning_rate": 9.997879756500752e-06,
+      "loss": 1.2678,
+      "step": 478
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.661062102561084,
+      "learning_rate": 9.997841034179323e-06,
+      "loss": 0.879,
+      "step": 479
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.5448288381588355,
+      "learning_rate": 9.997801961530989e-06,
+      "loss": 0.8834,
+      "step": 480
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.746344735056418,
+      "learning_rate": 9.997762538558488e-06,
+      "loss": 1.4655,
+      "step": 481
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.7462834947998007,
+      "learning_rate": 9.997722765264582e-06,
+      "loss": 0.7281,
+      "step": 482
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.7286846769815325,
+      "learning_rate": 9.997682641652062e-06,
+      "loss": 1.1582,
+      "step": 483
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.03167073385437,
+      "learning_rate": 9.997642167723737e-06,
+      "loss": 0.9327,
+      "step": 484
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.232339007741685,
+      "learning_rate": 9.997601343482448e-06,
+      "loss": 1.1108,
+      "step": 485
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.245419578495987,
+      "learning_rate": 9.997560168931053e-06,
+      "loss": 0.8701,
+      "step": 486
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.9573609801548555,
+      "learning_rate": 9.99751864407244e-06,
+      "loss": 1.6929,
+      "step": 487
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.6903216396074745,
+      "learning_rate": 9.99747676890952e-06,
+      "loss": 1.0227,
+      "step": 488
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.879451561124423,
+      "learning_rate": 9.997434543445227e-06,
+      "loss": 1.1881,
+      "step": 489
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.016960042499101,
+      "learning_rate": 9.997391967682522e-06,
+      "loss": 0.6582,
+      "step": 490
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.015495969337913,
+      "learning_rate": 9.997349041624387e-06,
+      "loss": 1.2102,
+      "step": 491
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.627930447954648,
+      "learning_rate": 9.997305765273834e-06,
+      "loss": 0.6894,
+      "step": 492
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.403979572529196,
+      "learning_rate": 9.997262138633895e-06,
+      "loss": 0.9797,
+      "step": 493
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.264874126549467,
+      "learning_rate": 9.99721816170763e-06,
+      "loss": 1.0349,
+      "step": 494
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.456314950766531,
+      "learning_rate": 9.997173834498118e-06,
+      "loss": 0.2876,
+      "step": 495
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 7.390226223155865,
+      "learning_rate": 9.997129157008467e-06,
+      "loss": 0.9932,
+      "step": 496
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.220155764196953,
+      "learning_rate": 9.99708412924181e-06,
+      "loss": 0.9571,
+      "step": 497
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.090257690615265,
+      "learning_rate": 9.997038751201305e-06,
+      "loss": 0.6411,
+      "step": 498
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.4198073238913955,
+      "learning_rate": 9.996993022890129e-06,
+      "loss": 0.4335,
+      "step": 499
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.9754266714771402,
+      "learning_rate": 9.99694694431149e-06,
+      "loss": 0.5786,
+      "step": 500
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.862299420547906,
+      "learning_rate": 9.996900515468614e-06,
+      "loss": 0.4861,
+      "step": 501
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.749216334168513,
+      "learning_rate": 9.996853736364763e-06,
+      "loss": 1.0879,
+      "step": 502
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.357431072223263,
+      "learning_rate": 9.996806607003209e-06,
+      "loss": 1.2146,
+      "step": 503
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.591655212241847,
+      "learning_rate": 9.996759127387259e-06,
+      "loss": 0.7087,
+      "step": 504
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.45088704039877,
+      "learning_rate": 9.996711297520238e-06,
+      "loss": 1.3822,
+      "step": 505
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 7.018084335327446,
+      "learning_rate": 9.996663117405503e-06,
+      "loss": 1.5407,
+      "step": 506
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.427680298582249,
+      "learning_rate": 9.99661458704643e-06,
+      "loss": 0.5793,
+      "step": 507
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 7.765130560248742,
+      "learning_rate": 9.996565706446418e-06,
+      "loss": 1.67,
+      "step": 508
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.8745700844606406,
+      "learning_rate": 9.996516475608894e-06,
+      "loss": 1.421,
+      "step": 509
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.075923799198178,
+      "learning_rate": 9.996466894537311e-06,
+      "loss": 1.1038,
+      "step": 510
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.259111833152261,
+      "learning_rate": 9.996416963235144e-06,
+      "loss": 0.2819,
+      "step": 511
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9609336099359,
+      "learning_rate": 9.996366681705892e-06,
+      "loss": 0.4535,
+      "step": 512
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.9779750310059985,
+      "learning_rate": 9.99631604995308e-06,
+      "loss": 0.9532,
+      "step": 513
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.5643470817068112,
+      "learning_rate": 9.996265067980256e-06,
+      "loss": 0.4561,
+      "step": 514
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.865293599732065,
+      "learning_rate": 9.996213735790995e-06,
+      "loss": 1.4386,
+      "step": 515
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.698458945782838,
+      "learning_rate": 9.996162053388895e-06,
+      "loss": 1.2189,
+      "step": 516
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.947033920225605,
+      "learning_rate": 9.996110020777579e-06,
+      "loss": 0.9818,
+      "step": 517
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.38030957812177,
+      "learning_rate": 9.996057637960694e-06,
+      "loss": 0.8517,
+      "step": 518
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.3778991466873185,
+      "learning_rate": 9.996004904941911e-06,
+      "loss": 1.012,
+      "step": 519
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.495684553725476,
+      "learning_rate": 9.995951821724926e-06,
+      "loss": 1.1764,
+      "step": 520
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.828192139209808,
+      "learning_rate": 9.99589838831346e-06,
+      "loss": 1.6443,
+      "step": 521
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.584025659582305,
+      "learning_rate": 9.995844604711262e-06,
+      "loss": 0.8156,
+      "step": 522
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.378898352225901,
+      "learning_rate": 9.995790470922098e-06,
+      "loss": 0.7585,
+      "step": 523
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.815193156427421,
+      "learning_rate": 9.995735986949763e-06,
+      "loss": 1.1676,
+      "step": 524
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.110317832939963,
+      "learning_rate": 9.995681152798079e-06,
+      "loss": 0.6391,
+      "step": 525
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.39861470191864,
+      "learning_rate": 9.995625968470883e-06,
+      "loss": 0.7007,
+      "step": 526
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 7.833803015864922,
+      "learning_rate": 9.995570433972051e-06,
+      "loss": 1.3191,
+      "step": 527
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.314688180821015,
+      "learning_rate": 9.995514549305472e-06,
+      "loss": 1.2388,
+      "step": 528
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.017090408912602,
+      "learning_rate": 9.995458314475064e-06,
+      "loss": 0.9468,
+      "step": 529
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.001208522010045,
+      "learning_rate": 9.995401729484768e-06,
+      "loss": 0.831,
+      "step": 530
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.782985672211555,
+      "learning_rate": 9.995344794338551e-06,
+      "loss": 0.8729,
+      "step": 531
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.543829103793221,
+      "learning_rate": 9.995287509040403e-06,
+      "loss": 1.1005,
+      "step": 532
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.598299872795798,
+      "learning_rate": 9.99522987359434e-06,
+      "loss": 1.0555,
+      "step": 533
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.2787368892690845,
+      "learning_rate": 9.995171888004403e-06,
+      "loss": 1.141,
+      "step": 534
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.8012577004433523,
+      "learning_rate": 9.995113552274656e-06,
+      "loss": 0.7279,
+      "step": 535
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.805953330549673,
+      "learning_rate": 9.995054866409186e-06,
+      "loss": 1.3504,
+      "step": 536
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.725355607677452,
+      "learning_rate": 9.99499583041211e-06,
+      "loss": 1.1677,
+      "step": 537
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.409679139961245,
+      "learning_rate": 9.994936444287565e-06,
+      "loss": 1.1266,
+      "step": 538
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.340503058191312,
+      "learning_rate": 9.994876708039712e-06,
+      "loss": 0.924,
+      "step": 539
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.635724815694596,
+      "learning_rate": 9.99481662167274e-06,
+      "loss": 0.6758,
+      "step": 540
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.137018574224754,
+      "learning_rate": 9.99475618519086e-06,
+      "loss": 1.488,
+      "step": 541
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.69956991439474,
+      "learning_rate": 9.99469539859831e-06,
+      "loss": 0.823,
+      "step": 542
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.019347119841303,
+      "learning_rate": 9.994634261899347e-06,
+      "loss": 0.9059,
+      "step": 543
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.7141592038245035,
+      "learning_rate": 9.994572775098262e-06,
+      "loss": 1.0794,
+      "step": 544
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 7.810662155227189,
+      "learning_rate": 9.99451093819936e-06,
+      "loss": 1.2191,
+      "step": 545
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6379710591200758,
+      "learning_rate": 9.994448751206978e-06,
+      "loss": 0.2519,
+      "step": 546
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.276706835772696,
+      "learning_rate": 9.994386214125476e-06,
+      "loss": 1.4129,
+      "step": 547
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 7.014552652694709,
+      "learning_rate": 9.994323326959234e-06,
+      "loss": 1.2461,
+      "step": 548
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.620540826198909,
+      "learning_rate": 9.994260089712662e-06,
+      "loss": 1.0425,
+      "step": 549
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 8.080536553202029,
+      "learning_rate": 9.994196502390194e-06,
+      "loss": 1.7665,
+      "step": 550
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.637864424017921,
+      "learning_rate": 9.994132564996284e-06,
+      "loss": 0.9307,
+      "step": 551
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.781137741702556,
+      "learning_rate": 9.994068277535418e-06,
+      "loss": 0.8607,
+      "step": 552
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1510246907020736,
+      "learning_rate": 9.994003640012099e-06,
+      "loss": 0.2683,
+      "step": 553
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.250424983427129,
+      "learning_rate": 9.99393865243086e-06,
+      "loss": 0.6378,
+      "step": 554
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.319957939313974,
+      "learning_rate": 9.993873314796253e-06,
+      "loss": 0.8753,
+      "step": 555
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.405556982487025,
+      "learning_rate": 9.993807627112862e-06,
+      "loss": 1.3312,
+      "step": 556
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.160310787271504,
+      "learning_rate": 9.993741589385287e-06,
+      "loss": 0.7164,
+      "step": 557
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.749100097319363,
+      "learning_rate": 9.993675201618162e-06,
+      "loss": 1.0449,
+      "step": 558
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.796749991766679,
+      "learning_rate": 9.993608463816137e-06,
+      "loss": 1.0364,
+      "step": 559
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.0955792124834423,
+      "learning_rate": 9.99354137598389e-06,
+      "loss": 0.5138,
+      "step": 560
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.177971692772812,
+      "learning_rate": 9.993473938126126e-06,
+      "loss": 1.0678,
+      "step": 561
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.299457668753542,
+      "learning_rate": 9.99340615024757e-06,
+      "loss": 1.4612,
+      "step": 562
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.471041279551884,
+      "learning_rate": 9.993338012352973e-06,
+      "loss": 0.8609,
+      "step": 563
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.131374831065728,
+      "learning_rate": 9.993269524447115e-06,
+      "loss": 0.8387,
+      "step": 564
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.95767345172936,
+      "learning_rate": 9.993200686534793e-06,
+      "loss": 1.1389,
+      "step": 565
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.787905925023107,
+      "learning_rate": 9.993131498620833e-06,
+      "loss": 1.0385,
+      "step": 566
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.709708416684444,
+      "learning_rate": 9.993061960710084e-06,
+      "loss": 1.262,
+      "step": 567
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.7600247581497,
+      "learning_rate": 9.992992072807424e-06,
+      "loss": 0.7766,
+      "step": 568
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.040306502808157,
+      "learning_rate": 9.992921834917748e-06,
+      "loss": 1.0075,
+      "step": 569
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.765888597714929,
+      "learning_rate": 9.99285124704598e-06,
+      "loss": 1.1774,
+      "step": 570
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.251268554699921,
+      "learning_rate": 9.99278030919707e-06,
+      "loss": 0.8978,
+      "step": 571
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.687606443004777,
+      "learning_rate": 9.992709021375987e-06,
+      "loss": 0.6764,
+      "step": 572
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.7975790585142155,
+      "learning_rate": 9.992637383587731e-06,
+      "loss": 1.3707,
+      "step": 573
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.216275180993614,
+      "learning_rate": 9.992565395837323e-06,
+      "loss": 1.0971,
+      "step": 574
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.248741788140497,
+      "learning_rate": 9.992493058129808e-06,
+      "loss": 0.7766,
+      "step": 575
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.605246423497219,
+      "learning_rate": 9.992420370470257e-06,
+      "loss": 0.7862,
+      "step": 576
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.2123803620145885,
+      "learning_rate": 9.992347332863766e-06,
+      "loss": 1.1643,
+      "step": 577
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.017266738904277,
+      "learning_rate": 9.992273945315451e-06,
+      "loss": 0.9691,
+      "step": 578
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.856181346386722,
+      "learning_rate": 9.992200207830461e-06,
+      "loss": 0.6863,
+      "step": 579
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.208324579382292,
+      "learning_rate": 9.992126120413963e-06,
+      "loss": 0.8127,
+      "step": 580
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.758991816274172,
+      "learning_rate": 9.99205168307115e-06,
+      "loss": 0.7493,
+      "step": 581
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.657302028990941,
+      "learning_rate": 9.991976895807237e-06,
+      "loss": 0.9181,
+      "step": 582
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.7383748965946557,
+      "learning_rate": 9.99190175862747e-06,
+      "loss": 0.5278,
+      "step": 583
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.017967974387647,
+      "learning_rate": 9.991826271537115e-06,
+      "loss": 1.0367,
+      "step": 584
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.830582460586334,
+      "learning_rate": 9.991750434541463e-06,
+      "loss": 1.2415,
+      "step": 585
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.414455961267636,
+      "learning_rate": 9.99167424764583e-06,
+      "loss": 0.823,
+      "step": 586
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.008113109219425,
+      "learning_rate": 9.991597710855555e-06,
+      "loss": 1.4785,
+      "step": 587
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.6219971020775015,
+      "learning_rate": 9.991520824176004e-06,
+      "loss": 1.1196,
+      "step": 588
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.2924239603883967,
+      "learning_rate": 9.991443587612568e-06,
+      "loss": 0.4934,
+      "step": 589
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.8783937859474875,
+      "learning_rate": 9.991366001170656e-06,
+      "loss": 1.0913,
+      "step": 590
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.6596365717238735,
+      "learning_rate": 9.991288064855713e-06,
+      "loss": 0.8888,
+      "step": 591
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9574292806593845,
+      "learning_rate": 9.991209778673199e-06,
+      "loss": 0.241,
+      "step": 592
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.958682017079736,
+      "learning_rate": 9.991131142628601e-06,
+      "loss": 1.2773,
+      "step": 593
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.497872038290533,
+      "learning_rate": 9.99105215672743e-06,
+      "loss": 0.9163,
+      "step": 594
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.356815585248249,
+      "learning_rate": 9.990972820975224e-06,
+      "loss": 0.9554,
+      "step": 595
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.823751138948768,
+      "learning_rate": 9.990893135377544e-06,
+      "loss": 0.6938,
+      "step": 596
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.869332541879883,
+      "learning_rate": 9.990813099939977e-06,
+      "loss": 0.9904,
+      "step": 597
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.557426669936456,
+      "learning_rate": 9.990732714668132e-06,
+      "loss": 0.4088,
+      "step": 598
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.994165849060579,
+      "learning_rate": 9.99065197956764e-06,
+      "loss": 0.5122,
+      "step": 599
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.441289707355455,
+      "learning_rate": 9.990570894644168e-06,
+      "loss": 1.1079,
+      "step": 600
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.361083611031531,
+      "learning_rate": 9.990489459903391e-06,
+      "loss": 0.7912,
+      "step": 601
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.907837876833044,
+      "learning_rate": 9.990407675351027e-06,
+      "loss": 1.0607,
+      "step": 602
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.583072805757161,
+      "learning_rate": 9.990325540992798e-06,
+      "loss": 1.0303,
+      "step": 603
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.063504134304614,
+      "learning_rate": 9.99024305683447e-06,
+      "loss": 0.9656,
+      "step": 604
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.3739718312431455,
+      "learning_rate": 9.99016022288182e-06,
+      "loss": 1.9083,
+      "step": 605
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.3717803890143556,
+      "learning_rate": 9.990077039140655e-06,
+      "loss": 0.6516,
+      "step": 606
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.8385892151528815,
+      "learning_rate": 9.989993505616807e-06,
+      "loss": 0.7832,
+      "step": 607
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.415013525607286,
+      "learning_rate": 9.989909622316132e-06,
+      "loss": 0.9748,
+      "step": 608
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.30838058965553,
+      "learning_rate": 9.989825389244508e-06,
+      "loss": 1.1293,
+      "step": 609
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.9641337437991773,
+      "learning_rate": 9.989740806407839e-06,
+      "loss": 0.7073,
+      "step": 610
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.539025450895296,
+      "learning_rate": 9.989655873812054e-06,
+      "loss": 0.4529,
+      "step": 611
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 8.404028963526692,
+      "learning_rate": 9.98957059146311e-06,
+      "loss": 1.2211,
+      "step": 612
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.464389130013617,
+      "learning_rate": 9.989484959366981e-06,
+      "loss": 0.9611,
+      "step": 613
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.268794174288258,
+      "learning_rate": 9.98939897752967e-06,
+      "loss": 1.5645,
+      "step": 614
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.190325579863303,
+      "learning_rate": 9.989312645957206e-06,
+      "loss": 0.9776,
+      "step": 615
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.519508259471203,
+      "learning_rate": 9.989225964655638e-06,
+      "loss": 0.7953,
+      "step": 616
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.031059724485583,
+      "learning_rate": 9.989138933631042e-06,
+      "loss": 0.5061,
+      "step": 617
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.168095468868191,
+      "learning_rate": 9.989051552889521e-06,
+      "loss": 1.3652,
+      "step": 618
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.7765541732269625,
+      "learning_rate": 9.9889638224372e-06,
+      "loss": 0.8034,
+      "step": 619
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.109554857741848,
+      "learning_rate": 9.988875742280225e-06,
+      "loss": 0.9353,
+      "step": 620
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.082910298532423,
+      "learning_rate": 9.988787312424773e-06,
+      "loss": 0.8639,
+      "step": 621
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.8718090115890407,
+      "learning_rate": 9.98869853287704e-06,
+      "loss": 0.6346,
+      "step": 622
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.209133554282209,
+      "learning_rate": 9.988609403643254e-06,
+      "loss": 1.0097,
+      "step": 623
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.2825195486259995,
+      "learning_rate": 9.988519924729658e-06,
+      "loss": 1.0552,
+      "step": 624
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.090137938891286,
+      "learning_rate": 9.988430096142523e-06,
+      "loss": 0.9436,
+      "step": 625
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.669478879594225,
+      "learning_rate": 9.98833991788815e-06,
+      "loss": 0.4304,
+      "step": 626
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.7927076917738867,
+      "learning_rate": 9.988249389972859e-06,
+      "loss": 0.4305,
+      "step": 627
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.556812199763877,
+      "learning_rate": 9.988158512402993e-06,
+      "loss": 1.3901,
+      "step": 628
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.615049583677811,
+      "learning_rate": 9.988067285184924e-06,
+      "loss": 0.5666,
+      "step": 629
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.2011693520645546,
+      "learning_rate": 9.987975708325048e-06,
+      "loss": 0.7677,
+      "step": 630
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.2099903765384887,
+      "learning_rate": 9.98788378182978e-06,
+      "loss": 0.5586,
+      "step": 631
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2163063910389378,
+      "learning_rate": 9.987791505705568e-06,
+      "loss": 0.2413,
+      "step": 632
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.488636294376988,
+      "learning_rate": 9.987698879958879e-06,
+      "loss": 0.9507,
+      "step": 633
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.9033659480857534,
+      "learning_rate": 9.987605904596203e-06,
+      "loss": 0.2726,
+      "step": 634
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.480796139908746,
+      "learning_rate": 9.987512579624061e-06,
+      "loss": 0.8255,
+      "step": 635
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.527278591142934,
+      "learning_rate": 9.987418905048993e-06,
+      "loss": 0.9837,
+      "step": 636
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.120787858966511,
+      "learning_rate": 9.987324880877564e-06,
+      "loss": 0.6009,
+      "step": 637
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.656164287100769,
+      "learning_rate": 9.987230507116366e-06,
+      "loss": 0.951,
+      "step": 638
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.721155595685185,
+      "learning_rate": 9.987135783772014e-06,
+      "loss": 1.1902,
+      "step": 639
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.837561241031265,
+      "learning_rate": 9.987040710851148e-06,
+      "loss": 1.3766,
+      "step": 640
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.4364747725629443,
+      "learning_rate": 9.986945288360431e-06,
+      "loss": 0.5502,
+      "step": 641
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.505635828322947,
+      "learning_rate": 9.986849516306554e-06,
+      "loss": 1.3267,
+      "step": 642
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.644333592153027,
+      "learning_rate": 9.986753394696227e-06,
+      "loss": 0.6654,
+      "step": 643
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.0669290688594195,
+      "learning_rate": 9.98665692353619e-06,
+      "loss": 1.1344,
+      "step": 644
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.344739819105405,
+      "learning_rate": 9.986560102833206e-06,
+      "loss": 0.9141,
+      "step": 645
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.8933191624197305,
+      "learning_rate": 9.986462932594059e-06,
+      "loss": 0.487,
+      "step": 646
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.096586790266863,
+      "learning_rate": 9.986365412825562e-06,
+      "loss": 0.7367,
+      "step": 647
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.303554708517804,
+      "learning_rate": 9.98626754353455e-06,
+      "loss": 0.8811,
+      "step": 648
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.3709015554748842,
+      "learning_rate": 9.986169324727883e-06,
+      "loss": 0.617,
+      "step": 649
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.4762846409644337,
+      "learning_rate": 9.986070756412447e-06,
+      "loss": 0.4911,
+      "step": 650
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.601479155779438,
+      "learning_rate": 9.98597183859515e-06,
+      "loss": 0.6638,
+      "step": 651
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.004344661451242,
+      "learning_rate": 9.985872571282927e-06,
+      "loss": 0.6485,
+      "step": 652
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.256974623442099,
+      "learning_rate": 9.985772954482736e-06,
+      "loss": 1.076,
+      "step": 653
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.072351574263503,
+      "learning_rate": 9.985672988201557e-06,
+      "loss": 1.0198,
+      "step": 654
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.2702241501346734,
+      "learning_rate": 9.9855726724464e-06,
+      "loss": 0.6186,
+      "step": 655
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.057842319740918,
+      "learning_rate": 9.985472007224296e-06,
+      "loss": 1.0415,
+      "step": 656
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.149812927136185,
+      "learning_rate": 9.9853709925423e-06,
+      "loss": 1.3504,
+      "step": 657
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.477356519779247,
+      "learning_rate": 9.985269628407497e-06,
+      "loss": 0.957,
+      "step": 658
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.9499863604689995,
+      "learning_rate": 9.985167914826986e-06,
+      "loss": 0.9287,
+      "step": 659
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.666449716748098,
+      "learning_rate": 9.9850658518079e-06,
+      "loss": 1.204,
+      "step": 660
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.5396153481196504,
+      "learning_rate": 9.984963439357395e-06,
+      "loss": 0.4128,
+      "step": 661
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.9862693153077973,
+      "learning_rate": 9.984860677482646e-06,
+      "loss": 0.531,
+      "step": 662
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.3886584701639064,
+      "learning_rate": 9.984757566190856e-06,
+      "loss": 0.5735,
+      "step": 663
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.718073410688344,
+      "learning_rate": 9.984654105489258e-06,
+      "loss": 1.4752,
+      "step": 664
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6470470212912942,
+      "learning_rate": 9.984550295385097e-06,
+      "loss": 0.1962,
+      "step": 665
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.677375978952525,
+      "learning_rate": 9.984446135885657e-06,
+      "loss": 1.2061,
+      "step": 666
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.1751126586821545,
+      "learning_rate": 9.984341626998234e-06,
+      "loss": 1.6522,
+      "step": 667
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.198772872513048,
+      "learning_rate": 9.984236768730152e-06,
+      "loss": 0.9486,
+      "step": 668
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.550318005535857,
+      "learning_rate": 9.984131561088766e-06,
+      "loss": 0.8527,
+      "step": 669
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.8438636208886927,
+      "learning_rate": 9.98402600408145e-06,
+      "loss": 0.6577,
+      "step": 670
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.408862532659198,
+      "learning_rate": 9.983920097715599e-06,
+      "loss": 0.9216,
+      "step": 671
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.077809477522914,
+      "learning_rate": 9.983813841998639e-06,
+      "loss": 1.0509,
+      "step": 672
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.2653328768983725,
+      "learning_rate": 9.98370723693802e-06,
+      "loss": 0.567,
+      "step": 673
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.705621516826731,
+      "learning_rate": 9.983600282541213e-06,
+      "loss": 1.3581,
+      "step": 674
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.5134736946043725,
+      "learning_rate": 9.983492978815716e-06,
+      "loss": 0.7735,
+      "step": 675
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.973311993907203,
+      "learning_rate": 9.983385325769047e-06,
+      "loss": 0.7616,
+      "step": 676
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.666889227432155,
+      "learning_rate": 9.983277323408755e-06,
+      "loss": 1.2404,
+      "step": 677
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.986813884345034,
+      "learning_rate": 9.983168971742411e-06,
+      "loss": 1.295,
+      "step": 678
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.530210032015996,
+      "learning_rate": 9.983060270777607e-06,
+      "loss": 1.124,
+      "step": 679
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.079685519249418,
+      "learning_rate": 9.982951220521965e-06,
+      "loss": 1.0623,
+      "step": 680
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.617538825588403,
+      "learning_rate": 9.98284182098313e-06,
+      "loss": 1.4321,
+      "step": 681
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.736195546177727,
+      "learning_rate": 9.982732072168768e-06,
+      "loss": 0.9926,
+      "step": 682
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.911523379676589,
+      "learning_rate": 9.982621974086572e-06,
+      "loss": 1.1862,
+      "step": 683
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.679514520970694,
+      "learning_rate": 9.98251152674426e-06,
+      "loss": 1.266,
+      "step": 684
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.072991628720345,
+      "learning_rate": 9.982400730149574e-06,
+      "loss": 0.9329,
+      "step": 685
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.69136136575507,
+      "learning_rate": 9.98228958431028e-06,
+      "loss": 0.9406,
+      "step": 686
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.591915346611839,
+      "learning_rate": 9.982178089234171e-06,
+      "loss": 1.3627,
+      "step": 687
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.747909856781231,
+      "learning_rate": 9.982066244929058e-06,
+      "loss": 1.2625,
+      "step": 688
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 7.065528326230627,
+      "learning_rate": 9.981954051402785e-06,
+      "loss": 1.0663,
+      "step": 689
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.0001909990159215,
+      "learning_rate": 9.981841508663214e-06,
+      "loss": 0.7058,
+      "step": 690
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.3998445578928598,
+      "learning_rate": 9.981728616718234e-06,
+      "loss": 0.7326,
+      "step": 691
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.855144959078903,
+      "learning_rate": 9.98161537557576e-06,
+      "loss": 0.6568,
+      "step": 692
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.066439170079921,
+      "learning_rate": 9.981501785243725e-06,
+      "loss": 1.1099,
+      "step": 693
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.380398469657443,
+      "learning_rate": 9.981387845730097e-06,
+      "loss": 1.0579,
+      "step": 694
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.897195008908236,
+      "learning_rate": 9.981273557042861e-06,
+      "loss": 1.1187,
+      "step": 695
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.252493751213305,
+      "learning_rate": 9.981158919190024e-06,
+      "loss": 1.4001,
+      "step": 696
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.075321407345318,
+      "learning_rate": 9.981043932179629e-06,
+      "loss": 0.654,
+      "step": 697
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.737293080752598,
+      "learning_rate": 9.980928596019727e-06,
+      "loss": 0.9636,
+      "step": 698
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.220734829615724,
+      "learning_rate": 9.98081291071841e-06,
+      "loss": 1.1473,
+      "step": 699
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.338659674445185,
+      "learning_rate": 9.980696876283785e-06,
+      "loss": 0.7959,
+      "step": 700
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.4732476751320323,
+      "learning_rate": 9.980580492723984e-06,
+      "loss": 0.6809,
+      "step": 701
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.506473008904325,
+      "learning_rate": 9.980463760047167e-06,
+      "loss": 0.853,
+      "step": 702
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.132017256774828,
+      "learning_rate": 9.980346678261515e-06,
+      "loss": 0.9536,
+      "step": 703
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.583142094061532,
+      "learning_rate": 9.980229247375236e-06,
+      "loss": 0.447,
+      "step": 704
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.83052530122639,
+      "learning_rate": 9.980111467396561e-06,
+      "loss": 0.389,
+      "step": 705
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.794608506821916,
+      "learning_rate": 9.979993338333745e-06,
+      "loss": 1.4973,
+      "step": 706
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.703224751249668,
+      "learning_rate": 9.979874860195068e-06,
+      "loss": 0.8389,
+      "step": 707
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.1532072810420435,
+      "learning_rate": 9.979756032988837e-06,
+      "loss": 0.6571,
+      "step": 708
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.015780174494101,
+      "learning_rate": 9.979636856723379e-06,
+      "loss": 1.3718,
+      "step": 709
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.2193398459983555,
+      "learning_rate": 9.97951733140705e-06,
+      "loss": 0.8092,
+      "step": 710
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.705786850338479,
+      "learning_rate": 9.979397457048226e-06,
+      "loss": 0.9143,
+      "step": 711
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.9202053826916483,
+      "learning_rate": 9.97927723365531e-06,
+      "loss": 0.7494,
+      "step": 712
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.633324076574433,
+      "learning_rate": 9.979156661236733e-06,
+      "loss": 0.8238,
+      "step": 713
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.2712391089999926,
+      "learning_rate": 9.97903573980094e-06,
+      "loss": 0.4017,
+      "step": 714
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.71236457496792,
+      "learning_rate": 9.978914469356413e-06,
+      "loss": 0.8248,
+      "step": 715
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.412765845449585,
+      "learning_rate": 9.97879284991165e-06,
+      "loss": 1.0794,
+      "step": 716
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.969742110862581,
+      "learning_rate": 9.978670881475173e-06,
+      "loss": 0.832,
+      "step": 717
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.931801381819527,
+      "learning_rate": 9.978548564055537e-06,
+      "loss": 0.9876,
+      "step": 718
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.316852797200995,
+      "learning_rate": 9.978425897661312e-06,
+      "loss": 0.8552,
+      "step": 719
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2826681827328124,
+      "learning_rate": 9.978302882301098e-06,
+      "loss": 0.1575,
+      "step": 720
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.760416074010551,
+      "learning_rate": 9.97817951798352e-06,
+      "loss": 1.1668,
+      "step": 721
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.310256577274185,
+      "learning_rate": 9.97805580471722e-06,
+      "loss": 1.0728,
+      "step": 722
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.802609025604089,
+      "learning_rate": 9.977931742510873e-06,
+      "loss": 1.4633,
+      "step": 723
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8533623374374455,
+      "learning_rate": 9.977807331373176e-06,
+      "loss": 0.2901,
+      "step": 724
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.6339167914487835,
+      "learning_rate": 9.977682571312847e-06,
+      "loss": 0.5014,
+      "step": 725
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.062861416110086,
+      "learning_rate": 9.977557462338635e-06,
+      "loss": 0.7564,
+      "step": 726
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.059168335509616,
+      "learning_rate": 9.977432004459306e-06,
+      "loss": 0.549,
+      "step": 727
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.702152763867874,
+      "learning_rate": 9.977306197683656e-06,
+      "loss": 1.2333,
+      "step": 728
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.630310813593966,
+      "learning_rate": 9.977180042020502e-06,
+      "loss": 1.1659,
+      "step": 729
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.511468366268989,
+      "learning_rate": 9.977053537478686e-06,
+      "loss": 1.0232,
+      "step": 730
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.0047996416749823,
+      "learning_rate": 9.976926684067082e-06,
+      "loss": 0.4844,
+      "step": 731
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.247586868009353,
+      "learning_rate": 9.976799481794573e-06,
+      "loss": 0.4085,
+      "step": 732
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.131277338264275,
+      "learning_rate": 9.976671930670081e-06,
+      "loss": 0.6309,
+      "step": 733
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.2916927804561515,
+      "learning_rate": 9.976544030702546e-06,
+      "loss": 0.8517,
+      "step": 734
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.487087260109036,
+      "learning_rate": 9.97641578190093e-06,
+      "loss": 0.7722,
+      "step": 735
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.2674515774054473,
+      "learning_rate": 9.976287184274228e-06,
+      "loss": 0.5029,
+      "step": 736
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.887167598931923,
+      "learning_rate": 9.976158237831449e-06,
+      "loss": 0.5151,
+      "step": 737
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.117472870245781,
+      "learning_rate": 9.976028942581636e-06,
+      "loss": 0.9479,
+      "step": 738
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.499321395252827,
+      "learning_rate": 9.975899298533848e-06,
+      "loss": 0.5442,
+      "step": 739
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.571042528541903,
+      "learning_rate": 9.975769305697174e-06,
+      "loss": 0.5613,
+      "step": 740
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.101831596434222,
+      "learning_rate": 9.975638964080727e-06,
+      "loss": 0.9055,
+      "step": 741
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.5100041104182385,
+      "learning_rate": 9.975508273693643e-06,
+      "loss": 1.3315,
+      "step": 742
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.0554279292222346,
+      "learning_rate": 9.975377234545083e-06,
+      "loss": 0.6267,
+      "step": 743
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.89859289725527,
+      "learning_rate": 9.97524584664423e-06,
+      "loss": 1.6799,
+      "step": 744
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.430271664446078,
+      "learning_rate": 9.975114110000297e-06,
+      "loss": 0.8808,
+      "step": 745
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.731882166706876,
+      "learning_rate": 9.974982024622517e-06,
+      "loss": 1.1066,
+      "step": 746
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.459257443847804,
+      "learning_rate": 9.974849590520148e-06,
+      "loss": 0.7659,
+      "step": 747
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.939537799631192,
+      "learning_rate": 9.974716807702473e-06,
+      "loss": 0.8695,
+      "step": 748
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.520153583674207,
+      "learning_rate": 9.9745836761788e-06,
+      "loss": 0.8502,
+      "step": 749
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.8020858847345793,
+      "learning_rate": 9.974450195958459e-06,
+      "loss": 0.6382,
+      "step": 750
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.8753527561070307,
+      "learning_rate": 9.97431636705081e-06,
+      "loss": 0.7334,
+      "step": 751
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.723888177512531,
+      "learning_rate": 9.974182189465232e-06,
+      "loss": 0.7886,
+      "step": 752
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.128138154428111,
+      "learning_rate": 9.974047663211131e-06,
+      "loss": 1.7984,
+      "step": 753
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.773109744828223,
+      "learning_rate": 9.973912788297933e-06,
+      "loss": 1.306,
+      "step": 754
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.00723529489966,
+      "learning_rate": 9.973777564735097e-06,
+      "loss": 0.8355,
+      "step": 755
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.9820207602050672,
+      "learning_rate": 9.973641992532099e-06,
+      "loss": 0.4745,
+      "step": 756
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.715125249193091,
+      "learning_rate": 9.973506071698444e-06,
+      "loss": 1.0481,
+      "step": 757
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.439607792847339,
+      "learning_rate": 9.973369802243658e-06,
+      "loss": 0.4661,
+      "step": 758
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.530386457711915,
+      "learning_rate": 9.97323318417729e-06,
+      "loss": 0.9343,
+      "step": 759
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.740555661029211,
+      "learning_rate": 9.973096217508925e-06,
+      "loss": 0.7943,
+      "step": 760
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.7744732026579766,
+      "learning_rate": 9.972958902248153e-06,
+      "loss": 0.6054,
+      "step": 761
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.929079530607702,
+      "learning_rate": 9.972821238404607e-06,
+      "loss": 0.3151,
+      "step": 762
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.495765793088585,
+      "learning_rate": 9.972683225987933e-06,
+      "loss": 1.0715,
+      "step": 763
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.467686535685932,
+      "learning_rate": 9.972544865007807e-06,
+      "loss": 0.6673,
+      "step": 764
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.146818085401882,
+      "learning_rate": 9.972406155473925e-06,
+      "loss": 0.8161,
+      "step": 765
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.526183507517289,
+      "learning_rate": 9.972267097396013e-06,
+      "loss": 1.1339,
+      "step": 766
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.391115260744661,
+      "learning_rate": 9.972127690783815e-06,
+      "loss": 1.6191,
+      "step": 767
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.089685092999949,
+      "learning_rate": 9.971987935647106e-06,
+      "loss": 1.0668,
+      "step": 768
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.520289400789311,
+      "learning_rate": 9.97184783199568e-06,
+      "loss": 0.8992,
+      "step": 769
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.625631151177899,
+      "learning_rate": 9.97170737983936e-06,
+      "loss": 1.1493,
+      "step": 770
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.97478303005252,
+      "learning_rate": 9.971566579187988e-06,
+      "loss": 1.1023,
+      "step": 771
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.547320317403259,
+      "learning_rate": 9.971425430051437e-06,
+      "loss": 1.0619,
+      "step": 772
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.636700625448609,
+      "learning_rate": 9.971283932439597e-06,
+      "loss": 1.1449,
+      "step": 773
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.769386076301568,
+      "learning_rate": 9.971142086362392e-06,
+      "loss": 1.3077,
+      "step": 774
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.635243672503049,
+      "learning_rate": 9.970999891829757e-06,
+      "loss": 1.1862,
+      "step": 775
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.848083854458284,
+      "learning_rate": 9.970857348851667e-06,
+      "loss": 0.9362,
+      "step": 776
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.4745444676936006,
+      "learning_rate": 9.970714457438106e-06,
+      "loss": 0.5107,
+      "step": 777
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.489989337493255,
+      "learning_rate": 9.970571217599096e-06,
+      "loss": 0.9107,
+      "step": 778
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.096443541651896,
+      "learning_rate": 9.970427629344676e-06,
+      "loss": 0.4118,
+      "step": 779
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.530946936104318,
+      "learning_rate": 9.970283692684911e-06,
+      "loss": 0.8377,
+      "step": 780
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.22289394368659,
+      "learning_rate": 9.97013940762989e-06,
+      "loss": 1.3088,
+      "step": 781
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.453490008200965,
+      "learning_rate": 9.969994774189726e-06,
+      "loss": 1.451,
+      "step": 782
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.76093404920054,
+      "learning_rate": 9.969849792374558e-06,
+      "loss": 1.0252,
+      "step": 783
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.57953489925236,
+      "learning_rate": 9.969704462194549e-06,
+      "loss": 1.4084,
+      "step": 784
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.4399668508195123,
+      "learning_rate": 9.969558783659884e-06,
+      "loss": 0.5534,
+      "step": 785
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.7564780956274015,
+      "learning_rate": 9.969412756780776e-06,
+      "loss": 1.0238,
+      "step": 786
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.826815046400798,
+      "learning_rate": 9.969266381567462e-06,
+      "loss": 0.94,
+      "step": 787
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.080648998198349,
+      "learning_rate": 9.9691196580302e-06,
+      "loss": 0.897,
+      "step": 788
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.6470320822489994,
+      "learning_rate": 9.968972586179275e-06,
+      "loss": 0.8945,
+      "step": 789
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.408120703606715,
+      "learning_rate": 9.968825166024999e-06,
+      "loss": 0.4984,
+      "step": 790
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 7.859349302027108,
+      "learning_rate": 9.968677397577701e-06,
+      "loss": 1.8087,
+      "step": 791
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.3390940932775894,
+      "learning_rate": 9.968529280847743e-06,
+      "loss": 0.474,
+      "step": 792
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.1126987657491423,
+      "learning_rate": 9.968380815845504e-06,
+      "loss": 0.6885,
+      "step": 793
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.074268950114531,
+      "learning_rate": 9.968232002581394e-06,
+      "loss": 0.8573,
+      "step": 794
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.5693261133819005,
+      "learning_rate": 9.96808284106584e-06,
+      "loss": 1.0474,
+      "step": 795
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.392596789543054,
+      "learning_rate": 9.967933331309302e-06,
+      "loss": 0.7573,
+      "step": 796
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.7715217973867903,
+      "learning_rate": 9.967783473322258e-06,
+      "loss": 0.6199,
+      "step": 797
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.074185896638036,
+      "learning_rate": 9.96763326711521e-06,
+      "loss": 0.4578,
+      "step": 798
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.176116951214148,
+      "learning_rate": 9.967482712698694e-06,
+      "loss": 0.6773,
+      "step": 799
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.76439860919025,
+      "learning_rate": 9.967331810083254e-06,
+      "loss": 1.1874,
+      "step": 800
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.582020589446579,
+      "learning_rate": 9.967180559279472e-06,
+      "loss": 0.9202,
+      "step": 801
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0977116361481984,
+      "learning_rate": 9.967028960297954e-06,
+      "loss": 0.4643,
+      "step": 802
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.098250568478063,
+      "learning_rate": 9.966877013149319e-06,
+      "loss": 0.4353,
+      "step": 803
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.337788120873333,
+      "learning_rate": 9.966724717844222e-06,
+      "loss": 0.7149,
+      "step": 804
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.694222423633388,
+      "learning_rate": 9.966572074393337e-06,
+      "loss": 1.0502,
+      "step": 805
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.776286287183273,
+      "learning_rate": 9.966419082807366e-06,
+      "loss": 0.5987,
+      "step": 806
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.21966864717937,
+      "learning_rate": 9.96626574309703e-06,
+      "loss": 0.4904,
+      "step": 807
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.924492528369482,
+      "learning_rate": 9.966112055273078e-06,
+      "loss": 1.1346,
+      "step": 808
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.1712724961958045,
+      "learning_rate": 9.965958019346284e-06,
+      "loss": 0.6334,
+      "step": 809
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.8522795318151353,
+      "learning_rate": 9.965803635327445e-06,
+      "loss": 0.7694,
+      "step": 810
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.574605370309623,
+      "learning_rate": 9.965648903227383e-06,
+      "loss": 0.261,
+      "step": 811
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.705029403781401,
+      "learning_rate": 9.965493823056943e-06,
+      "loss": 0.6301,
+      "step": 812
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.734011167965443,
+      "learning_rate": 9.965338394826995e-06,
+      "loss": 1.5186,
+      "step": 813
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.9008709218185853,
+      "learning_rate": 9.965182618548437e-06,
+      "loss": 0.5122,
+      "step": 814
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.156045034777844,
+      "learning_rate": 9.965026494232184e-06,
+      "loss": 0.2935,
+      "step": 815
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.339119067796793,
+      "learning_rate": 9.964870021889181e-06,
+      "loss": 1.0165,
+      "step": 816
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.134637446860939,
+      "learning_rate": 9.964713201530399e-06,
+      "loss": 0.7808,
+      "step": 817
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.597641511494216,
+      "learning_rate": 9.964556033166826e-06,
+      "loss": 0.9558,
+      "step": 818
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.3933383102397845,
+      "learning_rate": 9.964398516809482e-06,
+      "loss": 0.7344,
+      "step": 819
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.0015608056776495,
+      "learning_rate": 9.964240652469408e-06,
+      "loss": 0.6732,
+      "step": 820
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.065452808847677,
+      "learning_rate": 9.964082440157668e-06,
+      "loss": 0.6751,
+      "step": 821
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.916773694484184,
+      "learning_rate": 9.963923879885351e-06,
+      "loss": 0.6768,
+      "step": 822
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.630978787822194,
+      "learning_rate": 9.963764971663575e-06,
+      "loss": 0.4979,
+      "step": 823
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.170772583289991,
+      "learning_rate": 9.963605715503477e-06,
+      "loss": 1.2307,
+      "step": 824
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.211658909815121,
+      "learning_rate": 9.96344611141622e-06,
+      "loss": 0.9501,
+      "step": 825
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.606969981147839,
+      "learning_rate": 9.96328615941299e-06,
+      "loss": 1.4042,
+      "step": 826
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.4391829172628925,
+      "learning_rate": 9.963125859505e-06,
+      "loss": 0.604,
+      "step": 827
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.6553164080327796,
+      "learning_rate": 9.96296521170349e-06,
+      "loss": 0.4192,
+      "step": 828
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.426383397582067,
+      "learning_rate": 9.962804216019715e-06,
+      "loss": 0.3705,
+      "step": 829
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.210417920800824,
+      "learning_rate": 9.962642872464964e-06,
+      "loss": 1.218,
+      "step": 830
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.627215980781415,
+      "learning_rate": 9.962481181050544e-06,
+      "loss": 1.1948,
+      "step": 831
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.6291433620745694,
+      "learning_rate": 9.96231914178779e-06,
+      "loss": 0.9308,
+      "step": 832
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.467086133481989,
+      "learning_rate": 9.962156754688062e-06,
+      "loss": 1.166,
+      "step": 833
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.5665160649933267,
+      "learning_rate": 9.96199401976274e-06,
+      "loss": 0.7913,
+      "step": 834
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.3983297626025517,
+      "learning_rate": 9.96183093702323e-06,
+      "loss": 0.6424,
+      "step": 835
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.069701322695037,
+      "learning_rate": 9.961667506480967e-06,
+      "loss": 1.5727,
+      "step": 836
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.402442563901685,
+      "learning_rate": 9.961503728147405e-06,
+      "loss": 1.4001,
+      "step": 837
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.720680137374831,
+      "learning_rate": 9.961339602034026e-06,
+      "loss": 1.0963,
+      "step": 838
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.0981566758358903,
+      "learning_rate": 9.96117512815233e-06,
+      "loss": 0.4581,
+      "step": 839
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.2511679821777086,
+      "learning_rate": 9.96101030651385e-06,
+      "loss": 0.4581,
+      "step": 840
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.790266578683795,
+      "learning_rate": 9.960845137130137e-06,
+      "loss": 0.7372,
+      "step": 841
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.982491338777422,
+      "learning_rate": 9.96067962001277e-06,
+      "loss": 0.972,
+      "step": 842
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.7105151457719914,
+      "learning_rate": 9.96051375517335e-06,
+      "loss": 0.8763,
+      "step": 843
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.185155104774775,
+      "learning_rate": 9.960347542623506e-06,
+      "loss": 0.5442,
+      "step": 844
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.156382246620666,
+      "learning_rate": 9.960180982374884e-06,
+      "loss": 0.4527,
+      "step": 845
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.4217843784477244,
+      "learning_rate": 9.960014074439164e-06,
+      "loss": 0.5007,
+      "step": 846
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.391806637780469,
+      "learning_rate": 9.959846818828041e-06,
+      "loss": 1.3905,
+      "step": 847
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.459019607298156,
+      "learning_rate": 9.959679215553244e-06,
+      "loss": 0.6361,
+      "step": 848
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.619323467927718,
+      "learning_rate": 9.959511264626518e-06,
+      "loss": 0.7784,
+      "step": 849
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.728782464117148,
+      "learning_rate": 9.959342966059636e-06,
+      "loss": 1.2141,
+      "step": 850
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.738622637108753,
+      "learning_rate": 9.959174319864395e-06,
+      "loss": 1.1925,
+      "step": 851
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.479658138287649,
+      "learning_rate": 9.959005326052615e-06,
+      "loss": 1.0201,
+      "step": 852
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.614085673476612,
+      "learning_rate": 9.958835984636146e-06,
+      "loss": 1.165,
+      "step": 853
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.5184482099574366,
+      "learning_rate": 9.958666295626854e-06,
+      "loss": 0.8741,
+      "step": 854
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 7.138268357473538,
+      "learning_rate": 9.958496259036635e-06,
+      "loss": 1.638,
+      "step": 855
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.042464483331498,
+      "learning_rate": 9.958325874877408e-06,
+      "loss": 0.7941,
+      "step": 856
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.184549738302962,
+      "learning_rate": 9.958155143161115e-06,
+      "loss": 0.915,
+      "step": 857
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.664162701019237,
+      "learning_rate": 9.957984063899727e-06,
+      "loss": 1.1962,
+      "step": 858
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8646671334584424,
+      "learning_rate": 9.95781263710523e-06,
+      "loss": 0.308,
+      "step": 859
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.639081617480649,
+      "learning_rate": 9.957640862789644e-06,
+      "loss": 1.092,
+      "step": 860
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.5841789111561875,
+      "learning_rate": 9.95746874096501e-06,
+      "loss": 0.5315,
+      "step": 861
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.234477795410627,
+      "learning_rate": 9.957296271643393e-06,
+      "loss": 0.769,
+      "step": 862
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.313150775350866,
+      "learning_rate": 9.957123454836882e-06,
+      "loss": 0.9755,
+      "step": 863
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.115863019680228,
+      "learning_rate": 9.95695029055759e-06,
+      "loss": 1.3551,
+      "step": 864
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.635816693720637,
+      "learning_rate": 9.956776778817654e-06,
+      "loss": 1.222,
+      "step": 865
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.4986707461344753,
+      "learning_rate": 9.956602919629239e-06,
+      "loss": 0.6475,
+      "step": 866
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.7358230723201142,
+      "learning_rate": 9.956428713004529e-06,
+      "loss": 0.5906,
+      "step": 867
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.528042168147081,
+      "learning_rate": 9.956254158955738e-06,
+      "loss": 1.3471,
+      "step": 868
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.613230901572128,
+      "learning_rate": 9.9560792574951e-06,
+      "loss": 0.9629,
+      "step": 869
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.86962822781077,
+      "learning_rate": 9.955904008634876e-06,
+      "loss": 1.3635,
+      "step": 870
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.695620277128768,
+      "learning_rate": 9.955728412387347e-06,
+      "loss": 1.0614,
+      "step": 871
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.794859671216474,
+      "learning_rate": 9.955552468764825e-06,
+      "loss": 1.0326,
+      "step": 872
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.291780152477322,
+      "learning_rate": 9.955376177779641e-06,
+      "loss": 0.6374,
+      "step": 873
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.777384653007664,
+      "learning_rate": 9.955199539444154e-06,
+      "loss": 0.5032,
+      "step": 874
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.269996080253249,
+      "learning_rate": 9.955022553770743e-06,
+      "loss": 0.7474,
+      "step": 875
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.790840369737783,
+      "learning_rate": 9.954845220771816e-06,
+      "loss": 0.9783,
+      "step": 876
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.356400418330578,
+      "learning_rate": 9.954667540459802e-06,
+      "loss": 1.2314,
+      "step": 877
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.2532838304563505,
+      "learning_rate": 9.954489512847156e-06,
+      "loss": 0.8081,
+      "step": 878
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.461855774922967,
+      "learning_rate": 9.954311137946358e-06,
+      "loss": 0.5668,
+      "step": 879
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.000711677503919,
+      "learning_rate": 9.954132415769911e-06,
+      "loss": 0.8234,
+      "step": 880
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.01348735820718,
+      "learning_rate": 9.95395334633034e-06,
+      "loss": 0.8112,
+      "step": 881
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.796895609166545,
+      "learning_rate": 9.953773929640202e-06,
+      "loss": 0.9285,
+      "step": 882
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.523348062987123,
+      "learning_rate": 9.953594165712068e-06,
+      "loss": 0.3327,
+      "step": 883
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.060091186298477,
+      "learning_rate": 9.953414054558543e-06,
+      "loss": 0.7849,
+      "step": 884
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.531720184448185,
+      "learning_rate": 9.95323359619225e-06,
+      "loss": 1.4627,
+      "step": 885
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.124767255143433,
+      "learning_rate": 9.953052790625835e-06,
+      "loss": 1.147,
+      "step": 886
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.174673608668578,
+      "learning_rate": 9.95287163787198e-06,
+      "loss": 1.0221,
+      "step": 887
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.759573418537057,
+      "learning_rate": 9.952690137943374e-06,
+      "loss": 0.7192,
+      "step": 888
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.32649581631083,
+      "learning_rate": 9.952508290852746e-06,
+      "loss": 0.9132,
+      "step": 889
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.49609553767445,
+      "learning_rate": 9.95232609661284e-06,
+      "loss": 0.9195,
+      "step": 890
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.744183694556523,
+      "learning_rate": 9.952143555236426e-06,
+      "loss": 0.5019,
+      "step": 891
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.346790453736447,
+      "learning_rate": 9.9519606667363e-06,
+      "loss": 1.061,
+      "step": 892
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.8526268520642506,
+      "learning_rate": 9.951777431125285e-06,
+      "loss": 0.7299,
+      "step": 893
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.280723702096715,
+      "learning_rate": 9.95159384841622e-06,
+      "loss": 0.8639,
+      "step": 894
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.5124974801196025,
+      "learning_rate": 9.951409918621977e-06,
+      "loss": 1.5656,
+      "step": 895
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.493827302226007,
+      "learning_rate": 9.951225641755447e-06,
+      "loss": 1.2438,
+      "step": 896
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.500693687023412,
+      "learning_rate": 9.951041017829546e-06,
+      "loss": 1.3041,
+      "step": 897
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.256144142745983,
+      "learning_rate": 9.950856046857218e-06,
+      "loss": 1.1518,
+      "step": 898
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.0082769764561474,
+      "learning_rate": 9.950670728851428e-06,
+      "loss": 1.2247,
+      "step": 899
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.9888071975117203,
+      "learning_rate": 9.950485063825164e-06,
+      "loss": 0.8726,
+      "step": 900
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.6636039422830637,
+      "learning_rate": 9.950299051791442e-06,
+      "loss": 0.6408,
+      "step": 901
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.611758379413739,
+      "learning_rate": 9.9501126927633e-06,
+      "loss": 1.0135,
+      "step": 902
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.193649777684185,
+      "learning_rate": 9.949925986753801e-06,
+      "loss": 0.975,
+      "step": 903
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2652496595795037,
+      "learning_rate": 9.949738933776034e-06,
+      "loss": 0.2191,
+      "step": 904
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.00394153281254,
+      "learning_rate": 9.949551533843108e-06,
+      "loss": 1.1315,
+      "step": 905
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.376078409437867,
+      "learning_rate": 9.949363786968161e-06,
+      "loss": 0.8514,
+      "step": 906
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.0947758925959357,
+      "learning_rate": 9.94917569316435e-06,
+      "loss": 0.4318,
+      "step": 907
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.8579514061113698,
+      "learning_rate": 9.948987252444863e-06,
+      "loss": 0.8145,
+      "step": 908
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.341532052086995,
+      "learning_rate": 9.948798464822908e-06,
+      "loss": 1.0982,
+      "step": 909
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.061424905844848,
+      "learning_rate": 9.948609330311717e-06,
+      "loss": 1.2016,
+      "step": 910
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.975840773315984,
+      "learning_rate": 9.94841984892455e-06,
+      "loss": 1.241,
+      "step": 911
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.617213604441644,
+      "learning_rate": 9.948230020674685e-06,
+      "loss": 0.409,
+      "step": 912
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.751113151861608,
+      "learning_rate": 9.948039845575433e-06,
+      "loss": 1.0532,
+      "step": 913
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.045895048439603,
+      "learning_rate": 9.947849323640119e-06,
+      "loss": 1.2309,
+      "step": 914
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.5548904890189315,
+      "learning_rate": 9.947658454882102e-06,
+      "loss": 1.2756,
+      "step": 915
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.567909285488109,
+      "learning_rate": 9.947467239314759e-06,
+      "loss": 1.6017,
+      "step": 916
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.686842726080642,
+      "learning_rate": 9.947275676951493e-06,
+      "loss": 0.9778,
+      "step": 917
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 6.261416529520622,
+      "learning_rate": 9.947083767805736e-06,
+      "loss": 1.2024,
+      "step": 918
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1007870673806903,
+      "learning_rate": 9.946891511890934e-06,
+      "loss": 0.3884,
+      "step": 919
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.098588168379796,
+      "learning_rate": 9.946698909220567e-06,
+      "loss": 0.5807,
+      "step": 920
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.367683093526055,
+      "learning_rate": 9.946505959808133e-06,
+      "loss": 0.9718,
+      "step": 921
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.9092380880050563,
+      "learning_rate": 9.94631266366716e-06,
+      "loss": 0.6939,
+      "step": 922
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 6.2437644983208935,
+      "learning_rate": 9.946119020811196e-06,
+      "loss": 1.6135,
+      "step": 923
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.559272956140806,
+      "learning_rate": 9.945925031253814e-06,
+      "loss": 0.5448,
+      "step": 924
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.824143064798661,
+      "learning_rate": 9.945730695008611e-06,
+      "loss": 0.9135,
+      "step": 925
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.722676137253339,
+      "learning_rate": 9.94553601208921e-06,
+      "loss": 1.153,
+      "step": 926
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.396639886547055,
+      "learning_rate": 9.94534098250926e-06,
+      "loss": 0.6553,
+      "step": 927
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.255260531499106,
+      "learning_rate": 9.945145606282427e-06,
+      "loss": 1.4183,
+      "step": 928
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.9560741446234613,
+      "learning_rate": 9.944949883422409e-06,
+      "loss": 0.8848,
+      "step": 929
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.44099530624096,
+      "learning_rate": 9.944753813942924e-06,
+      "loss": 1.0808,
+      "step": 930
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.612367341298937,
+      "learning_rate": 9.944557397857717e-06,
+      "loss": 1.7301,
+      "step": 931
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.774377825851068,
+      "learning_rate": 9.944360635180554e-06,
+      "loss": 1.2681,
+      "step": 932
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.4851455553756505,
+      "learning_rate": 9.94416352592523e-06,
+      "loss": 1.0088,
+      "step": 933
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.6727752379053444,
+      "learning_rate": 9.94396607010556e-06,
+      "loss": 0.799,
+      "step": 934
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.661571569893697,
+      "learning_rate": 9.943768267735384e-06,
+      "loss": 1.2836,
+      "step": 935
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.73506630653504,
+      "learning_rate": 9.943570118828569e-06,
+      "loss": 1.107,
+      "step": 936
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.216817841660681,
+      "learning_rate": 9.943371623399001e-06,
+      "loss": 1.3673,
+      "step": 937
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.588521825500617,
+      "learning_rate": 9.943172781460596e-06,
+      "loss": 0.9086,
+      "step": 938
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.840878788819319,
+      "learning_rate": 9.942973593027295e-06,
+      "loss": 0.8751,
+      "step": 939
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.159323698725976,
+      "learning_rate": 9.942774058113053e-06,
+      "loss": 0.969,
+      "step": 940
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.500036698511603,
+      "learning_rate": 9.942574176731863e-06,
+      "loss": 0.9783,
+      "step": 941
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.637897868620929,
+      "learning_rate": 9.942373948897732e-06,
+      "loss": 1.1139,
+      "step": 942
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.260102017631009,
+      "learning_rate": 9.942173374624697e-06,
+      "loss": 1.0704,
+      "step": 943
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.6440686740791235,
+      "learning_rate": 9.941972453926817e-06,
+      "loss": 0.3392,
+      "step": 944
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.4619149305092707,
+      "learning_rate": 9.941771186818176e-06,
+      "loss": 0.7033,
+      "step": 945
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.2853503596861815,
+      "learning_rate": 9.941569573312882e-06,
+      "loss": 0.7557,
+      "step": 946
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.4535644407032398,
+      "learning_rate": 9.941367613425064e-06,
+      "loss": 0.5132,
+      "step": 947
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 7.988273824309304,
+      "learning_rate": 9.941165307168883e-06,
+      "loss": 1.0222,
+      "step": 948
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.797267330365143,
+      "learning_rate": 9.94096265455852e-06,
+      "loss": 0.8054,
+      "step": 949
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.451973933847118,
+      "learning_rate": 9.940759655608174e-06,
+      "loss": 1.2013,
+      "step": 950
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.937150092892459,
+      "learning_rate": 9.940556310332081e-06,
+      "loss": 1.0607,
+      "step": 951
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.48017657805159,
+      "learning_rate": 9.94035261874449e-06,
+      "loss": 1.2165,
+      "step": 952
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 6.6979240049013375,
+      "learning_rate": 9.940148580859684e-06,
+      "loss": 1.169,
+      "step": 953
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.048990796591552,
+      "learning_rate": 9.93994419669196e-06,
+      "loss": 1.193,
+      "step": 954
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 7.120682486408936,
+      "learning_rate": 9.939739466255646e-06,
+      "loss": 1.006,
+      "step": 955
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.25295719086572,
+      "learning_rate": 9.939534389565096e-06,
+      "loss": 0.5601,
+      "step": 956
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.9677365404338754,
+      "learning_rate": 9.939328966634679e-06,
+      "loss": 0.5047,
+      "step": 957
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.720941728762174,
+      "learning_rate": 9.9391231974788e-06,
+      "loss": 1.2215,
+      "step": 958
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7910568012463464,
+      "learning_rate": 9.938917082111878e-06,
+      "loss": 0.2296,
+      "step": 959
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.461023834809202,
+      "learning_rate": 9.938710620548363e-06,
+      "loss": 0.8519,
+      "step": 960
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.880190533803183,
+      "learning_rate": 9.938503812802726e-06,
+      "loss": 0.8242,
+      "step": 961
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.93656669188537,
+      "learning_rate": 9.938296658889467e-06,
+      "loss": 0.7574,
+      "step": 962
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.2002274052516855,
+      "learning_rate": 9.938089158823101e-06,
+      "loss": 0.9766,
+      "step": 963
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.8287118106163005,
+      "learning_rate": 9.937881312618178e-06,
+      "loss": 0.984,
+      "step": 964
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 6.032433365734891,
+      "learning_rate": 9.937673120289264e-06,
+      "loss": 1.3364,
+      "step": 965
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.624076949651301,
+      "learning_rate": 9.937464581850952e-06,
+      "loss": 0.2272,
+      "step": 966
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.46448601887881,
+      "learning_rate": 9.937255697317862e-06,
+      "loss": 0.6506,
+      "step": 967
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.687183703503713,
+      "learning_rate": 9.937046466704635e-06,
+      "loss": 1.1185,
+      "step": 968
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.12344443076618,
+      "learning_rate": 9.936836890025934e-06,
+      "loss": 0.6552,
+      "step": 969
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.008246852671144,
+      "learning_rate": 9.936626967296454e-06,
+      "loss": 0.3855,
+      "step": 970
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.727741706894561,
+      "learning_rate": 9.936416698530908e-06,
+      "loss": 1.3848,
+      "step": 971
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.0058232533086335,
+      "learning_rate": 9.936206083744036e-06,
+      "loss": 0.607,
+      "step": 972
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.474463900134665,
+      "learning_rate": 9.935995122950597e-06,
+      "loss": 0.9998,
+      "step": 973
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.71840698955997,
+      "learning_rate": 9.935783816165384e-06,
+      "loss": 0.4255,
+      "step": 974
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.69033891376098,
+      "learning_rate": 9.935572163403205e-06,
+      "loss": 1.203,
+      "step": 975
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.058363801460739,
+      "learning_rate": 9.935360164678897e-06,
+      "loss": 0.9688,
+      "step": 976
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.1298057644289985,
+      "learning_rate": 9.93514782000732e-06,
+      "loss": 0.8337,
+      "step": 977
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.248683450242094,
+      "learning_rate": 9.93493512940336e-06,
+      "loss": 0.718,
+      "step": 978
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.004927639727986,
+      "learning_rate": 9.934722092881923e-06,
+      "loss": 0.3778,
+      "step": 979
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 6.349465302935186,
+      "learning_rate": 9.934508710457944e-06,
+      "loss": 1.3507,
+      "step": 980
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.6202710526830284,
+      "learning_rate": 9.934294982146379e-06,
+      "loss": 0.5398,
+      "step": 981
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.969030239131703,
+      "learning_rate": 9.93408090796221e-06,
+      "loss": 1.1796,
+      "step": 982
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.911435850540964,
+      "learning_rate": 9.933866487920443e-06,
+      "loss": 1.5327,
+      "step": 983
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.304625028599958,
+      "learning_rate": 9.933651722036106e-06,
+      "loss": 1.1508,
+      "step": 984
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.8855327016276147,
+      "learning_rate": 9.933436610324256e-06,
+      "loss": 0.7818,
+      "step": 985
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.284180123451619,
+      "learning_rate": 9.93322115279997e-06,
+      "loss": 0.505,
+      "step": 986
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.1541005528884596,
+      "learning_rate": 9.93300534947835e-06,
+      "loss": 0.7007,
+      "step": 987
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.915895095443689,
+      "learning_rate": 9.932789200374525e-06,
+      "loss": 1.5247,
+      "step": 988
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.442887963856919,
+      "learning_rate": 9.93257270550364e-06,
+      "loss": 0.6883,
+      "step": 989
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.345430763632468,
+      "learning_rate": 9.93235586488088e-06,
+      "loss": 1.083,
+      "step": 990
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.6329279194096324,
+      "learning_rate": 9.932138678521438e-06,
+      "loss": 1.4216,
+      "step": 991
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.8404128106622193,
+      "learning_rate": 9.93192114644054e-06,
+      "loss": 0.5213,
+      "step": 992
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.829742066244959,
+      "learning_rate": 9.931703268653431e-06,
+      "loss": 0.9371,
+      "step": 993
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.9845704625784375,
+      "learning_rate": 9.931485045175388e-06,
+      "loss": 1.1885,
+      "step": 994
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.3190081677989216,
+      "learning_rate": 9.931266476021704e-06,
+      "loss": 0.6295,
+      "step": 995
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.910239429035196,
+      "learning_rate": 9.931047561207702e-06,
+      "loss": 1.2699,
+      "step": 996
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.619849252844896,
+      "learning_rate": 9.930828300748726e-06,
+      "loss": 0.7838,
+      "step": 997
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.235519631028344,
+      "learning_rate": 9.930608694660144e-06,
+      "loss": 1.2416,
+      "step": 998
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.353341612052121,
+      "learning_rate": 9.930388742957351e-06,
+      "loss": 0.939,
+      "step": 999
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.028762046178202,
+      "learning_rate": 9.930168445655766e-06,
+      "loss": 0.4832,
+      "step": 1000
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.591371052059989,
+      "learning_rate": 9.929947802770827e-06,
+      "loss": 1.2762,
+      "step": 1001
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.697433120159279,
+      "learning_rate": 9.929726814318004e-06,
+      "loss": 0.2835,
+      "step": 1002
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.835924666363637,
+      "learning_rate": 9.929505480312785e-06,
+      "loss": 0.9394,
+      "step": 1003
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.3460503065278155,
+      "learning_rate": 9.929283800770684e-06,
+      "loss": 0.4421,
+      "step": 1004
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.235110616209546,
+      "learning_rate": 9.92906177570724e-06,
+      "loss": 0.7821,
+      "step": 1005
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.546859033042787,
+      "learning_rate": 9.92883940513802e-06,
+      "loss": 1.023,
+      "step": 1006
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.881188064635342,
+      "learning_rate": 9.928616689078605e-06,
+      "loss": 1.1696,
+      "step": 1007
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.414790877336852,
+      "learning_rate": 9.928393627544612e-06,
+      "loss": 0.9386,
+      "step": 1008
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.553507370975231,
+      "learning_rate": 9.928170220551671e-06,
+      "loss": 1.222,
+      "step": 1009
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.6732592042961465,
+      "learning_rate": 9.927946468115448e-06,
+      "loss": 0.7668,
+      "step": 1010
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.3890264197841145,
+      "learning_rate": 9.927722370251623e-06,
+      "loss": 0.8403,
+      "step": 1011
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.415266088346729,
+      "learning_rate": 9.927497926975906e-06,
+      "loss": 1.3521,
+      "step": 1012
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.0549480173712524,
+      "learning_rate": 9.927273138304028e-06,
+      "loss": 0.691,
+      "step": 1013
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.4185227553692865,
+      "learning_rate": 9.927048004251748e-06,
+      "loss": 0.7587,
+      "step": 1014
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5284307873609142,
+      "learning_rate": 9.926822524834845e-06,
+      "loss": 0.2633,
+      "step": 1015
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.855098239343265,
+      "learning_rate": 9.926596700069122e-06,
+      "loss": 0.8446,
+      "step": 1016
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.706309854028427,
+      "learning_rate": 9.926370529970414e-06,
+      "loss": 1.1894,
+      "step": 1017
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.751806675384341,
+      "learning_rate": 9.92614401455457e-06,
+      "loss": 0.7973,
+      "step": 1018
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.9276380261145714,
+      "learning_rate": 9.925917153837469e-06,
+      "loss": 0.792,
+      "step": 1019
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.120171889646719,
+      "learning_rate": 9.925689947835015e-06,
+      "loss": 0.412,
+      "step": 1020
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.509134466109772,
+      "learning_rate": 9.925462396563131e-06,
+      "loss": 0.7342,
+      "step": 1021
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.275264743662667,
+      "learning_rate": 9.925234500037768e-06,
+      "loss": 1.1055,
+      "step": 1022
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.706494826186237,
+      "learning_rate": 9.925006258274903e-06,
+      "loss": 0.6293,
+      "step": 1023
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.601024212740757,
+      "learning_rate": 9.924777671290532e-06,
+      "loss": 1.1678,
+      "step": 1024
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.604621793451188,
+      "learning_rate": 9.92454873910068e-06,
+      "loss": 0.443,
+      "step": 1025
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.552121826306486,
+      "learning_rate": 9.924319461721391e-06,
+      "loss": 0.7045,
+      "step": 1026
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.5575856771387238,
+      "learning_rate": 9.92408983916874e-06,
+      "loss": 0.4696,
+      "step": 1027
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.289654417823295,
+      "learning_rate": 9.92385987145882e-06,
+      "loss": 1.1612,
+      "step": 1028
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.014493264742926,
+      "learning_rate": 9.923629558607753e-06,
+      "loss": 1.0631,
+      "step": 1029
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.69656964091769,
+      "learning_rate": 9.923398900631681e-06,
+      "loss": 0.7831,
+      "step": 1030
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7779160202979185,
+      "learning_rate": 9.923167897546773e-06,
+      "loss": 0.4389,
+      "step": 1031
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.123388231016368,
+      "learning_rate": 9.92293654936922e-06,
+      "loss": 1.0849,
+      "step": 1032
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.652202727799504,
+      "learning_rate": 9.922704856115241e-06,
+      "loss": 0.7509,
+      "step": 1033
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9835718893039749,
+      "learning_rate": 9.922472817801075e-06,
+      "loss": 0.4422,
+      "step": 1034
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.67324732639257,
+      "learning_rate": 9.922240434442988e-06,
+      "loss": 1.484,
+      "step": 1035
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.1649733381598733,
+      "learning_rate": 9.922007706057266e-06,
+      "loss": 0.4445,
+      "step": 1036
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.977160755506409,
+      "learning_rate": 9.921774632660226e-06,
+      "loss": 0.6295,
+      "step": 1037
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.9528745949766493,
+      "learning_rate": 9.921541214268202e-06,
+      "loss": 0.5489,
+      "step": 1038
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.978548604532141,
+      "learning_rate": 9.92130745089756e-06,
+      "loss": 0.5643,
+      "step": 1039
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.542451724141814,
+      "learning_rate": 9.921073342564681e-06,
+      "loss": 0.781,
+      "step": 1040
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.3721141526352976,
+      "learning_rate": 9.920838889285979e-06,
+      "loss": 0.5458,
+      "step": 1041
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.7847159815128384,
+      "learning_rate": 9.920604091077886e-06,
+      "loss": 0.6578,
+      "step": 1042
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.9754391055259912,
+      "learning_rate": 9.92036894795686e-06,
+      "loss": 0.613,
+      "step": 1043
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.030356464555068,
+      "learning_rate": 9.920133459939385e-06,
+      "loss": 0.6361,
+      "step": 1044
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.880834558249092,
+      "learning_rate": 9.919897627041967e-06,
+      "loss": 0.8937,
+      "step": 1045
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2906676140310624,
+      "learning_rate": 9.919661449281136e-06,
+      "loss": 0.288,
+      "step": 1046
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.284278659461712,
+      "learning_rate": 9.919424926673449e-06,
+      "loss": 0.635,
+      "step": 1047
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.82577780947523,
+      "learning_rate": 9.919188059235483e-06,
+      "loss": 0.9913,
+      "step": 1048
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.758298021782454,
+      "learning_rate": 9.918950846983844e-06,
+      "loss": 0.4224,
+      "step": 1049
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.663385005526883,
+      "learning_rate": 9.918713289935156e-06,
+      "loss": 1.3562,
+      "step": 1050
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.746255654862295,
+      "learning_rate": 9.918475388106075e-06,
+      "loss": 0.7948,
+      "step": 1051
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.1445840998076156,
+      "learning_rate": 9.918237141513272e-06,
+      "loss": 0.687,
+      "step": 1052
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.729431954155433,
+      "learning_rate": 9.917998550173451e-06,
+      "loss": 0.6632,
+      "step": 1053
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.7846370807915215,
+      "learning_rate": 9.917759614103335e-06,
+      "loss": 0.6811,
+      "step": 1054
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.817056381788165,
+      "learning_rate": 9.917520333319671e-06,
+      "loss": 0.9256,
+      "step": 1055
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.366125843476706,
+      "learning_rate": 9.917280707839235e-06,
+      "loss": 0.4559,
+      "step": 1056
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.2462937271026755,
+      "learning_rate": 9.91704073767882e-06,
+      "loss": 0.5,
+      "step": 1057
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.598722852680578,
+      "learning_rate": 9.916800422855247e-06,
+      "loss": 0.8063,
+      "step": 1058
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 6.463919076864924,
+      "learning_rate": 9.916559763385364e-06,
+      "loss": 1.3674,
+      "step": 1059
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.0754104305945087,
+      "learning_rate": 9.916318759286039e-06,
+      "loss": 0.5969,
+      "step": 1060
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.634585435044843,
+      "learning_rate": 9.916077410574163e-06,
+      "loss": 1.1031,
+      "step": 1061
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.7528878203710736,
+      "learning_rate": 9.915835717266658e-06,
+      "loss": 0.7968,
+      "step": 1062
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.66701414858373,
+      "learning_rate": 9.91559367938046e-06,
+      "loss": 0.8395,
+      "step": 1063
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 6.168129947730295,
+      "learning_rate": 9.91535129693254e-06,
+      "loss": 1.4167,
+      "step": 1064
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.6400228917682234,
+      "learning_rate": 9.915108569939884e-06,
+      "loss": 0.8876,
+      "step": 1065
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.3615172096238357,
+      "learning_rate": 9.91486549841951e-06,
+      "loss": 0.3196,
+      "step": 1066
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 6.1590255485854595,
+      "learning_rate": 9.914622082388452e-06,
+      "loss": 1.7367,
+      "step": 1067
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5833506445418588,
+      "learning_rate": 9.914378321863776e-06,
+      "loss": 0.2484,
+      "step": 1068
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.9244870590910623,
+      "learning_rate": 9.914134216862568e-06,
+      "loss": 0.9296,
+      "step": 1069
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.118313589125725,
+      "learning_rate": 9.913889767401935e-06,
+      "loss": 1.171,
+      "step": 1070
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.112181123473643,
+      "learning_rate": 9.913644973499017e-06,
+      "loss": 0.726,
+      "step": 1071
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.9211444194643703,
+      "learning_rate": 9.913399835170969e-06,
+      "loss": 0.4475,
+      "step": 1072
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.2639064043344446,
+      "learning_rate": 9.913154352434977e-06,
+      "loss": 0.7201,
+      "step": 1073
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.9042396813762323,
+      "learning_rate": 9.912908525308246e-06,
+      "loss": 0.6094,
+      "step": 1074
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.129947465502224,
+      "learning_rate": 9.912662353808009e-06,
+      "loss": 0.783,
+      "step": 1075
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.217328334333304,
+      "learning_rate": 9.912415837951522e-06,
+      "loss": 0.9254,
+      "step": 1076
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.371558982797147,
+      "learning_rate": 9.912168977756061e-06,
+      "loss": 0.9842,
+      "step": 1077
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.104153333889138,
+      "learning_rate": 9.911921773238935e-06,
+      "loss": 0.9036,
+      "step": 1078
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.151522115800109,
+      "learning_rate": 9.911674224417469e-06,
+      "loss": 1.0193,
+      "step": 1079
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.746469156002981,
+      "learning_rate": 9.911426331309013e-06,
+      "loss": 1.1992,
+      "step": 1080
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.258407588455397,
+      "learning_rate": 9.911178093930949e-06,
+      "loss": 1.3186,
+      "step": 1081
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.235760757939858,
+      "learning_rate": 9.910929512300673e-06,
+      "loss": 1.0851,
+      "step": 1082
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.453081187996927,
+      "learning_rate": 9.91068058643561e-06,
+      "loss": 0.737,
+      "step": 1083
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.71536401568929,
+      "learning_rate": 9.91043131635321e-06,
+      "loss": 1.3866,
+      "step": 1084
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.547313676110811,
+      "learning_rate": 9.910181702070944e-06,
+      "loss": 1.1275,
+      "step": 1085
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.2128405071826625,
+      "learning_rate": 9.90993174360631e-06,
+      "loss": 0.5183,
+      "step": 1086
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.711441991976482,
+      "learning_rate": 9.90968144097683e-06,
+      "loss": 0.9171,
+      "step": 1087
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.8161551402219795,
+      "learning_rate": 9.909430794200047e-06,
+      "loss": 0.6247,
+      "step": 1088
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.613553679871445,
+      "learning_rate": 9.909179803293532e-06,
+      "loss": 0.9455,
+      "step": 1089
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.1355714333717275,
+      "learning_rate": 9.90892846827488e-06,
+      "loss": 1.2161,
+      "step": 1090
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.0312869401343745,
+      "learning_rate": 9.908676789161701e-06,
+      "loss": 1.0728,
+      "step": 1091
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.8283986450010477,
+      "learning_rate": 9.908424765971644e-06,
+      "loss": 0.5282,
+      "step": 1092
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.571969875090206,
+      "learning_rate": 9.908172398722374e-06,
+      "loss": 0.518,
+      "step": 1093
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.920075402725496,
+      "learning_rate": 9.907919687431578e-06,
+      "loss": 0.8735,
+      "step": 1094
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.4311968729395437,
+      "learning_rate": 9.90766663211697e-06,
+      "loss": 0.8832,
+      "step": 1095
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.73982103420756,
+      "learning_rate": 9.90741323279629e-06,
+      "loss": 1.2395,
+      "step": 1096
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.256664032664819,
+      "learning_rate": 9.907159489487301e-06,
+      "loss": 0.913,
+      "step": 1097
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.182530238530806,
+      "learning_rate": 9.906905402207786e-06,
+      "loss": 1.1338,
+      "step": 1098
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.10551835294458,
+      "learning_rate": 9.90665097097556e-06,
+      "loss": 0.9437,
+      "step": 1099
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.9631556450062604,
+      "learning_rate": 9.906396195808452e-06,
+      "loss": 0.6628,
+      "step": 1100
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.946634427167888,
+      "learning_rate": 9.906141076724324e-06,
+      "loss": 1.1265,
+      "step": 1101
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.369504579349011,
+      "learning_rate": 9.905885613741058e-06,
+      "loss": 0.6546,
+      "step": 1102
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.9024786940759517,
+      "learning_rate": 9.905629806876562e-06,
+      "loss": 0.6797,
+      "step": 1103
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.6359571441599225,
+      "learning_rate": 9.905373656148765e-06,
+      "loss": 0.3359,
+      "step": 1104
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.629817569105054,
+      "learning_rate": 9.905117161575622e-06,
+      "loss": 1.1071,
+      "step": 1105
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.820538418080594,
+      "learning_rate": 9.904860323175114e-06,
+      "loss": 0.6526,
+      "step": 1106
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.639546998032372,
+      "learning_rate": 9.904603140965244e-06,
+      "loss": 0.3731,
+      "step": 1107
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.60234746760394,
+      "learning_rate": 9.904345614964039e-06,
+      "loss": 1.2051,
+      "step": 1108
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.467763997576738,
+      "learning_rate": 9.90408774518955e-06,
+      "loss": 0.498,
+      "step": 1109
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.706513283245458,
+      "learning_rate": 9.903829531659853e-06,
+      "loss": 0.6948,
+      "step": 1110
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.228780493202378,
+      "learning_rate": 9.903570974393044e-06,
+      "loss": 0.6231,
+      "step": 1111
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.23269091182088,
+      "learning_rate": 9.903312073407255e-06,
+      "loss": 0.808,
+      "step": 1112
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.5909195757076335,
+      "learning_rate": 9.903052828720626e-06,
+      "loss": 0.9084,
+      "step": 1113
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.665407647763041,
+      "learning_rate": 9.902793240351332e-06,
+      "loss": 1.209,
+      "step": 1114
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.181500368016984,
+      "learning_rate": 9.90253330831757e-06,
+      "loss": 0.9321,
+      "step": 1115
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.007003219308571,
+      "learning_rate": 9.902273032637558e-06,
+      "loss": 0.6731,
+      "step": 1116
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.249991934849754,
+      "learning_rate": 9.902012413329541e-06,
+      "loss": 0.9785,
+      "step": 1117
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.636337888119685,
+      "learning_rate": 9.901751450411787e-06,
+      "loss": 0.5601,
+      "step": 1118
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.845419166719933,
+      "learning_rate": 9.90149014390259e-06,
+      "loss": 0.9011,
+      "step": 1119
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.201928266545957,
+      "learning_rate": 9.901228493820265e-06,
+      "loss": 1.3249,
+      "step": 1120
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.578795797853542,
+      "learning_rate": 9.900966500183153e-06,
+      "loss": 1.1478,
+      "step": 1121
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.6095756033428,
+      "learning_rate": 9.90070416300962e-06,
+      "loss": 0.6756,
+      "step": 1122
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.911041891172558,
+      "learning_rate": 9.900441482318051e-06,
+      "loss": 1.011,
+      "step": 1123
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.284294682149408,
+      "learning_rate": 9.900178458126862e-06,
+      "loss": 0.4008,
+      "step": 1124
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.7326206756730653,
+      "learning_rate": 9.899915090454487e-06,
+      "loss": 0.6352,
+      "step": 1125
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.428099213081668,
+      "learning_rate": 9.89965137931939e-06,
+      "loss": 0.8726,
+      "step": 1126
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.3340338283313966,
+      "learning_rate": 9.899387324740053e-06,
+      "loss": 0.813,
+      "step": 1127
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.746073944845481,
+      "learning_rate": 9.899122926734988e-06,
+      "loss": 0.6223,
+      "step": 1128
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.590965204756769,
+      "learning_rate": 9.898858185322728e-06,
+      "loss": 0.8611,
+      "step": 1129
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.894240050324802,
+      "learning_rate": 9.898593100521828e-06,
+      "loss": 0.8581,
+      "step": 1130
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.7572285946477177,
+      "learning_rate": 9.898327672350871e-06,
+      "loss": 0.6341,
+      "step": 1131
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.4296680237638593,
+      "learning_rate": 9.898061900828461e-06,
+      "loss": 0.7285,
+      "step": 1132
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.7157870658361665,
+      "learning_rate": 9.897795785973227e-06,
+      "loss": 0.9815,
+      "step": 1133
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.270557295796642,
+      "learning_rate": 9.897529327803825e-06,
+      "loss": 0.7319,
+      "step": 1134
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.906704857168554,
+      "learning_rate": 9.897262526338933e-06,
+      "loss": 0.436,
+      "step": 1135
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.3758294790859398,
+      "learning_rate": 9.896995381597248e-06,
+      "loss": 0.4856,
+      "step": 1136
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.349959256999474,
+      "learning_rate": 9.8967278935975e-06,
+      "loss": 0.6012,
+      "step": 1137
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.3521814313627614,
+      "learning_rate": 9.896460062358437e-06,
+      "loss": 0.3664,
+      "step": 1138
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.938674501026468,
+      "learning_rate": 9.896191887898833e-06,
+      "loss": 0.7762,
+      "step": 1139
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.476274030084168,
+      "learning_rate": 9.895923370237487e-06,
+      "loss": 0.8053,
+      "step": 1140
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.477353094729364,
+      "learning_rate": 9.895654509393219e-06,
+      "loss": 1.5004,
+      "step": 1141
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.7789096830346423,
+      "learning_rate": 9.895385305384875e-06,
+      "loss": 0.9327,
+      "step": 1142
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.052117060974299,
+      "learning_rate": 9.895115758231327e-06,
+      "loss": 0.7655,
+      "step": 1143
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.311851068484635,
+      "learning_rate": 9.894845867951468e-06,
+      "loss": 0.7396,
+      "step": 1144
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.7923414529988455,
+      "learning_rate": 9.894575634564216e-06,
+      "loss": 0.7855,
+      "step": 1145
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.945249613145775,
+      "learning_rate": 9.894305058088513e-06,
+      "loss": 0.9153,
+      "step": 1146
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.7840576172347395,
+      "learning_rate": 9.894034138543325e-06,
+      "loss": 1.0292,
+      "step": 1147
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.93236236975966,
+      "learning_rate": 9.893762875947643e-06,
+      "loss": 0.4059,
+      "step": 1148
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.58890168616701,
+      "learning_rate": 9.893491270320482e-06,
+      "loss": 1.0852,
+      "step": 1149
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.559899603280683,
+      "learning_rate": 9.89321932168088e-06,
+      "loss": 1.1237,
+      "step": 1150
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.60214704397996,
+      "learning_rate": 9.892947030047897e-06,
+      "loss": 0.9579,
+      "step": 1151
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.435288516889513,
+      "learning_rate": 9.892674395440623e-06,
+      "loss": 0.7672,
+      "step": 1152
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.966391987672413,
+      "learning_rate": 9.892401417878166e-06,
+      "loss": 0.494,
+      "step": 1153
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.5165349113611164,
+      "learning_rate": 9.89212809737966e-06,
+      "loss": 0.6433,
+      "step": 1154
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.964943785418872,
+      "learning_rate": 9.891854433964268e-06,
+      "loss": 1.0597,
+      "step": 1155
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.7599503203593763,
+      "learning_rate": 9.891580427651169e-06,
+      "loss": 0.6986,
+      "step": 1156
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.787335060739045,
+      "learning_rate": 9.891306078459569e-06,
+      "loss": 1.4073,
+      "step": 1157
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.2838187991127645,
+      "learning_rate": 9.8910313864087e-06,
+      "loss": 0.7886,
+      "step": 1158
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.4466029728678587,
+      "learning_rate": 9.890756351517816e-06,
+      "loss": 0.6352,
+      "step": 1159
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.956320751045181,
+      "learning_rate": 9.890480973806198e-06,
+      "loss": 1.3651,
+      "step": 1160
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.405188553619671,
+      "learning_rate": 9.890205253293145e-06,
+      "loss": 0.6927,
+      "step": 1161
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.331448218013707,
+      "learning_rate": 9.889929189997988e-06,
+      "loss": 1.1553,
+      "step": 1162
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.717470241828111,
+      "learning_rate": 9.889652783940075e-06,
+      "loss": 1.2707,
+      "step": 1163
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.3074133176826,
+      "learning_rate": 9.889376035138782e-06,
+      "loss": 0.8255,
+      "step": 1164
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.8692669892049047,
+      "learning_rate": 9.889098943613508e-06,
+      "loss": 0.669,
+      "step": 1165
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.0717586645649155,
+      "learning_rate": 9.888821509383676e-06,
+      "loss": 0.6628,
+      "step": 1166
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.781072878569787,
+      "learning_rate": 9.888543732468732e-06,
+      "loss": 1.0705,
+      "step": 1167
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.952219904396231,
+      "learning_rate": 9.888265612888145e-06,
+      "loss": 0.5253,
+      "step": 1168
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.000881351619064,
+      "learning_rate": 9.887987150661415e-06,
+      "loss": 1.2732,
+      "step": 1169
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.308052200362575,
+      "learning_rate": 9.887708345808059e-06,
+      "loss": 0.8583,
+      "step": 1170
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.8577523911579465,
+      "learning_rate": 9.887429198347617e-06,
+      "loss": 1.3483,
+      "step": 1171
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.898494233387562,
+      "learning_rate": 9.88714970829966e-06,
+      "loss": 0.9511,
+      "step": 1172
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.794041871561278,
+      "learning_rate": 9.886869875683776e-06,
+      "loss": 0.6661,
+      "step": 1173
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.9150265779956146,
+      "learning_rate": 9.886589700519583e-06,
+      "loss": 0.3034,
+      "step": 1174
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.368498537449657,
+      "learning_rate": 9.886309182826717e-06,
+      "loss": 0.5748,
+      "step": 1175
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.713244358718991,
+      "learning_rate": 9.886028322624843e-06,
+      "loss": 1.1596,
+      "step": 1176
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.28797089150776,
+      "learning_rate": 9.885747119933648e-06,
+      "loss": 1.0055,
+      "step": 1177
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.465500959031329,
+      "learning_rate": 9.885465574772842e-06,
+      "loss": 1.1441,
+      "step": 1178
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.163076994144076,
+      "learning_rate": 9.885183687162162e-06,
+      "loss": 0.4571,
+      "step": 1179
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.7490054918364697,
+      "learning_rate": 9.884901457121366e-06,
+      "loss": 0.6162,
+      "step": 1180
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.299431696370699,
+      "learning_rate": 9.884618884670236e-06,
+      "loss": 0.9722,
+      "step": 1181
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.29584811568989,
+      "learning_rate": 9.88433596982858e-06,
+      "loss": 0.6991,
+      "step": 1182
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.759772526614289,
+      "learning_rate": 9.88405271261623e-06,
+      "loss": 1.0675,
+      "step": 1183
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3720315128058174,
+      "learning_rate": 9.883769113053039e-06,
+      "loss": 0.149,
+      "step": 1184
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.66518376225917,
+      "learning_rate": 9.883485171158889e-06,
+      "loss": 0.8875,
+      "step": 1185
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.227971159935246,
+      "learning_rate": 9.883200886953682e-06,
+      "loss": 1.3924,
+      "step": 1186
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 6.02754405998497,
+      "learning_rate": 9.882916260457343e-06,
+      "loss": 1.6309,
+      "step": 1187
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.07747958412001,
+      "learning_rate": 9.882631291689823e-06,
+      "loss": 0.8492,
+      "step": 1188
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.2134197448319526,
+      "learning_rate": 9.882345980671102e-06,
+      "loss": 0.6647,
+      "step": 1189
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.346689217592907,
+      "learning_rate": 9.882060327421174e-06,
+      "loss": 1.1335,
+      "step": 1190
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.263765075179125,
+      "learning_rate": 9.881774331960065e-06,
+      "loss": 0.3391,
+      "step": 1191
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.641436335366062,
+      "learning_rate": 9.88148799430782e-06,
+      "loss": 0.4934,
+      "step": 1192
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 6.032362831368957,
+      "learning_rate": 9.881201314484513e-06,
+      "loss": 0.8404,
+      "step": 1193
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.9748639666149304,
+      "learning_rate": 9.880914292510233e-06,
+      "loss": 0.7072,
+      "step": 1194
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.8610638245991957,
+      "learning_rate": 9.880626928405106e-06,
+      "loss": 1.1508,
+      "step": 1195
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.3376495404426785,
+      "learning_rate": 9.88033922218927e-06,
+      "loss": 0.4453,
+      "step": 1196
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.859625847530836,
+      "learning_rate": 9.880051173882896e-06,
+      "loss": 1.4039,
+      "step": 1197
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.202714929903976,
+      "learning_rate": 9.879762783506172e-06,
+      "loss": 0.9029,
+      "step": 1198
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.083474294001203,
+      "learning_rate": 9.879474051079312e-06,
+      "loss": 0.9305,
+      "step": 1199
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.199315543955551,
+      "learning_rate": 9.879184976622557e-06,
+      "loss": 1.0684,
+      "step": 1200
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.131337594808835,
+      "learning_rate": 9.878895560156172e-06,
+      "loss": 0.9519,
+      "step": 1201
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.1636870069489524,
+      "learning_rate": 9.87860580170044e-06,
+      "loss": 0.4102,
+      "step": 1202
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.486130674054726,
+      "learning_rate": 9.878315701275671e-06,
+      "loss": 1.1242,
+      "step": 1203
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.040979726468128,
+      "learning_rate": 9.878025258902204e-06,
+      "loss": 0.8364,
+      "step": 1204
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.454378469994424,
+      "learning_rate": 9.877734474600395e-06,
+      "loss": 1.3415,
+      "step": 1205
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.1153828727308985,
+      "learning_rate": 9.877443348390627e-06,
+      "loss": 1.0923,
+      "step": 1206
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.157714411270862,
+      "learning_rate": 9.877151880293307e-06,
+      "loss": 1.1689,
+      "step": 1207
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.205116285973389,
+      "learning_rate": 9.876860070328867e-06,
+      "loss": 0.7162,
+      "step": 1208
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.125812556804762,
+      "learning_rate": 9.876567918517759e-06,
+      "loss": 0.6812,
+      "step": 1209
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.136997839119193,
+      "learning_rate": 9.876275424880463e-06,
+      "loss": 0.3998,
+      "step": 1210
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.9726971596754406,
+      "learning_rate": 9.875982589437481e-06,
+      "loss": 0.9348,
+      "step": 1211
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.295857980061875,
+      "learning_rate": 9.87568941220934e-06,
+      "loss": 0.5959,
+      "step": 1212
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 6.049165049333135,
+      "learning_rate": 9.87539589321659e-06,
+      "loss": 1.2373,
+      "step": 1213
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.855626558039712,
+      "learning_rate": 9.875102032479807e-06,
+      "loss": 0.7339,
+      "step": 1214
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.826138284182142,
+      "learning_rate": 9.874807830019586e-06,
+      "loss": 0.6591,
+      "step": 1215
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.83684207885743,
+      "learning_rate": 9.874513285856553e-06,
+      "loss": 0.7731,
+      "step": 1216
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.933104564450693,
+      "learning_rate": 9.874218400011352e-06,
+      "loss": 0.8364,
+      "step": 1217
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4001799848197887,
+      "learning_rate": 9.873923172504653e-06,
+      "loss": 0.18,
+      "step": 1218
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.252528244067028,
+      "learning_rate": 9.873627603357152e-06,
+      "loss": 0.8766,
+      "step": 1219
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.988190929247905,
+      "learning_rate": 9.873331692589566e-06,
+      "loss": 0.8168,
+      "step": 1220
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.0318100849354,
+      "learning_rate": 9.873035440222638e-06,
+      "loss": 0.8017,
+      "step": 1221
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.756505224009058,
+      "learning_rate": 9.872738846277133e-06,
+      "loss": 0.6414,
+      "step": 1222
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.069998346474113,
+      "learning_rate": 9.87244191077384e-06,
+      "loss": 0.7499,
+      "step": 1223
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.5809666241236684,
+      "learning_rate": 9.872144633733573e-06,
+      "loss": 0.739,
+      "step": 1224
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.0947168680195705,
+      "learning_rate": 9.871847015177173e-06,
+      "loss": 0.5898,
+      "step": 1225
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.1358159711006333,
+      "learning_rate": 9.8715490551255e-06,
+      "loss": 0.4362,
+      "step": 1226
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.202307254243295,
+      "learning_rate": 9.871250753599438e-06,
+      "loss": 0.8291,
+      "step": 1227
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.560667754171791,
+      "learning_rate": 9.870952110619899e-06,
+      "loss": 1.2004,
+      "step": 1228
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.839448570202385,
+      "learning_rate": 9.870653126207813e-06,
+      "loss": 1.1906,
+      "step": 1229
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.587820228326079,
+      "learning_rate": 9.870353800384142e-06,
+      "loss": 1.1577,
+      "step": 1230
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.369399402473251,
+      "learning_rate": 9.870054133169864e-06,
+      "loss": 0.3833,
+      "step": 1231
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.332705145152846,
+      "learning_rate": 9.869754124585988e-06,
+      "loss": 0.9675,
+      "step": 1232
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.892316115188248,
+      "learning_rate": 9.869453774653539e-06,
+      "loss": 1.351,
+      "step": 1233
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.125532292329167,
+      "learning_rate": 9.869153083393573e-06,
+      "loss": 0.9729,
+      "step": 1234
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.038649194505703,
+      "learning_rate": 9.868852050827167e-06,
+      "loss": 0.6525,
+      "step": 1235
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.5708776653671337,
+      "learning_rate": 9.868550676975422e-06,
+      "loss": 0.7557,
+      "step": 1236
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.9247024667032946,
+      "learning_rate": 9.86824896185946e-06,
+      "loss": 0.6781,
+      "step": 1237
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.0777651806845143,
+      "learning_rate": 9.867946905500437e-06,
+      "loss": 0.3866,
+      "step": 1238
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.8429779511171898,
+      "learning_rate": 9.867644507919518e-06,
+      "loss": 0.7838,
+      "step": 1239
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.9761613135796563,
+      "learning_rate": 9.867341769137902e-06,
+      "loss": 0.7101,
+      "step": 1240
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 6.379783725511916,
+      "learning_rate": 9.867038689176814e-06,
+      "loss": 1.2627,
+      "step": 1241
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.9441494398725743,
+      "learning_rate": 9.866735268057492e-06,
+      "loss": 0.6645,
+      "step": 1242
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.3048297328129035,
+      "learning_rate": 9.86643150580121e-06,
+      "loss": 0.7912,
+      "step": 1243
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.212674300762759,
+      "learning_rate": 9.866127402429257e-06,
+      "loss": 0.5092,
+      "step": 1244
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.867708233325816,
+      "learning_rate": 9.865822957962952e-06,
+      "loss": 1.2898,
+      "step": 1245
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.9895126564047425,
+      "learning_rate": 9.865518172423634e-06,
+      "loss": 0.8157,
+      "step": 1246
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.1574728111342276,
+      "learning_rate": 9.865213045832664e-06,
+      "loss": 0.613,
+      "step": 1247
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.308032016001108,
+      "learning_rate": 9.864907578211436e-06,
+      "loss": 0.9943,
+      "step": 1248
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.978772964634741,
+      "learning_rate": 9.864601769581357e-06,
+      "loss": 0.5542,
+      "step": 1249
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.5859393692610575,
+      "learning_rate": 9.864295619963866e-06,
+      "loss": 0.708,
+      "step": 1250
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.2253481480588952,
+      "learning_rate": 9.863989129380421e-06,
+      "loss": 0.5619,
+      "step": 1251
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.248573746610271,
+      "learning_rate": 9.863682297852506e-06,
+      "loss": 1.1753,
+      "step": 1252
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.7692726514908133,
+      "learning_rate": 9.86337512540163e-06,
+      "loss": 0.785,
+      "step": 1253
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.140073443579133,
+      "learning_rate": 9.863067612049321e-06,
+      "loss": 0.7547,
+      "step": 1254
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.219012600204615,
+      "learning_rate": 9.862759757817138e-06,
+      "loss": 0.2145,
+      "step": 1255
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.692923749310751,
+      "learning_rate": 9.862451562726659e-06,
+      "loss": 1.0327,
+      "step": 1256
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.142416334886868,
+      "learning_rate": 9.862143026799486e-06,
+      "loss": 1.1972,
+      "step": 1257
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.215708570787776,
+      "learning_rate": 9.861834150057247e-06,
+      "loss": 0.462,
+      "step": 1258
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.209753111231955,
+      "learning_rate": 9.861524932521595e-06,
+      "loss": 0.9367,
+      "step": 1259
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.4564385364514285,
+      "learning_rate": 9.8612153742142e-06,
+      "loss": 0.701,
+      "step": 1260
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.7263112565077634,
+      "learning_rate": 9.860905475156765e-06,
+      "loss": 0.8455,
+      "step": 1261
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.164659038297447,
+      "learning_rate": 9.86059523537101e-06,
+      "loss": 0.7103,
+      "step": 1262
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.4221185950858892,
+      "learning_rate": 9.860284654878682e-06,
+      "loss": 0.5031,
+      "step": 1263
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.8565077383119695,
+      "learning_rate": 9.859973733701553e-06,
+      "loss": 0.7379,
+      "step": 1264
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.610438004102319,
+      "learning_rate": 9.859662471861415e-06,
+      "loss": 1.1086,
+      "step": 1265
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.0726742543457335,
+      "learning_rate": 9.859350869380086e-06,
+      "loss": 0.6462,
+      "step": 1266
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.0570579775243845,
+      "learning_rate": 9.859038926279412e-06,
+      "loss": 1.3292,
+      "step": 1267
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.549728041559905,
+      "learning_rate": 9.858726642581253e-06,
+      "loss": 0.7154,
+      "step": 1268
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.013422156375804,
+      "learning_rate": 9.858414018307503e-06,
+      "loss": 1.2002,
+      "step": 1269
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.2470568492629837,
+      "learning_rate": 9.858101053480074e-06,
+      "loss": 0.5817,
+      "step": 1270
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.8005829520202603,
+      "learning_rate": 9.857787748120904e-06,
+      "loss": 0.9949,
+      "step": 1271
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.094339621776428,
+      "learning_rate": 9.857474102251955e-06,
+      "loss": 0.7578,
+      "step": 1272
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.525640812991137,
+      "learning_rate": 9.857160115895208e-06,
+      "loss": 0.7344,
+      "step": 1273
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.002900024886897,
+      "learning_rate": 9.856845789072678e-06,
+      "loss": 0.9144,
+      "step": 1274
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.72727273650834,
+      "learning_rate": 9.856531121806395e-06,
+      "loss": 1.2397,
+      "step": 1275
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.5075716444438942,
+      "learning_rate": 9.856216114118416e-06,
+      "loss": 0.5811,
+      "step": 1276
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.847247096507675,
+      "learning_rate": 9.855900766030819e-06,
+      "loss": 0.9566,
+      "step": 1277
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.5594697153204096,
+      "learning_rate": 9.855585077565714e-06,
+      "loss": 0.4128,
+      "step": 1278
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.821428414387075,
+      "learning_rate": 9.855269048745227e-06,
+      "loss": 0.9853,
+      "step": 1279
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.722415247680657,
+      "learning_rate": 9.854952679591508e-06,
+      "loss": 0.4608,
+      "step": 1280
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.2581275069199576,
+      "learning_rate": 9.854635970126738e-06,
+      "loss": 0.4616,
+      "step": 1281
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.600979724613779,
+      "learning_rate": 9.854318920373111e-06,
+      "loss": 0.869,
+      "step": 1282
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.5221907799442045,
+      "learning_rate": 9.854001530352855e-06,
+      "loss": 0.6486,
+      "step": 1283
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.6404075289214157,
+      "learning_rate": 9.853683800088217e-06,
+      "loss": 0.7015,
+      "step": 1284
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.650201614991193,
+      "learning_rate": 9.853365729601465e-06,
+      "loss": 0.7917,
+      "step": 1285
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.977376505203249,
+      "learning_rate": 9.853047318914898e-06,
+      "loss": 0.9086,
+      "step": 1286
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.186796749935245,
+      "learning_rate": 9.852728568050838e-06,
+      "loss": 0.7049,
+      "step": 1287
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.395589551710405,
+      "learning_rate": 9.852409477031621e-06,
+      "loss": 1.4272,
+      "step": 1288
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.224627049772095,
+      "learning_rate": 9.852090045879619e-06,
+      "loss": 0.9042,
+      "step": 1289
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.4495366289399074,
+      "learning_rate": 9.85177027461722e-06,
+      "loss": 0.6737,
+      "step": 1290
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.118977930027873,
+      "learning_rate": 9.851450163266843e-06,
+      "loss": 1.3165,
+      "step": 1291
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.946791989038946,
+      "learning_rate": 9.85112971185092e-06,
+      "loss": 0.5781,
+      "step": 1292
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.9952593447159943,
+      "learning_rate": 9.850808920391917e-06,
+      "loss": 0.9511,
+      "step": 1293
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.293125196572581,
+      "learning_rate": 9.850487788912319e-06,
+      "loss": 0.8557,
+      "step": 1294
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.410250168793054,
+      "learning_rate": 9.850166317434638e-06,
+      "loss": 0.6575,
+      "step": 1295
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.481123343866197,
+      "learning_rate": 9.849844505981405e-06,
+      "loss": 0.5825,
+      "step": 1296
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.3876454567069585,
+      "learning_rate": 9.849522354575178e-06,
+      "loss": 1.014,
+      "step": 1297
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.192263677148237,
+      "learning_rate": 9.84919986323854e-06,
+      "loss": 0.3586,
+      "step": 1298
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.3341819516293008,
+      "learning_rate": 9.848877031994095e-06,
+      "loss": 0.6102,
+      "step": 1299
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.2551287600956305,
+      "learning_rate": 9.848553860864474e-06,
+      "loss": 0.5703,
+      "step": 1300
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.7703916932902035,
+      "learning_rate": 9.848230349872326e-06,
+      "loss": 1.0168,
+      "step": 1301
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.053425404359997,
+      "learning_rate": 9.847906499040332e-06,
+      "loss": 0.9807,
+      "step": 1302
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.072379777925076,
+      "learning_rate": 9.847582308391189e-06,
+      "loss": 1.0596,
+      "step": 1303
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.207348153568482,
+      "learning_rate": 9.847257777947624e-06,
+      "loss": 0.5372,
+      "step": 1304
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.7962584417662635,
+      "learning_rate": 9.846932907732383e-06,
+      "loss": 1.4428,
+      "step": 1305
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.9570045662665962,
+      "learning_rate": 9.84660769776824e-06,
+      "loss": 0.9904,
+      "step": 1306
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.37422552493355,
+      "learning_rate": 9.84628214807799e-06,
+      "loss": 1.3554,
+      "step": 1307
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.344126050620689,
+      "learning_rate": 9.845956258684453e-06,
+      "loss": 1.1511,
+      "step": 1308
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.878686990867304,
+      "learning_rate": 9.84563002961047e-06,
+      "loss": 0.9766,
+      "step": 1309
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.2087546006341934,
+      "learning_rate": 9.845303460878913e-06,
+      "loss": 0.5185,
+      "step": 1310
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 6.443723713364844,
+      "learning_rate": 9.844976552512669e-06,
+      "loss": 2.064,
+      "step": 1311
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9195624220575043,
+      "learning_rate": 9.844649304534653e-06,
+      "loss": 0.2766,
+      "step": 1312
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.073045313682807,
+      "learning_rate": 9.844321716967805e-06,
+      "loss": 0.6306,
+      "step": 1313
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.6814448226780194,
+      "learning_rate": 9.843993789835088e-06,
+      "loss": 0.8284,
+      "step": 1314
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.174265067657053,
+      "learning_rate": 9.843665523159488e-06,
+      "loss": 0.9003,
+      "step": 1315
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.8093957573304404,
+      "learning_rate": 9.843336916964012e-06,
+      "loss": 1.1277,
+      "step": 1316
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.1501702688347955,
+      "learning_rate": 9.8430079712717e-06,
+      "loss": 0.9939,
+      "step": 1317
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.180878812579681,
+      "learning_rate": 9.842678686105603e-06,
+      "loss": 0.6883,
+      "step": 1318
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.9089167772488325,
+      "learning_rate": 9.842349061488805e-06,
+      "loss": 1.0128,
+      "step": 1319
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.7583039485954535,
+      "learning_rate": 9.842019097444414e-06,
+      "loss": 1.3914,
+      "step": 1320
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.575696603618623,
+      "learning_rate": 9.841688793995556e-06,
+      "loss": 1.1302,
+      "step": 1321
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.32329831084482,
+      "learning_rate": 9.841358151165385e-06,
+      "loss": 0.4465,
+      "step": 1322
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.344382550224537,
+      "learning_rate": 9.841027168977078e-06,
+      "loss": 1.4557,
+      "step": 1323
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.7431485492243355,
+      "learning_rate": 9.840695847453833e-06,
+      "loss": 0.901,
+      "step": 1324
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.841450732154974,
+      "learning_rate": 9.840364186618876e-06,
+      "loss": 0.8751,
+      "step": 1325
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.86898813254035,
+      "learning_rate": 9.840032186495457e-06,
+      "loss": 0.1148,
+      "step": 1326
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.354391039173452,
+      "learning_rate": 9.839699847106843e-06,
+      "loss": 0.9702,
+      "step": 1327
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.6741596542471933,
+      "learning_rate": 9.839367168476333e-06,
+      "loss": 0.7526,
+      "step": 1328
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.16879096917253,
+      "learning_rate": 9.839034150627245e-06,
+      "loss": 0.7054,
+      "step": 1329
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.765591481592283,
+      "learning_rate": 9.838700793582925e-06,
+      "loss": 0.8889,
+      "step": 1330
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.38145209692892,
+      "learning_rate": 9.838367097366734e-06,
+      "loss": 0.6064,
+      "step": 1331
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.854729782273505,
+      "learning_rate": 9.83803306200207e-06,
+      "loss": 1.133,
+      "step": 1332
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.116468267131302,
+      "learning_rate": 9.83769868751234e-06,
+      "loss": 1.1587,
+      "step": 1333
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 6.275829287323761,
+      "learning_rate": 9.837363973920989e-06,
+      "loss": 1.2242,
+      "step": 1334
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.645549953945881,
+      "learning_rate": 9.837028921251472e-06,
+      "loss": 0.4298,
+      "step": 1335
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.623616570720286,
+      "learning_rate": 9.836693529527281e-06,
+      "loss": 0.7962,
+      "step": 1336
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.405795229808539,
+      "learning_rate": 9.836357798771922e-06,
+      "loss": 1.0292,
+      "step": 1337
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.8284035944571864,
+      "learning_rate": 9.83602172900893e-06,
+      "loss": 0.7384,
+      "step": 1338
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.405684722902709,
+      "learning_rate": 9.83568532026186e-06,
+      "loss": 1.0122,
+      "step": 1339
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.241662354295038,
+      "learning_rate": 9.835348572554296e-06,
+      "loss": 0.8407,
+      "step": 1340
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.2711045608163785,
+      "learning_rate": 9.835011485909837e-06,
+      "loss": 0.9705,
+      "step": 1341
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.518128074176903,
+      "learning_rate": 9.834674060352119e-06,
+      "loss": 1.445,
+      "step": 1342
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.9578130578482336,
+      "learning_rate": 9.834336295904787e-06,
+      "loss": 0.5462,
+      "step": 1343
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.230932342502631,
+      "learning_rate": 9.83399819259152e-06,
+      "loss": 0.7917,
+      "step": 1344
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.904182482171247,
+      "learning_rate": 9.83365975043602e-06,
+      "loss": 0.8205,
+      "step": 1345
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.052625787870512,
+      "learning_rate": 9.833320969462006e-06,
+      "loss": 0.9524,
+      "step": 1346
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.9363570141585504,
+      "learning_rate": 9.832981849693226e-06,
+      "loss": 0.6908,
+      "step": 1347
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.845194021761832,
+      "learning_rate": 9.832642391153452e-06,
+      "loss": 0.9929,
+      "step": 1348
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.250043298251886,
+      "learning_rate": 9.832302593866478e-06,
+      "loss": 1.3876,
+      "step": 1349
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.9347958176876774,
+      "learning_rate": 9.831962457856124e-06,
+      "loss": 0.897,
+      "step": 1350
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.7234753317121525,
+      "learning_rate": 9.831621983146227e-06,
+      "loss": 1.291,
+      "step": 1351
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.657411952759672,
+      "learning_rate": 9.83128116976066e-06,
+      "loss": 0.6593,
+      "step": 1352
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.4986134293129334,
+      "learning_rate": 9.830940017723308e-06,
+      "loss": 0.6853,
+      "step": 1353
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.6022819684060194,
+      "learning_rate": 9.830598527058083e-06,
+      "loss": 0.9227,
+      "step": 1354
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.0415847920356756,
+      "learning_rate": 9.830256697788924e-06,
+      "loss": 0.6046,
+      "step": 1355
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.222301158214487,
+      "learning_rate": 9.829914529939794e-06,
+      "loss": 0.6136,
+      "step": 1356
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.498417460325489,
+      "learning_rate": 9.829572023534675e-06,
+      "loss": 0.8575,
+      "step": 1357
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.3827152069059965,
+      "learning_rate": 9.829229178597575e-06,
+      "loss": 0.9866,
+      "step": 1358
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.531474859604405,
+      "learning_rate": 9.828885995152525e-06,
+      "loss": 1.1519,
+      "step": 1359
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.3485362372809284,
+      "learning_rate": 9.828542473223586e-06,
+      "loss": 0.4419,
+      "step": 1360
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.76206496778344,
+      "learning_rate": 9.82819861283483e-06,
+      "loss": 0.9233,
+      "step": 1361
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 6.539558811679356,
+      "learning_rate": 9.827854414010366e-06,
+      "loss": 1.7233,
+      "step": 1362
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.815566721141583,
+      "learning_rate": 9.827509876774315e-06,
+      "loss": 0.6861,
+      "step": 1363
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.757901899383783,
+      "learning_rate": 9.827165001150834e-06,
+      "loss": 1.1095,
+      "step": 1364
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.886305903447551,
+      "learning_rate": 9.826819787164095e-06,
+      "loss": 0.44,
+      "step": 1365
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.926176324529388,
+      "learning_rate": 9.826474234838293e-06,
+      "loss": 0.6576,
+      "step": 1366
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.459558687055677,
+      "learning_rate": 9.826128344197653e-06,
+      "loss": 1.2845,
+      "step": 1367
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.384367792526328,
+      "learning_rate": 9.82578211526642e-06,
+      "loss": 0.9965,
+      "step": 1368
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.3568468480970686,
+      "learning_rate": 9.825435548068862e-06,
+      "loss": 0.6548,
+      "step": 1369
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.872095489295021,
+      "learning_rate": 9.825088642629271e-06,
+      "loss": 1.259,
+      "step": 1370
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.260165583117883,
+      "learning_rate": 9.824741398971966e-06,
+      "loss": 0.9074,
+      "step": 1371
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.568108760017305,
+      "learning_rate": 9.824393817121288e-06,
+      "loss": 1.2388,
+      "step": 1372
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.952084072267941,
+      "learning_rate": 9.824045897101598e-06,
+      "loss": 0.7787,
+      "step": 1373
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 6.423914169789909,
+      "learning_rate": 9.823697638937283e-06,
+      "loss": 1.3275,
+      "step": 1374
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.7116454839421893,
+      "learning_rate": 9.82334904265276e-06,
+      "loss": 0.7016,
+      "step": 1375
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.2828456033633255,
+      "learning_rate": 9.823000108272458e-06,
+      "loss": 0.397,
+      "step": 1376
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.5598140162103995,
+      "learning_rate": 9.82265083582084e-06,
+      "loss": 0.7259,
+      "step": 1377
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.4343896057203995,
+      "learning_rate": 9.822301225322384e-06,
+      "loss": 0.9835,
+      "step": 1378
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.0649876892148185,
+      "learning_rate": 9.8219512768016e-06,
+      "loss": 0.6065,
+      "step": 1379
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.373673238593863,
+      "learning_rate": 9.821600990283018e-06,
+      "loss": 0.3573,
+      "step": 1380
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.275665587049736,
+      "learning_rate": 9.821250365791189e-06,
+      "loss": 1.1575,
+      "step": 1381
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.8296355375812245,
+      "learning_rate": 9.820899403350693e-06,
+      "loss": 1.0401,
+      "step": 1382
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.274330476790679,
+      "learning_rate": 9.820548102986126e-06,
+      "loss": 0.7298,
+      "step": 1383
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.54174599316339,
+      "learning_rate": 9.820196464722118e-06,
+      "loss": 0.5905,
+      "step": 1384
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.140626710176374,
+      "learning_rate": 9.819844488583316e-06,
+      "loss": 0.9191,
+      "step": 1385
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9979127187144647,
+      "learning_rate": 9.819492174594391e-06,
+      "loss": 0.4385,
+      "step": 1386
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.5767283557263294,
+      "learning_rate": 9.819139522780038e-06,
+      "loss": 0.6833,
+      "step": 1387
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.796622823599924,
+      "learning_rate": 9.81878653316498e-06,
+      "loss": 0.5222,
+      "step": 1388
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.207140533365846,
+      "learning_rate": 9.818433205773957e-06,
+      "loss": 1.5635,
+      "step": 1389
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.294209540551426,
+      "learning_rate": 9.818079540631732e-06,
+      "loss": 0.4936,
+      "step": 1390
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.614697231475761,
+      "learning_rate": 9.817725537763105e-06,
+      "loss": 0.6669,
+      "step": 1391
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.692084220618287,
+      "learning_rate": 9.817371197192883e-06,
+      "loss": 1.22,
+      "step": 1392
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.752787361806504,
+      "learning_rate": 9.817016518945904e-06,
+      "loss": 0.7487,
+      "step": 1393
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.355449957476081,
+      "learning_rate": 9.816661503047032e-06,
+      "loss": 0.8125,
+      "step": 1394
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.128535214061091,
+      "learning_rate": 9.816306149521149e-06,
+      "loss": 0.8633,
+      "step": 1395
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.5263385505708444,
+      "learning_rate": 9.815950458393166e-06,
+      "loss": 0.932,
+      "step": 1396
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.112181680433037,
+      "learning_rate": 9.815594429688015e-06,
+      "loss": 0.9579,
+      "step": 1397
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 6.645733663557989,
+      "learning_rate": 9.815238063430655e-06,
+      "loss": 1.6471,
+      "step": 1398
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.1645593081286645,
+      "learning_rate": 9.81488135964606e-06,
+      "loss": 0.5587,
+      "step": 1399
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.099895044129971,
+      "learning_rate": 9.814524318359235e-06,
+      "loss": 0.3865,
+      "step": 1400
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.3226922818944016,
+      "learning_rate": 9.81416693959521e-06,
+      "loss": 0.6881,
+      "step": 1401
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.234298859212859,
+      "learning_rate": 9.813809223379035e-06,
+      "loss": 0.7269,
+      "step": 1402
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.8633112567954715,
+      "learning_rate": 9.813451169735781e-06,
+      "loss": 1.3484,
+      "step": 1403
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.249634519658076,
+      "learning_rate": 9.813092778690549e-06,
+      "loss": 1.1644,
+      "step": 1404
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9572073819275848,
+      "learning_rate": 9.81273405026846e-06,
+      "loss": 0.3299,
+      "step": 1405
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.362360907175296,
+      "learning_rate": 9.81237498449466e-06,
+      "loss": 0.374,
+      "step": 1406
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.593558326449794,
+      "learning_rate": 9.812015581394316e-06,
+      "loss": 0.8593,
+      "step": 1407
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.463677247987641,
+      "learning_rate": 9.811655840992621e-06,
+      "loss": 1.2366,
+      "step": 1408
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.32660737602931,
+      "learning_rate": 9.811295763314793e-06,
+      "loss": 0.9649,
+      "step": 1409
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.616667588532532,
+      "learning_rate": 9.810935348386071e-06,
+      "loss": 0.6619,
+      "step": 1410
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.5831618295972625,
+      "learning_rate": 9.810574596231717e-06,
+      "loss": 0.9357,
+      "step": 1411
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.107669813938388,
+      "learning_rate": 9.810213506877021e-06,
+      "loss": 0.4273,
+      "step": 1412
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.507060032810869,
+      "learning_rate": 9.80985208034729e-06,
+      "loss": 1.0523,
+      "step": 1413
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.778321036304939,
+      "learning_rate": 9.809490316667864e-06,
+      "loss": 1.1447,
+      "step": 1414
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.34623077029894,
+      "learning_rate": 9.809128215864096e-06,
+      "loss": 0.4395,
+      "step": 1415
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.176105641856664,
+      "learning_rate": 9.80876577796137e-06,
+      "loss": 1.2677,
+      "step": 1416
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.7614486871129627,
+      "learning_rate": 9.808403002985089e-06,
+      "loss": 0.5742,
+      "step": 1417
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.620567580821391,
+      "learning_rate": 9.808039890960687e-06,
+      "loss": 0.7539,
+      "step": 1418
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.6874968173664113,
+      "learning_rate": 9.807676441913611e-06,
+      "loss": 0.4365,
+      "step": 1419
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.541845640368986,
+      "learning_rate": 9.80731265586934e-06,
+      "loss": 0.9808,
+      "step": 1420
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 7.123859209612732,
+      "learning_rate": 9.806948532853373e-06,
+      "loss": 1.6505,
+      "step": 1421
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.068838188795062,
+      "learning_rate": 9.806584072891234e-06,
+      "loss": 0.8492,
+      "step": 1422
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.347877420451936,
+      "learning_rate": 9.80621927600847e-06,
+      "loss": 0.7774,
+      "step": 1423
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.895252251166327,
+      "learning_rate": 9.805854142230652e-06,
+      "loss": 1.3487,
+      "step": 1424
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.267175221265885,
+      "learning_rate": 9.805488671583372e-06,
+      "loss": 0.3279,
+      "step": 1425
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.513202645820931,
+      "learning_rate": 9.80512286409225e-06,
+      "loss": 0.8466,
+      "step": 1426
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.423363671659631,
+      "learning_rate": 9.80475671978293e-06,
+      "loss": 0.7481,
+      "step": 1427
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.215881653445727,
+      "learning_rate": 9.804390238681072e-06,
+      "loss": 1.2405,
+      "step": 1428
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3735270370900747,
+      "learning_rate": 9.804023420812368e-06,
+      "loss": 0.2282,
+      "step": 1429
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.832715397250428,
+      "learning_rate": 9.803656266202528e-06,
+      "loss": 0.7381,
+      "step": 1430
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.34847388319156,
+      "learning_rate": 9.80328877487729e-06,
+      "loss": 0.8616,
+      "step": 1431
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.360354162722593,
+      "learning_rate": 9.802920946862413e-06,
+      "loss": 0.9632,
+      "step": 1432
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.954599355658937,
+      "learning_rate": 9.802552782183682e-06,
+      "loss": 1.3467,
+      "step": 1433
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.53958496323328,
+      "learning_rate": 9.802184280866898e-06,
+      "loss": 1.011,
+      "step": 1434
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.434045092648887,
+      "learning_rate": 9.801815442937897e-06,
+      "loss": 1.3907,
+      "step": 1435
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.6551014885191977,
+      "learning_rate": 9.80144626842253e-06,
+      "loss": 0.8479,
+      "step": 1436
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.6799219829730974,
+      "learning_rate": 9.801076757346677e-06,
+      "loss": 0.9056,
+      "step": 1437
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.763827428568593,
+      "learning_rate": 9.800706909736237e-06,
+      "loss": 1.3745,
+      "step": 1438
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.907388733876458,
+      "learning_rate": 9.800336725617136e-06,
+      "loss": 1.1375,
+      "step": 1439
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.367826033412014,
+      "learning_rate": 9.79996620501532e-06,
+      "loss": 1.2576,
+      "step": 1440
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.256085637336182,
+      "learning_rate": 9.799595347956764e-06,
+      "loss": 1.3381,
+      "step": 1441
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.1059074280871393,
+      "learning_rate": 9.79922415446746e-06,
+      "loss": 0.7916,
+      "step": 1442
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.329508448979159,
+      "learning_rate": 9.798852624573432e-06,
+      "loss": 0.9223,
+      "step": 1443
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.8386067688568892,
+      "learning_rate": 9.79848075830072e-06,
+      "loss": 0.5353,
+      "step": 1444
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.926444301146794,
+      "learning_rate": 9.798108555675388e-06,
+      "loss": 0.9793,
+      "step": 1445
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.056237294722397,
+      "learning_rate": 9.797736016723527e-06,
+      "loss": 0.8944,
+      "step": 1446
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.7066633631885098,
+      "learning_rate": 9.797363141471252e-06,
+      "loss": 0.5836,
+      "step": 1447
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5653471821648113,
+      "learning_rate": 9.796989929944699e-06,
+      "loss": 0.2457,
+      "step": 1448
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.171565444891341,
+      "learning_rate": 9.796616382170028e-06,
+      "loss": 1.3466,
+      "step": 1449
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.8421857025707835,
+      "learning_rate": 9.796242498173425e-06,
+      "loss": 1.4725,
+      "step": 1450
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1240727792271181,
+      "learning_rate": 9.795868277981095e-06,
+      "loss": 0.1594,
+      "step": 1451
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.2964525728779703,
+      "learning_rate": 9.795493721619271e-06,
+      "loss": 0.7353,
+      "step": 1452
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.073169173071545,
+      "learning_rate": 9.795118829114205e-06,
+      "loss": 0.5664,
+      "step": 1453
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.921321716329079,
+      "learning_rate": 9.79474360049218e-06,
+      "loss": 0.8704,
+      "step": 1454
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.326213766913278,
+      "learning_rate": 9.794368035779496e-06,
+      "loss": 0.6958,
+      "step": 1455
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.670985328015172,
+      "learning_rate": 9.793992135002476e-06,
+      "loss": 1.1243,
+      "step": 1456
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.921221306493759,
+      "learning_rate": 9.793615898187473e-06,
+      "loss": 1.4096,
+      "step": 1457
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.652832396557077,
+      "learning_rate": 9.793239325360855e-06,
+      "loss": 1.0637,
+      "step": 1458
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.7588457089114455,
+      "learning_rate": 9.792862416549021e-06,
+      "loss": 0.6599,
+      "step": 1459
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.13310023429643,
+      "learning_rate": 9.792485171778389e-06,
+      "loss": 0.3432,
+      "step": 1460
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.7093931763306576,
+      "learning_rate": 9.792107591075406e-06,
+      "loss": 0.8885,
+      "step": 1461
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.21724107443512,
+      "learning_rate": 9.791729674466534e-06,
+      "loss": 0.6405,
+      "step": 1462
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.805812423017493,
+      "learning_rate": 9.791351421978269e-06,
+      "loss": 1.5855,
+      "step": 1463
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.1123464014190056,
+      "learning_rate": 9.790972833637118e-06,
+      "loss": 0.3719,
+      "step": 1464
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.7958239629043624,
+      "learning_rate": 9.790593909469623e-06,
+      "loss": 0.4876,
+      "step": 1465
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.37017892442726,
+      "learning_rate": 9.790214649502343e-06,
+      "loss": 1.2543,
+      "step": 1466
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.9239535160362804,
+      "learning_rate": 9.789835053761865e-06,
+      "loss": 0.8073,
+      "step": 1467
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.555633739107286,
+      "learning_rate": 9.789455122274793e-06,
+      "loss": 1.0265,
+      "step": 1468
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.869738418739143,
+      "learning_rate": 9.789074855067761e-06,
+      "loss": 1.1461,
+      "step": 1469
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.8066477350594425,
+      "learning_rate": 9.788694252167424e-06,
+      "loss": 0.6731,
+      "step": 1470
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.9441755106990506,
+      "learning_rate": 9.788313313600462e-06,
+      "loss": 0.6184,
+      "step": 1471
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.9209845177925553,
+      "learning_rate": 9.787932039393574e-06,
+      "loss": 0.7527,
+      "step": 1472
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.5411521898176197,
+      "learning_rate": 9.787550429573487e-06,
+      "loss": 0.6581,
+      "step": 1473
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.588217832482472,
+      "learning_rate": 9.78716848416695e-06,
+      "loss": 1.0138,
+      "step": 1474
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.908534892628834,
+      "learning_rate": 9.786786203200738e-06,
+      "loss": 1.1612,
+      "step": 1475
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.6631338624314527,
+      "learning_rate": 9.786403586701643e-06,
+      "loss": 0.9183,
+      "step": 1476
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.24287728720431,
+      "learning_rate": 9.786020634696489e-06,
+      "loss": 1.4494,
+      "step": 1477
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.355388150801754,
+      "learning_rate": 9.785637347212117e-06,
+      "loss": 1.1414,
+      "step": 1478
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.574465500450339,
+      "learning_rate": 9.785253724275394e-06,
+      "loss": 0.948,
+      "step": 1479
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.8645555875140043,
+      "learning_rate": 9.78486976591321e-06,
+      "loss": 1.345,
+      "step": 1480
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.0871937132066,
+      "learning_rate": 9.784485472152479e-06,
+      "loss": 0.7189,
+      "step": 1481
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.754552849267925,
+      "learning_rate": 9.784100843020139e-06,
+      "loss": 1.1957,
+      "step": 1482
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.231759982859807,
+      "learning_rate": 9.783715878543149e-06,
+      "loss": 0.5092,
+      "step": 1483
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.9490285385860218,
+      "learning_rate": 9.783330578748497e-06,
+      "loss": 0.8589,
+      "step": 1484
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.2352180597670688,
+      "learning_rate": 9.782944943663187e-06,
+      "loss": 0.6635,
+      "step": 1485
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.184309454595811,
+      "learning_rate": 9.782558973314254e-06,
+      "loss": 0.41,
+      "step": 1486
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.461287240576972,
+      "learning_rate": 9.782172667728748e-06,
+      "loss": 1.5705,
+      "step": 1487
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.9641417501280753,
+      "learning_rate": 9.781786026933752e-06,
+      "loss": 0.3327,
+      "step": 1488
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.7825304552740193,
+      "learning_rate": 9.781399050956364e-06,
+      "loss": 0.5386,
+      "step": 1489
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.403847438502365,
+      "learning_rate": 9.781011739823715e-06,
+      "loss": 0.9521,
+      "step": 1490
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.746341743522744,
+      "learning_rate": 9.780624093562944e-06,
+      "loss": 0.8428,
+      "step": 1491
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.220237980809123,
+      "learning_rate": 9.780236112201235e-06,
+      "loss": 0.9644,
+      "step": 1492
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9812539161163594,
+      "learning_rate": 9.779847795765776e-06,
+      "loss": 0.3911,
+      "step": 1493
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.3386624342418085,
+      "learning_rate": 9.779459144283788e-06,
+      "loss": 0.8402,
+      "step": 1494
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.488310991588463,
+      "learning_rate": 9.779070157782515e-06,
+      "loss": 0.7609,
+      "step": 1495
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.805387887932308,
+      "learning_rate": 9.778680836289222e-06,
+      "loss": 0.5685,
+      "step": 1496
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.6968205267244985,
+      "learning_rate": 9.778291179831201e-06,
+      "loss": 0.9227,
+      "step": 1497
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.1847305569872852,
+      "learning_rate": 9.777901188435762e-06,
+      "loss": 0.6676,
+      "step": 1498
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.209245828758046,
+      "learning_rate": 9.777510862130242e-06,
+      "loss": 1.0329,
+      "step": 1499
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.037052184090017,
+      "learning_rate": 9.777120200942004e-06,
+      "loss": 0.6037,
+      "step": 1500
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.672518759031155,
+      "learning_rate": 9.77672920489843e-06,
+      "loss": 0.5179,
+      "step": 1501
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.6466801675576104,
+      "learning_rate": 9.776337874026926e-06,
+      "loss": 0.7091,
+      "step": 1502
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.411916495301089,
+      "learning_rate": 9.775946208354924e-06,
+      "loss": 1.0544,
+      "step": 1503
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.060051829029946,
+      "learning_rate": 9.775554207909879e-06,
+      "loss": 0.9035,
+      "step": 1504
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.07841372268714,
+      "learning_rate": 9.775161872719268e-06,
+      "loss": 0.7718,
+      "step": 1505
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.4655063677047417,
+      "learning_rate": 9.77476920281059e-06,
+      "loss": 0.6952,
+      "step": 1506
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.008834167325511,
+      "learning_rate": 9.77437619821137e-06,
+      "loss": 0.4992,
+      "step": 1507
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.618330088193883,
+      "learning_rate": 9.773982858949157e-06,
+      "loss": 0.8979,
+      "step": 1508
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.409746682340359,
+      "learning_rate": 9.773589185051522e-06,
+      "loss": 1.0297,
+      "step": 1509
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.7310014040891994,
+      "learning_rate": 9.77319517654606e-06,
+      "loss": 1.2448,
+      "step": 1510
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.9619728725901817,
+      "learning_rate": 9.77280083346039e-06,
+      "loss": 0.5431,
+      "step": 1511
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.205388385519706,
+      "learning_rate": 9.772406155822152e-06,
+      "loss": 0.773,
+      "step": 1512
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.5967486073724833,
+      "learning_rate": 9.772011143659013e-06,
+      "loss": 0.805,
+      "step": 1513
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.3716225881025577,
+      "learning_rate": 9.77161579699866e-06,
+      "loss": 0.5813,
+      "step": 1514
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0722565727923925,
+      "learning_rate": 9.771220115868805e-06,
+      "loss": 0.3726,
+      "step": 1515
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.1345323779270755,
+      "learning_rate": 9.770824100297185e-06,
+      "loss": 1.6538,
+      "step": 1516
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.337949991423968,
+      "learning_rate": 9.770427750311557e-06,
+      "loss": 1.219,
+      "step": 1517
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.0302519135261194,
+      "learning_rate": 9.770031065939707e-06,
+      "loss": 1.1707,
+      "step": 1518
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.807159466088738,
+      "learning_rate": 9.769634047209437e-06,
+      "loss": 1.3163,
+      "step": 1519
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.4439580054897303,
+      "learning_rate": 9.769236694148579e-06,
+      "loss": 0.4129,
+      "step": 1520
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.4429376768979445,
+      "learning_rate": 9.768839006784983e-06,
+      "loss": 0.9111,
+      "step": 1521
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.056315866869216,
+      "learning_rate": 9.768440985146529e-06,
+      "loss": 0.9465,
+      "step": 1522
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.043077745088097,
+      "learning_rate": 9.768042629261112e-06,
+      "loss": 0.4399,
+      "step": 1523
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.3192423340805197,
+      "learning_rate": 9.767643939156658e-06,
+      "loss": 0.5992,
+      "step": 1524
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.724102578486323,
+      "learning_rate": 9.767244914861114e-06,
+      "loss": 0.7851,
+      "step": 1525
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.960661209852135,
+      "learning_rate": 9.766845556402447e-06,
+      "loss": 0.853,
+      "step": 1526
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.132502240121899,
+      "learning_rate": 9.766445863808652e-06,
+      "loss": 0.8542,
+      "step": 1527
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.464818791787063,
+      "learning_rate": 9.766045837107745e-06,
+      "loss": 0.6607,
+      "step": 1528
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.9151827316498458,
+      "learning_rate": 9.765645476327768e-06,
+      "loss": 0.6962,
+      "step": 1529
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.397921765887368,
+      "learning_rate": 9.765244781496783e-06,
+      "loss": 1.2484,
+      "step": 1530
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.647934850478616,
+      "learning_rate": 9.764843752642876e-06,
+      "loss": 1.2478,
+      "step": 1531
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.3059946410621546,
+      "learning_rate": 9.76444238979416e-06,
+      "loss": 0.4285,
+      "step": 1532
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.8796632727172904,
+      "learning_rate": 9.764040692978767e-06,
+      "loss": 1.0721,
+      "step": 1533
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.532900000889822,
+      "learning_rate": 9.763638662224854e-06,
+      "loss": 0.8132,
+      "step": 1534
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.640399343376495,
+      "learning_rate": 9.763236297560603e-06,
+      "loss": 0.7806,
+      "step": 1535
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.1418755348948055,
+      "learning_rate": 9.762833599014219e-06,
+      "loss": 0.5318,
+      "step": 1536
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.586786253813165,
+      "learning_rate": 9.762430566613925e-06,
+      "loss": 0.9011,
+      "step": 1537
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.8280442642483803,
+      "learning_rate": 9.762027200387974e-06,
+      "loss": 0.5989,
+      "step": 1538
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.916761256288874,
+      "learning_rate": 9.761623500364643e-06,
+      "loss": 1.084,
+      "step": 1539
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.1449798089696324,
+      "learning_rate": 9.761219466572227e-06,
+      "loss": 0.8712,
+      "step": 1540
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.0647901636925643,
+      "learning_rate": 9.760815099039045e-06,
+      "loss": 0.6931,
+      "step": 1541
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.645838963488648,
+      "learning_rate": 9.760410397793445e-06,
+      "loss": 1.1038,
+      "step": 1542
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.045311760642244,
+      "learning_rate": 9.760005362863794e-06,
+      "loss": 0.8072,
+      "step": 1543
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.1799765839075955,
+      "learning_rate": 9.759599994278481e-06,
+      "loss": 0.6287,
+      "step": 1544
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.814868528193712,
+      "learning_rate": 9.759194292065925e-06,
+      "loss": 0.8318,
+      "step": 1545
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.554545532458709,
+      "learning_rate": 9.758788256254559e-06,
+      "loss": 1.5856,
+      "step": 1546
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.196075660470384,
+      "learning_rate": 9.758381886872848e-06,
+      "loss": 0.958,
+      "step": 1547
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0660407285250322,
+      "learning_rate": 9.757975183949275e-06,
+      "loss": 0.1906,
+      "step": 1548
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.842620018989324,
+      "learning_rate": 9.757568147512347e-06,
+      "loss": 1.0968,
+      "step": 1549
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.776522241854018,
+      "learning_rate": 9.757160777590597e-06,
+      "loss": 0.9168,
+      "step": 1550
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.409198973311985,
+      "learning_rate": 9.756753074212581e-06,
+      "loss": 1.8172,
+      "step": 1551
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.1874379937760207,
+      "learning_rate": 9.756345037406876e-06,
+      "loss": 0.5296,
+      "step": 1552
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.191603881891276,
+      "learning_rate": 9.755936667202084e-06,
+      "loss": 1.0245,
+      "step": 1553
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.9092492108199655,
+      "learning_rate": 9.755527963626828e-06,
+      "loss": 0.8931,
+      "step": 1554
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.429272274630198,
+      "learning_rate": 9.755118926709757e-06,
+      "loss": 0.6388,
+      "step": 1555
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.9498809549310594,
+      "learning_rate": 9.754709556479546e-06,
+      "loss": 0.6189,
+      "step": 1556
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.36139121660757,
+      "learning_rate": 9.754299852964886e-06,
+      "loss": 0.8474,
+      "step": 1557
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.489779862984102,
+      "learning_rate": 9.753889816194498e-06,
+      "loss": 0.7664,
+      "step": 1558
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.718174282412993,
+      "learning_rate": 9.753479446197122e-06,
+      "loss": 0.9349,
+      "step": 1559
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.191140162879363,
+      "learning_rate": 9.753068743001525e-06,
+      "loss": 0.5427,
+      "step": 1560
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.3149836151747762,
+      "learning_rate": 9.752657706636494e-06,
+      "loss": 0.4822,
+      "step": 1561
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 6.986035668649894,
+      "learning_rate": 9.752246337130841e-06,
+      "loss": 0.9761,
+      "step": 1562
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.6889568491015656,
+      "learning_rate": 9.751834634513404e-06,
+      "loss": 0.6154,
+      "step": 1563
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.1432527726484345,
+      "learning_rate": 9.751422598813037e-06,
+      "loss": 0.5122,
+      "step": 1564
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.3942869359327195,
+      "learning_rate": 9.751010230058624e-06,
+      "loss": 0.7232,
+      "step": 1565
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.3100380532903135,
+      "learning_rate": 9.750597528279072e-06,
+      "loss": 0.5781,
+      "step": 1566
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.7308147712033355,
+      "learning_rate": 9.750184493503306e-06,
+      "loss": 1.463,
+      "step": 1567
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.8363396138495314,
+      "learning_rate": 9.749771125760281e-06,
+      "loss": 0.623,
+      "step": 1568
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.067053229913921,
+      "learning_rate": 9.74935742507897e-06,
+      "loss": 0.8514,
+      "step": 1569
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.582862830286583,
+      "learning_rate": 9.748943391488374e-06,
+      "loss": 1.1104,
+      "step": 1570
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.7410002095176718,
+      "learning_rate": 9.748529025017512e-06,
+      "loss": 0.649,
+      "step": 1571
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.59880905028096,
+      "learning_rate": 9.748114325695433e-06,
+      "loss": 0.8195,
+      "step": 1572
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.108543123341428,
+      "learning_rate": 9.747699293551203e-06,
+      "loss": 0.6387,
+      "step": 1573
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.757953511646958,
+      "learning_rate": 9.747283928613915e-06,
+      "loss": 1.1136,
+      "step": 1574
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.635770244508862,
+      "learning_rate": 9.746868230912683e-06,
+      "loss": 0.6198,
+      "step": 1575
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.8130446795283235,
+      "learning_rate": 9.746452200476647e-06,
+      "loss": 0.6654,
+      "step": 1576
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.229086650299244,
+      "learning_rate": 9.74603583733497e-06,
+      "loss": 0.8575,
+      "step": 1577
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.877197051217789,
+      "learning_rate": 9.745619141516833e-06,
+      "loss": 0.7622,
+      "step": 1578
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.832704704597798,
+      "learning_rate": 9.745202113051448e-06,
+      "loss": 1.442,
+      "step": 1579
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.83291036043407,
+      "learning_rate": 9.744784751968046e-06,
+      "loss": 1.1734,
+      "step": 1580
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.766668118698729,
+      "learning_rate": 9.744367058295881e-06,
+      "loss": 1.0921,
+      "step": 1581
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.7452921009672466,
+      "learning_rate": 9.743949032064235e-06,
+      "loss": 0.8583,
+      "step": 1582
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.5324161965107184,
+      "learning_rate": 9.743530673302407e-06,
+      "loss": 0.615,
+      "step": 1583
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.96485593458066,
+      "learning_rate": 9.743111982039721e-06,
+      "loss": 1.2621,
+      "step": 1584
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.404087488808251,
+      "learning_rate": 9.742692958305528e-06,
+      "loss": 0.6942,
+      "step": 1585
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.4955877818953915,
+      "learning_rate": 9.742273602129201e-06,
+      "loss": 1.2858,
+      "step": 1586
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.213928992509133,
+      "learning_rate": 9.741853913540132e-06,
+      "loss": 1.1965,
+      "step": 1587
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.190641501195003,
+      "learning_rate": 9.74143389256774e-06,
+      "loss": 0.7693,
+      "step": 1588
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.70693513925525,
+      "learning_rate": 9.741013539241467e-06,
+      "loss": 0.6004,
+      "step": 1589
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.8013575804530837,
+      "learning_rate": 9.740592853590776e-06,
+      "loss": 0.7414,
+      "step": 1590
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.125539338774779,
+      "learning_rate": 9.740171835645158e-06,
+      "loss": 0.7715,
+      "step": 1591
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8688054107462326,
+      "learning_rate": 9.739750485434126e-06,
+      "loss": 0.4439,
+      "step": 1592
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.3397393731341016,
+      "learning_rate": 9.739328802987209e-06,
+      "loss": 0.6942,
+      "step": 1593
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3694727764828176,
+      "learning_rate": 9.738906788333971e-06,
+      "loss": 0.2148,
+      "step": 1594
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.298421483587479,
+      "learning_rate": 9.738484441503989e-06,
+      "loss": 0.4418,
+      "step": 1595
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.0267036380367305,
+      "learning_rate": 9.738061762526871e-06,
+      "loss": 0.6679,
+      "step": 1596
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.346525534672905,
+      "learning_rate": 9.737638751432244e-06,
+      "loss": 0.7686,
+      "step": 1597
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.250728585271418,
+      "learning_rate": 9.737215408249757e-06,
+      "loss": 1.2046,
+      "step": 1598
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.722914124483091,
+      "learning_rate": 9.736791733009087e-06,
+      "loss": 0.9467,
+      "step": 1599
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.902253643134758,
+      "learning_rate": 9.736367725739932e-06,
+      "loss": 1.284,
+      "step": 1600
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.683735001274794,
+      "learning_rate": 9.735943386472012e-06,
+      "loss": 1.4378,
+      "step": 1601
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.20889208970552,
+      "learning_rate": 9.73551871523507e-06,
+      "loss": 0.7832,
+      "step": 1602
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.0226202867704597,
+      "learning_rate": 9.735093712058876e-06,
+      "loss": 0.5958,
+      "step": 1603
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.3969939040986987,
+      "learning_rate": 9.73466837697322e-06,
+      "loss": 0.3861,
+      "step": 1604
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.404473157535664,
+      "learning_rate": 9.734242710007918e-06,
+      "loss": 0.2924,
+      "step": 1605
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.237250360859097,
+      "learning_rate": 9.733816711192803e-06,
+      "loss": 0.3559,
+      "step": 1606
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.6456248141718715,
+      "learning_rate": 9.733390380557739e-06,
+      "loss": 0.4604,
+      "step": 1607
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.761247734859934,
+      "learning_rate": 9.732963718132609e-06,
+      "loss": 1.1434,
+      "step": 1608
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.707625528974655,
+      "learning_rate": 9.73253672394732e-06,
+      "loss": 0.8052,
+      "step": 1609
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.0816386035698535,
+      "learning_rate": 9.732109398031804e-06,
+      "loss": 0.5118,
+      "step": 1610
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.1755136938926745,
+      "learning_rate": 9.731681740416012e-06,
+      "loss": 1.3818,
+      "step": 1611
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.469392922393817,
+      "learning_rate": 9.731253751129923e-06,
+      "loss": 0.3968,
+      "step": 1612
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.3961943593224126,
+      "learning_rate": 9.730825430203536e-06,
+      "loss": 0.7839,
+      "step": 1613
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.042124620832441,
+      "learning_rate": 9.730396777666875e-06,
+      "loss": 0.8608,
+      "step": 1614
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.685312270195895,
+      "learning_rate": 9.729967793549987e-06,
+      "loss": 0.968,
+      "step": 1615
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.494774242668652,
+      "learning_rate": 9.729538477882942e-06,
+      "loss": 0.896,
+      "step": 1616
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.48705535782037,
+      "learning_rate": 9.729108830695833e-06,
+      "loss": 1.0858,
+      "step": 1617
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.079053530713893,
+      "learning_rate": 9.728678852018775e-06,
+      "loss": 0.3735,
+      "step": 1618
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.522198117801278,
+      "learning_rate": 9.728248541881909e-06,
+      "loss": 0.7978,
+      "step": 1619
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.6673248515297145,
+      "learning_rate": 9.727817900315399e-06,
+      "loss": 0.7038,
+      "step": 1620
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.3868342089022954,
+      "learning_rate": 9.727386927349427e-06,
+      "loss": 0.4164,
+      "step": 1621
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.096149179181914,
+      "learning_rate": 9.726955623014207e-06,
+      "loss": 0.7268,
+      "step": 1622
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.1892463351865405,
+      "learning_rate": 9.72652398733997e-06,
+      "loss": 0.999,
+      "step": 1623
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.111134640592722,
+      "learning_rate": 9.72609202035697e-06,
+      "loss": 0.8955,
+      "step": 1624
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.619508841593312,
+      "learning_rate": 9.725659722095488e-06,
+      "loss": 0.8145,
+      "step": 1625
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.284117540605217,
+      "learning_rate": 9.725227092585824e-06,
+      "loss": 0.7883,
+      "step": 1626
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3053035878084933,
+      "learning_rate": 9.724794131858309e-06,
+      "loss": 0.2017,
+      "step": 1627
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.787383239731883,
+      "learning_rate": 9.724360839943285e-06,
+      "loss": 0.7844,
+      "step": 1628
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.566361724602944,
+      "learning_rate": 9.723927216871127e-06,
+      "loss": 0.623,
+      "step": 1629
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4600819399127305,
+      "learning_rate": 9.723493262672229e-06,
+      "loss": 0.2211,
+      "step": 1630
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.008762197460526,
+      "learning_rate": 9.723058977377012e-06,
+      "loss": 1.1448,
+      "step": 1631
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 6.186521654143597,
+      "learning_rate": 9.722624361015913e-06,
+      "loss": 1.0704,
+      "step": 1632
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.255225640204374,
+      "learning_rate": 9.7221894136194e-06,
+      "loss": 1.6081,
+      "step": 1633
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 6.173596004247248,
+      "learning_rate": 9.72175413521796e-06,
+      "loss": 1.2399,
+      "step": 1634
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.385440698351537,
+      "learning_rate": 9.721318525842105e-06,
+      "loss": 0.467,
+      "step": 1635
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.450899861913443,
+      "learning_rate": 9.720882585522368e-06,
+      "loss": 0.7838,
+      "step": 1636
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.001750450563991,
+      "learning_rate": 9.720446314289309e-06,
+      "loss": 0.7508,
+      "step": 1637
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.809705500312421,
+      "learning_rate": 9.720009712173504e-06,
+      "loss": 0.7776,
+      "step": 1638
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.0593428034581707,
+      "learning_rate": 9.719572779205562e-06,
+      "loss": 0.4057,
+      "step": 1639
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 6.409124956173562,
+      "learning_rate": 9.719135515416107e-06,
+      "loss": 1.4807,
+      "step": 1640
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.9182037085275163,
+      "learning_rate": 9.71869792083579e-06,
+      "loss": 0.5006,
+      "step": 1641
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8982368084936982,
+      "learning_rate": 9.718259995495284e-06,
+      "loss": 0.3411,
+      "step": 1642
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.6506744249878254,
+      "learning_rate": 9.717821739425286e-06,
+      "loss": 0.6956,
+      "step": 1643
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.7552476156750574,
+      "learning_rate": 9.717383152656518e-06,
+      "loss": 1.046,
+      "step": 1644
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0608675911426966,
+      "learning_rate": 9.71694423521972e-06,
+      "loss": 0.4058,
+      "step": 1645
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.030380844163063,
+      "learning_rate": 9.716504987145658e-06,
+      "loss": 1.2439,
+      "step": 1646
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.742621411829504,
+      "learning_rate": 9.716065408465124e-06,
+      "loss": 0.8299,
+      "step": 1647
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.194510187995256,
+      "learning_rate": 9.715625499208931e-06,
+      "loss": 1.0898,
+      "step": 1648
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.2683273114298403,
+      "learning_rate": 9.715185259407911e-06,
+      "loss": 0.8357,
+      "step": 1649
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.4839620674400145,
+      "learning_rate": 9.714744689092925e-06,
+      "loss": 1.0901,
+      "step": 1650
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.131210682499025,
+      "learning_rate": 9.714303788294854e-06,
+      "loss": 0.8754,
+      "step": 1651
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.6719171196306264,
+      "learning_rate": 9.713862557044607e-06,
+      "loss": 0.4085,
+      "step": 1652
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.524196692690876,
+      "learning_rate": 9.713420995373108e-06,
+      "loss": 0.6716,
+      "step": 1653
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3654936297880624,
+      "learning_rate": 9.712979103311308e-06,
+      "loss": 0.2285,
+      "step": 1654
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.852407882522216,
+      "learning_rate": 9.712536880890186e-06,
+      "loss": 1.2415,
+      "step": 1655
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.363295234003041,
+      "learning_rate": 9.712094328140738e-06,
+      "loss": 1.155,
+      "step": 1656
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.119060952663525,
+      "learning_rate": 9.711651445093984e-06,
+      "loss": 1.4858,
+      "step": 1657
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 6.07263896623256,
+      "learning_rate": 9.711208231780969e-06,
+      "loss": 1.5487,
+      "step": 1658
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.417118269313506,
+      "learning_rate": 9.71076468823276e-06,
+      "loss": 1.0961,
+      "step": 1659
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.240636340765624,
+      "learning_rate": 9.710320814480448e-06,
+      "loss": 1.0508,
+      "step": 1660
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.7871496737720873,
+      "learning_rate": 9.709876610555148e-06,
+      "loss": 0.7364,
+      "step": 1661
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.7911961749043086,
+      "learning_rate": 9.709432076487991e-06,
+      "loss": 0.3766,
+      "step": 1662
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.285382931590067,
+      "learning_rate": 9.708987212310144e-06,
+      "loss": 0.7827,
+      "step": 1663
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.770461234555658,
+      "learning_rate": 9.708542018052786e-06,
+      "loss": 0.378,
+      "step": 1664
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.6635570174102163,
+      "learning_rate": 9.708096493747123e-06,
+      "loss": 0.6873,
+      "step": 1665
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 6.309510039579954,
+      "learning_rate": 9.70765063942439e-06,
+      "loss": 1.3979,
+      "step": 1666
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.5408722162646455,
+      "learning_rate": 9.707204455115829e-06,
+      "loss": 0.9899,
+      "step": 1667
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.983879844016031,
+      "learning_rate": 9.706757940852724e-06,
+      "loss": 0.6502,
+      "step": 1668
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.611735659887384,
+      "learning_rate": 9.706311096666372e-06,
+      "loss": 0.8234,
+      "step": 1669
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.342774436992476,
+      "learning_rate": 9.705863922588093e-06,
+      "loss": 0.5715,
+      "step": 1670
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.347075365810961,
+      "learning_rate": 9.705416418649233e-06,
+      "loss": 0.685,
+      "step": 1671
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.097322467099301,
+      "learning_rate": 9.704968584881163e-06,
+      "loss": 0.957,
+      "step": 1672
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.284651690740073,
+      "learning_rate": 9.704520421315268e-06,
+      "loss": 0.412,
+      "step": 1673
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.9964240169463348,
+      "learning_rate": 9.704071927982966e-06,
+      "loss": 0.9967,
+      "step": 1674
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8789215636177454,
+      "learning_rate": 9.703623104915696e-06,
+      "loss": 0.1869,
+      "step": 1675
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.3620035126073153,
+      "learning_rate": 9.703173952144915e-06,
+      "loss": 0.3587,
+      "step": 1676
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.0358459729420963,
+      "learning_rate": 9.702724469702107e-06,
+      "loss": 0.5392,
+      "step": 1677
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.952662389243149,
+      "learning_rate": 9.702274657618781e-06,
+      "loss": 1.0653,
+      "step": 1678
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.242149136790763,
+      "learning_rate": 9.701824515926469e-06,
+      "loss": 0.9788,
+      "step": 1679
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.708365030480298,
+      "learning_rate": 9.701374044656716e-06,
+      "loss": 0.793,
+      "step": 1680
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.764014596285324,
+      "learning_rate": 9.700923243841106e-06,
+      "loss": 1.0391,
+      "step": 1681
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.482193777342618,
+      "learning_rate": 9.700472113511234e-06,
+      "loss": 0.6562,
+      "step": 1682
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.047861768371731,
+      "learning_rate": 9.700020653698722e-06,
+      "loss": 1.6494,
+      "step": 1683
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.170079677755622,
+      "learning_rate": 9.699568864435219e-06,
+      "loss": 1.4652,
+      "step": 1684
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.076027692146769,
+      "learning_rate": 9.69911674575239e-06,
+      "loss": 0.9265,
+      "step": 1685
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.85699929880583,
+      "learning_rate": 9.698664297681929e-06,
+      "loss": 0.4537,
+      "step": 1686
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0338154983394117,
+      "learning_rate": 9.698211520255549e-06,
+      "loss": 0.2813,
+      "step": 1687
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.875842754777342,
+      "learning_rate": 9.697758413504987e-06,
+      "loss": 0.3438,
+      "step": 1688
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.029530453781569,
+      "learning_rate": 9.697304977462005e-06,
+      "loss": 0.9498,
+      "step": 1689
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.754246632963344,
+      "learning_rate": 9.696851212158388e-06,
+      "loss": 1.0082,
+      "step": 1690
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.549144461758611,
+      "learning_rate": 9.696397117625942e-06,
+      "loss": 0.6994,
+      "step": 1691
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.530806771898034,
+      "learning_rate": 9.695942693896495e-06,
+      "loss": 0.6279,
+      "step": 1692
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.794353861007372,
+      "learning_rate": 9.695487941001905e-06,
+      "loss": 0.8341,
+      "step": 1693
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.96207377085134,
+      "learning_rate": 9.695032858974042e-06,
+      "loss": 0.755,
+      "step": 1694
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.6405072518293555,
+      "learning_rate": 9.694577447844809e-06,
+      "loss": 1.0264,
+      "step": 1695
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.4635637170023017,
+      "learning_rate": 9.694121707646129e-06,
+      "loss": 0.8758,
+      "step": 1696
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.624796096775306,
+      "learning_rate": 9.693665638409946e-06,
+      "loss": 1.5846,
+      "step": 1697
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.3551310744837854,
+      "learning_rate": 9.693209240168227e-06,
+      "loss": 0.6217,
+      "step": 1698
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.6274131018894096,
+      "learning_rate": 9.692752512952968e-06,
+      "loss": 0.9201,
+      "step": 1699
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.944204078548009,
+      "learning_rate": 9.692295456796178e-06,
+      "loss": 1.2045,
+      "step": 1700
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.46614524346582,
+      "learning_rate": 9.691838071729899e-06,
+      "loss": 1.0025,
+      "step": 1701
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.707777143872066,
+      "learning_rate": 9.691380357786189e-06,
+      "loss": 0.8076,
+      "step": 1702
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.450657216952367,
+      "learning_rate": 9.690922314997132e-06,
+      "loss": 0.9325,
+      "step": 1703
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.1716327050722235,
+      "learning_rate": 9.69046394339484e-06,
+      "loss": 0.663,
+      "step": 1704
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.7394523838000957,
+      "learning_rate": 9.690005243011436e-06,
+      "loss": 0.7823,
+      "step": 1705
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.4636145269488483,
+      "learning_rate": 9.689546213879074e-06,
+      "loss": 0.6237,
+      "step": 1706
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.690978397307599,
+      "learning_rate": 9.689086856029931e-06,
+      "loss": 0.7688,
+      "step": 1707
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.86435667426149,
+      "learning_rate": 9.68862716949621e-06,
+      "loss": 0.8823,
+      "step": 1708
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 6.120164203164415,
+      "learning_rate": 9.688167154310127e-06,
+      "loss": 1.3421,
+      "step": 1709
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.418183466759219,
+      "learning_rate": 9.68770681050393e-06,
+      "loss": 0.8123,
+      "step": 1710
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.6492709004623847,
+      "learning_rate": 9.687246138109888e-06,
+      "loss": 0.5541,
+      "step": 1711
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.604298506319751,
+      "learning_rate": 9.686785137160287e-06,
+      "loss": 0.5288,
+      "step": 1712
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.706652364518963,
+      "learning_rate": 9.686323807687447e-06,
+      "loss": 1.2949,
+      "step": 1713
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.5539009429313473,
+      "learning_rate": 9.685862149723703e-06,
+      "loss": 0.8843,
+      "step": 1714
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.9510422841886905,
+      "learning_rate": 9.685400163301415e-06,
+      "loss": 0.7165,
+      "step": 1715
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4029927629261745,
+      "learning_rate": 9.684937848452966e-06,
+      "loss": 0.2267,
+      "step": 1716
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.8532270273503113,
+      "learning_rate": 9.684475205210764e-06,
+      "loss": 0.4159,
+      "step": 1717
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.683097077822043,
+      "learning_rate": 9.684012233607237e-06,
+      "loss": 0.7487,
+      "step": 1718
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.564688641828072,
+      "learning_rate": 9.683548933674837e-06,
+      "loss": 0.6828,
+      "step": 1719
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 6.669827615735416,
+      "learning_rate": 9.683085305446038e-06,
+      "loss": 1.514,
+      "step": 1720
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.390248451343604,
+      "learning_rate": 9.682621348953343e-06,
+      "loss": 0.8943,
+      "step": 1721
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.465928623189289,
+      "learning_rate": 9.682157064229268e-06,
+      "loss": 0.777,
+      "step": 1722
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.4567929831863875,
+      "learning_rate": 9.68169245130636e-06,
+      "loss": 0.9962,
+      "step": 1723
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.5170790192773405,
+      "learning_rate": 9.681227510217186e-06,
+      "loss": 0.7667,
+      "step": 1724
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.8671609039319743,
+      "learning_rate": 9.680762240994336e-06,
+      "loss": 0.8341,
+      "step": 1725
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.090436071858452,
+      "learning_rate": 9.680296643670425e-06,
+      "loss": 0.4865,
+      "step": 1726
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.830201007655817,
+      "learning_rate": 9.679830718278087e-06,
+      "loss": 0.884,
+      "step": 1727
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 6.80538832131407,
+      "learning_rate": 9.679364464849983e-06,
+      "loss": 1.8187,
+      "step": 1728
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.6906925076304757,
+      "learning_rate": 9.678897883418794e-06,
+      "loss": 0.4176,
+      "step": 1729
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.0343769521331945,
+      "learning_rate": 9.678430974017226e-06,
+      "loss": 1.252,
+      "step": 1730
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.8324493094817256,
+      "learning_rate": 9.677963736678007e-06,
+      "loss": 0.4308,
+      "step": 1731
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.4653845451836105,
+      "learning_rate": 9.677496171433889e-06,
+      "loss": 0.9912,
+      "step": 1732
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.9840595685560345,
+      "learning_rate": 9.677028278317646e-06,
+      "loss": 0.7753,
+      "step": 1733
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.8594742130196154,
+      "learning_rate": 9.676560057362076e-06,
+      "loss": 0.7411,
+      "step": 1734
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.5574909671573813,
+      "learning_rate": 9.676091508599995e-06,
+      "loss": 0.8224,
+      "step": 1735
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.469295454039877,
+      "learning_rate": 9.675622632064255e-06,
+      "loss": 1.0785,
+      "step": 1736
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.839232896804892,
+      "learning_rate": 9.675153427787713e-06,
+      "loss": 1.437,
+      "step": 1737
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.4990413210459708,
+      "learning_rate": 9.674683895803262e-06,
+      "loss": 0.7086,
+      "step": 1738
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 6.6088857858705525,
+      "learning_rate": 9.674214036143817e-06,
+      "loss": 1.5871,
+      "step": 1739
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.3806875521861675,
+      "learning_rate": 9.673743848842309e-06,
+      "loss": 0.291,
+      "step": 1740
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.384847468400944,
+      "learning_rate": 9.673273333931696e-06,
+      "loss": 1.3201,
+      "step": 1741
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.7623502944249165,
+      "learning_rate": 9.672802491444962e-06,
+      "loss": 0.4727,
+      "step": 1742
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.839881108249043,
+      "learning_rate": 9.672331321415109e-06,
+      "loss": 0.8002,
+      "step": 1743
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.267489571195822,
+      "learning_rate": 9.671859823875166e-06,
+      "loss": 0.6568,
+      "step": 1744
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.02003689875668,
+      "learning_rate": 9.671387998858178e-06,
+      "loss": 1.0033,
+      "step": 1745
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.334547483328481,
+      "learning_rate": 9.670915846397224e-06,
+      "loss": 1.0469,
+      "step": 1746
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.894229760592162,
+      "learning_rate": 9.670443366525396e-06,
+      "loss": 0.5112,
+      "step": 1747
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.335584294099776,
+      "learning_rate": 9.669970559275814e-06,
+      "loss": 1.3517,
+      "step": 1748
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.3717056903934015,
+      "learning_rate": 9.66949742468162e-06,
+      "loss": 0.7089,
+      "step": 1749
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.627215878970121,
+      "learning_rate": 9.669023962775976e-06,
+      "loss": 0.7062,
+      "step": 1750
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.289061741779872,
+      "learning_rate": 9.668550173592075e-06,
+      "loss": 0.8946,
+      "step": 1751
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.224035246700839,
+      "learning_rate": 9.668076057163122e-06,
+      "loss": 0.8385,
+      "step": 1752
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.429426490400277,
+      "learning_rate": 9.667601613522355e-06,
+      "loss": 0.8882,
+      "step": 1753
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.511297438492241,
+      "learning_rate": 9.667126842703027e-06,
+      "loss": 0.2118,
+      "step": 1754
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.094554570312316,
+      "learning_rate": 9.666651744738417e-06,
+      "loss": 0.7532,
+      "step": 1755
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.468245134556565,
+      "learning_rate": 9.666176319661833e-06,
+      "loss": 0.8349,
+      "step": 1756
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.9396583629732396,
+      "learning_rate": 9.665700567506594e-06,
+      "loss": 1.0694,
+      "step": 1757
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.172716573001196,
+      "learning_rate": 9.66522448830605e-06,
+      "loss": 1.1011,
+      "step": 1758
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.2948059164102803,
+      "learning_rate": 9.664748082093573e-06,
+      "loss": 0.84,
+      "step": 1759
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.6774929634029596,
+      "learning_rate": 9.664271348902558e-06,
+      "loss": 0.3558,
+      "step": 1760
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.293291501424655,
+      "learning_rate": 9.66379428876642e-06,
+      "loss": 1.0833,
+      "step": 1761
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.8630268243168615,
+      "learning_rate": 9.663316901718599e-06,
+      "loss": 1.419,
+      "step": 1762
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.144353130832779,
+      "learning_rate": 9.662839187792556e-06,
+      "loss": 0.5933,
+      "step": 1763
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.4750146714606323,
+      "learning_rate": 9.66236114702178e-06,
+      "loss": 0.888,
+      "step": 1764
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.391849057094597,
+      "learning_rate": 9.661882779439778e-06,
+      "loss": 0.5401,
+      "step": 1765
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.330429068652904,
+      "learning_rate": 9.661404085080082e-06,
+      "loss": 0.47,
+      "step": 1766
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.802510374591477,
+      "learning_rate": 9.660925063976247e-06,
+      "loss": 1.0848,
+      "step": 1767
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.8303554655188647,
+      "learning_rate": 9.660445716161849e-06,
+      "loss": 1.1645,
+      "step": 1768
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.332359385121991,
+      "learning_rate": 9.659966041670488e-06,
+      "loss": 1.3822,
+      "step": 1769
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.49826770270501,
+      "learning_rate": 9.659486040535788e-06,
+      "loss": 0.75,
+      "step": 1770
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.625414454564575,
+      "learning_rate": 9.659005712791394e-06,
+      "loss": 1.0039,
+      "step": 1771
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.5687891069349362,
+      "learning_rate": 9.658525058470977e-06,
+      "loss": 0.7928,
+      "step": 1772
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.0856531814552612,
+      "learning_rate": 9.658044077608227e-06,
+      "loss": 0.9806,
+      "step": 1773
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0481731703439325,
+      "learning_rate": 9.657562770236857e-06,
+      "loss": 0.3284,
+      "step": 1774
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.477448597774813,
+      "learning_rate": 9.65708113639061e-06,
+      "loss": 0.684,
+      "step": 1775
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.786223392707992,
+      "learning_rate": 9.656599176103241e-06,
+      "loss": 1.5706,
+      "step": 1776
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.4299279077369706,
+      "learning_rate": 9.656116889408536e-06,
+      "loss": 0.9167,
+      "step": 1777
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.845132154284454,
+      "learning_rate": 9.6556342763403e-06,
+      "loss": 0.6764,
+      "step": 1778
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.990118489197662,
+      "learning_rate": 9.655151336932362e-06,
+      "loss": 1.2797,
+      "step": 1779
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.883998603403723,
+      "learning_rate": 9.654668071218576e-06,
+      "loss": 0.5424,
+      "step": 1780
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.976081723771014,
+      "learning_rate": 9.654184479232815e-06,
+      "loss": 1.2058,
+      "step": 1781
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.612574873209907,
+      "learning_rate": 9.653700561008976e-06,
+      "loss": 1.0735,
+      "step": 1782
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.153755087512503,
+      "learning_rate": 9.653216316580983e-06,
+      "loss": 0.9932,
+      "step": 1783
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.250713760253444,
+      "learning_rate": 9.652731745982775e-06,
+      "loss": 1.0884,
+      "step": 1784
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.5887055213448864,
+      "learning_rate": 9.652246849248321e-06,
+      "loss": 0.9022,
+      "step": 1785
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.760788781957178,
+      "learning_rate": 9.651761626411608e-06,
+      "loss": 1.2023,
+      "step": 1786
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.00386777303116,
+      "learning_rate": 9.651276077506653e-06,
+      "loss": 0.6427,
+      "step": 1787
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.734213588146607,
+      "learning_rate": 9.650790202567484e-06,
+      "loss": 1.0691,
+      "step": 1788
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.153949352710521,
+      "learning_rate": 9.650304001628164e-06,
+      "loss": 1.1805,
+      "step": 1789
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.59076553094134,
+      "learning_rate": 9.649817474722772e-06,
+      "loss": 1.0221,
+      "step": 1790
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.26331738370133,
+      "learning_rate": 9.64933062188541e-06,
+      "loss": 0.7572,
+      "step": 1791
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.920409775213098,
+      "learning_rate": 9.648843443150203e-06,
+      "loss": 0.727,
+      "step": 1792
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.005900979192892,
+      "learning_rate": 9.648355938551307e-06,
+      "loss": 0.5756,
+      "step": 1793
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.8375256998421667,
+      "learning_rate": 9.647868108122887e-06,
+      "loss": 0.9936,
+      "step": 1794
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.9040876808362346,
+      "learning_rate": 9.64737995189914e-06,
+      "loss": 0.7834,
+      "step": 1795
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.041394694735733,
+      "learning_rate": 9.646891469914285e-06,
+      "loss": 0.8196,
+      "step": 1796
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.082808833427863,
+      "learning_rate": 9.646402662202562e-06,
+      "loss": 0.8898,
+      "step": 1797
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8436244812636793,
+      "learning_rate": 9.645913528798231e-06,
+      "loss": 0.4136,
+      "step": 1798
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.493558306583951,
+      "learning_rate": 9.645424069735582e-06,
+      "loss": 1.0796,
+      "step": 1799
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.195389897526096,
+      "learning_rate": 9.644934285048924e-06,
+      "loss": 0.8139,
+      "step": 1800
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.203540340494399,
+      "learning_rate": 9.644444174772586e-06,
+      "loss": 0.9193,
+      "step": 1801
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.9164410427299736,
+      "learning_rate": 9.643953738940926e-06,
+      "loss": 0.6488,
+      "step": 1802
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.565239558464744,
+      "learning_rate": 9.643462977588315e-06,
+      "loss": 0.9911,
+      "step": 1803
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.6880077819346773,
+      "learning_rate": 9.642971890749163e-06,
+      "loss": 0.8051,
+      "step": 1804
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.401719903746576,
+      "learning_rate": 9.642480478457883e-06,
+      "loss": 0.7522,
+      "step": 1805
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.275589998252871,
+      "learning_rate": 9.64198874074893e-06,
+      "loss": 0.9793,
+      "step": 1806
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.114652840279528,
+      "learning_rate": 9.641496677656766e-06,
+      "loss": 0.9217,
+      "step": 1807
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.0759207295978834,
+      "learning_rate": 9.641004289215884e-06,
+      "loss": 0.4599,
+      "step": 1808
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.237292245483329,
+      "learning_rate": 9.6405115754608e-06,
+      "loss": 0.6787,
+      "step": 1809
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.9325714220187704,
+      "learning_rate": 9.64001853642605e-06,
+      "loss": 0.708,
+      "step": 1810
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.4734493440378533,
+      "learning_rate": 9.639525172146194e-06,
+      "loss": 0.8007,
+      "step": 1811
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3238979248148512,
+      "learning_rate": 9.639031482655814e-06,
+      "loss": 0.1702,
+      "step": 1812
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.147979100312514,
+      "learning_rate": 9.638537467989517e-06,
+      "loss": 0.832,
+      "step": 1813
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.5306368535026618,
+      "learning_rate": 9.63804312818193e-06,
+      "loss": 0.7108,
+      "step": 1814
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.354011197916151,
+      "learning_rate": 9.637548463267705e-06,
+      "loss": 0.6388,
+      "step": 1815
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.3059057159340535,
+      "learning_rate": 9.637053473281517e-06,
+      "loss": 0.5239,
+      "step": 1816
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.2267324783997355,
+      "learning_rate": 9.636558158258057e-06,
+      "loss": 0.6933,
+      "step": 1817
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7180296092373988,
+      "learning_rate": 9.636062518232052e-06,
+      "loss": 0.3191,
+      "step": 1818
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.910215876952668,
+      "learning_rate": 9.63556655323824e-06,
+      "loss": 0.6652,
+      "step": 1819
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.5030198200823057,
+      "learning_rate": 9.635070263311386e-06,
+      "loss": 0.5872,
+      "step": 1820
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.245126438609335,
+      "learning_rate": 9.634573648486278e-06,
+      "loss": 0.8985,
+      "step": 1821
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.960095535317417,
+      "learning_rate": 9.634076708797728e-06,
+      "loss": 0.6347,
+      "step": 1822
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.189854565011895,
+      "learning_rate": 9.633579444280568e-06,
+      "loss": 0.7024,
+      "step": 1823
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.554196019272498,
+      "learning_rate": 9.633081854969655e-06,
+      "loss": 0.7781,
+      "step": 1824
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.304378751681649,
+      "learning_rate": 9.632583940899866e-06,
+      "loss": 1.4871,
+      "step": 1825
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.481617498792958,
+      "learning_rate": 9.632085702106104e-06,
+      "loss": 0.6789,
+      "step": 1826
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.4928801318769964,
+      "learning_rate": 9.631587138623295e-06,
+      "loss": 0.4831,
+      "step": 1827
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.0021844643847864,
+      "learning_rate": 9.631088250486383e-06,
+      "loss": 0.4137,
+      "step": 1828
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.4877779902389694,
+      "learning_rate": 9.630589037730338e-06,
+      "loss": 0.4901,
+      "step": 1829
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.6749886358582238,
+      "learning_rate": 9.630089500390154e-06,
+      "loss": 0.9204,
+      "step": 1830
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.7402582348689775,
+      "learning_rate": 9.629589638500849e-06,
+      "loss": 1.2322,
+      "step": 1831
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.4426550739107515,
+      "learning_rate": 9.629089452097455e-06,
+      "loss": 0.7329,
+      "step": 1832
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.6734999505376194,
+      "learning_rate": 9.628588941215037e-06,
+      "loss": 0.7417,
+      "step": 1833
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.942397568879872,
+      "learning_rate": 9.62808810588868e-06,
+      "loss": 0.9746,
+      "step": 1834
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.175304781579661,
+      "learning_rate": 9.627586946153487e-06,
+      "loss": 0.5455,
+      "step": 1835
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.486209116943911,
+      "learning_rate": 9.62708546204459e-06,
+      "loss": 0.4457,
+      "step": 1836
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.699947940635257,
+      "learning_rate": 9.626583653597136e-06,
+      "loss": 0.7648,
+      "step": 1837
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.563861331354613,
+      "learning_rate": 9.626081520846304e-06,
+      "loss": 0.9206,
+      "step": 1838
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.9746637208939366,
+      "learning_rate": 9.625579063827291e-06,
+      "loss": 0.9338,
+      "step": 1839
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.993335491725338,
+      "learning_rate": 9.625076282575317e-06,
+      "loss": 0.7197,
+      "step": 1840
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.7751650590973203,
+      "learning_rate": 9.624573177125623e-06,
+      "loss": 0.6854,
+      "step": 1841
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.6287066239927293,
+      "learning_rate": 9.624069747513474e-06,
+      "loss": 0.8383,
+      "step": 1842
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.466341283669408,
+      "learning_rate": 9.623565993774161e-06,
+      "loss": 1.379,
+      "step": 1843
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.9565044600451404,
+      "learning_rate": 9.623061915942994e-06,
+      "loss": 0.8075,
+      "step": 1844
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.353500384494615,
+      "learning_rate": 9.622557514055306e-06,
+      "loss": 0.8647,
+      "step": 1845
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.7802807636965956,
+      "learning_rate": 9.622052788146454e-06,
+      "loss": 0.7,
+      "step": 1846
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.7243261490814894,
+      "learning_rate": 9.621547738251816e-06,
+      "loss": 0.6569,
+      "step": 1847
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.5979273143338655,
+      "learning_rate": 9.621042364406796e-06,
+      "loss": 0.6535,
+      "step": 1848
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.4155454455842635,
+      "learning_rate": 9.620536666646816e-06,
+      "loss": 0.5346,
+      "step": 1849
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.346796067656894,
+      "learning_rate": 9.620030645007324e-06,
+      "loss": 1.1118,
+      "step": 1850
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.065844455634222,
+      "learning_rate": 9.61952429952379e-06,
+      "loss": 0.4093,
+      "step": 1851
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0422151702096683,
+      "learning_rate": 9.619017630231709e-06,
+      "loss": 0.3572,
+      "step": 1852
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.384252750841941,
+      "learning_rate": 9.618510637166591e-06,
+      "loss": 0.6136,
+      "step": 1853
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.710009165195783,
+      "learning_rate": 9.61800332036398e-06,
+      "loss": 0.9828,
+      "step": 1854
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.274612438483947,
+      "learning_rate": 9.617495679859432e-06,
+      "loss": 0.3935,
+      "step": 1855
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.685079377159412,
+      "learning_rate": 9.616987715688534e-06,
+      "loss": 1.2142,
+      "step": 1856
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.7327329404216423,
+      "learning_rate": 9.616479427886889e-06,
+      "loss": 1.0473,
+      "step": 1857
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.633754836253732,
+      "learning_rate": 9.615970816490127e-06,
+      "loss": 1.088,
+      "step": 1858
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.6267774185656303,
+      "learning_rate": 9.6154618815339e-06,
+      "loss": 0.3738,
+      "step": 1859
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.683276366996444,
+      "learning_rate": 9.61495262305388e-06,
+      "loss": 0.7404,
+      "step": 1860
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.4397339031584555,
+      "learning_rate": 9.614443041085768e-06,
+      "loss": 0.5179,
+      "step": 1861
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.080227084199175,
+      "learning_rate": 9.613933135665281e-06,
+      "loss": 0.8409,
+      "step": 1862
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.720888044583893,
+      "learning_rate": 9.613422906828161e-06,
+      "loss": 0.5713,
+      "step": 1863
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.605411201655714,
+      "learning_rate": 9.61291235461017e-06,
+      "loss": 1.1279,
+      "step": 1864
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.6743922503335473,
+      "learning_rate": 9.612401479047102e-06,
+      "loss": 0.5532,
+      "step": 1865
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.458010962051488,
+      "learning_rate": 9.611890280174761e-06,
+      "loss": 0.4904,
+      "step": 1866
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.590096886998366,
+      "learning_rate": 9.611378758028984e-06,
+      "loss": 0.5157,
+      "step": 1867
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.786726498871613,
+      "learning_rate": 9.610866912645624e-06,
+      "loss": 0.9906,
+      "step": 1868
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.520165208333526,
+      "learning_rate": 9.61035474406056e-06,
+      "loss": 0.49,
+      "step": 1869
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.2529485159125437,
+      "learning_rate": 9.609842252309694e-06,
+      "loss": 0.9277,
+      "step": 1870
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.702654732411725,
+      "learning_rate": 9.609329437428946e-06,
+      "loss": 0.6127,
+      "step": 1871
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.388100944422168,
+      "learning_rate": 9.608816299454267e-06,
+      "loss": 0.4555,
+      "step": 1872
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.987626503763048,
+      "learning_rate": 9.608302838421622e-06,
+      "loss": 1.1882,
+      "step": 1873
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.92680809173823,
+      "learning_rate": 9.607789054367e-06,
+      "loss": 1.6455,
+      "step": 1874
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.933180856469748,
+      "learning_rate": 9.607274947326423e-06,
+      "loss": 0.2325,
+      "step": 1875
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.7454435218935003,
+      "learning_rate": 9.606760517335923e-06,
+      "loss": 0.6886,
+      "step": 1876
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.7601473733544006,
+      "learning_rate": 9.606245764431557e-06,
+      "loss": 1.0919,
+      "step": 1877
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.963578919378046,
+      "learning_rate": 9.605730688649411e-06,
+      "loss": 0.4018,
+      "step": 1878
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.094419034446812,
+      "learning_rate": 9.605215290025587e-06,
+      "loss": 0.8573,
+      "step": 1879
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.232518157536914,
+      "learning_rate": 9.604699568596211e-06,
+      "loss": 1.1854,
+      "step": 1880
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.330752275162666,
+      "learning_rate": 9.604183524397439e-06,
+      "loss": 1.1544,
+      "step": 1881
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.738236543702836,
+      "learning_rate": 9.603667157465436e-06,
+      "loss": 1.2742,
+      "step": 1882
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.500833585139346,
+      "learning_rate": 9.603150467836403e-06,
+      "loss": 1.0557,
+      "step": 1883
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.348349354005987,
+      "learning_rate": 9.602633455546556e-06,
+      "loss": 0.8161,
+      "step": 1884
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.345114650968165,
+      "learning_rate": 9.602116120632133e-06,
+      "loss": 0.9003,
+      "step": 1885
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.535743444943094,
+      "learning_rate": 9.601598463129398e-06,
+      "loss": 1.003,
+      "step": 1886
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.9347754032019444,
+      "learning_rate": 9.601080483074637e-06,
+      "loss": 0.7201,
+      "step": 1887
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.512700028513524,
+      "learning_rate": 9.600562180504158e-06,
+      "loss": 0.6555,
+      "step": 1888
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.831953765720359,
+      "learning_rate": 9.600043555454292e-06,
+      "loss": 1.2459,
+      "step": 1889
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.908436927960402,
+      "learning_rate": 9.599524607961394e-06,
+      "loss": 1.0104,
+      "step": 1890
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.103109745089075,
+      "learning_rate": 9.599005338061835e-06,
+      "loss": 1.1343,
+      "step": 1891
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.718322584210091,
+      "learning_rate": 9.598485745792019e-06,
+      "loss": 0.8079,
+      "step": 1892
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.160552069027906,
+      "learning_rate": 9.597965831188365e-06,
+      "loss": 0.8904,
+      "step": 1893
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.893236221583511,
+      "learning_rate": 9.597445594287315e-06,
+      "loss": 0.8749,
+      "step": 1894
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.8989511257441816,
+      "learning_rate": 9.596925035125338e-06,
+      "loss": 0.8513,
+      "step": 1895
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.502391681117759,
+      "learning_rate": 9.596404153738922e-06,
+      "loss": 1.2835,
+      "step": 1896
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5096032919902738,
+      "learning_rate": 9.595882950164579e-06,
+      "loss": 0.2231,
+      "step": 1897
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.3940828284046365,
+      "learning_rate": 9.595361424438841e-06,
+      "loss": 0.8311,
+      "step": 1898
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.1565144371734566,
+      "learning_rate": 9.594839576598267e-06,
+      "loss": 0.7746,
+      "step": 1899
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.861681899339718,
+      "learning_rate": 9.594317406679436e-06,
+      "loss": 1.4082,
+      "step": 1900
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.3701934090314865,
+      "learning_rate": 9.593794914718948e-06,
+      "loss": 1.0056,
+      "step": 1901
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4361844931507608,
+      "learning_rate": 9.593272100753431e-06,
+      "loss": 0.183,
+      "step": 1902
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.341261409056521,
+      "learning_rate": 9.592748964819528e-06,
+      "loss": 0.9966,
+      "step": 1903
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.246802133865648,
+      "learning_rate": 9.592225506953911e-06,
+      "loss": 0.8687,
+      "step": 1904
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.536665860148006,
+      "learning_rate": 9.591701727193272e-06,
+      "loss": 0.4399,
+      "step": 1905
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.3123258067487007,
+      "learning_rate": 9.591177625574322e-06,
+      "loss": 0.4896,
+      "step": 1906
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.273373552745985,
+      "learning_rate": 9.590653202133804e-06,
+      "loss": 0.9065,
+      "step": 1907
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.014798967567046,
+      "learning_rate": 9.590128456908476e-06,
+      "loss": 0.96,
+      "step": 1908
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.4186995200654637,
+      "learning_rate": 9.589603389935118e-06,
+      "loss": 0.5375,
+      "step": 1909
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.0198594591627925,
+      "learning_rate": 9.589078001250537e-06,
+      "loss": 0.6707,
+      "step": 1910
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.1364010070572155,
+      "learning_rate": 9.588552290891557e-06,
+      "loss": 1.1784,
+      "step": 1911
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.158358601291325,
+      "learning_rate": 9.588026258895034e-06,
+      "loss": 0.9459,
+      "step": 1912
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.444787446353252,
+      "learning_rate": 9.587499905297836e-06,
+      "loss": 1.0501,
+      "step": 1913
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.357310424227119,
+      "learning_rate": 9.58697323013686e-06,
+      "loss": 0.906,
+      "step": 1914
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.340254000033742,
+      "learning_rate": 9.586446233449024e-06,
+      "loss": 0.539,
+      "step": 1915
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3421768612808074,
+      "learning_rate": 9.585918915271267e-06,
+      "loss": 0.1963,
+      "step": 1916
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.200976251872486,
+      "learning_rate": 9.58539127564055e-06,
+      "loss": 0.9611,
+      "step": 1917
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.14363328588231,
+      "learning_rate": 9.584863314593862e-06,
+      "loss": 1.1128,
+      "step": 1918
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.767959520821083,
+      "learning_rate": 9.584335032168209e-06,
+      "loss": 1.2102,
+      "step": 1919
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.0603362286037,
+      "learning_rate": 9.58380642840062e-06,
+      "loss": 0.8753,
+      "step": 1920
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.784097276200648,
+      "learning_rate": 9.583277503328152e-06,
+      "loss": 0.6181,
+      "step": 1921
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.3628590333982915,
+      "learning_rate": 9.582748256987875e-06,
+      "loss": 0.3996,
+      "step": 1922
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.1992436604635115,
+      "learning_rate": 9.58221868941689e-06,
+      "loss": 0.8356,
+      "step": 1923
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.5690561422258256,
+      "learning_rate": 9.581688800652317e-06,
+      "loss": 0.5881,
+      "step": 1924
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.9962006736267157,
+      "learning_rate": 9.581158590731298e-06,
+      "loss": 0.6964,
+      "step": 1925
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.838643693846644,
+      "learning_rate": 9.580628059691e-06,
+      "loss": 1.1755,
+      "step": 1926
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.645525539871668,
+      "learning_rate": 9.58009720756861e-06,
+      "loss": 1.3661,
+      "step": 1927
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.369531010137947,
+      "learning_rate": 9.57956603440134e-06,
+      "loss": 1.1669,
+      "step": 1928
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0261167417220345,
+      "learning_rate": 9.579034540226417e-06,
+      "loss": 0.4221,
+      "step": 1929
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.805961619998852,
+      "learning_rate": 9.578502725081105e-06,
+      "loss": 0.4036,
+      "step": 1930
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.7343364076028704,
+      "learning_rate": 9.577970589002674e-06,
+      "loss": 0.8297,
+      "step": 1931
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.115812274867747,
+      "learning_rate": 9.577438132028431e-06,
+      "loss": 0.5971,
+      "step": 1932
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.013164357851665,
+      "learning_rate": 9.576905354195695e-06,
+      "loss": 0.7369,
+      "step": 1933
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.044032284551135,
+      "learning_rate": 9.576372255541812e-06,
+      "loss": 1.0083,
+      "step": 1934
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.314036059653204,
+      "learning_rate": 9.575838836104152e-06,
+      "loss": 1.1176,
+      "step": 1935
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.9222150730303875,
+      "learning_rate": 9.575305095920101e-06,
+      "loss": 0.8157,
+      "step": 1936
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.5652621046635025,
+      "learning_rate": 9.574771035027077e-06,
+      "loss": 0.826,
+      "step": 1937
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.9630771554858146,
+      "learning_rate": 9.574236653462511e-06,
+      "loss": 0.4971,
+      "step": 1938
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.412336714047085,
+      "learning_rate": 9.573701951263862e-06,
+      "loss": 1.0939,
+      "step": 1939
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.6333530332442,
+      "learning_rate": 9.573166928468615e-06,
+      "loss": 1.4336,
+      "step": 1940
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.62204995393872,
+      "learning_rate": 9.572631585114267e-06,
+      "loss": 1.1644,
+      "step": 1941
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.1738069100273223,
+      "learning_rate": 9.572095921238343e-06,
+      "loss": 0.7047,
+      "step": 1942
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.8525663295063906,
+      "learning_rate": 9.571559936878394e-06,
+      "loss": 0.9256,
+      "step": 1943
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4944576959377527,
+      "learning_rate": 9.571023632071989e-06,
+      "loss": 0.2304,
+      "step": 1944
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.812644322997221,
+      "learning_rate": 9.570487006856722e-06,
+      "loss": 0.6744,
+      "step": 1945
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.1507359526587972,
+      "learning_rate": 9.569950061270204e-06,
+      "loss": 0.7059,
+      "step": 1946
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.624674249680327,
+      "learning_rate": 9.569412795350076e-06,
+      "loss": 0.6104,
+      "step": 1947
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.31256529328677,
+      "learning_rate": 9.568875209133999e-06,
+      "loss": 1.0901,
+      "step": 1948
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.861101663997485,
+      "learning_rate": 9.568337302659652e-06,
+      "loss": 0.7332,
+      "step": 1949
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.252967963194855,
+      "learning_rate": 9.567799075964743e-06,
+      "loss": 0.473,
+      "step": 1950
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.7954437726383152,
+      "learning_rate": 9.567260529086997e-06,
+      "loss": 0.8397,
+      "step": 1951
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0570487471770336,
+      "learning_rate": 9.566721662064164e-06,
+      "loss": 0.4776,
+      "step": 1952
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.598324280894434,
+      "learning_rate": 9.566182474934017e-06,
+      "loss": 1.0364,
+      "step": 1953
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.9893068037392205,
+      "learning_rate": 9.565642967734351e-06,
+      "loss": 0.8069,
+      "step": 1954
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.4689637147814354,
+      "learning_rate": 9.565103140502982e-06,
+      "loss": 0.4132,
+      "step": 1955
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.9962889936616355,
+      "learning_rate": 9.564562993277752e-06,
+      "loss": 1.2551,
+      "step": 1956
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.037837016283461,
+      "learning_rate": 9.564022526096521e-06,
+      "loss": 0.6973,
+      "step": 1957
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.623565841563198,
+      "learning_rate": 9.563481738997172e-06,
+      "loss": 1.2694,
+      "step": 1958
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.7142809741906966,
+      "learning_rate": 9.562940632017614e-06,
+      "loss": 0.565,
+      "step": 1959
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.5410216749125705,
+      "learning_rate": 9.562399205195775e-06,
+      "loss": 0.6941,
+      "step": 1960
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.9151566880983317,
+      "learning_rate": 9.56185745856961e-06,
+      "loss": 0.9364,
+      "step": 1961
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.682734437236148,
+      "learning_rate": 9.561315392177089e-06,
+      "loss": 0.767,
+      "step": 1962
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.6613759123423164,
+      "learning_rate": 9.56077300605621e-06,
+      "loss": 1.0541,
+      "step": 1963
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.110581249062556,
+      "learning_rate": 9.56023030024499e-06,
+      "loss": 0.8277,
+      "step": 1964
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.288487454511914,
+      "learning_rate": 9.559687274781475e-06,
+      "loss": 0.768,
+      "step": 1965
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.8223670341511733,
+      "learning_rate": 9.559143929703724e-06,
+      "loss": 0.9807,
+      "step": 1966
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.108908693851089,
+      "learning_rate": 9.558600265049825e-06,
+      "loss": 0.8094,
+      "step": 1967
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.630304005106419,
+      "learning_rate": 9.558056280857887e-06,
+      "loss": 0.7759,
+      "step": 1968
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.8460453116763738,
+      "learning_rate": 9.55751197716604e-06,
+      "loss": 0.8335,
+      "step": 1969
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.365489981874306,
+      "learning_rate": 9.556967354012438e-06,
+      "loss": 0.8814,
+      "step": 1970
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.194935912677049,
+      "learning_rate": 9.556422411435257e-06,
+      "loss": 1.0257,
+      "step": 1971
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.9574399155708813,
+      "learning_rate": 9.555877149472695e-06,
+      "loss": 0.8627,
+      "step": 1972
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.040316351809699,
+      "learning_rate": 9.555331568162972e-06,
+      "loss": 0.8235,
+      "step": 1973
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.273221383752521,
+      "learning_rate": 9.554785667544329e-06,
+      "loss": 0.6658,
+      "step": 1974
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.044878071923383,
+      "learning_rate": 9.554239447655036e-06,
+      "loss": 0.9372,
+      "step": 1975
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 7.246557282396511,
+      "learning_rate": 9.553692908533375e-06,
+      "loss": 1.4103,
+      "step": 1976
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.274490525562516,
+      "learning_rate": 9.55314605021766e-06,
+      "loss": 1.3265,
+      "step": 1977
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.784351877267063,
+      "learning_rate": 9.552598872746223e-06,
+      "loss": 0.7399,
+      "step": 1978
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.510740649222768,
+      "learning_rate": 9.552051376157417e-06,
+      "loss": 1.1919,
+      "step": 1979
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.682551788596266,
+      "learning_rate": 9.55150356048962e-06,
+      "loss": 1.2566,
+      "step": 1980
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.308435854041583,
+      "learning_rate": 9.550955425781234e-06,
+      "loss": 0.5514,
+      "step": 1981
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.5546470124450935,
+      "learning_rate": 9.550406972070676e-06,
+      "loss": 0.7517,
+      "step": 1982
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.847516559239803,
+      "learning_rate": 9.549858199396394e-06,
+      "loss": 0.8813,
+      "step": 1983
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.2468457030579745,
+      "learning_rate": 9.549309107796852e-06,
+      "loss": 0.9317,
+      "step": 1984
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.7988168500030888,
+      "learning_rate": 9.54875969731054e-06,
+      "loss": 0.5256,
+      "step": 1985
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.034857576281372,
+      "learning_rate": 9.54820996797597e-06,
+      "loss": 0.9549,
+      "step": 1986
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.8790834699851295,
+      "learning_rate": 9.547659919831676e-06,
+      "loss": 1.6624,
+      "step": 1987
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.05241404605833,
+      "learning_rate": 9.547109552916211e-06,
+      "loss": 1.2909,
+      "step": 1988
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.9117588210126377,
+      "learning_rate": 9.546558867268159e-06,
+      "loss": 0.6867,
+      "step": 1989
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.309118950701456,
+      "learning_rate": 9.546007862926115e-06,
+      "loss": 0.6384,
+      "step": 1990
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.116069205074591,
+      "learning_rate": 9.545456539928704e-06,
+      "loss": 1.443,
+      "step": 1991
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.9087098585219024,
+      "learning_rate": 9.544904898314572e-06,
+      "loss": 1.0374,
+      "step": 1992
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.9309719792696702,
+      "learning_rate": 9.544352938122384e-06,
+      "loss": 0.8559,
+      "step": 1993
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.469895891148484,
+      "learning_rate": 9.543800659390835e-06,
+      "loss": 0.8993,
+      "step": 1994
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.278454789623266,
+      "learning_rate": 9.543248062158632e-06,
+      "loss": 1.0097,
+      "step": 1995
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.529235977103577,
+      "learning_rate": 9.542695146464513e-06,
+      "loss": 1.0865,
+      "step": 1996
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.169795889365141,
+      "learning_rate": 9.542141912347236e-06,
+      "loss": 1.018,
+      "step": 1997
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.51563326069593,
+      "learning_rate": 9.541588359845575e-06,
+      "loss": 1.0636,
+      "step": 1998
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.479455931358969,
+      "learning_rate": 9.541034488998338e-06,
+      "loss": 0.9901,
+      "step": 1999
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.372716248240536,
+      "learning_rate": 9.540480299844345e-06,
+      "loss": 0.6926,
+      "step": 2000
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.747085234992996,
+      "learning_rate": 9.539925792422443e-06,
+      "loss": 1.0609,
+      "step": 2001
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.269631520809824,
+      "learning_rate": 9.5393709667715e-06,
+      "loss": 0.209,
+      "step": 2002
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.8841818602939,
+      "learning_rate": 9.53881582293041e-06,
+      "loss": 0.8058,
+      "step": 2003
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.405695660762923,
+      "learning_rate": 9.538260360938081e-06,
+      "loss": 0.8038,
+      "step": 2004
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.771381211989904,
+      "learning_rate": 9.537704580833453e-06,
+      "loss": 0.4068,
+      "step": 2005
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.26310212288867,
+      "learning_rate": 9.537148482655482e-06,
+      "loss": 0.9612,
+      "step": 2006
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.976865740042055,
+      "learning_rate": 9.536592066443147e-06,
+      "loss": 1.0408,
+      "step": 2007
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.71800834789916,
+      "learning_rate": 9.536035332235454e-06,
+      "loss": 0.9944,
+      "step": 2008
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.622149280558416,
+      "learning_rate": 9.535478280071422e-06,
+      "loss": 0.9308,
+      "step": 2009
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 6.105878039691027,
+      "learning_rate": 9.534920909990101e-06,
+      "loss": 1.3476,
+      "step": 2010
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.166972114310374,
+      "learning_rate": 9.534363222030563e-06,
+      "loss": 0.8512,
+      "step": 2011
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.628377013586853,
+      "learning_rate": 9.533805216231894e-06,
+      "loss": 0.9158,
+      "step": 2012
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.980529070798636,
+      "learning_rate": 9.533246892633213e-06,
+      "loss": 0.7445,
+      "step": 2013
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.4338879751047973,
+      "learning_rate": 9.532688251273654e-06,
+      "loss": 0.5876,
+      "step": 2014
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.2946132643396164,
+      "learning_rate": 9.532129292192373e-06,
+      "loss": 0.7603,
+      "step": 2015
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.900269383522834,
+      "learning_rate": 9.531570015428553e-06,
+      "loss": 0.7026,
+      "step": 2016
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.469620357210813,
+      "learning_rate": 9.531010421021396e-06,
+      "loss": 1.2024,
+      "step": 2017
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.3814539431713895,
+      "learning_rate": 9.530450509010128e-06,
+      "loss": 1.1572,
+      "step": 2018
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.214817152693931,
+      "learning_rate": 9.529890279433995e-06,
+      "loss": 0.72,
+      "step": 2019
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.520489716408855,
+      "learning_rate": 9.529329732332268e-06,
+      "loss": 0.3485,
+      "step": 2020
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.803091814303413,
+      "learning_rate": 9.52876886774424e-06,
+      "loss": 1.1041,
+      "step": 2021
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.6421407248221,
+      "learning_rate": 9.528207685709221e-06,
+      "loss": 1.2171,
+      "step": 2022
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.23653445974646,
+      "learning_rate": 9.52764618626655e-06,
+      "loss": 1.1003,
+      "step": 2023
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.885493299123106,
+      "learning_rate": 9.527084369455589e-06,
+      "loss": 1.0119,
+      "step": 2024
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.731147655154812,
+      "learning_rate": 9.526522235315713e-06,
+      "loss": 1.1316,
+      "step": 2025
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.8303744896202905,
+      "learning_rate": 9.525959783886329e-06,
+      "loss": 0.7168,
+      "step": 2026
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.79528547028082,
+      "learning_rate": 9.525397015206861e-06,
+      "loss": 0.8626,
+      "step": 2027
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.403137139314899,
+      "learning_rate": 9.524833929316758e-06,
+      "loss": 0.7673,
+      "step": 2028
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.632791389070294,
+      "learning_rate": 9.524270526255486e-06,
+      "loss": 1.6007,
+      "step": 2029
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.910686152455089,
+      "learning_rate": 9.523706806062541e-06,
+      "loss": 1.383,
+      "step": 2030
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.2941777100821295,
+      "learning_rate": 9.523142768777435e-06,
+      "loss": 1.4508,
+      "step": 2031
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.2257729295585085,
+      "learning_rate": 9.522578414439707e-06,
+      "loss": 0.7555,
+      "step": 2032
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.177167550722004,
+      "learning_rate": 9.522013743088916e-06,
+      "loss": 0.9737,
+      "step": 2033
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.878369880410306,
+      "learning_rate": 9.52144875476464e-06,
+      "loss": 1.3448,
+      "step": 2034
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.307328908557558,
+      "learning_rate": 9.520883449506483e-06,
+      "loss": 1.0423,
+      "step": 2035
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.7938378285887997,
+      "learning_rate": 9.52031782735407e-06,
+      "loss": 0.5762,
+      "step": 2036
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.410318130606425,
+      "learning_rate": 9.519751888347053e-06,
+      "loss": 0.9626,
+      "step": 2037
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.689113064551304,
+      "learning_rate": 9.519185632525097e-06,
+      "loss": 1.0112,
+      "step": 2038
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.771765049703681,
+      "learning_rate": 9.518619059927895e-06,
+      "loss": 0.4313,
+      "step": 2039
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.524515596196058,
+      "learning_rate": 9.518052170595165e-06,
+      "loss": 0.9909,
+      "step": 2040
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.1098921131187756,
+      "learning_rate": 9.517484964566637e-06,
+      "loss": 0.4388,
+      "step": 2041
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.840815725983443,
+      "learning_rate": 9.516917441882074e-06,
+      "loss": 0.8934,
+      "step": 2042
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.737988001113275,
+      "learning_rate": 9.516349602581256e-06,
+      "loss": 0.9319,
+      "step": 2043
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.233006946267547,
+      "learning_rate": 9.515781446703988e-06,
+      "loss": 0.9586,
+      "step": 2044
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.338332469127597,
+      "learning_rate": 9.51521297429009e-06,
+      "loss": 1.1064,
+      "step": 2045
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.181487637050818,
+      "learning_rate": 9.514644185379416e-06,
+      "loss": 0.3214,
+      "step": 2046
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.820616144216788,
+      "learning_rate": 9.51407508001183e-06,
+      "loss": 1.0967,
+      "step": 2047
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.3995605186361595,
+      "learning_rate": 9.51350565822723e-06,
+      "loss": 0.46,
+      "step": 2048
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.189027154695834,
+      "learning_rate": 9.512935920065523e-06,
+      "loss": 1.0905,
+      "step": 2049
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.208339630062005,
+      "learning_rate": 9.512365865566648e-06,
+      "loss": 0.7666,
+      "step": 2050
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.44626051074097,
+      "learning_rate": 9.511795494770563e-06,
+      "loss": 0.9391,
+      "step": 2051
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.014287763227923,
+      "learning_rate": 9.511224807717253e-06,
+      "loss": 0.4553,
+      "step": 2052
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.5272840701887276,
+      "learning_rate": 9.510653804446714e-06,
+      "loss": 0.98,
+      "step": 2053
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.5804379646472637,
+      "learning_rate": 9.510082484998975e-06,
+      "loss": 0.5724,
+      "step": 2054
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.0664109603373664,
+      "learning_rate": 9.50951084941408e-06,
+      "loss": 0.777,
+      "step": 2055
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.9548121149971385,
+      "learning_rate": 9.508938897732101e-06,
+      "loss": 1.1975,
+      "step": 2056
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.704230560766375,
+      "learning_rate": 9.508366629993129e-06,
+      "loss": 0.8394,
+      "step": 2057
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.0182372497951966,
+      "learning_rate": 9.507794046237275e-06,
+      "loss": 0.4947,
+      "step": 2058
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.3287807453200715,
+      "learning_rate": 9.507221146504679e-06,
+      "loss": 0.5632,
+      "step": 2059
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.986043890091539,
+      "learning_rate": 9.506647930835494e-06,
+      "loss": 1.1423,
+      "step": 2060
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.958680811383781,
+      "learning_rate": 9.506074399269902e-06,
+      "loss": 1.0541,
+      "step": 2061
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.1924867873842455,
+      "learning_rate": 9.505500551848105e-06,
+      "loss": 0.7383,
+      "step": 2062
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.192658274030753,
+      "learning_rate": 9.50492638861033e-06,
+      "loss": 0.4288,
+      "step": 2063
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.0884543370481534,
+      "learning_rate": 9.504351909596818e-06,
+      "loss": 0.6757,
+      "step": 2064
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.299401261090491,
+      "learning_rate": 9.503777114847841e-06,
+      "loss": 0.5118,
+      "step": 2065
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9253161949191624,
+      "learning_rate": 9.503202004403688e-06,
+      "loss": 0.4369,
+      "step": 2066
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.028626104070394,
+      "learning_rate": 9.502626578304673e-06,
+      "loss": 0.7785,
+      "step": 2067
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.1122624328146475,
+      "learning_rate": 9.50205083659113e-06,
+      "loss": 0.7224,
+      "step": 2068
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.2818337720691417,
+      "learning_rate": 9.501474779303416e-06,
+      "loss": 0.6204,
+      "step": 2069
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.771811140932043,
+      "learning_rate": 9.500898406481911e-06,
+      "loss": 1.1285,
+      "step": 2070
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.8057931128703726,
+      "learning_rate": 9.500321718167017e-06,
+      "loss": 0.8784,
+      "step": 2071
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.5515639183361767,
+      "learning_rate": 9.499744714399155e-06,
+      "loss": 0.899,
+      "step": 2072
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.122862374078744,
+      "learning_rate": 9.499167395218772e-06,
+      "loss": 0.9491,
+      "step": 2073
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.7491156377139907,
+      "learning_rate": 9.498589760666333e-06,
+      "loss": 1.0864,
+      "step": 2074
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.448237666496974,
+      "learning_rate": 9.498011810782332e-06,
+      "loss": 0.7363,
+      "step": 2075
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.865420210192231,
+      "learning_rate": 9.497433545607278e-06,
+      "loss": 0.6435,
+      "step": 2076
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.8746356846623153,
+      "learning_rate": 9.496854965181705e-06,
+      "loss": 1.1468,
+      "step": 2077
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.349451863770051,
+      "learning_rate": 9.496276069546169e-06,
+      "loss": 0.998,
+      "step": 2078
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.444977958151714,
+      "learning_rate": 9.495696858741249e-06,
+      "loss": 0.9305,
+      "step": 2079
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.643233439483295,
+      "learning_rate": 9.495117332807542e-06,
+      "loss": 0.9184,
+      "step": 2080
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.447022066674226,
+      "learning_rate": 9.494537491785676e-06,
+      "loss": 0.7931,
+      "step": 2081
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.9125196189107183,
+      "learning_rate": 9.493957335716291e-06,
+      "loss": 0.9897,
+      "step": 2082
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 6.285592320779023,
+      "learning_rate": 9.493376864640054e-06,
+      "loss": 1.3372,
+      "step": 2083
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.0882900515413283,
+      "learning_rate": 9.492796078597655e-06,
+      "loss": 0.527,
+      "step": 2084
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.573554407494425,
+      "learning_rate": 9.492214977629804e-06,
+      "loss": 0.9002,
+      "step": 2085
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.834869538459952,
+      "learning_rate": 9.491633561777232e-06,
+      "loss": 1.1973,
+      "step": 2086
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.591982689006695,
+      "learning_rate": 9.491051831080695e-06,
+      "loss": 1.2376,
+      "step": 2087
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.410080452072863,
+      "learning_rate": 9.490469785580971e-06,
+      "loss": 0.77,
+      "step": 2088
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.4696812104534214,
+      "learning_rate": 9.489887425318856e-06,
+      "loss": 0.8371,
+      "step": 2089
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.6395549603734527,
+      "learning_rate": 9.489304750335173e-06,
+      "loss": 0.8124,
+      "step": 2090
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.3541758421205494,
+      "learning_rate": 9.488721760670766e-06,
+      "loss": 0.7443,
+      "step": 2091
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.157586521603289,
+      "learning_rate": 9.488138456366497e-06,
+      "loss": 0.681,
+      "step": 2092
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.117454253879517,
+      "learning_rate": 9.487554837463255e-06,
+      "loss": 0.4913,
+      "step": 2093
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 6.380817613738605,
+      "learning_rate": 9.48697090400195e-06,
+      "loss": 1.092,
+      "step": 2094
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.4747229179837382,
+      "learning_rate": 9.486386656023509e-06,
+      "loss": 0.5649,
+      "step": 2095
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.3807338815393715,
+      "learning_rate": 9.485802093568892e-06,
+      "loss": 0.7537,
+      "step": 2096
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.1771176947496484,
+      "learning_rate": 9.485217216679068e-06,
+      "loss": 0.9096,
+      "step": 2097
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.282633156917091,
+      "learning_rate": 9.484632025395037e-06,
+      "loss": 0.9372,
+      "step": 2098
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.095661517054039,
+      "learning_rate": 9.48404651975782e-06,
+      "loss": 1.414,
+      "step": 2099
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.529039630511247,
+      "learning_rate": 9.483460699808454e-06,
+      "loss": 1.2644,
+      "step": 2100
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.867831198743964,
+      "learning_rate": 9.482874565588008e-06,
+      "loss": 1.2301,
+      "step": 2101
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.908320279178014,
+      "learning_rate": 9.482288117137561e-06,
+      "loss": 0.5719,
+      "step": 2102
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.22276681186549,
+      "learning_rate": 9.481701354498227e-06,
+      "loss": 0.7356,
+      "step": 2103
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.8726519472165974,
+      "learning_rate": 9.48111427771113e-06,
+      "loss": 1.0238,
+      "step": 2104
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.545740432284491,
+      "learning_rate": 9.480526886817425e-06,
+      "loss": 0.5745,
+      "step": 2105
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.525664610056473,
+      "learning_rate": 9.479939181858286e-06,
+      "loss": 0.841,
+      "step": 2106
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.119819181857357,
+      "learning_rate": 9.479351162874905e-06,
+      "loss": 1.0244,
+      "step": 2107
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.480209901923949,
+      "learning_rate": 9.478762829908503e-06,
+      "loss": 1.0946,
+      "step": 2108
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.06730290748614,
+      "learning_rate": 9.478174183000319e-06,
+      "loss": 0.9159,
+      "step": 2109
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.50710093602591,
+      "learning_rate": 9.477585222191612e-06,
+      "loss": 1.0762,
+      "step": 2110
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.821870746791981,
+      "learning_rate": 9.476995947523668e-06,
+      "loss": 1.1686,
+      "step": 2111
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.4036271881923335,
+      "learning_rate": 9.476406359037792e-06,
+      "loss": 0.6,
+      "step": 2112
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.135499123303508,
+      "learning_rate": 9.475816456775313e-06,
+      "loss": 0.8012,
+      "step": 2113
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.674933703337597,
+      "learning_rate": 9.475226240777577e-06,
+      "loss": 1.1691,
+      "step": 2114
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 6.075280978313644,
+      "learning_rate": 9.47463571108596e-06,
+      "loss": 1.0191,
+      "step": 2115
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.4166952874689573,
+      "learning_rate": 9.474044867741852e-06,
+      "loss": 0.5438,
+      "step": 2116
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.991601282845236,
+      "learning_rate": 9.47345371078667e-06,
+      "loss": 1.1868,
+      "step": 2117
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.387621946592111,
+      "learning_rate": 9.47286224026185e-06,
+      "loss": 0.9359,
+      "step": 2118
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.120682905250976,
+      "learning_rate": 9.472270456208856e-06,
+      "loss": 0.682,
+      "step": 2119
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.9720525101079796,
+      "learning_rate": 9.471678358669164e-06,
+      "loss": 0.6188,
+      "step": 2120
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.7381808096544877,
+      "learning_rate": 9.47108594768428e-06,
+      "loss": 0.8853,
+      "step": 2121
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.8959848975983835,
+      "learning_rate": 9.47049322329573e-06,
+      "loss": 0.9168,
+      "step": 2122
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2368095730765394,
+      "learning_rate": 9.469900185545061e-06,
+      "loss": 0.1946,
+      "step": 2123
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.286467425215324,
+      "learning_rate": 9.469306834473841e-06,
+      "loss": 0.7813,
+      "step": 2124
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.914048061704354,
+      "learning_rate": 9.468713170123664e-06,
+      "loss": 0.4727,
+      "step": 2125
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.035570216006248,
+      "learning_rate": 9.46811919253614e-06,
+      "loss": 0.4793,
+      "step": 2126
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.383252038597668,
+      "learning_rate": 9.467524901752906e-06,
+      "loss": 1.0153,
+      "step": 2127
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.9088761095690225,
+      "learning_rate": 9.466930297815622e-06,
+      "loss": 1.2265,
+      "step": 2128
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.659464819329504,
+      "learning_rate": 9.46633538076596e-06,
+      "loss": 0.8317,
+      "step": 2129
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.897382976866003,
+      "learning_rate": 9.465740150645629e-06,
+      "loss": 0.6749,
+      "step": 2130
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.286261516991291,
+      "learning_rate": 9.465144607496347e-06,
+      "loss": 0.9856,
+      "step": 2131
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.8718066416881376,
+      "learning_rate": 9.46454875135986e-06,
+      "loss": 0.9202,
+      "step": 2132
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.9295301699594405,
+      "learning_rate": 9.463952582277936e-06,
+      "loss": 0.637,
+      "step": 2133
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.7624787002724838,
+      "learning_rate": 9.463356100292363e-06,
+      "loss": 0.6039,
+      "step": 2134
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.45137275432468,
+      "learning_rate": 9.462759305444951e-06,
+      "loss": 1.5769,
+      "step": 2135
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.043486054594081,
+      "learning_rate": 9.462162197777533e-06,
+      "loss": 0.8245,
+      "step": 2136
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.4508470611822823,
+      "learning_rate": 9.461564777331966e-06,
+      "loss": 0.6273,
+      "step": 2137
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.550464396618127,
+      "learning_rate": 9.460967044150125e-06,
+      "loss": 0.4785,
+      "step": 2138
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.220157500165512,
+      "learning_rate": 9.460368998273908e-06,
+      "loss": 0.8692,
+      "step": 2139
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.451446644164049,
+      "learning_rate": 9.459770639745235e-06,
+      "loss": 0.8494,
+      "step": 2140
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.649836832615487,
+      "learning_rate": 9.459171968606051e-06,
+      "loss": 1.0236,
+      "step": 2141
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.986127517748761,
+      "learning_rate": 9.458572984898318e-06,
+      "loss": 0.5701,
+      "step": 2142
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.979887171050888,
+      "learning_rate": 9.457973688664021e-06,
+      "loss": 0.865,
+      "step": 2143
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2761285963995945,
+      "learning_rate": 9.45737407994517e-06,
+      "loss": 0.2562,
+      "step": 2144
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.330361544442108,
+      "learning_rate": 9.456774158783795e-06,
+      "loss": 1.2154,
+      "step": 2145
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.612898726359513,
+      "learning_rate": 9.456173925221948e-06,
+      "loss": 0.7982,
+      "step": 2146
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.966569355397138,
+      "learning_rate": 9.455573379301702e-06,
+      "loss": 1.5651,
+      "step": 2147
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.4979505753209525,
+      "learning_rate": 9.45497252106515e-06,
+      "loss": 0.3068,
+      "step": 2148
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8643227063850487,
+      "learning_rate": 9.454371350554417e-06,
+      "loss": 0.1983,
+      "step": 2149
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.8998501733971542,
+      "learning_rate": 9.453769867811636e-06,
+      "loss": 0.7906,
+      "step": 2150
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.334813471461509,
+      "learning_rate": 9.45316807287897e-06,
+      "loss": 0.5067,
+      "step": 2151
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.6296558987705527,
+      "learning_rate": 9.4525659657986e-06,
+      "loss": 0.4986,
+      "step": 2152
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.9692009490803257,
+      "learning_rate": 9.451963546612737e-06,
+      "loss": 0.9403,
+      "step": 2153
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.384935016423835,
+      "learning_rate": 9.451360815363601e-06,
+      "loss": 0.9806,
+      "step": 2154
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.7487390901287174,
+      "learning_rate": 9.450757772093447e-06,
+      "loss": 0.6419,
+      "step": 2155
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.7973741872239026,
+      "learning_rate": 9.450154416844543e-06,
+      "loss": 0.7477,
+      "step": 2156
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.490270775900813,
+      "learning_rate": 9.44955074965918e-06,
+      "loss": 1.148,
+      "step": 2157
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.100769635829848,
+      "learning_rate": 9.448946770579675e-06,
+      "loss": 0.9803,
+      "step": 2158
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.324000798766789,
+      "learning_rate": 9.448342479648362e-06,
+      "loss": 0.6454,
+      "step": 2159
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.757169125636011,
+      "learning_rate": 9.447737876907602e-06,
+      "loss": 1.2184,
+      "step": 2160
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.7057613162293004,
+      "learning_rate": 9.447132962399772e-06,
+      "loss": 0.8668,
+      "step": 2161
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.352280279037213,
+      "learning_rate": 9.446527736167277e-06,
+      "loss": 0.3339,
+      "step": 2162
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.4131389889547674,
+      "learning_rate": 9.445922198252538e-06,
+      "loss": 0.8509,
+      "step": 2163
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.1302691719793336,
+      "learning_rate": 9.445316348698002e-06,
+      "loss": 0.9048,
+      "step": 2164
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.184514928247789,
+      "learning_rate": 9.444710187546136e-06,
+      "loss": 1.2745,
+      "step": 2165
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.3023991330152445,
+      "learning_rate": 9.444103714839427e-06,
+      "loss": 0.8257,
+      "step": 2166
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.4559356624211106,
+      "learning_rate": 9.443496930620392e-06,
+      "loss": 0.7127,
+      "step": 2167
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.0009784342285326,
+      "learning_rate": 9.442889834931558e-06,
+      "loss": 0.5808,
+      "step": 2168
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.271613359566975,
+      "learning_rate": 9.442282427815483e-06,
+      "loss": 0.8843,
+      "step": 2169
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.768172855722143,
+      "learning_rate": 9.441674709314743e-06,
+      "loss": 1.0149,
+      "step": 2170
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.393967273531713,
+      "learning_rate": 9.441066679471935e-06,
+      "loss": 0.5938,
+      "step": 2171
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.7071192673095257,
+      "learning_rate": 9.440458338329681e-06,
+      "loss": 0.5606,
+      "step": 2172
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.286003794630861,
+      "learning_rate": 9.439849685930623e-06,
+      "loss": 0.9005,
+      "step": 2173
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.8373323976564766,
+      "learning_rate": 9.439240722317423e-06,
+      "loss": 0.4884,
+      "step": 2174
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.8342604024352474,
+      "learning_rate": 9.43863144753277e-06,
+      "loss": 0.6117,
+      "step": 2175
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.5346372794679,
+      "learning_rate": 9.438021861619367e-06,
+      "loss": 0.8834,
+      "step": 2176
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.630996103152386,
+      "learning_rate": 9.437411964619947e-06,
+      "loss": 0.8089,
+      "step": 2177
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.55704600360632,
+      "learning_rate": 9.43680175657726e-06,
+      "loss": 0.8158,
+      "step": 2178
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.547727769767378,
+      "learning_rate": 9.43619123753408e-06,
+      "loss": 1.5137,
+      "step": 2179
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.6197319803485417,
+      "learning_rate": 9.435580407533198e-06,
+      "loss": 0.4518,
+      "step": 2180
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.264578431627168,
+      "learning_rate": 9.434969266617436e-06,
+      "loss": 1.4073,
+      "step": 2181
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.085161116879681,
+      "learning_rate": 9.434357814829627e-06,
+      "loss": 0.7944,
+      "step": 2182
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.8200278738459086,
+      "learning_rate": 9.433746052212636e-06,
+      "loss": 0.8745,
+      "step": 2183
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.073425916795223,
+      "learning_rate": 9.43313397880934e-06,
+      "loss": 0.9598,
+      "step": 2184
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.8512311184225396,
+      "learning_rate": 9.432521594662648e-06,
+      "loss": 0.7672,
+      "step": 2185
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.9337318171462963,
+      "learning_rate": 9.431908899815479e-06,
+      "loss": 0.8608,
+      "step": 2186
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8885296919716454,
+      "learning_rate": 9.431295894310786e-06,
+      "loss": 0.4857,
+      "step": 2187
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.903393735231428,
+      "learning_rate": 9.430682578191537e-06,
+      "loss": 0.6505,
+      "step": 2188
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6887859434219867,
+      "learning_rate": 9.43006895150072e-06,
+      "loss": 0.336,
+      "step": 2189
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.153251818217103,
+      "learning_rate": 9.429455014281349e-06,
+      "loss": 1.4917,
+      "step": 2190
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.706061847628713,
+      "learning_rate": 9.428840766576459e-06,
+      "loss": 1.0367,
+      "step": 2191
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.457412294800602,
+      "learning_rate": 9.428226208429106e-06,
+      "loss": 1.1445,
+      "step": 2192
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.555600596305112,
+      "learning_rate": 9.427611339882368e-06,
+      "loss": 1.197,
+      "step": 2193
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.005253708947643,
+      "learning_rate": 9.426996160979342e-06,
+      "loss": 1.0339,
+      "step": 2194
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.977210166365334,
+      "learning_rate": 9.426380671763154e-06,
+      "loss": 1.0448,
+      "step": 2195
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.1991074719897274,
+      "learning_rate": 9.425764872276942e-06,
+      "loss": 0.5701,
+      "step": 2196
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.1006524749382405,
+      "learning_rate": 9.425148762563876e-06,
+      "loss": 0.9353,
+      "step": 2197
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.332035032248812,
+      "learning_rate": 9.424532342667138e-06,
+      "loss": 0.7351,
+      "step": 2198
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.6185299233730506,
+      "learning_rate": 9.42391561262994e-06,
+      "loss": 0.7285,
+      "step": 2199
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.5599192483980033,
+      "learning_rate": 9.42329857249551e-06,
+      "loss": 0.4709,
+      "step": 2200
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.312042537898623,
+      "learning_rate": 9.422681222307099e-06,
+      "loss": 0.9185,
+      "step": 2201
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.6335923426745067,
+      "learning_rate": 9.422063562107984e-06,
+      "loss": 0.8953,
+      "step": 2202
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 6.034931329079666,
+      "learning_rate": 9.421445591941456e-06,
+      "loss": 1.3278,
+      "step": 2203
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.7279775684316347,
+      "learning_rate": 9.420827311850836e-06,
+      "loss": 0.5268,
+      "step": 2204
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.636644980441379,
+      "learning_rate": 9.42020872187946e-06,
+      "loss": 0.543,
+      "step": 2205
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.8513882635748193,
+      "learning_rate": 9.41958982207069e-06,
+      "loss": 1.1868,
+      "step": 2206
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.599020465113644,
+      "learning_rate": 9.418970612467908e-06,
+      "loss": 0.5247,
+      "step": 2207
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.247518479537959,
+      "learning_rate": 9.418351093114517e-06,
+      "loss": 0.343,
+      "step": 2208
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.2366428056676835,
+      "learning_rate": 9.417731264053942e-06,
+      "loss": 1.409,
+      "step": 2209
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.424511385285495,
+      "learning_rate": 9.417111125329633e-06,
+      "loss": 0.891,
+      "step": 2210
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.854324920321562,
+      "learning_rate": 9.416490676985057e-06,
+      "loss": 0.57,
+      "step": 2211
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.814197414087323,
+      "learning_rate": 9.415869919063705e-06,
+      "loss": 0.957,
+      "step": 2212
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.1892540112475123,
+      "learning_rate": 9.415248851609089e-06,
+      "loss": 0.6602,
+      "step": 2213
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 6.012278285033315,
+      "learning_rate": 9.414627474664745e-06,
+      "loss": 1.3654,
+      "step": 2214
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.5956857349837175,
+      "learning_rate": 9.414005788274226e-06,
+      "loss": 0.7299,
+      "step": 2215
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.8920306479101443,
+      "learning_rate": 9.413383792481112e-06,
+      "loss": 0.5877,
+      "step": 2216
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.961917764585591,
+      "learning_rate": 9.412761487329e-06,
+      "loss": 0.3996,
+      "step": 2217
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.622972061370353,
+      "learning_rate": 9.412138872861514e-06,
+      "loss": 0.7335,
+      "step": 2218
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.8898421548794957,
+      "learning_rate": 9.411515949122295e-06,
+      "loss": 0.7544,
+      "step": 2219
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.542825331149399,
+      "learning_rate": 9.410892716155006e-06,
+      "loss": 1.1682,
+      "step": 2220
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.240458429700037,
+      "learning_rate": 9.410269174003333e-06,
+      "loss": 1.1765,
+      "step": 2221
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.717643764221059,
+      "learning_rate": 9.409645322710985e-06,
+      "loss": 1.0527,
+      "step": 2222
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.61581799462886,
+      "learning_rate": 9.409021162321692e-06,
+      "loss": 1.2727,
+      "step": 2223
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.902406104492035,
+      "learning_rate": 9.408396692879202e-06,
+      "loss": 1.1395,
+      "step": 2224
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.042305305793964,
+      "learning_rate": 9.40777191442729e-06,
+      "loss": 0.859,
+      "step": 2225
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0774316749315587,
+      "learning_rate": 9.40714682700975e-06,
+      "loss": 0.165,
+      "step": 2226
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.302788281417149,
+      "learning_rate": 9.406521430670397e-06,
+      "loss": 0.8008,
+      "step": 2227
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.7846222730904553,
+      "learning_rate": 9.405895725453069e-06,
+      "loss": 0.7245,
+      "step": 2228
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.0857113770735145,
+      "learning_rate": 9.405269711401625e-06,
+      "loss": 0.8109,
+      "step": 2229
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9188680752732636,
+      "learning_rate": 9.404643388559945e-06,
+      "loss": 0.3626,
+      "step": 2230
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.1849555018977895,
+      "learning_rate": 9.404016756971934e-06,
+      "loss": 0.9315,
+      "step": 2231
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.7608332507637434,
+      "learning_rate": 9.403389816681514e-06,
+      "loss": 0.7762,
+      "step": 2232
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.9655397836011237,
+      "learning_rate": 9.402762567732632e-06,
+      "loss": 0.9902,
+      "step": 2233
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.1218309956038155,
+      "learning_rate": 9.402135010169254e-06,
+      "loss": 0.5033,
+      "step": 2234
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.366445790921517,
+      "learning_rate": 9.401507144035371e-06,
+      "loss": 0.222,
+      "step": 2235
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.903752718477064,
+      "learning_rate": 9.400878969374991e-06,
+      "loss": 0.7701,
+      "step": 2236
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.751144292776653,
+      "learning_rate": 9.40025048623215e-06,
+      "loss": 0.7907,
+      "step": 2237
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.825728154383655,
+      "learning_rate": 9.399621694650898e-06,
+      "loss": 1.3445,
+      "step": 2238
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.068955811651021,
+      "learning_rate": 9.398992594675314e-06,
+      "loss": 0.751,
+      "step": 2239
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.658416537459373,
+      "learning_rate": 9.398363186349493e-06,
+      "loss": 1.0305,
+      "step": 2240
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.329549125699289,
+      "learning_rate": 9.397733469717554e-06,
+      "loss": 1.2354,
+      "step": 2241
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.069811292045692,
+      "learning_rate": 9.397103444823638e-06,
+      "loss": 0.5039,
+      "step": 2242
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.574464473522762,
+      "learning_rate": 9.396473111711908e-06,
+      "loss": 1.2865,
+      "step": 2243
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.1815009387994815,
+      "learning_rate": 9.395842470426545e-06,
+      "loss": 0.3662,
+      "step": 2244
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.596516254714554,
+      "learning_rate": 9.395211521011756e-06,
+      "loss": 1.0371,
+      "step": 2245
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.278370333633337,
+      "learning_rate": 9.394580263511765e-06,
+      "loss": 0.9215,
+      "step": 2246
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.5781826380201975,
+      "learning_rate": 9.393948697970827e-06,
+      "loss": 0.6672,
+      "step": 2247
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.692094398353298,
+      "learning_rate": 9.393316824433204e-06,
+      "loss": 0.6842,
+      "step": 2248
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.611932153208114,
+      "learning_rate": 9.392684642943195e-06,
+      "loss": 0.9484,
+      "step": 2249
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.1271003219389075,
+      "learning_rate": 9.392052153545108e-06,
+      "loss": 0.6765,
+      "step": 2250
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.2203921452166915,
+      "learning_rate": 9.391419356283277e-06,
+      "loss": 1.0756,
+      "step": 2251
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.809427884781176,
+      "learning_rate": 9.390786251202064e-06,
+      "loss": 1.1822,
+      "step": 2252
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.7567251671685584,
+      "learning_rate": 9.390152838345842e-06,
+      "loss": 0.6625,
+      "step": 2253
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.694270323963662,
+      "learning_rate": 9.389519117759012e-06,
+      "loss": 1.3287,
+      "step": 2254
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.555233870715096,
+      "learning_rate": 9.388885089485995e-06,
+      "loss": 0.9989,
+      "step": 2255
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.243426813783271,
+      "learning_rate": 9.388250753571235e-06,
+      "loss": 0.8727,
+      "step": 2256
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1269155707380953,
+      "learning_rate": 9.387616110059194e-06,
+      "loss": 0.1509,
+      "step": 2257
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.1492173789558664,
+      "learning_rate": 9.386981158994359e-06,
+      "loss": 0.8127,
+      "step": 2258
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.778123070227287,
+      "learning_rate": 9.386345900421236e-06,
+      "loss": 0.6753,
+      "step": 2259
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.126349253955989,
+      "learning_rate": 9.385710334384357e-06,
+      "loss": 0.3218,
+      "step": 2260
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.1294988384899938,
+      "learning_rate": 9.385074460928267e-06,
+      "loss": 0.4438,
+      "step": 2261
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.9584109911097376,
+      "learning_rate": 9.384438280097543e-06,
+      "loss": 0.8727,
+      "step": 2262
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.5563394754276163,
+      "learning_rate": 9.383801791936777e-06,
+      "loss": 0.6102,
+      "step": 2263
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.360677635526274,
+      "learning_rate": 9.383164996490583e-06,
+      "loss": 0.4224,
+      "step": 2264
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.006397861225834,
+      "learning_rate": 9.382527893803599e-06,
+      "loss": 1.1627,
+      "step": 2265
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.465562048660857,
+      "learning_rate": 9.381890483920482e-06,
+      "loss": 0.5598,
+      "step": 2266
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.389436228854544,
+      "learning_rate": 9.381252766885914e-06,
+      "loss": 1.3719,
+      "step": 2267
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.650189768224478,
+      "learning_rate": 9.380614742744592e-06,
+      "loss": 1.073,
+      "step": 2268
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.940624357871344,
+      "learning_rate": 9.379976411541241e-06,
+      "loss": 1.0487,
+      "step": 2269
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.0691136895443245,
+      "learning_rate": 9.379337773320604e-06,
+      "loss": 0.9991,
+      "step": 2270
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.671113323218031,
+      "learning_rate": 9.378698828127451e-06,
+      "loss": 0.7108,
+      "step": 2271
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.094792417803656,
+      "learning_rate": 9.378059576006567e-06,
+      "loss": 0.6131,
+      "step": 2272
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.097438813925161,
+      "learning_rate": 9.377420017002756e-06,
+      "loss": 1.0312,
+      "step": 2273
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.892784583139424,
+      "learning_rate": 9.376780151160856e-06,
+      "loss": 1.0126,
+      "step": 2274
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.8083537051345573,
+      "learning_rate": 9.376139978525713e-06,
+      "loss": 0.9419,
+      "step": 2275
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.641312816729417,
+      "learning_rate": 9.375499499142204e-06,
+      "loss": 0.8278,
+      "step": 2276
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.562235260630101,
+      "learning_rate": 9.374858713055221e-06,
+      "loss": 1.0192,
+      "step": 2277
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2153746964649463,
+      "learning_rate": 9.374217620309684e-06,
+      "loss": 0.2192,
+      "step": 2278
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.6305760677433625,
+      "learning_rate": 9.373576220950527e-06,
+      "loss": 0.452,
+      "step": 2279
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.230674908813215,
+      "learning_rate": 9.37293451502271e-06,
+      "loss": 1.0268,
+      "step": 2280
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.652602389587261,
+      "learning_rate": 9.372292502571217e-06,
+      "loss": 1.3541,
+      "step": 2281
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5384697420557394,
+      "learning_rate": 9.371650183641046e-06,
+      "loss": 0.2499,
+      "step": 2282
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.051187199744434,
+      "learning_rate": 9.371007558277221e-06,
+      "loss": 1.079,
+      "step": 2283
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.6611226742779315,
+      "learning_rate": 9.370364626524791e-06,
+      "loss": 0.5498,
+      "step": 2284
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.9110461646193837,
+      "learning_rate": 9.36972138842882e-06,
+      "loss": 0.9785,
+      "step": 2285
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.815306567877642,
+      "learning_rate": 9.369077844034398e-06,
+      "loss": 1.5498,
+      "step": 2286
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.8516878045698872,
+      "learning_rate": 9.368433993386632e-06,
+      "loss": 0.7308,
+      "step": 2287
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.22649733569183,
+      "learning_rate": 9.367789836530655e-06,
+      "loss": 1.3369,
+      "step": 2288
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.188081935922202,
+      "learning_rate": 9.36714537351162e-06,
+      "loss": 1.3739,
+      "step": 2289
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.836611550160533,
+      "learning_rate": 9.366500604374699e-06,
+      "loss": 1.4048,
+      "step": 2290
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.68458182019333,
+      "learning_rate": 9.365855529165089e-06,
+      "loss": 0.8023,
+      "step": 2291
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.937118676656062,
+      "learning_rate": 9.365210147928006e-06,
+      "loss": 0.607,
+      "step": 2292
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.0959864170466402,
+      "learning_rate": 9.364564460708689e-06,
+      "loss": 0.5608,
+      "step": 2293
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.279504397544929,
+      "learning_rate": 9.3639184675524e-06,
+      "loss": 0.8709,
+      "step": 2294
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.814079321161493,
+      "learning_rate": 9.363272168504417e-06,
+      "loss": 0.4674,
+      "step": 2295
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.61620991804245,
+      "learning_rate": 9.362625563610044e-06,
+      "loss": 1.1514,
+      "step": 2296
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.758909968343346,
+      "learning_rate": 9.361978652914605e-06,
+      "loss": 0.8233,
+      "step": 2297
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.2385771274175483,
+      "learning_rate": 9.361331436463446e-06,
+      "loss": 0.9019,
+      "step": 2298
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.896629749654269,
+      "learning_rate": 9.360683914301934e-06,
+      "loss": 1.5023,
+      "step": 2299
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.312029356565346,
+      "learning_rate": 9.360036086475457e-06,
+      "loss": 1.4424,
+      "step": 2300
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.150712940690037,
+      "learning_rate": 9.359387953029425e-06,
+      "loss": 0.8739,
+      "step": 2301
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.363421170051736,
+      "learning_rate": 9.358739514009271e-06,
+      "loss": 0.9312,
+      "step": 2302
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.79247579777736,
+      "learning_rate": 9.358090769460446e-06,
+      "loss": 0.598,
+      "step": 2303
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.185708394371197,
+      "learning_rate": 9.357441719428423e-06,
+      "loss": 0.5437,
+      "step": 2304
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.29604390462049,
+      "learning_rate": 9.3567923639587e-06,
+      "loss": 0.872,
+      "step": 2305
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.621954214424412,
+      "learning_rate": 9.356142703096793e-06,
+      "loss": 1.0147,
+      "step": 2306
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.776210916732992,
+      "learning_rate": 9.355492736888242e-06,
+      "loss": 0.8019,
+      "step": 2307
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.737184950318045,
+      "learning_rate": 9.354842465378604e-06,
+      "loss": 1.0243,
+      "step": 2308
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.7661736156524657,
+      "learning_rate": 9.354191888613462e-06,
+      "loss": 0.9402,
+      "step": 2309
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.3941009941822626,
+      "learning_rate": 9.353541006638417e-06,
+      "loss": 0.7118,
+      "step": 2310
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.008294593308233,
+      "learning_rate": 9.352889819499096e-06,
+      "loss": 1.2041,
+      "step": 2311
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.497801080060148,
+      "learning_rate": 9.35223832724114e-06,
+      "loss": 1.1584,
+      "step": 2312
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.5206679515786554,
+      "learning_rate": 9.35158652991022e-06,
+      "loss": 1.1447,
+      "step": 2313
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.118739250240711,
+      "learning_rate": 9.350934427552023e-06,
+      "loss": 1.476,
+      "step": 2314
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.5796514622825963,
+      "learning_rate": 9.350282020212256e-06,
+      "loss": 0.5229,
+      "step": 2315
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.519285126458806,
+      "learning_rate": 9.349629307936653e-06,
+      "loss": 1.039,
+      "step": 2316
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.409487761550421,
+      "learning_rate": 9.348976290770965e-06,
+      "loss": 1.0798,
+      "step": 2317
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.566681383204464,
+      "learning_rate": 9.348322968760965e-06,
+      "loss": 0.8468,
+      "step": 2318
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.167042571796452,
+      "learning_rate": 9.34766934195245e-06,
+      "loss": 0.7639,
+      "step": 2319
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.0381343904833593,
+      "learning_rate": 9.347015410391235e-06,
+      "loss": 0.3901,
+      "step": 2320
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.5957839378494025,
+      "learning_rate": 9.34636117412316e-06,
+      "loss": 1.4191,
+      "step": 2321
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.3712381446725423,
+      "learning_rate": 9.345706633194078e-06,
+      "loss": 0.6917,
+      "step": 2322
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.525764739475181,
+      "learning_rate": 9.345051787649877e-06,
+      "loss": 0.6715,
+      "step": 2323
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.4365910876302572,
+      "learning_rate": 9.344396637536453e-06,
+      "loss": 0.8326,
+      "step": 2324
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.244381507411723,
+      "learning_rate": 9.343741182899733e-06,
+      "loss": 1.1729,
+      "step": 2325
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.6725681990302252,
+      "learning_rate": 9.34308542378566e-06,
+      "loss": 0.8226,
+      "step": 2326
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.8662307678886147,
+      "learning_rate": 9.3424293602402e-06,
+      "loss": 0.6593,
+      "step": 2327
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.578416090268908,
+      "learning_rate": 9.34177299230934e-06,
+      "loss": 0.7326,
+      "step": 2328
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.1837032657349225,
+      "learning_rate": 9.341116320039088e-06,
+      "loss": 0.9397,
+      "step": 2329
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.3667517386706205,
+      "learning_rate": 9.340459343475475e-06,
+      "loss": 1.2654,
+      "step": 2330
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.3000469622278645,
+      "learning_rate": 9.339802062664553e-06,
+      "loss": 1.3045,
+      "step": 2331
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.2957218695500545,
+      "learning_rate": 9.339144477652391e-06,
+      "loss": 1.0857,
+      "step": 2332
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.68864267751406,
+      "learning_rate": 9.338486588485087e-06,
+      "loss": 1.3811,
+      "step": 2333
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.372649789054265,
+      "learning_rate": 9.337828395208755e-06,
+      "loss": 0.5209,
+      "step": 2334
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.8499680904005222,
+      "learning_rate": 9.337169897869528e-06,
+      "loss": 0.8476,
+      "step": 2335
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.961242137201628,
+      "learning_rate": 9.336511096513568e-06,
+      "loss": 1.528,
+      "step": 2336
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.693661806290446,
+      "learning_rate": 9.335851991187053e-06,
+      "loss": 0.873,
+      "step": 2337
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.508153442359315,
+      "learning_rate": 9.335192581936183e-06,
+      "loss": 1.0223,
+      "step": 2338
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.4103159948747335,
+      "learning_rate": 9.334532868807179e-06,
+      "loss": 1.2319,
+      "step": 2339
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.19185231767955,
+      "learning_rate": 9.333872851846285e-06,
+      "loss": 1.0721,
+      "step": 2340
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.111140255013692,
+      "learning_rate": 9.333212531099767e-06,
+      "loss": 0.9359,
+      "step": 2341
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.5008772177759786,
+      "learning_rate": 9.332551906613908e-06,
+      "loss": 0.4058,
+      "step": 2342
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.1415338749073785,
+      "learning_rate": 9.331890978435014e-06,
+      "loss": 1.0631,
+      "step": 2343
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.80184772378031,
+      "learning_rate": 9.331229746609416e-06,
+      "loss": 0.6319,
+      "step": 2344
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.3313134792011,
+      "learning_rate": 9.330568211183461e-06,
+      "loss": 0.4148,
+      "step": 2345
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.914408637597205,
+      "learning_rate": 9.329906372203523e-06,
+      "loss": 0.4193,
+      "step": 2346
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.765654467492976,
+      "learning_rate": 9.329244229715992e-06,
+      "loss": 0.4101,
+      "step": 2347
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.550335647297338,
+      "learning_rate": 9.328581783767281e-06,
+      "loss": 0.6818,
+      "step": 2348
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.9500036638624088,
+      "learning_rate": 9.327919034403825e-06,
+      "loss": 0.3719,
+      "step": 2349
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.501275834383856,
+      "learning_rate": 9.327255981672082e-06,
+      "loss": 1.395,
+      "step": 2350
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.279179026210534,
+      "learning_rate": 9.326592625618523e-06,
+      "loss": 1.2283,
+      "step": 2351
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.262990927460758,
+      "learning_rate": 9.325928966289652e-06,
+      "loss": 1.0392,
+      "step": 2352
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.8054019782508375,
+      "learning_rate": 9.325265003731988e-06,
+      "loss": 0.5625,
+      "step": 2353
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.065618563471102,
+      "learning_rate": 9.324600737992069e-06,
+      "loss": 0.6746,
+      "step": 2354
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.9175758949154034,
+      "learning_rate": 9.323936169116461e-06,
+      "loss": 0.8539,
+      "step": 2355
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.403739334261655,
+      "learning_rate": 9.323271297151743e-06,
+      "loss": 1.1677,
+      "step": 2356
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.28527399597012,
+      "learning_rate": 9.322606122144524e-06,
+      "loss": 0.8743,
+      "step": 2357
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.158860499691112,
+      "learning_rate": 9.321940644141427e-06,
+      "loss": 0.3042,
+      "step": 2358
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.485617366183805,
+      "learning_rate": 9.3212748631891e-06,
+      "loss": 1.0994,
+      "step": 2359
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.1845245822430632,
+      "learning_rate": 9.320608779334212e-06,
+      "loss": 0.401,
+      "step": 2360
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.408107084724933,
+      "learning_rate": 9.319942392623451e-06,
+      "loss": 0.5151,
+      "step": 2361
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.550807236446376,
+      "learning_rate": 9.319275703103529e-06,
+      "loss": 0.5964,
+      "step": 2362
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.59677790334423,
+      "learning_rate": 9.318608710821179e-06,
+      "loss": 1.0998,
+      "step": 2363
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.2999653007697263,
+      "learning_rate": 9.317941415823151e-06,
+      "loss": 0.8447,
+      "step": 2364
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0328133919182665,
+      "learning_rate": 9.317273818156223e-06,
+      "loss": 0.1885,
+      "step": 2365
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.439082118673476,
+      "learning_rate": 9.316605917867189e-06,
+      "loss": 1.244,
+      "step": 2366
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.144453026630218,
+      "learning_rate": 9.315937715002865e-06,
+      "loss": 0.8939,
+      "step": 2367
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.661134589171982,
+      "learning_rate": 9.315269209610092e-06,
+      "loss": 0.727,
+      "step": 2368
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.0121522655064643,
+      "learning_rate": 9.314600401735727e-06,
+      "loss": 0.6767,
+      "step": 2369
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.9143730505910006,
+      "learning_rate": 9.31393129142665e-06,
+      "loss": 0.7601,
+      "step": 2370
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.5397444675924405,
+      "learning_rate": 9.313261878729765e-06,
+      "loss": 0.7476,
+      "step": 2371
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.4541352720937573,
+      "learning_rate": 9.312592163691991e-06,
+      "loss": 1.0122,
+      "step": 2372
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.368894813193451,
+      "learning_rate": 9.311922146360276e-06,
+      "loss": 1.2218,
+      "step": 2373
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.196780787673087,
+      "learning_rate": 9.311251826781587e-06,
+      "loss": 0.7043,
+      "step": 2374
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.937401888424667,
+      "learning_rate": 9.310581205002905e-06,
+      "loss": 0.7497,
+      "step": 2375
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.7329755870495376,
+      "learning_rate": 9.30991028107124e-06,
+      "loss": 0.9845,
+      "step": 2376
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.1875849682133115,
+      "learning_rate": 9.309239055033623e-06,
+      "loss": 0.7722,
+      "step": 2377
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.936687169393056,
+      "learning_rate": 9.3085675269371e-06,
+      "loss": 1.018,
+      "step": 2378
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.192478312596038,
+      "learning_rate": 9.307895696828746e-06,
+      "loss": 1.0972,
+      "step": 2379
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0847921517774552,
+      "learning_rate": 9.307223564755649e-06,
+      "loss": 0.3726,
+      "step": 2380
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.861499554696102,
+      "learning_rate": 9.306551130764929e-06,
+      "loss": 0.6663,
+      "step": 2381
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.132355188229489,
+      "learning_rate": 9.305878394903714e-06,
+      "loss": 0.8424,
+      "step": 2382
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.9622667131259264,
+      "learning_rate": 9.305205357219165e-06,
+      "loss": 0.88,
+      "step": 2383
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.162565828964038,
+      "learning_rate": 9.304532017758454e-06,
+      "loss": 0.8993,
+      "step": 2384
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.720231045728556,
+      "learning_rate": 9.303858376568784e-06,
+      "loss": 0.8852,
+      "step": 2385
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.6436473738437174,
+      "learning_rate": 9.303184433697371e-06,
+      "loss": 0.7395,
+      "step": 2386
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.7860005278747586,
+      "learning_rate": 9.302510189191458e-06,
+      "loss": 0.974,
+      "step": 2387
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.741299005535924,
+      "learning_rate": 9.301835643098305e-06,
+      "loss": 1.4295,
+      "step": 2388
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.393585118933961,
+      "learning_rate": 9.301160795465196e-06,
+      "loss": 1.1355,
+      "step": 2389
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.619503406329702,
+      "learning_rate": 9.300485646339431e-06,
+      "loss": 0.9534,
+      "step": 2390
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.111439895035622,
+      "learning_rate": 9.299810195768341e-06,
+      "loss": 0.8473,
+      "step": 2391
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.223696444269811,
+      "learning_rate": 9.299134443799267e-06,
+      "loss": 0.4557,
+      "step": 2392
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.174862259989276,
+      "learning_rate": 9.298458390479579e-06,
+      "loss": 1.2313,
+      "step": 2393
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.049120303711942,
+      "learning_rate": 9.297782035856667e-06,
+      "loss": 0.9009,
+      "step": 2394
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.607973996729967,
+      "learning_rate": 9.297105379977935e-06,
+      "loss": 1.311,
+      "step": 2395
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.804514523499376,
+      "learning_rate": 9.296428422890817e-06,
+      "loss": 0.5371,
+      "step": 2396
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.5712502914725803,
+      "learning_rate": 9.295751164642767e-06,
+      "loss": 0.3081,
+      "step": 2397
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.9385847868834265,
+      "learning_rate": 9.295073605281255e-06,
+      "loss": 1.3572,
+      "step": 2398
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.953183154224133,
+      "learning_rate": 9.294395744853775e-06,
+      "loss": 0.7451,
+      "step": 2399
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.006366705979045,
+      "learning_rate": 9.293717583407843e-06,
+      "loss": 1.0411,
+      "step": 2400
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.9360113632346465,
+      "learning_rate": 9.293039120990995e-06,
+      "loss": 0.7905,
+      "step": 2401
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.970447326489623,
+      "learning_rate": 9.292360357650785e-06,
+      "loss": 0.9431,
+      "step": 2402
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.742640339977302,
+      "learning_rate": 9.291681293434797e-06,
+      "loss": 0.8987,
+      "step": 2403
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.5528572102404032,
+      "learning_rate": 9.291001928390629e-06,
+      "loss": 0.8615,
+      "step": 2404
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.186330963000171,
+      "learning_rate": 9.290322262565897e-06,
+      "loss": 1.3626,
+      "step": 2405
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.9948835802395184,
+      "learning_rate": 9.289642296008248e-06,
+      "loss": 0.7315,
+      "step": 2406
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.8638274177086402,
+      "learning_rate": 9.288962028765342e-06,
+      "loss": 0.5873,
+      "step": 2407
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.20550751601823,
+      "learning_rate": 9.288281460884864e-06,
+      "loss": 1.2367,
+      "step": 2408
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.6467454307293434,
+      "learning_rate": 9.287600592414517e-06,
+      "loss": 0.6496,
+      "step": 2409
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0927102373684074,
+      "learning_rate": 9.28691942340203e-06,
+      "loss": 0.2213,
+      "step": 2410
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.9930871543106923,
+      "learning_rate": 9.286237953895148e-06,
+      "loss": 0.5662,
+      "step": 2411
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.3265903378205772,
+      "learning_rate": 9.285556183941637e-06,
+      "loss": 0.9091,
+      "step": 2412
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.955673187066376,
+      "learning_rate": 9.28487411358929e-06,
+      "loss": 0.6372,
+      "step": 2413
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.006607505537565,
+      "learning_rate": 9.284191742885915e-06,
+      "loss": 1.1425,
+      "step": 2414
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.078956744021499,
+      "learning_rate": 9.283509071879344e-06,
+      "loss": 1.2454,
+      "step": 2415
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.392870823843581,
+      "learning_rate": 9.282826100617429e-06,
+      "loss": 0.5376,
+      "step": 2416
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.467695605250697,
+      "learning_rate": 9.282142829148043e-06,
+      "loss": 1.0673,
+      "step": 2417
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1094007382106783,
+      "learning_rate": 9.28145925751908e-06,
+      "loss": 0.1823,
+      "step": 2418
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.416330381903597,
+      "learning_rate": 9.280775385778458e-06,
+      "loss": 1.1083,
+      "step": 2419
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.8890851270704814,
+      "learning_rate": 9.280091213974109e-06,
+      "loss": 0.974,
+      "step": 2420
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.781892576166949,
+      "learning_rate": 9.279406742153996e-06,
+      "loss": 0.4089,
+      "step": 2421
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.066866788110671,
+      "learning_rate": 9.278721970366092e-06,
+      "loss": 1.3843,
+      "step": 2422
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.9425551802733314,
+      "learning_rate": 9.278036898658401e-06,
+      "loss": 0.7035,
+      "step": 2423
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.68743343050816,
+      "learning_rate": 9.277351527078938e-06,
+      "loss": 1.1653,
+      "step": 2424
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.845281486402816,
+      "learning_rate": 9.276665855675751e-06,
+      "loss": 1.1039,
+      "step": 2425
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.201882004264333,
+      "learning_rate": 9.275979884496898e-06,
+      "loss": 1.107,
+      "step": 2426
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.02109199001339,
+      "learning_rate": 9.275293613590465e-06,
+      "loss": 0.9791,
+      "step": 2427
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9454430982583224,
+      "learning_rate": 9.274607043004556e-06,
+      "loss": 0.4404,
+      "step": 2428
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.627297027829158,
+      "learning_rate": 9.273920172787297e-06,
+      "loss": 0.7501,
+      "step": 2429
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.4906400608114176,
+      "learning_rate": 9.273233002986833e-06,
+      "loss": 0.929,
+      "step": 2430
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.8919170121454916,
+      "learning_rate": 9.27254553365133e-06,
+      "loss": 1.2215,
+      "step": 2431
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.729171809737526,
+      "learning_rate": 9.271857764828985e-06,
+      "loss": 1.419,
+      "step": 2432
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.387967287590401,
+      "learning_rate": 9.271169696567999e-06,
+      "loss": 0.7166,
+      "step": 2433
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.7446962517953417,
+      "learning_rate": 9.270481328916605e-06,
+      "loss": 0.509,
+      "step": 2434
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.3337688838086725,
+      "learning_rate": 9.269792661923055e-06,
+      "loss": 0.8205,
+      "step": 2435
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.070855586850423,
+      "learning_rate": 9.269103695635622e-06,
+      "loss": 1.2684,
+      "step": 2436
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.072157256597414,
+      "learning_rate": 9.2684144301026e-06,
+      "loss": 1.0799,
+      "step": 2437
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.4439968585400815,
+      "learning_rate": 9.2677248653723e-06,
+      "loss": 0.7584,
+      "step": 2438
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.0997069644786572,
+      "learning_rate": 9.267035001493064e-06,
+      "loss": 0.7358,
+      "step": 2439
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.321923940828505,
+      "learning_rate": 9.266344838513241e-06,
+      "loss": 0.911,
+      "step": 2440
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.797153143640582,
+      "learning_rate": 9.265654376481214e-06,
+      "loss": 0.7008,
+      "step": 2441
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.137346063274294,
+      "learning_rate": 9.264963615445378e-06,
+      "loss": 0.5624,
+      "step": 2442
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.329038814070353,
+      "learning_rate": 9.264272555454156e-06,
+      "loss": 1.1483,
+      "step": 2443
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.846667627346543,
+      "learning_rate": 9.263581196555984e-06,
+      "loss": 1.1069,
+      "step": 2444
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.0885308129338425,
+      "learning_rate": 9.262889538799327e-06,
+      "loss": 0.6528,
+      "step": 2445
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.984212107188813,
+      "learning_rate": 9.262197582232665e-06,
+      "loss": 0.7519,
+      "step": 2446
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.6441126736579625,
+      "learning_rate": 9.2615053269045e-06,
+      "loss": 0.5124,
+      "step": 2447
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.801090724259183,
+      "learning_rate": 9.260812772863362e-06,
+      "loss": 0.5342,
+      "step": 2448
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.0335952212132895,
+      "learning_rate": 9.260119920157786e-06,
+      "loss": 1.2109,
+      "step": 2449
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.3807668935028627,
+      "learning_rate": 9.259426768836347e-06,
+      "loss": 0.7047,
+      "step": 2450
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.588787739876182,
+      "learning_rate": 9.258733318947627e-06,
+      "loss": 1.5709,
+      "step": 2451
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.1994010042190997,
+      "learning_rate": 9.258039570540238e-06,
+      "loss": 0.8593,
+      "step": 2452
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.7504222258312465,
+      "learning_rate": 9.257345523662804e-06,
+      "loss": 0.9261,
+      "step": 2453
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.411878854734679,
+      "learning_rate": 9.256651178363978e-06,
+      "loss": 1.0292,
+      "step": 2454
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.609824834430926,
+      "learning_rate": 9.255956534692428e-06,
+      "loss": 1.5112,
+      "step": 2455
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.657719639064696,
+      "learning_rate": 9.255261592696849e-06,
+      "loss": 1.0546,
+      "step": 2456
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.9287243102542893,
+      "learning_rate": 9.254566352425949e-06,
+      "loss": 1.0827,
+      "step": 2457
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.406667014838754,
+      "learning_rate": 9.253870813928465e-06,
+      "loss": 1.0112,
+      "step": 2458
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.7527747181267754,
+      "learning_rate": 9.25317497725315e-06,
+      "loss": 0.9954,
+      "step": 2459
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9851742175340266,
+      "learning_rate": 9.252478842448778e-06,
+      "loss": 0.4253,
+      "step": 2460
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.928820661023417,
+      "learning_rate": 9.251782409564146e-06,
+      "loss": 0.4773,
+      "step": 2461
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.542990238697449,
+      "learning_rate": 9.251085678648072e-06,
+      "loss": 0.5688,
+      "step": 2462
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.9259639052018427,
+      "learning_rate": 9.250388649749391e-06,
+      "loss": 0.6107,
+      "step": 2463
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.511230033717203,
+      "learning_rate": 9.249691322916965e-06,
+      "loss": 1.7268,
+      "step": 2464
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.015230401762193,
+      "learning_rate": 9.248993698199672e-06,
+      "loss": 0.9483,
+      "step": 2465
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.568010511496238,
+      "learning_rate": 9.248295775646412e-06,
+      "loss": 0.9769,
+      "step": 2466
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.767556720925712,
+      "learning_rate": 9.247597555306107e-06,
+      "loss": 0.7468,
+      "step": 2467
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6824470971410319,
+      "learning_rate": 9.246899037227698e-06,
+      "loss": 0.1253,
+      "step": 2468
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.556352405481264,
+      "learning_rate": 9.246200221460148e-06,
+      "loss": 0.4267,
+      "step": 2469
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.5598987471209234,
+      "learning_rate": 9.245501108052447e-06,
+      "loss": 0.654,
+      "step": 2470
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.1694604413817036,
+      "learning_rate": 9.24480169705359e-06,
+      "loss": 0.3556,
+      "step": 2471
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.614876497302162,
+      "learning_rate": 9.244101988512608e-06,
+      "loss": 0.8888,
+      "step": 2472
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.11782195067912,
+      "learning_rate": 9.243401982478548e-06,
+      "loss": 1.0358,
+      "step": 2473
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.0826329130323074,
+      "learning_rate": 9.242701679000477e-06,
+      "loss": 0.4509,
+      "step": 2474
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.449403101466768,
+      "learning_rate": 9.242001078127483e-06,
+      "loss": 0.4125,
+      "step": 2475
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0237449018883034,
+      "learning_rate": 9.241300179908672e-06,
+      "loss": 0.4052,
+      "step": 2476
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.948681216804102,
+      "learning_rate": 9.240598984393179e-06,
+      "loss": 0.3693,
+      "step": 2477
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.3261854739380206,
+      "learning_rate": 9.239897491630152e-06,
+      "loss": 0.8982,
+      "step": 2478
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.033930031952647,
+      "learning_rate": 9.239195701668762e-06,
+      "loss": 0.6192,
+      "step": 2479
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.718121732200636,
+      "learning_rate": 9.238493614558203e-06,
+      "loss": 0.944,
+      "step": 2480
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.9801540794617747,
+      "learning_rate": 9.237791230347688e-06,
+      "loss": 1.0094,
+      "step": 2481
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.3348847550539866,
+      "learning_rate": 9.237088549086449e-06,
+      "loss": 0.7,
+      "step": 2482
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.9710795346239567,
+      "learning_rate": 9.236385570823746e-06,
+      "loss": 0.5839,
+      "step": 2483
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.174006026845121,
+      "learning_rate": 9.235682295608848e-06,
+      "loss": 1.3029,
+      "step": 2484
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.230100928597258,
+      "learning_rate": 9.234978723491054e-06,
+      "loss": 1.027,
+      "step": 2485
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.648003166348003,
+      "learning_rate": 9.234274854519685e-06,
+      "loss": 1.0663,
+      "step": 2486
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.8057104068829544,
+      "learning_rate": 9.233570688744076e-06,
+      "loss": 0.4592,
+      "step": 2487
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.906143424329132,
+      "learning_rate": 9.232866226213586e-06,
+      "loss": 0.5144,
+      "step": 2488
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.8208585072714936,
+      "learning_rate": 9.232161466977595e-06,
+      "loss": 0.5398,
+      "step": 2489
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.105264891709723,
+      "learning_rate": 9.231456411085502e-06,
+      "loss": 0.3397,
+      "step": 2490
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.9134449975356302,
+      "learning_rate": 9.23075105858673e-06,
+      "loss": 0.9787,
+      "step": 2491
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.549363434492431,
+      "learning_rate": 9.230045409530724e-06,
+      "loss": 1.1104,
+      "step": 2492
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.741785394799542,
+      "learning_rate": 9.229339463966942e-06,
+      "loss": 0.4278,
+      "step": 2493
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.3875035043855193,
+      "learning_rate": 9.228633221944869e-06,
+      "loss": 0.4966,
+      "step": 2494
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.079424147406122,
+      "learning_rate": 9.227926683514012e-06,
+      "loss": 0.4968,
+      "step": 2495
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.319924544614497,
+      "learning_rate": 9.227219848723893e-06,
+      "loss": 0.6642,
+      "step": 2496
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.25907882646003,
+      "learning_rate": 9.226512717624062e-06,
+      "loss": 0.7904,
+      "step": 2497
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.497068125726394,
+      "learning_rate": 9.22580529026408e-06,
+      "loss": 0.7181,
+      "step": 2498
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.802893634393092,
+      "learning_rate": 9.225097566693539e-06,
+      "loss": 0.7675,
+      "step": 2499
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.095308909812506,
+      "learning_rate": 9.224389546962047e-06,
+      "loss": 0.1666,
+      "step": 2500
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.395290691931256,
+      "learning_rate": 9.223681231119232e-06,
+      "loss": 0.6342,
+      "step": 2501
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.544834977506785,
+      "learning_rate": 9.222972619214745e-06,
+      "loss": 1.0556,
+      "step": 2502
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.257132580663881,
+      "learning_rate": 9.222263711298256e-06,
+      "loss": 1.2329,
+      "step": 2503
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.952517170894928,
+      "learning_rate": 9.221554507419455e-06,
+      "loss": 0.4574,
+      "step": 2504
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.894079523365917,
+      "learning_rate": 9.220845007628055e-06,
+      "loss": 0.94,
+      "step": 2505
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.166460250726755,
+      "learning_rate": 9.22013521197379e-06,
+      "loss": 0.8148,
+      "step": 2506
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.795346323399248,
+      "learning_rate": 9.219425120506414e-06,
+      "loss": 0.6581,
+      "step": 2507
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.856522992182047,
+      "learning_rate": 9.218714733275698e-06,
+      "loss": 0.8704,
+      "step": 2508
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.514542570030826,
+      "learning_rate": 9.21800405033144e-06,
+      "loss": 1.3371,
+      "step": 2509
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.5945254947430785,
+      "learning_rate": 9.217293071723455e-06,
+      "loss": 1.136,
+      "step": 2510
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.7277842663523812,
+      "learning_rate": 9.216581797501578e-06,
+      "loss": 0.686,
+      "step": 2511
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.369640029754952,
+      "learning_rate": 9.215870227715669e-06,
+      "loss": 1.2206,
+      "step": 2512
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.127736755897348,
+      "learning_rate": 9.215158362415604e-06,
+      "loss": 0.8052,
+      "step": 2513
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.604555276458949,
+      "learning_rate": 9.21444620165128e-06,
+      "loss": 1.0992,
+      "step": 2514
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.653818385651074,
+      "learning_rate": 9.213733745472623e-06,
+      "loss": 0.9455,
+      "step": 2515
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.9787448891704096,
+      "learning_rate": 9.213020993929566e-06,
+      "loss": 0.6354,
+      "step": 2516
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.8866749246671595,
+      "learning_rate": 9.212307947072074e-06,
+      "loss": 0.541,
+      "step": 2517
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 6.339381202051915,
+      "learning_rate": 9.211594604950127e-06,
+      "loss": 1.2898,
+      "step": 2518
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.961553529245296,
+      "learning_rate": 9.210880967613724e-06,
+      "loss": 1.0144,
+      "step": 2519
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.234675316545422,
+      "learning_rate": 9.210167035112894e-06,
+      "loss": 1.2231,
+      "step": 2520
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.358216145490436,
+      "learning_rate": 9.209452807497677e-06,
+      "loss": 1.1559,
+      "step": 2521
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.5686683887336494,
+      "learning_rate": 9.208738284818138e-06,
+      "loss": 0.7985,
+      "step": 2522
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9181533771460326,
+      "learning_rate": 9.20802346712436e-06,
+      "loss": 0.4145,
+      "step": 2523
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.231466180884634,
+      "learning_rate": 9.20730835446645e-06,
+      "loss": 0.9182,
+      "step": 2524
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.7500185297541964,
+      "learning_rate": 9.206592946894538e-06,
+      "loss": 0.6986,
+      "step": 2525
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.414215285507461,
+      "learning_rate": 9.205877244458765e-06,
+      "loss": 0.6744,
+      "step": 2526
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.395880282392656,
+      "learning_rate": 9.205161247209303e-06,
+      "loss": 0.317,
+      "step": 2527
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.427669095406331,
+      "learning_rate": 9.204444955196337e-06,
+      "loss": 0.8707,
+      "step": 2528
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.7214519772571037,
+      "learning_rate": 9.203728368470077e-06,
+      "loss": 0.8583,
+      "step": 2529
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.423959878964401,
+      "learning_rate": 9.203011487080755e-06,
+      "loss": 0.5747,
+      "step": 2530
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.98099678270369,
+      "learning_rate": 9.202294311078618e-06,
+      "loss": 0.7567,
+      "step": 2531
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.4053424148101112,
+      "learning_rate": 9.201576840513939e-06,
+      "loss": 0.737,
+      "step": 2532
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.3427163106192985,
+      "learning_rate": 9.200859075437008e-06,
+      "loss": 0.6799,
+      "step": 2533
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.32925968832063,
+      "learning_rate": 9.200141015898138e-06,
+      "loss": 1.209,
+      "step": 2534
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.265874001661272,
+      "learning_rate": 9.199422661947662e-06,
+      "loss": 0.8002,
+      "step": 2535
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.08560714492616,
+      "learning_rate": 9.198704013635934e-06,
+      "loss": 0.7652,
+      "step": 2536
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.8415414566991486,
+      "learning_rate": 9.197985071013326e-06,
+      "loss": 0.7683,
+      "step": 2537
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.849674167020621,
+      "learning_rate": 9.197265834130235e-06,
+      "loss": 0.8824,
+      "step": 2538
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.2292221455199654,
+      "learning_rate": 9.196546303037077e-06,
+      "loss": 0.9428,
+      "step": 2539
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.3977499377383547,
+      "learning_rate": 9.195826477784286e-06,
+      "loss": 0.8224,
+      "step": 2540
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.407183252962648,
+      "learning_rate": 9.19510635842232e-06,
+      "loss": 1.0697,
+      "step": 2541
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.555713809852212,
+      "learning_rate": 9.194385945001652e-06,
+      "loss": 1.1947,
+      "step": 2542
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.5305475089934677,
+      "learning_rate": 9.193665237572785e-06,
+      "loss": 0.6139,
+      "step": 2543
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.9543541968597733,
+      "learning_rate": 9.192944236186237e-06,
+      "loss": 0.5635,
+      "step": 2544
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.804833370830699,
+      "learning_rate": 9.192222940892543e-06,
+      "loss": 0.7743,
+      "step": 2545
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.051494233217439,
+      "learning_rate": 9.191501351742269e-06,
+      "loss": 0.6582,
+      "step": 2546
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.79343864325062,
+      "learning_rate": 9.19077946878599e-06,
+      "loss": 1.1509,
+      "step": 2547
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 6.22126397598874,
+      "learning_rate": 9.190057292074308e-06,
+      "loss": 1.5538,
+      "step": 2548
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.317098774130853,
+      "learning_rate": 9.189334821657846e-06,
+      "loss": 0.6657,
+      "step": 2549
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.7015628351357672,
+      "learning_rate": 9.188612057587246e-06,
+      "loss": 0.6748,
+      "step": 2550
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.6132826447826805,
+      "learning_rate": 9.187888999913166e-06,
+      "loss": 0.7137,
+      "step": 2551
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.2849372692578647,
+      "learning_rate": 9.187165648686296e-06,
+      "loss": 0.8883,
+      "step": 2552
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7770415993893534,
+      "learning_rate": 9.186442003957337e-06,
+      "loss": 0.4326,
+      "step": 2553
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.206890972843402,
+      "learning_rate": 9.185718065777011e-06,
+      "loss": 1.0782,
+      "step": 2554
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.855430118818454,
+      "learning_rate": 9.184993834196065e-06,
+      "loss": 0.9108,
+      "step": 2555
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.224854598296555,
+      "learning_rate": 9.184269309265266e-06,
+      "loss": 0.9526,
+      "step": 2556
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.649110311640722,
+      "learning_rate": 9.183544491035396e-06,
+      "loss": 0.5603,
+      "step": 2557
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.6275348993531513,
+      "learning_rate": 9.182819379557266e-06,
+      "loss": 0.634,
+      "step": 2558
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.656912091187971,
+      "learning_rate": 9.1820939748817e-06,
+      "loss": 0.8573,
+      "step": 2559
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.174479933286031,
+      "learning_rate": 9.181368277059548e-06,
+      "loss": 0.6829,
+      "step": 2560
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.5872657569719784,
+      "learning_rate": 9.180642286141678e-06,
+      "loss": 0.3411,
+      "step": 2561
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.789949702148944,
+      "learning_rate": 9.179916002178976e-06,
+      "loss": 0.6619,
+      "step": 2562
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.534928294722872,
+      "learning_rate": 9.179189425222354e-06,
+      "loss": 0.8082,
+      "step": 2563
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.851233261542717,
+      "learning_rate": 9.178462555322742e-06,
+      "loss": 1.2742,
+      "step": 2564
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.1864289822875618,
+      "learning_rate": 9.177735392531088e-06,
+      "loss": 0.6699,
+      "step": 2565
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.5049285283905367,
+      "learning_rate": 9.177007936898365e-06,
+      "loss": 0.804,
+      "step": 2566
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.372758267623915,
+      "learning_rate": 9.176280188475565e-06,
+      "loss": 0.8771,
+      "step": 2567
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.0264069966642824,
+      "learning_rate": 9.175552147313698e-06,
+      "loss": 0.458,
+      "step": 2568
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.921229189159816,
+      "learning_rate": 9.174823813463799e-06,
+      "loss": 1.0942,
+      "step": 2569
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.393971440135176,
+      "learning_rate": 9.174095186976916e-06,
+      "loss": 0.8405,
+      "step": 2570
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.680965342339284,
+      "learning_rate": 9.17336626790413e-06,
+      "loss": 1.5422,
+      "step": 2571
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.681309457351905,
+      "learning_rate": 9.172637056296529e-06,
+      "loss": 0.7138,
+      "step": 2572
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.449514166310241,
+      "learning_rate": 9.17190755220523e-06,
+      "loss": 1.0039,
+      "step": 2573
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.251873711827825,
+      "learning_rate": 9.17117775568137e-06,
+      "loss": 0.8907,
+      "step": 2574
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.1650591046235372,
+      "learning_rate": 9.170447666776101e-06,
+      "loss": 0.5766,
+      "step": 2575
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.4827679981366515,
+      "learning_rate": 9.1697172855406e-06,
+      "loss": 0.5803,
+      "step": 2576
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.405827340101964,
+      "learning_rate": 9.168986612026063e-06,
+      "loss": 1.0401,
+      "step": 2577
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.108557485440849,
+      "learning_rate": 9.16825564628371e-06,
+      "loss": 0.9864,
+      "step": 2578
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.72691957372029,
+      "learning_rate": 9.167524388364775e-06,
+      "loss": 1.2002,
+      "step": 2579
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.206603025302393,
+      "learning_rate": 9.166792838320517e-06,
+      "loss": 0.6778,
+      "step": 2580
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.857778089354351,
+      "learning_rate": 9.166060996202218e-06,
+      "loss": 0.5272,
+      "step": 2581
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.0259298682451554,
+      "learning_rate": 9.165328862061172e-06,
+      "loss": 0.5766,
+      "step": 2582
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.414197217782002,
+      "learning_rate": 9.164596435948699e-06,
+      "loss": 1.0155,
+      "step": 2583
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.908278810961928,
+      "learning_rate": 9.163863717916142e-06,
+      "loss": 1.0146,
+      "step": 2584
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.712167202066841,
+      "learning_rate": 9.163130708014858e-06,
+      "loss": 0.5501,
+      "step": 2585
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.0649722491996014,
+      "learning_rate": 9.16239740629623e-06,
+      "loss": 0.5382,
+      "step": 2586
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.3574318547925315,
+      "learning_rate": 9.16166381281166e-06,
+      "loss": 0.2507,
+      "step": 2587
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.015733310555574,
+      "learning_rate": 9.160929927612567e-06,
+      "loss": 1.1375,
+      "step": 2588
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.1669045648508427,
+      "learning_rate": 9.160195750750396e-06,
+      "loss": 0.3214,
+      "step": 2589
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.127864108182433,
+      "learning_rate": 9.159461282276605e-06,
+      "loss": 1.0092,
+      "step": 2590
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.848476885058125,
+      "learning_rate": 9.158726522242684e-06,
+      "loss": 0.6853,
+      "step": 2591
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.3804941378319917,
+      "learning_rate": 9.15799147070013e-06,
+      "loss": 0.4897,
+      "step": 2592
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.053977480378836,
+      "learning_rate": 9.157256127700472e-06,
+      "loss": 0.8833,
+      "step": 2593
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.75654008244524,
+      "learning_rate": 9.156520493295249e-06,
+      "loss": 0.7025,
+      "step": 2594
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.2543139923423734,
+      "learning_rate": 9.15578456753603e-06,
+      "loss": 0.8477,
+      "step": 2595
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.4459278163166998,
+      "learning_rate": 9.155048350474398e-06,
+      "loss": 0.4061,
+      "step": 2596
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.4250556753826427,
+      "learning_rate": 9.15431184216196e-06,
+      "loss": 0.559,
+      "step": 2597
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.7293760100346875,
+      "learning_rate": 9.153575042650342e-06,
+      "loss": 0.4459,
+      "step": 2598
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.88752240281919,
+      "learning_rate": 9.15283795199119e-06,
+      "loss": 0.702,
+      "step": 2599
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.850180234109402,
+      "learning_rate": 9.152100570236172e-06,
+      "loss": 1.0422,
+      "step": 2600
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.1384254357693853,
+      "learning_rate": 9.15136289743697e-06,
+      "loss": 1.0971,
+      "step": 2601
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.1876046760275427,
+      "learning_rate": 9.150624933645297e-06,
+      "loss": 0.9115,
+      "step": 2602
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.909756335673938,
+      "learning_rate": 9.149886678912883e-06,
+      "loss": 0.7705,
+      "step": 2603
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.100639887160335,
+      "learning_rate": 9.14914813329147e-06,
+      "loss": 0.8299,
+      "step": 2604
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.270330154313572,
+      "learning_rate": 9.148409296832832e-06,
+      "loss": 0.9597,
+      "step": 2605
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.054076412924293,
+      "learning_rate": 9.147670169588754e-06,
+      "loss": 1.1523,
+      "step": 2606
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.716632979901013,
+      "learning_rate": 9.14693075161105e-06,
+      "loss": 1.0254,
+      "step": 2607
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.6393434225559393,
+      "learning_rate": 9.146191042951546e-06,
+      "loss": 0.7487,
+      "step": 2608
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.467533295509262,
+      "learning_rate": 9.145451043662095e-06,
+      "loss": 0.7654,
+      "step": 2609
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.3948877468965932,
+      "learning_rate": 9.144710753794567e-06,
+      "loss": 0.3971,
+      "step": 2610
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.4801815731732444,
+      "learning_rate": 9.143970173400853e-06,
+      "loss": 0.3545,
+      "step": 2611
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.6644744590683365,
+      "learning_rate": 9.143229302532866e-06,
+      "loss": 1.3814,
+      "step": 2612
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.417698106094416,
+      "learning_rate": 9.142488141242534e-06,
+      "loss": 1.0687,
+      "step": 2613
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.8191416836094914,
+      "learning_rate": 9.141746689581811e-06,
+      "loss": 0.8562,
+      "step": 2614
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.273085690695707,
+      "learning_rate": 9.141004947602672e-06,
+      "loss": 0.5464,
+      "step": 2615
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.001040918183898,
+      "learning_rate": 9.140262915357107e-06,
+      "loss": 0.5869,
+      "step": 2616
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.510086130850804,
+      "learning_rate": 9.139520592897131e-06,
+      "loss": 1.0375,
+      "step": 2617
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.0965008392075584,
+      "learning_rate": 9.138777980274776e-06,
+      "loss": 0.564,
+      "step": 2618
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.2354925236562706,
+      "learning_rate": 9.138035077542096e-06,
+      "loss": 0.5104,
+      "step": 2619
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.003837214181366,
+      "learning_rate": 9.137291884751165e-06,
+      "loss": 1.6036,
+      "step": 2620
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.132618628152957,
+      "learning_rate": 9.13654840195408e-06,
+      "loss": 0.9808,
+      "step": 2621
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.189334795046342,
+      "learning_rate": 9.135804629202955e-06,
+      "loss": 0.7809,
+      "step": 2622
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.956989408432872,
+      "learning_rate": 9.135060566549924e-06,
+      "loss": 1.608,
+      "step": 2623
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.4074585256132326,
+      "learning_rate": 9.134316214047144e-06,
+      "loss": 0.3921,
+      "step": 2624
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.742860665175202,
+      "learning_rate": 9.133571571746786e-06,
+      "loss": 1.0199,
+      "step": 2625
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.452758980350587,
+      "learning_rate": 9.132826639701055e-06,
+      "loss": 0.8877,
+      "step": 2626
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.949992256628996,
+      "learning_rate": 9.13208141796216e-06,
+      "loss": 0.7818,
+      "step": 2627
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.4071015775853146,
+      "learning_rate": 9.13133590658234e-06,
+      "loss": 0.4309,
+      "step": 2628
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.843805103301418,
+      "learning_rate": 9.130590105613854e-06,
+      "loss": 0.7782,
+      "step": 2629
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.859908067407454,
+      "learning_rate": 9.129844015108978e-06,
+      "loss": 1.06,
+      "step": 2630
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.143641660915089,
+      "learning_rate": 9.12909763512001e-06,
+      "loss": 1.1174,
+      "step": 2631
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.0003721523478255,
+      "learning_rate": 9.128350965699267e-06,
+      "loss": 1.1409,
+      "step": 2632
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.3123734082883107,
+      "learning_rate": 9.127604006899088e-06,
+      "loss": 0.6201,
+      "step": 2633
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.8886274054005017,
+      "learning_rate": 9.126856758771832e-06,
+      "loss": 0.9564,
+      "step": 2634
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.363089925833701,
+      "learning_rate": 9.126109221369877e-06,
+      "loss": 0.8048,
+      "step": 2635
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.597459332169226,
+      "learning_rate": 9.125361394745621e-06,
+      "loss": 0.5441,
+      "step": 2636
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.121212674843806,
+      "learning_rate": 9.124613278951486e-06,
+      "loss": 0.6559,
+      "step": 2637
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.34419329799521,
+      "learning_rate": 9.12386487403991e-06,
+      "loss": 0.8502,
+      "step": 2638
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.4451716397751406,
+      "learning_rate": 9.123116180063356e-06,
+      "loss": 0.5108,
+      "step": 2639
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.473863821303688,
+      "learning_rate": 9.1223671970743e-06,
+      "loss": 0.9417,
+      "step": 2640
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.6621722801899974,
+      "learning_rate": 9.121617925125244e-06,
+      "loss": 0.8106,
+      "step": 2641
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.6666378806199424,
+      "learning_rate": 9.12086836426871e-06,
+      "loss": 0.932,
+      "step": 2642
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4977537563590089,
+      "learning_rate": 9.120118514557235e-06,
+      "loss": 0.2099,
+      "step": 2643
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.113942022339024,
+      "learning_rate": 9.119368376043384e-06,
+      "loss": 0.774,
+      "step": 2644
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.816855066586145,
+      "learning_rate": 9.118617948779738e-06,
+      "loss": 1.1958,
+      "step": 2645
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.335000089257285,
+      "learning_rate": 9.117867232818897e-06,
+      "loss": 1.0147,
+      "step": 2646
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.171377852877847,
+      "learning_rate": 9.117116228213485e-06,
+      "loss": 0.9544,
+      "step": 2647
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.662025483828039,
+      "learning_rate": 9.116364935016144e-06,
+      "loss": 1.0558,
+      "step": 2648
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.5688211631492845,
+      "learning_rate": 9.115613353279533e-06,
+      "loss": 0.6812,
+      "step": 2649
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.267943032312493,
+      "learning_rate": 9.11486148305634e-06,
+      "loss": 1.248,
+      "step": 2650
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.542385422783675,
+      "learning_rate": 9.114109324399263e-06,
+      "loss": 1.3974,
+      "step": 2651
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.4433326699791102,
+      "learning_rate": 9.113356877361027e-06,
+      "loss": 0.4743,
+      "step": 2652
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.6224171332006985,
+      "learning_rate": 9.112604141994376e-06,
+      "loss": 0.7523,
+      "step": 2653
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.3394163890362543,
+      "learning_rate": 9.111851118352074e-06,
+      "loss": 0.6131,
+      "step": 2654
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.7817423764246874,
+      "learning_rate": 9.111097806486901e-06,
+      "loss": 1.0332,
+      "step": 2655
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.691615566474534,
+      "learning_rate": 9.110344206451665e-06,
+      "loss": 0.697,
+      "step": 2656
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.4666547510013435,
+      "learning_rate": 9.109590318299189e-06,
+      "loss": 0.9713,
+      "step": 2657
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.537850282404029,
+      "learning_rate": 9.108836142082316e-06,
+      "loss": 1.2781,
+      "step": 2658
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.533021480613041,
+      "learning_rate": 9.108081677853911e-06,
+      "loss": 1.2023,
+      "step": 2659
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7972539953327008,
+      "learning_rate": 9.107326925666857e-06,
+      "loss": 0.1528,
+      "step": 2660
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.7799263906140883,
+      "learning_rate": 9.106571885574062e-06,
+      "loss": 0.9498,
+      "step": 2661
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.73423685102641,
+      "learning_rate": 9.105816557628451e-06,
+      "loss": 0.9331,
+      "step": 2662
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.762462442085408,
+      "learning_rate": 9.105060941882966e-06,
+      "loss": 1.0003,
+      "step": 2663
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.012753402102101,
+      "learning_rate": 9.104305038390575e-06,
+      "loss": 0.6783,
+      "step": 2664
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.68841305451672,
+      "learning_rate": 9.103548847204263e-06,
+      "loss": 1.2537,
+      "step": 2665
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.846154686958786,
+      "learning_rate": 9.102792368377036e-06,
+      "loss": 0.3033,
+      "step": 2666
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.785629135154621,
+      "learning_rate": 9.102035601961919e-06,
+      "loss": 0.5163,
+      "step": 2667
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.528580565267492,
+      "learning_rate": 9.101278548011959e-06,
+      "loss": 0.8711,
+      "step": 2668
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.3211863555450694,
+      "learning_rate": 9.100521206580219e-06,
+      "loss": 0.6148,
+      "step": 2669
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.636657691789844,
+      "learning_rate": 9.099763577719788e-06,
+      "loss": 1.4335,
+      "step": 2670
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.589107135519192,
+      "learning_rate": 9.099005661483776e-06,
+      "loss": 1.2736,
+      "step": 2671
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.473990332404505,
+      "learning_rate": 9.098247457925304e-06,
+      "loss": 0.9545,
+      "step": 2672
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.984610233768719,
+      "learning_rate": 9.09748896709752e-06,
+      "loss": 0.5687,
+      "step": 2673
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 6.398528093033693,
+      "learning_rate": 9.096730189053594e-06,
+      "loss": 1.6897,
+      "step": 2674
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.561569204175784,
+      "learning_rate": 9.09597112384671e-06,
+      "loss": 0.7358,
+      "step": 2675
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 6.1036865093347785,
+      "learning_rate": 9.095211771530074e-06,
+      "loss": 1.1493,
+      "step": 2676
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3760146537329225,
+      "learning_rate": 9.094452132156917e-06,
+      "loss": 0.2076,
+      "step": 2677
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.098394907191026,
+      "learning_rate": 9.093692205780485e-06,
+      "loss": 0.8469,
+      "step": 2678
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.467224924945489,
+      "learning_rate": 9.092931992454044e-06,
+      "loss": 1.1177,
+      "step": 2679
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.06386919862057,
+      "learning_rate": 9.092171492230883e-06,
+      "loss": 0.9505,
+      "step": 2680
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.5510675849780027,
+      "learning_rate": 9.091410705164312e-06,
+      "loss": 0.4873,
+      "step": 2681
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.255174006776468,
+      "learning_rate": 9.090649631307653e-06,
+      "loss": 1.1771,
+      "step": 2682
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.599542010746827,
+      "learning_rate": 9.08988827071426e-06,
+      "loss": 0.8635,
+      "step": 2683
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.673085770773519,
+      "learning_rate": 9.089126623437496e-06,
+      "loss": 1.5034,
+      "step": 2684
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.117986430311266,
+      "learning_rate": 9.088364689530753e-06,
+      "loss": 1.0058,
+      "step": 2685
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.5266863984789554,
+      "learning_rate": 9.087602469047438e-06,
+      "loss": 0.4884,
+      "step": 2686
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.577416855872609,
+      "learning_rate": 9.08683996204098e-06,
+      "loss": 0.9433,
+      "step": 2687
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.629953813099042,
+      "learning_rate": 9.086077168564825e-06,
+      "loss": 1.1885,
+      "step": 2688
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.596494599376898,
+      "learning_rate": 9.085314088672443e-06,
+      "loss": 0.797,
+      "step": 2689
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.020907218922748,
+      "learning_rate": 9.084550722417324e-06,
+      "loss": 1.3782,
+      "step": 2690
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.125555556634683,
+      "learning_rate": 9.083787069852976e-06,
+      "loss": 0.8511,
+      "step": 2691
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.513382843467977,
+      "learning_rate": 9.083023131032926e-06,
+      "loss": 0.6633,
+      "step": 2692
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.869475350461378,
+      "learning_rate": 9.082258906010724e-06,
+      "loss": 0.7274,
+      "step": 2693
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.066640667462265,
+      "learning_rate": 9.081494394839937e-06,
+      "loss": 0.8265,
+      "step": 2694
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.9031623988993083,
+      "learning_rate": 9.080729597574159e-06,
+      "loss": 0.7738,
+      "step": 2695
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.210771030915046,
+      "learning_rate": 9.079964514266993e-06,
+      "loss": 1.0799,
+      "step": 2696
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.746004952544171,
+      "learning_rate": 9.079199144972072e-06,
+      "loss": 0.8076,
+      "step": 2697
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.211666453404126,
+      "learning_rate": 9.078433489743044e-06,
+      "loss": 0.8715,
+      "step": 2698
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.725053966461781,
+      "learning_rate": 9.077667548633576e-06,
+      "loss": 0.7972,
+      "step": 2699
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.9799110669409408,
+      "learning_rate": 9.07690132169736e-06,
+      "loss": 0.6411,
+      "step": 2700
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.3679873723537055,
+      "learning_rate": 9.076134808988104e-06,
+      "loss": 1.417,
+      "step": 2701
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.1280222277258822,
+      "learning_rate": 9.075368010559538e-06,
+      "loss": 0.2649,
+      "step": 2702
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.6689600488903578,
+      "learning_rate": 9.07460092646541e-06,
+      "loss": 0.9392,
+      "step": 2703
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.193899358666189,
+      "learning_rate": 9.073833556759489e-06,
+      "loss": 0.801,
+      "step": 2704
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.847355714154215,
+      "learning_rate": 9.073065901495565e-06,
+      "loss": 1.1343,
+      "step": 2705
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.2529862008053017,
+      "learning_rate": 9.072297960727449e-06,
+      "loss": 0.7906,
+      "step": 2706
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.6042894685662863,
+      "learning_rate": 9.071529734508968e-06,
+      "loss": 0.704,
+      "step": 2707
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.9368839985787356,
+      "learning_rate": 9.070761222893972e-06,
+      "loss": 0.5706,
+      "step": 2708
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.7028095547180655,
+      "learning_rate": 9.06999242593633e-06,
+      "loss": 0.8101,
+      "step": 2709
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.8895209253771186,
+      "learning_rate": 9.06922334368993e-06,
+      "loss": 0.9076,
+      "step": 2710
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.513508298678548,
+      "learning_rate": 9.068453976208685e-06,
+      "loss": 1.2225,
+      "step": 2711
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.7300236102942743,
+      "learning_rate": 9.067684323546522e-06,
+      "loss": 0.5809,
+      "step": 2712
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.737520002700896,
+      "learning_rate": 9.066914385757391e-06,
+      "loss": 0.5947,
+      "step": 2713
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.500085553474255,
+      "learning_rate": 9.066144162895259e-06,
+      "loss": 0.5553,
+      "step": 2714
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.173765965653176,
+      "learning_rate": 9.065373655014118e-06,
+      "loss": 0.9426,
+      "step": 2715
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.0139284654348866,
+      "learning_rate": 9.064602862167978e-06,
+      "loss": 0.9988,
+      "step": 2716
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0631356101532043,
+      "learning_rate": 9.063831784410864e-06,
+      "loss": 0.233,
+      "step": 2717
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.3993921934621922,
+      "learning_rate": 9.06306042179683e-06,
+      "loss": 0.6685,
+      "step": 2718
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.032180623334634,
+      "learning_rate": 9.06228877437994e-06,
+      "loss": 0.5603,
+      "step": 2719
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.139258586810135,
+      "learning_rate": 9.061516842214289e-06,
+      "loss": 1.2063,
+      "step": 2720
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.178060190654128,
+      "learning_rate": 9.060744625353981e-06,
+      "loss": 1.0683,
+      "step": 2721
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.040238592557097,
+      "learning_rate": 9.059972123853147e-06,
+      "loss": 1.5402,
+      "step": 2722
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.555367406515826,
+      "learning_rate": 9.059199337765938e-06,
+      "loss": 1.0049,
+      "step": 2723
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.609256108481837,
+      "learning_rate": 9.05842626714652e-06,
+      "loss": 1.0136,
+      "step": 2724
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.6825562606320705,
+      "learning_rate": 9.057652912049084e-06,
+      "loss": 0.443,
+      "step": 2725
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 6.359237350151824,
+      "learning_rate": 9.056879272527837e-06,
+      "loss": 1.1107,
+      "step": 2726
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.4778726856530744,
+      "learning_rate": 9.05610534863701e-06,
+      "loss": 0.9013,
+      "step": 2727
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.147177221249099,
+      "learning_rate": 9.05533114043085e-06,
+      "loss": 1.0685,
+      "step": 2728
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.9314829273281657,
+      "learning_rate": 9.054556647963624e-06,
+      "loss": 1.0849,
+      "step": 2729
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.4745032042371378,
+      "learning_rate": 9.053781871289624e-06,
+      "loss": 0.7362,
+      "step": 2730
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.777530136657484,
+      "learning_rate": 9.053006810463156e-06,
+      "loss": 1.4634,
+      "step": 2731
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.414922717015416,
+      "learning_rate": 9.05223146553855e-06,
+      "loss": 1.0374,
+      "step": 2732
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.492283894983976,
+      "learning_rate": 9.051455836570154e-06,
+      "loss": 1.0562,
+      "step": 2733
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8660740691181457,
+      "learning_rate": 9.050679923612334e-06,
+      "loss": 0.3383,
+      "step": 2734
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.992589365919628,
+      "learning_rate": 9.049903726719482e-06,
+      "loss": 1.4697,
+      "step": 2735
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.2639832225852814,
+      "learning_rate": 9.049127245946003e-06,
+      "loss": 0.5715,
+      "step": 2736
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.503803510304516,
+      "learning_rate": 9.048350481346326e-06,
+      "loss": 0.793,
+      "step": 2737
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.2856682173269554,
+      "learning_rate": 9.047573432974898e-06,
+      "loss": 0.9469,
+      "step": 2738
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.985965444532201,
+      "learning_rate": 9.04679610088619e-06,
+      "loss": 0.3888,
+      "step": 2739
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.433476556382269,
+      "learning_rate": 9.046018485134684e-06,
+      "loss": 1.0396,
+      "step": 2740
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.1370696305822245,
+      "learning_rate": 9.045240585774893e-06,
+      "loss": 0.5518,
+      "step": 2741
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.337055499474134,
+      "learning_rate": 9.04446240286134e-06,
+      "loss": 0.7675,
+      "step": 2742
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.000323836971652,
+      "learning_rate": 9.043683936448576e-06,
+      "loss": 0.4838,
+      "step": 2743
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.88456891213328,
+      "learning_rate": 9.042905186591165e-06,
+      "loss": 1.1856,
+      "step": 2744
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.4318484148804034,
+      "learning_rate": 9.042126153343696e-06,
+      "loss": 0.6447,
+      "step": 2745
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.4363596168029864,
+      "learning_rate": 9.041346836760774e-06,
+      "loss": 0.6539,
+      "step": 2746
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.8489378324494146,
+      "learning_rate": 9.040567236897027e-06,
+      "loss": 0.6987,
+      "step": 2747
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.333163031387533,
+      "learning_rate": 9.039787353807101e-06,
+      "loss": 0.8948,
+      "step": 2748
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.691612764942628,
+      "learning_rate": 9.039007187545663e-06,
+      "loss": 1.2254,
+      "step": 2749
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.395036206555651,
+      "learning_rate": 9.0382267381674e-06,
+      "loss": 1.1411,
+      "step": 2750
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.264749249109578,
+      "learning_rate": 9.037446005727015e-06,
+      "loss": 0.8549,
+      "step": 2751
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.3635051911212206,
+      "learning_rate": 9.036664990279238e-06,
+      "loss": 0.6115,
+      "step": 2752
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.539616537868484,
+      "learning_rate": 9.035883691878811e-06,
+      "loss": 0.627,
+      "step": 2753
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.6299450090948415,
+      "learning_rate": 9.035102110580503e-06,
+      "loss": 0.5744,
+      "step": 2754
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.660952668616124,
+      "learning_rate": 9.034320246439099e-06,
+      "loss": 0.5867,
+      "step": 2755
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.7375965940589637,
+      "learning_rate": 9.0335380995094e-06,
+      "loss": 0.7727,
+      "step": 2756
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.536191405126242,
+      "learning_rate": 9.032755669846236e-06,
+      "loss": 1.3356,
+      "step": 2757
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.17031370829871,
+      "learning_rate": 9.03197295750445e-06,
+      "loss": 1.2032,
+      "step": 2758
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.026485263691201,
+      "learning_rate": 9.031189962538906e-06,
+      "loss": 0.837,
+      "step": 2759
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.423354422955292,
+      "learning_rate": 9.030406685004491e-06,
+      "loss": 0.9258,
+      "step": 2760
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.055087678512803,
+      "learning_rate": 9.029623124956107e-06,
+      "loss": 0.6416,
+      "step": 2761
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.190891435470464,
+      "learning_rate": 9.028839282448678e-06,
+      "loss": 0.5773,
+      "step": 2762
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.819205835772208,
+      "learning_rate": 9.02805515753715e-06,
+      "loss": 0.6063,
+      "step": 2763
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.024035874945257,
+      "learning_rate": 9.027270750276486e-06,
+      "loss": 0.5309,
+      "step": 2764
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.9569835244011022,
+      "learning_rate": 9.026486060721668e-06,
+      "loss": 0.521,
+      "step": 2765
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.359550607675502,
+      "learning_rate": 9.0257010889277e-06,
+      "loss": 1.5101,
+      "step": 2766
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.8025381643603797,
+      "learning_rate": 9.024915834949607e-06,
+      "loss": 1.0569,
+      "step": 2767
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.5427223532062015,
+      "learning_rate": 9.02413029884243e-06,
+      "loss": 1.044,
+      "step": 2768
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.606930909658853,
+      "learning_rate": 9.023344480661234e-06,
+      "loss": 0.8584,
+      "step": 2769
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.4269758171149345,
+      "learning_rate": 9.022558380461097e-06,
+      "loss": 1.2617,
+      "step": 2770
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.947679373528407,
+      "learning_rate": 9.021771998297124e-06,
+      "loss": 0.7401,
+      "step": 2771
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.6036071767627855,
+      "learning_rate": 9.020985334224437e-06,
+      "loss": 0.987,
+      "step": 2772
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.958509943303449,
+      "learning_rate": 9.020198388298179e-06,
+      "loss": 1.4065,
+      "step": 2773
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.673660488009326,
+      "learning_rate": 9.019411160573508e-06,
+      "loss": 0.7963,
+      "step": 2774
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.650622270596607,
+      "learning_rate": 9.018623651105607e-06,
+      "loss": 0.5745,
+      "step": 2775
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.216886938088694,
+      "learning_rate": 9.017835859949677e-06,
+      "loss": 0.8929,
+      "step": 2776
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8007191532303202,
+      "learning_rate": 9.01704778716094e-06,
+      "loss": 0.3384,
+      "step": 2777
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.657644897341523,
+      "learning_rate": 9.016259432794637e-06,
+      "loss": 0.811,
+      "step": 2778
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.512963732012924,
+      "learning_rate": 9.015470796906024e-06,
+      "loss": 1.1201,
+      "step": 2779
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.010832411307142,
+      "learning_rate": 9.014681879550385e-06,
+      "loss": 0.7525,
+      "step": 2780
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.486459322409384,
+      "learning_rate": 9.013892680783016e-06,
+      "loss": 1.0253,
+      "step": 2781
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.2106168850862566,
+      "learning_rate": 9.01310320065924e-06,
+      "loss": 0.8443,
+      "step": 2782
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.4301667912614193,
+      "learning_rate": 9.012313439234395e-06,
+      "loss": 0.637,
+      "step": 2783
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.212204675584937,
+      "learning_rate": 9.011523396563839e-06,
+      "loss": 0.8935,
+      "step": 2784
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.044338879716831,
+      "learning_rate": 9.010733072702952e-06,
+      "loss": 1.2117,
+      "step": 2785
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.859085867712923,
+      "learning_rate": 9.009942467707131e-06,
+      "loss": 0.5685,
+      "step": 2786
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.7258987707406455,
+      "learning_rate": 9.009151581631795e-06,
+      "loss": 0.8138,
+      "step": 2787
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.3186899853149128,
+      "learning_rate": 9.00836041453238e-06,
+      "loss": 0.5666,
+      "step": 2788
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.6209517043383737,
+      "learning_rate": 9.007568966464345e-06,
+      "loss": 0.6105,
+      "step": 2789
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.3931208874085,
+      "learning_rate": 9.006777237483165e-06,
+      "loss": 0.4942,
+      "step": 2790
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.544807705087439,
+      "learning_rate": 9.00598522764434e-06,
+      "loss": 0.7926,
+      "step": 2791
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.005000496520486,
+      "learning_rate": 9.005192937003384e-06,
+      "loss": 0.7464,
+      "step": 2792
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.872935081002712,
+      "learning_rate": 9.004400365615834e-06,
+      "loss": 1.3192,
+      "step": 2793
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.7811303859401684,
+      "learning_rate": 9.003607513537245e-06,
+      "loss": 0.7219,
+      "step": 2794
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.354983640897277,
+      "learning_rate": 9.002814380823192e-06,
+      "loss": 0.757,
+      "step": 2795
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.254025619919019,
+      "learning_rate": 9.002020967529272e-06,
+      "loss": 0.7183,
+      "step": 2796
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.057649153469898,
+      "learning_rate": 9.0012272737111e-06,
+      "loss": 1.0508,
+      "step": 2797
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.561540618546092,
+      "learning_rate": 9.000433299424308e-06,
+      "loss": 0.5899,
+      "step": 2798
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.940995656792812,
+      "learning_rate": 8.999639044724555e-06,
+      "loss": 0.9293,
+      "step": 2799
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.6829296806974505,
+      "learning_rate": 8.998844509667508e-06,
+      "loss": 0.6437,
+      "step": 2800
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.433040218782608,
+      "learning_rate": 8.998049694308866e-06,
+      "loss": 0.6815,
+      "step": 2801
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.4827796949727086,
+      "learning_rate": 8.99725459870434e-06,
+      "loss": 0.8859,
+      "step": 2802
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.9475617906007505,
+      "learning_rate": 8.996459222909662e-06,
+      "loss": 1.2561,
+      "step": 2803
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.340861492907865,
+      "learning_rate": 8.995663566980583e-06,
+      "loss": 0.4739,
+      "step": 2804
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.983847979765691,
+      "learning_rate": 8.994867630972878e-06,
+      "loss": 0.9867,
+      "step": 2805
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.4039614230101574,
+      "learning_rate": 8.99407141494234e-06,
+      "loss": 0.7996,
+      "step": 2806
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.665702582646373,
+      "learning_rate": 8.993274918944777e-06,
+      "loss": 1.0746,
+      "step": 2807
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.010574700798193,
+      "learning_rate": 8.992478143036022e-06,
+      "loss": 0.1999,
+      "step": 2808
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.811328637907539,
+      "learning_rate": 8.991681087271922e-06,
+      "loss": 0.9865,
+      "step": 2809
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.8568309546794937,
+      "learning_rate": 8.990883751708353e-06,
+      "loss": 0.812,
+      "step": 2810
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.1161826007469795,
+      "learning_rate": 8.990086136401199e-06,
+      "loss": 0.6223,
+      "step": 2811
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.456885999014682,
+      "learning_rate": 8.989288241406371e-06,
+      "loss": 0.8013,
+      "step": 2812
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.374121412700305,
+      "learning_rate": 8.9884900667798e-06,
+      "loss": 0.4702,
+      "step": 2813
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.5488691936882972,
+      "learning_rate": 8.987691612577433e-06,
+      "loss": 0.9326,
+      "step": 2814
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.828785058948604,
+      "learning_rate": 8.986892878855238e-06,
+      "loss": 1.2953,
+      "step": 2815
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.465452757305021,
+      "learning_rate": 8.986093865669205e-06,
+      "loss": 0.8868,
+      "step": 2816
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.891282308557172,
+      "learning_rate": 8.985294573075338e-06,
+      "loss": 1.0112,
+      "step": 2817
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.890800750650286,
+      "learning_rate": 8.984495001129667e-06,
+      "loss": 0.9078,
+      "step": 2818
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 6.05677774070146,
+      "learning_rate": 8.983695149888234e-06,
+      "loss": 1.4547,
+      "step": 2819
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.250995322790718,
+      "learning_rate": 8.982895019407112e-06,
+      "loss": 1.038,
+      "step": 2820
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.0724656980668823,
+      "learning_rate": 8.98209460974238e-06,
+      "loss": 0.7388,
+      "step": 2821
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.340758261958266,
+      "learning_rate": 8.981293920950147e-06,
+      "loss": 1.0972,
+      "step": 2822
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.462736754135804,
+      "learning_rate": 8.980492953086535e-06,
+      "loss": 0.5185,
+      "step": 2823
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.2096517402684297,
+      "learning_rate": 8.979691706207692e-06,
+      "loss": 0.6595,
+      "step": 2824
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.346508184568805,
+      "learning_rate": 8.97889018036978e-06,
+      "loss": 0.9841,
+      "step": 2825
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.848452650091313,
+      "learning_rate": 8.978088375628983e-06,
+      "loss": 0.7972,
+      "step": 2826
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.526629528918161,
+      "learning_rate": 8.977286292041503e-06,
+      "loss": 0.9327,
+      "step": 2827
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.9604609578722085,
+      "learning_rate": 8.976483929663562e-06,
+      "loss": 0.5696,
+      "step": 2828
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.259698795083026,
+      "learning_rate": 8.975681288551405e-06,
+      "loss": 0.8955,
+      "step": 2829
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.713873183280796,
+      "learning_rate": 8.97487836876129e-06,
+      "loss": 0.3443,
+      "step": 2830
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.466880037688497,
+      "learning_rate": 8.974075170349502e-06,
+      "loss": 0.6895,
+      "step": 2831
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.5967277234419393,
+      "learning_rate": 8.973271693372338e-06,
+      "loss": 0.411,
+      "step": 2832
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.029576229429126,
+      "learning_rate": 8.972467937886122e-06,
+      "loss": 0.733,
+      "step": 2833
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9498422274837677,
+      "learning_rate": 8.971663903947191e-06,
+      "loss": 0.3991,
+      "step": 2834
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.3772068633274634,
+      "learning_rate": 8.970859591611904e-06,
+      "loss": 0.5426,
+      "step": 2835
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.5590228128457717,
+      "learning_rate": 8.970055000936643e-06,
+      "loss": 0.9203,
+      "step": 2836
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.6475066147276056,
+      "learning_rate": 8.969250131977803e-06,
+      "loss": 0.9788,
+      "step": 2837
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.114364170867909,
+      "learning_rate": 8.968444984791801e-06,
+      "loss": 0.7102,
+      "step": 2838
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.201567434821648,
+      "learning_rate": 8.967639559435079e-06,
+      "loss": 0.7126,
+      "step": 2839
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.8625761671129175,
+      "learning_rate": 8.96683385596409e-06,
+      "loss": 1.3315,
+      "step": 2840
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.57336965684408,
+      "learning_rate": 8.966027874435313e-06,
+      "loss": 1.3707,
+      "step": 2841
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.4775308239244853,
+      "learning_rate": 8.965221614905241e-06,
+      "loss": 0.6265,
+      "step": 2842
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.850705463406801,
+      "learning_rate": 8.964415077430391e-06,
+      "loss": 0.6261,
+      "step": 2843
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.021140145526666,
+      "learning_rate": 8.963608262067296e-06,
+      "loss": 0.4525,
+      "step": 2844
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.147160395871126,
+      "learning_rate": 8.962801168872513e-06,
+      "loss": 0.9202,
+      "step": 2845
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6818190008786265,
+      "learning_rate": 8.961993797902613e-06,
+      "loss": 0.2499,
+      "step": 2846
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.281047026662951,
+      "learning_rate": 8.961186149214191e-06,
+      "loss": 0.8977,
+      "step": 2847
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.101860188112433,
+      "learning_rate": 8.96037822286386e-06,
+      "loss": 0.6484,
+      "step": 2848
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.623832576872076,
+      "learning_rate": 8.959570018908248e-06,
+      "loss": 0.6869,
+      "step": 2849
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.0541378484556665,
+      "learning_rate": 8.958761537404012e-06,
+      "loss": 0.7,
+      "step": 2850
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.779380648621428,
+      "learning_rate": 8.957952778407822e-06,
+      "loss": 0.6621,
+      "step": 2851
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.634388934331967,
+      "learning_rate": 8.957143741976366e-06,
+      "loss": 1.0359,
+      "step": 2852
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.745980430150055,
+      "learning_rate": 8.956334428166355e-06,
+      "loss": 1.0663,
+      "step": 2853
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0527854010053523,
+      "learning_rate": 8.95552483703452e-06,
+      "loss": 0.1956,
+      "step": 2854
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.2624856816156536,
+      "learning_rate": 8.954714968637606e-06,
+      "loss": 0.8258,
+      "step": 2855
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.8966623314578803,
+      "learning_rate": 8.953904823032384e-06,
+      "loss": 0.9205,
+      "step": 2856
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.640423997502576,
+      "learning_rate": 8.953094400275644e-06,
+      "loss": 1.499,
+      "step": 2857
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.986840018375436,
+      "learning_rate": 8.952283700424188e-06,
+      "loss": 1.0383,
+      "step": 2858
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.606151699801707,
+      "learning_rate": 8.951472723534846e-06,
+      "loss": 1.1475,
+      "step": 2859
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.788270601765476,
+      "learning_rate": 8.950661469664462e-06,
+      "loss": 1.0722,
+      "step": 2860
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.836919142856838,
+      "learning_rate": 8.949849938869904e-06,
+      "loss": 0.9889,
+      "step": 2861
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.563519527237157,
+      "learning_rate": 8.949038131208054e-06,
+      "loss": 0.8877,
+      "step": 2862
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.721883252612204,
+      "learning_rate": 8.948226046735817e-06,
+      "loss": 0.5332,
+      "step": 2863
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.448798099577989,
+      "learning_rate": 8.947413685510118e-06,
+      "loss": 1.032,
+      "step": 2864
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.160199095042135,
+      "learning_rate": 8.946601047587898e-06,
+      "loss": 0.5859,
+      "step": 2865
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.415316077485991,
+      "learning_rate": 8.94578813302612e-06,
+      "loss": 1.6669,
+      "step": 2866
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.28215722241667,
+      "learning_rate": 8.944974941881766e-06,
+      "loss": 0.883,
+      "step": 2867
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.2851793502192725,
+      "learning_rate": 8.94416147421184e-06,
+      "loss": 0.9386,
+      "step": 2868
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7798034305693526,
+      "learning_rate": 8.943347730073355e-06,
+      "loss": 0.3629,
+      "step": 2869
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.029526509580939,
+      "learning_rate": 8.942533709523358e-06,
+      "loss": 1.2032,
+      "step": 2870
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.715867193032858,
+      "learning_rate": 8.941719412618905e-06,
+      "loss": 0.2987,
+      "step": 2871
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.280487822475332,
+      "learning_rate": 8.940904839417078e-06,
+      "loss": 0.7824,
+      "step": 2872
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.987478844312189,
+      "learning_rate": 8.94008998997497e-06,
+      "loss": 0.7334,
+      "step": 2873
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.791659519165454,
+      "learning_rate": 8.939274864349702e-06,
+      "loss": 0.9929,
+      "step": 2874
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.081963403891954,
+      "learning_rate": 8.93845946259841e-06,
+      "loss": 1.318,
+      "step": 2875
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.7244126800865307,
+      "learning_rate": 8.93764378477825e-06,
+      "loss": 0.9154,
+      "step": 2876
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.4485265914698013,
+      "learning_rate": 8.936827830946397e-06,
+      "loss": 0.7187,
+      "step": 2877
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.302352810307197,
+      "learning_rate": 8.936011601160046e-06,
+      "loss": 0.7165,
+      "step": 2878
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.303497388254325,
+      "learning_rate": 8.935195095476412e-06,
+      "loss": 0.9351,
+      "step": 2879
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.9582810507637536,
+      "learning_rate": 8.934378313952727e-06,
+      "loss": 0.8128,
+      "step": 2880
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.63455467974814,
+      "learning_rate": 8.933561256646247e-06,
+      "loss": 0.599,
+      "step": 2881
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.817216225830816,
+      "learning_rate": 8.932743923614237e-06,
+      "loss": 0.8345,
+      "step": 2882
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.2788123295098903,
+      "learning_rate": 8.931926314913998e-06,
+      "loss": 0.495,
+      "step": 2883
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9451263961915763,
+      "learning_rate": 8.931108430602834e-06,
+      "loss": 0.3557,
+      "step": 2884
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.674895514486894,
+      "learning_rate": 8.930290270738079e-06,
+      "loss": 0.8228,
+      "step": 2885
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.7542288766120713,
+      "learning_rate": 8.929471835377078e-06,
+      "loss": 0.4755,
+      "step": 2886
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.057165314865615,
+      "learning_rate": 8.928653124577204e-06,
+      "loss": 1.569,
+      "step": 2887
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.9347074343546335,
+      "learning_rate": 8.927834138395843e-06,
+      "loss": 0.6106,
+      "step": 2888
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.248265115238616,
+      "learning_rate": 8.927014876890402e-06,
+      "loss": 1.0259,
+      "step": 2889
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.6666573587433353,
+      "learning_rate": 8.926195340118312e-06,
+      "loss": 0.4648,
+      "step": 2890
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.867031223985313,
+      "learning_rate": 8.925375528137012e-06,
+      "loss": 1.0118,
+      "step": 2891
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.6026790307902745,
+      "learning_rate": 8.924555441003973e-06,
+      "loss": 0.8052,
+      "step": 2892
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.857068725782363,
+      "learning_rate": 8.923735078776676e-06,
+      "loss": 1.1488,
+      "step": 2893
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.9381131716176174,
+      "learning_rate": 8.922914441512626e-06,
+      "loss": 0.3417,
+      "step": 2894
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4132180157242291,
+      "learning_rate": 8.922093529269347e-06,
+      "loss": 0.1973,
+      "step": 2895
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.9565676205158287,
+      "learning_rate": 8.921272342104382e-06,
+      "loss": 1.0563,
+      "step": 2896
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.9575448163406683,
+      "learning_rate": 8.92045088007529e-06,
+      "loss": 1.0864,
+      "step": 2897
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.0103830571098324,
+      "learning_rate": 8.919629143239652e-06,
+      "loss": 0.8211,
+      "step": 2898
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.6310097637805145,
+      "learning_rate": 8.918807131655069e-06,
+      "loss": 0.8162,
+      "step": 2899
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.6579277723192414,
+      "learning_rate": 8.917984845379162e-06,
+      "loss": 0.765,
+      "step": 2900
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.447466899662024,
+      "learning_rate": 8.917162284469569e-06,
+      "loss": 0.8084,
+      "step": 2901
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.345413652998398,
+      "learning_rate": 8.916339448983945e-06,
+      "loss": 1.2272,
+      "step": 2902
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.388461703716796,
+      "learning_rate": 8.915516338979973e-06,
+      "loss": 0.6305,
+      "step": 2903
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.687060236470035,
+      "learning_rate": 8.914692954515344e-06,
+      "loss": 0.4097,
+      "step": 2904
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.6100909371539154,
+      "learning_rate": 8.913869295647777e-06,
+      "loss": 0.7575,
+      "step": 2905
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.97719742428522,
+      "learning_rate": 8.913045362435004e-06,
+      "loss": 0.938,
+      "step": 2906
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.1183558522117183,
+      "learning_rate": 8.91222115493478e-06,
+      "loss": 0.7863,
+      "step": 2907
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.9833914186492243,
+      "learning_rate": 8.911396673204881e-06,
+      "loss": 0.5393,
+      "step": 2908
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.16799800442453,
+      "learning_rate": 8.910571917303098e-06,
+      "loss": 0.7233,
+      "step": 2909
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.8362192782466056,
+      "learning_rate": 8.90974688728724e-06,
+      "loss": 0.9734,
+      "step": 2910
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.6294280620231945,
+      "learning_rate": 8.908921583215143e-06,
+      "loss": 1.0333,
+      "step": 2911
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.8164965009412533,
+      "learning_rate": 8.908096005144654e-06,
+      "loss": 0.7351,
+      "step": 2912
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.405646643100329,
+      "learning_rate": 8.907270153133643e-06,
+      "loss": 1.1334,
+      "step": 2913
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.044455880720274,
+      "learning_rate": 8.906444027239999e-06,
+      "loss": 1.2822,
+      "step": 2914
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.3033739123756076,
+      "learning_rate": 8.90561762752163e-06,
+      "loss": 0.5727,
+      "step": 2915
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.408990250089634,
+      "learning_rate": 8.90479095403646e-06,
+      "loss": 0.7527,
+      "step": 2916
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.810162697001605,
+      "learning_rate": 8.90396400684244e-06,
+      "loss": 1.1846,
+      "step": 2917
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.771676233576919,
+      "learning_rate": 8.903136785997533e-06,
+      "loss": 1.3764,
+      "step": 2918
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.133198302514898,
+      "learning_rate": 8.902309291559724e-06,
+      "loss": 1.001,
+      "step": 2919
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.1451373989928895,
+      "learning_rate": 8.901481523587017e-06,
+      "loss": 0.9517,
+      "step": 2920
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.9171169131457897,
+      "learning_rate": 8.900653482137434e-06,
+      "loss": 0.7557,
+      "step": 2921
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.120715791842282,
+      "learning_rate": 8.899825167269016e-06,
+      "loss": 0.5538,
+      "step": 2922
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.0165744064583526,
+      "learning_rate": 8.898996579039829e-06,
+      "loss": 0.783,
+      "step": 2923
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.9787819731271474,
+      "learning_rate": 8.898167717507949e-06,
+      "loss": 0.3642,
+      "step": 2924
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.9979649908032515,
+      "learning_rate": 8.897338582731476e-06,
+      "loss": 0.694,
+      "step": 2925
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.465345306276673,
+      "learning_rate": 8.896509174768528e-06,
+      "loss": 1.2069,
+      "step": 2926
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.001986170825165,
+      "learning_rate": 8.895679493677247e-06,
+      "loss": 0.5428,
+      "step": 2927
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.970549933403323,
+      "learning_rate": 8.894849539515788e-06,
+      "loss": 0.5073,
+      "step": 2928
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.581485075523803,
+      "learning_rate": 8.894019312342325e-06,
+      "loss": 0.5156,
+      "step": 2929
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.758865803490577,
+      "learning_rate": 8.893188812215057e-06,
+      "loss": 1.1096,
+      "step": 2930
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.242292637061147,
+      "learning_rate": 8.892358039192196e-06,
+      "loss": 0.9125,
+      "step": 2931
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.851855379448961,
+      "learning_rate": 8.891526993331974e-06,
+      "loss": 0.658,
+      "step": 2932
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.7730340149059307,
+      "learning_rate": 8.890695674692648e-06,
+      "loss": 0.839,
+      "step": 2933
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.593971328906433,
+      "learning_rate": 8.889864083332486e-06,
+      "loss": 0.6572,
+      "step": 2934
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.509712803796424,
+      "learning_rate": 8.889032219309781e-06,
+      "loss": 0.2217,
+      "step": 2935
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.189866905519617,
+      "learning_rate": 8.888200082682842e-06,
+      "loss": 1.0506,
+      "step": 2936
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.721479794619482,
+      "learning_rate": 8.887367673510002e-06,
+      "loss": 1.4691,
+      "step": 2937
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.901042227359208,
+      "learning_rate": 8.886534991849603e-06,
+      "loss": 1.2325,
+      "step": 2938
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6920310456009662,
+      "learning_rate": 8.885702037760016e-06,
+      "loss": 0.2909,
+      "step": 2939
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.1842414345062067,
+      "learning_rate": 8.884868811299627e-06,
+      "loss": 0.858,
+      "step": 2940
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.217988813255178,
+      "learning_rate": 8.88403531252684e-06,
+      "loss": 0.7523,
+      "step": 2941
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.585574617111233,
+      "learning_rate": 8.883201541500082e-06,
+      "loss": 1.0042,
+      "step": 2942
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.334108757662184,
+      "learning_rate": 8.882367498277795e-06,
+      "loss": 1.034,
+      "step": 2943
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.8751675188151795,
+      "learning_rate": 8.881533182918444e-06,
+      "loss": 0.7488,
+      "step": 2944
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.5142236100060154,
+      "learning_rate": 8.880698595480509e-06,
+      "loss": 0.6338,
+      "step": 2945
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.212057409312506,
+      "learning_rate": 8.87986373602249e-06,
+      "loss": 1.2324,
+      "step": 2946
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.402148710685035,
+      "learning_rate": 8.879028604602908e-06,
+      "loss": 1.2733,
+      "step": 2947
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.195936383889245,
+      "learning_rate": 8.878193201280305e-06,
+      "loss": 0.8655,
+      "step": 2948
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.733486040762455,
+      "learning_rate": 8.877357526113234e-06,
+      "loss": 0.5531,
+      "step": 2949
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.142029788504823,
+      "learning_rate": 8.876521579160275e-06,
+      "loss": 0.4354,
+      "step": 2950
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.48630056663649,
+      "learning_rate": 8.875685360480027e-06,
+      "loss": 1.3183,
+      "step": 2951
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.994076738007516,
+      "learning_rate": 8.874848870131098e-06,
+      "loss": 1.7964,
+      "step": 2952
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.7150255903944416,
+      "learning_rate": 8.87401210817213e-06,
+      "loss": 0.5186,
+      "step": 2953
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.9136979849392772,
+      "learning_rate": 8.87317507466177e-06,
+      "loss": 0.4953,
+      "step": 2954
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.3771356342933405,
+      "learning_rate": 8.872337769658696e-06,
+      "loss": 0.991,
+      "step": 2955
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8083950829827165,
+      "learning_rate": 8.871500193221596e-06,
+      "loss": 0.3512,
+      "step": 2956
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.5684383495880447,
+      "learning_rate": 8.87066234540918e-06,
+      "loss": 0.7557,
+      "step": 2957
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.452996042800122,
+      "learning_rate": 8.86982422628018e-06,
+      "loss": 0.9597,
+      "step": 2958
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.9979832047022694,
+      "learning_rate": 8.868985835893344e-06,
+      "loss": 0.6404,
+      "step": 2959
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.537443193679681,
+      "learning_rate": 8.868147174307438e-06,
+      "loss": 1.2665,
+      "step": 2960
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.3265675796338345,
+      "learning_rate": 8.86730824158125e-06,
+      "loss": 0.648,
+      "step": 2961
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.830979842557535,
+      "learning_rate": 8.866469037773582e-06,
+      "loss": 1.4799,
+      "step": 2962
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.010278861649628,
+      "learning_rate": 8.865629562943265e-06,
+      "loss": 0.3216,
+      "step": 2963
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.616138757300956,
+      "learning_rate": 8.864789817149137e-06,
+      "loss": 0.7531,
+      "step": 2964
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.790458959821874,
+      "learning_rate": 8.863949800450063e-06,
+      "loss": 1.4917,
+      "step": 2965
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.9789882643523837,
+      "learning_rate": 8.863109512904924e-06,
+      "loss": 0.7569,
+      "step": 2966
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.5722174289876714,
+      "learning_rate": 8.862268954572618e-06,
+      "loss": 0.6274,
+      "step": 2967
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.0627508713733618,
+      "learning_rate": 8.861428125512071e-06,
+      "loss": 0.6932,
+      "step": 2968
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.5744792073861236,
+      "learning_rate": 8.860587025782215e-06,
+      "loss": 0.4992,
+      "step": 2969
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.43717337203305,
+      "learning_rate": 8.859745655442012e-06,
+      "loss": 0.503,
+      "step": 2970
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.287648098546942,
+      "learning_rate": 8.858904014550434e-06,
+      "loss": 1.156,
+      "step": 2971
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.6511168741579314,
+      "learning_rate": 8.858062103166479e-06,
+      "loss": 0.817,
+      "step": 2972
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.5830778351070243,
+      "learning_rate": 8.85721992134916e-06,
+      "loss": 1.1603,
+      "step": 2973
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.135773229473065,
+      "learning_rate": 8.856377469157513e-06,
+      "loss": 0.3204,
+      "step": 2974
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.242353578298476,
+      "learning_rate": 8.855534746650586e-06,
+      "loss": 0.4531,
+      "step": 2975
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.21442194749558,
+      "learning_rate": 8.854691753887455e-06,
+      "loss": 0.9405,
+      "step": 2976
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.836501269058079,
+      "learning_rate": 8.853848490927207e-06,
+      "loss": 0.8754,
+      "step": 2977
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.5125849385426475,
+      "learning_rate": 8.853004957828952e-06,
+      "loss": 0.5242,
+      "step": 2978
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.7242216451581576,
+      "learning_rate": 8.852161154651817e-06,
+      "loss": 0.4244,
+      "step": 2979
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.802380953813256,
+      "learning_rate": 8.851317081454951e-06,
+      "loss": 0.9277,
+      "step": 2980
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.049767752733464,
+      "learning_rate": 8.850472738297518e-06,
+      "loss": 0.6299,
+      "step": 2981
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.7215201474495654,
+      "learning_rate": 8.849628125238703e-06,
+      "loss": 0.7794,
+      "step": 2982
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.735496235244078,
+      "learning_rate": 8.848783242337711e-06,
+      "loss": 1.0313,
+      "step": 2983
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.8805762816472638,
+      "learning_rate": 8.847938089653763e-06,
+      "loss": 0.5554,
+      "step": 2984
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2182610153442475,
+      "learning_rate": 8.847092667246101e-06,
+      "loss": 0.1831,
+      "step": 2985
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.886218186327842,
+      "learning_rate": 8.846246975173985e-06,
+      "loss": 1.7733,
+      "step": 2986
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.01633070745938,
+      "learning_rate": 8.845401013496697e-06,
+      "loss": 0.5916,
+      "step": 2987
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.820860752067593,
+      "learning_rate": 8.84455478227353e-06,
+      "loss": 1.1315,
+      "step": 2988
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.0352922168496095,
+      "learning_rate": 8.843708281563808e-06,
+      "loss": 1.539,
+      "step": 2989
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.562737207595621,
+      "learning_rate": 8.842861511426862e-06,
+      "loss": 0.8672,
+      "step": 2990
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.0376963816397025,
+      "learning_rate": 8.842014471922046e-06,
+      "loss": 1.1109,
+      "step": 2991
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.132946137148141,
+      "learning_rate": 8.84116716310874e-06,
+      "loss": 1.2181,
+      "step": 2992
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.607972489288798,
+      "learning_rate": 8.84031958504633e-06,
+      "loss": 0.3004,
+      "step": 2993
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.1621884343551834,
+      "learning_rate": 8.839471737794232e-06,
+      "loss": 0.5174,
+      "step": 2994
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.320221171032136,
+      "learning_rate": 8.838623621411871e-06,
+      "loss": 0.5463,
+      "step": 2995
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.7088125465197876,
+      "learning_rate": 8.837775235958704e-06,
+      "loss": 0.5495,
+      "step": 2996
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.8556123789911116,
+      "learning_rate": 8.836926581494191e-06,
+      "loss": 0.8726,
+      "step": 2997
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.258349986279272,
+      "learning_rate": 8.836077658077825e-06,
+      "loss": 1.0929,
+      "step": 2998
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.886977236391661,
+      "learning_rate": 8.835228465769113e-06,
+      "loss": 1.1653,
+      "step": 2999
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.1490792321452905,
+      "learning_rate": 8.834379004627572e-06,
+      "loss": 0.5612,
+      "step": 3000
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.537597675669336,
+      "learning_rate": 8.833529274712751e-06,
+      "loss": 0.8386,
+      "step": 3001
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.7338484716835243,
+      "learning_rate": 8.832679276084213e-06,
+      "loss": 0.5664,
+      "step": 3002
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.834281332933508,
+      "learning_rate": 8.831829008801536e-06,
+      "loss": 0.8881,
+      "step": 3003
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.8698804980984405,
+      "learning_rate": 8.830978472924323e-06,
+      "loss": 0.8899,
+      "step": 3004
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.6275615582211356,
+      "learning_rate": 8.830127668512191e-06,
+      "loss": 0.8097,
+      "step": 3005
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.176712118230684,
+      "learning_rate": 8.829276595624778e-06,
+      "loss": 0.7396,
+      "step": 3006
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.450354431030108,
+      "learning_rate": 8.828425254321742e-06,
+      "loss": 0.8398,
+      "step": 3007
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.842755616061428,
+      "learning_rate": 8.827573644662756e-06,
+      "loss": 0.8975,
+      "step": 3008
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.422657524771481,
+      "learning_rate": 8.826721766707514e-06,
+      "loss": 1.2179,
+      "step": 3009
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.9833365920739645,
+      "learning_rate": 8.825869620515732e-06,
+      "loss": 0.5358,
+      "step": 3010
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.879167488092526,
+      "learning_rate": 8.825017206147139e-06,
+      "loss": 0.7018,
+      "step": 3011
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.355405824595915,
+      "learning_rate": 8.824164523661485e-06,
+      "loss": 0.4947,
+      "step": 3012
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.352312149931795,
+      "learning_rate": 8.823311573118543e-06,
+      "loss": 1.0323,
+      "step": 3013
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.7924492631796953,
+      "learning_rate": 8.822458354578098e-06,
+      "loss": 0.8462,
+      "step": 3014
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.319453911353385,
+      "learning_rate": 8.821604868099957e-06,
+      "loss": 1.1634,
+      "step": 3015
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.4177468723900803,
+      "learning_rate": 8.820751113743948e-06,
+      "loss": 0.4041,
+      "step": 3016
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.6565659236706436,
+      "learning_rate": 8.819897091569911e-06,
+      "loss": 0.8552,
+      "step": 3017
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.6235116887935432,
+      "learning_rate": 8.819042801637715e-06,
+      "loss": 0.5085,
+      "step": 3018
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.061247292842075,
+      "learning_rate": 8.81818824400724e-06,
+      "loss": 0.8747,
+      "step": 3019
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.807535007879584,
+      "learning_rate": 8.817333418738382e-06,
+      "loss": 1.2014,
+      "step": 3020
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.0249120389592665,
+      "learning_rate": 8.816478325891067e-06,
+      "loss": 0.7394,
+      "step": 3021
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.8172430367905124,
+      "learning_rate": 8.815622965525231e-06,
+      "loss": 0.6459,
+      "step": 3022
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.6977222583715434,
+      "learning_rate": 8.814767337700829e-06,
+      "loss": 1.14,
+      "step": 3023
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.945164214725052,
+      "learning_rate": 8.81391144247784e-06,
+      "loss": 0.6226,
+      "step": 3024
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.2322993706947263,
+      "learning_rate": 8.813055279916258e-06,
+      "loss": 0.5395,
+      "step": 3025
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.352276850776771,
+      "learning_rate": 8.812198850076092e-06,
+      "loss": 1.2339,
+      "step": 3026
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.269106853685705,
+      "learning_rate": 8.81134215301738e-06,
+      "loss": 0.3062,
+      "step": 3027
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.073530103864496,
+      "learning_rate": 8.81048518880017e-06,
+      "loss": 1.0182,
+      "step": 3028
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.73778326088764,
+      "learning_rate": 8.80962795748453e-06,
+      "loss": 0.8068,
+      "step": 3029
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.8308118516350893,
+      "learning_rate": 8.80877045913055e-06,
+      "loss": 0.8938,
+      "step": 3030
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.9154835262683707,
+      "learning_rate": 8.80791269379834e-06,
+      "loss": 0.5688,
+      "step": 3031
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.591077187198804,
+      "learning_rate": 8.807054661548019e-06,
+      "loss": 1.3252,
+      "step": 3032
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.6776801143872526,
+      "learning_rate": 8.806196362439734e-06,
+      "loss": 1.0447,
+      "step": 3033
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.2212908361446635,
+      "learning_rate": 8.80533779653365e-06,
+      "loss": 0.7886,
+      "step": 3034
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.1280264324817724,
+      "learning_rate": 8.804478963889948e-06,
+      "loss": 0.7596,
+      "step": 3035
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.078134161585271,
+      "learning_rate": 8.803619864568827e-06,
+      "loss": 0.5037,
+      "step": 3036
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.325914291845112,
+      "learning_rate": 8.802760498630507e-06,
+      "loss": 0.921,
+      "step": 3037
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.079137764634465,
+      "learning_rate": 8.801900866135225e-06,
+      "loss": 0.8713,
+      "step": 3038
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.2472378005541467,
+      "learning_rate": 8.80104096714324e-06,
+      "loss": 0.7648,
+      "step": 3039
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.168245278168702,
+      "learning_rate": 8.800180801714824e-06,
+      "loss": 1.0513,
+      "step": 3040
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.625803081651678,
+      "learning_rate": 8.799320369910273e-06,
+      "loss": 1.5099,
+      "step": 3041
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.16058040833108,
+      "learning_rate": 8.798459671789898e-06,
+      "loss": 0.9635,
+      "step": 3042
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.308757760875945,
+      "learning_rate": 8.79759870741403e-06,
+      "loss": 0.9304,
+      "step": 3043
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.187234819766285,
+      "learning_rate": 8.796737476843023e-06,
+      "loss": 0.4738,
+      "step": 3044
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.226740566813347,
+      "learning_rate": 8.795875980137238e-06,
+      "loss": 0.8863,
+      "step": 3045
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.51213011831428,
+      "learning_rate": 8.79501421735707e-06,
+      "loss": 1.4908,
+      "step": 3046
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.458799182603262,
+      "learning_rate": 8.79415218856292e-06,
+      "loss": 0.4516,
+      "step": 3047
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.033543389356406,
+      "learning_rate": 8.793289893815213e-06,
+      "loss": 0.2532,
+      "step": 3048
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.1917699404856252,
+      "learning_rate": 8.792427333174395e-06,
+      "loss": 0.7104,
+      "step": 3049
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.856959628270824,
+      "learning_rate": 8.791564506700925e-06,
+      "loss": 0.6704,
+      "step": 3050
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.470122110245426,
+      "learning_rate": 8.790701414455283e-06,
+      "loss": 0.9739,
+      "step": 3051
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.3464298401882315,
+      "learning_rate": 8.789838056497969e-06,
+      "loss": 0.8312,
+      "step": 3052
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.007427352617898,
+      "learning_rate": 8.7889744328895e-06,
+      "loss": 1.0857,
+      "step": 3053
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.173821069060673,
+      "learning_rate": 8.788110543690415e-06,
+      "loss": 0.8516,
+      "step": 3054
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.9007425602595327,
+      "learning_rate": 8.787246388961265e-06,
+      "loss": 1.0306,
+      "step": 3055
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.8175766762413605,
+      "learning_rate": 8.786381968762628e-06,
+      "loss": 0.344,
+      "step": 3056
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.714141728356981,
+      "learning_rate": 8.785517283155092e-06,
+      "loss": 1.0833,
+      "step": 3057
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.4920510483411547,
+      "learning_rate": 8.784652332199269e-06,
+      "loss": 0.9534,
+      "step": 3058
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.201057136856587,
+      "learning_rate": 8.783787115955787e-06,
+      "loss": 0.6107,
+      "step": 3059
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.96564825037283,
+      "learning_rate": 8.782921634485297e-06,
+      "loss": 0.5547,
+      "step": 3060
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.1921970486894624,
+      "learning_rate": 8.782055887848462e-06,
+      "loss": 0.6019,
+      "step": 3061
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.4492343149468145,
+      "learning_rate": 8.78118987610597e-06,
+      "loss": 0.8935,
+      "step": 3062
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.7783810161720743,
+      "learning_rate": 8.780323599318524e-06,
+      "loss": 0.8023,
+      "step": 3063
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.9544689623425415,
+      "learning_rate": 8.779457057546844e-06,
+      "loss": 0.7754,
+      "step": 3064
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.225298130462849,
+      "learning_rate": 8.778590250851674e-06,
+      "loss": 1.0747,
+      "step": 3065
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2969341365306017,
+      "learning_rate": 8.777723179293772e-06,
+      "loss": 0.2045,
+      "step": 3066
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.418704482740044,
+      "learning_rate": 8.776855842933915e-06,
+      "loss": 0.6294,
+      "step": 3067
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.8206797529615666,
+      "learning_rate": 8.7759882418329e-06,
+      "loss": 0.6849,
+      "step": 3068
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.607499970750639,
+      "learning_rate": 8.775120376051544e-06,
+      "loss": 0.5084,
+      "step": 3069
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.946479721945572,
+      "learning_rate": 8.774252245650678e-06,
+      "loss": 1.2328,
+      "step": 3070
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.305464030242199,
+      "learning_rate": 8.773383850691155e-06,
+      "loss": 1.336,
+      "step": 3071
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.521782900215057,
+      "learning_rate": 8.772515191233846e-06,
+      "loss": 0.3972,
+      "step": 3072
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.130677228668136,
+      "learning_rate": 8.771646267339641e-06,
+      "loss": 0.7908,
+      "step": 3073
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.3666015604730046,
+      "learning_rate": 8.770777079069446e-06,
+      "loss": 0.9013,
+      "step": 3074
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.1417450167135086,
+      "learning_rate": 8.769907626484189e-06,
+      "loss": 0.8646,
+      "step": 3075
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1039939450121492,
+      "learning_rate": 8.769037909644813e-06,
+      "loss": 0.1571,
+      "step": 3076
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.381687631869932,
+      "learning_rate": 8.768167928612283e-06,
+      "loss": 0.714,
+      "step": 3077
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.028985577124985,
+      "learning_rate": 8.76729768344758e-06,
+      "loss": 1.0472,
+      "step": 3078
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.4696662022938765,
+      "learning_rate": 8.766427174211704e-06,
+      "loss": 0.4696,
+      "step": 3079
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.270044585126113,
+      "learning_rate": 8.765556400965677e-06,
+      "loss": 0.8139,
+      "step": 3080
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.621035925749316,
+      "learning_rate": 8.764685363770534e-06,
+      "loss": 1.0925,
+      "step": 3081
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.025670248824383,
+      "learning_rate": 8.763814062687329e-06,
+      "loss": 0.4951,
+      "step": 3082
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.154694676655994,
+      "learning_rate": 8.762942497777138e-06,
+      "loss": 0.6662,
+      "step": 3083
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.17545478000675,
+      "learning_rate": 8.762070669101054e-06,
+      "loss": 0.9638,
+      "step": 3084
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0777362421774552,
+      "learning_rate": 8.761198576720191e-06,
+      "loss": 0.3896,
+      "step": 3085
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.8485534506189727,
+      "learning_rate": 8.760326220695677e-06,
+      "loss": 0.5518,
+      "step": 3086
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.4570503971498185,
+      "learning_rate": 8.759453601088658e-06,
+      "loss": 0.9435,
+      "step": 3087
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.762459736007317,
+      "learning_rate": 8.758580717960303e-06,
+      "loss": 0.6803,
+      "step": 3088
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.072988444518039,
+      "learning_rate": 8.757707571371795e-06,
+      "loss": 1.2577,
+      "step": 3089
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.0319532056540286,
+      "learning_rate": 8.756834161384344e-06,
+      "loss": 0.6772,
+      "step": 3090
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0531323444970426,
+      "learning_rate": 8.755960488059167e-06,
+      "loss": 0.3901,
+      "step": 3091
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.8435794741136617,
+      "learning_rate": 8.755086551457504e-06,
+      "loss": 1.0897,
+      "step": 3092
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.1029982378400565,
+      "learning_rate": 8.754212351640618e-06,
+      "loss": 0.9717,
+      "step": 3093
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.467093227327127,
+      "learning_rate": 8.753337888669783e-06,
+      "loss": 0.7102,
+      "step": 3094
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.026197241645423,
+      "learning_rate": 8.752463162606297e-06,
+      "loss": 0.6473,
+      "step": 3095
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.915327712366391,
+      "learning_rate": 8.751588173511474e-06,
+      "loss": 1.1337,
+      "step": 3096
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.9382618749318565,
+      "learning_rate": 8.750712921446647e-06,
+      "loss": 1.0953,
+      "step": 3097
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.2741886506684112,
+      "learning_rate": 8.749837406473168e-06,
+      "loss": 0.7295,
+      "step": 3098
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.701676756960232,
+      "learning_rate": 8.748961628652406e-06,
+      "loss": 0.7643,
+      "step": 3099
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.241600773286563,
+      "learning_rate": 8.74808558804575e-06,
+      "loss": 1.0252,
+      "step": 3100
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.367775399977547,
+      "learning_rate": 8.747209284714604e-06,
+      "loss": 0.6707,
+      "step": 3101
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.5932625088710335,
+      "learning_rate": 8.746332718720395e-06,
+      "loss": 1.2799,
+      "step": 3102
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.2022569176912743,
+      "learning_rate": 8.745455890124567e-06,
+      "loss": 0.803,
+      "step": 3103
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.9448983739541603,
+      "learning_rate": 8.744578798988584e-06,
+      "loss": 0.7613,
+      "step": 3104
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.885468151504446,
+      "learning_rate": 8.743701445373922e-06,
+      "loss": 1.1849,
+      "step": 3105
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.584814906714775,
+      "learning_rate": 8.74282382934208e-06,
+      "loss": 1.3323,
+      "step": 3106
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.7150379669348883,
+      "learning_rate": 8.741945950954577e-06,
+      "loss": 0.6995,
+      "step": 3107
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.3623560606552503,
+      "learning_rate": 8.741067810272949e-06,
+      "loss": 0.4057,
+      "step": 3108
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.7438001086287556,
+      "learning_rate": 8.740189407358747e-06,
+      "loss": 1.0674,
+      "step": 3109
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.386810913733393,
+      "learning_rate": 8.739310742273546e-06,
+      "loss": 0.4845,
+      "step": 3110
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.908178175809527,
+      "learning_rate": 8.738431815078937e-06,
+      "loss": 1.0802,
+      "step": 3111
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.133124078726872,
+      "learning_rate": 8.737552625836525e-06,
+      "loss": 0.3782,
+      "step": 3112
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.1146653144289034,
+      "learning_rate": 8.73667317460794e-06,
+      "loss": 0.7028,
+      "step": 3113
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.8189764064404965,
+      "learning_rate": 8.735793461454828e-06,
+      "loss": 1.0162,
+      "step": 3114
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.21556316598591,
+      "learning_rate": 8.734913486438854e-06,
+      "loss": 1.6151,
+      "step": 3115
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.610434382709641,
+      "learning_rate": 8.734033249621695e-06,
+      "loss": 0.6186,
+      "step": 3116
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.1136830203208943,
+      "learning_rate": 8.73315275106506e-06,
+      "loss": 0.4801,
+      "step": 3117
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.8810962077461704,
+      "learning_rate": 8.732271990830663e-06,
+      "loss": 0.6271,
+      "step": 3118
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.8009657466197377,
+      "learning_rate": 8.731390968980242e-06,
+      "loss": 0.5574,
+      "step": 3119
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.5106276997584525,
+      "learning_rate": 8.730509685575552e-06,
+      "loss": 0.8657,
+      "step": 3120
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.70797696118993,
+      "learning_rate": 8.72962814067837e-06,
+      "loss": 0.9346,
+      "step": 3121
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.1023948752405675,
+      "learning_rate": 8.728746334350483e-06,
+      "loss": 1.2921,
+      "step": 3122
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.1629856076308487,
+      "learning_rate": 8.72786426665371e-06,
+      "loss": 0.2938,
+      "step": 3123
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.098831316671046,
+      "learning_rate": 8.726981937649875e-06,
+      "loss": 1.134,
+      "step": 3124
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.927581384352747,
+      "learning_rate": 8.726099347400824e-06,
+      "loss": 0.6837,
+      "step": 3125
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.745472576702596,
+      "learning_rate": 8.725216495968426e-06,
+      "loss": 1.5312,
+      "step": 3126
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.4741765877770887,
+      "learning_rate": 8.724333383414563e-06,
+      "loss": 0.7448,
+      "step": 3127
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.3200229506989993,
+      "learning_rate": 8.723450009801139e-06,
+      "loss": 0.5416,
+      "step": 3128
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.1467770194927835,
+      "learning_rate": 8.722566375190073e-06,
+      "loss": 0.9865,
+      "step": 3129
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.1764658181726544,
+      "learning_rate": 8.721682479643307e-06,
+      "loss": 0.679,
+      "step": 3130
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.045637342760714,
+      "learning_rate": 8.720798323222795e-06,
+      "loss": 1.0718,
+      "step": 3131
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 6.108791828817261,
+      "learning_rate": 8.719913905990511e-06,
+      "loss": 1.3495,
+      "step": 3132
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.581562699895055,
+      "learning_rate": 8.719029228008454e-06,
+      "loss": 0.9609,
+      "step": 3133
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.905464031884023,
+      "learning_rate": 8.718144289338632e-06,
+      "loss": 1.1061,
+      "step": 3134
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.038760696296255,
+      "learning_rate": 8.717259090043078e-06,
+      "loss": 1.6582,
+      "step": 3135
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.8090502903294605,
+      "learning_rate": 8.716373630183839e-06,
+      "loss": 0.5046,
+      "step": 3136
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.019665012589466,
+      "learning_rate": 8.715487909822982e-06,
+      "loss": 0.7241,
+      "step": 3137
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.342658470523938,
+      "learning_rate": 8.714601929022591e-06,
+      "loss": 0.7319,
+      "step": 3138
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.2930299554632527,
+      "learning_rate": 8.713715687844772e-06,
+      "loss": 0.6454,
+      "step": 3139
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.7485307418029983,
+      "learning_rate": 8.712829186351646e-06,
+      "loss": 0.8893,
+      "step": 3140
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.609508442458764,
+      "learning_rate": 8.711942424605352e-06,
+      "loss": 0.903,
+      "step": 3141
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.4121316309959697,
+      "learning_rate": 8.711055402668049e-06,
+      "loss": 0.8174,
+      "step": 3142
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.6279918762098293,
+      "learning_rate": 8.710168120601912e-06,
+      "loss": 0.6543,
+      "step": 3143
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.403460609793017,
+      "learning_rate": 8.709280578469135e-06,
+      "loss": 1.2444,
+      "step": 3144
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.7833886357786386,
+      "learning_rate": 8.708392776331935e-06,
+      "loss": 0.5886,
+      "step": 3145
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.3529510963178852,
+      "learning_rate": 8.707504714252539e-06,
+      "loss": 0.8862,
+      "step": 3146
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0977778285785627,
+      "learning_rate": 8.706616392293199e-06,
+      "loss": 0.1751,
+      "step": 3147
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 6.1552747267107994,
+      "learning_rate": 8.705727810516179e-06,
+      "loss": 1.0112,
+      "step": 3148
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.988284675168184,
+      "learning_rate": 8.70483896898377e-06,
+      "loss": 1.4558,
+      "step": 3149
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.146370997195073,
+      "learning_rate": 8.703949867758269e-06,
+      "loss": 0.1939,
+      "step": 3150
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.518287166379679,
+      "learning_rate": 8.703060506902004e-06,
+      "loss": 0.7558,
+      "step": 3151
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.253671816177628,
+      "learning_rate": 8.702170886477312e-06,
+      "loss": 1.1437,
+      "step": 3152
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.8170806354964486,
+      "learning_rate": 8.701281006546554e-06,
+      "loss": 0.9219,
+      "step": 3153
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.031382707502849,
+      "learning_rate": 8.700390867172104e-06,
+      "loss": 1.4201,
+      "step": 3154
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.142443824282733,
+      "learning_rate": 8.699500468416359e-06,
+      "loss": 0.5597,
+      "step": 3155
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.268972095381594,
+      "learning_rate": 8.698609810341733e-06,
+      "loss": 1.1795,
+      "step": 3156
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.245915832669328,
+      "learning_rate": 8.697718893010654e-06,
+      "loss": 1.0289,
+      "step": 3157
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.3672476490029912,
+      "learning_rate": 8.696827716485575e-06,
+      "loss": 0.6437,
+      "step": 3158
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.32873091167265,
+      "learning_rate": 8.69593628082896e-06,
+      "loss": 0.7401,
+      "step": 3159
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.05422206738384,
+      "learning_rate": 8.695044586103297e-06,
+      "loss": 0.7442,
+      "step": 3160
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.8301976158364446,
+      "learning_rate": 8.69415263237109e-06,
+      "loss": 0.8611,
+      "step": 3161
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.4771740486546956,
+      "learning_rate": 8.693260419694858e-06,
+      "loss": 0.4881,
+      "step": 3162
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.354569593164627,
+      "learning_rate": 8.692367948137146e-06,
+      "loss": 0.8609,
+      "step": 3163
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.3948040475678503,
+      "learning_rate": 8.69147521776051e-06,
+      "loss": 0.7847,
+      "step": 3164
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.952371981072322,
+      "learning_rate": 8.690582228627525e-06,
+      "loss": 1.2305,
+      "step": 3165
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.9741205651636946,
+      "learning_rate": 8.68968898080079e-06,
+      "loss": 0.4751,
+      "step": 3166
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.558786018301694,
+      "learning_rate": 8.688795474342913e-06,
+      "loss": 0.3973,
+      "step": 3167
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.3154057654178395,
+      "learning_rate": 8.687901709316526e-06,
+      "loss": 1.5187,
+      "step": 3168
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.86788167062856,
+      "learning_rate": 8.68700768578428e-06,
+      "loss": 0.3617,
+      "step": 3169
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 7.8911866823110985,
+      "learning_rate": 8.686113403808843e-06,
+      "loss": 0.997,
+      "step": 3170
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.869882483356167,
+      "learning_rate": 8.685218863452896e-06,
+      "loss": 1.2136,
+      "step": 3171
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.496630214403798,
+      "learning_rate": 8.684324064779147e-06,
+      "loss": 0.5059,
+      "step": 3172
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.0616387089416395,
+      "learning_rate": 8.683429007850313e-06,
+      "loss": 0.6251,
+      "step": 3173
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.2089421498861,
+      "learning_rate": 8.682533692729137e-06,
+      "loss": 0.7262,
+      "step": 3174
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.707763156712534,
+      "learning_rate": 8.681638119478375e-06,
+      "loss": 0.8375,
+      "step": 3175
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.622146077316868,
+      "learning_rate": 8.680742288160803e-06,
+      "loss": 0.5689,
+      "step": 3176
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.846788125256505,
+      "learning_rate": 8.679846198839216e-06,
+      "loss": 1.0026,
+      "step": 3177
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.110639668818378,
+      "learning_rate": 8.678949851576425e-06,
+      "loss": 0.5267,
+      "step": 3178
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5945107075834242,
+      "learning_rate": 8.678053246435261e-06,
+      "loss": 0.3912,
+      "step": 3179
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0778832980751147,
+      "learning_rate": 8.677156383478571e-06,
+      "loss": 0.5534,
+      "step": 3180
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.170325352138206,
+      "learning_rate": 8.676259262769222e-06,
+      "loss": 1.2933,
+      "step": 3181
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.6725061220038397,
+      "learning_rate": 8.675361884370097e-06,
+      "loss": 0.5551,
+      "step": 3182
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.1924954050152907,
+      "learning_rate": 8.6744642483441e-06,
+      "loss": 0.7139,
+      "step": 3183
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.9889298140945484,
+      "learning_rate": 8.67356635475415e-06,
+      "loss": 0.4657,
+      "step": 3184
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.434549748705281,
+      "learning_rate": 8.672668203663185e-06,
+      "loss": 0.6748,
+      "step": 3185
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.549624402623971,
+      "learning_rate": 8.671769795134164e-06,
+      "loss": 1.0978,
+      "step": 3186
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.962116235820115,
+      "learning_rate": 8.67087112923006e-06,
+      "loss": 0.6544,
+      "step": 3187
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.522077343959333,
+      "learning_rate": 8.669972206013864e-06,
+      "loss": 1.3273,
+      "step": 3188
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.2457425570795433,
+      "learning_rate": 8.669073025548588e-06,
+      "loss": 0.7193,
+      "step": 3189
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.7196750818422535,
+      "learning_rate": 8.668173587897261e-06,
+      "loss": 0.7571,
+      "step": 3190
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.3977177249844335,
+      "learning_rate": 8.667273893122932e-06,
+      "loss": 0.9859,
+      "step": 3191
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.8157013770108463,
+      "learning_rate": 8.666373941288661e-06,
+      "loss": 1.2364,
+      "step": 3192
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.493289730627912,
+      "learning_rate": 8.665473732457533e-06,
+      "loss": 0.9903,
+      "step": 3193
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.98117903788595,
+      "learning_rate": 8.664573266692648e-06,
+      "loss": 1.0363,
+      "step": 3194
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.7274376036047885,
+      "learning_rate": 8.663672544057126e-06,
+      "loss": 0.9828,
+      "step": 3195
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.84076112404563,
+      "learning_rate": 8.662771564614102e-06,
+      "loss": 0.6248,
+      "step": 3196
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.890432405279825,
+      "learning_rate": 8.661870328426734e-06,
+      "loss": 0.7854,
+      "step": 3197
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.554546599450026,
+      "learning_rate": 8.66096883555819e-06,
+      "loss": 0.5484,
+      "step": 3198
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.383581001219526,
+      "learning_rate": 8.660067086071665e-06,
+      "loss": 1.0188,
+      "step": 3199
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.5078927578736643,
+      "learning_rate": 8.659165080030366e-06,
+      "loss": 0.6404,
+      "step": 3200
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.906240123006569,
+      "learning_rate": 8.658262817497517e-06,
+      "loss": 0.6517,
+      "step": 3201
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.2438586396263625,
+      "learning_rate": 8.657360298536368e-06,
+      "loss": 0.3485,
+      "step": 3202
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.6901225925717074,
+      "learning_rate": 8.656457523210178e-06,
+      "loss": 0.7426,
+      "step": 3203
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.135785463653785,
+      "learning_rate": 8.65555449158223e-06,
+      "loss": 1.2519,
+      "step": 3204
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.197218607509568,
+      "learning_rate": 8.65465120371582e-06,
+      "loss": 0.9433,
+      "step": 3205
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.414861530515555,
+      "learning_rate": 8.653747659674265e-06,
+      "loss": 1.0128,
+      "step": 3206
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.7227671150350234,
+      "learning_rate": 8.6528438595209e-06,
+      "loss": 0.9727,
+      "step": 3207
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.859963917459991,
+      "learning_rate": 8.651939803319079e-06,
+      "loss": 0.8443,
+      "step": 3208
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.026595312703798,
+      "learning_rate": 8.651035491132171e-06,
+      "loss": 0.6055,
+      "step": 3209
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.250457816338681,
+      "learning_rate": 8.650130923023564e-06,
+      "loss": 0.8802,
+      "step": 3210
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.4885457304896335,
+      "learning_rate": 8.649226099056664e-06,
+      "loss": 0.8172,
+      "step": 3211
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.726001747173857,
+      "learning_rate": 8.648321019294896e-06,
+      "loss": 0.5996,
+      "step": 3212
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.501357604415449,
+      "learning_rate": 8.647415683801703e-06,
+      "loss": 0.8746,
+      "step": 3213
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.668083160728214,
+      "learning_rate": 8.646510092640546e-06,
+      "loss": 0.7768,
+      "step": 3214
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.0515484249074465,
+      "learning_rate": 8.645604245874898e-06,
+      "loss": 1.1065,
+      "step": 3215
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.044574041841505,
+      "learning_rate": 8.64469814356826e-06,
+      "loss": 1.0122,
+      "step": 3216
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.206242620906617,
+      "learning_rate": 8.643791785784143e-06,
+      "loss": 1.2983,
+      "step": 3217
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.788662775672696,
+      "learning_rate": 8.642885172586079e-06,
+      "loss": 0.6405,
+      "step": 3218
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.771380323773392,
+      "learning_rate": 8.641978304037619e-06,
+      "loss": 0.502,
+      "step": 3219
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.196793597191887,
+      "learning_rate": 8.64107118020233e-06,
+      "loss": 0.4887,
+      "step": 3220
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.197875163689073,
+      "learning_rate": 8.640163801143797e-06,
+      "loss": 0.8519,
+      "step": 3221
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.267636049624269,
+      "learning_rate": 8.639256166925623e-06,
+      "loss": 0.4885,
+      "step": 3222
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.219769417655235,
+      "learning_rate": 8.638348277611431e-06,
+      "loss": 1.0352,
+      "step": 3223
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.300249351364667,
+      "learning_rate": 8.637440133264858e-06,
+      "loss": 0.8545,
+      "step": 3224
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.122445723263637,
+      "learning_rate": 8.636531733949562e-06,
+      "loss": 0.9217,
+      "step": 3225
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.6354846730557955,
+      "learning_rate": 8.635623079729218e-06,
+      "loss": 0.53,
+      "step": 3226
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.037083101704012,
+      "learning_rate": 8.63471417066752e-06,
+      "loss": 0.5501,
+      "step": 3227
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.815458380227235,
+      "learning_rate": 8.633805006828175e-06,
+      "loss": 0.7468,
+      "step": 3228
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.5727780826689126,
+      "learning_rate": 8.632895588274913e-06,
+      "loss": 0.7856,
+      "step": 3229
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.2251162220840035,
+      "learning_rate": 8.63198591507148e-06,
+      "loss": 0.6999,
+      "step": 3230
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.1361439942367513,
+      "learning_rate": 8.631075987281645e-06,
+      "loss": 0.7396,
+      "step": 3231
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.4658354764042856,
+      "learning_rate": 8.630165804969181e-06,
+      "loss": 1.0458,
+      "step": 3232
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.45617644112384,
+      "learning_rate": 8.629255368197894e-06,
+      "loss": 0.8709,
+      "step": 3233
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.986012130449948,
+      "learning_rate": 8.628344677031602e-06,
+      "loss": 1.3019,
+      "step": 3234
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.327627012636433,
+      "learning_rate": 8.627433731534138e-06,
+      "loss": 1.1655,
+      "step": 3235
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.5669462906154594,
+      "learning_rate": 8.626522531769356e-06,
+      "loss": 0.571,
+      "step": 3236
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.7131241591546362,
+      "learning_rate": 8.625611077801127e-06,
+      "loss": 0.7743,
+      "step": 3237
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.0749190601802012,
+      "learning_rate": 8.624699369693338e-06,
+      "loss": 1.0538,
+      "step": 3238
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.836550605964662,
+      "learning_rate": 8.623787407509902e-06,
+      "loss": 0.7836,
+      "step": 3239
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.759929928706973,
+      "learning_rate": 8.622875191314737e-06,
+      "loss": 0.6143,
+      "step": 3240
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.6827245495757204,
+      "learning_rate": 8.621962721171789e-06,
+      "loss": 0.6333,
+      "step": 3241
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.017636572954006,
+      "learning_rate": 8.621049997145016e-06,
+      "loss": 0.4061,
+      "step": 3242
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.046091184963527,
+      "learning_rate": 8.620137019298397e-06,
+      "loss": 1.274,
+      "step": 3243
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.630981230839866,
+      "learning_rate": 8.61922378769593e-06,
+      "loss": 0.8632,
+      "step": 3244
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.857283660257242,
+      "learning_rate": 8.618310302401624e-06,
+      "loss": 0.8118,
+      "step": 3245
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.8626741829376483,
+      "learning_rate": 8.617396563479512e-06,
+      "loss": 0.7905,
+      "step": 3246
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.421926641303325,
+      "learning_rate": 8.616482570993648e-06,
+      "loss": 1.2504,
+      "step": 3247
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.546897594665816,
+      "learning_rate": 8.61556832500809e-06,
+      "loss": 1.514,
+      "step": 3248
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.060131968138134,
+      "learning_rate": 8.614653825586929e-06,
+      "loss": 1.1402,
+      "step": 3249
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.1715649425333967,
+      "learning_rate": 8.613739072794268e-06,
+      "loss": 0.6246,
+      "step": 3250
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.404068495336512,
+      "learning_rate": 8.612824066694223e-06,
+      "loss": 1.0936,
+      "step": 3251
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.484444207292898,
+      "learning_rate": 8.611908807350934e-06,
+      "loss": 0.9757,
+      "step": 3252
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.423023227701489,
+      "learning_rate": 8.610993294828557e-06,
+      "loss": 0.4086,
+      "step": 3253
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.349545026797728,
+      "learning_rate": 8.610077529191265e-06,
+      "loss": 0.9609,
+      "step": 3254
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.389377079746561,
+      "learning_rate": 8.60916151050325e-06,
+      "loss": 1.0769,
+      "step": 3255
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.202524400456061,
+      "learning_rate": 8.60824523882872e-06,
+      "loss": 0.7872,
+      "step": 3256
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.505244005208422,
+      "learning_rate": 8.607328714231901e-06,
+      "loss": 0.7738,
+      "step": 3257
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.492210496875677,
+      "learning_rate": 8.60641193677704e-06,
+      "loss": 0.1749,
+      "step": 3258
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.211838371020348,
+      "learning_rate": 8.605494906528395e-06,
+      "loss": 0.9311,
+      "step": 3259
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.423420650706527,
+      "learning_rate": 8.604577623550249e-06,
+      "loss": 0.9224,
+      "step": 3260
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.158714351825901,
+      "learning_rate": 8.603660087906901e-06,
+      "loss": 1.0484,
+      "step": 3261
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.102169056933126,
+      "learning_rate": 8.602742299662662e-06,
+      "loss": 1.0258,
+      "step": 3262
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.499002455252406,
+      "learning_rate": 8.601824258881868e-06,
+      "loss": 1.3883,
+      "step": 3263
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.4907076006441042,
+      "learning_rate": 8.600905965628867e-06,
+      "loss": 0.5417,
+      "step": 3264
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.8531022936766153,
+      "learning_rate": 8.59998741996803e-06,
+      "loss": 0.877,
+      "step": 3265
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.370751097548014,
+      "learning_rate": 8.599068621963741e-06,
+      "loss": 1.0476,
+      "step": 3266
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.508339914520264,
+      "learning_rate": 8.598149571680407e-06,
+      "loss": 1.4139,
+      "step": 3267
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.3080389351523425,
+      "learning_rate": 8.597230269182446e-06,
+      "loss": 0.7074,
+      "step": 3268
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.562432608615904,
+      "learning_rate": 8.596310714534299e-06,
+      "loss": 0.4951,
+      "step": 3269
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.437605479178292,
+      "learning_rate": 8.59539090780042e-06,
+      "loss": 1.112,
+      "step": 3270
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.508607042973389,
+      "learning_rate": 8.59447084904529e-06,
+      "loss": 0.4823,
+      "step": 3271
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.228500440504984,
+      "learning_rate": 8.593550538333392e-06,
+      "loss": 1.4294,
+      "step": 3272
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.767059649852298,
+      "learning_rate": 8.592629975729242e-06,
+      "loss": 0.9167,
+      "step": 3273
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.269612391619527,
+      "learning_rate": 8.591709161297366e-06,
+      "loss": 0.6488,
+      "step": 3274
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.846307992233512,
+      "learning_rate": 8.59078809510231e-06,
+      "loss": 0.9216,
+      "step": 3275
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.826302142300191,
+      "learning_rate": 8.589866777208632e-06,
+      "loss": 1.4859,
+      "step": 3276
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.9845686316737927,
+      "learning_rate": 8.588945207680922e-06,
+      "loss": 0.5955,
+      "step": 3277
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.516434459764189,
+      "learning_rate": 8.588023386583766e-06,
+      "loss": 1.227,
+      "step": 3278
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.000628683865046,
+      "learning_rate": 8.587101313981789e-06,
+      "loss": 0.429,
+      "step": 3279
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.149387629155255,
+      "learning_rate": 8.586178989939622e-06,
+      "loss": 0.2166,
+      "step": 3280
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.674756287794339,
+      "learning_rate": 8.585256414521912e-06,
+      "loss": 1.4346,
+      "step": 3281
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.814894563094977,
+      "learning_rate": 8.584333587793334e-06,
+      "loss": 1.093,
+      "step": 3282
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.2099645602466165,
+      "learning_rate": 8.583410509818567e-06,
+      "loss": 0.8124,
+      "step": 3283
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.7785344671897114,
+      "learning_rate": 8.58248718066232e-06,
+      "loss": 0.9876,
+      "step": 3284
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.827799117265657,
+      "learning_rate": 8.581563600389313e-06,
+      "loss": 1.1371,
+      "step": 3285
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.027555913927014,
+      "learning_rate": 8.580639769064283e-06,
+      "loss": 0.9466,
+      "step": 3286
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.4193502984889474,
+      "learning_rate": 8.57971568675199e-06,
+      "loss": 0.5155,
+      "step": 3287
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.811872675998102,
+      "learning_rate": 8.578791353517205e-06,
+      "loss": 0.6685,
+      "step": 3288
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.0447413505751895,
+      "learning_rate": 8.577866769424722e-06,
+      "loss": 0.7002,
+      "step": 3289
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.67709329214413,
+      "learning_rate": 8.57694193453935e-06,
+      "loss": 1.059,
+      "step": 3290
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.9639687943958597,
+      "learning_rate": 8.576016848925914e-06,
+      "loss": 0.9254,
+      "step": 3291
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.756176769055125,
+      "learning_rate": 8.57509151264926e-06,
+      "loss": 0.689,
+      "step": 3292
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.9754738372845124,
+      "learning_rate": 8.574165925774249e-06,
+      "loss": 0.8303,
+      "step": 3293
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.7067266196734256,
+      "learning_rate": 8.573240088365764e-06,
+      "loss": 0.5972,
+      "step": 3294
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.318813164893715,
+      "learning_rate": 8.572314000488697e-06,
+      "loss": 0.9459,
+      "step": 3295
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.176265666423131,
+      "learning_rate": 8.571387662207966e-06,
+      "loss": 1.0976,
+      "step": 3296
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.027574899271077,
+      "learning_rate": 8.570461073588503e-06,
+      "loss": 1.4062,
+      "step": 3297
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.5759335168554047,
+      "learning_rate": 8.569534234695258e-06,
+      "loss": 0.7524,
+      "step": 3298
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.98753427562043,
+      "learning_rate": 8.568607145593197e-06,
+      "loss": 0.5578,
+      "step": 3299
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.824454307674025,
+      "learning_rate": 8.567679806347307e-06,
+      "loss": 1.1906,
+      "step": 3300
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.8811060287963155,
+      "learning_rate": 8.566752217022587e-06,
+      "loss": 0.9038,
+      "step": 3301
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.1295118533835895,
+      "learning_rate": 8.56582437768406e-06,
+      "loss": 0.5656,
+      "step": 3302
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.189873115925194,
+      "learning_rate": 8.564896288396762e-06,
+      "loss": 0.7928,
+      "step": 3303
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.9550707861543692,
+      "learning_rate": 8.56396794922575e-06,
+      "loss": 0.9126,
+      "step": 3304
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.734076565039832,
+      "learning_rate": 8.563039360236097e-06,
+      "loss": 1.3105,
+      "step": 3305
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.7954476971871696,
+      "learning_rate": 8.562110521492888e-06,
+      "loss": 0.5605,
+      "step": 3306
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.913998182189051,
+      "learning_rate": 8.561181433061234e-06,
+      "loss": 1.4244,
+      "step": 3307
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.009470227880041,
+      "learning_rate": 8.56025209500626e-06,
+      "loss": 0.7917,
+      "step": 3308
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.665922735380563,
+      "learning_rate": 8.55932250739311e-06,
+      "loss": 0.3222,
+      "step": 3309
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.7198380743455095,
+      "learning_rate": 8.558392670286942e-06,
+      "loss": 1.1366,
+      "step": 3310
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.178748702987763,
+      "learning_rate": 8.557462583752934e-06,
+      "loss": 0.7711,
+      "step": 3311
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.438568163220989,
+      "learning_rate": 8.55653224785628e-06,
+      "loss": 0.4922,
+      "step": 3312
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.330476916402365,
+      "learning_rate": 8.555601662662194e-06,
+      "loss": 1.2982,
+      "step": 3313
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.6678253271213106,
+      "learning_rate": 8.554670828235905e-06,
+      "loss": 0.5943,
+      "step": 3314
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.1007843639528807,
+      "learning_rate": 8.553739744642662e-06,
+      "loss": 0.3446,
+      "step": 3315
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5255229994819723,
+      "learning_rate": 8.552808411947727e-06,
+      "loss": 0.0892,
+      "step": 3316
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.245052826510659,
+      "learning_rate": 8.551876830216385e-06,
+      "loss": 0.2156,
+      "step": 3317
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.4013826162449154,
+      "learning_rate": 8.550944999513936e-06,
+      "loss": 0.8355,
+      "step": 3318
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9915229079069694,
+      "learning_rate": 8.550012919905696e-06,
+      "loss": 0.3716,
+      "step": 3319
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.177043562454798,
+      "learning_rate": 8.549080591457e-06,
+      "loss": 0.8958,
+      "step": 3320
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.4423048906076823,
+      "learning_rate": 8.5481480142332e-06,
+      "loss": 0.4812,
+      "step": 3321
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.207681663736577,
+      "learning_rate": 8.547215188299664e-06,
+      "loss": 0.5496,
+      "step": 3322
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3400518150108705,
+      "learning_rate": 8.546282113721785e-06,
+      "loss": 0.2233,
+      "step": 3323
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.521282854097466,
+      "learning_rate": 8.54534879056496e-06,
+      "loss": 0.7035,
+      "step": 3324
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.027767335477696,
+      "learning_rate": 8.544415218894615e-06,
+      "loss": 0.6919,
+      "step": 3325
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.016276645524086,
+      "learning_rate": 8.543481398776188e-06,
+      "loss": 0.6924,
+      "step": 3326
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.264797794444411,
+      "learning_rate": 8.542547330275138e-06,
+      "loss": 0.9471,
+      "step": 3327
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.2930852371271047,
+      "learning_rate": 8.541613013456935e-06,
+      "loss": 0.8721,
+      "step": 3328
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.3653326087802973,
+      "learning_rate": 8.540678448387075e-06,
+      "loss": 0.7226,
+      "step": 3329
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.672402910371198,
+      "learning_rate": 8.539743635131064e-06,
+      "loss": 1.0184,
+      "step": 3330
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.302459581678297,
+      "learning_rate": 8.538808573754428e-06,
+      "loss": 1.2515,
+      "step": 3331
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.9777333277068743,
+      "learning_rate": 8.537873264322714e-06,
+      "loss": 0.8424,
+      "step": 3332
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.2118263151992004,
+      "learning_rate": 8.53693770690148e-06,
+      "loss": 0.736,
+      "step": 3333
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.505969676427309,
+      "learning_rate": 8.536001901556306e-06,
+      "loss": 0.6689,
+      "step": 3334
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.2270756392293136,
+      "learning_rate": 8.535065848352785e-06,
+      "loss": 0.6879,
+      "step": 3335
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.698836210287015,
+      "learning_rate": 8.534129547356536e-06,
+      "loss": 0.6287,
+      "step": 3336
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.961390143155534,
+      "learning_rate": 8.533192998633184e-06,
+      "loss": 0.9411,
+      "step": 3337
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.4226150856526845,
+      "learning_rate": 8.532256202248382e-06,
+      "loss": 1.0278,
+      "step": 3338
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1314777414437935,
+      "learning_rate": 8.53131915826779e-06,
+      "loss": 0.3731,
+      "step": 3339
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.7013183433744246,
+      "learning_rate": 8.530381866757096e-06,
+      "loss": 0.6113,
+      "step": 3340
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.1012435037945,
+      "learning_rate": 8.529444327781995e-06,
+      "loss": 1.0921,
+      "step": 3341
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.828225586633404,
+      "learning_rate": 8.528506541408208e-06,
+      "loss": 0.471,
+      "step": 3342
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.270768950776447,
+      "learning_rate": 8.527568507701467e-06,
+      "loss": 0.6118,
+      "step": 3343
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.133735752099045,
+      "learning_rate": 8.526630226727528e-06,
+      "loss": 0.8534,
+      "step": 3344
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.2286250088537605,
+      "learning_rate": 8.525691698552157e-06,
+      "loss": 1.2391,
+      "step": 3345
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.674507236090259,
+      "learning_rate": 8.52475292324114e-06,
+      "loss": 0.9262,
+      "step": 3346
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.251410586765895,
+      "learning_rate": 8.523813900860285e-06,
+      "loss": 0.7037,
+      "step": 3347
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.489785459479057,
+      "learning_rate": 8.52287463147541e-06,
+      "loss": 0.8188,
+      "step": 3348
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.275572233721359,
+      "learning_rate": 8.521935115152357e-06,
+      "loss": 0.8396,
+      "step": 3349
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.0843978814099873,
+      "learning_rate": 8.520995351956977e-06,
+      "loss": 0.416,
+      "step": 3350
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.3651970008548,
+      "learning_rate": 8.52005534195515e-06,
+      "loss": 1.806,
+      "step": 3351
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.6903373871188485,
+      "learning_rate": 8.51911508521276e-06,
+      "loss": 0.562,
+      "step": 3352
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.201406982156682,
+      "learning_rate": 8.518174581795718e-06,
+      "loss": 0.6777,
+      "step": 3353
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.8101727232527187,
+      "learning_rate": 8.517233831769951e-06,
+      "loss": 0.641,
+      "step": 3354
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.201730798911782,
+      "learning_rate": 8.516292835201396e-06,
+      "loss": 0.7662,
+      "step": 3355
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.016548229239503,
+      "learning_rate": 8.51535159215602e-06,
+      "loss": 0.9928,
+      "step": 3356
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.594045982399574,
+      "learning_rate": 8.514410102699794e-06,
+      "loss": 0.3828,
+      "step": 3357
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.063304614421123,
+      "learning_rate": 8.513468366898714e-06,
+      "loss": 0.965,
+      "step": 3358
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.613244946794997,
+      "learning_rate": 8.512526384818794e-06,
+      "loss": 0.9638,
+      "step": 3359
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.133040387545648,
+      "learning_rate": 8.511584156526059e-06,
+      "loss": 0.8531,
+      "step": 3360
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.5220936512990617,
+      "learning_rate": 8.510641682086557e-06,
+      "loss": 0.782,
+      "step": 3361
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.2675450795257666,
+      "learning_rate": 8.509698961566356e-06,
+      "loss": 0.6078,
+      "step": 3362
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.7808428333412145,
+      "learning_rate": 8.508755995031527e-06,
+      "loss": 0.9535,
+      "step": 3363
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.202309335801023,
+      "learning_rate": 8.507812782548174e-06,
+      "loss": 1.0453,
+      "step": 3364
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.494824079311042,
+      "learning_rate": 8.506869324182411e-06,
+      "loss": 1.143,
+      "step": 3365
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.022960989901094,
+      "learning_rate": 8.505925620000373e-06,
+      "loss": 0.625,
+      "step": 3366
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.417571085131394,
+      "learning_rate": 8.504981670068204e-06,
+      "loss": 1.2577,
+      "step": 3367
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.54414024670515,
+      "learning_rate": 8.504037474452073e-06,
+      "loss": 1.224,
+      "step": 3368
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.328681112964479,
+      "learning_rate": 8.503093033218168e-06,
+      "loss": 0.7479,
+      "step": 3369
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.581582710005775,
+      "learning_rate": 8.502148346432683e-06,
+      "loss": 0.6942,
+      "step": 3370
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.630141365785073,
+      "learning_rate": 8.501203414161844e-06,
+      "loss": 0.8597,
+      "step": 3371
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.4991407245673773,
+      "learning_rate": 8.50025823647188e-06,
+      "loss": 0.5979,
+      "step": 3372
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.313086390452513,
+      "learning_rate": 8.499312813429047e-06,
+      "loss": 0.8321,
+      "step": 3373
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.098169768904149,
+      "learning_rate": 8.498367145099618e-06,
+      "loss": 0.3456,
+      "step": 3374
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.49992191921518,
+      "learning_rate": 8.497421231549873e-06,
+      "loss": 1.3791,
+      "step": 3375
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.2400728584587615,
+      "learning_rate": 8.496475072846125e-06,
+      "loss": 0.5226,
+      "step": 3376
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.5645603139882116,
+      "learning_rate": 8.495528669054688e-06,
+      "loss": 0.579,
+      "step": 3377
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.016893323985373,
+      "learning_rate": 8.494582020241905e-06,
+      "loss": 1.2116,
+      "step": 3378
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.0975535418531996,
+      "learning_rate": 8.493635126474128e-06,
+      "loss": 0.8675,
+      "step": 3379
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.967808603430733,
+      "learning_rate": 8.492687987817736e-06,
+      "loss": 1.4063,
+      "step": 3380
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.069151077502174,
+      "learning_rate": 8.491740604339115e-06,
+      "loss": 0.5528,
+      "step": 3381
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.29931499894164,
+      "learning_rate": 8.490792976104676e-06,
+      "loss": 0.8807,
+      "step": 3382
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.600143238803147,
+      "learning_rate": 8.48984510318084e-06,
+      "loss": 0.886,
+      "step": 3383
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.7937467755769947,
+      "learning_rate": 8.48889698563405e-06,
+      "loss": 1.1959,
+      "step": 3384
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.2159369615732913,
+      "learning_rate": 8.487948623530765e-06,
+      "loss": 0.3218,
+      "step": 3385
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.567853307581026,
+      "learning_rate": 8.48700001693746e-06,
+      "loss": 0.6045,
+      "step": 3386
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.204283266026797,
+      "learning_rate": 8.48605116592063e-06,
+      "loss": 0.8285,
+      "step": 3387
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.522280887836711,
+      "learning_rate": 8.485102070546786e-06,
+      "loss": 0.9568,
+      "step": 3388
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.688467153997024,
+      "learning_rate": 8.484152730882453e-06,
+      "loss": 0.5945,
+      "step": 3389
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.046191178978568,
+      "learning_rate": 8.483203146994174e-06,
+      "loss": 0.4337,
+      "step": 3390
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.301048854694765,
+      "learning_rate": 8.482253318948516e-06,
+      "loss": 1.2326,
+      "step": 3391
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.0677386148310584,
+      "learning_rate": 8.481303246812056e-06,
+      "loss": 0.6228,
+      "step": 3392
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.9716547402830846,
+      "learning_rate": 8.480352930651387e-06,
+      "loss": 0.9784,
+      "step": 3393
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.63926742491477,
+      "learning_rate": 8.479402370533127e-06,
+      "loss": 0.9952,
+      "step": 3394
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.2588996439244484,
+      "learning_rate": 8.478451566523902e-06,
+      "loss": 0.7426,
+      "step": 3395
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.91189519136294,
+      "learning_rate": 8.47750051869036e-06,
+      "loss": 0.6852,
+      "step": 3396
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.392249566403095,
+      "learning_rate": 8.476549227099164e-06,
+      "loss": 0.7342,
+      "step": 3397
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.808611635862418,
+      "learning_rate": 8.475597691817e-06,
+      "loss": 0.7754,
+      "step": 3398
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.1065486891336858,
+      "learning_rate": 8.474645912910562e-06,
+      "loss": 0.7521,
+      "step": 3399
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.844818704294859,
+      "learning_rate": 8.473693890446568e-06,
+      "loss": 1.1064,
+      "step": 3400
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.4964248348112372,
+      "learning_rate": 8.472741624491749e-06,
+      "loss": 0.829,
+      "step": 3401
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.813746484461821,
+      "learning_rate": 8.471789115112857e-06,
+      "loss": 1.6135,
+      "step": 3402
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 6.04943564553509,
+      "learning_rate": 8.470836362376657e-06,
+      "loss": 1.5515,
+      "step": 3403
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.974791909097457,
+      "learning_rate": 8.469883366349932e-06,
+      "loss": 0.6906,
+      "step": 3404
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.369584392055278,
+      "learning_rate": 8.468930127099486e-06,
+      "loss": 1.2062,
+      "step": 3405
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.0886044700598396,
+      "learning_rate": 8.467976644692131e-06,
+      "loss": 0.7033,
+      "step": 3406
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.00103892834991,
+      "learning_rate": 8.46702291919471e-06,
+      "loss": 0.9827,
+      "step": 3407
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.134494477289156,
+      "learning_rate": 8.466068950674068e-06,
+      "loss": 1.0653,
+      "step": 3408
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.323040388604154,
+      "learning_rate": 8.465114739197079e-06,
+      "loss": 1.0076,
+      "step": 3409
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.5968184604234805,
+      "learning_rate": 8.464160284830627e-06,
+      "loss": 1.2016,
+      "step": 3410
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.7621565259260055,
+      "learning_rate": 8.463205587641614e-06,
+      "loss": 0.2877,
+      "step": 3411
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.938771720970346,
+      "learning_rate": 8.462250647696962e-06,
+      "loss": 1.5023,
+      "step": 3412
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.9458559677127023,
+      "learning_rate": 8.461295465063605e-06,
+      "loss": 0.6964,
+      "step": 3413
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.4891976793370416,
+      "learning_rate": 8.460340039808504e-06,
+      "loss": 1.0164,
+      "step": 3414
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.8104414866522105,
+      "learning_rate": 8.459384371998622e-06,
+      "loss": 0.7668,
+      "step": 3415
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.732488335406169,
+      "learning_rate": 8.458428461700951e-06,
+      "loss": 0.841,
+      "step": 3416
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.7949169717721754,
+      "learning_rate": 8.457472308982498e-06,
+      "loss": 0.6091,
+      "step": 3417
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.3442232976274724,
+      "learning_rate": 8.456515913910282e-06,
+      "loss": 0.4215,
+      "step": 3418
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.660297546667241,
+      "learning_rate": 8.455559276551343e-06,
+      "loss": 0.3241,
+      "step": 3419
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.1989636832644086,
+      "learning_rate": 8.454602396972737e-06,
+      "loss": 0.6127,
+      "step": 3420
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.3692484321829754,
+      "learning_rate": 8.453645275241538e-06,
+      "loss": 0.4041,
+      "step": 3421
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.5031828824437254,
+      "learning_rate": 8.452687911424836e-06,
+      "loss": 0.8936,
+      "step": 3422
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.276633330790473,
+      "learning_rate": 8.451730305589737e-06,
+      "loss": 0.5023,
+      "step": 3423
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.379893635875519,
+      "learning_rate": 8.450772457803365e-06,
+      "loss": 0.5751,
+      "step": 3424
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.221177909343238,
+      "learning_rate": 8.449814368132862e-06,
+      "loss": 1.0798,
+      "step": 3425
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.1564530399167605,
+      "learning_rate": 8.448856036645386e-06,
+      "loss": 1.1126,
+      "step": 3426
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.136210138891469,
+      "learning_rate": 8.447897463408113e-06,
+      "loss": 1.3091,
+      "step": 3427
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.802072054536583,
+      "learning_rate": 8.44693864848823e-06,
+      "loss": 0.9722,
+      "step": 3428
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.438403174790669,
+      "learning_rate": 8.44597959195295e-06,
+      "loss": 1.4541,
+      "step": 3429
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.858668859920182,
+      "learning_rate": 8.445020293869497e-06,
+      "loss": 0.7488,
+      "step": 3430
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.384226705250362,
+      "learning_rate": 8.444060754305115e-06,
+      "loss": 0.6044,
+      "step": 3431
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.210435131363074,
+      "learning_rate": 8.443100973327063e-06,
+      "loss": 0.8752,
+      "step": 3432
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.611685456780522,
+      "learning_rate": 8.442140951002616e-06,
+      "loss": 0.7336,
+      "step": 3433
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.968448930951749,
+      "learning_rate": 8.441180687399068e-06,
+      "loss": 0.3585,
+      "step": 3434
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.1026752437935223,
+      "learning_rate": 8.440220182583731e-06,
+      "loss": 0.4757,
+      "step": 3435
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.6772016795236184,
+      "learning_rate": 8.439259436623933e-06,
+      "loss": 0.8092,
+      "step": 3436
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.929545885760317,
+      "learning_rate": 8.438298449587014e-06,
+      "loss": 0.4462,
+      "step": 3437
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.8387804784920716,
+      "learning_rate": 8.437337221540337e-06,
+      "loss": 0.638,
+      "step": 3438
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 6.15289257004559,
+      "learning_rate": 8.436375752551282e-06,
+      "loss": 1.1346,
+      "step": 3439
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.3656847914305503,
+      "learning_rate": 8.43541404268724e-06,
+      "loss": 0.7933,
+      "step": 3440
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.8082542811817923,
+      "learning_rate": 8.434452092015624e-06,
+      "loss": 0.6105,
+      "step": 3441
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.434561121019493,
+      "learning_rate": 8.433489900603866e-06,
+      "loss": 1.2958,
+      "step": 3442
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.036759914134969,
+      "learning_rate": 8.432527468519405e-06,
+      "loss": 1.3316,
+      "step": 3443
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.488323094554122,
+      "learning_rate": 8.43156479582971e-06,
+      "loss": 0.746,
+      "step": 3444
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.2517439549549003,
+      "learning_rate": 8.430601882602256e-06,
+      "loss": 0.6266,
+      "step": 3445
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.502813425517545,
+      "learning_rate": 8.429638728904538e-06,
+      "loss": 1.0791,
+      "step": 3446
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.22014969866628,
+      "learning_rate": 8.428675334804073e-06,
+      "loss": 1.3465,
+      "step": 3447
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.552856686401315,
+      "learning_rate": 8.42771170036839e-06,
+      "loss": 0.4467,
+      "step": 3448
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.423996933534739,
+      "learning_rate": 8.426747825665032e-06,
+      "loss": 0.6019,
+      "step": 3449
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.81325929155305,
+      "learning_rate": 8.425783710761565e-06,
+      "loss": 0.8259,
+      "step": 3450
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.4597466616695405,
+      "learning_rate": 8.42481935572557e-06,
+      "loss": 1.0784,
+      "step": 3451
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.190375688125212,
+      "learning_rate": 8.423854760624641e-06,
+      "loss": 1.4251,
+      "step": 3452
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.350206830956461,
+      "learning_rate": 8.422889925526396e-06,
+      "loss": 1.0581,
+      "step": 3453
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.675596277735707,
+      "learning_rate": 8.421924850498464e-06,
+      "loss": 0.318,
+      "step": 3454
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.567263639604266,
+      "learning_rate": 8.420959535608491e-06,
+      "loss": 0.8752,
+      "step": 3455
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.188249554694329,
+      "learning_rate": 8.419993980924141e-06,
+      "loss": 1.2736,
+      "step": 3456
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.804489030767843,
+      "learning_rate": 8.4190281865131e-06,
+      "loss": 1.0392,
+      "step": 3457
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.910357957576581,
+      "learning_rate": 8.41806215244306e-06,
+      "loss": 0.9776,
+      "step": 3458
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.0388462868255,
+      "learning_rate": 8.417095878781742e-06,
+      "loss": 1.0865,
+      "step": 3459
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.290045246104023,
+      "learning_rate": 8.41612936559687e-06,
+      "loss": 0.4962,
+      "step": 3460
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.627539575503532,
+      "learning_rate": 8.4151626129562e-06,
+      "loss": 0.6425,
+      "step": 3461
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.649946228294108,
+      "learning_rate": 8.414195620927491e-06,
+      "loss": 1.0886,
+      "step": 3462
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.51517747858038,
+      "learning_rate": 8.41322838957853e-06,
+      "loss": 0.8784,
+      "step": 3463
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.330803877844067,
+      "learning_rate": 8.412260918977112e-06,
+      "loss": 0.8254,
+      "step": 3464
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3595172213056332,
+      "learning_rate": 8.411293209191054e-06,
+      "loss": 0.1939,
+      "step": 3465
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.892873810491101,
+      "learning_rate": 8.410325260288188e-06,
+      "loss": 1.0629,
+      "step": 3466
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.569474704149108,
+      "learning_rate": 8.409357072336363e-06,
+      "loss": 0.5773,
+      "step": 3467
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.545847930059436,
+      "learning_rate": 8.408388645403445e-06,
+      "loss": 0.6537,
+      "step": 3468
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.18436426463238,
+      "learning_rate": 8.407419979557317e-06,
+      "loss": 1.4713,
+      "step": 3469
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.442562603742148,
+      "learning_rate": 8.406451074865875e-06,
+      "loss": 1.0713,
+      "step": 3470
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.6746294802722166,
+      "learning_rate": 8.405481931397042e-06,
+      "loss": 0.8966,
+      "step": 3471
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.414364511955644,
+      "learning_rate": 8.404512549218741e-06,
+      "loss": 0.6344,
+      "step": 3472
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.5411981015933045,
+      "learning_rate": 8.40354292839893e-06,
+      "loss": 0.5917,
+      "step": 3473
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.4104669592059573,
+      "learning_rate": 8.402573069005573e-06,
+      "loss": 0.3866,
+      "step": 3474
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.8931161075128085,
+      "learning_rate": 8.40160297110665e-06,
+      "loss": 0.559,
+      "step": 3475
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.6508457374642984,
+      "learning_rate": 8.400632634770163e-06,
+      "loss": 0.5797,
+      "step": 3476
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.405419805992831,
+      "learning_rate": 8.399662060064126e-06,
+      "loss": 0.4704,
+      "step": 3477
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.5565941443329865,
+      "learning_rate": 8.398691247056577e-06,
+      "loss": 0.8638,
+      "step": 3478
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.3152786905128035,
+      "learning_rate": 8.397720195815561e-06,
+      "loss": 0.9719,
+      "step": 3479
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.155128845995398,
+      "learning_rate": 8.396748906409147e-06,
+      "loss": 1.1606,
+      "step": 3480
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.602097103846587,
+      "learning_rate": 8.395777378905417e-06,
+      "loss": 0.6643,
+      "step": 3481
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.734829924575268,
+      "learning_rate": 8.394805613372471e-06,
+      "loss": 0.6499,
+      "step": 3482
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.0622166922148346,
+      "learning_rate": 8.393833609878426e-06,
+      "loss": 0.5855,
+      "step": 3483
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.236309198802574,
+      "learning_rate": 8.392861368491415e-06,
+      "loss": 0.7262,
+      "step": 3484
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.706892030666024,
+      "learning_rate": 8.391888889279589e-06,
+      "loss": 0.8424,
+      "step": 3485
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.003666146621652,
+      "learning_rate": 8.390916172311113e-06,
+      "loss": 0.8174,
+      "step": 3486
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.7056523946650533,
+      "learning_rate": 8.389943217654169e-06,
+      "loss": 0.687,
+      "step": 3487
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.099632712182386,
+      "learning_rate": 8.38897002537696e-06,
+      "loss": 1.0527,
+      "step": 3488
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.0084346188837983,
+      "learning_rate": 8.387996595547699e-06,
+      "loss": 0.7832,
+      "step": 3489
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.1426806013722053,
+      "learning_rate": 8.387022928234623e-06,
+      "loss": 0.7562,
+      "step": 3490
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.9478691188997646,
+      "learning_rate": 8.38604902350598e-06,
+      "loss": 0.7023,
+      "step": 3491
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.9094567058010594,
+      "learning_rate": 8.385074881430036e-06,
+      "loss": 0.8844,
+      "step": 3492
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.2910983396938285,
+      "learning_rate": 8.384100502075076e-06,
+      "loss": 1.015,
+      "step": 3493
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.628523203688381,
+      "learning_rate": 8.383125885509398e-06,
+      "loss": 1.1924,
+      "step": 3494
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.288493469772743,
+      "learning_rate": 8.382151031801318e-06,
+      "loss": 0.5543,
+      "step": 3495
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.178578709877618,
+      "learning_rate": 8.381175941019171e-06,
+      "loss": 1.2796,
+      "step": 3496
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.1525444590197744,
+      "learning_rate": 8.380200613231305e-06,
+      "loss": 0.6884,
+      "step": 3497
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.9848281841139404,
+      "learning_rate": 8.379225048506085e-06,
+      "loss": 0.6297,
+      "step": 3498
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.5167439822359685,
+      "learning_rate": 8.378249246911898e-06,
+      "loss": 0.7095,
+      "step": 3499
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.7118260838690333,
+      "learning_rate": 8.377273208517138e-06,
+      "loss": 0.9257,
+      "step": 3500
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.303460637133216,
+      "learning_rate": 8.376296933390227e-06,
+      "loss": 0.9383,
+      "step": 3501
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.786574910586223,
+      "learning_rate": 8.375320421599595e-06,
+      "loss": 0.3783,
+      "step": 3502
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.643732629168323,
+      "learning_rate": 8.374343673213689e-06,
+      "loss": 0.8174,
+      "step": 3503
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.9863507062348122,
+      "learning_rate": 8.373366688300978e-06,
+      "loss": 0.7188,
+      "step": 3504
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.426477659359341,
+      "learning_rate": 8.37238946692994e-06,
+      "loss": 0.8369,
+      "step": 3505
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.820979127414353,
+      "learning_rate": 8.37141200916908e-06,
+      "loss": 0.7249,
+      "step": 3506
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.560419011175925,
+      "learning_rate": 8.37043431508691e-06,
+      "loss": 0.9008,
+      "step": 3507
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.813495080021344,
+      "learning_rate": 8.369456384751963e-06,
+      "loss": 0.5069,
+      "step": 3508
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9553754956034446,
+      "learning_rate": 8.368478218232787e-06,
+      "loss": 0.3191,
+      "step": 3509
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.3542405637871937,
+      "learning_rate": 8.367499815597947e-06,
+      "loss": 0.6272,
+      "step": 3510
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.501836087508141,
+      "learning_rate": 8.366521176916027e-06,
+      "loss": 1.3518,
+      "step": 3511
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.610952413356651,
+      "learning_rate": 8.365542302255623e-06,
+      "loss": 0.5242,
+      "step": 3512
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.8167889077572505,
+      "learning_rate": 8.364563191685348e-06,
+      "loss": 0.9545,
+      "step": 3513
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.0670028406576355,
+      "learning_rate": 8.363583845273839e-06,
+      "loss": 1.4594,
+      "step": 3514
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.519355454572228,
+      "learning_rate": 8.362604263089739e-06,
+      "loss": 0.5279,
+      "step": 3515
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.5810225460305403,
+      "learning_rate": 8.361624445201715e-06,
+      "loss": 0.5254,
+      "step": 3516
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9515789420431564,
+      "learning_rate": 8.360644391678448e-06,
+      "loss": 0.4145,
+      "step": 3517
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.796650732563646,
+      "learning_rate": 8.359664102588633e-06,
+      "loss": 1.2567,
+      "step": 3518
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.5323364328106757,
+      "learning_rate": 8.358683578000987e-06,
+      "loss": 0.7193,
+      "step": 3519
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.897656459189634,
+      "learning_rate": 8.357702817984239e-06,
+      "loss": 0.6481,
+      "step": 3520
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.38233537413345,
+      "learning_rate": 8.356721822607136e-06,
+      "loss": 0.8116,
+      "step": 3521
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.32681111980777,
+      "learning_rate": 8.355740591938445e-06,
+      "loss": 0.8616,
+      "step": 3522
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.2162808086280132,
+      "learning_rate": 8.354759126046937e-06,
+      "loss": 0.5248,
+      "step": 3523
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.434389364827244,
+      "learning_rate": 8.353777425001417e-06,
+      "loss": 1.2069,
+      "step": 3524
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.824519616989892,
+      "learning_rate": 8.352795488870695e-06,
+      "loss": 0.6385,
+      "step": 3525
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.8981735074257484,
+      "learning_rate": 8.351813317723601e-06,
+      "loss": 0.6773,
+      "step": 3526
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.326576867153255,
+      "learning_rate": 8.35083091162898e-06,
+      "loss": 1.1598,
+      "step": 3527
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.421449357052687,
+      "learning_rate": 8.349848270655696e-06,
+      "loss": 0.7413,
+      "step": 3528
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.4048118640068776,
+      "learning_rate": 8.348865394872625e-06,
+      "loss": 0.6517,
+      "step": 3529
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.8012794215788777,
+      "learning_rate": 8.347882284348665e-06,
+      "loss": 1.1257,
+      "step": 3530
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.094074924068292,
+      "learning_rate": 8.346898939152728e-06,
+      "loss": 0.7916,
+      "step": 3531
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.981656102798751,
+      "learning_rate": 8.34591535935374e-06,
+      "loss": 0.8597,
+      "step": 3532
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.362189708212367,
+      "learning_rate": 8.344931545020646e-06,
+      "loss": 1.3468,
+      "step": 3533
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.011845342498781,
+      "learning_rate": 8.343947496222409e-06,
+      "loss": 0.6193,
+      "step": 3534
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.339480083175476,
+      "learning_rate": 8.342963213028005e-06,
+      "loss": 1.1736,
+      "step": 3535
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.9558061142349534,
+      "learning_rate": 8.34197869550643e-06,
+      "loss": 0.7738,
+      "step": 3536
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.154641996932149,
+      "learning_rate": 8.340993943726692e-06,
+      "loss": 1.3008,
+      "step": 3537
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.497783457109824,
+      "learning_rate": 8.340008957757817e-06,
+      "loss": 0.9407,
+      "step": 3538
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.138359581887209,
+      "learning_rate": 8.339023737668851e-06,
+      "loss": 1.0942,
+      "step": 3539
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.806807753455482,
+      "learning_rate": 8.338038283528852e-06,
+      "loss": 0.9245,
+      "step": 3540
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.201461202068064,
+      "learning_rate": 8.337052595406896e-06,
+      "loss": 0.8242,
+      "step": 3541
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.7793788421250945,
+      "learning_rate": 8.336066673372079e-06,
+      "loss": 0.8631,
+      "step": 3542
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.354344052829089,
+      "learning_rate": 8.335080517493503e-06,
+      "loss": 1.5768,
+      "step": 3543
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.216675886871792,
+      "learning_rate": 8.3340941278403e-06,
+      "loss": 0.7338,
+      "step": 3544
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.86311833127007,
+      "learning_rate": 8.333107504481606e-06,
+      "loss": 1.0233,
+      "step": 3545
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9121852510210247,
+      "learning_rate": 8.332120647486583e-06,
+      "loss": 0.402,
+      "step": 3546
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7083309623026617,
+      "learning_rate": 8.331133556924404e-06,
+      "loss": 0.466,
+      "step": 3547
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.756683675731017,
+      "learning_rate": 8.33014623286426e-06,
+      "loss": 1.0057,
+      "step": 3548
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.173372658028368,
+      "learning_rate": 8.32915867537536e-06,
+      "loss": 0.7373,
+      "step": 3549
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.611139422628014,
+      "learning_rate": 8.32817088452692e-06,
+      "loss": 1.2505,
+      "step": 3550
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.8996910690245694,
+      "learning_rate": 8.32718286038819e-06,
+      "loss": 0.8201,
+      "step": 3551
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.909677954824557,
+      "learning_rate": 8.326194603028419e-06,
+      "loss": 0.8776,
+      "step": 3552
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.2155694907097425,
+      "learning_rate": 8.325206112516883e-06,
+      "loss": 1.5689,
+      "step": 3553
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.877852476122359,
+      "learning_rate": 8.324217388922871e-06,
+      "loss": 0.8479,
+      "step": 3554
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.7553956611936457,
+      "learning_rate": 8.323228432315684e-06,
+      "loss": 0.8556,
+      "step": 3555
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.6617576423421254,
+      "learning_rate": 8.32223924276465e-06,
+      "loss": 1.0142,
+      "step": 3556
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.0526872916953827,
+      "learning_rate": 8.321249820339102e-06,
+      "loss": 0.3802,
+      "step": 3557
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.872143189106044,
+      "learning_rate": 8.320260165108395e-06,
+      "loss": 0.6965,
+      "step": 3558
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.806996264415691,
+      "learning_rate": 8.3192702771419e-06,
+      "loss": 0.287,
+      "step": 3559
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.699748900635862,
+      "learning_rate": 8.318280156509007e-06,
+      "loss": 0.5149,
+      "step": 3560
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.74006251254508,
+      "learning_rate": 8.317289803279112e-06,
+      "loss": 0.615,
+      "step": 3561
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.9784451434125705,
+      "learning_rate": 8.316299217521641e-06,
+      "loss": 1.4028,
+      "step": 3562
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.017270239393142,
+      "learning_rate": 8.315308399306027e-06,
+      "loss": 0.9775,
+      "step": 3563
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.42080669867183,
+      "learning_rate": 8.314317348701724e-06,
+      "loss": 0.7133,
+      "step": 3564
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.893028050718947,
+      "learning_rate": 8.313326065778198e-06,
+      "loss": 1.2777,
+      "step": 3565
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.55452023460235,
+      "learning_rate": 8.312334550604934e-06,
+      "loss": 1.2704,
+      "step": 3566
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.1051621769345426,
+      "learning_rate": 8.311342803251431e-06,
+      "loss": 1.1103,
+      "step": 3567
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.133309248809495,
+      "learning_rate": 8.310350823787213e-06,
+      "loss": 0.9868,
+      "step": 3568
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.6100634341728,
+      "learning_rate": 8.309358612281805e-06,
+      "loss": 0.905,
+      "step": 3569
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.6992294908098087,
+      "learning_rate": 8.308366168804764e-06,
+      "loss": 0.5854,
+      "step": 3570
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.193632411771907,
+      "learning_rate": 8.30737349342565e-06,
+      "loss": 0.6192,
+      "step": 3571
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.956331830855571,
+      "learning_rate": 8.30638058621405e-06,
+      "loss": 0.8116,
+      "step": 3572
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.4776215746659096,
+      "learning_rate": 8.305387447239559e-06,
+      "loss": 0.5024,
+      "step": 3573
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.9348804554264616,
+      "learning_rate": 8.304394076571794e-06,
+      "loss": 0.8389,
+      "step": 3574
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.9240920418295615,
+      "learning_rate": 8.303400474280384e-06,
+      "loss": 0.8246,
+      "step": 3575
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.604727604154122,
+      "learning_rate": 8.302406640434978e-06,
+      "loss": 1.1539,
+      "step": 3576
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.3571536012994,
+      "learning_rate": 8.301412575105238e-06,
+      "loss": 1.0719,
+      "step": 3577
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.614080600263977,
+      "learning_rate": 8.300418278360844e-06,
+      "loss": 0.4737,
+      "step": 3578
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.5897488641657365,
+      "learning_rate": 8.299423750271493e-06,
+      "loss": 0.9158,
+      "step": 3579
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.6018999721838405,
+      "learning_rate": 8.298428990906896e-06,
+      "loss": 0.7084,
+      "step": 3580
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9521900164481667,
+      "learning_rate": 8.297434000336781e-06,
+      "loss": 0.1663,
+      "step": 3581
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.416897765599965,
+      "learning_rate": 8.296438778630893e-06,
+      "loss": 0.7562,
+      "step": 3582
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.0830365100554555,
+      "learning_rate": 8.295443325858994e-06,
+      "loss": 0.9708,
+      "step": 3583
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5648254742585337,
+      "learning_rate": 8.294447642090857e-06,
+      "loss": 0.2603,
+      "step": 3584
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.5890801278182125,
+      "learning_rate": 8.293451727396283e-06,
+      "loss": 0.84,
+      "step": 3585
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.596402155220595,
+      "learning_rate": 8.29245558184507e-06,
+      "loss": 0.5958,
+      "step": 3586
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.645888567946884,
+      "learning_rate": 8.291459205507054e-06,
+      "loss": 1.1768,
+      "step": 3587
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.7688250555200846,
+      "learning_rate": 8.29046259845207e-06,
+      "loss": 0.6303,
+      "step": 3588
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.681310621502814,
+      "learning_rate": 8.289465760749977e-06,
+      "loss": 1.7021,
+      "step": 3589
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.613193942592307,
+      "learning_rate": 8.288468692470652e-06,
+      "loss": 1.0909,
+      "step": 3590
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.8875252487754874,
+      "learning_rate": 8.287471393683984e-06,
+      "loss": 1.1162,
+      "step": 3591
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.9841346177304655,
+      "learning_rate": 8.286473864459876e-06,
+      "loss": 0.7506,
+      "step": 3592
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.508869631913947,
+      "learning_rate": 8.285476104868252e-06,
+      "loss": 1.1238,
+      "step": 3593
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.837820743640624,
+      "learning_rate": 8.284478114979056e-06,
+      "loss": 0.7888,
+      "step": 3594
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.836434081449681,
+      "learning_rate": 8.283479894862233e-06,
+      "loss": 0.5704,
+      "step": 3595
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.2250188976590355,
+      "learning_rate": 8.282481444587764e-06,
+      "loss": 0.6061,
+      "step": 3596
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.100264445911385,
+      "learning_rate": 8.281482764225628e-06,
+      "loss": 0.5391,
+      "step": 3597
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.284825992082469,
+      "learning_rate": 8.280483853845831e-06,
+      "loss": 1.4993,
+      "step": 3598
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.683705715157677,
+      "learning_rate": 8.279484713518395e-06,
+      "loss": 1.1575,
+      "step": 3599
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.422930359517672,
+      "learning_rate": 8.27848534331335e-06,
+      "loss": 0.5458,
+      "step": 3600
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.337143182155289,
+      "learning_rate": 8.277485743300753e-06,
+      "loss": 0.6768,
+      "step": 3601
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.951202647514589,
+      "learning_rate": 8.27648591355067e-06,
+      "loss": 1.0886,
+      "step": 3602
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.1163177763316767,
+      "learning_rate": 8.27548585413318e-06,
+      "loss": 0.4608,
+      "step": 3603
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.473268162565548,
+      "learning_rate": 8.274485565118389e-06,
+      "loss": 0.9527,
+      "step": 3604
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.2134953665323875,
+      "learning_rate": 8.27348504657641e-06,
+      "loss": 1.02,
+      "step": 3605
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.865807389021216,
+      "learning_rate": 8.272484298577375e-06,
+      "loss": 1.2196,
+      "step": 3606
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.4482018269506804,
+      "learning_rate": 8.271483321191433e-06,
+      "loss": 1.0805,
+      "step": 3607
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.789927696800828,
+      "learning_rate": 8.270482114488748e-06,
+      "loss": 0.7778,
+      "step": 3608
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.3632231589887605,
+      "learning_rate": 8.269480678539498e-06,
+      "loss": 0.8847,
+      "step": 3609
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.1027736871674576,
+      "learning_rate": 8.268479013413883e-06,
+      "loss": 0.679,
+      "step": 3610
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9596146365430251,
+      "learning_rate": 8.267477119182115e-06,
+      "loss": 0.3198,
+      "step": 3611
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3201503867779574,
+      "learning_rate": 8.266474995914419e-06,
+      "loss": 0.1946,
+      "step": 3612
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.4227795740721945,
+      "learning_rate": 8.265472643681042e-06,
+      "loss": 0.5886,
+      "step": 3613
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9999825595731449,
+      "learning_rate": 8.264470062552246e-06,
+      "loss": 0.501,
+      "step": 3614
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.466292481798756,
+      "learning_rate": 8.263467252598303e-06,
+      "loss": 0.3048,
+      "step": 3615
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.3983750894590083,
+      "learning_rate": 8.26246421388951e-06,
+      "loss": 0.8234,
+      "step": 3616
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.9845882181511003,
+      "learning_rate": 8.261460946496172e-06,
+      "loss": 0.7159,
+      "step": 3617
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.963902341128361,
+      "learning_rate": 8.260457450488617e-06,
+      "loss": 0.5597,
+      "step": 3618
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.58947166693942,
+      "learning_rate": 8.259453725937184e-06,
+      "loss": 0.9475,
+      "step": 3619
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8608228824499164,
+      "learning_rate": 8.25844977291223e-06,
+      "loss": 0.4321,
+      "step": 3620
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.1629799771517217,
+      "learning_rate": 8.257445591484128e-06,
+      "loss": 0.7059,
+      "step": 3621
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.4020188663153315,
+      "learning_rate": 8.256441181723265e-06,
+      "loss": 0.5953,
+      "step": 3622
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.750190553423603,
+      "learning_rate": 8.255436543700048e-06,
+      "loss": 0.9565,
+      "step": 3623
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.420792324678391,
+      "learning_rate": 8.254431677484898e-06,
+      "loss": 0.9565,
+      "step": 3624
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.610989213133846,
+      "learning_rate": 8.25342658314825e-06,
+      "loss": 1.0059,
+      "step": 3625
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.5987006785906437,
+      "learning_rate": 8.252421260760558e-06,
+      "loss": 0.9466,
+      "step": 3626
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.658086668796395,
+      "learning_rate": 8.25141571039229e-06,
+      "loss": 1.2601,
+      "step": 3627
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.424797321212776,
+      "learning_rate": 8.25040993211393e-06,
+      "loss": 0.8074,
+      "step": 3628
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.945071389194256,
+      "learning_rate": 8.24940392599598e-06,
+      "loss": 0.6433,
+      "step": 3629
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.658612357174698,
+      "learning_rate": 8.248397692108957e-06,
+      "loss": 1.1627,
+      "step": 3630
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.488384567872834,
+      "learning_rate": 8.247391230523393e-06,
+      "loss": 1.3089,
+      "step": 3631
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.272377230796433,
+      "learning_rate": 8.246384541309835e-06,
+      "loss": 0.4514,
+      "step": 3632
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.105369688695962,
+      "learning_rate": 8.24537762453885e-06,
+      "loss": 0.8004,
+      "step": 3633
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.537957023672104,
+      "learning_rate": 8.244370480281018e-06,
+      "loss": 0.6889,
+      "step": 3634
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.014372334228167,
+      "learning_rate": 8.243363108606934e-06,
+      "loss": 1.0708,
+      "step": 3635
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.644010528781357,
+      "learning_rate": 8.24235550958721e-06,
+      "loss": 0.6836,
+      "step": 3636
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.8196684083357106,
+      "learning_rate": 8.241347683292478e-06,
+      "loss": 0.9738,
+      "step": 3637
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.07292046837355,
+      "learning_rate": 8.240339629793379e-06,
+      "loss": 1.3078,
+      "step": 3638
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.013052169089353,
+      "learning_rate": 8.239331349160573e-06,
+      "loss": 0.5815,
+      "step": 3639
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.653944549068703,
+      "learning_rate": 8.238322841464738e-06,
+      "loss": 0.3109,
+      "step": 3640
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.115651595817726,
+      "learning_rate": 8.237314106776563e-06,
+      "loss": 0.8189,
+      "step": 3641
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.3410805462475524,
+      "learning_rate": 8.236305145166761e-06,
+      "loss": 0.808,
+      "step": 3642
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.9670019774770364,
+      "learning_rate": 8.23529595670605e-06,
+      "loss": 0.6854,
+      "step": 3643
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.958379765225836,
+      "learning_rate": 8.234286541465175e-06,
+      "loss": 0.534,
+      "step": 3644
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.855724854233526,
+      "learning_rate": 8.233276899514887e-06,
+      "loss": 1.0768,
+      "step": 3645
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.519939370491934,
+      "learning_rate": 8.232267030925963e-06,
+      "loss": 1.2004,
+      "step": 3646
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.796719534626249,
+      "learning_rate": 8.231256935769183e-06,
+      "loss": 1.1672,
+      "step": 3647
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.58716913978193,
+      "learning_rate": 8.230246614115357e-06,
+      "loss": 1.191,
+      "step": 3648
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.5302816001180433,
+      "learning_rate": 8.2292360660353e-06,
+      "loss": 0.461,
+      "step": 3649
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.9954865922410283,
+      "learning_rate": 8.228225291599849e-06,
+      "loss": 0.8287,
+      "step": 3650
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.078826919973614,
+      "learning_rate": 8.227214290879855e-06,
+      "loss": 0.6676,
+      "step": 3651
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.6359386984785216,
+      "learning_rate": 8.226203063946184e-06,
+      "loss": 0.5838,
+      "step": 3652
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.600655242064704,
+      "learning_rate": 8.225191610869717e-06,
+      "loss": 0.8791,
+      "step": 3653
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.1916381293937146,
+      "learning_rate": 8.224179931721354e-06,
+      "loss": 0.5714,
+      "step": 3654
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.793635684996372,
+      "learning_rate": 8.22316802657201e-06,
+      "loss": 0.542,
+      "step": 3655
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.6517913737323298,
+      "learning_rate": 8.222155895492616e-06,
+      "loss": 0.8904,
+      "step": 3656
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.309634828087789,
+      "learning_rate": 8.221143538554116e-06,
+      "loss": 0.9436,
+      "step": 3657
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.59453233275118,
+      "learning_rate": 8.220130955827472e-06,
+      "loss": 1.6515,
+      "step": 3658
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.0694475479979255,
+      "learning_rate": 8.219118147383663e-06,
+      "loss": 1.2232,
+      "step": 3659
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.7378378994927486,
+      "learning_rate": 8.218105113293681e-06,
+      "loss": 1.0414,
+      "step": 3660
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.278096243544778,
+      "learning_rate": 8.217091853628535e-06,
+      "loss": 0.4086,
+      "step": 3661
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.204213682458238,
+      "learning_rate": 8.216078368459253e-06,
+      "loss": 0.7492,
+      "step": 3662
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.626511416048154,
+      "learning_rate": 8.21506465785687e-06,
+      "loss": 1.181,
+      "step": 3663
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.263531538490547,
+      "learning_rate": 8.21405072189245e-06,
+      "loss": 0.8094,
+      "step": 3664
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.842607134945213,
+      "learning_rate": 8.213036560637062e-06,
+      "loss": 0.9559,
+      "step": 3665
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.6295866100852026,
+      "learning_rate": 8.21202217416179e-06,
+      "loss": 0.6185,
+      "step": 3666
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.613191946253642,
+      "learning_rate": 8.211007562537747e-06,
+      "loss": 0.4489,
+      "step": 3667
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9761898696424298,
+      "learning_rate": 8.209992725836047e-06,
+      "loss": 0.4938,
+      "step": 3668
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.3974288514895035,
+      "learning_rate": 8.208977664127827e-06,
+      "loss": 1.1194,
+      "step": 3669
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.378116796132323,
+      "learning_rate": 8.207962377484237e-06,
+      "loss": 0.4233,
+      "step": 3670
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.093985355862017,
+      "learning_rate": 8.206946865976446e-06,
+      "loss": 1.0518,
+      "step": 3671
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.9070968664524504,
+      "learning_rate": 8.205931129675637e-06,
+      "loss": 0.7833,
+      "step": 3672
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.719407576076978,
+      "learning_rate": 8.204915168653007e-06,
+      "loss": 0.6622,
+      "step": 3673
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.7822855101196815,
+      "learning_rate": 8.203898982979773e-06,
+      "loss": 0.6128,
+      "step": 3674
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.4931324684841822,
+      "learning_rate": 8.202882572727161e-06,
+      "loss": 0.5935,
+      "step": 3675
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.591226514045569,
+      "learning_rate": 8.201865937966423e-06,
+      "loss": 1.1802,
+      "step": 3676
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.043022550682876,
+      "learning_rate": 8.200849078768816e-06,
+      "loss": 0.6867,
+      "step": 3677
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.6300983589675506,
+      "learning_rate": 8.199831995205619e-06,
+      "loss": 0.9301,
+      "step": 3678
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.276523837771925,
+      "learning_rate": 8.198814687348123e-06,
+      "loss": 0.5072,
+      "step": 3679
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.306168790115253,
+      "learning_rate": 8.19779715526764e-06,
+      "loss": 1.0553,
+      "step": 3680
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.0588477007410275,
+      "learning_rate": 8.196779399035492e-06,
+      "loss": 0.5845,
+      "step": 3681
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.9399757382701606,
+      "learning_rate": 8.195761418723023e-06,
+      "loss": 0.6021,
+      "step": 3682
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.9964090262739225,
+      "learning_rate": 8.194743214401587e-06,
+      "loss": 0.7194,
+      "step": 3683
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.500776219734207,
+      "learning_rate": 8.193724786142552e-06,
+      "loss": 0.9013,
+      "step": 3684
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.1845474908110916,
+      "learning_rate": 8.192706134017312e-06,
+      "loss": 1.0675,
+      "step": 3685
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.336093403098695,
+      "learning_rate": 8.191687258097264e-06,
+      "loss": 1.2922,
+      "step": 3686
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.091688178982156,
+      "learning_rate": 8.19066815845383e-06,
+      "loss": 1.1675,
+      "step": 3687
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.095007486460994,
+      "learning_rate": 8.189648835158445e-06,
+      "loss": 0.9352,
+      "step": 3688
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.983663848468049,
+      "learning_rate": 8.188629288282557e-06,
+      "loss": 0.9149,
+      "step": 3689
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.755782459984909,
+      "learning_rate": 8.187609517897634e-06,
+      "loss": 1.4777,
+      "step": 3690
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.4191866976219076,
+      "learning_rate": 8.186589524075155e-06,
+      "loss": 1.014,
+      "step": 3691
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.8368571027938874,
+      "learning_rate": 8.18556930688662e-06,
+      "loss": 1.0577,
+      "step": 3692
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.6395547430488384,
+      "learning_rate": 8.18454886640354e-06,
+      "loss": 0.9151,
+      "step": 3693
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.425943955850182,
+      "learning_rate": 8.183528202697441e-06,
+      "loss": 1.344,
+      "step": 3694
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.573148618469356,
+      "learning_rate": 8.18250731583987e-06,
+      "loss": 0.2741,
+      "step": 3695
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.766066715900613,
+      "learning_rate": 8.181486205902391e-06,
+      "loss": 0.8259,
+      "step": 3696
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4938051351655979,
+      "learning_rate": 8.180464872956572e-06,
+      "loss": 0.2153,
+      "step": 3697
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.04645686085355,
+      "learning_rate": 8.179443317074008e-06,
+      "loss": 0.897,
+      "step": 3698
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.281302380409822,
+      "learning_rate": 8.178421538326303e-06,
+      "loss": 0.8353,
+      "step": 3699
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.739685727324987,
+      "learning_rate": 8.17739953678508e-06,
+      "loss": 1.5031,
+      "step": 3700
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.517623643422315,
+      "learning_rate": 8.17637731252198e-06,
+      "loss": 1.1821,
+      "step": 3701
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.058673056478612,
+      "learning_rate": 8.175354865608651e-06,
+      "loss": 0.5239,
+      "step": 3702
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.307110932872007,
+      "learning_rate": 8.174332196116767e-06,
+      "loss": 1.2173,
+      "step": 3703
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.667206926341169,
+      "learning_rate": 8.173309304118011e-06,
+      "loss": 0.812,
+      "step": 3704
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.5122682057515293,
+      "learning_rate": 8.172286189684082e-06,
+      "loss": 0.5349,
+      "step": 3705
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.206192713863934,
+      "learning_rate": 8.171262852886697e-06,
+      "loss": 0.8859,
+      "step": 3706
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.745896340372273,
+      "learning_rate": 8.170239293797588e-06,
+      "loss": 0.6659,
+      "step": 3707
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.472030233684138,
+      "learning_rate": 8.169215512488502e-06,
+      "loss": 0.937,
+      "step": 3708
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.4137168251087555,
+      "learning_rate": 8.1681915090312e-06,
+      "loss": 0.5604,
+      "step": 3709
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.9059515725854372,
+      "learning_rate": 8.167167283497462e-06,
+      "loss": 1.3055,
+      "step": 3710
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.6243245126837236,
+      "learning_rate": 8.16614283595908e-06,
+      "loss": 1.3093,
+      "step": 3711
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.7052226321091295,
+      "learning_rate": 8.165118166487866e-06,
+      "loss": 0.563,
+      "step": 3712
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.6276544526695704,
+      "learning_rate": 8.164093275155642e-06,
+      "loss": 0.4162,
+      "step": 3713
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.414009897639559,
+      "learning_rate": 8.16306816203425e-06,
+      "loss": 0.5708,
+      "step": 3714
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.191779412183785,
+      "learning_rate": 8.162042827195545e-06,
+      "loss": 0.7869,
+      "step": 3715
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.966228928181594,
+      "learning_rate": 8.1610172707114e-06,
+      "loss": 1.0857,
+      "step": 3716
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.416053641244488,
+      "learning_rate": 8.1599914926537e-06,
+      "loss": 0.3816,
+      "step": 3717
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.4520616146089678,
+      "learning_rate": 8.158965493094349e-06,
+      "loss": 0.8468,
+      "step": 3718
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.866070693077448,
+      "learning_rate": 8.157939272105265e-06,
+      "loss": 0.3885,
+      "step": 3719
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.3887081665511363,
+      "learning_rate": 8.15691282975838e-06,
+      "loss": 0.5784,
+      "step": 3720
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.593655781114713,
+      "learning_rate": 8.155886166125647e-06,
+      "loss": 0.381,
+      "step": 3721
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1882794822789533,
+      "learning_rate": 8.154859281279028e-06,
+      "loss": 0.2174,
+      "step": 3722
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.8400830066215126,
+      "learning_rate": 8.153832175290503e-06,
+      "loss": 0.8367,
+      "step": 3723
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.357996923956879,
+      "learning_rate": 8.152804848232067e-06,
+      "loss": 0.7642,
+      "step": 3724
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.0239486314492354,
+      "learning_rate": 8.151777300175733e-06,
+      "loss": 0.3608,
+      "step": 3725
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.4587234054487888,
+      "learning_rate": 8.150749531193527e-06,
+      "loss": 0.8073,
+      "step": 3726
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.8268776728496308,
+      "learning_rate": 8.14972154135749e-06,
+      "loss": 0.8313,
+      "step": 3727
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.592494371196147,
+      "learning_rate": 8.14869333073968e-06,
+      "loss": 0.527,
+      "step": 3728
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.116864773442245,
+      "learning_rate": 8.147664899412174e-06,
+      "loss": 0.7039,
+      "step": 3729
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.763498367287853,
+      "learning_rate": 8.146636247447053e-06,
+      "loss": 0.766,
+      "step": 3730
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.474528023608184,
+      "learning_rate": 8.145607374916428e-06,
+      "loss": 1.0169,
+      "step": 3731
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.734071201766395,
+      "learning_rate": 8.144578281892414e-06,
+      "loss": 0.9553,
+      "step": 3732
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.092806982413093,
+      "learning_rate": 8.14354896844715e-06,
+      "loss": 1.055,
+      "step": 3733
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.42314577570228,
+      "learning_rate": 8.142519434652782e-06,
+      "loss": 0.8633,
+      "step": 3734
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.467308023579122,
+      "learning_rate": 8.141489680581481e-06,
+      "loss": 1.3204,
+      "step": 3735
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.970938266610763,
+      "learning_rate": 8.140459706305422e-06,
+      "loss": 0.6799,
+      "step": 3736
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.45068394944959,
+      "learning_rate": 8.139429511896806e-06,
+      "loss": 1.2216,
+      "step": 3737
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.082640105344123,
+      "learning_rate": 8.138399097427843e-06,
+      "loss": 0.5311,
+      "step": 3738
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.065166150068355,
+      "learning_rate": 8.137368462970762e-06,
+      "loss": 0.995,
+      "step": 3739
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3988745842381092,
+      "learning_rate": 8.136337608597807e-06,
+      "loss": 0.1561,
+      "step": 3740
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.16548411325666,
+      "learning_rate": 8.135306534381233e-06,
+      "loss": 0.4723,
+      "step": 3741
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.435847542644207,
+      "learning_rate": 8.134275240393318e-06,
+      "loss": 1.2246,
+      "step": 3742
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.052208610377437,
+      "learning_rate": 8.133243726706348e-06,
+      "loss": 1.0259,
+      "step": 3743
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.9003020589030672,
+      "learning_rate": 8.132211993392629e-06,
+      "loss": 1.0583,
+      "step": 3744
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.149256679099094,
+      "learning_rate": 8.131180040524482e-06,
+      "loss": 1.1742,
+      "step": 3745
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8161975702048256,
+      "learning_rate": 8.130147868174238e-06,
+      "loss": 0.354,
+      "step": 3746
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.830405910264299,
+      "learning_rate": 8.129115476414253e-06,
+      "loss": 0.9095,
+      "step": 3747
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.274315406313386,
+      "learning_rate": 8.12808286531689e-06,
+      "loss": 1.2045,
+      "step": 3748
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.008066219732132,
+      "learning_rate": 8.127050034954533e-06,
+      "loss": 1.0741,
+      "step": 3749
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.177893875464701,
+      "learning_rate": 8.126016985399576e-06,
+      "loss": 0.7071,
+      "step": 3750
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.260607866286649,
+      "learning_rate": 8.124983716724434e-06,
+      "loss": 1.4082,
+      "step": 3751
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.137610931060111,
+      "learning_rate": 8.123950229001533e-06,
+      "loss": 1.4752,
+      "step": 3752
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9728338746060647,
+      "learning_rate": 8.122916522303314e-06,
+      "loss": 0.3583,
+      "step": 3753
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.642077452337778,
+      "learning_rate": 8.121882596702241e-06,
+      "loss": 0.7528,
+      "step": 3754
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.488596401108742,
+      "learning_rate": 8.120848452270784e-06,
+      "loss": 1.0414,
+      "step": 3755
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.429872932359214,
+      "learning_rate": 8.119814089081433e-06,
+      "loss": 0.9902,
+      "step": 3756
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.132381624984531,
+      "learning_rate": 8.118779507206692e-06,
+      "loss": 1.2691,
+      "step": 3757
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.051652764278989,
+      "learning_rate": 8.117744706719077e-06,
+      "loss": 0.7505,
+      "step": 3758
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.7541486157549886,
+      "learning_rate": 8.11670968769113e-06,
+      "loss": 0.8725,
+      "step": 3759
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.2188128012707553,
+      "learning_rate": 8.115674450195397e-06,
+      "loss": 0.4125,
+      "step": 3760
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.8775421294704433,
+      "learning_rate": 8.114638994304442e-06,
+      "loss": 0.9549,
+      "step": 3761
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.24804227327714,
+      "learning_rate": 8.113603320090852e-06,
+      "loss": 0.5652,
+      "step": 3762
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.378678797966362,
+      "learning_rate": 8.112567427627218e-06,
+      "loss": 0.5553,
+      "step": 3763
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9313481307698672,
+      "learning_rate": 8.111531316986155e-06,
+      "loss": 0.3617,
+      "step": 3764
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.841068469149828,
+      "learning_rate": 8.110494988240287e-06,
+      "loss": 0.8996,
+      "step": 3765
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.5605387433865445,
+      "learning_rate": 8.109458441462257e-06,
+      "loss": 0.6492,
+      "step": 3766
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.577317967150967,
+      "learning_rate": 8.108421676724721e-06,
+      "loss": 0.9732,
+      "step": 3767
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.808841322327526,
+      "learning_rate": 8.107384694100355e-06,
+      "loss": 0.9966,
+      "step": 3768
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.424400232186057,
+      "learning_rate": 8.106347493661846e-06,
+      "loss": 0.9183,
+      "step": 3769
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.6727410221654617,
+      "learning_rate": 8.105310075481895e-06,
+      "loss": 0.7357,
+      "step": 3770
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.092483526607276,
+      "learning_rate": 8.104272439633225e-06,
+      "loss": 1.1123,
+      "step": 3771
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.640259166857718,
+      "learning_rate": 8.103234586188564e-06,
+      "loss": 1.1403,
+      "step": 3772
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.7035333161123165,
+      "learning_rate": 8.102196515220664e-06,
+      "loss": 0.6131,
+      "step": 3773
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.32058535500473,
+      "learning_rate": 8.10115822680229e-06,
+      "loss": 0.6775,
+      "step": 3774
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.5768608057991385,
+      "learning_rate": 8.100119721006221e-06,
+      "loss": 0.7513,
+      "step": 3775
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.461663425433065,
+      "learning_rate": 8.09908099790525e-06,
+      "loss": 1.1056,
+      "step": 3776
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.728957538433035,
+      "learning_rate": 8.098042057572188e-06,
+      "loss": 1.4042,
+      "step": 3777
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.687046450091034,
+      "learning_rate": 8.097002900079862e-06,
+      "loss": 1.29,
+      "step": 3778
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.511627277637122,
+      "learning_rate": 8.095963525501111e-06,
+      "loss": 1.322,
+      "step": 3779
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.8622060976057337,
+      "learning_rate": 8.094923933908789e-06,
+      "loss": 0.5612,
+      "step": 3780
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.995257314579237,
+      "learning_rate": 8.093884125375769e-06,
+      "loss": 0.9435,
+      "step": 3781
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.2454494244850616,
+      "learning_rate": 8.092844099974936e-06,
+      "loss": 0.3857,
+      "step": 3782
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.781446530884619,
+      "learning_rate": 8.091803857779194e-06,
+      "loss": 0.8944,
+      "step": 3783
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.7753792270276065,
+      "learning_rate": 8.090763398861455e-06,
+      "loss": 0.5534,
+      "step": 3784
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.6585854485006717,
+      "learning_rate": 8.089722723294654e-06,
+      "loss": 0.5455,
+      "step": 3785
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.2715132195590995,
+      "learning_rate": 8.088681831151737e-06,
+      "loss": 0.9291,
+      "step": 3786
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.6559595986554685,
+      "learning_rate": 8.087640722505663e-06,
+      "loss": 0.7922,
+      "step": 3787
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.9979353328556613,
+      "learning_rate": 8.086599397429413e-06,
+      "loss": 0.6418,
+      "step": 3788
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.749128142979223,
+      "learning_rate": 8.085557855995979e-06,
+      "loss": 0.5899,
+      "step": 3789
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.23753924224674,
+      "learning_rate": 8.084516098278367e-06,
+      "loss": 1.0294,
+      "step": 3790
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.2067365626434885,
+      "learning_rate": 8.083474124349602e-06,
+      "loss": 1.2521,
+      "step": 3791
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.053171316466954,
+      "learning_rate": 8.082431934282717e-06,
+      "loss": 0.6479,
+      "step": 3792
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.178116731766731,
+      "learning_rate": 8.081389528150772e-06,
+      "loss": 1.078,
+      "step": 3793
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.8930725769871146,
+      "learning_rate": 8.080346906026827e-06,
+      "loss": 0.9376,
+      "step": 3794
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.874489840565244,
+      "learning_rate": 8.079304067983972e-06,
+      "loss": 0.5448,
+      "step": 3795
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.0668608875262424,
+      "learning_rate": 8.078261014095303e-06,
+      "loss": 0.8477,
+      "step": 3796
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.441895416564226,
+      "learning_rate": 8.077217744433934e-06,
+      "loss": 1.211,
+      "step": 3797
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.2771698264465536,
+      "learning_rate": 8.076174259072994e-06,
+      "loss": 0.5761,
+      "step": 3798
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.614567455901114,
+      "learning_rate": 8.075130558085625e-06,
+      "loss": 1.0329,
+      "step": 3799
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.913152393460195,
+      "learning_rate": 8.074086641544985e-06,
+      "loss": 0.7162,
+      "step": 3800
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.877290036991181,
+      "learning_rate": 8.073042509524252e-06,
+      "loss": 0.6069,
+      "step": 3801
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.6871844033861536,
+      "learning_rate": 8.071998162096613e-06,
+      "loss": 0.6831,
+      "step": 3802
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.592996585961057,
+      "learning_rate": 8.070953599335271e-06,
+      "loss": 1.1962,
+      "step": 3803
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.986885576237931,
+      "learning_rate": 8.069908821313446e-06,
+      "loss": 0.8866,
+      "step": 3804
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.7512065263487377,
+      "learning_rate": 8.068863828104377e-06,
+      "loss": 0.6925,
+      "step": 3805
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.0276423770928598,
+      "learning_rate": 8.067818619781304e-06,
+      "loss": 0.6365,
+      "step": 3806
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.167495623338151,
+      "learning_rate": 8.0667731964175e-06,
+      "loss": 0.6901,
+      "step": 3807
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.766886335698619,
+      "learning_rate": 8.065727558086241e-06,
+      "loss": 0.4542,
+      "step": 3808
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.171617751108701,
+      "learning_rate": 8.064681704860824e-06,
+      "loss": 0.8883,
+      "step": 3809
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.7588444811396613,
+      "learning_rate": 8.063635636814555e-06,
+      "loss": 0.7518,
+      "step": 3810
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.447026050799633,
+      "learning_rate": 8.06258935402076e-06,
+      "loss": 1.1046,
+      "step": 3811
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.0596379802357623,
+      "learning_rate": 8.061542856552782e-06,
+      "loss": 0.6165,
+      "step": 3812
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.767255884519906,
+      "learning_rate": 8.06049614448397e-06,
+      "loss": 1.2185,
+      "step": 3813
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.873121207979868,
+      "learning_rate": 8.059449217887703e-06,
+      "loss": 0.7159,
+      "step": 3814
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.283891610711034,
+      "learning_rate": 8.058402076837357e-06,
+      "loss": 0.5596,
+      "step": 3815
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.062392687744952,
+      "learning_rate": 8.057354721406336e-06,
+      "loss": 0.8439,
+      "step": 3816
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.557955428306261,
+      "learning_rate": 8.056307151668056e-06,
+      "loss": 1.1224,
+      "step": 3817
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.138074034895307,
+      "learning_rate": 8.055259367695944e-06,
+      "loss": 1.0132,
+      "step": 3818
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.846901873436069,
+      "learning_rate": 8.054211369563448e-06,
+      "loss": 0.7708,
+      "step": 3819
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.566826926447175,
+      "learning_rate": 8.053163157344025e-06,
+      "loss": 1.1192,
+      "step": 3820
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.438548471422388,
+      "learning_rate": 8.052114731111154e-06,
+      "loss": 1.3369,
+      "step": 3821
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.2034914657844618,
+      "learning_rate": 8.051066090938325e-06,
+      "loss": 0.6045,
+      "step": 3822
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.354183372517673,
+      "learning_rate": 8.050017236899038e-06,
+      "loss": 0.6668,
+      "step": 3823
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.856461759994505,
+      "learning_rate": 8.048968169066817e-06,
+      "loss": 1.0358,
+      "step": 3824
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.1394448375714346,
+      "learning_rate": 8.047918887515199e-06,
+      "loss": 0.6115,
+      "step": 3825
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.307289435541749,
+      "learning_rate": 8.046869392317729e-06,
+      "loss": 0.4689,
+      "step": 3826
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.4991103377960795,
+      "learning_rate": 8.045819683547976e-06,
+      "loss": 0.6016,
+      "step": 3827
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.902513823352388,
+      "learning_rate": 8.044769761279516e-06,
+      "loss": 0.8296,
+      "step": 3828
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7915524409926955,
+      "learning_rate": 8.04371962558595e-06,
+      "loss": 0.4232,
+      "step": 3829
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.357581688423257,
+      "learning_rate": 8.042669276540882e-06,
+      "loss": 0.7239,
+      "step": 3830
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.034115483169221,
+      "learning_rate": 8.041618714217941e-06,
+      "loss": 1.0947,
+      "step": 3831
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.4102344394295265,
+      "learning_rate": 8.040567938690764e-06,
+      "loss": 0.7779,
+      "step": 3832
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.9394698900528162,
+      "learning_rate": 8.039516950033008e-06,
+      "loss": 0.7769,
+      "step": 3833
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.7683839440198894,
+      "learning_rate": 8.038465748318342e-06,
+      "loss": 0.7318,
+      "step": 3834
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.805228173708322,
+      "learning_rate": 8.03741433362045e-06,
+      "loss": 0.9185,
+      "step": 3835
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.114508005145592,
+      "learning_rate": 8.036362706013033e-06,
+      "loss": 1.2705,
+      "step": 3836
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.760045632096391,
+      "learning_rate": 8.035310865569806e-06,
+      "loss": 0.7342,
+      "step": 3837
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.133031716360296,
+      "learning_rate": 8.034258812364492e-06,
+      "loss": 0.6147,
+      "step": 3838
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.8580255170415314,
+      "learning_rate": 8.033206546470846e-06,
+      "loss": 0.6132,
+      "step": 3839
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.789984994703072,
+      "learning_rate": 8.032154067962617e-06,
+      "loss": 0.9953,
+      "step": 3840
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.252557561048338,
+      "learning_rate": 8.031101376913586e-06,
+      "loss": 0.693,
+      "step": 3841
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.403266683476786,
+      "learning_rate": 8.03004847339754e-06,
+      "loss": 1.0276,
+      "step": 3842
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.288651909984002,
+      "learning_rate": 8.028995357488284e-06,
+      "loss": 0.1872,
+      "step": 3843
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.199625220667626,
+      "learning_rate": 8.027942029259633e-06,
+      "loss": 0.5979,
+      "step": 3844
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.0408324902372055,
+      "learning_rate": 8.026888488785426e-06,
+      "loss": 0.9398,
+      "step": 3845
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.251594099668823,
+      "learning_rate": 8.025834736139506e-06,
+      "loss": 0.8767,
+      "step": 3846
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.90119502716278,
+      "learning_rate": 8.024780771395741e-06,
+      "loss": 0.7032,
+      "step": 3847
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.3691113441484335,
+      "learning_rate": 8.023726594628008e-06,
+      "loss": 1.0865,
+      "step": 3848
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.272557560423606,
+      "learning_rate": 8.0226722059102e-06,
+      "loss": 0.9431,
+      "step": 3849
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.1018996779982,
+      "learning_rate": 8.021617605316225e-06,
+      "loss": 1.1198,
+      "step": 3850
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.34270314985889,
+      "learning_rate": 8.020562792920007e-06,
+      "loss": 1.4252,
+      "step": 3851
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.6576903851374674,
+      "learning_rate": 8.019507768795481e-06,
+      "loss": 0.5995,
+      "step": 3852
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.676769230408065,
+      "learning_rate": 8.018452533016604e-06,
+      "loss": 0.6054,
+      "step": 3853
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.304160513568824,
+      "learning_rate": 8.017397085657337e-06,
+      "loss": 1.1721,
+      "step": 3854
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.834167523884977,
+      "learning_rate": 8.016341426791669e-06,
+      "loss": 1.001,
+      "step": 3855
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.332024439659001,
+      "learning_rate": 8.015285556493592e-06,
+      "loss": 0.4388,
+      "step": 3856
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.011369652265117,
+      "learning_rate": 8.014229474837122e-06,
+      "loss": 1.1744,
+      "step": 3857
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.196051480762647,
+      "learning_rate": 8.013173181896283e-06,
+      "loss": 0.4594,
+      "step": 3858
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.8879096058266005,
+      "learning_rate": 8.012116677745116e-06,
+      "loss": 1.116,
+      "step": 3859
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.1754368357717064,
+      "learning_rate": 8.011059962457681e-06,
+      "loss": 0.605,
+      "step": 3860
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.411006122417595,
+      "learning_rate": 8.010003036108045e-06,
+      "loss": 0.402,
+      "step": 3861
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.7769095221665934,
+      "learning_rate": 8.008945898770298e-06,
+      "loss": 0.5581,
+      "step": 3862
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.4745966802453316,
+      "learning_rate": 8.007888550518536e-06,
+      "loss": 0.5202,
+      "step": 3863
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.0437840671956957,
+      "learning_rate": 8.006830991426879e-06,
+      "loss": 0.8083,
+      "step": 3864
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.4753333783047267,
+      "learning_rate": 8.005773221569453e-06,
+      "loss": 0.8619,
+      "step": 3865
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.218536691525492,
+      "learning_rate": 8.004715241020407e-06,
+      "loss": 1.2444,
+      "step": 3866
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.707391376778872,
+      "learning_rate": 8.0036570498539e-06,
+      "loss": 1.2269,
+      "step": 3867
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.269846290227814,
+      "learning_rate": 8.002598648144104e-06,
+      "loss": 0.7239,
+      "step": 3868
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.416247878295309,
+      "learning_rate": 8.00154003596521e-06,
+      "loss": 0.8834,
+      "step": 3869
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.61950119442241,
+      "learning_rate": 8.000481213391422e-06,
+      "loss": 0.5118,
+      "step": 3870
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.452753986663066,
+      "learning_rate": 7.999422180496961e-06,
+      "loss": 1.2589,
+      "step": 3871
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.1217098336646676,
+      "learning_rate": 7.998362937356057e-06,
+      "loss": 0.7483,
+      "step": 3872
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.665134235943946,
+      "learning_rate": 7.99730348404296e-06,
+      "loss": 0.57,
+      "step": 3873
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.284385585767888,
+      "learning_rate": 7.996243820631936e-06,
+      "loss": 1.1521,
+      "step": 3874
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.167941665748867,
+      "learning_rate": 7.995183947197256e-06,
+      "loss": 0.8497,
+      "step": 3875
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.1403467646995615,
+      "learning_rate": 7.994123863813217e-06,
+      "loss": 0.8655,
+      "step": 3876
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.602304879232952,
+      "learning_rate": 7.993063570554128e-06,
+      "loss": 0.6588,
+      "step": 3877
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.783080663293601,
+      "learning_rate": 7.992003067494307e-06,
+      "loss": 0.8908,
+      "step": 3878
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.8033073441225937,
+      "learning_rate": 7.990942354708093e-06,
+      "loss": 0.6443,
+      "step": 3879
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.077322224246064,
+      "learning_rate": 7.989881432269835e-06,
+      "loss": 0.8065,
+      "step": 3880
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.7208368668094294,
+      "learning_rate": 7.988820300253902e-06,
+      "loss": 0.755,
+      "step": 3881
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.6569711380052623,
+      "learning_rate": 7.987758958734672e-06,
+      "loss": 1.0911,
+      "step": 3882
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.7112812737009873,
+      "learning_rate": 7.986697407786543e-06,
+      "loss": 1.043,
+      "step": 3883
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.831832105716206,
+      "learning_rate": 7.985635647483922e-06,
+      "loss": 1.0517,
+      "step": 3884
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.195462499961567,
+      "learning_rate": 7.984573677901238e-06,
+      "loss": 0.6196,
+      "step": 3885
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.0927161199014352,
+      "learning_rate": 7.983511499112927e-06,
+      "loss": 0.747,
+      "step": 3886
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.1604091144922686,
+      "learning_rate": 7.982449111193445e-06,
+      "loss": 0.7857,
+      "step": 3887
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.184526063278354,
+      "learning_rate": 7.98138651421726e-06,
+      "loss": 0.7998,
+      "step": 3888
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.414539531775396,
+      "learning_rate": 7.980323708258854e-06,
+      "loss": 1.2143,
+      "step": 3889
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.0841779869278705,
+      "learning_rate": 7.979260693392726e-06,
+      "loss": 1.0425,
+      "step": 3890
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.97584223991958,
+      "learning_rate": 7.978197469693392e-06,
+      "loss": 0.7008,
+      "step": 3891
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.040826383131776,
+      "learning_rate": 7.977134037235375e-06,
+      "loss": 0.5522,
+      "step": 3892
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.171089779923319,
+      "learning_rate": 7.976070396093217e-06,
+      "loss": 0.5813,
+      "step": 3893
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.3970568499228593,
+      "learning_rate": 7.975006546341478e-06,
+      "loss": 0.366,
+      "step": 3894
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.1551192192534327,
+      "learning_rate": 7.973942488054726e-06,
+      "loss": 0.6433,
+      "step": 3895
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.4197719603725343,
+      "learning_rate": 7.97287822130755e-06,
+      "loss": 0.5926,
+      "step": 3896
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.6648438074557177,
+      "learning_rate": 7.971813746174548e-06,
+      "loss": 0.5191,
+      "step": 3897
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.228504648321421,
+      "learning_rate": 7.970749062730336e-06,
+      "loss": 0.8926,
+      "step": 3898
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.388096166572964,
+      "learning_rate": 7.969684171049543e-06,
+      "loss": 0.7807,
+      "step": 3899
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.4729634968726595,
+      "learning_rate": 7.968619071206813e-06,
+      "loss": 1.2072,
+      "step": 3900
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.345484389252216,
+      "learning_rate": 7.967553763276808e-06,
+      "loss": 0.8457,
+      "step": 3901
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.7935286469224163,
+      "learning_rate": 7.966488247334196e-06,
+      "loss": 0.8782,
+      "step": 3902
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.29533451863232,
+      "learning_rate": 7.96542252345367e-06,
+      "loss": 0.58,
+      "step": 3903
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.365676149628274,
+      "learning_rate": 7.96435659170993e-06,
+      "loss": 1.2744,
+      "step": 3904
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.6534210181216107,
+      "learning_rate": 7.963290452177695e-06,
+      "loss": 0.9971,
+      "step": 3905
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6606709479755615,
+      "learning_rate": 7.962224104931692e-06,
+      "loss": 0.2075,
+      "step": 3906
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.692509435384667,
+      "learning_rate": 7.961157550046674e-06,
+      "loss": 0.8298,
+      "step": 3907
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.3064776406188825,
+      "learning_rate": 7.9600907875974e-06,
+      "loss": 0.6335,
+      "step": 3908
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.2577188657508187,
+      "learning_rate": 7.959023817658642e-06,
+      "loss": 0.9199,
+      "step": 3909
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.065028697756866,
+      "learning_rate": 7.957956640305192e-06,
+      "loss": 0.4054,
+      "step": 3910
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.275405899307535,
+      "learning_rate": 7.956889255611855e-06,
+      "loss": 0.1956,
+      "step": 3911
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.874605605122154,
+      "learning_rate": 7.955821663653448e-06,
+      "loss": 0.5685,
+      "step": 3912
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.4400953132304206,
+      "learning_rate": 7.95475386450481e-06,
+      "loss": 0.8722,
+      "step": 3913
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.918856789489256,
+      "learning_rate": 7.953685858240783e-06,
+      "loss": 1.1118,
+      "step": 3914
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.0210892735943915,
+      "learning_rate": 7.952617644936231e-06,
+      "loss": 1.0947,
+      "step": 3915
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.330205720746106,
+      "learning_rate": 7.951549224666034e-06,
+      "loss": 1.2001,
+      "step": 3916
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.272606217468375,
+      "learning_rate": 7.950480597505079e-06,
+      "loss": 0.834,
+      "step": 3917
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.6360691348276872,
+      "learning_rate": 7.949411763528276e-06,
+      "loss": 0.7878,
+      "step": 3918
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.357258381230591,
+      "learning_rate": 7.948342722810544e-06,
+      "loss": 0.6788,
+      "step": 3919
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.790471200518028,
+      "learning_rate": 7.94727347542682e-06,
+      "loss": 1.0941,
+      "step": 3920
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.880762080944756,
+      "learning_rate": 7.946204021452049e-06,
+      "loss": 0.5526,
+      "step": 3921
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.906717414163718,
+      "learning_rate": 7.9451343609612e-06,
+      "loss": 1.049,
+      "step": 3922
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.4014886526948636,
+      "learning_rate": 7.94406449402925e-06,
+      "loss": 0.7849,
+      "step": 3923
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.5413370736232705,
+      "learning_rate": 7.942994420731189e-06,
+      "loss": 0.402,
+      "step": 3924
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.1905916450375775,
+      "learning_rate": 7.94192414114203e-06,
+      "loss": 0.9411,
+      "step": 3925
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.654706974498102,
+      "learning_rate": 7.940853655336791e-06,
+      "loss": 0.5907,
+      "step": 3926
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.672345696377473,
+      "learning_rate": 7.93978296339051e-06,
+      "loss": 0.7108,
+      "step": 3927
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.6847837849040816,
+      "learning_rate": 7.938712065378235e-06,
+      "loss": 0.9222,
+      "step": 3928
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.046999168227711,
+      "learning_rate": 7.937640961375036e-06,
+      "loss": 0.8271,
+      "step": 3929
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.5800151195645653,
+      "learning_rate": 7.936569651455987e-06,
+      "loss": 0.8926,
+      "step": 3930
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.6666933200400034,
+      "learning_rate": 7.93549813569619e-06,
+      "loss": 0.6696,
+      "step": 3931
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.425401258509315,
+      "learning_rate": 7.934426414170747e-06,
+      "loss": 1.005,
+      "step": 3932
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.191697810533967,
+      "learning_rate": 7.933354486954781e-06,
+      "loss": 0.9103,
+      "step": 3933
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.755872248436024,
+      "learning_rate": 7.932282354123433e-06,
+      "loss": 0.4972,
+      "step": 3934
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.279827015704212,
+      "learning_rate": 7.931210015751854e-06,
+      "loss": 0.9444,
+      "step": 3935
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8322074800850854,
+      "learning_rate": 7.93013747191521e-06,
+      "loss": 0.3787,
+      "step": 3936
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.503534226337361,
+      "learning_rate": 7.92906472268868e-06,
+      "loss": 1.0117,
+      "step": 3937
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.9678540703978524,
+      "learning_rate": 7.92799176814746e-06,
+      "loss": 0.7385,
+      "step": 3938
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.4197151347721753,
+      "learning_rate": 7.92691860836676e-06,
+      "loss": 0.9999,
+      "step": 3939
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.05089929572889,
+      "learning_rate": 7.925845243421803e-06,
+      "loss": 1.1175,
+      "step": 3940
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.487274292627951,
+      "learning_rate": 7.924771673387828e-06,
+      "loss": 0.7277,
+      "step": 3941
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.203089793147763,
+      "learning_rate": 7.92369789834009e-06,
+      "loss": 0.6635,
+      "step": 3942
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.680738296993968,
+      "learning_rate": 7.92262391835385e-06,
+      "loss": 1.0936,
+      "step": 3943
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.252418062665237,
+      "learning_rate": 7.921549733504394e-06,
+      "loss": 1.4216,
+      "step": 3944
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.721036769804235,
+      "learning_rate": 7.920475343867016e-06,
+      "loss": 0.8752,
+      "step": 3945
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.060450737205802,
+      "learning_rate": 7.919400749517026e-06,
+      "loss": 0.9217,
+      "step": 3946
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.1952284512374827,
+      "learning_rate": 7.91832595052975e-06,
+      "loss": 0.6315,
+      "step": 3947
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.6845820229727635,
+      "learning_rate": 7.917250946980525e-06,
+      "loss": 0.6788,
+      "step": 3948
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.2111912846058237,
+      "learning_rate": 7.916175738944703e-06,
+      "loss": 0.962,
+      "step": 3949
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.310775447454021,
+      "learning_rate": 7.915100326497655e-06,
+      "loss": 0.5742,
+      "step": 3950
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.454874359101468,
+      "learning_rate": 7.91402470971476e-06,
+      "loss": 1.0579,
+      "step": 3951
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.777140529355507,
+      "learning_rate": 7.912948888671415e-06,
+      "loss": 0.848,
+      "step": 3952
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.917375205195893,
+      "learning_rate": 7.911872863443028e-06,
+      "loss": 1.0811,
+      "step": 3953
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.20008656594547,
+      "learning_rate": 7.910796634105029e-06,
+      "loss": 0.8757,
+      "step": 3954
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.3461871368791574,
+      "learning_rate": 7.90972020073285e-06,
+      "loss": 0.6789,
+      "step": 3955
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.726416038016359,
+      "learning_rate": 7.90864356340195e-06,
+      "loss": 0.8677,
+      "step": 3956
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.5161194158107385,
+      "learning_rate": 7.907566722187796e-06,
+      "loss": 0.8401,
+      "step": 3957
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.44494707113555,
+      "learning_rate": 7.906489677165868e-06,
+      "loss": 0.765,
+      "step": 3958
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.571669862485657,
+      "learning_rate": 7.905412428411663e-06,
+      "loss": 0.9774,
+      "step": 3959
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.7726222805080134,
+      "learning_rate": 7.90433497600069e-06,
+      "loss": 0.8009,
+      "step": 3960
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.904481349413059,
+      "learning_rate": 7.903257320008475e-06,
+      "loss": 0.6383,
+      "step": 3961
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.329707161356198,
+      "learning_rate": 7.902179460510558e-06,
+      "loss": 1.1996,
+      "step": 3962
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.031137913745396,
+      "learning_rate": 7.901101397582489e-06,
+      "loss": 1.0229,
+      "step": 3963
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.233129840323534,
+      "learning_rate": 7.90002313129984e-06,
+      "loss": 1.5545,
+      "step": 3964
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.5120920452991,
+      "learning_rate": 7.898944661738188e-06,
+      "loss": 0.8688,
+      "step": 3965
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.8309770680435657,
+      "learning_rate": 7.897865988973133e-06,
+      "loss": 0.6724,
+      "step": 3966
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.735828399323141,
+      "learning_rate": 7.896787113080284e-06,
+      "loss": 0.6444,
+      "step": 3967
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.959729189806863,
+      "learning_rate": 7.895708034135265e-06,
+      "loss": 1.0035,
+      "step": 3968
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.190551313321349,
+      "learning_rate": 7.894628752213715e-06,
+      "loss": 0.6192,
+      "step": 3969
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.8311685842134064,
+      "learning_rate": 7.893549267391287e-06,
+      "loss": 0.8306,
+      "step": 3970
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.819702650845054,
+      "learning_rate": 7.892469579743647e-06,
+      "loss": 0.7833,
+      "step": 3971
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.5640523641815736,
+      "learning_rate": 7.891389689346479e-06,
+      "loss": 0.3882,
+      "step": 3972
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.3126765364576305,
+      "learning_rate": 7.890309596275476e-06,
+      "loss": 0.5957,
+      "step": 3973
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.89337526452459,
+      "learning_rate": 7.88922930060635e-06,
+      "loss": 0.6068,
+      "step": 3974
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.134076773870964,
+      "learning_rate": 7.888148802414824e-06,
+      "loss": 0.8467,
+      "step": 3975
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.898634541280175,
+      "learning_rate": 7.887068101776633e-06,
+      "loss": 0.9021,
+      "step": 3976
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.921800857438866,
+      "learning_rate": 7.885987198767538e-06,
+      "loss": 0.4761,
+      "step": 3977
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.193341180831775,
+      "learning_rate": 7.884906093463295e-06,
+      "loss": 0.585,
+      "step": 3978
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.8833964995321897,
+      "learning_rate": 7.883824785939694e-06,
+      "loss": 0.4738,
+      "step": 3979
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.5041822053988834,
+      "learning_rate": 7.882743276272524e-06,
+      "loss": 0.5757,
+      "step": 3980
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.479653639530403,
+      "learning_rate": 7.881661564537598e-06,
+      "loss": 1.0564,
+      "step": 3981
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.7477370824532636,
+      "learning_rate": 7.880579650810736e-06,
+      "loss": 0.7131,
+      "step": 3982
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.8493434819124897,
+      "learning_rate": 7.879497535167776e-06,
+      "loss": 0.5089,
+      "step": 3983
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.462570167528221,
+      "learning_rate": 7.87841521768457e-06,
+      "loss": 0.9516,
+      "step": 3984
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.7688333887424483,
+      "learning_rate": 7.877332698436986e-06,
+      "loss": 0.4017,
+      "step": 3985
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9950498856678862,
+      "learning_rate": 7.876249977500902e-06,
+      "loss": 0.3658,
+      "step": 3986
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.095597838558024,
+      "learning_rate": 7.875167054952211e-06,
+      "loss": 1.2446,
+      "step": 3987
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.749989836918884,
+      "learning_rate": 7.874083930866822e-06,
+      "loss": 1.0899,
+      "step": 3988
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.9102664158338305,
+      "learning_rate": 7.873000605320658e-06,
+      "loss": 0.6445,
+      "step": 3989
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.269179065194626,
+      "learning_rate": 7.871917078389655e-06,
+      "loss": 1.2587,
+      "step": 3990
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.166224022227581,
+      "learning_rate": 7.870833350149764e-06,
+      "loss": 0.941,
+      "step": 3991
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.300253791870946,
+      "learning_rate": 7.869749420676949e-06,
+      "loss": 0.9979,
+      "step": 3992
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.4175409176355664,
+      "learning_rate": 7.868665290047189e-06,
+      "loss": 0.7883,
+      "step": 3993
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.183480038731687,
+      "learning_rate": 7.867580958336474e-06,
+      "loss": 1.0306,
+      "step": 3994
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.974127330270877,
+      "learning_rate": 7.866496425620816e-06,
+      "loss": 0.9258,
+      "step": 3995
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.185511981526038,
+      "learning_rate": 7.865411691976232e-06,
+      "loss": 1.3501,
+      "step": 3996
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2284723128580735,
+      "learning_rate": 7.864326757478762e-06,
+      "loss": 0.1758,
+      "step": 3997
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.218909079487965,
+      "learning_rate": 7.86324162220445e-06,
+      "loss": 0.8576,
+      "step": 3998
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.174994398303952,
+      "learning_rate": 7.86215628622936e-06,
+      "loss": 1.5497,
+      "step": 3999
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.383769075116413,
+      "learning_rate": 7.86107074962957e-06,
+      "loss": 0.6589,
+      "step": 4000
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.153776451524846,
+      "learning_rate": 7.859985012481174e-06,
+      "loss": 0.5676,
+      "step": 4001
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.212803702193593,
+      "learning_rate": 7.858899074860276e-06,
+      "loss": 0.6885,
+      "step": 4002
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.179651835380731,
+      "learning_rate": 7.857812936842992e-06,
+      "loss": 1.0661,
+      "step": 4003
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.44607333850051,
+      "learning_rate": 7.856726598505459e-06,
+      "loss": 0.5944,
+      "step": 4004
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.3703513487093195,
+      "learning_rate": 7.855640059923826e-06,
+      "loss": 0.9042,
+      "step": 4005
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.3136300965208285,
+      "learning_rate": 7.85455332117425e-06,
+      "loss": 1.0606,
+      "step": 4006
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.5904783749373816,
+      "learning_rate": 7.85346638233291e-06,
+      "loss": 0.2983,
+      "step": 4007
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.986458941824462,
+      "learning_rate": 7.852379243475998e-06,
+      "loss": 1.0047,
+      "step": 4008
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.4523716432618428,
+      "learning_rate": 7.851291904679712e-06,
+      "loss": 0.8155,
+      "step": 4009
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.525587156791062,
+      "learning_rate": 7.850204366020271e-06,
+      "loss": 0.4656,
+      "step": 4010
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.1498400625389915,
+      "learning_rate": 7.84911662757391e-06,
+      "loss": 0.9032,
+      "step": 4011
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.740206562939246,
+      "learning_rate": 7.848028689416872e-06,
+      "loss": 1.2643,
+      "step": 4012
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.0050724670801054,
+      "learning_rate": 7.846940551625417e-06,
+      "loss": 0.8215,
+      "step": 4013
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.3537970585947905,
+      "learning_rate": 7.845852214275821e-06,
+      "loss": 1.0797,
+      "step": 4014
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.0278943481498746,
+      "learning_rate": 7.84476367744437e-06,
+      "loss": 0.69,
+      "step": 4015
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.741402674292105,
+      "learning_rate": 7.843674941207364e-06,
+      "loss": 0.9737,
+      "step": 4016
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.492878113334946,
+      "learning_rate": 7.842586005641123e-06,
+      "loss": 0.8798,
+      "step": 4017
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.533598324291426,
+      "learning_rate": 7.84149687082197e-06,
+      "loss": 0.7147,
+      "step": 4018
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.769338538009378,
+      "learning_rate": 7.840407536826256e-06,
+      "loss": 0.8957,
+      "step": 4019
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.860169527682215,
+      "learning_rate": 7.839318003730332e-06,
+      "loss": 0.9229,
+      "step": 4020
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.5413965574010757,
+      "learning_rate": 7.838228271610575e-06,
+      "loss": 0.3215,
+      "step": 4021
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.664172854252663,
+      "learning_rate": 7.837138340543368e-06,
+      "loss": 0.9405,
+      "step": 4022
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.704049865569673,
+      "learning_rate": 7.836048210605109e-06,
+      "loss": 1.0005,
+      "step": 4023
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.0229149140174143,
+      "learning_rate": 7.834957881872215e-06,
+      "loss": 0.5615,
+      "step": 4024
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.293901840144184,
+      "learning_rate": 7.83386735442111e-06,
+      "loss": 0.8158,
+      "step": 4025
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.9068154726580633,
+      "learning_rate": 7.832776628328237e-06,
+      "loss": 0.6568,
+      "step": 4026
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.3477590814031055,
+      "learning_rate": 7.831685703670052e-06,
+      "loss": 0.8672,
+      "step": 4027
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.9376357638709742,
+      "learning_rate": 7.83059458052302e-06,
+      "loss": 0.6062,
+      "step": 4028
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.25054189403912,
+      "learning_rate": 7.829503258963629e-06,
+      "loss": 1.1562,
+      "step": 4029
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.477460824502643,
+      "learning_rate": 7.82841173906837e-06,
+      "loss": 1.5191,
+      "step": 4030
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.789155502998915,
+      "learning_rate": 7.827320020913762e-06,
+      "loss": 1.2872,
+      "step": 4031
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.208558129868596,
+      "learning_rate": 7.826228104576324e-06,
+      "loss": 0.9987,
+      "step": 4032
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.821842829169687,
+      "learning_rate": 7.825135990132592e-06,
+      "loss": 1.5145,
+      "step": 4033
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.6215811257063426,
+      "learning_rate": 7.824043677659127e-06,
+      "loss": 1.0486,
+      "step": 4034
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.879403051650573,
+      "learning_rate": 7.822951167232488e-06,
+      "loss": 1.1389,
+      "step": 4035
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.3982667736288654,
+      "learning_rate": 7.821858458929256e-06,
+      "loss": 0.6997,
+      "step": 4036
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.189642795215071,
+      "learning_rate": 7.82076555282603e-06,
+      "loss": 1.3593,
+      "step": 4037
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.261351402183224,
+      "learning_rate": 7.819672448999413e-06,
+      "loss": 1.0783,
+      "step": 4038
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.0582346994452942,
+      "learning_rate": 7.81857914752603e-06,
+      "loss": 0.8212,
+      "step": 4039
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.128775511789336,
+      "learning_rate": 7.817485648482514e-06,
+      "loss": 1.4036,
+      "step": 4040
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.1862480949901624,
+      "learning_rate": 7.816391951945517e-06,
+      "loss": 0.5946,
+      "step": 4041
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.5393890511012724,
+      "learning_rate": 7.8152980579917e-06,
+      "loss": 0.5955,
+      "step": 4042
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.291111423854695,
+      "learning_rate": 7.814203966697742e-06,
+      "loss": 1.3313,
+      "step": 4043
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.198816817119858,
+      "learning_rate": 7.813109678140333e-06,
+      "loss": 0.6167,
+      "step": 4044
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.178105362579311,
+      "learning_rate": 7.812015192396178e-06,
+      "loss": 1.0517,
+      "step": 4045
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.2557654770972406,
+      "learning_rate": 7.810920509541997e-06,
+      "loss": 0.6081,
+      "step": 4046
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.158509861411965,
+      "learning_rate": 7.80982562965452e-06,
+      "loss": 1.0021,
+      "step": 4047
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.6075015660901797,
+      "learning_rate": 7.808730552810496e-06,
+      "loss": 0.5844,
+      "step": 4048
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.118174912930761,
+      "learning_rate": 7.807635279086682e-06,
+      "loss": 0.5398,
+      "step": 4049
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.087999613701567,
+      "learning_rate": 7.806539808559856e-06,
+      "loss": 0.7191,
+      "step": 4050
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2024083038375055,
+      "learning_rate": 7.805444141306804e-06,
+      "loss": 0.2323,
+      "step": 4051
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.287541535163385,
+      "learning_rate": 7.804348277404324e-06,
+      "loss": 0.802,
+      "step": 4052
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.3546151266802453,
+      "learning_rate": 7.803252216929236e-06,
+      "loss": 0.7602,
+      "step": 4053
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.237050962673722,
+      "learning_rate": 7.802155959958368e-06,
+      "loss": 0.6012,
+      "step": 4054
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.6069184684330327,
+      "learning_rate": 7.80105950656856e-06,
+      "loss": 0.8598,
+      "step": 4055
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.9434142461101187,
+      "learning_rate": 7.799962856836674e-06,
+      "loss": 0.2683,
+      "step": 4056
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.135090704148495,
+      "learning_rate": 7.798866010839577e-06,
+      "loss": 0.4931,
+      "step": 4057
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.8851883209140317,
+      "learning_rate": 7.79776896865415e-06,
+      "loss": 0.4785,
+      "step": 4058
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.578701701147169,
+      "learning_rate": 7.796671730357296e-06,
+      "loss": 1.5124,
+      "step": 4059
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.839867870120802,
+      "learning_rate": 7.795574296025925e-06,
+      "loss": 1.0868,
+      "step": 4060
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.327365478228468,
+      "learning_rate": 7.794476665736964e-06,
+      "loss": 1.2185,
+      "step": 4061
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.7570060148375393,
+      "learning_rate": 7.793378839567348e-06,
+      "loss": 0.7382,
+      "step": 4062
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.8834551320663824,
+      "learning_rate": 7.792280817594033e-06,
+      "loss": 0.775,
+      "step": 4063
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.0132087833361663,
+      "learning_rate": 7.791182599893985e-06,
+      "loss": 0.5297,
+      "step": 4064
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 6.041447540688191,
+      "learning_rate": 7.790084186544183e-06,
+      "loss": 1.5578,
+      "step": 4065
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3766809671719165,
+      "learning_rate": 7.788985577621623e-06,
+      "loss": 0.2116,
+      "step": 4066
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.3972620375857447,
+      "learning_rate": 7.78788677320331e-06,
+      "loss": 0.7537,
+      "step": 4067
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.9631226229157055,
+      "learning_rate": 7.78678777336627e-06,
+      "loss": 1.0203,
+      "step": 4068
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.6928746766462877,
+      "learning_rate": 7.785688578187534e-06,
+      "loss": 0.6483,
+      "step": 4069
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.1103911778195075,
+      "learning_rate": 7.784589187744151e-06,
+      "loss": 0.2568,
+      "step": 4070
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.080635885354421,
+      "learning_rate": 7.783489602113185e-06,
+      "loss": 0.7317,
+      "step": 4071
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.883177504009893,
+      "learning_rate": 7.782389821371712e-06,
+      "loss": 0.5972,
+      "step": 4072
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.327020316159079,
+      "learning_rate": 7.781289845596821e-06,
+      "loss": 0.9701,
+      "step": 4073
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.7278052835176005,
+      "learning_rate": 7.780189674865617e-06,
+      "loss": 0.5983,
+      "step": 4074
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.160715692148022,
+      "learning_rate": 7.779089309255217e-06,
+      "loss": 0.7912,
+      "step": 4075
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.248414112566387,
+      "learning_rate": 7.777988748842748e-06,
+      "loss": 1.1924,
+      "step": 4076
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.836771315743967,
+      "learning_rate": 7.77688799370536e-06,
+      "loss": 1.2628,
+      "step": 4077
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.348685058036905,
+      "learning_rate": 7.775787043920209e-06,
+      "loss": 0.9582,
+      "step": 4078
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.85058506058486,
+      "learning_rate": 7.774685899564465e-06,
+      "loss": 0.8381,
+      "step": 4079
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.9380355388535526,
+      "learning_rate": 7.773584560715315e-06,
+      "loss": 0.8012,
+      "step": 4080
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.2877638211599227,
+      "learning_rate": 7.772483027449957e-06,
+      "loss": 0.7378,
+      "step": 4081
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.58337541740198,
+      "learning_rate": 7.771381299845607e-06,
+      "loss": 0.7549,
+      "step": 4082
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.431569297930181,
+      "learning_rate": 7.770279377979488e-06,
+      "loss": 0.349,
+      "step": 4083
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.677928014296555,
+      "learning_rate": 7.76917726192884e-06,
+      "loss": 1.2377,
+      "step": 4084
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.046569984878376,
+      "learning_rate": 7.768074951770919e-06,
+      "loss": 1.0108,
+      "step": 4085
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.008142292994773,
+      "learning_rate": 7.76697244758299e-06,
+      "loss": 0.5098,
+      "step": 4086
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.4447675454103566,
+      "learning_rate": 7.765869749442334e-06,
+      "loss": 0.7362,
+      "step": 4087
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.213593544623172,
+      "learning_rate": 7.764766857426247e-06,
+      "loss": 1.1057,
+      "step": 4088
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.7927402365139145,
+      "learning_rate": 7.763663771612033e-06,
+      "loss": 0.7684,
+      "step": 4089
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.515848562656643,
+      "learning_rate": 7.762560492077018e-06,
+      "loss": 0.8077,
+      "step": 4090
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.687411542813131,
+      "learning_rate": 7.761457018898536e-06,
+      "loss": 0.7698,
+      "step": 4091
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.173316022459569,
+      "learning_rate": 7.760353352153933e-06,
+      "loss": 0.8834,
+      "step": 4092
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.5145776395485453,
+      "learning_rate": 7.759249491920573e-06,
+      "loss": 0.7676,
+      "step": 4093
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.379993757189861,
+      "learning_rate": 7.758145438275834e-06,
+      "loss": 0.9946,
+      "step": 4094
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.071372107617467,
+      "learning_rate": 7.757041191297102e-06,
+      "loss": 0.9479,
+      "step": 4095
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.354744362202982,
+      "learning_rate": 7.755936751061781e-06,
+      "loss": 0.6599,
+      "step": 4096
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.615773819555657,
+      "learning_rate": 7.75483211764729e-06,
+      "loss": 0.7168,
+      "step": 4097
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.520254715236886,
+      "learning_rate": 7.753727291131052e-06,
+      "loss": 0.8517,
+      "step": 4098
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.185083001280689,
+      "learning_rate": 7.75262227159052e-06,
+      "loss": 0.9201,
+      "step": 4099
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.808745561079665,
+      "learning_rate": 7.751517059103142e-06,
+      "loss": 0.9437,
+      "step": 4100
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7571468364421476,
+      "learning_rate": 7.750411653746395e-06,
+      "loss": 0.4474,
+      "step": 4101
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.549902143678165,
+      "learning_rate": 7.74930605559776e-06,
+      "loss": 0.8425,
+      "step": 4102
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.1746358152158685,
+      "learning_rate": 7.748200264734735e-06,
+      "loss": 1.2278,
+      "step": 4103
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.987321493415824,
+      "learning_rate": 7.747094281234833e-06,
+      "loss": 0.4285,
+      "step": 4104
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.0173645566645195,
+      "learning_rate": 7.745988105175577e-06,
+      "loss": 0.4033,
+      "step": 4105
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.354547097155099,
+      "learning_rate": 7.744881736634506e-06,
+      "loss": 1.3625,
+      "step": 4106
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.332371613972039,
+      "learning_rate": 7.74377517568917e-06,
+      "loss": 0.7414,
+      "step": 4107
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.1920844245645,
+      "learning_rate": 7.742668422417137e-06,
+      "loss": 0.3366,
+      "step": 4108
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.4605536417478153,
+      "learning_rate": 7.74156147689598e-06,
+      "loss": 0.7064,
+      "step": 4109
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.4351461374391827,
+      "learning_rate": 7.740454339203298e-06,
+      "loss": 0.6305,
+      "step": 4110
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.341168273700383,
+      "learning_rate": 7.739347009416693e-06,
+      "loss": 0.9239,
+      "step": 4111
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.1896550569997317,
+      "learning_rate": 7.738239487613783e-06,
+      "loss": 0.4106,
+      "step": 4112
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.772165868649677,
+      "learning_rate": 7.737131773872202e-06,
+      "loss": 0.7461,
+      "step": 4113
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.172701606361016,
+      "learning_rate": 7.736023868269598e-06,
+      "loss": 0.572,
+      "step": 4114
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.2955721559098285,
+      "learning_rate": 7.734915770883624e-06,
+      "loss": 0.6784,
+      "step": 4115
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 5.013100997652645,
+      "learning_rate": 7.733807481791959e-06,
+      "loss": 1.4583,
+      "step": 4116
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.8655354170040597,
+      "learning_rate": 7.732699001072288e-06,
+      "loss": 1.1896,
+      "step": 4117
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9395350071445177,
+      "learning_rate": 7.73159032880231e-06,
+      "loss": 0.4108,
+      "step": 4118
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.7818826869175437,
+      "learning_rate": 7.730481465059736e-06,
+      "loss": 0.7055,
+      "step": 4119
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 5.15552039543157,
+      "learning_rate": 7.729372409922295e-06,
+      "loss": 1.4362,
+      "step": 4120
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.2196597029324914,
+      "learning_rate": 7.728263163467727e-06,
+      "loss": 0.6513,
+      "step": 4121
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.255860561163699,
+      "learning_rate": 7.727153725773783e-06,
+      "loss": 0.8975,
+      "step": 4122
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.043423662983045,
+      "learning_rate": 7.726044096918232e-06,
+      "loss": 0.6755,
+      "step": 4123
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.680203207964453,
+      "learning_rate": 7.724934276978855e-06,
+      "loss": 1.0033,
+      "step": 4124
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.746901777325029,
+      "learning_rate": 7.723824266033444e-06,
+      "loss": 0.5562,
+      "step": 4125
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.272360035766652,
+      "learning_rate": 7.722714064159805e-06,
+      "loss": 0.3457,
+      "step": 4126
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.7240643115152325,
+      "learning_rate": 7.72160367143576e-06,
+      "loss": 0.7799,
+      "step": 4127
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.591078813080817,
+      "learning_rate": 7.720493087939143e-06,
+      "loss": 1.2883,
+      "step": 4128
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.210409977210689,
+      "learning_rate": 7.719382313747799e-06,
+      "loss": 0.7096,
+      "step": 4129
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.6817128508281134,
+      "learning_rate": 7.71827134893959e-06,
+      "loss": 0.6766,
+      "step": 4130
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7634823470854484,
+      "learning_rate": 7.717160193592387e-06,
+      "loss": 0.4359,
+      "step": 4131
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.9420876984077293,
+      "learning_rate": 7.716048847784082e-06,
+      "loss": 0.7967,
+      "step": 4132
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.148358437503598,
+      "learning_rate": 7.714937311592573e-06,
+      "loss": 0.1711,
+      "step": 4133
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.250724443509678,
+      "learning_rate": 7.713825585095775e-06,
+      "loss": 1.2042,
+      "step": 4134
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.758507214850353,
+      "learning_rate": 7.712713668371612e-06,
+      "loss": 1.0143,
+      "step": 4135
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.7488317621633676,
+      "learning_rate": 7.711601561498027e-06,
+      "loss": 0.5658,
+      "step": 4136
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.397262375524663,
+      "learning_rate": 7.710489264552974e-06,
+      "loss": 0.8973,
+      "step": 4137
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.83739358311308,
+      "learning_rate": 7.709376777614418e-06,
+      "loss": 0.5373,
+      "step": 4138
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.5305706042930973,
+      "learning_rate": 7.708264100760343e-06,
+      "loss": 0.4427,
+      "step": 4139
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.6652675405330037,
+      "learning_rate": 7.707151234068741e-06,
+      "loss": 0.735,
+      "step": 4140
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.988081527589675,
+      "learning_rate": 7.706038177617618e-06,
+      "loss": 0.7667,
+      "step": 4141
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.477405865454765,
+      "learning_rate": 7.704924931484997e-06,
+      "loss": 0.613,
+      "step": 4142
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.063093742168863,
+      "learning_rate": 7.703811495748908e-06,
+      "loss": 1.087,
+      "step": 4143
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.503004822250554,
+      "learning_rate": 7.702697870487401e-06,
+      "loss": 0.4957,
+      "step": 4144
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 5.030254929001747,
+      "learning_rate": 7.701584055778536e-06,
+      "loss": 1.126,
+      "step": 4145
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.257522661426769,
+      "learning_rate": 7.700470051700385e-06,
+      "loss": 0.8968,
+      "step": 4146
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.9125424198868415,
+      "learning_rate": 7.699355858331038e-06,
+      "loss": 0.5521,
+      "step": 4147
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.844248185026583,
+      "learning_rate": 7.698241475748592e-06,
+      "loss": 0.6196,
+      "step": 4148
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.9157845125845605,
+      "learning_rate": 7.69712690403116e-06,
+      "loss": 0.7622,
+      "step": 4149
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.053419733377103,
+      "learning_rate": 7.696012143256873e-06,
+      "loss": 0.1924,
+      "step": 4150
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.6960704618009794,
+      "learning_rate": 7.694897193503865e-06,
+      "loss": 1.0025,
+      "step": 4151
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.4272805195111458,
+      "learning_rate": 7.693782054850293e-06,
+      "loss": 0.8893,
+      "step": 4152
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.636670957115418,
+      "learning_rate": 7.692666727374321e-06,
+      "loss": 0.7311,
+      "step": 4153
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.4288545667039547,
+      "learning_rate": 7.691551211154132e-06,
+      "loss": 0.7569,
+      "step": 4154
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.996161453088842,
+      "learning_rate": 7.690435506267916e-06,
+      "loss": 1.0075,
+      "step": 4155
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.473680046059575,
+      "learning_rate": 7.689319612793878e-06,
+      "loss": 0.964,
+      "step": 4156
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 5.043930889076822,
+      "learning_rate": 7.68820353081024e-06,
+      "loss": 1.343,
+      "step": 4157
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.8010199690411755,
+      "learning_rate": 7.687087260395237e-06,
+      "loss": 0.962,
+      "step": 4158
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.846913545191014,
+      "learning_rate": 7.685970801627108e-06,
+      "loss": 0.9527,
+      "step": 4159
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.7844959560021967,
+      "learning_rate": 7.684854154584117e-06,
+      "loss": 0.5735,
+      "step": 4160
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.320667753043801,
+      "learning_rate": 7.68373731934453e-06,
+      "loss": 1.2572,
+      "step": 4161
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.6840322336012488,
+      "learning_rate": 7.682620295986642e-06,
+      "loss": 0.7284,
+      "step": 4162
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.9959981512529237,
+      "learning_rate": 7.681503084588743e-06,
+      "loss": 0.673,
+      "step": 4163
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.1346317197548474,
+      "learning_rate": 7.680385685229148e-06,
+      "loss": 0.389,
+      "step": 4164
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 5.405414090668094,
+      "learning_rate": 7.679268097986183e-06,
+      "loss": 1.4987,
+      "step": 4165
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.157420352861571,
+      "learning_rate": 7.678150322938183e-06,
+      "loss": 0.5769,
+      "step": 4166
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.3666285293683225,
+      "learning_rate": 7.6770323601635e-06,
+      "loss": 0.6687,
+      "step": 4167
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.277406369736445,
+      "learning_rate": 7.675914209740503e-06,
+      "loss": 0.7282,
+      "step": 4168
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9396861983972804,
+      "learning_rate": 7.67479587174756e-06,
+      "loss": 0.2985,
+      "step": 4169
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.241210008332523,
+      "learning_rate": 7.67367734626307e-06,
+      "loss": 1.2679,
+      "step": 4170
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.780035706850577,
+      "learning_rate": 7.672558633365434e-06,
+      "loss": 0.8957,
+      "step": 4171
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.875053635380888,
+      "learning_rate": 7.67143973313307e-06,
+      "loss": 1.4268,
+      "step": 4172
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.7887924259311805,
+      "learning_rate": 7.670320645644404e-06,
+      "loss": 0.6755,
+      "step": 4173
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.04293308169156,
+      "learning_rate": 7.669201370977885e-06,
+      "loss": 0.9438,
+      "step": 4174
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.2374797347176925,
+      "learning_rate": 7.668081909211964e-06,
+      "loss": 0.8673,
+      "step": 4175
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.8677793397185134,
+      "learning_rate": 7.666962260425113e-06,
+      "loss": 0.903,
+      "step": 4176
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.348804643728993,
+      "learning_rate": 7.665842424695815e-06,
+      "loss": 0.9382,
+      "step": 4177
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.9757059499309753,
+      "learning_rate": 7.664722402102564e-06,
+      "loss": 0.5758,
+      "step": 4178
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.5385242534408654,
+      "learning_rate": 7.663602192723871e-06,
+      "loss": 0.6308,
+      "step": 4179
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.389637679225512,
+      "learning_rate": 7.662481796638254e-06,
+      "loss": 1.2565,
+      "step": 4180
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.035115559574408,
+      "learning_rate": 7.661361213924252e-06,
+      "loss": 0.8521,
+      "step": 4181
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.824584879682278,
+      "learning_rate": 7.660240444660411e-06,
+      "loss": 0.8049,
+      "step": 4182
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.586953981275565,
+      "learning_rate": 7.659119488925292e-06,
+      "loss": 1.082,
+      "step": 4183
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.9302524382227393,
+      "learning_rate": 7.65799834679747e-06,
+      "loss": 0.7177,
+      "step": 4184
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.019542010284757,
+      "learning_rate": 7.656877018355533e-06,
+      "loss": 0.9594,
+      "step": 4185
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.68219286689825,
+      "learning_rate": 7.655755503678075e-06,
+      "loss": 0.4558,
+      "step": 4186
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.0411454882488655,
+      "learning_rate": 7.654633802843718e-06,
+      "loss": 0.9238,
+      "step": 4187
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.1291071195040256,
+      "learning_rate": 7.653511915931082e-06,
+      "loss": 0.485,
+      "step": 4188
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.2706829407485616,
+      "learning_rate": 7.65238984301881e-06,
+      "loss": 0.9747,
+      "step": 4189
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.027229745479082,
+      "learning_rate": 7.651267584185554e-06,
+      "loss": 1.2054,
+      "step": 4190
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.778073601555418,
+      "learning_rate": 7.650145139509978e-06,
+      "loss": 0.7593,
+      "step": 4191
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.897854125935847,
+      "learning_rate": 7.649022509070761e-06,
+      "loss": 0.6984,
+      "step": 4192
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.233075336679899,
+      "learning_rate": 7.647899692946594e-06,
+      "loss": 0.4863,
+      "step": 4193
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.915603508098223,
+      "learning_rate": 7.646776691216185e-06,
+      "loss": 0.681,
+      "step": 4194
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.258355781356614,
+      "learning_rate": 7.645653503958246e-06,
+      "loss": 0.7473,
+      "step": 4195
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7927823875566085,
+      "learning_rate": 7.64453013125151e-06,
+      "loss": 0.2872,
+      "step": 4196
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.508724132851226,
+      "learning_rate": 7.643406573174724e-06,
+      "loss": 0.7924,
+      "step": 4197
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 5.050262457545431,
+      "learning_rate": 7.642282829806639e-06,
+      "loss": 1.0736,
+      "step": 4198
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.029013736020635,
+      "learning_rate": 7.641158901226026e-06,
+      "loss": 0.8475,
+      "step": 4199
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.696466610682216,
+      "learning_rate": 7.64003478751167e-06,
+      "loss": 0.6701,
+      "step": 4200
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.143479928983292,
+      "learning_rate": 7.638910488742364e-06,
+      "loss": 0.729,
+      "step": 4201
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.727763069861252,
+      "learning_rate": 7.637786004996918e-06,
+      "loss": 1.209,
+      "step": 4202
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.851506905628155,
+      "learning_rate": 7.636661336354152e-06,
+      "loss": 1.0456,
+      "step": 4203
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.9358162237417544,
+      "learning_rate": 7.635536482892902e-06,
+      "loss": 0.6617,
+      "step": 4204
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 5.672038616038138,
+      "learning_rate": 7.634411444692014e-06,
+      "loss": 1.3754,
+      "step": 4205
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.3060864557234773,
+      "learning_rate": 7.63328622183035e-06,
+      "loss": 0.5077,
+      "step": 4206
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 5.1549988023402165,
+      "learning_rate": 7.63216081438678e-06,
+      "loss": 1.3093,
+      "step": 4207
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.4609699622097643,
+      "learning_rate": 7.631035222440192e-06,
+      "loss": 0.3794,
+      "step": 4208
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.602474189020254,
+      "learning_rate": 7.629909446069487e-06,
+      "loss": 0.7048,
+      "step": 4209
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.785531242848229,
+      "learning_rate": 7.628783485353573e-06,
+      "loss": 0.8278,
+      "step": 4210
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.5722162724562665,
+      "learning_rate": 7.6276573403713796e-06,
+      "loss": 0.8092,
+      "step": 4211
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.458727519165083,
+      "learning_rate": 7.626531011201841e-06,
+      "loss": 0.9704,
+      "step": 4212
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.8927187042148437,
+      "learning_rate": 7.625404497923909e-06,
+      "loss": 1.0258,
+      "step": 4213
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.848434988336024,
+      "learning_rate": 7.6242778006165484e-06,
+      "loss": 0.7578,
+      "step": 4214
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.755430091464229,
+      "learning_rate": 7.623150919358734e-06,
+      "loss": 0.4261,
+      "step": 4215
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.131447734058058,
+      "learning_rate": 7.6220238542294565e-06,
+      "loss": 1.1711,
+      "step": 4216
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.4337178438962437,
+      "learning_rate": 7.620896605307717e-06,
+      "loss": 0.6572,
+      "step": 4217
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.2111164733993345,
+      "learning_rate": 7.619769172672533e-06,
+      "loss": 0.9089,
+      "step": 4218
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 5.3934535906184164,
+      "learning_rate": 7.61864155640293e-06,
+      "loss": 1.6321,
+      "step": 4219
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.380431114335264,
+      "learning_rate": 7.6175137565779524e-06,
+      "loss": 0.9619,
+      "step": 4220
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.5949226198374613,
+      "learning_rate": 7.616385773276651e-06,
+      "loss": 0.707,
+      "step": 4221
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.375986072874407,
+      "learning_rate": 7.615257606578093e-06,
+      "loss": 0.9955,
+      "step": 4222
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.0945232970753622,
+      "learning_rate": 7.6141292565613574e-06,
+      "loss": 0.6822,
+      "step": 4223
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.121369107003071,
+      "learning_rate": 7.613000723305539e-06,
+      "loss": 0.7806,
+      "step": 4224
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.842882239042579,
+      "learning_rate": 7.611872006889741e-06,
+      "loss": 1.1877,
+      "step": 4225
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.235155562415778,
+      "learning_rate": 7.610743107393083e-06,
+      "loss": 0.68,
+      "step": 4226
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.197630275967312,
+      "learning_rate": 7.609614024894694e-06,
+      "loss": 0.9361,
+      "step": 4227
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8510247652966472,
+      "learning_rate": 7.60848475947372e-06,
+      "loss": 0.3807,
+      "step": 4228
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3552331105985607,
+      "learning_rate": 7.607355311209317e-06,
+      "loss": 0.1851,
+      "step": 4229
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.2471461785944435,
+      "learning_rate": 7.606225680180652e-06,
+      "loss": 0.3754,
+      "step": 4230
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.5961883792714393,
+      "learning_rate": 7.605095866466912e-06,
+      "loss": 0.6708,
+      "step": 4231
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.8820743215669724,
+      "learning_rate": 7.603965870147285e-06,
+      "loss": 0.8477,
+      "step": 4232
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.420318348105564,
+      "learning_rate": 7.602835691300986e-06,
+      "loss": 0.2926,
+      "step": 4233
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.7725992633943646,
+      "learning_rate": 7.601705330007232e-06,
+      "loss": 0.8095,
+      "step": 4234
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.561808281233337,
+      "learning_rate": 7.600574786345257e-06,
+      "loss": 0.5723,
+      "step": 4235
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.3420721342527875,
+      "learning_rate": 7.599444060394308e-06,
+      "loss": 1.0933,
+      "step": 4236
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.303114674088338,
+      "learning_rate": 7.598313152233643e-06,
+      "loss": 0.5405,
+      "step": 4237
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.639902615736915,
+      "learning_rate": 7.597182061942533e-06,
+      "loss": 1.2133,
+      "step": 4238
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.119971811972484,
+      "learning_rate": 7.5960507896002636e-06,
+      "loss": 1.0088,
+      "step": 4239
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.9814954955260253,
+      "learning_rate": 7.594919335286133e-06,
+      "loss": 0.8402,
+      "step": 4240
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.235203233182246,
+      "learning_rate": 7.593787699079449e-06,
+      "loss": 0.7345,
+      "step": 4241
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.8363860198831965,
+      "learning_rate": 7.592655881059536e-06,
+      "loss": 0.6156,
+      "step": 4242
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.137444392466568,
+      "learning_rate": 7.591523881305728e-06,
+      "loss": 1.2916,
+      "step": 4243
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.376223279509348,
+      "learning_rate": 7.5903916998973745e-06,
+      "loss": 0.7987,
+      "step": 4244
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.541874168567518,
+      "learning_rate": 7.589259336913839e-06,
+      "loss": 0.552,
+      "step": 4245
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.0811400669956592,
+      "learning_rate": 7.588126792434489e-06,
+      "loss": 0.777,
+      "step": 4246
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.6244134270135215,
+      "learning_rate": 7.586994066538715e-06,
+      "loss": 1.0474,
+      "step": 4247
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.294865922920346,
+      "learning_rate": 7.585861159305917e-06,
+      "loss": 0.6377,
+      "step": 4248
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.027424272181276,
+      "learning_rate": 7.584728070815504e-06,
+      "loss": 1.2547,
+      "step": 4249
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.073784678794884,
+      "learning_rate": 7.583594801146903e-06,
+      "loss": 0.9806,
+      "step": 4250
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.7825420370232674,
+      "learning_rate": 7.582461350379551e-06,
+      "loss": 0.8304,
+      "step": 4251
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.0319950587374103,
+      "learning_rate": 7.581327718592896e-06,
+      "loss": 0.778,
+      "step": 4252
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0521053776449354,
+      "learning_rate": 7.580193905866402e-06,
+      "loss": 0.415,
+      "step": 4253
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.885777423711224,
+      "learning_rate": 7.579059912279545e-06,
+      "loss": 0.78,
+      "step": 4254
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.715809617328955,
+      "learning_rate": 7.577925737911811e-06,
+      "loss": 1.4035,
+      "step": 4255
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.577394588141058,
+      "learning_rate": 7.576791382842702e-06,
+      "loss": 0.5588,
+      "step": 4256
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.162114084015856,
+      "learning_rate": 7.575656847151732e-06,
+      "loss": 0.7035,
+      "step": 4257
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.5025398505743817,
+      "learning_rate": 7.574522130918426e-06,
+      "loss": 0.9669,
+      "step": 4258
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.689264485842858,
+      "learning_rate": 7.5733872342223235e-06,
+      "loss": 0.985,
+      "step": 4259
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.1512518368625324,
+      "learning_rate": 7.572252157142976e-06,
+      "loss": 0.5427,
+      "step": 4260
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.136962873078472,
+      "learning_rate": 7.571116899759945e-06,
+      "loss": 0.5495,
+      "step": 4261
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.657164795694336,
+      "learning_rate": 7.56998146215281e-06,
+      "loss": 0.7432,
+      "step": 4262
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.5593972705527133,
+      "learning_rate": 7.568845844401158e-06,
+      "loss": 0.9293,
+      "step": 4263
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.104600859719557,
+      "learning_rate": 7.567710046584593e-06,
+      "loss": 0.937,
+      "step": 4264
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.3107309830162897,
+      "learning_rate": 7.566574068782728e-06,
+      "loss": 0.8032,
+      "step": 4265
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.597493941109041,
+      "learning_rate": 7.565437911075191e-06,
+      "loss": 1.1043,
+      "step": 4266
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.8071170485716794,
+      "learning_rate": 7.564301573541621e-06,
+      "loss": 0.8017,
+      "step": 4267
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9680658468583032,
+      "learning_rate": 7.563165056261671e-06,
+      "loss": 0.4377,
+      "step": 4268
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.4083963853423915,
+      "learning_rate": 7.562028359315005e-06,
+      "loss": 0.9813,
+      "step": 4269
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.065016615189242,
+      "learning_rate": 7.560891482781301e-06,
+      "loss": 0.3513,
+      "step": 4270
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.22180599717921,
+      "learning_rate": 7.559754426740249e-06,
+      "loss": 0.9598,
+      "step": 4271
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.177975680038039,
+      "learning_rate": 7.558617191271551e-06,
+      "loss": 1.0541,
+      "step": 4272
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.206282755330498,
+      "learning_rate": 7.557479776454923e-06,
+      "loss": 0.9705,
+      "step": 4273
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.3772672282559864,
+      "learning_rate": 7.556342182370095e-06,
+      "loss": 0.7269,
+      "step": 4274
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.815793144389638,
+      "learning_rate": 7.5552044090968035e-06,
+      "loss": 1.1305,
+      "step": 4275
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.397789333338332,
+      "learning_rate": 7.554066456714804e-06,
+      "loss": 0.7765,
+      "step": 4276
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.169891416483646,
+      "learning_rate": 7.552928325303861e-06,
+      "loss": 0.3282,
+      "step": 4277
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.905886108680131,
+      "learning_rate": 7.551790014943752e-06,
+      "loss": 0.6804,
+      "step": 4278
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.8025974684265913,
+      "learning_rate": 7.55065152571427e-06,
+      "loss": 0.7483,
+      "step": 4279
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.7342712449571205,
+      "learning_rate": 7.549512857695216e-06,
+      "loss": 0.7249,
+      "step": 4280
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.2459565813784925,
+      "learning_rate": 7.5483740109664036e-06,
+      "loss": 0.5911,
+      "step": 4281
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.5258338374757656,
+      "learning_rate": 7.547234985607668e-06,
+      "loss": 0.5009,
+      "step": 4282
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.486670744801288,
+      "learning_rate": 7.546095781698842e-06,
+      "loss": 1.0581,
+      "step": 4283
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.9298991349988945,
+      "learning_rate": 7.544956399319785e-06,
+      "loss": 1.068,
+      "step": 4284
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.172582064142233,
+      "learning_rate": 7.543816838550359e-06,
+      "loss": 1.1742,
+      "step": 4285
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.201667663325446,
+      "learning_rate": 7.542677099470441e-06,
+      "loss": 0.8472,
+      "step": 4286
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.1995176189199728,
+      "learning_rate": 7.541537182159926e-06,
+      "loss": 0.7892,
+      "step": 4287
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.328526607337528,
+      "learning_rate": 7.540397086698716e-06,
+      "loss": 0.7107,
+      "step": 4288
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.353118807139896,
+      "learning_rate": 7.5392568131667265e-06,
+      "loss": 0.8643,
+      "step": 4289
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.1174315933767485,
+      "learning_rate": 7.538116361643883e-06,
+      "loss": 0.5618,
+      "step": 4290
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 7.299874282257006,
+      "learning_rate": 7.53697573221013e-06,
+      "loss": 1.828,
+      "step": 4291
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.045071910752177,
+      "learning_rate": 7.53583492494542e-06,
+      "loss": 0.3556,
+      "step": 4292
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.386906087901318,
+      "learning_rate": 7.534693939929716e-06,
+      "loss": 0.9769,
+      "step": 4293
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.1939267691785265,
+      "learning_rate": 7.533552777243e-06,
+      "loss": 1.0888,
+      "step": 4294
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.7922405769680005,
+      "learning_rate": 7.532411436965258e-06,
+      "loss": 0.7094,
+      "step": 4295
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.0949378799917553,
+      "learning_rate": 7.531269919176496e-06,
+      "loss": 0.7065,
+      "step": 4296
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.3404622426864967,
+      "learning_rate": 7.530128223956729e-06,
+      "loss": 0.5209,
+      "step": 4297
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.6828821634625255,
+      "learning_rate": 7.528986351385985e-06,
+      "loss": 1.0418,
+      "step": 4298
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.3181216173240107,
+      "learning_rate": 7.527844301544304e-06,
+      "loss": 0.5293,
+      "step": 4299
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.96720934548505,
+      "learning_rate": 7.526702074511738e-06,
+      "loss": 1.2176,
+      "step": 4300
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.8152740806384444,
+      "learning_rate": 7.5255596703683535e-06,
+      "loss": 0.6621,
+      "step": 4301
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.3656367937880436,
+      "learning_rate": 7.524417089194227e-06,
+      "loss": 0.77,
+      "step": 4302
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0161219249370117,
+      "learning_rate": 7.523274331069449e-06,
+      "loss": 0.1678,
+      "step": 4303
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.962457911789098,
+      "learning_rate": 7.522131396074122e-06,
+      "loss": 1.2677,
+      "step": 4304
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.7386987611806384,
+      "learning_rate": 7.52098828428836e-06,
+      "loss": 0.7596,
+      "step": 4305
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.930736277089549,
+      "learning_rate": 7.5198449957922895e-06,
+      "loss": 0.955,
+      "step": 4306
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.174764922991015,
+      "learning_rate": 7.5187015306660524e-06,
+      "loss": 0.7774,
+      "step": 4307
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.139106980850027,
+      "learning_rate": 7.5175578889898016e-06,
+      "loss": 0.7638,
+      "step": 4308
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.277844030429609,
+      "learning_rate": 7.516414070843696e-06,
+      "loss": 0.9127,
+      "step": 4309
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.377312607254998,
+      "learning_rate": 7.515270076307917e-06,
+      "loss": 0.88,
+      "step": 4310
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.753351823698591,
+      "learning_rate": 7.514125905462651e-06,
+      "loss": 1.3045,
+      "step": 4311
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.380965467013008,
+      "learning_rate": 7.512981558388101e-06,
+      "loss": 0.7882,
+      "step": 4312
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.472155010468368,
+      "learning_rate": 7.51183703516448e-06,
+      "loss": 0.7921,
+      "step": 4313
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.192913037162404,
+      "learning_rate": 7.510692335872016e-06,
+      "loss": 1.0474,
+      "step": 4314
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.2134565473107704,
+      "learning_rate": 7.509547460590945e-06,
+      "loss": 0.5618,
+      "step": 4315
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.918952899825416,
+      "learning_rate": 7.508402409401519e-06,
+      "loss": 1.2301,
+      "step": 4316
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.286764533220852,
+      "learning_rate": 7.507257182384e-06,
+      "loss": 0.6964,
+      "step": 4317
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.4183487003326625,
+      "learning_rate": 7.506111779618663e-06,
+      "loss": 0.8308,
+      "step": 4318
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.622961051435708,
+      "learning_rate": 7.504966201185798e-06,
+      "loss": 0.9083,
+      "step": 4319
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.51977662181751,
+      "learning_rate": 7.503820447165705e-06,
+      "loss": 0.6034,
+      "step": 4320
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.032466768649381,
+      "learning_rate": 7.502674517638694e-06,
+      "loss": 0.7514,
+      "step": 4321
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.2389907475732223,
+      "learning_rate": 7.5015284126850915e-06,
+      "loss": 0.6963,
+      "step": 4322
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.987600811240381,
+      "learning_rate": 7.500382132385234e-06,
+      "loss": 1.0207,
+      "step": 4323
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.066603624193605,
+      "learning_rate": 7.499235676819471e-06,
+      "loss": 0.8864,
+      "step": 4324
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.769839249998782,
+      "learning_rate": 7.498089046068163e-06,
+      "loss": 0.5207,
+      "step": 4325
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.4436628569051204,
+      "learning_rate": 7.496942240211685e-06,
+      "loss": 1.22,
+      "step": 4326
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0388291037490998,
+      "learning_rate": 7.495795259330422e-06,
+      "loss": 0.6388,
+      "step": 4327
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.42893247686087,
+      "learning_rate": 7.494648103504774e-06,
+      "loss": 0.8951,
+      "step": 4328
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.7612822548289877,
+      "learning_rate": 7.49350077281515e-06,
+      "loss": 0.5806,
+      "step": 4329
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.706151510063128,
+      "learning_rate": 7.492353267341974e-06,
+      "loss": 0.8698,
+      "step": 4330
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.6983960971622984,
+      "learning_rate": 7.491205587165682e-06,
+      "loss": 1.1328,
+      "step": 4331
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.319249197795334,
+      "learning_rate": 7.490057732366718e-06,
+      "loss": 0.9426,
+      "step": 4332
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.5025170873982514,
+      "learning_rate": 7.4889097030255445e-06,
+      "loss": 0.5951,
+      "step": 4333
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.4219007121333407,
+      "learning_rate": 7.487761499222632e-06,
+      "loss": 0.339,
+      "step": 4334
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.367668946465433,
+      "learning_rate": 7.486613121038466e-06,
+      "loss": 1.0931,
+      "step": 4335
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.9355104713653724,
+      "learning_rate": 7.485464568553541e-06,
+      "loss": 0.5703,
+      "step": 4336
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.17807422897983,
+      "learning_rate": 7.484315841848368e-06,
+      "loss": 0.9853,
+      "step": 4337
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.043315819974853,
+      "learning_rate": 7.483166941003466e-06,
+      "loss": 1.1789,
+      "step": 4338
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.530950037640052,
+      "learning_rate": 7.482017866099367e-06,
+      "loss": 0.8051,
+      "step": 4339
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.8380019274677557,
+      "learning_rate": 7.480868617216619e-06,
+      "loss": 0.3348,
+      "step": 4340
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.6038676133727834,
+      "learning_rate": 7.479719194435776e-06,
+      "loss": 0.8044,
+      "step": 4341
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.0925376126716975,
+      "learning_rate": 7.478569597837411e-06,
+      "loss": 0.9227,
+      "step": 4342
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.696213275373554,
+      "learning_rate": 7.4774198275021014e-06,
+      "loss": 0.8679,
+      "step": 4343
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.7817964848829764,
+      "learning_rate": 7.476269883510445e-06,
+      "loss": 0.935,
+      "step": 4344
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.406957888921831,
+      "learning_rate": 7.475119765943049e-06,
+      "loss": 0.9647,
+      "step": 4345
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.5175991949471355,
+      "learning_rate": 7.473969474880527e-06,
+      "loss": 1.0671,
+      "step": 4346
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.442572440305692,
+      "learning_rate": 7.47281901040351e-06,
+      "loss": 1.2769,
+      "step": 4347
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.039003798061688,
+      "learning_rate": 7.471668372592644e-06,
+      "loss": 0.7816,
+      "step": 4348
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.356063338687145,
+      "learning_rate": 7.470517561528582e-06,
+      "loss": 0.8141,
+      "step": 4349
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.627272893200747,
+      "learning_rate": 7.46936657729199e-06,
+      "loss": 0.8789,
+      "step": 4350
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.165112986350278,
+      "learning_rate": 7.4682154199635475e-06,
+      "loss": 0.9576,
+      "step": 4351
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.786959812117681,
+      "learning_rate": 7.467064089623945e-06,
+      "loss": 0.676,
+      "step": 4352
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.344544708423713,
+      "learning_rate": 7.465912586353888e-06,
+      "loss": 1.0466,
+      "step": 4353
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.141295381304714,
+      "learning_rate": 7.464760910234091e-06,
+      "loss": 1.1653,
+      "step": 4354
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.106223819266664,
+      "learning_rate": 7.463609061345279e-06,
+      "loss": 1.2457,
+      "step": 4355
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.190753808346917,
+      "learning_rate": 7.462457039768194e-06,
+      "loss": 1.1127,
+      "step": 4356
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.536070663602797,
+      "learning_rate": 7.461304845583588e-06,
+      "loss": 0.5238,
+      "step": 4357
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.067446549558834,
+      "learning_rate": 7.460152478872224e-06,
+      "loss": 0.7506,
+      "step": 4358
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.767927207227846,
+      "learning_rate": 7.458999939714876e-06,
+      "loss": 0.849,
+      "step": 4359
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.1358705863995686,
+      "learning_rate": 7.4578472281923356e-06,
+      "loss": 0.3897,
+      "step": 4360
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.149045683865217,
+      "learning_rate": 7.456694344385401e-06,
+      "loss": 1.053,
+      "step": 4361
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.297878415611068,
+      "learning_rate": 7.455541288374885e-06,
+      "loss": 0.5006,
+      "step": 4362
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.493178783840786,
+      "learning_rate": 7.45438806024161e-06,
+      "loss": 0.8101,
+      "step": 4363
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.92095481738485,
+      "learning_rate": 7.453234660066413e-06,
+      "loss": 0.9906,
+      "step": 4364
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.954136088552313,
+      "learning_rate": 7.452081087930143e-06,
+      "loss": 0.81,
+      "step": 4365
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.68244091356062,
+      "learning_rate": 7.450927343913661e-06,
+      "loss": 0.541,
+      "step": 4366
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.627231427015078,
+      "learning_rate": 7.449773428097838e-06,
+      "loss": 0.9492,
+      "step": 4367
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.8169653497783305,
+      "learning_rate": 7.44861934056356e-06,
+      "loss": 0.6861,
+      "step": 4368
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.8882242598826835,
+      "learning_rate": 7.447465081391722e-06,
+      "loss": 1.1398,
+      "step": 4369
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.067463057219432,
+      "learning_rate": 7.446310650663234e-06,
+      "loss": 0.7872,
+      "step": 4370
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0416188880316226,
+      "learning_rate": 7.445156048459016e-06,
+      "loss": 0.4258,
+      "step": 4371
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.8461591247025497,
+      "learning_rate": 7.444001274859999e-06,
+      "loss": 1.0479,
+      "step": 4372
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.6675086056877957,
+      "learning_rate": 7.4428463299471285e-06,
+      "loss": 0.4093,
+      "step": 4373
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.8086978839082044,
+      "learning_rate": 7.441691213801363e-06,
+      "loss": 0.7284,
+      "step": 4374
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.849463434709465,
+      "learning_rate": 7.440535926503669e-06,
+      "loss": 0.4524,
+      "step": 4375
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.878760788811572,
+      "learning_rate": 7.439380468135029e-06,
+      "loss": 0.7494,
+      "step": 4376
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.46426717186742,
+      "learning_rate": 7.4382248387764335e-06,
+      "loss": 0.8812,
+      "step": 4377
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.307437375750391,
+      "learning_rate": 7.437069038508888e-06,
+      "loss": 0.6284,
+      "step": 4378
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.498403905595587,
+      "learning_rate": 7.435913067413409e-06,
+      "loss": 1.0276,
+      "step": 4379
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.338473989667431,
+      "learning_rate": 7.4347569255710254e-06,
+      "loss": 0.4613,
+      "step": 4380
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.099509016808251,
+      "learning_rate": 7.433600613062777e-06,
+      "loss": 0.9861,
+      "step": 4381
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.289778871357129,
+      "learning_rate": 7.432444129969717e-06,
+      "loss": 0.823,
+      "step": 4382
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.002809802943864,
+      "learning_rate": 7.431287476372909e-06,
+      "loss": 0.9164,
+      "step": 4383
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.526139381340818,
+      "learning_rate": 7.43013065235343e-06,
+      "loss": 0.7189,
+      "step": 4384
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.7695362431841057,
+      "learning_rate": 7.4289736579923685e-06,
+      "loss": 0.7519,
+      "step": 4385
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.6250804019629017,
+      "learning_rate": 7.427816493370825e-06,
+      "loss": 0.8639,
+      "step": 4386
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.744309483105921,
+      "learning_rate": 7.426659158569911e-06,
+      "loss": 0.9766,
+      "step": 4387
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.399109856726638,
+      "learning_rate": 7.425501653670751e-06,
+      "loss": 1.0083,
+      "step": 4388
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.538105144605974,
+      "learning_rate": 7.4243439787544805e-06,
+      "loss": 1.0412,
+      "step": 4389
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.200788192853599,
+      "learning_rate": 7.423186133902247e-06,
+      "loss": 1.1422,
+      "step": 4390
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.5083238263558525,
+      "learning_rate": 7.422028119195213e-06,
+      "loss": 0.6474,
+      "step": 4391
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.88018038458345,
+      "learning_rate": 7.420869934714548e-06,
+      "loss": 0.6862,
+      "step": 4392
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.0792825826252725,
+      "learning_rate": 7.419711580541436e-06,
+      "loss": 1.0469,
+      "step": 4393
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1807818355369477,
+      "learning_rate": 7.418553056757072e-06,
+      "loss": 0.1271,
+      "step": 4394
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.50969090143382,
+      "learning_rate": 7.417394363442665e-06,
+      "loss": 0.5851,
+      "step": 4395
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.4359477881273057,
+      "learning_rate": 7.416235500679433e-06,
+      "loss": 0.5199,
+      "step": 4396
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.6176376746670478,
+      "learning_rate": 7.41507646854861e-06,
+      "loss": 0.7529,
+      "step": 4397
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.4386064681048185,
+      "learning_rate": 7.4139172671314344e-06,
+      "loss": 0.3244,
+      "step": 4398
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.6355116762246937,
+      "learning_rate": 7.412757896509164e-06,
+      "loss": 0.9841,
+      "step": 4399
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.632425199290458,
+      "learning_rate": 7.411598356763068e-06,
+      "loss": 1.095,
+      "step": 4400
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.1940908728026596,
+      "learning_rate": 7.41043864797442e-06,
+      "loss": 0.568,
+      "step": 4401
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.46000689781777,
+      "learning_rate": 7.409278770224515e-06,
+      "loss": 1.1823,
+      "step": 4402
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.912980192437,
+      "learning_rate": 7.4081187235946515e-06,
+      "loss": 0.8146,
+      "step": 4403
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.9160335596719977,
+      "learning_rate": 7.406958508166147e-06,
+      "loss": 0.8216,
+      "step": 4404
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.55346932823823,
+      "learning_rate": 7.405798124020326e-06,
+      "loss": 0.9,
+      "step": 4405
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.8839523637298212,
+      "learning_rate": 7.4046375712385256e-06,
+      "loss": 0.6489,
+      "step": 4406
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.26364676153241,
+      "learning_rate": 7.403476849902096e-06,
+      "loss": 0.5996,
+      "step": 4407
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.13212097104461,
+      "learning_rate": 7.402315960092401e-06,
+      "loss": 1.1889,
+      "step": 4408
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.462942185439612,
+      "learning_rate": 7.401154901890812e-06,
+      "loss": 0.8067,
+      "step": 4409
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.815555272242939,
+      "learning_rate": 7.399993675378714e-06,
+      "loss": 0.7739,
+      "step": 4410
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.445708292478904,
+      "learning_rate": 7.398832280637504e-06,
+      "loss": 1.3827,
+      "step": 4411
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.00619628143818,
+      "learning_rate": 7.397670717748591e-06,
+      "loss": 0.4527,
+      "step": 4412
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.454879259467784,
+      "learning_rate": 7.396508986793393e-06,
+      "loss": 0.7629,
+      "step": 4413
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.8921905481022745,
+      "learning_rate": 7.395347087853349e-06,
+      "loss": 0.7617,
+      "step": 4414
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.457760151179242,
+      "learning_rate": 7.394185021009895e-06,
+      "loss": 0.8917,
+      "step": 4415
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.1954274713742685,
+      "learning_rate": 7.393022786344492e-06,
+      "loss": 0.4957,
+      "step": 4416
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.4738524987507375,
+      "learning_rate": 7.391860383938607e-06,
+      "loss": 0.8411,
+      "step": 4417
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.476168980159589,
+      "learning_rate": 7.390697813873718e-06,
+      "loss": 0.6591,
+      "step": 4418
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.836743690795182,
+      "learning_rate": 7.389535076231315e-06,
+      "loss": 1.1051,
+      "step": 4419
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.70651116649257,
+      "learning_rate": 7.3883721710929045e-06,
+      "loss": 1.0457,
+      "step": 4420
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7965196368932832,
+      "learning_rate": 7.387209098539998e-06,
+      "loss": 0.2635,
+      "step": 4421
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.633654664167891,
+      "learning_rate": 7.386045858654123e-06,
+      "loss": 1.1022,
+      "step": 4422
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.2827126308298826,
+      "learning_rate": 7.384882451516817e-06,
+      "loss": 0.7331,
+      "step": 4423
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.420810694777215,
+      "learning_rate": 7.383718877209631e-06,
+      "loss": 1.0322,
+      "step": 4424
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.4038382834630596,
+      "learning_rate": 7.382555135814126e-06,
+      "loss": 0.4581,
+      "step": 4425
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.085944848338009,
+      "learning_rate": 7.381391227411875e-06,
+      "loss": 1.3082,
+      "step": 4426
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.6484173790981114,
+      "learning_rate": 7.380227152084461e-06,
+      "loss": 0.6663,
+      "step": 4427
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.625304518689477,
+      "learning_rate": 7.379062909913484e-06,
+      "loss": 0.3384,
+      "step": 4428
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.806506871861837,
+      "learning_rate": 7.37789850098055e-06,
+      "loss": 1.1255,
+      "step": 4429
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0405473109459544,
+      "learning_rate": 7.37673392536728e-06,
+      "loss": 0.5441,
+      "step": 4430
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.217603664239242,
+      "learning_rate": 7.375569183155306e-06,
+      "loss": 0.6702,
+      "step": 4431
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.465887043027505,
+      "learning_rate": 7.37440427442627e-06,
+      "loss": 0.8486,
+      "step": 4432
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.5703859822195945,
+      "learning_rate": 7.373239199261828e-06,
+      "loss": 0.9924,
+      "step": 4433
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.86403338734041,
+      "learning_rate": 7.372073957743646e-06,
+      "loss": 1.0849,
+      "step": 4434
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.8492373869585315,
+      "learning_rate": 7.370908549953404e-06,
+      "loss": 0.9398,
+      "step": 4435
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.111972033867977,
+      "learning_rate": 7.369742975972789e-06,
+      "loss": 0.7484,
+      "step": 4436
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.7084509760487485,
+      "learning_rate": 7.368577235883508e-06,
+      "loss": 0.7273,
+      "step": 4437
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.40040600697199,
+      "learning_rate": 7.367411329767267e-06,
+      "loss": 0.8338,
+      "step": 4438
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.070106174285915,
+      "learning_rate": 7.366245257705798e-06,
+      "loss": 0.5525,
+      "step": 4439
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.343881665537808,
+      "learning_rate": 7.365079019780832e-06,
+      "loss": 0.5944,
+      "step": 4440
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.12321303109076,
+      "learning_rate": 7.36391261607412e-06,
+      "loss": 1.0556,
+      "step": 4441
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.297429956371091,
+      "learning_rate": 7.3627460466674215e-06,
+      "loss": 0.8856,
+      "step": 4442
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.004084573771542,
+      "learning_rate": 7.361579311642508e-06,
+      "loss": 0.8428,
+      "step": 4443
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.137291119020999,
+      "learning_rate": 7.360412411081163e-06,
+      "loss": 0.8561,
+      "step": 4444
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.425871984712271,
+      "learning_rate": 7.35924534506518e-06,
+      "loss": 1.0632,
+      "step": 4445
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.263514033800204,
+      "learning_rate": 7.3580781136763656e-06,
+      "loss": 0.9843,
+      "step": 4446
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3638618591766134,
+      "learning_rate": 7.356910716996538e-06,
+      "loss": 0.2298,
+      "step": 4447
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.188339227415263,
+      "learning_rate": 7.355743155107526e-06,
+      "loss": 0.3663,
+      "step": 4448
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.7041795126120665,
+      "learning_rate": 7.354575428091172e-06,
+      "loss": 0.7665,
+      "step": 4449
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.401883725066224,
+      "learning_rate": 7.353407536029327e-06,
+      "loss": 0.7469,
+      "step": 4450
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.841773200375602,
+      "learning_rate": 7.352239479003857e-06,
+      "loss": 0.9938,
+      "step": 4451
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.715651076061036,
+      "learning_rate": 7.351071257096634e-06,
+      "loss": 1.2306,
+      "step": 4452
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.563175322043322,
+      "learning_rate": 7.349902870389549e-06,
+      "loss": 1.028,
+      "step": 4453
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.465074526425466,
+      "learning_rate": 7.3487343189645e-06,
+      "loss": 0.3103,
+      "step": 4454
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.055709724006543,
+      "learning_rate": 7.347565602903397e-06,
+      "loss": 0.6235,
+      "step": 4455
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.656519516509362,
+      "learning_rate": 7.346396722288162e-06,
+      "loss": 0.679,
+      "step": 4456
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.9727904646221956,
+      "learning_rate": 7.345227677200728e-06,
+      "loss": 0.5582,
+      "step": 4457
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.824112013113073,
+      "learning_rate": 7.34405846772304e-06,
+      "loss": 0.798,
+      "step": 4458
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1629289610763556,
+      "learning_rate": 7.3428890939370545e-06,
+      "loss": 0.1277,
+      "step": 4459
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.477409798493978,
+      "learning_rate": 7.341719555924741e-06,
+      "loss": 0.7387,
+      "step": 4460
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.1155639227406398,
+      "learning_rate": 7.3405498537680765e-06,
+      "loss": 0.5929,
+      "step": 4461
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.3179173250112637,
+      "learning_rate": 7.339379987549054e-06,
+      "loss": 0.5323,
+      "step": 4462
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.758447610320383,
+      "learning_rate": 7.338209957349677e-06,
+      "loss": 1.1001,
+      "step": 4463
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.9869551230787956,
+      "learning_rate": 7.337039763251956e-06,
+      "loss": 0.5098,
+      "step": 4464
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.228207987999737,
+      "learning_rate": 7.335869405337919e-06,
+      "loss": 1.0302,
+      "step": 4465
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.227773063389464,
+      "learning_rate": 7.334698883689601e-06,
+      "loss": 1.0155,
+      "step": 4466
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9999135138495103,
+      "learning_rate": 7.333528198389053e-06,
+      "loss": 0.3648,
+      "step": 4467
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.16060025922408,
+      "learning_rate": 7.332357349518334e-06,
+      "loss": 0.6752,
+      "step": 4468
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.04076814093737,
+      "learning_rate": 7.331186337159515e-06,
+      "loss": 1.1512,
+      "step": 4469
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.613831500731791,
+      "learning_rate": 7.3300151613946805e-06,
+      "loss": 0.9678,
+      "step": 4470
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.1838442374567535,
+      "learning_rate": 7.328843822305922e-06,
+      "loss": 0.7081,
+      "step": 4471
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.4458421972312783,
+      "learning_rate": 7.327672319975348e-06,
+      "loss": 0.8107,
+      "step": 4472
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.2583030157807977,
+      "learning_rate": 7.326500654485071e-06,
+      "loss": 0.5454,
+      "step": 4473
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.715040107956233,
+      "learning_rate": 7.325328825917226e-06,
+      "loss": 1.1163,
+      "step": 4474
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.0790611672845545,
+      "learning_rate": 7.3241568343539505e-06,
+      "loss": 0.5023,
+      "step": 4475
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.4918898891044106,
+      "learning_rate": 7.322984679877394e-06,
+      "loss": 0.4306,
+      "step": 4476
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8286276573382796,
+      "learning_rate": 7.3218123625697225e-06,
+      "loss": 0.4407,
+      "step": 4477
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8752323357289453,
+      "learning_rate": 7.320639882513108e-06,
+      "loss": 0.4156,
+      "step": 4478
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.1699152957715624,
+      "learning_rate": 7.319467239789738e-06,
+      "loss": 0.7109,
+      "step": 4479
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.585118381742184,
+      "learning_rate": 7.318294434481808e-06,
+      "loss": 1.0924,
+      "step": 4480
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.345711180050005,
+      "learning_rate": 7.317121466671528e-06,
+      "loss": 0.827,
+      "step": 4481
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.9439381191619036,
+      "learning_rate": 7.3159483364411175e-06,
+      "loss": 0.6872,
+      "step": 4482
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.0173118862391557,
+      "learning_rate": 7.314775043872807e-06,
+      "loss": 0.4207,
+      "step": 4483
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.989630248205145,
+      "learning_rate": 7.31360158904884e-06,
+      "loss": 0.7898,
+      "step": 4484
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.960866870194185,
+      "learning_rate": 7.3124279720514715e-06,
+      "loss": 0.8696,
+      "step": 4485
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.2405716519452077,
+      "learning_rate": 7.311254192962964e-06,
+      "loss": 0.716,
+      "step": 4486
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.4706557443470807,
+      "learning_rate": 7.3100802518655975e-06,
+      "loss": 0.8487,
+      "step": 4487
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.719720262071688,
+      "learning_rate": 7.308906148841659e-06,
+      "loss": 1.1216,
+      "step": 4488
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.172493649847887,
+      "learning_rate": 7.307731883973447e-06,
+      "loss": 0.6986,
+      "step": 4489
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.934165336710921,
+      "learning_rate": 7.306557457343273e-06,
+      "loss": 0.554,
+      "step": 4490
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.820002123608001,
+      "learning_rate": 7.30538286903346e-06,
+      "loss": 0.6866,
+      "step": 4491
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.646670289136569,
+      "learning_rate": 7.3042081191263415e-06,
+      "loss": 1.1311,
+      "step": 4492
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.7813068574299105,
+      "learning_rate": 7.30303320770426e-06,
+      "loss": 0.4627,
+      "step": 4493
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7240575636139102,
+      "learning_rate": 7.301858134849575e-06,
+      "loss": 0.2563,
+      "step": 4494
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.305120771122525,
+      "learning_rate": 7.3006829006446535e-06,
+      "loss": 0.8042,
+      "step": 4495
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.261801668048653,
+      "learning_rate": 7.299507505171871e-06,
+      "loss": 1.3942,
+      "step": 4496
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.204142139929167,
+      "learning_rate": 7.298331948513622e-06,
+      "loss": 1.1927,
+      "step": 4497
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.044871850015931,
+      "learning_rate": 7.297156230752303e-06,
+      "loss": 1.2654,
+      "step": 4498
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.337850294767246,
+      "learning_rate": 7.295980351970331e-06,
+      "loss": 0.908,
+      "step": 4499
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.053857052514148,
+      "learning_rate": 7.29480431225013e-06,
+      "loss": 0.6775,
+      "step": 4500
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.949137459079822,
+      "learning_rate": 7.2936281116741314e-06,
+      "loss": 0.9686,
+      "step": 4501
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.037736967540822,
+      "learning_rate": 7.292451750324785e-06,
+      "loss": 0.7931,
+      "step": 4502
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.3779809612700555,
+      "learning_rate": 7.291275228284549e-06,
+      "loss": 0.9467,
+      "step": 4503
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.418346099332919,
+      "learning_rate": 7.290098545635889e-06,
+      "loss": 0.7067,
+      "step": 4504
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.232539189776654,
+      "learning_rate": 7.288921702461289e-06,
+      "loss": 0.3651,
+      "step": 4505
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.5749360072395358,
+      "learning_rate": 7.287744698843237e-06,
+      "loss": 0.7257,
+      "step": 4506
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.0033083083553005,
+      "learning_rate": 7.28656753486424e-06,
+      "loss": 0.3874,
+      "step": 4507
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.740670590431987,
+      "learning_rate": 7.285390210606809e-06,
+      "loss": 1.3633,
+      "step": 4508
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.5485675390190594,
+      "learning_rate": 7.28421272615347e-06,
+      "loss": 0.8501,
+      "step": 4509
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.9958369179935214,
+      "learning_rate": 7.283035081586761e-06,
+      "loss": 0.9872,
+      "step": 4510
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.60533465076516,
+      "learning_rate": 7.281857276989228e-06,
+      "loss": 1.1645,
+      "step": 4511
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.431379925508597,
+      "learning_rate": 7.28067931244343e-06,
+      "loss": 0.8068,
+      "step": 4512
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.452343209098598,
+      "learning_rate": 7.279501188031939e-06,
+      "loss": 1.1694,
+      "step": 4513
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.6836445048120234,
+      "learning_rate": 7.278322903837334e-06,
+      "loss": 0.7585,
+      "step": 4514
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.9059582544576688,
+      "learning_rate": 7.2771444599422096e-06,
+      "loss": 0.6519,
+      "step": 4515
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.883169272294483,
+      "learning_rate": 7.275965856429167e-06,
+      "loss": 0.7735,
+      "step": 4516
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.639805399214749,
+      "learning_rate": 7.274787093380825e-06,
+      "loss": 0.7228,
+      "step": 4517
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.0174849334069562,
+      "learning_rate": 7.273608170879807e-06,
+      "loss": 0.8193,
+      "step": 4518
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.095551984163171,
+      "learning_rate": 7.27242908900875e-06,
+      "loss": 0.9427,
+      "step": 4519
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.704871089442316,
+      "learning_rate": 7.271249847850306e-06,
+      "loss": 0.429,
+      "step": 4520
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.820630439574544,
+      "learning_rate": 7.27007044748713e-06,
+      "loss": 0.9946,
+      "step": 4521
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.82734560676249,
+      "learning_rate": 7.268890888001896e-06,
+      "loss": 0.9962,
+      "step": 4522
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.9606092407357973,
+      "learning_rate": 7.267711169477284e-06,
+      "loss": 0.8341,
+      "step": 4523
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.245772471440562,
+      "learning_rate": 7.266531291995989e-06,
+      "loss": 0.6565,
+      "step": 4524
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.211696537207009,
+      "learning_rate": 7.265351255640713e-06,
+      "loss": 0.983,
+      "step": 4525
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.468724734753304,
+      "learning_rate": 7.2641710604941754e-06,
+      "loss": 0.8992,
+      "step": 4526
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.586107456241061,
+      "learning_rate": 7.262990706639097e-06,
+      "loss": 1.0833,
+      "step": 4527
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.473517836636879,
+      "learning_rate": 7.261810194158221e-06,
+      "loss": 0.8026,
+      "step": 4528
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.502267514552507,
+      "learning_rate": 7.260629523134293e-06,
+      "loss": 0.8131,
+      "step": 4529
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.220683416304137,
+      "learning_rate": 7.259448693650073e-06,
+      "loss": 0.42,
+      "step": 4530
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.8593087635128094,
+      "learning_rate": 7.258267705788334e-06,
+      "loss": 0.3321,
+      "step": 4531
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.6463886519185356,
+      "learning_rate": 7.2570865596318565e-06,
+      "loss": 0.5994,
+      "step": 4532
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.4815018579928054,
+      "learning_rate": 7.255905255263434e-06,
+      "loss": 0.6225,
+      "step": 4533
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.056716362932539,
+      "learning_rate": 7.254723792765872e-06,
+      "loss": 1.0677,
+      "step": 4534
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.332364971141776,
+      "learning_rate": 7.253542172221982e-06,
+      "loss": 0.7248,
+      "step": 4535
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.471393689780005,
+      "learning_rate": 7.252360393714595e-06,
+      "loss": 0.8082,
+      "step": 4536
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.980817164222064,
+      "learning_rate": 7.251178457326547e-06,
+      "loss": 0.6879,
+      "step": 4537
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.319603041963911,
+      "learning_rate": 7.249996363140686e-06,
+      "loss": 0.7517,
+      "step": 4538
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.220525399600135,
+      "learning_rate": 7.24881411123987e-06,
+      "loss": 0.9474,
+      "step": 4539
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.4638507168662445,
+      "learning_rate": 7.247631701706974e-06,
+      "loss": 0.6042,
+      "step": 4540
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.2917495569563706,
+      "learning_rate": 7.246449134624878e-06,
+      "loss": 0.2501,
+      "step": 4541
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.9933586455885357,
+      "learning_rate": 7.2452664100764725e-06,
+      "loss": 0.5591,
+      "step": 4542
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.604732548760846,
+      "learning_rate": 7.244083528144663e-06,
+      "loss": 0.6803,
+      "step": 4543
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.220440325230384,
+      "learning_rate": 7.242900488912364e-06,
+      "loss": 0.7553,
+      "step": 4544
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.861147346123085,
+      "learning_rate": 7.241717292462505e-06,
+      "loss": 1.7752,
+      "step": 4545
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.233779493000437,
+      "learning_rate": 7.240533938878016e-06,
+      "loss": 0.6548,
+      "step": 4546
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.0855895495993035,
+      "learning_rate": 7.239350428241851e-06,
+      "loss": 1.3626,
+      "step": 4547
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.136242258149596,
+      "learning_rate": 7.238166760636966e-06,
+      "loss": 1.0945,
+      "step": 4548
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.536209408736868,
+      "learning_rate": 7.236982936146332e-06,
+      "loss": 0.3704,
+      "step": 4549
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.679206527422662,
+      "learning_rate": 7.235798954852929e-06,
+      "loss": 0.4141,
+      "step": 4550
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.609558168860129,
+      "learning_rate": 7.23461481683975e-06,
+      "loss": 1.0022,
+      "step": 4551
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.9874756521860313,
+      "learning_rate": 7.233430522189797e-06,
+      "loss": 0.8539,
+      "step": 4552
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.117818020127755,
+      "learning_rate": 7.232246070986084e-06,
+      "loss": 0.8997,
+      "step": 4553
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.8538029708155297,
+      "learning_rate": 7.2310614633116376e-06,
+      "loss": 0.7597,
+      "step": 4554
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.0481157383577426,
+      "learning_rate": 7.22987669924949e-06,
+      "loss": 0.6272,
+      "step": 4555
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.4563412200333063,
+      "learning_rate": 7.2286917788826926e-06,
+      "loss": 0.8794,
+      "step": 4556
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.077233686569749,
+      "learning_rate": 7.2275067022943005e-06,
+      "loss": 1.1632,
+      "step": 4557
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.0498537414819245,
+      "learning_rate": 7.226321469567381e-06,
+      "loss": 0.4963,
+      "step": 4558
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.4327224589889855,
+      "learning_rate": 7.225136080785016e-06,
+      "loss": 1.0744,
+      "step": 4559
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.911606630606663,
+      "learning_rate": 7.223950536030297e-06,
+      "loss": 1.0902,
+      "step": 4560
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.122772441786447,
+      "learning_rate": 7.2227648353863225e-06,
+      "loss": 1.3166,
+      "step": 4561
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.281493251686354,
+      "learning_rate": 7.221578978936207e-06,
+      "loss": 0.7545,
+      "step": 4562
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.8442070070178835,
+      "learning_rate": 7.220392966763072e-06,
+      "loss": 0.7848,
+      "step": 4563
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.704511580743091,
+      "learning_rate": 7.219206798950056e-06,
+      "loss": 0.9293,
+      "step": 4564
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.8800793017540376,
+      "learning_rate": 7.218020475580301e-06,
+      "loss": 0.7855,
+      "step": 4565
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.285617047839879,
+      "learning_rate": 7.216833996736963e-06,
+      "loss": 0.8195,
+      "step": 4566
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.473580101703317,
+      "learning_rate": 7.2156473625032075e-06,
+      "loss": 1.4403,
+      "step": 4567
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.510327755197747,
+      "learning_rate": 7.2144605729622166e-06,
+      "loss": 1.1779,
+      "step": 4568
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.0337100328946316,
+      "learning_rate": 7.213273628197176e-06,
+      "loss": 0.642,
+      "step": 4569
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.993271610598595,
+      "learning_rate": 7.212086528291286e-06,
+      "loss": 0.9385,
+      "step": 4570
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.5938483151970684,
+      "learning_rate": 7.2108992733277595e-06,
+      "loss": 0.5875,
+      "step": 4571
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.905467746768345,
+      "learning_rate": 7.209711863389815e-06,
+      "loss": 0.7408,
+      "step": 4572
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.3494342996801905,
+      "learning_rate": 7.208524298560684e-06,
+      "loss": 0.7769,
+      "step": 4573
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.3883729246932255,
+      "learning_rate": 7.207336578923613e-06,
+      "loss": 0.5969,
+      "step": 4574
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1952773938067043,
+      "learning_rate": 7.206148704561853e-06,
+      "loss": 0.2003,
+      "step": 4575
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.2361353989913852,
+      "learning_rate": 7.20496067555867e-06,
+      "loss": 0.7028,
+      "step": 4576
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.4319513704107956,
+      "learning_rate": 7.20377249199734e-06,
+      "loss": 0.6851,
+      "step": 4577
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.76818205154338,
+      "learning_rate": 7.2025841539611485e-06,
+      "loss": 0.7971,
+      "step": 4578
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.088168301961054,
+      "learning_rate": 7.201395661533395e-06,
+      "loss": 0.3357,
+      "step": 4579
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.965093420537595,
+      "learning_rate": 7.200207014797385e-06,
+      "loss": 0.8798,
+      "step": 4580
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.1078580794068604,
+      "learning_rate": 7.19901821383644e-06,
+      "loss": 0.4883,
+      "step": 4581
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.9574296718617967,
+      "learning_rate": 7.197829258733886e-06,
+      "loss": 1.0293,
+      "step": 4582
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.736266789641212,
+      "learning_rate": 7.1966401495730675e-06,
+      "loss": 0.6452,
+      "step": 4583
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.842006382916651,
+      "learning_rate": 7.195450886437334e-06,
+      "loss": 0.3723,
+      "step": 4584
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.885450874992221,
+      "learning_rate": 7.1942614694100476e-06,
+      "loss": 0.8921,
+      "step": 4585
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.33205324545829,
+      "learning_rate": 7.1930718985745815e-06,
+      "loss": 1.0317,
+      "step": 4586
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.603549135962034,
+      "learning_rate": 7.1918821740143196e-06,
+      "loss": 0.88,
+      "step": 4587
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.2702414115235223,
+      "learning_rate": 7.190692295812658e-06,
+      "loss": 0.4342,
+      "step": 4588
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.7315755603115504,
+      "learning_rate": 7.189502264053e-06,
+      "loss": 0.8002,
+      "step": 4589
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.350066189445501,
+      "learning_rate": 7.188312078818761e-06,
+      "loss": 1.4728,
+      "step": 4590
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.6486634707171044,
+      "learning_rate": 7.18712174019337e-06,
+      "loss": 0.8281,
+      "step": 4591
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.2919211998095594,
+      "learning_rate": 7.185931248260262e-06,
+      "loss": 0.78,
+      "step": 4592
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.5585559243371536,
+      "learning_rate": 7.1847406031028866e-06,
+      "loss": 0.4944,
+      "step": 4593
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.0813044061523454,
+      "learning_rate": 7.183549804804704e-06,
+      "loss": 0.3836,
+      "step": 4594
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.3181687589362014,
+      "learning_rate": 7.182358853449183e-06,
+      "loss": 0.4667,
+      "step": 4595
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.826918066825633,
+      "learning_rate": 7.181167749119804e-06,
+      "loss": 0.5078,
+      "step": 4596
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.630806612184872,
+      "learning_rate": 7.179976491900058e-06,
+      "loss": 1.3874,
+      "step": 4597
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.0602457597906483,
+      "learning_rate": 7.178785081873446e-06,
+      "loss": 0.5455,
+      "step": 4598
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.5222743372079512,
+      "learning_rate": 7.177593519123483e-06,
+      "loss": 0.4905,
+      "step": 4599
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.188427047383891,
+      "learning_rate": 7.176401803733691e-06,
+      "loss": 0.7658,
+      "step": 4600
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.7162178644039856,
+      "learning_rate": 7.175209935787605e-06,
+      "loss": 0.9036,
+      "step": 4601
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.602496845755888,
+      "learning_rate": 7.174017915368769e-06,
+      "loss": 0.9856,
+      "step": 4602
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.381026091099391,
+      "learning_rate": 7.172825742560737e-06,
+      "loss": 0.6102,
+      "step": 4603
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.774135585966469,
+      "learning_rate": 7.171633417447078e-06,
+      "loss": 0.8705,
+      "step": 4604
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.7188104057779228,
+      "learning_rate": 7.170440940111367e-06,
+      "loss": 0.6872,
+      "step": 4605
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.4582576816064883,
+      "learning_rate": 7.169248310637192e-06,
+      "loss": 0.9781,
+      "step": 4606
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.6689730999616215,
+      "learning_rate": 7.168055529108151e-06,
+      "loss": 0.8557,
+      "step": 4607
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.716459932166737,
+      "learning_rate": 7.166862595607853e-06,
+      "loss": 0.8339,
+      "step": 4608
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.144471309239014,
+      "learning_rate": 7.165669510219917e-06,
+      "loss": 1.0876,
+      "step": 4609
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.411698165275617,
+      "learning_rate": 7.164476273027973e-06,
+      "loss": 0.961,
+      "step": 4610
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.257966379763284,
+      "learning_rate": 7.163282884115662e-06,
+      "loss": 0.6124,
+      "step": 4611
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.2342756996710795,
+      "learning_rate": 7.1620893435666375e-06,
+      "loss": 0.8242,
+      "step": 4612
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9126104647795454,
+      "learning_rate": 7.160895651464557e-06,
+      "loss": 0.4754,
+      "step": 4613
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.460995081548863,
+      "learning_rate": 7.159701807893097e-06,
+      "loss": 1.2291,
+      "step": 4614
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.141431103320525,
+      "learning_rate": 7.1585078129359385e-06,
+      "loss": 0.9833,
+      "step": 4615
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.277701753729068,
+      "learning_rate": 7.157313666676775e-06,
+      "loss": 0.9546,
+      "step": 4616
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.7258342660221495,
+      "learning_rate": 7.156119369199315e-06,
+      "loss": 0.5565,
+      "step": 4617
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.75708470979647,
+      "learning_rate": 7.154924920587269e-06,
+      "loss": 0.8413,
+      "step": 4618
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.422246019971207,
+      "learning_rate": 7.153730320924365e-06,
+      "loss": 0.9829,
+      "step": 4619
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.067999875100732,
+      "learning_rate": 7.152535570294339e-06,
+      "loss": 0.9425,
+      "step": 4620
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.242856394561573,
+      "learning_rate": 7.151340668780935e-06,
+      "loss": 0.8913,
+      "step": 4621
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.493206777102948,
+      "learning_rate": 7.150145616467916e-06,
+      "loss": 0.5349,
+      "step": 4622
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.618796024916159,
+      "learning_rate": 7.148950413439044e-06,
+      "loss": 0.8845,
+      "step": 4623
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.8203092018052547,
+      "learning_rate": 7.147755059778101e-06,
+      "loss": 0.4983,
+      "step": 4624
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.312465249288361,
+      "learning_rate": 7.146559555568876e-06,
+      "loss": 0.6341,
+      "step": 4625
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.007985232158787,
+      "learning_rate": 7.145363900895168e-06,
+      "loss": 0.8231,
+      "step": 4626
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.36413579721577,
+      "learning_rate": 7.144168095840786e-06,
+      "loss": 0.6309,
+      "step": 4627
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.760414558968082,
+      "learning_rate": 7.142972140489555e-06,
+      "loss": 0.9974,
+      "step": 4628
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.611181049430095,
+      "learning_rate": 7.1417760349253005e-06,
+      "loss": 0.9156,
+      "step": 4629
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.496988257817793,
+      "learning_rate": 7.140579779231866e-06,
+      "loss": 0.5388,
+      "step": 4630
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.9493649618885285,
+      "learning_rate": 7.139383373493107e-06,
+      "loss": 0.5925,
+      "step": 4631
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.630837403273287,
+      "learning_rate": 7.1381868177928834e-06,
+      "loss": 0.7775,
+      "step": 4632
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.618547196665511,
+      "learning_rate": 7.1369901122150694e-06,
+      "loss": 0.8912,
+      "step": 4633
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.5624447846221794,
+      "learning_rate": 7.1357932568435496e-06,
+      "loss": 0.9665,
+      "step": 4634
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.767445364020535,
+      "learning_rate": 7.134596251762217e-06,
+      "loss": 1.3131,
+      "step": 4635
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.482513277230487,
+      "learning_rate": 7.1333990970549764e-06,
+      "loss": 1.0189,
+      "step": 4636
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.301321079555022,
+      "learning_rate": 7.132201792805744e-06,
+      "loss": 0.8224,
+      "step": 4637
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.05399806054526,
+      "learning_rate": 7.131004339098445e-06,
+      "loss": 0.5469,
+      "step": 4638
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.792266429876065,
+      "learning_rate": 7.129806736017015e-06,
+      "loss": 0.8165,
+      "step": 4639
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.72525028465202,
+      "learning_rate": 7.128608983645404e-06,
+      "loss": 0.6924,
+      "step": 4640
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.587819774226734,
+      "learning_rate": 7.127411082067566e-06,
+      "loss": 1.0171,
+      "step": 4641
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.247042866915213,
+      "learning_rate": 7.126213031367471e-06,
+      "loss": 0.5919,
+      "step": 4642
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.3236833483018917,
+      "learning_rate": 7.1250148316290936e-06,
+      "loss": 0.5729,
+      "step": 4643
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.0220944780262355,
+      "learning_rate": 7.123816482936425e-06,
+      "loss": 0.9583,
+      "step": 4644
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.455372502724746,
+      "learning_rate": 7.122617985373466e-06,
+      "loss": 0.6043,
+      "step": 4645
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.784670669336466,
+      "learning_rate": 7.12141933902422e-06,
+      "loss": 1.4889,
+      "step": 4646
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.5993468761311838,
+      "learning_rate": 7.120220543972714e-06,
+      "loss": 0.7301,
+      "step": 4647
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.180978435085142,
+      "learning_rate": 7.119021600302973e-06,
+      "loss": 0.9515,
+      "step": 4648
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.3315021353161502,
+      "learning_rate": 7.117822508099042e-06,
+      "loss": 0.7091,
+      "step": 4649
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.7542941945456128,
+      "learning_rate": 7.116623267444969e-06,
+      "loss": 0.7678,
+      "step": 4650
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.664733599625183,
+      "learning_rate": 7.115423878424817e-06,
+      "loss": 1.1193,
+      "step": 4651
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.516391532806129,
+      "learning_rate": 7.114224341122655e-06,
+      "loss": 1.5861,
+      "step": 4652
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.9343289636014185,
+      "learning_rate": 7.113024655622571e-06,
+      "loss": 0.4885,
+      "step": 4653
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.356400620423507,
+      "learning_rate": 7.111824822008653e-06,
+      "loss": 0.6609,
+      "step": 4654
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.128175145715575,
+      "learning_rate": 7.110624840365005e-06,
+      "loss": 0.9746,
+      "step": 4655
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.2156393306671833,
+      "learning_rate": 7.109424710775742e-06,
+      "loss": 0.484,
+      "step": 4656
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.1678849959490245,
+      "learning_rate": 7.108224433324987e-06,
+      "loss": 0.555,
+      "step": 4657
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.5124720518160513,
+      "learning_rate": 7.107024008096874e-06,
+      "loss": 0.5308,
+      "step": 4658
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.8215022406943118,
+      "learning_rate": 7.105823435175549e-06,
+      "loss": 0.4427,
+      "step": 4659
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.791148483114985,
+      "learning_rate": 7.104622714645165e-06,
+      "loss": 1.3376,
+      "step": 4660
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.2189910061655698,
+      "learning_rate": 7.103421846589888e-06,
+      "loss": 0.5059,
+      "step": 4661
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.2099581856724844,
+      "learning_rate": 7.102220831093893e-06,
+      "loss": 0.5016,
+      "step": 4662
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.551899396842402,
+      "learning_rate": 7.101019668241368e-06,
+      "loss": 0.7423,
+      "step": 4663
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.839274902607117,
+      "learning_rate": 7.099818358116506e-06,
+      "loss": 0.8159,
+      "step": 4664
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.091721458837044,
+      "learning_rate": 7.0986169008035175e-06,
+      "loss": 0.7227,
+      "step": 4665
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.497212176209524,
+      "learning_rate": 7.097415296386617e-06,
+      "loss": 0.9619,
+      "step": 4666
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.133435638548798,
+      "learning_rate": 7.096213544950032e-06,
+      "loss": 0.9201,
+      "step": 4667
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7868724618128153,
+      "learning_rate": 7.095011646578001e-06,
+      "loss": 0.3228,
+      "step": 4668
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.6219947054937496,
+      "learning_rate": 7.093809601354769e-06,
+      "loss": 0.615,
+      "step": 4669
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.4273523018159104,
+      "learning_rate": 7.092607409364597e-06,
+      "loss": 0.6523,
+      "step": 4670
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.217834982538416,
+      "learning_rate": 7.0914050706917536e-06,
+      "loss": 0.7628,
+      "step": 4671
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.023922645696487,
+      "learning_rate": 7.090202585420516e-06,
+      "loss": 1.1023,
+      "step": 4672
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.1929940805848345,
+      "learning_rate": 7.088999953635174e-06,
+      "loss": 0.6617,
+      "step": 4673
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.27375063412841,
+      "learning_rate": 7.087797175420028e-06,
+      "loss": 1.1346,
+      "step": 4674
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.14383253909754,
+      "learning_rate": 7.086594250859383e-06,
+      "loss": 1.0806,
+      "step": 4675
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.798510194907834,
+      "learning_rate": 7.085391180037564e-06,
+      "loss": 0.8079,
+      "step": 4676
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.836075392491869,
+      "learning_rate": 7.084187963038899e-06,
+      "loss": 0.9677,
+      "step": 4677
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.100173145603222,
+      "learning_rate": 7.082984599947727e-06,
+      "loss": 0.8584,
+      "step": 4678
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6888219478857118,
+      "learning_rate": 7.0817810908483995e-06,
+      "loss": 0.2592,
+      "step": 4679
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.1288526522734137,
+      "learning_rate": 7.080577435825279e-06,
+      "loss": 0.8195,
+      "step": 4680
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.1526606107911697,
+      "learning_rate": 7.079373634962735e-06,
+      "loss": 0.8578,
+      "step": 4681
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.7193182479352145,
+      "learning_rate": 7.0781696883451486e-06,
+      "loss": 0.9729,
+      "step": 4682
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8929230926532878,
+      "learning_rate": 7.076965596056911e-06,
+      "loss": 0.4022,
+      "step": 4683
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.720322440920711,
+      "learning_rate": 7.075761358182423e-06,
+      "loss": 0.5376,
+      "step": 4684
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.45190880528166,
+      "learning_rate": 7.074556974806098e-06,
+      "loss": 0.9728,
+      "step": 4685
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.503468316807604,
+      "learning_rate": 7.073352446012357e-06,
+      "loss": 0.7926,
+      "step": 4686
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.194545270931051,
+      "learning_rate": 7.072147771885633e-06,
+      "loss": 1.4915,
+      "step": 4687
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.941519706504127,
+      "learning_rate": 7.07094295251037e-06,
+      "loss": 1.0565,
+      "step": 4688
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.2702531812199265,
+      "learning_rate": 7.069737987971017e-06,
+      "loss": 0.771,
+      "step": 4689
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.864746525153197,
+      "learning_rate": 7.06853287835204e-06,
+      "loss": 0.6395,
+      "step": 4690
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.320869939847818,
+      "learning_rate": 7.06732762373791e-06,
+      "loss": 0.8968,
+      "step": 4691
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.8603919227263654,
+      "learning_rate": 7.06612222421311e-06,
+      "loss": 0.5715,
+      "step": 4692
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.815299350955375,
+      "learning_rate": 7.064916679862134e-06,
+      "loss": 0.2505,
+      "step": 4693
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.416622876392271,
+      "learning_rate": 7.0637109907694855e-06,
+      "loss": 0.6977,
+      "step": 4694
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.325136418083608,
+      "learning_rate": 7.062505157019678e-06,
+      "loss": 1.155,
+      "step": 4695
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.29304724423045,
+      "learning_rate": 7.061299178697234e-06,
+      "loss": 0.8864,
+      "step": 4696
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.2348968803133,
+      "learning_rate": 7.06009305588669e-06,
+      "loss": 1.2064,
+      "step": 4697
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.325225458239991,
+      "learning_rate": 7.058886788672588e-06,
+      "loss": 0.816,
+      "step": 4698
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.7366570370091066,
+      "learning_rate": 7.057680377139482e-06,
+      "loss": 0.549,
+      "step": 4699
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.6770758782743433,
+      "learning_rate": 7.056473821371936e-06,
+      "loss": 1.0477,
+      "step": 4700
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.9101510171746092,
+      "learning_rate": 7.055267121454525e-06,
+      "loss": 1.0068,
+      "step": 4701
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.387053921551004,
+      "learning_rate": 7.054060277471834e-06,
+      "loss": 1.2587,
+      "step": 4702
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.8716796704049807,
+      "learning_rate": 7.052853289508458e-06,
+      "loss": 0.7437,
+      "step": 4703
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.101094078305496,
+      "learning_rate": 7.051646157648998e-06,
+      "loss": 0.5624,
+      "step": 4704
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.2949686093473995,
+      "learning_rate": 7.050438881978073e-06,
+      "loss": 0.5994,
+      "step": 4705
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.680571162710411,
+      "learning_rate": 7.049231462580306e-06,
+      "loss": 0.984,
+      "step": 4706
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.764122359448531,
+      "learning_rate": 7.0480238995403305e-06,
+      "loss": 1.0566,
+      "step": 4707
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.3283288351922162,
+      "learning_rate": 7.046816192942794e-06,
+      "loss": 0.867,
+      "step": 4708
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.8178504026156137,
+      "learning_rate": 7.045608342872349e-06,
+      "loss": 0.6565,
+      "step": 4709
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.594460551811602,
+      "learning_rate": 7.044400349413661e-06,
+      "loss": 1.315,
+      "step": 4710
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.49430484143887,
+      "learning_rate": 7.043192212651407e-06,
+      "loss": 0.326,
+      "step": 4711
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.924263522486229,
+      "learning_rate": 7.041983932670271e-06,
+      "loss": 1.0455,
+      "step": 4712
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.046648117170458,
+      "learning_rate": 7.040775509554948e-06,
+      "loss": 0.531,
+      "step": 4713
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.7765304854224135,
+      "learning_rate": 7.039566943390144e-06,
+      "loss": 0.717,
+      "step": 4714
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.7657904788378427,
+      "learning_rate": 7.038358234260572e-06,
+      "loss": 0.9202,
+      "step": 4715
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.952201527018048,
+      "learning_rate": 7.037149382250959e-06,
+      "loss": 0.5853,
+      "step": 4716
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.1458126725943245,
+      "learning_rate": 7.035940387446041e-06,
+      "loss": 0.6387,
+      "step": 4717
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.345177420459916,
+      "learning_rate": 7.03473124993056e-06,
+      "loss": 0.7602,
+      "step": 4718
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9731307704986518,
+      "learning_rate": 7.033521969789275e-06,
+      "loss": 0.3163,
+      "step": 4719
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.9952518683111844,
+      "learning_rate": 7.03231254710695e-06,
+      "loss": 0.7432,
+      "step": 4720
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.6574995418647163,
+      "learning_rate": 7.031102981968361e-06,
+      "loss": 0.5555,
+      "step": 4721
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.1569025298242925,
+      "learning_rate": 7.029893274458291e-06,
+      "loss": 0.4483,
+      "step": 4722
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.535936462310264,
+      "learning_rate": 7.028683424661538e-06,
+      "loss": 0.9622,
+      "step": 4723
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.7605565413311677,
+      "learning_rate": 7.0274734326629035e-06,
+      "loss": 0.6126,
+      "step": 4724
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.8761513436527633,
+      "learning_rate": 7.026263298547207e-06,
+      "loss": 0.7268,
+      "step": 4725
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.849127956931701,
+      "learning_rate": 7.025053022399271e-06,
+      "loss": 0.9776,
+      "step": 4726
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.2426848629555685,
+      "learning_rate": 7.02384260430393e-06,
+      "loss": 0.4053,
+      "step": 4727
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.547617522634692,
+      "learning_rate": 7.022632044346032e-06,
+      "loss": 1.2811,
+      "step": 4728
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.7078112908147984,
+      "learning_rate": 7.0214213426104295e-06,
+      "loss": 0.6024,
+      "step": 4729
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.5512443500532593,
+      "learning_rate": 7.020210499181988e-06,
+      "loss": 0.7873,
+      "step": 4730
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.8017442415960017,
+      "learning_rate": 7.0189995141455836e-06,
+      "loss": 0.6312,
+      "step": 4731
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.198275952054718,
+      "learning_rate": 7.017788387586097e-06,
+      "loss": 0.9264,
+      "step": 4732
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.9401947156048407,
+      "learning_rate": 7.016577119588428e-06,
+      "loss": 0.8761,
+      "step": 4733
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.407470671577025,
+      "learning_rate": 7.015365710237479e-06,
+      "loss": 0.8748,
+      "step": 4734
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.3772780839984895,
+      "learning_rate": 7.0141541596181654e-06,
+      "loss": 1.3098,
+      "step": 4735
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.125854112287967,
+      "learning_rate": 7.01294246781541e-06,
+      "loss": 0.606,
+      "step": 4736
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.953373275840162,
+      "learning_rate": 7.0117306349141485e-06,
+      "loss": 0.5249,
+      "step": 4737
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.831197244885152,
+      "learning_rate": 7.010518660999324e-06,
+      "loss": 1.3313,
+      "step": 4738
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.08326069466575,
+      "learning_rate": 7.009306546155889e-06,
+      "loss": 0.6701,
+      "step": 4739
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.6870448741867894,
+      "learning_rate": 7.008094290468813e-06,
+      "loss": 0.9529,
+      "step": 4740
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.124650538400718,
+      "learning_rate": 7.006881894023065e-06,
+      "loss": 0.8456,
+      "step": 4741
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.983367153795829,
+      "learning_rate": 7.005669356903631e-06,
+      "loss": 1.1963,
+      "step": 4742
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.0868385441047606,
+      "learning_rate": 7.004456679195503e-06,
+      "loss": 0.8454,
+      "step": 4743
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.0532609433396725,
+      "learning_rate": 7.003243860983686e-06,
+      "loss": 0.6778,
+      "step": 4744
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.0415860533912085,
+      "learning_rate": 7.002030902353193e-06,
+      "loss": 0.8216,
+      "step": 4745
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.584568112521553,
+      "learning_rate": 7.000817803389045e-06,
+      "loss": 0.9213,
+      "step": 4746
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.4185218446827954,
+      "learning_rate": 6.999604564176277e-06,
+      "loss": 0.8299,
+      "step": 4747
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.722918145577284,
+      "learning_rate": 6.998391184799932e-06,
+      "loss": 1.7364,
+      "step": 4748
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.229748668312871,
+      "learning_rate": 6.99717766534506e-06,
+      "loss": 0.8441,
+      "step": 4749
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.1861083971798494,
+      "learning_rate": 6.995964005896727e-06,
+      "loss": 0.4733,
+      "step": 4750
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.2484214364248305,
+      "learning_rate": 6.994750206540004e-06,
+      "loss": 0.7676,
+      "step": 4751
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.449813091236215,
+      "learning_rate": 6.993536267359974e-06,
+      "loss": 0.9787,
+      "step": 4752
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.6520084624347993,
+      "learning_rate": 6.992322188441725e-06,
+      "loss": 0.9034,
+      "step": 4753
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.1529575535292254,
+      "learning_rate": 6.991107969870363e-06,
+      "loss": 0.4434,
+      "step": 4754
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.303898302529973,
+      "learning_rate": 6.989893611730996e-06,
+      "loss": 0.6798,
+      "step": 4755
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.822231187335992,
+      "learning_rate": 6.988679114108747e-06,
+      "loss": 1.0074,
+      "step": 4756
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.588487698818553,
+      "learning_rate": 6.987464477088748e-06,
+      "loss": 0.8526,
+      "step": 4757
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.1697715748788413,
+      "learning_rate": 6.986249700756138e-06,
+      "loss": 0.3875,
+      "step": 4758
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.4852047598958436,
+      "learning_rate": 6.985034785196069e-06,
+      "loss": 1.0415,
+      "step": 4759
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.216136059924069,
+      "learning_rate": 6.983819730493699e-06,
+      "loss": 0.6872,
+      "step": 4760
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.5697743111338145,
+      "learning_rate": 6.9826045367342e-06,
+      "loss": 0.6744,
+      "step": 4761
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.286634141739833,
+      "learning_rate": 6.981389204002751e-06,
+      "loss": 0.7238,
+      "step": 4762
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.5571073222616736,
+      "learning_rate": 6.980173732384543e-06,
+      "loss": 0.5277,
+      "step": 4763
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.818274543221162,
+      "learning_rate": 6.978958121964773e-06,
+      "loss": 1.4853,
+      "step": 4764
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.9200206363249097,
+      "learning_rate": 6.977742372828652e-06,
+      "loss": 0.9199,
+      "step": 4765
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.176623938334342,
+      "learning_rate": 6.976526485061397e-06,
+      "loss": 0.6575,
+      "step": 4766
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.174931941643446,
+      "learning_rate": 6.975310458748238e-06,
+      "loss": 1.1418,
+      "step": 4767
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.2719845268830756,
+      "learning_rate": 6.9740942939744125e-06,
+      "loss": 0.5741,
+      "step": 4768
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.788764754495829,
+      "learning_rate": 6.972877990825167e-06,
+      "loss": 0.7534,
+      "step": 4769
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.6073263780058684,
+      "learning_rate": 6.971661549385762e-06,
+      "loss": 0.6959,
+      "step": 4770
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.604762311779483,
+      "learning_rate": 6.970444969741462e-06,
+      "loss": 1.6464,
+      "step": 4771
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.7151129366464,
+      "learning_rate": 6.969228251977545e-06,
+      "loss": 0.552,
+      "step": 4772
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.972356743488502,
+      "learning_rate": 6.968011396179297e-06,
+      "loss": 0.7818,
+      "step": 4773
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.365531384342218,
+      "learning_rate": 6.966794402432017e-06,
+      "loss": 0.469,
+      "step": 4774
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.091976058428769,
+      "learning_rate": 6.965577270821008e-06,
+      "loss": 0.8084,
+      "step": 4775
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.840710225808364,
+      "learning_rate": 6.964360001431586e-06,
+      "loss": 1.1058,
+      "step": 4776
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.307707946660783,
+      "learning_rate": 6.963142594349077e-06,
+      "loss": 1.0741,
+      "step": 4777
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.7821674714992835,
+      "learning_rate": 6.961925049658816e-06,
+      "loss": 1.1528,
+      "step": 4778
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.610844148978979,
+      "learning_rate": 6.960707367446147e-06,
+      "loss": 0.9325,
+      "step": 4779
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.9451373481729353,
+      "learning_rate": 6.959489547796426e-06,
+      "loss": 0.6211,
+      "step": 4780
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.697845570706351,
+      "learning_rate": 6.958271590795014e-06,
+      "loss": 0.5125,
+      "step": 4781
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.993889001188721,
+      "learning_rate": 6.957053496527286e-06,
+      "loss": 1.0208,
+      "step": 4782
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.9045862533892004,
+      "learning_rate": 6.955835265078626e-06,
+      "loss": 1.0204,
+      "step": 4783
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.574727633134075,
+      "learning_rate": 6.954616896534424e-06,
+      "loss": 0.8613,
+      "step": 4784
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.0174056463238688,
+      "learning_rate": 6.953398390980086e-06,
+      "loss": 0.5144,
+      "step": 4785
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.568291338854759,
+      "learning_rate": 6.95217974850102e-06,
+      "loss": 0.7935,
+      "step": 4786
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.635262649776755,
+      "learning_rate": 6.950960969182649e-06,
+      "loss": 0.6803,
+      "step": 4787
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.877628339574649,
+      "learning_rate": 6.949742053110408e-06,
+      "loss": 0.9124,
+      "step": 4788
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.122801296062488,
+      "learning_rate": 6.9485230003697325e-06,
+      "loss": 0.6189,
+      "step": 4789
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.4939221191245924,
+      "learning_rate": 6.947303811046074e-06,
+      "loss": 0.6689,
+      "step": 4790
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.140755038586852,
+      "learning_rate": 6.946084485224895e-06,
+      "loss": 0.6994,
+      "step": 4791
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.215834949502133,
+      "learning_rate": 6.944865022991661e-06,
+      "loss": 0.986,
+      "step": 4792
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.4420112654537522,
+      "learning_rate": 6.943645424431854e-06,
+      "loss": 0.6215,
+      "step": 4793
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.709803801079052,
+      "learning_rate": 6.942425689630962e-06,
+      "loss": 0.7751,
+      "step": 4794
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.797934906655287,
+      "learning_rate": 6.941205818674482e-06,
+      "loss": 1.1691,
+      "step": 4795
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.187981917818049,
+      "learning_rate": 6.93998581164792e-06,
+      "loss": 1.2553,
+      "step": 4796
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.89920836804541,
+      "learning_rate": 6.9387656686367985e-06,
+      "loss": 1.244,
+      "step": 4797
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.9961056564406827,
+      "learning_rate": 6.93754538972664e-06,
+      "loss": 0.5682,
+      "step": 4798
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.6553136476349426,
+      "learning_rate": 6.936324975002983e-06,
+      "loss": 0.8373,
+      "step": 4799
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.015444678582228,
+      "learning_rate": 6.935104424551372e-06,
+      "loss": 0.7753,
+      "step": 4800
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.835661988834962,
+      "learning_rate": 6.933883738457361e-06,
+      "loss": 0.7163,
+      "step": 4801
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.730630702598074,
+      "learning_rate": 6.932662916806516e-06,
+      "loss": 0.9613,
+      "step": 4802
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.070507776187825,
+      "learning_rate": 6.931441959684414e-06,
+      "loss": 0.7724,
+      "step": 4803
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.472275772354918,
+      "learning_rate": 6.930220867176633e-06,
+      "loss": 0.8967,
+      "step": 4804
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.17464067657258,
+      "learning_rate": 6.928999639368773e-06,
+      "loss": 1.3153,
+      "step": 4805
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.7710448689759484,
+      "learning_rate": 6.927778276346431e-06,
+      "loss": 0.7007,
+      "step": 4806
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.7912421280625472,
+      "learning_rate": 6.926556778195224e-06,
+      "loss": 0.7555,
+      "step": 4807
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.6303058642059765,
+      "learning_rate": 6.925335145000769e-06,
+      "loss": 0.4378,
+      "step": 4808
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.552899346086931,
+      "learning_rate": 6.9241133768487005e-06,
+      "loss": 0.8362,
+      "step": 4809
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.691047824294737,
+      "learning_rate": 6.922891473824655e-06,
+      "loss": 0.8027,
+      "step": 4810
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.868971910272877,
+      "learning_rate": 6.92166943601429e-06,
+      "loss": 1.427,
+      "step": 4811
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.197455573046435,
+      "learning_rate": 6.9204472635032586e-06,
+      "loss": 0.8535,
+      "step": 4812
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.943066512717961,
+      "learning_rate": 6.9192249563772304e-06,
+      "loss": 0.7115,
+      "step": 4813
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.7013442623838446,
+      "learning_rate": 6.918002514721887e-06,
+      "loss": 0.6296,
+      "step": 4814
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.3205044878653345,
+      "learning_rate": 6.916779938622916e-06,
+      "loss": 0.6106,
+      "step": 4815
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.576051596597386,
+      "learning_rate": 6.9155572281660114e-06,
+      "loss": 0.4611,
+      "step": 4816
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.3346066642862535,
+      "learning_rate": 6.914334383436882e-06,
+      "loss": 1.1357,
+      "step": 4817
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.324620550091956,
+      "learning_rate": 6.913111404521242e-06,
+      "loss": 0.8622,
+      "step": 4818
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.50650194741537,
+      "learning_rate": 6.9118882915048204e-06,
+      "loss": 1.2789,
+      "step": 4819
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.085677384828208,
+      "learning_rate": 6.9106650444733495e-06,
+      "loss": 0.3836,
+      "step": 4820
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.459250967794102,
+      "learning_rate": 6.909441663512574e-06,
+      "loss": 0.9297,
+      "step": 4821
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.723734873755664,
+      "learning_rate": 6.908218148708248e-06,
+      "loss": 1.0447,
+      "step": 4822
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.082525983452769,
+      "learning_rate": 6.906994500146134e-06,
+      "loss": 0.7147,
+      "step": 4823
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.72503396605867,
+      "learning_rate": 6.905770717912006e-06,
+      "loss": 1.0701,
+      "step": 4824
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.2688936455944555,
+      "learning_rate": 6.904546802091644e-06,
+      "loss": 1.1725,
+      "step": 4825
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.991151581247189,
+      "learning_rate": 6.903322752770839e-06,
+      "loss": 0.8043,
+      "step": 4826
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.291152192141413,
+      "learning_rate": 6.9020985700353925e-06,
+      "loss": 1.0158,
+      "step": 4827
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.054654981801179,
+      "learning_rate": 6.900874253971115e-06,
+      "loss": 0.7262,
+      "step": 4828
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.352789982756976,
+      "learning_rate": 6.899649804663824e-06,
+      "loss": 0.4879,
+      "step": 4829
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.69679378842051,
+      "learning_rate": 6.898425222199349e-06,
+      "loss": 0.7028,
+      "step": 4830
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.316293810375166,
+      "learning_rate": 6.897200506663531e-06,
+      "loss": 1.0383,
+      "step": 4831
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.9655308923279713,
+      "learning_rate": 6.89597565814221e-06,
+      "loss": 0.7912,
+      "step": 4832
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.3248114459066294,
+      "learning_rate": 6.894750676721248e-06,
+      "loss": 0.6337,
+      "step": 4833
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.877496016078539,
+      "learning_rate": 6.893525562486511e-06,
+      "loss": 1.5636,
+      "step": 4834
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.0601769129749345,
+      "learning_rate": 6.89230031552387e-06,
+      "loss": 0.6569,
+      "step": 4835
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.5687417670290493,
+      "learning_rate": 6.891074935919213e-06,
+      "loss": 0.9939,
+      "step": 4836
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.141206013039817,
+      "learning_rate": 6.889849423758435e-06,
+      "loss": 1.0678,
+      "step": 4837
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.782425765103793,
+      "learning_rate": 6.888623779127436e-06,
+      "loss": 0.8534,
+      "step": 4838
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.249219968307424,
+      "learning_rate": 6.887398002112129e-06,
+      "loss": 1.0996,
+      "step": 4839
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.089008116773115,
+      "learning_rate": 6.886172092798436e-06,
+      "loss": 1.0721,
+      "step": 4840
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.914132826221444,
+      "learning_rate": 6.8849460512722874e-06,
+      "loss": 0.4047,
+      "step": 4841
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2686723684403072,
+      "learning_rate": 6.883719877619625e-06,
+      "loss": 0.2041,
+      "step": 4842
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.565081196331538,
+      "learning_rate": 6.882493571926396e-06,
+      "loss": 1.047,
+      "step": 4843
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.948217418062479,
+      "learning_rate": 6.881267134278562e-06,
+      "loss": 0.5319,
+      "step": 4844
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.177452726225325,
+      "learning_rate": 6.880040564762089e-06,
+      "loss": 1.0535,
+      "step": 4845
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.215548927000322,
+      "learning_rate": 6.878813863462953e-06,
+      "loss": 0.9454,
+      "step": 4846
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.090912258099464,
+      "learning_rate": 6.877587030467142e-06,
+      "loss": 0.7074,
+      "step": 4847
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.7783187665986246,
+      "learning_rate": 6.876360065860653e-06,
+      "loss": 0.7078,
+      "step": 4848
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.5361133983410573,
+      "learning_rate": 6.875132969729488e-06,
+      "loss": 0.7058,
+      "step": 4849
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.377776290693036,
+      "learning_rate": 6.873905742159661e-06,
+      "loss": 0.8566,
+      "step": 4850
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.8147401543582333,
+      "learning_rate": 6.872678383237199e-06,
+      "loss": 0.5288,
+      "step": 4851
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.080223092160162,
+      "learning_rate": 6.871450893048132e-06,
+      "loss": 1.1785,
+      "step": 4852
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.8363511837021016,
+      "learning_rate": 6.8702232716785015e-06,
+      "loss": 1.0499,
+      "step": 4853
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.670029911106436,
+      "learning_rate": 6.868995519214359e-06,
+      "loss": 1.0351,
+      "step": 4854
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.94734432167073,
+      "learning_rate": 6.867767635741765e-06,
+      "loss": 1.0276,
+      "step": 4855
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.5924768730461567,
+      "learning_rate": 6.866539621346786e-06,
+      "loss": 0.4919,
+      "step": 4856
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6637855824823422,
+      "learning_rate": 6.865311476115506e-06,
+      "loss": 0.3427,
+      "step": 4857
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.41025917393053,
+      "learning_rate": 6.864083200134007e-06,
+      "loss": 0.7974,
+      "step": 4858
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.333879761950064,
+      "learning_rate": 6.862854793488388e-06,
+      "loss": 0.4166,
+      "step": 4859
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.7012765849009965,
+      "learning_rate": 6.861626256264757e-06,
+      "loss": 0.5799,
+      "step": 4860
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.0504789769281935,
+      "learning_rate": 6.860397588549227e-06,
+      "loss": 1.1706,
+      "step": 4861
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.9441842501381243,
+      "learning_rate": 6.859168790427921e-06,
+      "loss": 0.3816,
+      "step": 4862
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.2684089653925614,
+      "learning_rate": 6.857939861986976e-06,
+      "loss": 0.7132,
+      "step": 4863
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.049869237519961,
+      "learning_rate": 6.856710803312531e-06,
+      "loss": 0.99,
+      "step": 4864
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.98686435202232,
+      "learning_rate": 6.85548161449074e-06,
+      "loss": 0.5555,
+      "step": 4865
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.66733054774265,
+      "learning_rate": 6.854252295607761e-06,
+      "loss": 0.9083,
+      "step": 4866
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.0811161745650453,
+      "learning_rate": 6.8530228467497685e-06,
+      "loss": 0.6796,
+      "step": 4867
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.4239574539894875,
+      "learning_rate": 6.8517932680029374e-06,
+      "loss": 0.9727,
+      "step": 4868
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.419282291306946,
+      "learning_rate": 6.850563559453458e-06,
+      "loss": 0.8212,
+      "step": 4869
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.033435941073197,
+      "learning_rate": 6.849333721187525e-06,
+      "loss": 0.8775,
+      "step": 4870
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.107942841853004,
+      "learning_rate": 6.848103753291349e-06,
+      "loss": 1.1675,
+      "step": 4871
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.517224443664284,
+      "learning_rate": 6.8468736558511405e-06,
+      "loss": 1.021,
+      "step": 4872
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.4227119961292236,
+      "learning_rate": 6.845643428953127e-06,
+      "loss": 0.9713,
+      "step": 4873
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.666843444544057,
+      "learning_rate": 6.844413072683542e-06,
+      "loss": 0.7418,
+      "step": 4874
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.315576569835842,
+      "learning_rate": 6.8431825871286275e-06,
+      "loss": 0.7384,
+      "step": 4875
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.1521853040591217,
+      "learning_rate": 6.841951972374636e-06,
+      "loss": 0.5464,
+      "step": 4876
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.3331402848579645,
+      "learning_rate": 6.840721228507826e-06,
+      "loss": 0.9008,
+      "step": 4877
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.730462628966787,
+      "learning_rate": 6.839490355614468e-06,
+      "loss": 1.1883,
+      "step": 4878
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.152897251668137,
+      "learning_rate": 6.838259353780843e-06,
+      "loss": 0.6253,
+      "step": 4879
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4914931988995255,
+      "learning_rate": 6.8370282230932375e-06,
+      "loss": 0.2271,
+      "step": 4880
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.220985605585712,
+      "learning_rate": 6.835796963637947e-06,
+      "loss": 0.5998,
+      "step": 4881
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.224144863216822,
+      "learning_rate": 6.83456557550128e-06,
+      "loss": 1.1766,
+      "step": 4882
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.6436117844737055,
+      "learning_rate": 6.83333405876955e-06,
+      "loss": 1.1108,
+      "step": 4883
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.825779692682705,
+      "learning_rate": 6.832102413529082e-06,
+      "loss": 0.8362,
+      "step": 4884
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.0609627060824423,
+      "learning_rate": 6.830870639866207e-06,
+      "loss": 0.5508,
+      "step": 4885
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.765868409693761,
+      "learning_rate": 6.829638737867268e-06,
+      "loss": 1.0675,
+      "step": 4886
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8755726868889399,
+      "learning_rate": 6.828406707618616e-06,
+      "loss": 0.1566,
+      "step": 4887
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.431109974323921,
+      "learning_rate": 6.827174549206612e-06,
+      "loss": 0.9188,
+      "step": 4888
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.6741186744873215,
+      "learning_rate": 6.825942262717623e-06,
+      "loss": 1.0093,
+      "step": 4889
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.414868717211905,
+      "learning_rate": 6.824709848238028e-06,
+      "loss": 0.3435,
+      "step": 4890
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.4953640571943225,
+      "learning_rate": 6.823477305854215e-06,
+      "loss": 0.7189,
+      "step": 4891
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.071480562210039,
+      "learning_rate": 6.822244635652579e-06,
+      "loss": 0.5268,
+      "step": 4892
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.9820593782485667,
+      "learning_rate": 6.821011837719522e-06,
+      "loss": 0.8389,
+      "step": 4893
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.7425034378072586,
+      "learning_rate": 6.819778912141461e-06,
+      "loss": 0.7943,
+      "step": 4894
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5328573063544362,
+      "learning_rate": 6.818545859004819e-06,
+      "loss": 0.1973,
+      "step": 4895
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.588916469323275,
+      "learning_rate": 6.817312678396026e-06,
+      "loss": 1.0831,
+      "step": 4896
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.555749972340232,
+      "learning_rate": 6.8160793704015234e-06,
+      "loss": 0.6709,
+      "step": 4897
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.8676428553444286,
+      "learning_rate": 6.81484593510776e-06,
+      "loss": 0.7756,
+      "step": 4898
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.780370674790176,
+      "learning_rate": 6.813612372601196e-06,
+      "loss": 0.5553,
+      "step": 4899
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.940664038442654,
+      "learning_rate": 6.812378682968297e-06,
+      "loss": 0.9143,
+      "step": 4900
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.548389415177721,
+      "learning_rate": 6.811144866295541e-06,
+      "loss": 0.7957,
+      "step": 4901
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.0121381310770756,
+      "learning_rate": 6.8099109226694095e-06,
+      "loss": 0.6322,
+      "step": 4902
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.40491945442787,
+      "learning_rate": 6.8086768521764006e-06,
+      "loss": 1.3037,
+      "step": 4903
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.4569409355456053,
+      "learning_rate": 6.807442654903015e-06,
+      "loss": 0.7545,
+      "step": 4904
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.398706886637341,
+      "learning_rate": 6.806208330935766e-06,
+      "loss": 1.2965,
+      "step": 4905
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.009674249846179,
+      "learning_rate": 6.804973880361172e-06,
+      "loss": 0.4626,
+      "step": 4906
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.770078543481889,
+      "learning_rate": 6.8037393032657665e-06,
+      "loss": 1.2375,
+      "step": 4907
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.9386078500281094,
+      "learning_rate": 6.802504599736085e-06,
+      "loss": 1.0183,
+      "step": 4908
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.705145761727528,
+      "learning_rate": 6.801269769858676e-06,
+      "loss": 0.773,
+      "step": 4909
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.7872201645622794,
+      "learning_rate": 6.800034813720093e-06,
+      "loss": 0.5834,
+      "step": 4910
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5884509932720416,
+      "learning_rate": 6.798799731406906e-06,
+      "loss": 0.222,
+      "step": 4911
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.9759992294640036,
+      "learning_rate": 6.797564523005685e-06,
+      "loss": 0.691,
+      "step": 4912
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.130717058700437,
+      "learning_rate": 6.796329188603015e-06,
+      "loss": 1.0033,
+      "step": 4913
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.3091944709110024,
+      "learning_rate": 6.795093728285485e-06,
+      "loss": 0.5313,
+      "step": 4914
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.183029532855425,
+      "learning_rate": 6.793858142139698e-06,
+      "loss": 0.9509,
+      "step": 4915
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.924495972395091,
+      "learning_rate": 6.792622430252263e-06,
+      "loss": 0.9398,
+      "step": 4916
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.319719235075423,
+      "learning_rate": 6.791386592709795e-06,
+      "loss": 0.906,
+      "step": 4917
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.862168877419175,
+      "learning_rate": 6.790150629598924e-06,
+      "loss": 0.8313,
+      "step": 4918
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.5343487433565732,
+      "learning_rate": 6.788914541006284e-06,
+      "loss": 0.8099,
+      "step": 4919
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 6.140203940052827,
+      "learning_rate": 6.787678327018521e-06,
+      "loss": 1.2256,
+      "step": 4920
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.050709535933509,
+      "learning_rate": 6.786441987722288e-06,
+      "loss": 0.8507,
+      "step": 4921
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.55419400867438,
+      "learning_rate": 6.785205523204244e-06,
+      "loss": 0.4897,
+      "step": 4922
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.2000464736683516,
+      "learning_rate": 6.783968933551064e-06,
+      "loss": 0.3428,
+      "step": 4923
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.170422918668265,
+      "learning_rate": 6.782732218849425e-06,
+      "loss": 1.2579,
+      "step": 4924
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.0477546435397898,
+      "learning_rate": 6.781495379186016e-06,
+      "loss": 0.7338,
+      "step": 4925
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.076146588457267,
+      "learning_rate": 6.780258414647534e-06,
+      "loss": 0.9534,
+      "step": 4926
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.711725811895843,
+      "learning_rate": 6.779021325320684e-06,
+      "loss": 0.6164,
+      "step": 4927
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.6802883358122496,
+      "learning_rate": 6.7777841112921825e-06,
+      "loss": 0.3153,
+      "step": 4928
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.8613066773022577,
+      "learning_rate": 6.776546772648751e-06,
+      "loss": 0.7101,
+      "step": 4929
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.823045677823471,
+      "learning_rate": 6.775309309477123e-06,
+      "loss": 0.401,
+      "step": 4930
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.2928797004917074,
+      "learning_rate": 6.7740717218640374e-06,
+      "loss": 0.6771,
+      "step": 4931
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.8684327129382496,
+      "learning_rate": 6.772834009896248e-06,
+      "loss": 0.9103,
+      "step": 4932
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.648521213162987,
+      "learning_rate": 6.771596173660506e-06,
+      "loss": 0.7071,
+      "step": 4933
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.804648349814525,
+      "learning_rate": 6.770358213243584e-06,
+      "loss": 1.1199,
+      "step": 4934
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.1312186278869145,
+      "learning_rate": 6.769120128732256e-06,
+      "loss": 0.3414,
+      "step": 4935
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.662486515215794,
+      "learning_rate": 6.7678819202133054e-06,
+      "loss": 0.2188,
+      "step": 4936
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.5205019556917874,
+      "learning_rate": 6.766643587773527e-06,
+      "loss": 0.5434,
+      "step": 4937
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.6374269679557463,
+      "learning_rate": 6.76540513149972e-06,
+      "loss": 0.878,
+      "step": 4938
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.7359192378862724,
+      "learning_rate": 6.764166551478699e-06,
+      "loss": 0.842,
+      "step": 4939
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.778568949153722,
+      "learning_rate": 6.762927847797279e-06,
+      "loss": 1.1776,
+      "step": 4940
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.048195079758534,
+      "learning_rate": 6.761689020542288e-06,
+      "loss": 0.6763,
+      "step": 4941
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.911603612677543,
+      "learning_rate": 6.760450069800565e-06,
+      "loss": 0.4051,
+      "step": 4942
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.2709982715047503,
+      "learning_rate": 6.759210995658953e-06,
+      "loss": 1.0407,
+      "step": 4943
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.101935661507507,
+      "learning_rate": 6.757971798204307e-06,
+      "loss": 0.6369,
+      "step": 4944
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.350693122627702,
+      "learning_rate": 6.756732477523489e-06,
+      "loss": 0.9884,
+      "step": 4945
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.426764541479174,
+      "learning_rate": 6.755493033703367e-06,
+      "loss": 0.6377,
+      "step": 4946
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.511797157726579,
+      "learning_rate": 6.754253466830827e-06,
+      "loss": 1.0663,
+      "step": 4947
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.986812080177097,
+      "learning_rate": 6.753013776992752e-06,
+      "loss": 1.2665,
+      "step": 4948
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.622379462419554,
+      "learning_rate": 6.751773964276039e-06,
+      "loss": 0.523,
+      "step": 4949
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.52878001591544,
+      "learning_rate": 6.750534028767596e-06,
+      "loss": 0.6513,
+      "step": 4950
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.057689733164265,
+      "learning_rate": 6.7492939705543355e-06,
+      "loss": 1.0922,
+      "step": 4951
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.326429923094338,
+      "learning_rate": 6.748053789723181e-06,
+      "loss": 1.0292,
+      "step": 4952
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.5933440981978935,
+      "learning_rate": 6.746813486361063e-06,
+      "loss": 0.7417,
+      "step": 4953
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.0535524993846352,
+      "learning_rate": 6.745573060554922e-06,
+      "loss": 0.6412,
+      "step": 4954
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.09612405979385,
+      "learning_rate": 6.744332512391707e-06,
+      "loss": 1.0326,
+      "step": 4955
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.9193694002324544,
+      "learning_rate": 6.743091841958373e-06,
+      "loss": 0.8915,
+      "step": 4956
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.4841669402172886,
+      "learning_rate": 6.741851049341888e-06,
+      "loss": 0.7924,
+      "step": 4957
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.163209288640059,
+      "learning_rate": 6.740610134629224e-06,
+      "loss": 1.0377,
+      "step": 4958
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.709614902603784,
+      "learning_rate": 6.739369097907365e-06,
+      "loss": 0.6028,
+      "step": 4959
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.164803894961905,
+      "learning_rate": 6.7381279392633025e-06,
+      "loss": 0.6679,
+      "step": 4960
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.2438255695122615,
+      "learning_rate": 6.736886658784034e-06,
+      "loss": 0.5955,
+      "step": 4961
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.618511131153098,
+      "learning_rate": 6.735645256556572e-06,
+      "loss": 1.2125,
+      "step": 4962
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.0953174041968565,
+      "learning_rate": 6.734403732667931e-06,
+      "loss": 0.6487,
+      "step": 4963
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.3826746576217044,
+      "learning_rate": 6.733162087205135e-06,
+      "loss": 0.7166,
+      "step": 4964
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.7470232526694387,
+      "learning_rate": 6.731920320255221e-06,
+      "loss": 0.7716,
+      "step": 4965
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.0457182226061286,
+      "learning_rate": 6.730678431905228e-06,
+      "loss": 0.7626,
+      "step": 4966
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.626770885537802,
+      "learning_rate": 6.7294364222422104e-06,
+      "loss": 1.2037,
+      "step": 4967
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.745128912279462,
+      "learning_rate": 6.728194291353226e-06,
+      "loss": 1.2067,
+      "step": 4968
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.6009434566161396,
+      "learning_rate": 6.7269520393253414e-06,
+      "loss": 0.3551,
+      "step": 4969
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.0433053176479525,
+      "learning_rate": 6.725709666245637e-06,
+      "loss": 0.8137,
+      "step": 4970
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.612255664240316,
+      "learning_rate": 6.7244671722011946e-06,
+      "loss": 0.5242,
+      "step": 4971
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.227490852490809,
+      "learning_rate": 6.723224557279107e-06,
+      "loss": 1.0358,
+      "step": 4972
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.9207008712420093,
+      "learning_rate": 6.721981821566476e-06,
+      "loss": 0.8894,
+      "step": 4973
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.6462179783860873,
+      "learning_rate": 6.7207389651504175e-06,
+      "loss": 1.1477,
+      "step": 4974
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.202301954200226,
+      "learning_rate": 6.719495988118043e-06,
+      "loss": 0.7103,
+      "step": 4975
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.6300639065375386,
+      "learning_rate": 6.718252890556485e-06,
+      "loss": 0.5941,
+      "step": 4976
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.548896662632624,
+      "learning_rate": 6.717009672552877e-06,
+      "loss": 0.6483,
+      "step": 4977
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.513029543800947,
+      "learning_rate": 6.715766334194362e-06,
+      "loss": 0.7638,
+      "step": 4978
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.354263444583725,
+      "learning_rate": 6.714522875568095e-06,
+      "loss": 0.8984,
+      "step": 4979
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.937833410720314,
+      "learning_rate": 6.713279296761237e-06,
+      "loss": 0.4593,
+      "step": 4980
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.705053803134201,
+      "learning_rate": 6.712035597860955e-06,
+      "loss": 0.6847,
+      "step": 4981
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.107071080317969,
+      "learning_rate": 6.710791778954429e-06,
+      "loss": 1.013,
+      "step": 4982
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.549944846724847,
+      "learning_rate": 6.709547840128844e-06,
+      "loss": 0.8946,
+      "step": 4983
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 9.743906409309483,
+      "learning_rate": 6.708303781471396e-06,
+      "loss": 0.6407,
+      "step": 4984
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.9118164028408535,
+      "learning_rate": 6.707059603069288e-06,
+      "loss": 0.947,
+      "step": 4985
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.0791859941056,
+      "learning_rate": 6.705815305009731e-06,
+      "loss": 0.6561,
+      "step": 4986
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.574974857317388,
+      "learning_rate": 6.7045708873799435e-06,
+      "loss": 0.6531,
+      "step": 4987
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.563827132306292,
+      "learning_rate": 6.703326350267157e-06,
+      "loss": 1.0643,
+      "step": 4988
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.784500592588589,
+      "learning_rate": 6.7020816937586046e-06,
+      "loss": 0.7969,
+      "step": 4989
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.9799773996493015,
+      "learning_rate": 6.7008369179415324e-06,
+      "loss": 0.7725,
+      "step": 4990
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.510790901955249,
+      "learning_rate": 6.699592022903197e-06,
+      "loss": 0.9653,
+      "step": 4991
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.180373643015554,
+      "learning_rate": 6.698347008730854e-06,
+      "loss": 0.676,
+      "step": 4992
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.9645250432962915,
+      "learning_rate": 6.697101875511779e-06,
+      "loss": 0.6309,
+      "step": 4993
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.2167881441671975,
+      "learning_rate": 6.695856623333249e-06,
+      "loss": 0.5562,
+      "step": 4994
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7854948379887672,
+      "learning_rate": 6.694611252282549e-06,
+      "loss": 0.3688,
+      "step": 4995
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.119302845105482,
+      "learning_rate": 6.693365762446975e-06,
+      "loss": 1.0407,
+      "step": 4996
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.19548604602088,
+      "learning_rate": 6.692120153913831e-06,
+      "loss": 0.5511,
+      "step": 4997
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.9396715204049366,
+      "learning_rate": 6.690874426770428e-06,
+      "loss": 0.7435,
+      "step": 4998
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.067757458057823,
+      "learning_rate": 6.6896285811040865e-06,
+      "loss": 0.8161,
+      "step": 4999
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.027032846338074,
+      "learning_rate": 6.688382617002135e-06,
+      "loss": 1.0669,
+      "step": 5000
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.804959960690094,
+      "learning_rate": 6.687136534551909e-06,
+      "loss": 1.174,
+      "step": 5001
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.4409453582294325,
+      "learning_rate": 6.685890333840757e-06,
+      "loss": 0.5652,
+      "step": 5002
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.6361198348967276,
+      "learning_rate": 6.6846440149560276e-06,
+      "loss": 0.7007,
+      "step": 5003
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.93626377499621,
+      "learning_rate": 6.683397577985084e-06,
+      "loss": 0.9234,
+      "step": 5004
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.49474660728893,
+      "learning_rate": 6.6821510230152975e-06,
+      "loss": 0.4029,
+      "step": 5005
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.615337291292612,
+      "learning_rate": 6.680904350134044e-06,
+      "loss": 0.5586,
+      "step": 5006
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4573510011122839,
+      "learning_rate": 6.679657559428712e-06,
+      "loss": 0.4093,
+      "step": 5007
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.085463953356308,
+      "learning_rate": 6.678410650986694e-06,
+      "loss": 0.7689,
+      "step": 5008
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.620326708654164,
+      "learning_rate": 6.677163624895393e-06,
+      "loss": 0.4192,
+      "step": 5009
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.145993087796699,
+      "learning_rate": 6.6759164812422225e-06,
+      "loss": 0.6969,
+      "step": 5010
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.507549152610467,
+      "learning_rate": 6.674669220114601e-06,
+      "loss": 1.1422,
+      "step": 5011
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.4986053398734898,
+      "learning_rate": 6.673421841599954e-06,
+      "loss": 0.6505,
+      "step": 5012
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.8082322958596393,
+      "learning_rate": 6.672174345785718e-06,
+      "loss": 0.3428,
+      "step": 5013
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.544467121237856,
+      "learning_rate": 6.6709267327593396e-06,
+      "loss": 0.625,
+      "step": 5014
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.9603559962796795,
+      "learning_rate": 6.669679002608267e-06,
+      "loss": 0.6185,
+      "step": 5015
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.217658841777538,
+      "learning_rate": 6.668431155419963e-06,
+      "loss": 0.9385,
+      "step": 5016
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.1587212821777464,
+      "learning_rate": 6.6671831912818985e-06,
+      "loss": 0.5103,
+      "step": 5017
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0508066768483084,
+      "learning_rate": 6.6659351102815475e-06,
+      "loss": 0.1753,
+      "step": 5018
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.1266872495955336,
+      "learning_rate": 6.664686912506393e-06,
+      "loss": 0.3366,
+      "step": 5019
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.31249644910684,
+      "learning_rate": 6.663438598043932e-06,
+      "loss": 0.6046,
+      "step": 5020
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.1792768803257747,
+      "learning_rate": 6.662190166981665e-06,
+      "loss": 0.8227,
+      "step": 5021
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.0539249160343376,
+      "learning_rate": 6.6609416194071e-06,
+      "loss": 0.3148,
+      "step": 5022
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.899219170184907,
+      "learning_rate": 6.659692955407757e-06,
+      "loss": 0.6692,
+      "step": 5023
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.542525499890373,
+      "learning_rate": 6.65844417507116e-06,
+      "loss": 0.9865,
+      "step": 5024
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.938901666456711,
+      "learning_rate": 6.657195278484845e-06,
+      "loss": 0.8797,
+      "step": 5025
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.461312687276681,
+      "learning_rate": 6.6559462657363525e-06,
+      "loss": 0.8987,
+      "step": 5026
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.123690030958761,
+      "learning_rate": 6.654697136913233e-06,
+      "loss": 0.9183,
+      "step": 5027
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.4249287381361073,
+      "learning_rate": 6.653447892103047e-06,
+      "loss": 0.4914,
+      "step": 5028
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.428473580726565,
+      "learning_rate": 6.652198531393358e-06,
+      "loss": 0.8932,
+      "step": 5029
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.264837748812746,
+      "learning_rate": 6.650949054871742e-06,
+      "loss": 1.3844,
+      "step": 5030
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.8526595143421294,
+      "learning_rate": 6.649699462625784e-06,
+      "loss": 0.6511,
+      "step": 5031
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.886729359381155,
+      "learning_rate": 6.648449754743072e-06,
+      "loss": 0.951,
+      "step": 5032
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.10215164550763,
+      "learning_rate": 6.647199931311207e-06,
+      "loss": 0.7978,
+      "step": 5033
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.4323786311499314,
+      "learning_rate": 6.645949992417795e-06,
+      "loss": 0.9083,
+      "step": 5034
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.0846317512520756,
+      "learning_rate": 6.644699938150452e-06,
+      "loss": 0.333,
+      "step": 5035
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.03665806830898,
+      "learning_rate": 6.6434497685968e-06,
+      "loss": 0.7356,
+      "step": 5036
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.1071041268719446,
+      "learning_rate": 6.642199483844473e-06,
+      "loss": 0.4502,
+      "step": 5037
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.7463856308086165,
+      "learning_rate": 6.640949083981108e-06,
+      "loss": 0.9525,
+      "step": 5038
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.950570863817365,
+      "learning_rate": 6.639698569094353e-06,
+      "loss": 1.06,
+      "step": 5039
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.304358286996472,
+      "learning_rate": 6.638447939271866e-06,
+      "loss": 0.7415,
+      "step": 5040
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.653097435243491,
+      "learning_rate": 6.637197194601309e-06,
+      "loss": 1.3601,
+      "step": 5041
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.391416070719546,
+      "learning_rate": 6.635946335170352e-06,
+      "loss": 0.8694,
+      "step": 5042
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.807569406655368,
+      "learning_rate": 6.634695361066679e-06,
+      "loss": 0.643,
+      "step": 5043
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.5645027321889433,
+      "learning_rate": 6.633444272377974e-06,
+      "loss": 0.7314,
+      "step": 5044
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8905815390481095,
+      "learning_rate": 6.632193069191934e-06,
+      "loss": 0.4264,
+      "step": 5045
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.678692969410927,
+      "learning_rate": 6.630941751596264e-06,
+      "loss": 0.9051,
+      "step": 5046
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.0461687288099317,
+      "learning_rate": 6.629690319678674e-06,
+      "loss": 0.3569,
+      "step": 5047
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.4543454838653753,
+      "learning_rate": 6.6284387735268865e-06,
+      "loss": 0.8512,
+      "step": 5048
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.1412107825446887,
+      "learning_rate": 6.627187113228627e-06,
+      "loss": 0.7086,
+      "step": 5049
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.3819860329330944,
+      "learning_rate": 6.625935338871632e-06,
+      "loss": 0.4417,
+      "step": 5050
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.70029031761103,
+      "learning_rate": 6.624683450543647e-06,
+      "loss": 0.7293,
+      "step": 5051
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.5089283676811505,
+      "learning_rate": 6.623431448332421e-06,
+      "loss": 0.3025,
+      "step": 5052
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.9728192523556984,
+      "learning_rate": 6.622179332325718e-06,
+      "loss": 0.7548,
+      "step": 5053
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.8230702830268473,
+      "learning_rate": 6.620927102611302e-06,
+      "loss": 0.7466,
+      "step": 5054
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.773342740786333,
+      "learning_rate": 6.619674759276951e-06,
+      "loss": 1.38,
+      "step": 5055
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.359982031316425,
+      "learning_rate": 6.6184223024104474e-06,
+      "loss": 1.1946,
+      "step": 5056
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.312823363429766,
+      "learning_rate": 6.6171697320995855e-06,
+      "loss": 1.1666,
+      "step": 5057
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.862686956689697,
+      "learning_rate": 6.615917048432161e-06,
+      "loss": 1.3333,
+      "step": 5058
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.1330910039124475,
+      "learning_rate": 6.614664251495986e-06,
+      "loss": 1.6148,
+      "step": 5059
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.906527461127112,
+      "learning_rate": 6.613411341378872e-06,
+      "loss": 0.7501,
+      "step": 5060
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.627117957234848,
+      "learning_rate": 6.6121583181686466e-06,
+      "loss": 1.197,
+      "step": 5061
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.1981226308634,
+      "learning_rate": 6.610905181953138e-06,
+      "loss": 0.9891,
+      "step": 5062
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.7675032094973404,
+      "learning_rate": 6.609651932820187e-06,
+      "loss": 0.4652,
+      "step": 5063
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.3054489860527765,
+      "learning_rate": 6.608398570857642e-06,
+      "loss": 0.9609,
+      "step": 5064
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.036451784571845,
+      "learning_rate": 6.607145096153355e-06,
+      "loss": 0.8836,
+      "step": 5065
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.7342217509523232,
+      "learning_rate": 6.605891508795193e-06,
+      "loss": 0.4318,
+      "step": 5066
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.884013713134834,
+      "learning_rate": 6.604637808871023e-06,
+      "loss": 0.915,
+      "step": 5067
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.360546766458889,
+      "learning_rate": 6.603383996468727e-06,
+      "loss": 0.3974,
+      "step": 5068
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9703620324064468,
+      "learning_rate": 6.602130071676191e-06,
+      "loss": 0.4134,
+      "step": 5069
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.6625200237642317,
+      "learning_rate": 6.600876034581308e-06,
+      "loss": 0.9043,
+      "step": 5070
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.4910179839017763,
+      "learning_rate": 6.599621885271984e-06,
+      "loss": 0.6769,
+      "step": 5071
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.013895703767678,
+      "learning_rate": 6.5983676238361284e-06,
+      "loss": 1.139,
+      "step": 5072
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.950931574451492,
+      "learning_rate": 6.5971132503616554e-06,
+      "loss": 0.5599,
+      "step": 5073
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.736885945190558,
+      "learning_rate": 6.5958587649364955e-06,
+      "loss": 1.0086,
+      "step": 5074
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.896035336353438,
+      "learning_rate": 6.5946041676485815e-06,
+      "loss": 1.1428,
+      "step": 5075
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.677453534155014,
+      "learning_rate": 6.593349458585855e-06,
+      "loss": 0.864,
+      "step": 5076
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7658904975909975,
+      "learning_rate": 6.592094637836266e-06,
+      "loss": 0.4789,
+      "step": 5077
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.702326669703679,
+      "learning_rate": 6.5908397054877715e-06,
+      "loss": 0.8829,
+      "step": 5078
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.204164682234299,
+      "learning_rate": 6.589584661628338e-06,
+      "loss": 0.7511,
+      "step": 5079
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.0722343576460274,
+      "learning_rate": 6.588329506345936e-06,
+      "loss": 0.3892,
+      "step": 5080
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3100602640670984,
+      "learning_rate": 6.587074239728549e-06,
+      "loss": 0.2382,
+      "step": 5081
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.720024571628034,
+      "learning_rate": 6.585818861864164e-06,
+      "loss": 0.9731,
+      "step": 5082
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.2744365581092225,
+      "learning_rate": 6.584563372840779e-06,
+      "loss": 1.0198,
+      "step": 5083
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.324752110388792,
+      "learning_rate": 6.583307772746397e-06,
+      "loss": 0.5498,
+      "step": 5084
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.008820777817378,
+      "learning_rate": 6.582052061669032e-06,
+      "loss": 1.1406,
+      "step": 5085
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.971044782311768,
+      "learning_rate": 6.580796239696701e-06,
+      "loss": 1.0095,
+      "step": 5086
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.7736435523199208,
+      "learning_rate": 6.579540306917434e-06,
+      "loss": 0.9718,
+      "step": 5087
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.449587139629432,
+      "learning_rate": 6.578284263419266e-06,
+      "loss": 0.1805,
+      "step": 5088
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.2797921454680115,
+      "learning_rate": 6.5770281092902385e-06,
+      "loss": 0.7411,
+      "step": 5089
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.885074877905152,
+      "learning_rate": 6.575771844618405e-06,
+      "loss": 0.5535,
+      "step": 5090
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.8296362782063436,
+      "learning_rate": 6.574515469491823e-06,
+      "loss": 0.7637,
+      "step": 5091
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.861826826046879,
+      "learning_rate": 6.573258983998558e-06,
+      "loss": 0.612,
+      "step": 5092
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.555017757308554,
+      "learning_rate": 6.572002388226686e-06,
+      "loss": 1.2151,
+      "step": 5093
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.497326683154025,
+      "learning_rate": 6.570745682264288e-06,
+      "loss": 0.7634,
+      "step": 5094
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.187260057733963,
+      "learning_rate": 6.569488866199454e-06,
+      "loss": 0.5992,
+      "step": 5095
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.917652628497841,
+      "learning_rate": 6.568231940120279e-06,
+      "loss": 0.9032,
+      "step": 5096
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.302888393164354,
+      "learning_rate": 6.566974904114871e-06,
+      "loss": 0.9961,
+      "step": 5097
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9869652572565435,
+      "learning_rate": 6.565717758271342e-06,
+      "loss": 0.488,
+      "step": 5098
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.894664576934961,
+      "learning_rate": 6.5644605026778115e-06,
+      "loss": 0.9216,
+      "step": 5099
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 6.1909555246544485,
+      "learning_rate": 6.563203137422409e-06,
+      "loss": 1.1481,
+      "step": 5100
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.448891006913339,
+      "learning_rate": 6.561945662593268e-06,
+      "loss": 1.0079,
+      "step": 5101
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.68256971253895,
+      "learning_rate": 6.5606880782785365e-06,
+      "loss": 0.9523,
+      "step": 5102
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.2626027984745805,
+      "learning_rate": 6.559430384566361e-06,
+      "loss": 0.4791,
+      "step": 5103
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.2812898718664627,
+      "learning_rate": 6.558172581544904e-06,
+      "loss": 0.6137,
+      "step": 5104
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.9689861187037994,
+      "learning_rate": 6.5569146693023285e-06,
+      "loss": 0.4125,
+      "step": 5105
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.3895446517519527,
+      "learning_rate": 6.5556566479268105e-06,
+      "loss": 0.7106,
+      "step": 5106
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.13831278802137,
+      "learning_rate": 6.5543985175065315e-06,
+      "loss": 1.0889,
+      "step": 5107
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.307667915755631,
+      "learning_rate": 6.553140278129683e-06,
+      "loss": 0.4637,
+      "step": 5108
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.620203642781845,
+      "learning_rate": 6.551881929884458e-06,
+      "loss": 1.3451,
+      "step": 5109
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.818227667612198,
+      "learning_rate": 6.550623472859063e-06,
+      "loss": 0.4817,
+      "step": 5110
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.744101031433427,
+      "learning_rate": 6.549364907141713e-06,
+      "loss": 0.4434,
+      "step": 5111
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.6282621358430904,
+      "learning_rate": 6.5481062328206265e-06,
+      "loss": 0.5363,
+      "step": 5112
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.7443038737454217,
+      "learning_rate": 6.546847449984028e-06,
+      "loss": 0.8433,
+      "step": 5113
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.5481004467431294,
+      "learning_rate": 6.5455885587201574e-06,
+      "loss": 0.5598,
+      "step": 5114
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.827802691559211,
+      "learning_rate": 6.544329559117254e-06,
+      "loss": 0.896,
+      "step": 5115
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.65077598090812,
+      "learning_rate": 6.543070451263569e-06,
+      "loss": 0.4954,
+      "step": 5116
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.689678199470181,
+      "learning_rate": 6.5418112352473616e-06,
+      "loss": 0.6008,
+      "step": 5117
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.3606194456689513,
+      "learning_rate": 6.540551911156896e-06,
+      "loss": 0.4282,
+      "step": 5118
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.317531169578305,
+      "learning_rate": 6.5392924790804475e-06,
+      "loss": 0.7708,
+      "step": 5119
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.7304560986881934,
+      "learning_rate": 6.538032939106295e-06,
+      "loss": 0.6627,
+      "step": 5120
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.198971090333942,
+      "learning_rate": 6.536773291322726e-06,
+      "loss": 0.3177,
+      "step": 5121
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.217417961186106,
+      "learning_rate": 6.5355135358180365e-06,
+      "loss": 0.5996,
+      "step": 5122
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.115620532143598,
+      "learning_rate": 6.5342536726805325e-06,
+      "loss": 0.7306,
+      "step": 5123
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.446784294414177,
+      "learning_rate": 6.532993701998522e-06,
+      "loss": 0.4508,
+      "step": 5124
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.8874074345338547,
+      "learning_rate": 6.531733623860326e-06,
+      "loss": 0.4286,
+      "step": 5125
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.661011299093553,
+      "learning_rate": 6.5304734383542664e-06,
+      "loss": 1.1787,
+      "step": 5126
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.880231091993628,
+      "learning_rate": 6.5292131455686825e-06,
+      "loss": 0.6272,
+      "step": 5127
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.639270364858419,
+      "learning_rate": 6.527952745591911e-06,
+      "loss": 1.0348,
+      "step": 5128
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.2309889550281254,
+      "learning_rate": 6.526692238512301e-06,
+      "loss": 0.7215,
+      "step": 5129
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.240868644277337,
+      "learning_rate": 6.5254316244182096e-06,
+      "loss": 0.4796,
+      "step": 5130
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.479293937527797,
+      "learning_rate": 6.524170903398001e-06,
+      "loss": 0.9256,
+      "step": 5131
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.3407275596290247,
+      "learning_rate": 6.522910075540043e-06,
+      "loss": 0.9344,
+      "step": 5132
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.091374171071857,
+      "learning_rate": 6.5216491409327174e-06,
+      "loss": 0.3943,
+      "step": 5133
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.700049334269497,
+      "learning_rate": 6.5203880996644105e-06,
+      "loss": 1.0703,
+      "step": 5134
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.604011669234343,
+      "learning_rate": 6.519126951823516e-06,
+      "loss": 0.4824,
+      "step": 5135
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.522206485399803,
+      "learning_rate": 6.51786569749843e-06,
+      "loss": 0.4807,
+      "step": 5136
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.886140827984647,
+      "learning_rate": 6.516604336777565e-06,
+      "loss": 0.8842,
+      "step": 5137
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.020753694642666,
+      "learning_rate": 6.515342869749337e-06,
+      "loss": 0.9779,
+      "step": 5138
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.913893544134915,
+      "learning_rate": 6.5140812965021685e-06,
+      "loss": 0.6798,
+      "step": 5139
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.839134577631026,
+      "learning_rate": 6.512819617124491e-06,
+      "loss": 0.9442,
+      "step": 5140
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.5516407268408408,
+      "learning_rate": 6.511557831704741e-06,
+      "loss": 0.7877,
+      "step": 5141
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8731293715755902,
+      "learning_rate": 6.510295940331367e-06,
+      "loss": 0.1423,
+      "step": 5142
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.0190805563422805,
+      "learning_rate": 6.509033943092819e-06,
+      "loss": 0.8424,
+      "step": 5143
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.14674799879688,
+      "learning_rate": 6.50777184007756e-06,
+      "loss": 0.6919,
+      "step": 5144
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.113716663688054,
+      "learning_rate": 6.506509631374056e-06,
+      "loss": 0.419,
+      "step": 5145
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.937709065235011,
+      "learning_rate": 6.5052473170707844e-06,
+      "loss": 1.1737,
+      "step": 5146
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.277353913858537,
+      "learning_rate": 6.5039848972562246e-06,
+      "loss": 0.7087,
+      "step": 5147
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.3910715221425365,
+      "learning_rate": 6.50272237201887e-06,
+      "loss": 0.7598,
+      "step": 5148
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.939373388917265,
+      "learning_rate": 6.501459741447217e-06,
+      "loss": 0.8829,
+      "step": 5149
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.0479883517194137,
+      "learning_rate": 6.500197005629772e-06,
+      "loss": 0.5932,
+      "step": 5150
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.306912013821991,
+      "learning_rate": 6.498934164655044e-06,
+      "loss": 1.0467,
+      "step": 5151
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.8791463716466894,
+      "learning_rate": 6.4976712186115545e-06,
+      "loss": 0.5539,
+      "step": 5152
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.319894481396109,
+      "learning_rate": 6.49640816758783e-06,
+      "loss": 0.813,
+      "step": 5153
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.2850483337946716,
+      "learning_rate": 6.495145011672406e-06,
+      "loss": 0.8408,
+      "step": 5154
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.7827482983780927,
+      "learning_rate": 6.493881750953823e-06,
+      "loss": 0.9218,
+      "step": 5155
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.6812886117281316,
+      "learning_rate": 6.49261838552063e-06,
+      "loss": 0.7084,
+      "step": 5156
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.422633886344834,
+      "learning_rate": 6.491354915461387e-06,
+      "loss": 0.907,
+      "step": 5157
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.8343098758964165,
+      "learning_rate": 6.490091340864654e-06,
+      "loss": 1.0196,
+      "step": 5158
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.662496099515557,
+      "learning_rate": 6.488827661819002e-06,
+      "loss": 0.9483,
+      "step": 5159
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.168576346105122,
+      "learning_rate": 6.48756387841301e-06,
+      "loss": 1.0537,
+      "step": 5160
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.17476297436785,
+      "learning_rate": 6.486299990735263e-06,
+      "loss": 1.0399,
+      "step": 5161
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.3953438108264753,
+      "learning_rate": 6.485035998874356e-06,
+      "loss": 0.3847,
+      "step": 5162
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.156931565798611,
+      "learning_rate": 6.48377190291889e-06,
+      "loss": 0.9419,
+      "step": 5163
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.7204080529191734,
+      "learning_rate": 6.482507702957469e-06,
+      "loss": 0.804,
+      "step": 5164
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.306466886058456,
+      "learning_rate": 6.481243399078712e-06,
+      "loss": 0.8862,
+      "step": 5165
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9172894698382654,
+      "learning_rate": 6.479978991371239e-06,
+      "loss": 0.3973,
+      "step": 5166
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.6456024896313104,
+      "learning_rate": 6.478714479923677e-06,
+      "loss": 0.7912,
+      "step": 5167
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.752028880061259,
+      "learning_rate": 6.4774498648246675e-06,
+      "loss": 0.3674,
+      "step": 5168
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.2718369483666523,
+      "learning_rate": 6.4761851461628514e-06,
+      "loss": 0.5094,
+      "step": 5169
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.5861334535536775,
+      "learning_rate": 6.47492032402688e-06,
+      "loss": 0.8346,
+      "step": 5170
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.514568973146496,
+      "learning_rate": 6.473655398505414e-06,
+      "loss": 0.9248,
+      "step": 5171
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.151764429495408,
+      "learning_rate": 6.472390369687118e-06,
+      "loss": 0.8877,
+      "step": 5172
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.7227006795204702,
+      "learning_rate": 6.471125237660665e-06,
+      "loss": 0.532,
+      "step": 5173
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.548160455981264,
+      "learning_rate": 6.469860002514736e-06,
+      "loss": 0.7771,
+      "step": 5174
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.223658539674634,
+      "learning_rate": 6.468594664338016e-06,
+      "loss": 0.4935,
+      "step": 5175
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4522936109620626,
+      "learning_rate": 6.467329223219201e-06,
+      "loss": 0.2192,
+      "step": 5176
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.913587802323358,
+      "learning_rate": 6.4660636792469955e-06,
+      "loss": 1.3289,
+      "step": 5177
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.732764437814705,
+      "learning_rate": 6.464798032510104e-06,
+      "loss": 0.9273,
+      "step": 5178
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.9186169441628973,
+      "learning_rate": 6.4635322830972465e-06,
+      "loss": 0.9121,
+      "step": 5179
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.9273731689932845,
+      "learning_rate": 6.462266431097146e-06,
+      "loss": 0.7626,
+      "step": 5180
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.427628339205233,
+      "learning_rate": 6.461000476598532e-06,
+      "loss": 1.3493,
+      "step": 5181
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.0967838122151363,
+      "learning_rate": 6.459734419690143e-06,
+      "loss": 0.2514,
+      "step": 5182
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.88698206894186,
+      "learning_rate": 6.458468260460724e-06,
+      "loss": 0.4262,
+      "step": 5183
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.9352789215009603,
+      "learning_rate": 6.457201998999025e-06,
+      "loss": 1.0131,
+      "step": 5184
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.9065289517117976,
+      "learning_rate": 6.455935635393811e-06,
+      "loss": 0.7477,
+      "step": 5185
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.289868895267918,
+      "learning_rate": 6.454669169733843e-06,
+      "loss": 0.8929,
+      "step": 5186
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.8693499507499465,
+      "learning_rate": 6.4534026021078966e-06,
+      "loss": 0.5043,
+      "step": 5187
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.947055324633199,
+      "learning_rate": 6.452135932604755e-06,
+      "loss": 1.3338,
+      "step": 5188
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.990863485491423,
+      "learning_rate": 6.450869161313205e-06,
+      "loss": 1.1023,
+      "step": 5189
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.6732958963792015,
+      "learning_rate": 6.4496022883220376e-06,
+      "loss": 0.4853,
+      "step": 5190
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.5877289700187065,
+      "learning_rate": 6.448335313720061e-06,
+      "loss": 1.2412,
+      "step": 5191
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.922699463637487,
+      "learning_rate": 6.447068237596081e-06,
+      "loss": 0.7607,
+      "step": 5192
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.6481712138131246,
+      "learning_rate": 6.445801060038915e-06,
+      "loss": 0.5285,
+      "step": 5193
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.319125009354408,
+      "learning_rate": 6.444533781137387e-06,
+      "loss": 0.9255,
+      "step": 5194
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.0171710686961,
+      "learning_rate": 6.443266400980328e-06,
+      "loss": 0.8706,
+      "step": 5195
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.344644252269986,
+      "learning_rate": 6.441998919656575e-06,
+      "loss": 1.2445,
+      "step": 5196
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.909068690829836,
+      "learning_rate": 6.440731337254975e-06,
+      "loss": 0.7838,
+      "step": 5197
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7922868575018684,
+      "learning_rate": 6.439463653864376e-06,
+      "loss": 0.318,
+      "step": 5198
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.713543045431059,
+      "learning_rate": 6.43819586957364e-06,
+      "loss": 0.693,
+      "step": 5199
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.86145422753484,
+      "learning_rate": 6.436927984471634e-06,
+      "loss": 1.1019,
+      "step": 5200
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.9079994710944934,
+      "learning_rate": 6.435659998647228e-06,
+      "loss": 0.8679,
+      "step": 5201
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.172939734126654,
+      "learning_rate": 6.434391912189304e-06,
+      "loss": 0.6581,
+      "step": 5202
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.971402010277261,
+      "learning_rate": 6.433123725186752e-06,
+      "loss": 0.8514,
+      "step": 5203
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.570102971289309,
+      "learning_rate": 6.431855437728463e-06,
+      "loss": 1.2454,
+      "step": 5204
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.120223974832511,
+      "learning_rate": 6.430587049903336e-06,
+      "loss": 0.743,
+      "step": 5205
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.162178464204561,
+      "learning_rate": 6.429318561800286e-06,
+      "loss": 1.0354,
+      "step": 5206
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.501800227045963,
+      "learning_rate": 6.428049973508225e-06,
+      "loss": 0.7937,
+      "step": 5207
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.94871567949659,
+      "learning_rate": 6.426781285116075e-06,
+      "loss": 0.6766,
+      "step": 5208
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.466894583246494,
+      "learning_rate": 6.4255124967127665e-06,
+      "loss": 0.4047,
+      "step": 5209
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.43354768501383,
+      "learning_rate": 6.424243608387235e-06,
+      "loss": 0.734,
+      "step": 5210
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.646302110542977,
+      "learning_rate": 6.422974620228426e-06,
+      "loss": 1.5648,
+      "step": 5211
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.5900228242446808,
+      "learning_rate": 6.421705532325289e-06,
+      "loss": 0.6069,
+      "step": 5212
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.402590196811873,
+      "learning_rate": 6.420436344766781e-06,
+      "loss": 0.8138,
+      "step": 5213
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.417897697444654,
+      "learning_rate": 6.419167057641868e-06,
+      "loss": 0.6705,
+      "step": 5214
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.35920113079431,
+      "learning_rate": 6.417897671039519e-06,
+      "loss": 0.4583,
+      "step": 5215
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.834644706331093,
+      "learning_rate": 6.4166281850487135e-06,
+      "loss": 0.9184,
+      "step": 5216
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.201190354437537,
+      "learning_rate": 6.415358599758439e-06,
+      "loss": 0.6057,
+      "step": 5217
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.498817582268024,
+      "learning_rate": 6.414088915257686e-06,
+      "loss": 0.4364,
+      "step": 5218
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.101896399093929,
+      "learning_rate": 6.4128191316354525e-06,
+      "loss": 0.4627,
+      "step": 5219
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.7323505431124886,
+      "learning_rate": 6.411549248980748e-06,
+      "loss": 0.8131,
+      "step": 5220
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.0423803714767366,
+      "learning_rate": 6.410279267382585e-06,
+      "loss": 0.7822,
+      "step": 5221
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.115267896822398,
+      "learning_rate": 6.409009186929982e-06,
+      "loss": 0.4279,
+      "step": 5222
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.012951896962024,
+      "learning_rate": 6.407739007711969e-06,
+      "loss": 0.3176,
+      "step": 5223
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.608909053480998,
+      "learning_rate": 6.406468729817574e-06,
+      "loss": 0.5814,
+      "step": 5224
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.4885622809302306,
+      "learning_rate": 6.405198353335844e-06,
+      "loss": 0.7136,
+      "step": 5225
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.275506002331668,
+      "learning_rate": 6.403927878355825e-06,
+      "loss": 0.5962,
+      "step": 5226
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.760382784091925,
+      "learning_rate": 6.402657304966572e-06,
+      "loss": 1.0417,
+      "step": 5227
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.418888169136374,
+      "learning_rate": 6.401386633257146e-06,
+      "loss": 0.6734,
+      "step": 5228
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.440478521109242,
+      "learning_rate": 6.400115863316616e-06,
+      "loss": 0.4783,
+      "step": 5229
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.641165967006061,
+      "learning_rate": 6.398844995234057e-06,
+      "loss": 0.8536,
+      "step": 5230
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.7431611684166457,
+      "learning_rate": 6.397574029098552e-06,
+      "loss": 0.5524,
+      "step": 5231
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.040278877345919,
+      "learning_rate": 6.39630296499919e-06,
+      "loss": 0.9451,
+      "step": 5232
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.409153375979241,
+      "learning_rate": 6.3950318030250654e-06,
+      "loss": 0.6427,
+      "step": 5233
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.7850491565280184,
+      "learning_rate": 6.393760543265285e-06,
+      "loss": 0.8784,
+      "step": 5234
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.452661896158122,
+      "learning_rate": 6.392489185808954e-06,
+      "loss": 0.7516,
+      "step": 5235
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.916918257978019,
+      "learning_rate": 6.391217730745193e-06,
+      "loss": 0.9683,
+      "step": 5236
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.432841017110807,
+      "learning_rate": 6.3899461781631225e-06,
+      "loss": 1.1874,
+      "step": 5237
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.5810353135567894,
+      "learning_rate": 6.388674528151875e-06,
+      "loss": 0.8334,
+      "step": 5238
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.7601699256718675,
+      "learning_rate": 6.387402780800585e-06,
+      "loss": 0.7938,
+      "step": 5239
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.490327124441589,
+      "learning_rate": 6.386130936198399e-06,
+      "loss": 0.8184,
+      "step": 5240
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.657950700172878,
+      "learning_rate": 6.384858994434467e-06,
+      "loss": 0.7991,
+      "step": 5241
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.360273681567106,
+      "learning_rate": 6.383586955597945e-06,
+      "loss": 1.0199,
+      "step": 5242
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.221993123701766,
+      "learning_rate": 6.382314819778e-06,
+      "loss": 0.4936,
+      "step": 5243
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.0920500320981583,
+      "learning_rate": 6.381042587063803e-06,
+      "loss": 0.5727,
+      "step": 5244
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2347677618701909,
+      "learning_rate": 6.379770257544529e-06,
+      "loss": 0.1976,
+      "step": 5245
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.033352049988148,
+      "learning_rate": 6.378497831309367e-06,
+      "loss": 0.6842,
+      "step": 5246
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.4052660445723,
+      "learning_rate": 6.377225308447503e-06,
+      "loss": 0.8753,
+      "step": 5247
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.323082370180678,
+      "learning_rate": 6.375952689048141e-06,
+      "loss": 1.2976,
+      "step": 5248
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.142553012636817,
+      "learning_rate": 6.374679973200484e-06,
+      "loss": 0.3026,
+      "step": 5249
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.790216347839119,
+      "learning_rate": 6.373407160993742e-06,
+      "loss": 0.6913,
+      "step": 5250
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.9260276299107923,
+      "learning_rate": 6.372134252517136e-06,
+      "loss": 1.0458,
+      "step": 5251
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.4639919518325244,
+      "learning_rate": 6.370861247859891e-06,
+      "loss": 0.4821,
+      "step": 5252
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.3230537659127353,
+      "learning_rate": 6.369588147111236e-06,
+      "loss": 0.5219,
+      "step": 5253
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.368258655052682,
+      "learning_rate": 6.368314950360416e-06,
+      "loss": 0.8872,
+      "step": 5254
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.3266571921352304,
+      "learning_rate": 6.36704165769667e-06,
+      "loss": 0.3647,
+      "step": 5255
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.475615810880121,
+      "learning_rate": 6.365768269209254e-06,
+      "loss": 1.0257,
+      "step": 5256
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.926406979750642,
+      "learning_rate": 6.364494784987427e-06,
+      "loss": 1.1206,
+      "step": 5257
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.387279390308762,
+      "learning_rate": 6.363221205120452e-06,
+      "loss": 0.5091,
+      "step": 5258
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.051417370757732,
+      "learning_rate": 6.361947529697605e-06,
+      "loss": 1.0578,
+      "step": 5259
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.161788346493241,
+      "learning_rate": 6.360673758808163e-06,
+      "loss": 0.781,
+      "step": 5260
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.948150103760491,
+      "learning_rate": 6.359399892541412e-06,
+      "loss": 0.726,
+      "step": 5261
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.113461144271678,
+      "learning_rate": 6.358125930986645e-06,
+      "loss": 0.3269,
+      "step": 5262
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.04931175796063,
+      "learning_rate": 6.356851874233161e-06,
+      "loss": 0.5767,
+      "step": 5263
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.9176332686041966,
+      "learning_rate": 6.355577722370264e-06,
+      "loss": 0.7563,
+      "step": 5264
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.59725461401139,
+      "learning_rate": 6.354303475487269e-06,
+      "loss": 1.1339,
+      "step": 5265
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.889559779448481,
+      "learning_rate": 6.353029133673496e-06,
+      "loss": 0.9163,
+      "step": 5266
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.026137753654765,
+      "learning_rate": 6.351754697018269e-06,
+      "loss": 1.1465,
+      "step": 5267
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.1978137486717033,
+      "learning_rate": 6.3504801656109195e-06,
+      "loss": 0.6188,
+      "step": 5268
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.443129720076294,
+      "learning_rate": 6.349205539540786e-06,
+      "loss": 0.444,
+      "step": 5269
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.5323700976741597,
+      "learning_rate": 6.3479308188972175e-06,
+      "loss": 0.6203,
+      "step": 5270
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.506653631760839,
+      "learning_rate": 6.346656003769565e-06,
+      "loss": 0.8935,
+      "step": 5271
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.51929605166405,
+      "learning_rate": 6.345381094247188e-06,
+      "loss": 0.8745,
+      "step": 5272
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.7618715703681103,
+      "learning_rate": 6.34410609041945e-06,
+      "loss": 0.7381,
+      "step": 5273
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.9031600701897586,
+      "learning_rate": 6.342830992375725e-06,
+      "loss": 0.8362,
+      "step": 5274
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.069650157218168,
+      "learning_rate": 6.341555800205392e-06,
+      "loss": 1.0269,
+      "step": 5275
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.452049394631239,
+      "learning_rate": 6.340280513997835e-06,
+      "loss": 0.2514,
+      "step": 5276
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.893706571124117,
+      "learning_rate": 6.3390051338424485e-06,
+      "loss": 1.5367,
+      "step": 5277
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.659535833865813,
+      "learning_rate": 6.337729659828627e-06,
+      "loss": 0.9222,
+      "step": 5278
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.2514738051998213,
+      "learning_rate": 6.33645409204578e-06,
+      "loss": 0.7972,
+      "step": 5279
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.6843164318194446,
+      "learning_rate": 6.3351784305833175e-06,
+      "loss": 0.6152,
+      "step": 5280
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.5362471559949156,
+      "learning_rate": 6.333902675530657e-06,
+      "loss": 1.0854,
+      "step": 5281
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.556924473924575,
+      "learning_rate": 6.332626826977224e-06,
+      "loss": 1.0025,
+      "step": 5282
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.0381830710652,
+      "learning_rate": 6.33135088501245e-06,
+      "loss": 0.6271,
+      "step": 5283
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.457149041496543,
+      "learning_rate": 6.330074849725774e-06,
+      "loss": 0.7812,
+      "step": 5284
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.794346012715579,
+      "learning_rate": 6.328798721206638e-06,
+      "loss": 0.4939,
+      "step": 5285
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.158328439921532,
+      "learning_rate": 6.327522499544496e-06,
+      "loss": 0.4347,
+      "step": 5286
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.111621810371288,
+      "learning_rate": 6.3262461848288034e-06,
+      "loss": 1.3857,
+      "step": 5287
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.8505774832006874,
+      "learning_rate": 6.324969777149026e-06,
+      "loss": 0.6636,
+      "step": 5288
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.506604162373283,
+      "learning_rate": 6.323693276594632e-06,
+      "loss": 0.5305,
+      "step": 5289
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.8964055593354967,
+      "learning_rate": 6.322416683255103e-06,
+      "loss": 0.5667,
+      "step": 5290
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.76693745022548,
+      "learning_rate": 6.321139997219917e-06,
+      "loss": 0.5249,
+      "step": 5291
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.3431205458148905,
+      "learning_rate": 6.319863218578568e-06,
+      "loss": 0.6691,
+      "step": 5292
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.317627231571605,
+      "learning_rate": 6.31858634742055e-06,
+      "loss": 1.1668,
+      "step": 5293
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 6.154345697947195,
+      "learning_rate": 6.317309383835368e-06,
+      "loss": 1.2021,
+      "step": 5294
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.437607805714208,
+      "learning_rate": 6.316032327912532e-06,
+      "loss": 0.4886,
+      "step": 5295
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.6188452267712865,
+      "learning_rate": 6.314755179741556e-06,
+      "loss": 0.4352,
+      "step": 5296
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.2353476662810428,
+      "learning_rate": 6.313477939411965e-06,
+      "loss": 0.4746,
+      "step": 5297
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.243223467921312,
+      "learning_rate": 6.312200607013287e-06,
+      "loss": 1.0017,
+      "step": 5298
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.310191103900768,
+      "learning_rate": 6.310923182635056e-06,
+      "loss": 1.0909,
+      "step": 5299
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.4573373782516903,
+      "learning_rate": 6.309645666366816e-06,
+      "loss": 0.9911,
+      "step": 5300
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.5494617356408997,
+      "learning_rate": 6.308368058298114e-06,
+      "loss": 0.6491,
+      "step": 5301
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0475818745256136,
+      "learning_rate": 6.307090358518504e-06,
+      "loss": 0.1533,
+      "step": 5302
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.288289841775113,
+      "learning_rate": 6.30581256711755e-06,
+      "loss": 0.9734,
+      "step": 5303
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.3351673026973065,
+      "learning_rate": 6.304534684184816e-06,
+      "loss": 0.5136,
+      "step": 5304
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.0374178588832,
+      "learning_rate": 6.303256709809879e-06,
+      "loss": 0.8494,
+      "step": 5305
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5981949932303068,
+      "learning_rate": 6.301978644082321e-06,
+      "loss": 0.3312,
+      "step": 5306
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8521095385310853,
+      "learning_rate": 6.300700487091723e-06,
+      "loss": 0.3675,
+      "step": 5307
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.542418793742306,
+      "learning_rate": 6.299422238927683e-06,
+      "loss": 1.0892,
+      "step": 5308
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.363744970945015,
+      "learning_rate": 6.298143899679798e-06,
+      "loss": 0.6085,
+      "step": 5309
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.358168052292183,
+      "learning_rate": 6.296865469437675e-06,
+      "loss": 0.8132,
+      "step": 5310
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.9215527271968007,
+      "learning_rate": 6.295586948290928e-06,
+      "loss": 1.1532,
+      "step": 5311
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.222417314593268,
+      "learning_rate": 6.294308336329174e-06,
+      "loss": 0.7616,
+      "step": 5312
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.7126852976017393,
+      "learning_rate": 6.293029633642038e-06,
+      "loss": 0.5614,
+      "step": 5313
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.361404869047012,
+      "learning_rate": 6.291750840319152e-06,
+      "loss": 0.495,
+      "step": 5314
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.676970942660408,
+      "learning_rate": 6.2904719564501545e-06,
+      "loss": 0.5076,
+      "step": 5315
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.7089502354975656,
+      "learning_rate": 6.2891929821246875e-06,
+      "loss": 0.7755,
+      "step": 5316
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 6.006691414153541,
+      "learning_rate": 6.287913917432405e-06,
+      "loss": 1.0736,
+      "step": 5317
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.859693190142713,
+      "learning_rate": 6.286634762462961e-06,
+      "loss": 0.6757,
+      "step": 5318
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.372561771884962,
+      "learning_rate": 6.285355517306019e-06,
+      "loss": 0.9634,
+      "step": 5319
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.077137875843918,
+      "learning_rate": 6.2840761820512505e-06,
+      "loss": 1.1677,
+      "step": 5320
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1370738600649346,
+      "learning_rate": 6.282796756788328e-06,
+      "loss": 0.1414,
+      "step": 5321
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.220375852691493,
+      "learning_rate": 6.281517241606938e-06,
+      "loss": 0.4744,
+      "step": 5322
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.5808255674719076,
+      "learning_rate": 6.280237636596765e-06,
+      "loss": 0.5416,
+      "step": 5323
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.384436992824313,
+      "learning_rate": 6.278957941847506e-06,
+      "loss": 0.708,
+      "step": 5324
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.3301604915033787,
+      "learning_rate": 6.27767815744886e-06,
+      "loss": 0.8014,
+      "step": 5325
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.8885651689869096,
+      "learning_rate": 6.276398283490537e-06,
+      "loss": 0.6231,
+      "step": 5326
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.355467328726323,
+      "learning_rate": 6.275118320062248e-06,
+      "loss": 1.1165,
+      "step": 5327
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.197077206685935,
+      "learning_rate": 6.273838267253716e-06,
+      "loss": 1.0601,
+      "step": 5328
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.5896109313530795,
+      "learning_rate": 6.272558125154663e-06,
+      "loss": 1.3009,
+      "step": 5329
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.9650801372699567,
+      "learning_rate": 6.271277893854825e-06,
+      "loss": 0.6123,
+      "step": 5330
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.2756477223518687,
+      "learning_rate": 6.26999757344394e-06,
+      "loss": 0.5752,
+      "step": 5331
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.172137842304576,
+      "learning_rate": 6.268717164011751e-06,
+      "loss": 0.44,
+      "step": 5332
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.484492362826107,
+      "learning_rate": 6.2674366656480105e-06,
+      "loss": 1.3716,
+      "step": 5333
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.409549960962839,
+      "learning_rate": 6.266156078442476e-06,
+      "loss": 1.2073,
+      "step": 5334
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.4379919669246224,
+      "learning_rate": 6.264875402484909e-06,
+      "loss": 0.6082,
+      "step": 5335
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.4813981703057624,
+      "learning_rate": 6.263594637865081e-06,
+      "loss": 0.5548,
+      "step": 5336
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.174097987149481,
+      "learning_rate": 6.262313784672771e-06,
+      "loss": 0.7561,
+      "step": 5337
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.338098544074793,
+      "learning_rate": 6.261032842997756e-06,
+      "loss": 0.9866,
+      "step": 5338
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.5408474628918287,
+      "learning_rate": 6.259751812929829e-06,
+      "loss": 0.7631,
+      "step": 5339
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.8321054319008008,
+      "learning_rate": 6.25847069455878e-06,
+      "loss": 0.6125,
+      "step": 5340
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.8415935691356777,
+      "learning_rate": 6.257189487974414e-06,
+      "loss": 0.6153,
+      "step": 5341
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.96059793607359,
+      "learning_rate": 6.255908193266533e-06,
+      "loss": 0.7248,
+      "step": 5342
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.367845793763824,
+      "learning_rate": 6.254626810524956e-06,
+      "loss": 0.8438,
+      "step": 5343
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.9224084609222465,
+      "learning_rate": 6.2533453398395e-06,
+      "loss": 0.6552,
+      "step": 5344
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.990580969392164,
+      "learning_rate": 6.25206378129999e-06,
+      "loss": 0.9274,
+      "step": 5345
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.6908493507615985,
+      "learning_rate": 6.250782134996257e-06,
+      "loss": 0.6803,
+      "step": 5346
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.576204374110863,
+      "learning_rate": 6.24950040101814e-06,
+      "loss": 1.2582,
+      "step": 5347
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4265933546341147,
+      "learning_rate": 6.248218579455484e-06,
+      "loss": 0.2474,
+      "step": 5348
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.1413602211082585,
+      "learning_rate": 6.246936670398136e-06,
+      "loss": 0.9275,
+      "step": 5349
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.004104808189576,
+      "learning_rate": 6.245654673935955e-06,
+      "loss": 0.7209,
+      "step": 5350
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.363007538770154,
+      "learning_rate": 6.244372590158802e-06,
+      "loss": 1.1466,
+      "step": 5351
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.362132366906052,
+      "learning_rate": 6.243090419156547e-06,
+      "loss": 1.5448,
+      "step": 5352
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.5829805256999094,
+      "learning_rate": 6.241808161019063e-06,
+      "loss": 0.5098,
+      "step": 5353
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.414527589241339,
+      "learning_rate": 6.2405258158362315e-06,
+      "loss": 0.7025,
+      "step": 5354
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.349450647296863,
+      "learning_rate": 6.239243383697938e-06,
+      "loss": 1.1286,
+      "step": 5355
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.115393927866742,
+      "learning_rate": 6.237960864694077e-06,
+      "loss": 0.913,
+      "step": 5356
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.3695423402768703,
+      "learning_rate": 6.236678258914548e-06,
+      "loss": 1.0175,
+      "step": 5357
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.415987708355679,
+      "learning_rate": 6.235395566449253e-06,
+      "loss": 0.9599,
+      "step": 5358
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.2467334640139955,
+      "learning_rate": 6.234112787388107e-06,
+      "loss": 0.448,
+      "step": 5359
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.8813058250264225,
+      "learning_rate": 6.232829921821025e-06,
+      "loss": 0.8177,
+      "step": 5360
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.876710756602804,
+      "learning_rate": 6.2315469698379316e-06,
+      "loss": 0.6011,
+      "step": 5361
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.4807173168245664,
+      "learning_rate": 6.230263931528755e-06,
+      "loss": 0.5266,
+      "step": 5362
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.781292797993209,
+      "learning_rate": 6.2289808069834315e-06,
+      "loss": 0.7123,
+      "step": 5363
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.1198830841349747,
+      "learning_rate": 6.227697596291899e-06,
+      "loss": 0.4698,
+      "step": 5364
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.6855322557577477,
+      "learning_rate": 6.22641429954411e-06,
+      "loss": 0.73,
+      "step": 5365
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.8384085488713446,
+      "learning_rate": 6.225130916830017e-06,
+      "loss": 0.8108,
+      "step": 5366
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.4435267183396716,
+      "learning_rate": 6.223847448239577e-06,
+      "loss": 0.5509,
+      "step": 5367
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.270433996869262,
+      "learning_rate": 6.222563893862758e-06,
+      "loss": 0.9696,
+      "step": 5368
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.5873863164036046,
+      "learning_rate": 6.22128025378953e-06,
+      "loss": 0.6469,
+      "step": 5369
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.631306525128808,
+      "learning_rate": 6.219996528109872e-06,
+      "loss": 0.6957,
+      "step": 5370
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.8230625464072774,
+      "learning_rate": 6.218712716913766e-06,
+      "loss": 0.7993,
+      "step": 5371
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.009789421544894,
+      "learning_rate": 6.217428820291202e-06,
+      "loss": 0.3,
+      "step": 5372
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.322712395684982,
+      "learning_rate": 6.2161448383321766e-06,
+      "loss": 0.7037,
+      "step": 5373
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.018911762920762,
+      "learning_rate": 6.214860771126692e-06,
+      "loss": 1.1273,
+      "step": 5374
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.427465808351938,
+      "learning_rate": 6.213576618764752e-06,
+      "loss": 1.2406,
+      "step": 5375
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.7019203466619572,
+      "learning_rate": 6.2122923813363746e-06,
+      "loss": 0.6044,
+      "step": 5376
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.4004793599648444,
+      "learning_rate": 6.211008058931577e-06,
+      "loss": 0.5663,
+      "step": 5377
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.592096067149256,
+      "learning_rate": 6.209723651640383e-06,
+      "loss": 1.1116,
+      "step": 5378
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.6651568911112204,
+      "learning_rate": 6.208439159552826e-06,
+      "loss": 0.7627,
+      "step": 5379
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.456885830028118,
+      "learning_rate": 6.207154582758945e-06,
+      "loss": 0.4029,
+      "step": 5380
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2173491774046696,
+      "learning_rate": 6.205869921348779e-06,
+      "loss": 0.1786,
+      "step": 5381
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.2110055439490357,
+      "learning_rate": 6.204585175412381e-06,
+      "loss": 0.7859,
+      "step": 5382
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.813704688609762,
+      "learning_rate": 6.203300345039804e-06,
+      "loss": 1.3696,
+      "step": 5383
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.059360791876233,
+      "learning_rate": 6.202015430321111e-06,
+      "loss": 0.2938,
+      "step": 5384
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.339930811234842,
+      "learning_rate": 6.200730431346366e-06,
+      "loss": 1.0934,
+      "step": 5385
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.948427648322673,
+      "learning_rate": 6.1994453482056436e-06,
+      "loss": 0.642,
+      "step": 5386
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.2495473070357863,
+      "learning_rate": 6.198160180989022e-06,
+      "loss": 0.8292,
+      "step": 5387
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.579186010590315,
+      "learning_rate": 6.196874929786587e-06,
+      "loss": 0.5365,
+      "step": 5388
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.12675507132254,
+      "learning_rate": 6.195589594688428e-06,
+      "loss": 0.8934,
+      "step": 5389
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.3617677981629064,
+      "learning_rate": 6.194304175784641e-06,
+      "loss": 0.6774,
+      "step": 5390
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.848370066636426,
+      "learning_rate": 6.19301867316533e-06,
+      "loss": 1.2827,
+      "step": 5391
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.417159823089846,
+      "learning_rate": 6.1917330869206015e-06,
+      "loss": 1.1245,
+      "step": 5392
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.440488032675081,
+      "learning_rate": 6.190447417140569e-06,
+      "loss": 0.7665,
+      "step": 5393
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.7880391211632882,
+      "learning_rate": 6.189161663915355e-06,
+      "loss": 0.6494,
+      "step": 5394
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.784454644022496,
+      "learning_rate": 6.187875827335082e-06,
+      "loss": 0.8974,
+      "step": 5395
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.6564929135237456,
+      "learning_rate": 6.186589907489884e-06,
+      "loss": 0.4535,
+      "step": 5396
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.6595258116723364,
+      "learning_rate": 6.185303904469898e-06,
+      "loss": 0.7704,
+      "step": 5397
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.558221882463503,
+      "learning_rate": 6.1840178183652665e-06,
+      "loss": 0.9053,
+      "step": 5398
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.009016678164901,
+      "learning_rate": 6.1827316492661395e-06,
+      "loss": 0.934,
+      "step": 5399
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.933120148917259,
+      "learning_rate": 6.181445397262671e-06,
+      "loss": 0.5614,
+      "step": 5400
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.563426634437689,
+      "learning_rate": 6.180159062445021e-06,
+      "loss": 1.2745,
+      "step": 5401
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.930257212761975,
+      "learning_rate": 6.178872644903355e-06,
+      "loss": 0.8114,
+      "step": 5402
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.4791722485318206,
+      "learning_rate": 6.177586144727851e-06,
+      "loss": 0.85,
+      "step": 5403
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.404753109780831,
+      "learning_rate": 6.17629956200868e-06,
+      "loss": 0.9254,
+      "step": 5404
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.9481351872811214,
+      "learning_rate": 6.17501289683603e-06,
+      "loss": 0.8583,
+      "step": 5405
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.9426272635727204,
+      "learning_rate": 6.17372614930009e-06,
+      "loss": 0.9742,
+      "step": 5406
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.7767531685817355,
+      "learning_rate": 6.172439319491055e-06,
+      "loss": 1.239,
+      "step": 5407
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.140978720203091,
+      "learning_rate": 6.171152407499127e-06,
+      "loss": 1.0022,
+      "step": 5408
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.5839408873445144,
+      "learning_rate": 6.16986541341451e-06,
+      "loss": 0.4733,
+      "step": 5409
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.6971096050248082,
+      "learning_rate": 6.168578337327419e-06,
+      "loss": 0.6547,
+      "step": 5410
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.608322605111382,
+      "learning_rate": 6.167291179328074e-06,
+      "loss": 0.6168,
+      "step": 5411
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.6416010801486935,
+      "learning_rate": 6.166003939506696e-06,
+      "loss": 0.7506,
+      "step": 5412
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.090256949971821,
+      "learning_rate": 6.164716617953515e-06,
+      "loss": 0.8866,
+      "step": 5413
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.722911518681974,
+      "learning_rate": 6.163429214758772e-06,
+      "loss": 0.7578,
+      "step": 5414
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.9354120613448753,
+      "learning_rate": 6.1621417300127015e-06,
+      "loss": 0.6877,
+      "step": 5415
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.216105671503533,
+      "learning_rate": 6.160854163805554e-06,
+      "loss": 0.9651,
+      "step": 5416
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.631811758347738,
+      "learning_rate": 6.159566516227582e-06,
+      "loss": 1.1827,
+      "step": 5417
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.0660770200994407,
+      "learning_rate": 6.158278787369043e-06,
+      "loss": 0.7919,
+      "step": 5418
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.432485108761104,
+      "learning_rate": 6.156990977320201e-06,
+      "loss": 1.1963,
+      "step": 5419
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.591645022195573,
+      "learning_rate": 6.155703086171328e-06,
+      "loss": 0.9741,
+      "step": 5420
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.2928527957734675,
+      "learning_rate": 6.154415114012697e-06,
+      "loss": 0.5989,
+      "step": 5421
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.5588751254931115,
+      "learning_rate": 6.1531270609345915e-06,
+      "loss": 1.0163,
+      "step": 5422
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.696751098953714,
+      "learning_rate": 6.151838927027299e-06,
+      "loss": 1.1034,
+      "step": 5423
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.01743567800604,
+      "learning_rate": 6.150550712381109e-06,
+      "loss": 0.7967,
+      "step": 5424
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.214745265293675,
+      "learning_rate": 6.149262417086321e-06,
+      "loss": 0.9011,
+      "step": 5425
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.5396276536538127,
+      "learning_rate": 6.14797404123324e-06,
+      "loss": 0.9816,
+      "step": 5426
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.9307304526840565,
+      "learning_rate": 6.146685584912174e-06,
+      "loss": 0.3536,
+      "step": 5427
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.9738710837402182,
+      "learning_rate": 6.1453970482134395e-06,
+      "loss": 1.1024,
+      "step": 5428
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.634119645787071,
+      "learning_rate": 6.1441084312273555e-06,
+      "loss": 0.6975,
+      "step": 5429
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.9498673560304196,
+      "learning_rate": 6.142819734044251e-06,
+      "loss": 1.0039,
+      "step": 5430
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.557682544770897,
+      "learning_rate": 6.141530956754457e-06,
+      "loss": 0.8699,
+      "step": 5431
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.506189181916139,
+      "learning_rate": 6.1402420994483104e-06,
+      "loss": 0.7495,
+      "step": 5432
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.374658163440085,
+      "learning_rate": 6.138953162216154e-06,
+      "loss": 0.6774,
+      "step": 5433
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.5169033704099713,
+      "learning_rate": 6.137664145148339e-06,
+      "loss": 0.6085,
+      "step": 5434
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.7937381584883245,
+      "learning_rate": 6.136375048335217e-06,
+      "loss": 0.6849,
+      "step": 5435
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.520608587272846,
+      "learning_rate": 6.13508587186715e-06,
+      "loss": 0.9342,
+      "step": 5436
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.906253332892987,
+      "learning_rate": 6.133796615834504e-06,
+      "loss": 1.0895,
+      "step": 5437
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.032434857952008,
+      "learning_rate": 6.13250728032765e-06,
+      "loss": 1.4765,
+      "step": 5438
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.800388318250968,
+      "learning_rate": 6.131217865436964e-06,
+      "loss": 1.2376,
+      "step": 5439
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.600806621811746,
+      "learning_rate": 6.129928371252829e-06,
+      "loss": 0.8767,
+      "step": 5440
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.008348225563147,
+      "learning_rate": 6.128638797865631e-06,
+      "loss": 1.0546,
+      "step": 5441
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.12472089271931,
+      "learning_rate": 6.127349145365766e-06,
+      "loss": 1.1961,
+      "step": 5442
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.319833628907693,
+      "learning_rate": 6.126059413843633e-06,
+      "loss": 1.1684,
+      "step": 5443
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.48468306527288,
+      "learning_rate": 6.1247696033896345e-06,
+      "loss": 0.9538,
+      "step": 5444
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.1059854657164463,
+      "learning_rate": 6.123479714094181e-06,
+      "loss": 0.3268,
+      "step": 5445
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.9646551102691605,
+      "learning_rate": 6.122189746047691e-06,
+      "loss": 0.7663,
+      "step": 5446
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.4451680586833144,
+      "learning_rate": 6.1208996993405835e-06,
+      "loss": 1.1906,
+      "step": 5447
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.69245122401176,
+      "learning_rate": 6.119609574063285e-06,
+      "loss": 0.9925,
+      "step": 5448
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.7749565835361505,
+      "learning_rate": 6.118319370306227e-06,
+      "loss": 1.0335,
+      "step": 5449
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.74924286006014,
+      "learning_rate": 6.117029088159849e-06,
+      "loss": 0.7541,
+      "step": 5450
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.710865070074655,
+      "learning_rate": 6.115738727714593e-06,
+      "loss": 0.7989,
+      "step": 5451
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.197289043979481,
+      "learning_rate": 6.114448289060908e-06,
+      "loss": 0.539,
+      "step": 5452
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.880621264997749,
+      "learning_rate": 6.113157772289246e-06,
+      "loss": 0.8333,
+      "step": 5453
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.72748977598308,
+      "learning_rate": 6.111867177490072e-06,
+      "loss": 0.9621,
+      "step": 5454
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.43236946265601,
+      "learning_rate": 6.1105765047538465e-06,
+      "loss": 0.9359,
+      "step": 5455
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.8075159373402525,
+      "learning_rate": 6.10928575417104e-06,
+      "loss": 1.1357,
+      "step": 5456
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.3112380522958835,
+      "learning_rate": 6.107994925832131e-06,
+      "loss": 0.6632,
+      "step": 5457
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.6203833931869545,
+      "learning_rate": 6.106704019827599e-06,
+      "loss": 1.3663,
+      "step": 5458
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.6398419742734265,
+      "learning_rate": 6.105413036247933e-06,
+      "loss": 0.6478,
+      "step": 5459
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.119880358959511,
+      "learning_rate": 6.104121975183623e-06,
+      "loss": 0.9093,
+      "step": 5460
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.932584298045523,
+      "learning_rate": 6.102830836725167e-06,
+      "loss": 1.1778,
+      "step": 5461
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.117612576846759,
+      "learning_rate": 6.10153962096307e-06,
+      "loss": 0.9118,
+      "step": 5462
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.6671893745898414,
+      "learning_rate": 6.100248327987839e-06,
+      "loss": 1.0006,
+      "step": 5463
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 6.39738255545773,
+      "learning_rate": 6.0989569578899885e-06,
+      "loss": 1.1004,
+      "step": 5464
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.3645479788769417,
+      "learning_rate": 6.097665510760037e-06,
+      "loss": 0.4135,
+      "step": 5465
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.8817906212282347,
+      "learning_rate": 6.096373986688512e-06,
+      "loss": 0.499,
+      "step": 5466
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.431268970812689,
+      "learning_rate": 6.09508238576594e-06,
+      "loss": 0.9452,
+      "step": 5467
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.3163539589645685,
+      "learning_rate": 6.093790708082861e-06,
+      "loss": 0.4327,
+      "step": 5468
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.8907058897093836,
+      "learning_rate": 6.092498953729812e-06,
+      "loss": 1.168,
+      "step": 5469
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.771897691342272,
+      "learning_rate": 6.091207122797341e-06,
+      "loss": 0.4317,
+      "step": 5470
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.400698623834575,
+      "learning_rate": 6.089915215376001e-06,
+      "loss": 0.9681,
+      "step": 5471
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.2667063679311665,
+      "learning_rate": 6.088623231556345e-06,
+      "loss": 0.5894,
+      "step": 5472
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.1379010219634256,
+      "learning_rate": 6.087331171428941e-06,
+      "loss": 0.581,
+      "step": 5473
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.992031232407942,
+      "learning_rate": 6.086039035084353e-06,
+      "loss": 0.6718,
+      "step": 5474
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.954322977592264,
+      "learning_rate": 6.084746822613154e-06,
+      "loss": 1.194,
+      "step": 5475
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.707176933117051,
+      "learning_rate": 6.083454534105924e-06,
+      "loss": 0.4213,
+      "step": 5476
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 6.327659835343956,
+      "learning_rate": 6.082162169653247e-06,
+      "loss": 1.5028,
+      "step": 5477
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.502937909209937,
+      "learning_rate": 6.080869729345712e-06,
+      "loss": 0.5243,
+      "step": 5478
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.2022708381766205,
+      "learning_rate": 6.079577213273911e-06,
+      "loss": 0.9597,
+      "step": 5479
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.902687279888713,
+      "learning_rate": 6.078284621528448e-06,
+      "loss": 1.1178,
+      "step": 5480
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.3045093198536692,
+      "learning_rate": 6.076991954199923e-06,
+      "loss": 0.3561,
+      "step": 5481
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.5323094331506057,
+      "learning_rate": 6.0756992113789514e-06,
+      "loss": 0.7133,
+      "step": 5482
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.904239608362607,
+      "learning_rate": 6.074406393156146e-06,
+      "loss": 1.0177,
+      "step": 5483
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.155375095499139,
+      "learning_rate": 6.073113499622127e-06,
+      "loss": 0.3451,
+      "step": 5484
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.6957454743955696,
+      "learning_rate": 6.071820530867524e-06,
+      "loss": 0.4902,
+      "step": 5485
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.7827946076692256,
+      "learning_rate": 6.070527486982965e-06,
+      "loss": 0.782,
+      "step": 5486
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.670415124750267,
+      "learning_rate": 6.0692343680590894e-06,
+      "loss": 1.0284,
+      "step": 5487
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.632010818389444,
+      "learning_rate": 6.067941174186537e-06,
+      "loss": 0.72,
+      "step": 5488
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.421765823790791,
+      "learning_rate": 6.066647905455955e-06,
+      "loss": 1.3704,
+      "step": 5489
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.4884903471133044,
+      "learning_rate": 6.065354561957998e-06,
+      "loss": 0.9789,
+      "step": 5490
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.738458251237575,
+      "learning_rate": 6.064061143783323e-06,
+      "loss": 0.7934,
+      "step": 5491
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.025182361176615,
+      "learning_rate": 6.0627676510225915e-06,
+      "loss": 0.8666,
+      "step": 5492
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.467433330714114,
+      "learning_rate": 6.061474083766475e-06,
+      "loss": 0.6233,
+      "step": 5493
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.3534751064256008,
+      "learning_rate": 6.060180442105643e-06,
+      "loss": 0.5152,
+      "step": 5494
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.071042799742737,
+      "learning_rate": 6.058886726130776e-06,
+      "loss": 1.062,
+      "step": 5495
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.362853383119304,
+      "learning_rate": 6.057592935932557e-06,
+      "loss": 0.4955,
+      "step": 5496
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.536563141462291,
+      "learning_rate": 6.056299071601678e-06,
+      "loss": 0.9823,
+      "step": 5497
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.8697421361151076,
+      "learning_rate": 6.055005133228829e-06,
+      "loss": 0.6875,
+      "step": 5498
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.109542289630285,
+      "learning_rate": 6.0537111209047115e-06,
+      "loss": 0.4757,
+      "step": 5499
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.5212861195477325,
+      "learning_rate": 6.052417034720032e-06,
+      "loss": 0.8824,
+      "step": 5500
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.002153356204538,
+      "learning_rate": 6.0511228747654985e-06,
+      "loss": 0.9321,
+      "step": 5501
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.0857396379637962,
+      "learning_rate": 6.0498286411318255e-06,
+      "loss": 0.9694,
+      "step": 5502
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.853963101535627,
+      "learning_rate": 6.0485343339097326e-06,
+      "loss": 0.8505,
+      "step": 5503
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.3368021160183443,
+      "learning_rate": 6.047239953189947e-06,
+      "loss": 0.9013,
+      "step": 5504
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.310716106274053,
+      "learning_rate": 6.045945499063197e-06,
+      "loss": 0.8569,
+      "step": 5505
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.6831018651989345,
+      "learning_rate": 6.044650971620222e-06,
+      "loss": 1.0725,
+      "step": 5506
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.9558960430298806,
+      "learning_rate": 6.043356370951757e-06,
+      "loss": 0.5028,
+      "step": 5507
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8734062377687183,
+      "learning_rate": 6.042061697148555e-06,
+      "loss": 0.1543,
+      "step": 5508
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.0318544937798455,
+      "learning_rate": 6.040766950301361e-06,
+      "loss": 0.8827,
+      "step": 5509
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.356019493312616,
+      "learning_rate": 6.039472130500933e-06,
+      "loss": 0.6415,
+      "step": 5510
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9896690571284249,
+      "learning_rate": 6.038177237838034e-06,
+      "loss": 0.1408,
+      "step": 5511
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.614899518715243,
+      "learning_rate": 6.036882272403426e-06,
+      "loss": 0.8803,
+      "step": 5512
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.004401352590362,
+      "learning_rate": 6.035587234287884e-06,
+      "loss": 0.7291,
+      "step": 5513
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.7206832528416385,
+      "learning_rate": 6.034292123582185e-06,
+      "loss": 1.5733,
+      "step": 5514
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.143042002035592,
+      "learning_rate": 6.032996940377108e-06,
+      "loss": 1.1401,
+      "step": 5515
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.4042392691211365,
+      "learning_rate": 6.031701684763443e-06,
+      "loss": 1.0504,
+      "step": 5516
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.4200691856860685,
+      "learning_rate": 6.030406356831979e-06,
+      "loss": 0.9079,
+      "step": 5517
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.323923364132357,
+      "learning_rate": 6.029110956673513e-06,
+      "loss": 0.6282,
+      "step": 5518
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.098465470441731,
+      "learning_rate": 6.027815484378848e-06,
+      "loss": 0.9255,
+      "step": 5519
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.5134793061531457,
+      "learning_rate": 6.0265199400387904e-06,
+      "loss": 0.6663,
+      "step": 5520
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.517252099688381,
+      "learning_rate": 6.025224323744153e-06,
+      "loss": 0.9122,
+      "step": 5521
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.236853991100827,
+      "learning_rate": 6.023928635585752e-06,
+      "loss": 0.5387,
+      "step": 5522
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.0916988080878007,
+      "learning_rate": 6.0226328756544105e-06,
+      "loss": 0.3724,
+      "step": 5523
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.777582158576558,
+      "learning_rate": 6.021337044040954e-06,
+      "loss": 0.889,
+      "step": 5524
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.314986998203574,
+      "learning_rate": 6.020041140836217e-06,
+      "loss": 0.3922,
+      "step": 5525
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.836377061609312,
+      "learning_rate": 6.0187451661310345e-06,
+      "loss": 0.7884,
+      "step": 5526
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.28583804011794,
+      "learning_rate": 6.017449120016249e-06,
+      "loss": 0.9602,
+      "step": 5527
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.8415936348387962,
+      "learning_rate": 6.016153002582708e-06,
+      "loss": 0.3775,
+      "step": 5528
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.3430676896119875,
+      "learning_rate": 6.014856813921264e-06,
+      "loss": 1.0213,
+      "step": 5529
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.3657222141846157,
+      "learning_rate": 6.013560554122773e-06,
+      "loss": 0.3764,
+      "step": 5530
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.845682187224199,
+      "learning_rate": 6.0122642232781e-06,
+      "loss": 0.472,
+      "step": 5531
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.7079274159398072,
+      "learning_rate": 6.01096782147811e-06,
+      "loss": 0.469,
+      "step": 5532
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.345960476458339,
+      "learning_rate": 6.009671348813675e-06,
+      "loss": 0.2141,
+      "step": 5533
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.052975998276666,
+      "learning_rate": 6.008374805375674e-06,
+      "loss": 0.595,
+      "step": 5534
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.3505565706139375,
+      "learning_rate": 6.0070781912549855e-06,
+      "loss": 1.0399,
+      "step": 5535
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.2468596412968216,
+      "learning_rate": 6.005781506542498e-06,
+      "loss": 0.6398,
+      "step": 5536
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.073450939851316,
+      "learning_rate": 6.004484751329107e-06,
+      "loss": 0.8609,
+      "step": 5537
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.93717082896763,
+      "learning_rate": 6.003187925705704e-06,
+      "loss": 1.2229,
+      "step": 5538
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.492738577572085,
+      "learning_rate": 6.001891029763194e-06,
+      "loss": 0.7971,
+      "step": 5539
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.101721588026555,
+      "learning_rate": 6.000594063592484e-06,
+      "loss": 0.7578,
+      "step": 5540
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.8164989870699646,
+      "learning_rate": 5.999297027284484e-06,
+      "loss": 0.7848,
+      "step": 5541
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.4856305895260355,
+      "learning_rate": 5.997999920930111e-06,
+      "loss": 0.8738,
+      "step": 5542
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.6225289921817803,
+      "learning_rate": 5.9967027446202885e-06,
+      "loss": 0.7648,
+      "step": 5543
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.766273112541162,
+      "learning_rate": 5.995405498445939e-06,
+      "loss": 0.3674,
+      "step": 5544
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.7646141585134343,
+      "learning_rate": 5.994108182497997e-06,
+      "loss": 0.7503,
+      "step": 5545
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.8461361156409706,
+      "learning_rate": 5.992810796867398e-06,
+      "loss": 0.7506,
+      "step": 5546
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.003624900489704,
+      "learning_rate": 5.991513341645082e-06,
+      "loss": 0.9627,
+      "step": 5547
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.417943959323835,
+      "learning_rate": 5.990215816921998e-06,
+      "loss": 1.0619,
+      "step": 5548
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.733192583725168,
+      "learning_rate": 5.988918222789093e-06,
+      "loss": 0.8055,
+      "step": 5549
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.890782699748269,
+      "learning_rate": 5.987620559337325e-06,
+      "loss": 0.7575,
+      "step": 5550
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.791496139999864,
+      "learning_rate": 5.9863228266576535e-06,
+      "loss": 1.1452,
+      "step": 5551
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8890858540861521,
+      "learning_rate": 5.985025024841043e-06,
+      "loss": 0.3557,
+      "step": 5552
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.912391990086859,
+      "learning_rate": 5.983727153978467e-06,
+      "loss": 0.5329,
+      "step": 5553
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.0390393911337132,
+      "learning_rate": 5.982429214160899e-06,
+      "loss": 0.5287,
+      "step": 5554
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.299533614708435,
+      "learning_rate": 5.981131205479317e-06,
+      "loss": 0.3959,
+      "step": 5555
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.61051813457963,
+      "learning_rate": 5.9798331280247094e-06,
+      "loss": 1.3886,
+      "step": 5556
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.8045804554973506,
+      "learning_rate": 5.9785349818880626e-06,
+      "loss": 0.7375,
+      "step": 5557
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.170660867053804,
+      "learning_rate": 5.9772367671603715e-06,
+      "loss": 0.5208,
+      "step": 5558
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.786348074967104,
+      "learning_rate": 5.975938483932636e-06,
+      "loss": 0.7823,
+      "step": 5559
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.6829597297233416,
+      "learning_rate": 5.974640132295862e-06,
+      "loss": 0.5209,
+      "step": 5560
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.311060390431268,
+      "learning_rate": 5.973341712341054e-06,
+      "loss": 0.5302,
+      "step": 5561
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.122646154918164,
+      "learning_rate": 5.9720432241592285e-06,
+      "loss": 0.835,
+      "step": 5562
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.768375739659902,
+      "learning_rate": 5.970744667841404e-06,
+      "loss": 1.0994,
+      "step": 5563
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.145289986517489,
+      "learning_rate": 5.9694460434786035e-06,
+      "loss": 0.7075,
+      "step": 5564
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.2116496176499525,
+      "learning_rate": 5.968147351161854e-06,
+      "loss": 0.3714,
+      "step": 5565
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.668740981570834,
+      "learning_rate": 5.9668485909821886e-06,
+      "loss": 0.9293,
+      "step": 5566
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.489957152442339,
+      "learning_rate": 5.965549763030643e-06,
+      "loss": 0.438,
+      "step": 5567
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2246574479366155,
+      "learning_rate": 5.9642508673982634e-06,
+      "loss": 0.225,
+      "step": 5568
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3514524458334403,
+      "learning_rate": 5.9629519041760934e-06,
+      "loss": 0.1995,
+      "step": 5569
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.902275863099431,
+      "learning_rate": 5.961652873455186e-06,
+      "loss": 0.7944,
+      "step": 5570
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.4541717446544973,
+      "learning_rate": 5.9603537753265975e-06,
+      "loss": 0.6926,
+      "step": 5571
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.036517138546423,
+      "learning_rate": 5.959054609881388e-06,
+      "loss": 0.7789,
+      "step": 5572
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.843901626875687,
+      "learning_rate": 5.957755377210624e-06,
+      "loss": 0.7952,
+      "step": 5573
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.240485021356367,
+      "learning_rate": 5.956456077405378e-06,
+      "loss": 0.4053,
+      "step": 5574
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.836154350544739,
+      "learning_rate": 5.955156710556722e-06,
+      "loss": 1.0861,
+      "step": 5575
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.504086906465851,
+      "learning_rate": 5.953857276755737e-06,
+      "loss": 1.4191,
+      "step": 5576
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.627717249324198,
+      "learning_rate": 5.95255777609351e-06,
+      "loss": 0.4167,
+      "step": 5577
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.0684441155294637,
+      "learning_rate": 5.951258208661126e-06,
+      "loss": 0.4953,
+      "step": 5578
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.802080977568901,
+      "learning_rate": 5.949958574549683e-06,
+      "loss": 0.5834,
+      "step": 5579
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.077596077455358,
+      "learning_rate": 5.948658873850279e-06,
+      "loss": 0.9969,
+      "step": 5580
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.116797514430086,
+      "learning_rate": 5.947359106654016e-06,
+      "loss": 0.7344,
+      "step": 5581
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.413224521799694,
+      "learning_rate": 5.946059273052001e-06,
+      "loss": 1.4048,
+      "step": 5582
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.876379618779488,
+      "learning_rate": 5.944759373135349e-06,
+      "loss": 0.8151,
+      "step": 5583
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.721767456197241,
+      "learning_rate": 5.943459406995177e-06,
+      "loss": 0.7432,
+      "step": 5584
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.366332111754268,
+      "learning_rate": 5.942159374722606e-06,
+      "loss": 0.7185,
+      "step": 5585
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.0003025244338435,
+      "learning_rate": 5.940859276408764e-06,
+      "loss": 0.8708,
+      "step": 5586
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.4982341148824743,
+      "learning_rate": 5.939559112144781e-06,
+      "loss": 0.7583,
+      "step": 5587
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.4006679307302226,
+      "learning_rate": 5.938258882021793e-06,
+      "loss": 0.3947,
+      "step": 5588
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.548102068329933,
+      "learning_rate": 5.936958586130941e-06,
+      "loss": 0.8525,
+      "step": 5589
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.541643842186359,
+      "learning_rate": 5.935658224563369e-06,
+      "loss": 1.1481,
+      "step": 5590
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.6352525308965022,
+      "learning_rate": 5.934357797410229e-06,
+      "loss": 0.6617,
+      "step": 5591
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.956310662055934,
+      "learning_rate": 5.933057304762672e-06,
+      "loss": 0.7801,
+      "step": 5592
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.0028046951097123,
+      "learning_rate": 5.9317567467118585e-06,
+      "loss": 0.5619,
+      "step": 5593
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.959436889256036,
+      "learning_rate": 5.930456123348953e-06,
+      "loss": 0.9281,
+      "step": 5594
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.907500813877162,
+      "learning_rate": 5.929155434765122e-06,
+      "loss": 0.4004,
+      "step": 5595
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.106542104042912,
+      "learning_rate": 5.927854681051539e-06,
+      "loss": 0.622,
+      "step": 5596
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5786093027769945,
+      "learning_rate": 5.926553862299382e-06,
+      "loss": 0.3419,
+      "step": 5597
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.5457708150946003,
+      "learning_rate": 5.92525297859983e-06,
+      "loss": 0.7155,
+      "step": 5598
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.684123186030232,
+      "learning_rate": 5.923952030044071e-06,
+      "loss": 0.6645,
+      "step": 5599
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.085103688310422,
+      "learning_rate": 5.922651016723298e-06,
+      "loss": 0.4691,
+      "step": 5600
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.366456362481405,
+      "learning_rate": 5.9213499387287025e-06,
+      "loss": 0.4627,
+      "step": 5601
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.945239115251609,
+      "learning_rate": 5.9200487961514855e-06,
+      "loss": 1.3532,
+      "step": 5602
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.0888492208054945,
+      "learning_rate": 5.918747589082853e-06,
+      "loss": 1.133,
+      "step": 5603
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.8174920053002035,
+      "learning_rate": 5.917446317614012e-06,
+      "loss": 0.9695,
+      "step": 5604
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.6787831396291586,
+      "learning_rate": 5.916144981836177e-06,
+      "loss": 0.5809,
+      "step": 5605
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2059337241792771,
+      "learning_rate": 5.914843581840566e-06,
+      "loss": 0.2091,
+      "step": 5606
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.134426566217191,
+      "learning_rate": 5.913542117718401e-06,
+      "loss": 0.556,
+      "step": 5607
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.488473168302814,
+      "learning_rate": 5.91224058956091e-06,
+      "loss": 0.4394,
+      "step": 5608
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.674172646517885,
+      "learning_rate": 5.9109389974593234e-06,
+      "loss": 1.0567,
+      "step": 5609
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.098756970166008,
+      "learning_rate": 5.909637341504878e-06,
+      "loss": 0.731,
+      "step": 5610
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.020079420394541,
+      "learning_rate": 5.908335621788814e-06,
+      "loss": 0.7689,
+      "step": 5611
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.9346032106283833,
+      "learning_rate": 5.907033838402375e-06,
+      "loss": 0.8292,
+      "step": 5612
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.3910071745667127,
+      "learning_rate": 5.90573199143681e-06,
+      "loss": 0.8922,
+      "step": 5613
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.5611325907497147,
+      "learning_rate": 5.904430080983378e-06,
+      "loss": 0.5135,
+      "step": 5614
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.210320865357581,
+      "learning_rate": 5.9031281071333305e-06,
+      "loss": 0.87,
+      "step": 5615
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.9804731507309135,
+      "learning_rate": 5.901826069977933e-06,
+      "loss": 0.5034,
+      "step": 5616
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7768985879771326,
+      "learning_rate": 5.900523969608454e-06,
+      "loss": 0.3578,
+      "step": 5617
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.0134309205375125,
+      "learning_rate": 5.8992218061161645e-06,
+      "loss": 0.6261,
+      "step": 5618
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.525221805410668,
+      "learning_rate": 5.897919579592337e-06,
+      "loss": 1.1076,
+      "step": 5619
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.80410281650793,
+      "learning_rate": 5.896617290128258e-06,
+      "loss": 1.1153,
+      "step": 5620
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.8945697187673267,
+      "learning_rate": 5.895314937815206e-06,
+      "loss": 0.9155,
+      "step": 5621
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.4692131851393646,
+      "learning_rate": 5.894012522744474e-06,
+      "loss": 0.3466,
+      "step": 5622
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.439566624699234,
+      "learning_rate": 5.892710045007357e-06,
+      "loss": 1.1799,
+      "step": 5623
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3540267704180882,
+      "learning_rate": 5.891407504695149e-06,
+      "loss": 0.2322,
+      "step": 5624
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.8034982849943217,
+      "learning_rate": 5.8901049018991564e-06,
+      "loss": 0.7363,
+      "step": 5625
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.844189826181136,
+      "learning_rate": 5.888802236710681e-06,
+      "loss": 0.568,
+      "step": 5626
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.288461366974695,
+      "learning_rate": 5.88749950922104e-06,
+      "loss": 1.1385,
+      "step": 5627
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.448119701622255,
+      "learning_rate": 5.886196719521544e-06,
+      "loss": 1.1424,
+      "step": 5628
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.2093745245117726,
+      "learning_rate": 5.884893867703515e-06,
+      "loss": 0.6699,
+      "step": 5629
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.6306971758348965,
+      "learning_rate": 5.883590953858276e-06,
+      "loss": 0.4229,
+      "step": 5630
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.804745123512171,
+      "learning_rate": 5.882287978077158e-06,
+      "loss": 0.4848,
+      "step": 5631
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.331107191514581,
+      "learning_rate": 5.880984940451491e-06,
+      "loss": 1.6346,
+      "step": 5632
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.233822282708715,
+      "learning_rate": 5.879681841072614e-06,
+      "loss": 0.7985,
+      "step": 5633
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.374574431792657,
+      "learning_rate": 5.87837868003187e-06,
+      "loss": 0.5632,
+      "step": 5634
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.36991586919913,
+      "learning_rate": 5.877075457420602e-06,
+      "loss": 0.7271,
+      "step": 5635
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.221505425581795,
+      "learning_rate": 5.875772173330162e-06,
+      "loss": 0.1695,
+      "step": 5636
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.6084175650892627,
+      "learning_rate": 5.874468827851903e-06,
+      "loss": 0.5352,
+      "step": 5637
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.3842514466431557,
+      "learning_rate": 5.873165421077186e-06,
+      "loss": 0.6894,
+      "step": 5638
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.907343104798318,
+      "learning_rate": 5.871861953097372e-06,
+      "loss": 1.0439,
+      "step": 5639
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8433580518949468,
+      "learning_rate": 5.87055842400383e-06,
+      "loss": 0.3318,
+      "step": 5640
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.193401909883628,
+      "learning_rate": 5.869254833887931e-06,
+      "loss": 0.6984,
+      "step": 5641
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.215790357501764,
+      "learning_rate": 5.867951182841052e-06,
+      "loss": 0.9167,
+      "step": 5642
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.2390874229381965,
+      "learning_rate": 5.866647470954572e-06,
+      "loss": 0.4804,
+      "step": 5643
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.5895580562867244,
+      "learning_rate": 5.8653436983198755e-06,
+      "loss": 0.4501,
+      "step": 5644
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.5130031528387895,
+      "learning_rate": 5.864039865028351e-06,
+      "loss": 0.64,
+      "step": 5645
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.5725355905213174,
+      "learning_rate": 5.862735971171394e-06,
+      "loss": 0.8459,
+      "step": 5646
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.600874218765597,
+      "learning_rate": 5.8614320168403986e-06,
+      "loss": 0.5834,
+      "step": 5647
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.7204376203085254,
+      "learning_rate": 5.860128002126769e-06,
+      "loss": 0.8484,
+      "step": 5648
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.766384604117466,
+      "learning_rate": 5.858823927121908e-06,
+      "loss": 0.5566,
+      "step": 5649
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 6.210756108565881,
+      "learning_rate": 5.85751979191723e-06,
+      "loss": 1.1673,
+      "step": 5650
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.199498809687956,
+      "learning_rate": 5.856215596604146e-06,
+      "loss": 0.7733,
+      "step": 5651
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.9279068313710535,
+      "learning_rate": 5.854911341274074e-06,
+      "loss": 0.9522,
+      "step": 5652
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.0719494374336302,
+      "learning_rate": 5.853607026018435e-06,
+      "loss": 0.519,
+      "step": 5653
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.687955377889996,
+      "learning_rate": 5.852302650928663e-06,
+      "loss": 0.8059,
+      "step": 5654
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.218504325638327,
+      "learning_rate": 5.850998216096181e-06,
+      "loss": 0.8211,
+      "step": 5655
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.226698442187045,
+      "learning_rate": 5.849693721612428e-06,
+      "loss": 1.1432,
+      "step": 5656
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.8147250107784245,
+      "learning_rate": 5.848389167568845e-06,
+      "loss": 1.0348,
+      "step": 5657
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.9113101562177413,
+      "learning_rate": 5.847084554056873e-06,
+      "loss": 0.8063,
+      "step": 5658
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.512599912384629,
+      "learning_rate": 5.845779881167959e-06,
+      "loss": 0.9087,
+      "step": 5659
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.254628577072085,
+      "learning_rate": 5.844475148993558e-06,
+      "loss": 0.7667,
+      "step": 5660
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.965014543590636,
+      "learning_rate": 5.843170357625122e-06,
+      "loss": 0.6832,
+      "step": 5661
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.922702768592842,
+      "learning_rate": 5.8418655071541145e-06,
+      "loss": 0.5666,
+      "step": 5662
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.1475215219336645,
+      "learning_rate": 5.840560597671999e-06,
+      "loss": 0.3893,
+      "step": 5663
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.9612076985975904,
+      "learning_rate": 5.8392556292702425e-06,
+      "loss": 0.5127,
+      "step": 5664
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.8902169840323704,
+      "learning_rate": 5.837950602040321e-06,
+      "loss": 0.6394,
+      "step": 5665
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.313920807203621,
+      "learning_rate": 5.836645516073709e-06,
+      "loss": 1.2423,
+      "step": 5666
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.533505413128344,
+      "learning_rate": 5.835340371461886e-06,
+      "loss": 0.7282,
+      "step": 5667
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.180132481219037,
+      "learning_rate": 5.83403516829634e-06,
+      "loss": 0.5309,
+      "step": 5668
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.42936288406953,
+      "learning_rate": 5.832729906668556e-06,
+      "loss": 0.7588,
+      "step": 5669
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.619655411931857,
+      "learning_rate": 5.83142458667003e-06,
+      "loss": 1.2244,
+      "step": 5670
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.3241952225594344,
+      "learning_rate": 5.83011920839226e-06,
+      "loss": 0.6462,
+      "step": 5671
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.3524127689735126,
+      "learning_rate": 5.828813771926746e-06,
+      "loss": 0.563,
+      "step": 5672
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.7958909598404076,
+      "learning_rate": 5.827508277364994e-06,
+      "loss": 0.9086,
+      "step": 5673
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.6682499503574486,
+      "learning_rate": 5.826202724798513e-06,
+      "loss": 1.0416,
+      "step": 5674
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.167000503301076,
+      "learning_rate": 5.824897114318815e-06,
+      "loss": 0.9161,
+      "step": 5675
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.1926145434582027,
+      "learning_rate": 5.82359144601742e-06,
+      "loss": 0.4624,
+      "step": 5676
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.785094822337679,
+      "learning_rate": 5.8222857199858495e-06,
+      "loss": 1.0595,
+      "step": 5677
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.5222900636213748,
+      "learning_rate": 5.820979936315628e-06,
+      "loss": 0.3336,
+      "step": 5678
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.0167878091794575,
+      "learning_rate": 5.819674095098286e-06,
+      "loss": 0.4347,
+      "step": 5679
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.9541820770494716,
+      "learning_rate": 5.818368196425358e-06,
+      "loss": 0.5592,
+      "step": 5680
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.438792495249288,
+      "learning_rate": 5.8170622403883815e-06,
+      "loss": 1.0044,
+      "step": 5681
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.7398885866053737,
+      "learning_rate": 5.815756227078896e-06,
+      "loss": 0.4634,
+      "step": 5682
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.352284171432852,
+      "learning_rate": 5.814450156588451e-06,
+      "loss": 0.8685,
+      "step": 5683
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.1546835878369515,
+      "learning_rate": 5.813144029008593e-06,
+      "loss": 0.7393,
+      "step": 5684
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.0705438955166504,
+      "learning_rate": 5.811837844430877e-06,
+      "loss": 0.6707,
+      "step": 5685
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.250858309091342,
+      "learning_rate": 5.810531602946863e-06,
+      "loss": 0.9572,
+      "step": 5686
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.595430117641484,
+      "learning_rate": 5.8092253046481095e-06,
+      "loss": 0.2611,
+      "step": 5687
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.0958849916024,
+      "learning_rate": 5.807918949626184e-06,
+      "loss": 0.7519,
+      "step": 5688
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.858938120379014,
+      "learning_rate": 5.806612537972658e-06,
+      "loss": 0.3707,
+      "step": 5689
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.710846347541895,
+      "learning_rate": 5.805306069779102e-06,
+      "loss": 0.7015,
+      "step": 5690
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.6940713544638384,
+      "learning_rate": 5.803999545137096e-06,
+      "loss": 0.8145,
+      "step": 5691
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.9538177719718988,
+      "learning_rate": 5.80269296413822e-06,
+      "loss": 0.5931,
+      "step": 5692
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.266763180046466,
+      "learning_rate": 5.80138632687406e-06,
+      "loss": 0.7787,
+      "step": 5693
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.5477191117174707,
+      "learning_rate": 5.8000796334362074e-06,
+      "loss": 0.5738,
+      "step": 5694
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.7226809880218266,
+      "learning_rate": 5.798772883916254e-06,
+      "loss": 0.6034,
+      "step": 5695
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7352513352651564,
+      "learning_rate": 5.797466078405798e-06,
+      "loss": 0.4274,
+      "step": 5696
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.044027972306153,
+      "learning_rate": 5.796159216996441e-06,
+      "loss": 0.4223,
+      "step": 5697
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.2023661997985142,
+      "learning_rate": 5.794852299779787e-06,
+      "loss": 0.3714,
+      "step": 5698
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.603670980340674,
+      "learning_rate": 5.7935453268474454e-06,
+      "loss": 1.1581,
+      "step": 5699
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.592738406110725,
+      "learning_rate": 5.792238298291031e-06,
+      "loss": 1.0247,
+      "step": 5700
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.7736571140746995,
+      "learning_rate": 5.790931214202159e-06,
+      "loss": 0.7871,
+      "step": 5701
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.3027765022500475,
+      "learning_rate": 5.7896240746724505e-06,
+      "loss": 0.7136,
+      "step": 5702
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.1869716893594364,
+      "learning_rate": 5.788316879793533e-06,
+      "loss": 0.5946,
+      "step": 5703
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.43303866822753,
+      "learning_rate": 5.787009629657032e-06,
+      "loss": 0.6631,
+      "step": 5704
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.369661125838084,
+      "learning_rate": 5.78570232435458e-06,
+      "loss": 0.8293,
+      "step": 5705
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.144263937869944,
+      "learning_rate": 5.784394963977815e-06,
+      "loss": 0.9573,
+      "step": 5706
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.1146427071289295,
+      "learning_rate": 5.783087548618377e-06,
+      "loss": 0.6525,
+      "step": 5707
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.547824438666656,
+      "learning_rate": 5.78178007836791e-06,
+      "loss": 0.966,
+      "step": 5708
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.950261807048099,
+      "learning_rate": 5.7804725533180615e-06,
+      "loss": 1.0529,
+      "step": 5709
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.9747306031892378,
+      "learning_rate": 5.779164973560483e-06,
+      "loss": 0.6283,
+      "step": 5710
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.529311563326823,
+      "learning_rate": 5.777857339186832e-06,
+      "loss": 0.7368,
+      "step": 5711
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.246827561234806,
+      "learning_rate": 5.776549650288767e-06,
+      "loss": 1.0219,
+      "step": 5712
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.408726329418388,
+      "learning_rate": 5.775241906957949e-06,
+      "loss": 0.8025,
+      "step": 5713
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.653669605183592,
+      "learning_rate": 5.7739341092860505e-06,
+      "loss": 1.1074,
+      "step": 5714
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.2976226358559413,
+      "learning_rate": 5.772626257364736e-06,
+      "loss": 0.503,
+      "step": 5715
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.2205688378314914,
+      "learning_rate": 5.771318351285684e-06,
+      "loss": 0.4389,
+      "step": 5716
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.054750227078472,
+      "learning_rate": 5.7700103911405735e-06,
+      "loss": 0.8102,
+      "step": 5717
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.245210152402887,
+      "learning_rate": 5.7687023770210835e-06,
+      "loss": 0.57,
+      "step": 5718
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.907066907446562,
+      "learning_rate": 5.767394309018905e-06,
+      "loss": 0.8984,
+      "step": 5719
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.286741875986198,
+      "learning_rate": 5.766086187225725e-06,
+      "loss": 0.9761,
+      "step": 5720
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.224340502273579,
+      "learning_rate": 5.764778011733235e-06,
+      "loss": 1.224,
+      "step": 5721
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.559377456310667,
+      "learning_rate": 5.763469782633136e-06,
+      "loss": 0.9564,
+      "step": 5722
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.497098893514583,
+      "learning_rate": 5.762161500017128e-06,
+      "loss": 0.5805,
+      "step": 5723
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.434473663235255,
+      "learning_rate": 5.760853163976915e-06,
+      "loss": 0.8909,
+      "step": 5724
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.9448167601439676,
+      "learning_rate": 5.759544774604207e-06,
+      "loss": 0.8595,
+      "step": 5725
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.743227996145874,
+      "learning_rate": 5.758236331990717e-06,
+      "loss": 0.7532,
+      "step": 5726
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.024226223243954,
+      "learning_rate": 5.756927836228158e-06,
+      "loss": 0.8766,
+      "step": 5727
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.5246596586966805,
+      "learning_rate": 5.755619287408253e-06,
+      "loss": 0.4241,
+      "step": 5728
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.005558144090475,
+      "learning_rate": 5.754310685622724e-06,
+      "loss": 0.2407,
+      "step": 5729
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.4679414027430684,
+      "learning_rate": 5.753002030963298e-06,
+      "loss": 0.7409,
+      "step": 5730
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.7261723466337284,
+      "learning_rate": 5.751693323521709e-06,
+      "loss": 0.7915,
+      "step": 5731
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.9821460241633777,
+      "learning_rate": 5.750384563389687e-06,
+      "loss": 0.8028,
+      "step": 5732
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.857185247382448,
+      "learning_rate": 5.749075750658973e-06,
+      "loss": 0.8735,
+      "step": 5733
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.846167569519588,
+      "learning_rate": 5.747766885421309e-06,
+      "loss": 0.5309,
+      "step": 5734
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.7474054280595746,
+      "learning_rate": 5.7464579677684415e-06,
+      "loss": 0.8645,
+      "step": 5735
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.295387737306311,
+      "learning_rate": 5.745148997792119e-06,
+      "loss": 0.8833,
+      "step": 5736
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.1438553701194087,
+      "learning_rate": 5.743839975584096e-06,
+      "loss": 0.4879,
+      "step": 5737
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.732617652793853,
+      "learning_rate": 5.7425309012361255e-06,
+      "loss": 1.003,
+      "step": 5738
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.8092772920083044,
+      "learning_rate": 5.741221774839971e-06,
+      "loss": 0.903,
+      "step": 5739
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.2358834200477284,
+      "learning_rate": 5.739912596487396e-06,
+      "loss": 0.7134,
+      "step": 5740
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.775149360538276,
+      "learning_rate": 5.738603366270168e-06,
+      "loss": 0.436,
+      "step": 5741
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.6307972809973466,
+      "learning_rate": 5.737294084280058e-06,
+      "loss": 0.7147,
+      "step": 5742
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.3848807124944083,
+      "learning_rate": 5.735984750608843e-06,
+      "loss": 1.0273,
+      "step": 5743
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.759591221597147,
+      "learning_rate": 5.734675365348299e-06,
+      "loss": 0.5455,
+      "step": 5744
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.205319482526289,
+      "learning_rate": 5.733365928590208e-06,
+      "loss": 0.9674,
+      "step": 5745
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.888076685253536,
+      "learning_rate": 5.732056440426359e-06,
+      "loss": 0.8197,
+      "step": 5746
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.375645006216542,
+      "learning_rate": 5.730746900948538e-06,
+      "loss": 1.0402,
+      "step": 5747
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.742123547621015,
+      "learning_rate": 5.729437310248541e-06,
+      "loss": 0.4881,
+      "step": 5748
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.354753309388454,
+      "learning_rate": 5.728127668418162e-06,
+      "loss": 0.9882,
+      "step": 5749
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.141328133970809,
+      "learning_rate": 5.726817975549201e-06,
+      "loss": 1.0981,
+      "step": 5750
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.362439718903292,
+      "learning_rate": 5.7255082317334665e-06,
+      "loss": 0.895,
+      "step": 5751
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.465193722877705,
+      "learning_rate": 5.72419843706276e-06,
+      "loss": 0.9494,
+      "step": 5752
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.394136592870638,
+      "learning_rate": 5.722888591628895e-06,
+      "loss": 0.6267,
+      "step": 5753
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.902125420421986,
+      "learning_rate": 5.7215786955236865e-06,
+      "loss": 0.6764,
+      "step": 5754
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 7.066351910581949,
+      "learning_rate": 5.72026874883895e-06,
+      "loss": 1.0081,
+      "step": 5755
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.120000950336103,
+      "learning_rate": 5.7189587516665105e-06,
+      "loss": 0.5777,
+      "step": 5756
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.7120891216701657,
+      "learning_rate": 5.717648704098191e-06,
+      "loss": 0.8645,
+      "step": 5757
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.5467828767445257,
+      "learning_rate": 5.716338606225821e-06,
+      "loss": 0.5317,
+      "step": 5758
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.9582341832144654,
+      "learning_rate": 5.715028458141232e-06,
+      "loss": 0.948,
+      "step": 5759
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.9280643667834787,
+      "learning_rate": 5.71371825993626e-06,
+      "loss": 0.7351,
+      "step": 5760
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.238562911035449,
+      "learning_rate": 5.7124080117027435e-06,
+      "loss": 0.6953,
+      "step": 5761
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.01743293033794,
+      "learning_rate": 5.711097713532525e-06,
+      "loss": 0.92,
+      "step": 5762
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9946077563575941,
+      "learning_rate": 5.709787365517453e-06,
+      "loss": 0.2586,
+      "step": 5763
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.9150789608805256,
+      "learning_rate": 5.708476967749375e-06,
+      "loss": 0.6597,
+      "step": 5764
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8986227363890973,
+      "learning_rate": 5.7071665203201444e-06,
+      "loss": 0.4035,
+      "step": 5765
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.552870877519428,
+      "learning_rate": 5.70585602332162e-06,
+      "loss": 1.4425,
+      "step": 5766
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.336830414117216,
+      "learning_rate": 5.704545476845659e-06,
+      "loss": 0.8516,
+      "step": 5767
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.513803551281201,
+      "learning_rate": 5.703234880984126e-06,
+      "loss": 0.454,
+      "step": 5768
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.9759675928845812,
+      "learning_rate": 5.70192423582889e-06,
+      "loss": 0.5834,
+      "step": 5769
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.596271191049582,
+      "learning_rate": 5.700613541471818e-06,
+      "loss": 0.5559,
+      "step": 5770
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.074366001164237,
+      "learning_rate": 5.6993027980047866e-06,
+      "loss": 0.6883,
+      "step": 5771
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.394754836340234,
+      "learning_rate": 5.6979920055196725e-06,
+      "loss": 1.3086,
+      "step": 5772
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.611169159436164,
+      "learning_rate": 5.696681164108355e-06,
+      "loss": 0.5912,
+      "step": 5773
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.738071640705117,
+      "learning_rate": 5.6953702738627215e-06,
+      "loss": 0.8901,
+      "step": 5774
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.026026575450983,
+      "learning_rate": 5.694059334874658e-06,
+      "loss": 0.8682,
+      "step": 5775
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.7881691134910755,
+      "learning_rate": 5.692748347236055e-06,
+      "loss": 0.5597,
+      "step": 5776
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.318696413394652,
+      "learning_rate": 5.69143731103881e-06,
+      "loss": 1.2268,
+      "step": 5777
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.482657348304525,
+      "learning_rate": 5.6901262263748155e-06,
+      "loss": 0.7588,
+      "step": 5778
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.181824507035459,
+      "learning_rate": 5.6888150933359765e-06,
+      "loss": 1.1801,
+      "step": 5779
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.050095910926002,
+      "learning_rate": 5.687503912014199e-06,
+      "loss": 0.9628,
+      "step": 5780
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.63451128404456,
+      "learning_rate": 5.686192682501388e-06,
+      "loss": 1.1779,
+      "step": 5781
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.2602723144577888,
+      "learning_rate": 5.684881404889456e-06,
+      "loss": 0.4366,
+      "step": 5782
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.9475423541631125,
+      "learning_rate": 5.68357007927032e-06,
+      "loss": 0.968,
+      "step": 5783
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.903104647376368,
+      "learning_rate": 5.682258705735895e-06,
+      "loss": 0.5192,
+      "step": 5784
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.125000797006016,
+      "learning_rate": 5.680947284378102e-06,
+      "loss": 1.1176,
+      "step": 5785
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.9315500926892883,
+      "learning_rate": 5.679635815288871e-06,
+      "loss": 0.7608,
+      "step": 5786
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.367998612990884,
+      "learning_rate": 5.678324298560125e-06,
+      "loss": 0.9233,
+      "step": 5787
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.4874247781562575,
+      "learning_rate": 5.677012734283799e-06,
+      "loss": 0.6668,
+      "step": 5788
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.371404386512218,
+      "learning_rate": 5.675701122551827e-06,
+      "loss": 0.4671,
+      "step": 5789
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.789388752179784,
+      "learning_rate": 5.674389463456146e-06,
+      "loss": 0.5927,
+      "step": 5790
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.8202185280672785,
+      "learning_rate": 5.6730777570887e-06,
+      "loss": 0.6714,
+      "step": 5791
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.733175702563129,
+      "learning_rate": 5.67176600354143e-06,
+      "loss": 0.8003,
+      "step": 5792
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.072108180130317,
+      "learning_rate": 5.670454202906288e-06,
+      "loss": 0.7722,
+      "step": 5793
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.9407350730100346,
+      "learning_rate": 5.669142355275225e-06,
+      "loss": 0.5922,
+      "step": 5794
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.7537621502854615,
+      "learning_rate": 5.6678304607401934e-06,
+      "loss": 0.7926,
+      "step": 5795
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.5741520687458026,
+      "learning_rate": 5.6665185193931535e-06,
+      "loss": 0.6796,
+      "step": 5796
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.1724419860360955,
+      "learning_rate": 5.6652065313260675e-06,
+      "loss": 0.8818,
+      "step": 5797
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.183680157685205,
+      "learning_rate": 5.663894496630898e-06,
+      "loss": 1.269,
+      "step": 5798
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.298214517678515,
+      "learning_rate": 5.662582415399612e-06,
+      "loss": 0.6199,
+      "step": 5799
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.2366070249034333,
+      "learning_rate": 5.661270287724184e-06,
+      "loss": 0.5728,
+      "step": 5800
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.9758275887289702,
+      "learning_rate": 5.6599581136965855e-06,
+      "loss": 0.4086,
+      "step": 5801
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.251838524436833,
+      "learning_rate": 5.658645893408795e-06,
+      "loss": 0.6051,
+      "step": 5802
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.1888373776623515,
+      "learning_rate": 5.657333626952796e-06,
+      "loss": 0.893,
+      "step": 5803
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.505630791060191,
+      "learning_rate": 5.656021314420568e-06,
+      "loss": 0.8322,
+      "step": 5804
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.5915373058970728,
+      "learning_rate": 5.6547089559041025e-06,
+      "loss": 0.6709,
+      "step": 5805
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.9530410208834805,
+      "learning_rate": 5.65339655149539e-06,
+      "loss": 0.6155,
+      "step": 5806
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.0412472608721006,
+      "learning_rate": 5.652084101286419e-06,
+      "loss": 0.553,
+      "step": 5807
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7818611388354708,
+      "learning_rate": 5.6507716053691916e-06,
+      "loss": 0.2103,
+      "step": 5808
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8957158313893403,
+      "learning_rate": 5.649459063835708e-06,
+      "loss": 0.3198,
+      "step": 5809
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.033966939716265,
+      "learning_rate": 5.648146476777969e-06,
+      "loss": 0.5737,
+      "step": 5810
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.0922371158650925,
+      "learning_rate": 5.646833844287985e-06,
+      "loss": 1.1988,
+      "step": 5811
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.005342843854818,
+      "learning_rate": 5.6455211664577615e-06,
+      "loss": 0.8087,
+      "step": 5812
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.9302402031710386,
+      "learning_rate": 5.644208443379315e-06,
+      "loss": 0.5335,
+      "step": 5813
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.093569074041109,
+      "learning_rate": 5.642895675144659e-06,
+      "loss": 0.9548,
+      "step": 5814
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.564522932067168,
+      "learning_rate": 5.641582861845815e-06,
+      "loss": 1.2332,
+      "step": 5815
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.296403499703234,
+      "learning_rate": 5.640270003574804e-06,
+      "loss": 0.3742,
+      "step": 5816
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.5994409059679557,
+      "learning_rate": 5.638957100423652e-06,
+      "loss": 0.3849,
+      "step": 5817
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.550878428964589,
+      "learning_rate": 5.637644152484389e-06,
+      "loss": 1.1475,
+      "step": 5818
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.33694740793171,
+      "learning_rate": 5.6363311598490444e-06,
+      "loss": 0.8275,
+      "step": 5819
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.4951912625209527,
+      "learning_rate": 5.635018122609656e-06,
+      "loss": 0.5433,
+      "step": 5820
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.888249678559443,
+      "learning_rate": 5.633705040858262e-06,
+      "loss": 0.4817,
+      "step": 5821
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9934104155114218,
+      "learning_rate": 5.6323919146869e-06,
+      "loss": 0.4254,
+      "step": 5822
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.9285482506530185,
+      "learning_rate": 5.631078744187618e-06,
+      "loss": 0.6149,
+      "step": 5823
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.701760545520466,
+      "learning_rate": 5.629765529452463e-06,
+      "loss": 1.0096,
+      "step": 5824
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.097522815975042,
+      "learning_rate": 5.628452270573483e-06,
+      "loss": 0.8303,
+      "step": 5825
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.130182625184013,
+      "learning_rate": 5.6271389676427365e-06,
+      "loss": 0.6292,
+      "step": 5826
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.911329866629304,
+      "learning_rate": 5.625825620752277e-06,
+      "loss": 0.9716,
+      "step": 5827
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.1791773291730485,
+      "learning_rate": 5.624512229994165e-06,
+      "loss": 0.5456,
+      "step": 5828
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.022555284755302,
+      "learning_rate": 5.623198795460463e-06,
+      "loss": 0.7795,
+      "step": 5829
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.251633314084586,
+      "learning_rate": 5.621885317243238e-06,
+      "loss": 0.9052,
+      "step": 5830
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.1305021462405858,
+      "learning_rate": 5.620571795434559e-06,
+      "loss": 0.8813,
+      "step": 5831
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.017284324772185,
+      "learning_rate": 5.619258230126497e-06,
+      "loss": 0.7426,
+      "step": 5832
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.6937448691485972,
+      "learning_rate": 5.617944621411128e-06,
+      "loss": 1.0898,
+      "step": 5833
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.990890063867452,
+      "learning_rate": 5.616630969380532e-06,
+      "loss": 1.0355,
+      "step": 5834
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.1770042370908653,
+      "learning_rate": 5.615317274126787e-06,
+      "loss": 0.4759,
+      "step": 5835
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.3327796202482585,
+      "learning_rate": 5.614003535741979e-06,
+      "loss": 0.4879,
+      "step": 5836
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.616353182769598,
+      "learning_rate": 5.612689754318196e-06,
+      "loss": 0.899,
+      "step": 5837
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.9915988752037297,
+      "learning_rate": 5.611375929947528e-06,
+      "loss": 0.5136,
+      "step": 5838
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.015772413753597,
+      "learning_rate": 5.610062062722067e-06,
+      "loss": 0.9687,
+      "step": 5839
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.8811532612639135,
+      "learning_rate": 5.608748152733911e-06,
+      "loss": 0.8158,
+      "step": 5840
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.7737396464658843,
+      "learning_rate": 5.607434200075159e-06,
+      "loss": 0.5984,
+      "step": 5841
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.641508917195879,
+      "learning_rate": 5.6061202048379125e-06,
+      "loss": 0.6941,
+      "step": 5842
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.8913397236093714,
+      "learning_rate": 5.6048061671142784e-06,
+      "loss": 0.5939,
+      "step": 5843
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.8629156087235854,
+      "learning_rate": 5.603492086996362e-06,
+      "loss": 0.7298,
+      "step": 5844
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.405977959179416,
+      "learning_rate": 5.602177964576279e-06,
+      "loss": 0.5936,
+      "step": 5845
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.454230608280289,
+      "learning_rate": 5.600863799946142e-06,
+      "loss": 0.7115,
+      "step": 5846
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.032858118379392,
+      "learning_rate": 5.599549593198066e-06,
+      "loss": 0.7614,
+      "step": 5847
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8293281378783064,
+      "learning_rate": 5.598235344424172e-06,
+      "loss": 0.3763,
+      "step": 5848
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.9752145053909422,
+      "learning_rate": 5.596921053716585e-06,
+      "loss": 0.5239,
+      "step": 5849
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.8362708691057366,
+      "learning_rate": 5.59560672116743e-06,
+      "loss": 0.8695,
+      "step": 5850
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.093146394011919,
+      "learning_rate": 5.594292346868836e-06,
+      "loss": 0.3435,
+      "step": 5851
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.654731671404846,
+      "learning_rate": 5.592977930912934e-06,
+      "loss": 0.9871,
+      "step": 5852
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.566857605599453,
+      "learning_rate": 5.5916634733918604e-06,
+      "loss": 0.3263,
+      "step": 5853
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.165932241309278,
+      "learning_rate": 5.590348974397754e-06,
+      "loss": 0.7467,
+      "step": 5854
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.921543603924678,
+      "learning_rate": 5.589034434022751e-06,
+      "loss": 0.878,
+      "step": 5855
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.4275927453103914,
+      "learning_rate": 5.587719852358998e-06,
+      "loss": 0.7094,
+      "step": 5856
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.42932349890855,
+      "learning_rate": 5.586405229498641e-06,
+      "loss": 0.717,
+      "step": 5857
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.4348472387267925,
+      "learning_rate": 5.58509056553383e-06,
+      "loss": 0.3761,
+      "step": 5858
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.504870214130669,
+      "learning_rate": 5.583775860556717e-06,
+      "loss": 1.402,
+      "step": 5859
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.994133857225235,
+      "learning_rate": 5.582461114659456e-06,
+      "loss": 0.6329,
+      "step": 5860
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.549347230680468,
+      "learning_rate": 5.581146327934207e-06,
+      "loss": 0.5622,
+      "step": 5861
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.511334130940213,
+      "learning_rate": 5.579831500473129e-06,
+      "loss": 0.3354,
+      "step": 5862
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.567918857164272,
+      "learning_rate": 5.578516632368387e-06,
+      "loss": 0.7942,
+      "step": 5863
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.184583826317246,
+      "learning_rate": 5.577201723712145e-06,
+      "loss": 0.9476,
+      "step": 5864
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.20165301221488,
+      "learning_rate": 5.575886774596574e-06,
+      "loss": 0.5796,
+      "step": 5865
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.294679591078616,
+      "learning_rate": 5.574571785113848e-06,
+      "loss": 0.7292,
+      "step": 5866
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.3593511678914845,
+      "learning_rate": 5.57325675535614e-06,
+      "loss": 0.7505,
+      "step": 5867
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.96636084138706,
+      "learning_rate": 5.571941685415628e-06,
+      "loss": 0.9361,
+      "step": 5868
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.014532399988285,
+      "learning_rate": 5.570626575384494e-06,
+      "loss": 1.1275,
+      "step": 5869
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.261551974745053,
+      "learning_rate": 5.569311425354918e-06,
+      "loss": 0.4729,
+      "step": 5870
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.3052957847977624,
+      "learning_rate": 5.567996235419092e-06,
+      "loss": 0.428,
+      "step": 5871
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.925265009700976,
+      "learning_rate": 5.566681005669199e-06,
+      "loss": 0.7657,
+      "step": 5872
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.014876049573892,
+      "learning_rate": 5.565365736197434e-06,
+      "loss": 0.904,
+      "step": 5873
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.136471844997838,
+      "learning_rate": 5.564050427095993e-06,
+      "loss": 0.6476,
+      "step": 5874
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.9040217539038466,
+      "learning_rate": 5.56273507845707e-06,
+      "loss": 1.1665,
+      "step": 5875
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.9449207137628903,
+      "learning_rate": 5.561419690372869e-06,
+      "loss": 0.5986,
+      "step": 5876
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.207911771550932,
+      "learning_rate": 5.56010426293559e-06,
+      "loss": 0.3904,
+      "step": 5877
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.728565173368823,
+      "learning_rate": 5.55878879623744e-06,
+      "loss": 1.4535,
+      "step": 5878
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.8645212034986347,
+      "learning_rate": 5.557473290370626e-06,
+      "loss": 0.552,
+      "step": 5879
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 6.05123044463225,
+      "learning_rate": 5.556157745427362e-06,
+      "loss": 1.2705,
+      "step": 5880
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.530542306620343,
+      "learning_rate": 5.554842161499859e-06,
+      "loss": 0.3485,
+      "step": 5881
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.2538968371779498,
+      "learning_rate": 5.553526538680336e-06,
+      "loss": 0.4511,
+      "step": 5882
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.075364276320732,
+      "learning_rate": 5.552210877061013e-06,
+      "loss": 0.8227,
+      "step": 5883
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.914795662493378,
+      "learning_rate": 5.550895176734109e-06,
+      "loss": 0.7729,
+      "step": 5884
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.610903928405641,
+      "learning_rate": 5.549579437791851e-06,
+      "loss": 0.959,
+      "step": 5885
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9201182277283213,
+      "learning_rate": 5.548263660326466e-06,
+      "loss": 0.3125,
+      "step": 5886
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.6869314822047303,
+      "learning_rate": 5.546947844430185e-06,
+      "loss": 0.9981,
+      "step": 5887
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.61344356519624,
+      "learning_rate": 5.5456319901952395e-06,
+      "loss": 1.2744,
+      "step": 5888
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.566978454501438,
+      "learning_rate": 5.5443160977138665e-06,
+      "loss": 1.2574,
+      "step": 5889
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.499805201646772,
+      "learning_rate": 5.543000167078304e-06,
+      "loss": 1.0968,
+      "step": 5890
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.3004879299099406,
+      "learning_rate": 5.541684198380793e-06,
+      "loss": 0.7849,
+      "step": 5891
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.486444310642127,
+      "learning_rate": 5.5403681917135785e-06,
+      "loss": 0.2836,
+      "step": 5892
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.346837915740214,
+      "learning_rate": 5.539052147168903e-06,
+      "loss": 0.4871,
+      "step": 5893
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.654557335604403,
+      "learning_rate": 5.53773606483902e-06,
+      "loss": 0.5943,
+      "step": 5894
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.12528844521019,
+      "learning_rate": 5.536419944816177e-06,
+      "loss": 0.8638,
+      "step": 5895
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.340640069400724,
+      "learning_rate": 5.535103787192631e-06,
+      "loss": 0.8913,
+      "step": 5896
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.315891110519818,
+      "learning_rate": 5.53378759206064e-06,
+      "loss": 0.6577,
+      "step": 5897
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.025794917047234,
+      "learning_rate": 5.53247135951246e-06,
+      "loss": 0.851,
+      "step": 5898
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.9692707729717536,
+      "learning_rate": 5.531155089640357e-06,
+      "loss": 0.8836,
+      "step": 5899
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.5622011013882013,
+      "learning_rate": 5.529838782536591e-06,
+      "loss": 0.5705,
+      "step": 5900
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.680175791348038,
+      "learning_rate": 5.528522438293434e-06,
+      "loss": 0.7379,
+      "step": 5901
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.8857911704118298,
+      "learning_rate": 5.527206057003154e-06,
+      "loss": 0.9183,
+      "step": 5902
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.635674392930614,
+      "learning_rate": 5.525889638758024e-06,
+      "loss": 0.3244,
+      "step": 5903
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.16657170536579,
+      "learning_rate": 5.524573183650318e-06,
+      "loss": 1.1687,
+      "step": 5904
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.6165556998196884,
+      "learning_rate": 5.523256691772315e-06,
+      "loss": 0.9416,
+      "step": 5905
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.649624359146059,
+      "learning_rate": 5.521940163216296e-06,
+      "loss": 0.7414,
+      "step": 5906
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.278093059658078,
+      "learning_rate": 5.5206235980745435e-06,
+      "loss": 0.8448,
+      "step": 5907
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.904017485476948,
+      "learning_rate": 5.519306996439342e-06,
+      "loss": 0.5803,
+      "step": 5908
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.9049980254893626,
+      "learning_rate": 5.5179903584029805e-06,
+      "loss": 0.4991,
+      "step": 5909
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.362250538158217,
+      "learning_rate": 5.516673684057747e-06,
+      "loss": 0.8793,
+      "step": 5910
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.8059950212682,
+      "learning_rate": 5.515356973495939e-06,
+      "loss": 0.9024,
+      "step": 5911
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.2889491763754126,
+      "learning_rate": 5.514040226809849e-06,
+      "loss": 0.8642,
+      "step": 5912
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.014061500307177,
+      "learning_rate": 5.512723444091776e-06,
+      "loss": 0.8543,
+      "step": 5913
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.317930909121491,
+      "learning_rate": 5.5114066254340215e-06,
+      "loss": 0.3756,
+      "step": 5914
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.296118836384158,
+      "learning_rate": 5.510089770928889e-06,
+      "loss": 0.7751,
+      "step": 5915
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.17047831966923,
+      "learning_rate": 5.508772880668682e-06,
+      "loss": 0.9673,
+      "step": 5916
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.36608965530601,
+      "learning_rate": 5.507455954745712e-06,
+      "loss": 0.7936,
+      "step": 5917
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.3478663979393133,
+      "learning_rate": 5.506138993252285e-06,
+      "loss": 0.7502,
+      "step": 5918
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.997363442657447,
+      "learning_rate": 5.504821996280719e-06,
+      "loss": 0.8172,
+      "step": 5919
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.639148310727434,
+      "learning_rate": 5.50350496392333e-06,
+      "loss": 0.7692,
+      "step": 5920
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.705950501431014,
+      "learning_rate": 5.502187896272432e-06,
+      "loss": 0.8484,
+      "step": 5921
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.888276588328582,
+      "learning_rate": 5.500870793420349e-06,
+      "loss": 1.4463,
+      "step": 5922
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 6.138729432535756,
+      "learning_rate": 5.4995536554594035e-06,
+      "loss": 1.5104,
+      "step": 5923
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.424694881068357,
+      "learning_rate": 5.498236482481919e-06,
+      "loss": 0.6552,
+      "step": 5924
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.7475488823187413,
+      "learning_rate": 5.496919274580226e-06,
+      "loss": 0.449,
+      "step": 5925
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.8115374734444227,
+      "learning_rate": 5.495602031846655e-06,
+      "loss": 0.8435,
+      "step": 5926
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.8308256770443916,
+      "learning_rate": 5.494284754373538e-06,
+      "loss": 0.56,
+      "step": 5927
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.414466118736038,
+      "learning_rate": 5.492967442253211e-06,
+      "loss": 0.9561,
+      "step": 5928
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.5556606784233122,
+      "learning_rate": 5.491650095578013e-06,
+      "loss": 0.5788,
+      "step": 5929
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.00760104064909,
+      "learning_rate": 5.4903327144402814e-06,
+      "loss": 0.9597,
+      "step": 5930
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.699846439981498,
+      "learning_rate": 5.489015298932362e-06,
+      "loss": 0.8504,
+      "step": 5931
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.5765579870648723,
+      "learning_rate": 5.487697849146596e-06,
+      "loss": 0.7532,
+      "step": 5932
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.194937056636086,
+      "learning_rate": 5.4863803651753345e-06,
+      "loss": 0.5019,
+      "step": 5933
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.387994818204489,
+      "learning_rate": 5.485062847110927e-06,
+      "loss": 0.3907,
+      "step": 5934
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.7829552978663625,
+      "learning_rate": 5.483745295045724e-06,
+      "loss": 0.8088,
+      "step": 5935
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.194828419721978,
+      "learning_rate": 5.48242770907208e-06,
+      "loss": 0.3704,
+      "step": 5936
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.16157216863073,
+      "learning_rate": 5.481110089282355e-06,
+      "loss": 1.063,
+      "step": 5937
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.462414446566765,
+      "learning_rate": 5.4797924357689045e-06,
+      "loss": 0.9265,
+      "step": 5938
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.1440044922376944,
+      "learning_rate": 5.478474748624095e-06,
+      "loss": 0.6075,
+      "step": 5939
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.715496593630727,
+      "learning_rate": 5.477157027940286e-06,
+      "loss": 0.5834,
+      "step": 5940
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.899814377063645,
+      "learning_rate": 5.475839273809846e-06,
+      "loss": 1.2268,
+      "step": 5941
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.6442247505119956,
+      "learning_rate": 5.474521486325145e-06,
+      "loss": 0.5574,
+      "step": 5942
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.2709048047438967,
+      "learning_rate": 5.473203665578553e-06,
+      "loss": 0.411,
+      "step": 5943
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.2479962450882995,
+      "learning_rate": 5.471885811662442e-06,
+      "loss": 0.5546,
+      "step": 5944
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.4924594522157797,
+      "learning_rate": 5.470567924669189e-06,
+      "loss": 0.9061,
+      "step": 5945
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.707626979273108,
+      "learning_rate": 5.469250004691174e-06,
+      "loss": 1.1444,
+      "step": 5946
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.7467684133691357,
+      "learning_rate": 5.467932051820776e-06,
+      "loss": 0.8675,
+      "step": 5947
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.129339723411537,
+      "learning_rate": 5.466614066150375e-06,
+      "loss": 1.0813,
+      "step": 5948
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.2643072965740694,
+      "learning_rate": 5.465296047772362e-06,
+      "loss": 0.6233,
+      "step": 5949
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.1355986530047826,
+      "learning_rate": 5.463977996779119e-06,
+      "loss": 0.8856,
+      "step": 5950
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.443501834756944,
+      "learning_rate": 5.4626599132630384e-06,
+      "loss": 1.0806,
+      "step": 5951
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.627958424176721,
+      "learning_rate": 5.46134179731651e-06,
+      "loss": 0.659,
+      "step": 5952
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.309435183852325,
+      "learning_rate": 5.4600236490319305e-06,
+      "loss": 0.7406,
+      "step": 5953
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.533187208048489,
+      "learning_rate": 5.458705468501696e-06,
+      "loss": 0.6347,
+      "step": 5954
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5210903979167285,
+      "learning_rate": 5.457387255818204e-06,
+      "loss": 0.39,
+      "step": 5955
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.4005814640182117,
+      "learning_rate": 5.456069011073854e-06,
+      "loss": 0.753,
+      "step": 5956
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.28949496031527,
+      "learning_rate": 5.454750734361054e-06,
+      "loss": 1.3707,
+      "step": 5957
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.28649999961873,
+      "learning_rate": 5.453432425772205e-06,
+      "loss": 0.6708,
+      "step": 5958
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.050799746874854,
+      "learning_rate": 5.4521140853997166e-06,
+      "loss": 0.8119,
+      "step": 5959
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.9711602718109136,
+      "learning_rate": 5.450795713335999e-06,
+      "loss": 0.8896,
+      "step": 5960
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.5967403665073654,
+      "learning_rate": 5.449477309673462e-06,
+      "loss": 0.4791,
+      "step": 5961
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.957273264260848,
+      "learning_rate": 5.4481588745045245e-06,
+      "loss": 1.1205,
+      "step": 5962
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.5226503055402056,
+      "learning_rate": 5.446840407921599e-06,
+      "loss": 0.3371,
+      "step": 5963
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.6492716318407035,
+      "learning_rate": 5.445521910017104e-06,
+      "loss": 0.7335,
+      "step": 5964
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.534855989455732,
+      "learning_rate": 5.444203380883464e-06,
+      "loss": 0.5681,
+      "step": 5965
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.8778627977130533,
+      "learning_rate": 5.442884820613099e-06,
+      "loss": 0.5447,
+      "step": 5966
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.641037162912989,
+      "learning_rate": 5.441566229298436e-06,
+      "loss": 0.8469,
+      "step": 5967
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.211550326231087,
+      "learning_rate": 5.440247607031901e-06,
+      "loss": 1.1026,
+      "step": 5968
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.769302466784294,
+      "learning_rate": 5.438928953905926e-06,
+      "loss": 1.2086,
+      "step": 5969
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.350234579846934,
+      "learning_rate": 5.437610270012943e-06,
+      "loss": 0.889,
+      "step": 5970
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.712562852846676,
+      "learning_rate": 5.436291555445383e-06,
+      "loss": 1.0308,
+      "step": 5971
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.5171631949861393,
+      "learning_rate": 5.434972810295683e-06,
+      "loss": 0.671,
+      "step": 5972
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.662786074086821,
+      "learning_rate": 5.433654034656283e-06,
+      "loss": 0.932,
+      "step": 5973
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.598717293441438,
+      "learning_rate": 5.4323352286196215e-06,
+      "loss": 0.695,
+      "step": 5974
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.191268318394292,
+      "learning_rate": 5.431016392278142e-06,
+      "loss": 1.0992,
+      "step": 5975
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7364795990950455,
+      "learning_rate": 5.429697525724289e-06,
+      "loss": 0.3564,
+      "step": 5976
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.4752863215518532,
+      "learning_rate": 5.428378629050511e-06,
+      "loss": 0.5215,
+      "step": 5977
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.6326295764189065,
+      "learning_rate": 5.427059702349255e-06,
+      "loss": 0.7813,
+      "step": 5978
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.4500554392551557,
+      "learning_rate": 5.425740745712972e-06,
+      "loss": 0.5694,
+      "step": 5979
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.664928275209959,
+      "learning_rate": 5.4244217592341165e-06,
+      "loss": 0.7722,
+      "step": 5980
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.2221681323965026,
+      "learning_rate": 5.423102743005141e-06,
+      "loss": 0.4881,
+      "step": 5981
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.414440890088172,
+      "learning_rate": 5.421783697118506e-06,
+      "loss": 1.3149,
+      "step": 5982
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.183253117450436,
+      "learning_rate": 5.420464621666669e-06,
+      "loss": 1.2179,
+      "step": 5983
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.7352526455230937,
+      "learning_rate": 5.4191455167420905e-06,
+      "loss": 0.581,
+      "step": 5984
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.8724811227526703,
+      "learning_rate": 5.417826382437238e-06,
+      "loss": 0.9634,
+      "step": 5985
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 6.3110195488406875,
+      "learning_rate": 5.4165072188445734e-06,
+      "loss": 1.7039,
+      "step": 5986
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.598222072883564,
+      "learning_rate": 5.415188026056565e-06,
+      "loss": 0.6402,
+      "step": 5987
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.553321277530451,
+      "learning_rate": 5.413868804165682e-06,
+      "loss": 1.097,
+      "step": 5988
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.4588707286392295,
+      "learning_rate": 5.412549553264399e-06,
+      "loss": 0.6014,
+      "step": 5989
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.733512858828318,
+      "learning_rate": 5.411230273445186e-06,
+      "loss": 1.325,
+      "step": 5990
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.39984972508648,
+      "learning_rate": 5.409910964800522e-06,
+      "loss": 0.8622,
+      "step": 5991
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.2789657345602468,
+      "learning_rate": 5.4085916274228825e-06,
+      "loss": 0.5902,
+      "step": 5992
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.1865544219152837,
+      "learning_rate": 5.407272261404748e-06,
+      "loss": 0.4441,
+      "step": 5993
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.593265950272095,
+      "learning_rate": 5.405952866838602e-06,
+      "loss": 0.384,
+      "step": 5994
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.557730997750645,
+      "learning_rate": 5.4046334438169245e-06,
+      "loss": 1.0818,
+      "step": 5995
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.413490038438569,
+      "learning_rate": 5.403313992432203e-06,
+      "loss": 0.5347,
+      "step": 5996
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 6.295959641215288,
+      "learning_rate": 5.401994512776928e-06,
+      "loss": 1.3955,
+      "step": 5997
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.796602048676265,
+      "learning_rate": 5.4006750049435864e-06,
+      "loss": 0.3992,
+      "step": 5998
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.21319372425049,
+      "learning_rate": 5.3993554690246695e-06,
+      "loss": 1.3535,
+      "step": 5999
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.4363423550782217,
+      "learning_rate": 5.398035905112675e-06,
+      "loss": 0.9095,
+      "step": 6000
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.160294196177696,
+      "learning_rate": 5.396716313300094e-06,
+      "loss": 0.6002,
+      "step": 6001
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.867758171193,
+      "learning_rate": 5.395396693679427e-06,
+      "loss": 0.987,
+      "step": 6002
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9501210659651882,
+      "learning_rate": 5.394077046343172e-06,
+      "loss": 0.4098,
+      "step": 6003
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.628940380104001,
+      "learning_rate": 5.39275737138383e-06,
+      "loss": 0.8241,
+      "step": 6004
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.718599098761617,
+      "learning_rate": 5.3914376688939065e-06,
+      "loss": 0.919,
+      "step": 6005
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.045665200502083,
+      "learning_rate": 5.390117938965906e-06,
+      "loss": 0.7792,
+      "step": 6006
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.8960198644132213,
+      "learning_rate": 5.388798181692335e-06,
+      "loss": 0.8319,
+      "step": 6007
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.9175820202923095,
+      "learning_rate": 5.387478397165704e-06,
+      "loss": 1.3319,
+      "step": 6008
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.8390832555859196,
+      "learning_rate": 5.386158585478525e-06,
+      "loss": 0.6011,
+      "step": 6009
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.63872663669692,
+      "learning_rate": 5.384838746723308e-06,
+      "loss": 0.6307,
+      "step": 6010
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.2631288297706647,
+      "learning_rate": 5.383518880992571e-06,
+      "loss": 0.45,
+      "step": 6011
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.06846177820315,
+      "learning_rate": 5.382198988378829e-06,
+      "loss": 0.9493,
+      "step": 6012
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.471610145989451,
+      "learning_rate": 5.380879068974599e-06,
+      "loss": 0.426,
+      "step": 6013
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.05154167912491,
+      "learning_rate": 5.3795591228724065e-06,
+      "loss": 0.1397,
+      "step": 6014
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.269178799749522,
+      "learning_rate": 5.37823915016477e-06,
+      "loss": 0.868,
+      "step": 6015
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.6985714026759426,
+      "learning_rate": 5.376919150944218e-06,
+      "loss": 0.3655,
+      "step": 6016
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.5445661109934754,
+      "learning_rate": 5.375599125303272e-06,
+      "loss": 0.5883,
+      "step": 6017
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.865373846194751,
+      "learning_rate": 5.3742790733344604e-06,
+      "loss": 0.3618,
+      "step": 6018
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.484862095770465,
+      "learning_rate": 5.372958995130315e-06,
+      "loss": 0.5057,
+      "step": 6019
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.6606923982642225,
+      "learning_rate": 5.37163889078337e-06,
+      "loss": 0.748,
+      "step": 6020
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.877526635623892,
+      "learning_rate": 5.3703187603861525e-06,
+      "loss": 0.7487,
+      "step": 6021
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.2999462796503423,
+      "learning_rate": 5.368998604031202e-06,
+      "loss": 0.8011,
+      "step": 6022
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.3706799983308726,
+      "learning_rate": 5.367678421811058e-06,
+      "loss": 0.4402,
+      "step": 6023
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.3353806291212935,
+      "learning_rate": 5.366358213818256e-06,
+      "loss": 0.4826,
+      "step": 6024
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.558573600171047,
+      "learning_rate": 5.365037980145337e-06,
+      "loss": 1.0699,
+      "step": 6025
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.7952480984435892,
+      "learning_rate": 5.3637177208848435e-06,
+      "loss": 0.5923,
+      "step": 6026
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.223010186727015,
+      "learning_rate": 5.362397436129321e-06,
+      "loss": 0.6329,
+      "step": 6027
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.1927538308642323,
+      "learning_rate": 5.361077125971316e-06,
+      "loss": 0.5016,
+      "step": 6028
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.116713591518483,
+      "learning_rate": 5.359756790503376e-06,
+      "loss": 0.6132,
+      "step": 6029
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.433978801843281,
+      "learning_rate": 5.358436429818049e-06,
+      "loss": 0.6,
+      "step": 6030
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.227780056767419,
+      "learning_rate": 5.357116044007889e-06,
+      "loss": 0.2047,
+      "step": 6031
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.257502964558977,
+      "learning_rate": 5.35579563316545e-06,
+      "loss": 0.3571,
+      "step": 6032
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.06788660150286,
+      "learning_rate": 5.354475197383284e-06,
+      "loss": 1.581,
+      "step": 6033
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.744145590496548,
+      "learning_rate": 5.353154736753951e-06,
+      "loss": 0.5749,
+      "step": 6034
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.09242606935554,
+      "learning_rate": 5.351834251370006e-06,
+      "loss": 0.9254,
+      "step": 6035
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.771394588947226,
+      "learning_rate": 5.350513741324011e-06,
+      "loss": 0.3295,
+      "step": 6036
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.021615405626378,
+      "learning_rate": 5.349193206708529e-06,
+      "loss": 1.1829,
+      "step": 6037
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.202720037583286,
+      "learning_rate": 5.347872647616122e-06,
+      "loss": 1.2606,
+      "step": 6038
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.6572563355449743,
+      "learning_rate": 5.3465520641393585e-06,
+      "loss": 0.7677,
+      "step": 6039
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.8552302348051444,
+      "learning_rate": 5.345231456370802e-06,
+      "loss": 0.7232,
+      "step": 6040
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.806366339065954,
+      "learning_rate": 5.3439108244030234e-06,
+      "loss": 1.2052,
+      "step": 6041
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.29929892416941,
+      "learning_rate": 5.342590168328592e-06,
+      "loss": 0.7961,
+      "step": 6042
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.1248953723976363,
+      "learning_rate": 5.341269488240082e-06,
+      "loss": 0.4013,
+      "step": 6043
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.355464824501211,
+      "learning_rate": 5.3399487842300646e-06,
+      "loss": 1.2155,
+      "step": 6044
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.124892612547487,
+      "learning_rate": 5.338628056391118e-06,
+      "loss": 1.2099,
+      "step": 6045
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.18600294308645,
+      "learning_rate": 5.337307304815817e-06,
+      "loss": 0.343,
+      "step": 6046
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.844076954517568,
+      "learning_rate": 5.335986529596743e-06,
+      "loss": 0.6046,
+      "step": 6047
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 6.3044142195789785,
+      "learning_rate": 5.334665730826476e-06,
+      "loss": 1.1133,
+      "step": 6048
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3205392762697725,
+      "learning_rate": 5.333344908597597e-06,
+      "loss": 0.1942,
+      "step": 6049
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.773085470119928,
+      "learning_rate": 5.332024063002691e-06,
+      "loss": 0.9705,
+      "step": 6050
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.0438150380204245,
+      "learning_rate": 5.330703194134342e-06,
+      "loss": 1.1824,
+      "step": 6051
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.62512093898664,
+      "learning_rate": 5.32938230208514e-06,
+      "loss": 1.051,
+      "step": 6052
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.944231217257914,
+      "learning_rate": 5.328061386947671e-06,
+      "loss": 0.9748,
+      "step": 6053
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.627745480283834,
+      "learning_rate": 5.326740448814527e-06,
+      "loss": 0.8678,
+      "step": 6054
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.288780733394661,
+      "learning_rate": 5.325419487778299e-06,
+      "loss": 0.7457,
+      "step": 6055
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.2709540185919006,
+      "learning_rate": 5.324098503931581e-06,
+      "loss": 0.4442,
+      "step": 6056
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.528635346577336,
+      "learning_rate": 5.3227774973669695e-06,
+      "loss": 1.0207,
+      "step": 6057
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.666482261079773,
+      "learning_rate": 5.3214564681770585e-06,
+      "loss": 0.9279,
+      "step": 6058
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.9690261119748724,
+      "learning_rate": 5.320135416454448e-06,
+      "loss": 0.9664,
+      "step": 6059
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.266908773209643,
+      "learning_rate": 5.3188143422917405e-06,
+      "loss": 0.8157,
+      "step": 6060
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.098677175359244,
+      "learning_rate": 5.317493245781533e-06,
+      "loss": 0.5497,
+      "step": 6061
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.276565373552879,
+      "learning_rate": 5.316172127016431e-06,
+      "loss": 0.7332,
+      "step": 6062
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2610912101476668,
+      "learning_rate": 5.31485098608904e-06,
+      "loss": 0.2074,
+      "step": 6063
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.128417890513426,
+      "learning_rate": 5.313529823091964e-06,
+      "loss": 0.3317,
+      "step": 6064
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.5045172622483607,
+      "learning_rate": 5.312208638117812e-06,
+      "loss": 1.0077,
+      "step": 6065
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.508505709824206,
+      "learning_rate": 5.310887431259194e-06,
+      "loss": 0.9134,
+      "step": 6066
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.28128242779539,
+      "learning_rate": 5.309566202608719e-06,
+      "loss": 0.3332,
+      "step": 6067
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.8888855207366695,
+      "learning_rate": 5.3082449522590005e-06,
+      "loss": 1.0126,
+      "step": 6068
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.5042120031476163,
+      "learning_rate": 5.306923680302654e-06,
+      "loss": 1.0205,
+      "step": 6069
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.3608982135572427,
+      "learning_rate": 5.30560238683229e-06,
+      "loss": 0.6756,
+      "step": 6070
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.1357296635241734,
+      "learning_rate": 5.304281071940532e-06,
+      "loss": 0.6604,
+      "step": 6071
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.487126975007773,
+      "learning_rate": 5.302959735719995e-06,
+      "loss": 0.7766,
+      "step": 6072
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.043631275556565,
+      "learning_rate": 5.301638378263296e-06,
+      "loss": 0.8359,
+      "step": 6073
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.251135154978727,
+      "learning_rate": 5.300316999663062e-06,
+      "loss": 0.7924,
+      "step": 6074
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.0227537491324274,
+      "learning_rate": 5.298995600011912e-06,
+      "loss": 0.5663,
+      "step": 6075
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.749566650478806,
+      "learning_rate": 5.2976741794024725e-06,
+      "loss": 0.4459,
+      "step": 6076
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.2032780087289554,
+      "learning_rate": 5.296352737927368e-06,
+      "loss": 0.2584,
+      "step": 6077
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.4978819867980775,
+      "learning_rate": 5.295031275679226e-06,
+      "loss": 0.5388,
+      "step": 6078
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.7285154432660215,
+      "learning_rate": 5.293709792750677e-06,
+      "loss": 0.7644,
+      "step": 6079
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.7192776482024046,
+      "learning_rate": 5.292388289234349e-06,
+      "loss": 0.9788,
+      "step": 6080
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.711842440350777,
+      "learning_rate": 5.2910667652228735e-06,
+      "loss": 0.7359,
+      "step": 6081
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.673850306310747,
+      "learning_rate": 5.289745220808885e-06,
+      "loss": 0.924,
+      "step": 6082
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.4479694632064355,
+      "learning_rate": 5.288423656085018e-06,
+      "loss": 0.8344,
+      "step": 6083
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.3718104486576355,
+      "learning_rate": 5.287102071143907e-06,
+      "loss": 0.3449,
+      "step": 6084
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9814963008856785,
+      "learning_rate": 5.28578046607819e-06,
+      "loss": 0.2176,
+      "step": 6085
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.0309721955972324,
+      "learning_rate": 5.284458840980507e-06,
+      "loss": 0.7822,
+      "step": 6086
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.8376790697897962,
+      "learning_rate": 5.283137195943499e-06,
+      "loss": 0.6084,
+      "step": 6087
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.032045832585798,
+      "learning_rate": 5.281815531059803e-06,
+      "loss": 0.7281,
+      "step": 6088
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.844734848963236,
+      "learning_rate": 5.280493846422066e-06,
+      "loss": 0.5976,
+      "step": 6089
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.2656608970985594,
+      "learning_rate": 5.27917214212293e-06,
+      "loss": 1.0984,
+      "step": 6090
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.5982297716675875,
+      "learning_rate": 5.2778504182550436e-06,
+      "loss": 1.1228,
+      "step": 6091
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.9160003755177946,
+      "learning_rate": 5.27652867491105e-06,
+      "loss": 0.9268,
+      "step": 6092
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.7159953624083575,
+      "learning_rate": 5.2752069121836e-06,
+      "loss": 0.7194,
+      "step": 6093
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.542153783435038,
+      "learning_rate": 5.273885130165345e-06,
+      "loss": 0.7737,
+      "step": 6094
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.4585385374304933,
+      "learning_rate": 5.2725633289489345e-06,
+      "loss": 0.5026,
+      "step": 6095
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.0917108234195774,
+      "learning_rate": 5.2712415086270185e-06,
+      "loss": 0.655,
+      "step": 6096
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.3591110112259255,
+      "learning_rate": 5.2699196692922546e-06,
+      "loss": 0.8652,
+      "step": 6097
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.300957812764537,
+      "learning_rate": 5.268597811037296e-06,
+      "loss": 0.538,
+      "step": 6098
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.2486258407781285,
+      "learning_rate": 5.2672759339547995e-06,
+      "loss": 0.6235,
+      "step": 6099
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.393035122156463,
+      "learning_rate": 5.265954038137424e-06,
+      "loss": 0.7664,
+      "step": 6100
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.9881075428557233,
+      "learning_rate": 5.264632123677827e-06,
+      "loss": 0.5406,
+      "step": 6101
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.355037993476339,
+      "learning_rate": 5.2633101906686715e-06,
+      "loss": 1.1116,
+      "step": 6102
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8713903279823412,
+      "learning_rate": 5.261988239202617e-06,
+      "loss": 0.396,
+      "step": 6103
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.868093306142094,
+      "learning_rate": 5.260666269372327e-06,
+      "loss": 0.3755,
+      "step": 6104
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.782449515322387,
+      "learning_rate": 5.259344281270464e-06,
+      "loss": 0.8487,
+      "step": 6105
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.327406827144394,
+      "learning_rate": 5.258022274989698e-06,
+      "loss": 0.8965,
+      "step": 6106
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.280276130996807,
+      "learning_rate": 5.256700250622692e-06,
+      "loss": 1.35,
+      "step": 6107
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.8376123252688825,
+      "learning_rate": 5.2553782082621155e-06,
+      "loss": 0.6479,
+      "step": 6108
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.390103833093076,
+      "learning_rate": 5.2540561480006395e-06,
+      "loss": 0.9966,
+      "step": 6109
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.1779116496556297,
+      "learning_rate": 5.252734069930933e-06,
+      "loss": 0.5502,
+      "step": 6110
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.844221005521417,
+      "learning_rate": 5.251411974145667e-06,
+      "loss": 1.1597,
+      "step": 6111
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.0666922562733765,
+      "learning_rate": 5.250089860737516e-06,
+      "loss": 0.7088,
+      "step": 6112
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.681093129070991,
+      "learning_rate": 5.248767729799153e-06,
+      "loss": 0.567,
+      "step": 6113
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.172353609353667,
+      "learning_rate": 5.247445581423257e-06,
+      "loss": 1.5891,
+      "step": 6114
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.01038345251461,
+      "learning_rate": 5.246123415702502e-06,
+      "loss": 1.059,
+      "step": 6115
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.3530754017556252,
+      "learning_rate": 5.244801232729566e-06,
+      "loss": 0.9773,
+      "step": 6116
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.0136245644424715,
+      "learning_rate": 5.2434790325971295e-06,
+      "loss": 0.901,
+      "step": 6117
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.13231353398788,
+      "learning_rate": 5.242156815397873e-06,
+      "loss": 0.3978,
+      "step": 6118
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7810045992983448,
+      "learning_rate": 5.240834581224476e-06,
+      "loss": 0.3124,
+      "step": 6119
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.194978801765924,
+      "learning_rate": 5.239512330169625e-06,
+      "loss": 0.5603,
+      "step": 6120
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.131501788834688,
+      "learning_rate": 5.238190062326001e-06,
+      "loss": 1.1013,
+      "step": 6121
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.673214352109646,
+      "learning_rate": 5.23686777778629e-06,
+      "loss": 0.3919,
+      "step": 6122
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.9256462722989505,
+      "learning_rate": 5.235545476643179e-06,
+      "loss": 0.8098,
+      "step": 6123
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.373331473673059,
+      "learning_rate": 5.234223158989354e-06,
+      "loss": 1.2181,
+      "step": 6124
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.1702997037447873,
+      "learning_rate": 5.232900824917507e-06,
+      "loss": 0.5741,
+      "step": 6125
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.186281271164072,
+      "learning_rate": 5.231578474520324e-06,
+      "loss": 0.8313,
+      "step": 6126
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.9746730564601425,
+      "learning_rate": 5.230256107890499e-06,
+      "loss": 0.8272,
+      "step": 6127
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.1480789590876506,
+      "learning_rate": 5.228933725120722e-06,
+      "loss": 0.2823,
+      "step": 6128
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.664596301551104,
+      "learning_rate": 5.227611326303688e-06,
+      "loss": 1.1076,
+      "step": 6129
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.57939184414976,
+      "learning_rate": 5.2262889115320895e-06,
+      "loss": 0.8099,
+      "step": 6130
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.7668009677088365,
+      "learning_rate": 5.224966480898624e-06,
+      "loss": 0.772,
+      "step": 6131
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.9670168505602925,
+      "learning_rate": 5.2236440344959875e-06,
+      "loss": 1.3492,
+      "step": 6132
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.797135588049277,
+      "learning_rate": 5.2223215724168764e-06,
+      "loss": 0.8794,
+      "step": 6133
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.8553143869273296,
+      "learning_rate": 5.220999094753992e-06,
+      "loss": 0.4619,
+      "step": 6134
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.0458033099846755,
+      "learning_rate": 5.2196766016000325e-06,
+      "loss": 0.7381,
+      "step": 6135
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.306552937733528,
+      "learning_rate": 5.218354093047697e-06,
+      "loss": 0.6208,
+      "step": 6136
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.4600195846023656,
+      "learning_rate": 5.217031569189692e-06,
+      "loss": 0.6653,
+      "step": 6137
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.174761429229127,
+      "learning_rate": 5.215709030118718e-06,
+      "loss": 0.1854,
+      "step": 6138
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.469828478962,
+      "learning_rate": 5.21438647592748e-06,
+      "loss": 0.425,
+      "step": 6139
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.028015358329215,
+      "learning_rate": 5.213063906708683e-06,
+      "loss": 0.7259,
+      "step": 6140
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.181936284915569,
+      "learning_rate": 5.211741322555034e-06,
+      "loss": 0.6182,
+      "step": 6141
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.5836489621305465,
+      "learning_rate": 5.2104187235592395e-06,
+      "loss": 0.797,
+      "step": 6142
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.5992488846652266,
+      "learning_rate": 5.209096109814008e-06,
+      "loss": 0.6928,
+      "step": 6143
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.4981195911857967,
+      "learning_rate": 5.207773481412049e-06,
+      "loss": 0.8769,
+      "step": 6144
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.6780818571798677,
+      "learning_rate": 5.206450838446072e-06,
+      "loss": 0.5248,
+      "step": 6145
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.7670184348930027,
+      "learning_rate": 5.205128181008791e-06,
+      "loss": 0.9068,
+      "step": 6146
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.5566362645193568,
+      "learning_rate": 5.203805509192917e-06,
+      "loss": 0.6849,
+      "step": 6147
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.4882825346129476,
+      "learning_rate": 5.202482823091165e-06,
+      "loss": 0.8195,
+      "step": 6148
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.611746586855792,
+      "learning_rate": 5.201160122796247e-06,
+      "loss": 0.3935,
+      "step": 6149
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.2290167217233137,
+      "learning_rate": 5.19983740840088e-06,
+      "loss": 0.3995,
+      "step": 6150
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.734378775660485,
+      "learning_rate": 5.198514679997782e-06,
+      "loss": 0.6287,
+      "step": 6151
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.602912475170529,
+      "learning_rate": 5.197191937679667e-06,
+      "loss": 0.9427,
+      "step": 6152
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.735567133731998,
+      "learning_rate": 5.195869181539255e-06,
+      "loss": 0.8277,
+      "step": 6153
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 6.010553948935826,
+      "learning_rate": 5.194546411669267e-06,
+      "loss": 0.8849,
+      "step": 6154
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.871751178871961,
+      "learning_rate": 5.193223628162421e-06,
+      "loss": 0.8736,
+      "step": 6155
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9287174402242686,
+      "learning_rate": 5.19190083111144e-06,
+      "loss": 0.1051,
+      "step": 6156
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.7633676823644064,
+      "learning_rate": 5.190578020609047e-06,
+      "loss": 0.7295,
+      "step": 6157
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.490580213260556,
+      "learning_rate": 5.189255196747964e-06,
+      "loss": 0.7785,
+      "step": 6158
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.2016378207556,
+      "learning_rate": 5.187932359620914e-06,
+      "loss": 0.7724,
+      "step": 6159
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.911124554425165,
+      "learning_rate": 5.186609509320625e-06,
+      "loss": 0.5837,
+      "step": 6160
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.103299272865019,
+      "learning_rate": 5.18528664593982e-06,
+      "loss": 0.3336,
+      "step": 6161
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.763653361213138,
+      "learning_rate": 5.183963769571227e-06,
+      "loss": 0.8515,
+      "step": 6162
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.0625643276009,
+      "learning_rate": 5.1826408803075765e-06,
+      "loss": 0.4117,
+      "step": 6163
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.418750418805527,
+      "learning_rate": 5.181317978241595e-06,
+      "loss": 1.5736,
+      "step": 6164
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.5187906029484766,
+      "learning_rate": 5.179995063466011e-06,
+      "loss": 0.7462,
+      "step": 6165
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.098670287517148,
+      "learning_rate": 5.178672136073558e-06,
+      "loss": 0.7015,
+      "step": 6166
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.194162905885539,
+      "learning_rate": 5.177349196156964e-06,
+      "loss": 0.7261,
+      "step": 6167
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.153913117205848,
+      "learning_rate": 5.1760262438089636e-06,
+      "loss": 0.8085,
+      "step": 6168
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.997762346294353,
+      "learning_rate": 5.174703279122291e-06,
+      "loss": 0.8826,
+      "step": 6169
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.476328227990508,
+      "learning_rate": 5.173380302189676e-06,
+      "loss": 0.9064,
+      "step": 6170
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.570371625510055,
+      "learning_rate": 5.172057313103859e-06,
+      "loss": 1.4817,
+      "step": 6171
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9306079358977413,
+      "learning_rate": 5.170734311957572e-06,
+      "loss": 0.3788,
+      "step": 6172
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.549389996259477,
+      "learning_rate": 5.169411298843554e-06,
+      "loss": 0.7813,
+      "step": 6173
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.499649837486986,
+      "learning_rate": 5.16808827385454e-06,
+      "loss": 0.7076,
+      "step": 6174
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.989391726629326,
+      "learning_rate": 5.16676523708327e-06,
+      "loss": 0.9742,
+      "step": 6175
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.67099707567816,
+      "learning_rate": 5.165442188622482e-06,
+      "loss": 0.879,
+      "step": 6176
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.558072060843167,
+      "learning_rate": 5.164119128564917e-06,
+      "loss": 0.8436,
+      "step": 6177
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8061979255149667,
+      "learning_rate": 5.162796057003316e-06,
+      "loss": 0.4058,
+      "step": 6178
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.1751579642441854,
+      "learning_rate": 5.161472974030418e-06,
+      "loss": 0.7461,
+      "step": 6179
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.176957644498074,
+      "learning_rate": 5.1601498797389695e-06,
+      "loss": 0.758,
+      "step": 6180
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.112975382854141,
+      "learning_rate": 5.158826774221711e-06,
+      "loss": 0.5926,
+      "step": 6181
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.8319967477483363,
+      "learning_rate": 5.157503657571386e-06,
+      "loss": 0.6073,
+      "step": 6182
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9127903859105102,
+      "learning_rate": 5.156180529880741e-06,
+      "loss": 0.1528,
+      "step": 6183
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.757988523389512,
+      "learning_rate": 5.15485739124252e-06,
+      "loss": 0.8195,
+      "step": 6184
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.149178222477253,
+      "learning_rate": 5.153534241749468e-06,
+      "loss": 1.2277,
+      "step": 6185
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.1744426702739177,
+      "learning_rate": 5.152211081494336e-06,
+      "loss": 0.5887,
+      "step": 6186
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.2991471817823994,
+      "learning_rate": 5.150887910569868e-06,
+      "loss": 0.3999,
+      "step": 6187
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.973632128969777,
+      "learning_rate": 5.149564729068816e-06,
+      "loss": 0.6487,
+      "step": 6188
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.4212412938391443,
+      "learning_rate": 5.148241537083928e-06,
+      "loss": 0.5501,
+      "step": 6189
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.9912273125515885,
+      "learning_rate": 5.146918334707952e-06,
+      "loss": 0.9781,
+      "step": 6190
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7338531351318034,
+      "learning_rate": 5.145595122033641e-06,
+      "loss": 0.5279,
+      "step": 6191
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.797698382201967,
+      "learning_rate": 5.144271899153743e-06,
+      "loss": 1.1141,
+      "step": 6192
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.051888552737093,
+      "learning_rate": 5.142948666161015e-06,
+      "loss": 0.962,
+      "step": 6193
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.695575880327883,
+      "learning_rate": 5.1416254231482075e-06,
+      "loss": 0.7048,
+      "step": 6194
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.149596836313373,
+      "learning_rate": 5.140302170208073e-06,
+      "loss": 0.7707,
+      "step": 6195
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.8069664635200313,
+      "learning_rate": 5.138978907433368e-06,
+      "loss": 0.4488,
+      "step": 6196
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8215426670677781,
+      "learning_rate": 5.137655634916847e-06,
+      "loss": 0.3643,
+      "step": 6197
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.3808154312253578,
+      "learning_rate": 5.136332352751264e-06,
+      "loss": 0.6463,
+      "step": 6198
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.444167956178485,
+      "learning_rate": 5.1350090610293765e-06,
+      "loss": 0.7728,
+      "step": 6199
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.866901181736248,
+      "learning_rate": 5.133685759843942e-06,
+      "loss": 0.7861,
+      "step": 6200
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.2801548514287724,
+      "learning_rate": 5.132362449287717e-06,
+      "loss": 0.8588,
+      "step": 6201
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.5811069303386147,
+      "learning_rate": 5.13103912945346e-06,
+      "loss": 0.9135,
+      "step": 6202
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.447064912064372,
+      "learning_rate": 5.129715800433931e-06,
+      "loss": 0.6497,
+      "step": 6203
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.5511670926001306,
+      "learning_rate": 5.128392462321889e-06,
+      "loss": 0.621,
+      "step": 6204
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.2851645891357455,
+      "learning_rate": 5.127069115210094e-06,
+      "loss": 0.6323,
+      "step": 6205
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.0718547433639816,
+      "learning_rate": 5.125745759191307e-06,
+      "loss": 0.4425,
+      "step": 6206
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.205788564576086,
+      "learning_rate": 5.124422394358289e-06,
+      "loss": 1.1922,
+      "step": 6207
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.570194497022711,
+      "learning_rate": 5.123099020803803e-06,
+      "loss": 1.1588,
+      "step": 6208
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.123753768682675,
+      "learning_rate": 5.1217756386206115e-06,
+      "loss": 0.9596,
+      "step": 6209
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.1374146278778574,
+      "learning_rate": 5.120452247901477e-06,
+      "loss": 0.7494,
+      "step": 6210
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.131343355026282,
+      "learning_rate": 5.119128848739165e-06,
+      "loss": 0.8035,
+      "step": 6211
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.085041690104158,
+      "learning_rate": 5.11780544122644e-06,
+      "loss": 0.8113,
+      "step": 6212
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.71463247425855,
+      "learning_rate": 5.116482025456066e-06,
+      "loss": 0.6242,
+      "step": 6213
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.9995805170970105,
+      "learning_rate": 5.115158601520807e-06,
+      "loss": 1.2192,
+      "step": 6214
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.687765821380948,
+      "learning_rate": 5.1138351695134325e-06,
+      "loss": 1.2597,
+      "step": 6215
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.4346026177724935,
+      "learning_rate": 5.112511729526708e-06,
+      "loss": 1.0865,
+      "step": 6216
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.6694227711778957,
+      "learning_rate": 5.111188281653401e-06,
+      "loss": 0.3491,
+      "step": 6217
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.883057852039323,
+      "learning_rate": 5.109864825986278e-06,
+      "loss": 0.4154,
+      "step": 6218
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.028272430272317,
+      "learning_rate": 5.1085413626181115e-06,
+      "loss": 1.1474,
+      "step": 6219
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 6.025093841644022,
+      "learning_rate": 5.107217891641666e-06,
+      "loss": 1.2059,
+      "step": 6220
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.402164700317038,
+      "learning_rate": 5.1058944131497136e-06,
+      "loss": 0.5818,
+      "step": 6221
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.7144367551172395,
+      "learning_rate": 5.104570927235022e-06,
+      "loss": 0.7003,
+      "step": 6222
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.4100266315263363,
+      "learning_rate": 5.103247433990366e-06,
+      "loss": 0.7946,
+      "step": 6223
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.265470392284644,
+      "learning_rate": 5.1019239335085125e-06,
+      "loss": 1.2367,
+      "step": 6224
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.9032990061558483,
+      "learning_rate": 5.100600425882235e-06,
+      "loss": 0.713,
+      "step": 6225
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.8693250297510664,
+      "learning_rate": 5.099276911204306e-06,
+      "loss": 0.7159,
+      "step": 6226
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.0870859928610015,
+      "learning_rate": 5.097953389567498e-06,
+      "loss": 0.8153,
+      "step": 6227
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.029878114306394,
+      "learning_rate": 5.096629861064582e-06,
+      "loss": 0.9994,
+      "step": 6228
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.951793258420478,
+      "learning_rate": 5.095306325788335e-06,
+      "loss": 0.7461,
+      "step": 6229
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.776917444102281,
+      "learning_rate": 5.093982783831528e-06,
+      "loss": 1.5689,
+      "step": 6230
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.812739997250496,
+      "learning_rate": 5.092659235286938e-06,
+      "loss": 0.5859,
+      "step": 6231
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.7644079508415316,
+      "learning_rate": 5.091335680247339e-06,
+      "loss": 0.6422,
+      "step": 6232
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9220303257829688,
+      "learning_rate": 5.090012118805505e-06,
+      "loss": 0.1435,
+      "step": 6233
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.62449968094465,
+      "learning_rate": 5.088688551054214e-06,
+      "loss": 0.5911,
+      "step": 6234
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.742055006600364,
+      "learning_rate": 5.0873649770862425e-06,
+      "loss": 0.7453,
+      "step": 6235
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.9706011107453003,
+      "learning_rate": 5.086041396994365e-06,
+      "loss": 0.9127,
+      "step": 6236
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.6868597009845514,
+      "learning_rate": 5.08471781087136e-06,
+      "loss": 0.8972,
+      "step": 6237
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.5694862211533263,
+      "learning_rate": 5.083394218810006e-06,
+      "loss": 0.6879,
+      "step": 6238
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.4643485370094247,
+      "learning_rate": 5.082070620903079e-06,
+      "loss": 0.5668,
+      "step": 6239
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.66966876203938,
+      "learning_rate": 5.0807470172433594e-06,
+      "loss": 1.0946,
+      "step": 6240
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.886788766329076,
+      "learning_rate": 5.079423407923625e-06,
+      "loss": 0.9192,
+      "step": 6241
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.5662657650469645,
+      "learning_rate": 5.078099793036656e-06,
+      "loss": 0.5687,
+      "step": 6242
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.2423877659804696,
+      "learning_rate": 5.076776172675232e-06,
+      "loss": 0.284,
+      "step": 6243
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.700604243524799,
+      "learning_rate": 5.0754525469321305e-06,
+      "loss": 0.6439,
+      "step": 6244
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.5641119785064723,
+      "learning_rate": 5.074128915900134e-06,
+      "loss": 1.022,
+      "step": 6245
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.13662079686435,
+      "learning_rate": 5.072805279672025e-06,
+      "loss": 0.6375,
+      "step": 6246
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.752477270526784,
+      "learning_rate": 5.071481638340581e-06,
+      "loss": 0.5996,
+      "step": 6247
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.72802022968144,
+      "learning_rate": 5.070157991998586e-06,
+      "loss": 0.9262,
+      "step": 6248
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.94412339728263,
+      "learning_rate": 5.06883434073882e-06,
+      "loss": 0.9916,
+      "step": 6249
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.737079520523553,
+      "learning_rate": 5.067510684654069e-06,
+      "loss": 0.459,
+      "step": 6250
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.5092207035448744,
+      "learning_rate": 5.06618702383711e-06,
+      "loss": 0.5877,
+      "step": 6251
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.1853215059093363,
+      "learning_rate": 5.06486335838073e-06,
+      "loss": 0.6728,
+      "step": 6252
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.408146409003843,
+      "learning_rate": 5.06353968837771e-06,
+      "loss": 0.8563,
+      "step": 6253
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.7415881893911864,
+      "learning_rate": 5.062216013920836e-06,
+      "loss": 0.6281,
+      "step": 6254
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.423419492453897,
+      "learning_rate": 5.060892335102888e-06,
+      "loss": 1.5519,
+      "step": 6255
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.1369108485311883,
+      "learning_rate": 5.0595686520166535e-06,
+      "loss": 0.4109,
+      "step": 6256
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.6276554918514528,
+      "learning_rate": 5.058244964754916e-06,
+      "loss": 0.4997,
+      "step": 6257
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.678298545740144,
+      "learning_rate": 5.056921273410459e-06,
+      "loss": 1.0561,
+      "step": 6258
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.04000145001499,
+      "learning_rate": 5.05559757807607e-06,
+      "loss": 0.6706,
+      "step": 6259
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.713966562253011,
+      "learning_rate": 5.054273878844532e-06,
+      "loss": 0.8137,
+      "step": 6260
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8093419343800716,
+      "learning_rate": 5.052950175808631e-06,
+      "loss": 0.217,
+      "step": 6261
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 8.562365238511848,
+      "learning_rate": 5.051626469061153e-06,
+      "loss": 0.5402,
+      "step": 6262
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.0469827554424525,
+      "learning_rate": 5.050302758694885e-06,
+      "loss": 1.1309,
+      "step": 6263
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.855241797223289,
+      "learning_rate": 5.048979044802611e-06,
+      "loss": 1.0927,
+      "step": 6264
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.7073773922637976,
+      "learning_rate": 5.047655327477119e-06,
+      "loss": 0.8027,
+      "step": 6265
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.4058276992984235,
+      "learning_rate": 5.0463316068111975e-06,
+      "loss": 0.8036,
+      "step": 6266
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.921621492746431,
+      "learning_rate": 5.0450078828976326e-06,
+      "loss": 1.1416,
+      "step": 6267
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.8577221966367548,
+      "learning_rate": 5.0436841558292096e-06,
+      "loss": 0.4846,
+      "step": 6268
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.471053589326682,
+      "learning_rate": 5.042360425698718e-06,
+      "loss": 0.9808,
+      "step": 6269
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 6.235702521038686,
+      "learning_rate": 5.041036692598944e-06,
+      "loss": 1.4737,
+      "step": 6270
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.378368906345673,
+      "learning_rate": 5.039712956622678e-06,
+      "loss": 1.4435,
+      "step": 6271
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.277791036639248,
+      "learning_rate": 5.038389217862705e-06,
+      "loss": 1.0241,
+      "step": 6272
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.796578394912698,
+      "learning_rate": 5.037065476411816e-06,
+      "loss": 0.8263,
+      "step": 6273
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.3092776679300604,
+      "learning_rate": 5.035741732362798e-06,
+      "loss": 0.8006,
+      "step": 6274
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.584746533139775,
+      "learning_rate": 5.0344179858084395e-06,
+      "loss": 1.0898,
+      "step": 6275
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.850175703684064,
+      "learning_rate": 5.033094236841531e-06,
+      "loss": 1.0502,
+      "step": 6276
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.6000543964963483,
+      "learning_rate": 5.03177048555486e-06,
+      "loss": 0.5403,
+      "step": 6277
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8518070760604197,
+      "learning_rate": 5.030446732041216e-06,
+      "loss": 0.6082,
+      "step": 6278
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5225956717460523,
+      "learning_rate": 5.029122976393388e-06,
+      "loss": 0.2299,
+      "step": 6279
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.0498073241758967,
+      "learning_rate": 5.027799218704168e-06,
+      "loss": 0.6301,
+      "step": 6280
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.102132833210754,
+      "learning_rate": 5.026475459066342e-06,
+      "loss": 0.5903,
+      "step": 6281
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.6469790492142407,
+      "learning_rate": 5.025151697572703e-06,
+      "loss": 0.8519,
+      "step": 6282
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.220063325470204,
+      "learning_rate": 5.023827934316039e-06,
+      "loss": 0.9139,
+      "step": 6283
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1852192562946298,
+      "learning_rate": 5.02250416938914e-06,
+      "loss": 0.2314,
+      "step": 6284
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.9635533768960713,
+      "learning_rate": 5.021180402884796e-06,
+      "loss": 0.386,
+      "step": 6285
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.860632087202993,
+      "learning_rate": 5.0198566348958e-06,
+      "loss": 0.4964,
+      "step": 6286
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.56821484008037,
+      "learning_rate": 5.018532865514938e-06,
+      "loss": 0.7766,
+      "step": 6287
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.670637243869879,
+      "learning_rate": 5.0172090948350036e-06,
+      "loss": 0.7336,
+      "step": 6288
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.17472118045534,
+      "learning_rate": 5.015885322948787e-06,
+      "loss": 1.0167,
+      "step": 6289
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.121084537108153,
+      "learning_rate": 5.01456154994908e-06,
+      "loss": 0.4099,
+      "step": 6290
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.323932905163277,
+      "learning_rate": 5.01323777592867e-06,
+      "loss": 0.7663,
+      "step": 6291
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.36491687728099,
+      "learning_rate": 5.011914000980349e-06,
+      "loss": 0.6422,
+      "step": 6292
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.666396995504404,
+      "learning_rate": 5.0105902251969084e-06,
+      "loss": 0.904,
+      "step": 6293
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.3936856557765145,
+      "learning_rate": 5.00926644867114e-06,
+      "loss": 0.9992,
+      "step": 6294
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.7203280943851595,
+      "learning_rate": 5.007942671495832e-06,
+      "loss": 1.2324,
+      "step": 6295
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.3110721080028003,
+      "learning_rate": 5.006618893763779e-06,
+      "loss": 0.6198,
+      "step": 6296
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.771613517384583,
+      "learning_rate": 5.005295115567771e-06,
+      "loss": 0.8101,
+      "step": 6297
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.2345002368899296,
+      "learning_rate": 5.003971337000597e-06,
+      "loss": 0.4533,
+      "step": 6298
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.523863241460051,
+      "learning_rate": 5.00264755815505e-06,
+      "loss": 0.8525,
+      "step": 6299
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.7989059343538045,
+      "learning_rate": 5.001323779123921e-06,
+      "loss": 0.5445,
+      "step": 6300
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.8013321983918775,
+      "learning_rate": 5e-06,
+      "loss": 0.8655,
+      "step": 6301
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.150377577679675,
+      "learning_rate": 4.99867622087608e-06,
+      "loss": 0.8885,
+      "step": 6302
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.902739896702521,
+      "learning_rate": 4.99735244184495e-06,
+      "loss": 0.8761,
+      "step": 6303
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.774687840814019,
+      "learning_rate": 4.996028662999405e-06,
+      "loss": 0.5963,
+      "step": 6304
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.410054218319989,
+      "learning_rate": 4.994704884432231e-06,
+      "loss": 1.0872,
+      "step": 6305
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0244463177855765,
+      "learning_rate": 4.9933811062362224e-06,
+      "loss": 0.1778,
+      "step": 6306
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.6620882328925575,
+      "learning_rate": 4.992057328504169e-06,
+      "loss": 0.884,
+      "step": 6307
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.403301419532233,
+      "learning_rate": 4.990733551328862e-06,
+      "loss": 0.5739,
+      "step": 6308
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.9845782164457884,
+      "learning_rate": 4.989409774803092e-06,
+      "loss": 0.7148,
+      "step": 6309
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.2284050123991626,
+      "learning_rate": 4.988085999019654e-06,
+      "loss": 0.5754,
+      "step": 6310
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.799427569150806,
+      "learning_rate": 4.9867622240713325e-06,
+      "loss": 0.585,
+      "step": 6311
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.501545703443735,
+      "learning_rate": 4.985438450050922e-06,
+      "loss": 0.901,
+      "step": 6312
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.6819769592865965,
+      "learning_rate": 4.984114677051214e-06,
+      "loss": 0.7882,
+      "step": 6313
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.440819132850262,
+      "learning_rate": 4.982790905164997e-06,
+      "loss": 0.7031,
+      "step": 6314
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.6972404685211484,
+      "learning_rate": 4.981467134485062e-06,
+      "loss": 1.0036,
+      "step": 6315
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.9079783686070577,
+      "learning_rate": 4.980143365104203e-06,
+      "loss": 0.5343,
+      "step": 6316
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.982005751028777,
+      "learning_rate": 4.978819597115205e-06,
+      "loss": 0.4965,
+      "step": 6317
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.893228389137094,
+      "learning_rate": 4.977495830610862e-06,
+      "loss": 0.4195,
+      "step": 6318
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.552228796695302,
+      "learning_rate": 4.976172065683963e-06,
+      "loss": 0.7378,
+      "step": 6319
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.264345436792313,
+      "learning_rate": 4.974848302427299e-06,
+      "loss": 0.8076,
+      "step": 6320
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.7932063119244797,
+      "learning_rate": 4.9735245409336586e-06,
+      "loss": 0.5244,
+      "step": 6321
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 5.227818897280464,
+      "learning_rate": 4.972200781295835e-06,
+      "loss": 1.6158,
+      "step": 6322
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.2963786876135655,
+      "learning_rate": 4.970877023606613e-06,
+      "loss": 0.9132,
+      "step": 6323
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 5.589089310372882,
+      "learning_rate": 4.969553267958785e-06,
+      "loss": 1.3736,
+      "step": 6324
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.6672605443108566,
+      "learning_rate": 4.9682295144451415e-06,
+      "loss": 1.0031,
+      "step": 6325
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.800499484512619,
+      "learning_rate": 4.96690576315847e-06,
+      "loss": 0.59,
+      "step": 6326
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.3517919780931447,
+      "learning_rate": 4.965582014191562e-06,
+      "loss": 0.5405,
+      "step": 6327
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.1593004177129544,
+      "learning_rate": 4.964258267637204e-06,
+      "loss": 0.4237,
+      "step": 6328
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.1968301978355282,
+      "learning_rate": 4.962934523588187e-06,
+      "loss": 0.3499,
+      "step": 6329
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.644271439604117,
+      "learning_rate": 4.961610782137297e-06,
+      "loss": 0.8678,
+      "step": 6330
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 5.526329531410269,
+      "learning_rate": 4.960287043377324e-06,
+      "loss": 0.9463,
+      "step": 6331
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.135163169342016,
+      "learning_rate": 4.958963307401056e-06,
+      "loss": 1.0386,
+      "step": 6332
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.074959112540754,
+      "learning_rate": 4.957639574301285e-06,
+      "loss": 0.588,
+      "step": 6333
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.475894793094254,
+      "learning_rate": 4.956315844170792e-06,
+      "loss": 0.4042,
+      "step": 6334
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.1667108394781,
+      "learning_rate": 4.954992117102369e-06,
+      "loss": 0.6566,
+      "step": 6335
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.6318038286029357,
+      "learning_rate": 4.953668393188803e-06,
+      "loss": 0.6653,
+      "step": 6336
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.5598441681408564,
+      "learning_rate": 4.9523446725228805e-06,
+      "loss": 0.9093,
+      "step": 6337
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.312639300371545,
+      "learning_rate": 4.95102095519739e-06,
+      "loss": 0.6447,
+      "step": 6338
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.2869108321149354,
+      "learning_rate": 4.949697241305118e-06,
+      "loss": 0.5326,
+      "step": 6339
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.2398638722023074,
+      "learning_rate": 4.948373530938849e-06,
+      "loss": 0.5487,
+      "step": 6340
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.6746133544438657,
+      "learning_rate": 4.947049824191371e-06,
+      "loss": 0.3259,
+      "step": 6341
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 5.023439343154598,
+      "learning_rate": 4.945726121155469e-06,
+      "loss": 0.7726,
+      "step": 6342
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.423088694007279,
+      "learning_rate": 4.9444024219239315e-06,
+      "loss": 0.599,
+      "step": 6343
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.657343174210334,
+      "learning_rate": 4.943078726589543e-06,
+      "loss": 0.8292,
+      "step": 6344
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.069730396690454,
+      "learning_rate": 4.941755035245087e-06,
+      "loss": 0.7747,
+      "step": 6345
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.907570960554669,
+      "learning_rate": 4.940431347983348e-06,
+      "loss": 0.5811,
+      "step": 6346
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.9121204580647655,
+      "learning_rate": 4.9391076648971135e-06,
+      "loss": 0.8157,
+      "step": 6347
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.920499319357153,
+      "learning_rate": 4.937783986079165e-06,
+      "loss": 0.8814,
+      "step": 6348
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.867359978831976,
+      "learning_rate": 4.93646031162229e-06,
+      "loss": 1.0379,
+      "step": 6349
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.988426299490411,
+      "learning_rate": 4.935136641619272e-06,
+      "loss": 1.4079,
+      "step": 6350
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.295221225986864,
+      "learning_rate": 4.933812976162892e-06,
+      "loss": 0.4422,
+      "step": 6351
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.7696669405353247,
+      "learning_rate": 4.932489315345933e-06,
+      "loss": 0.6909,
+      "step": 6352
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.1369585460443123,
+      "learning_rate": 4.9311656592611804e-06,
+      "loss": 0.645,
+      "step": 6353
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.3681224044111864,
+      "learning_rate": 4.929842008001415e-06,
+      "loss": 0.4803,
+      "step": 6354
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.98677924723652,
+      "learning_rate": 4.92851836165942e-06,
+      "loss": 1.2134,
+      "step": 6355
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.4940831075783816,
+      "learning_rate": 4.927194720327978e-06,
+      "loss": 0.6613,
+      "step": 6356
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.3360524861854586,
+      "learning_rate": 4.925871084099867e-06,
+      "loss": 0.5912,
+      "step": 6357
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.13829777093247,
+      "learning_rate": 4.924547453067871e-06,
+      "loss": 0.5976,
+      "step": 6358
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.864907012819576,
+      "learning_rate": 4.923223827324769e-06,
+      "loss": 0.7684,
+      "step": 6359
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.654007313074135,
+      "learning_rate": 4.921900206963345e-06,
+      "loss": 1.1836,
+      "step": 6360
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.7205381939259325,
+      "learning_rate": 4.920576592076375e-06,
+      "loss": 0.4543,
+      "step": 6361
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.207080349387025,
+      "learning_rate": 4.919252982756643e-06,
+      "loss": 1.1061,
+      "step": 6362
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.497492926051889,
+      "learning_rate": 4.9179293790969225e-06,
+      "loss": 1.2731,
+      "step": 6363
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.9592866021194957,
+      "learning_rate": 4.916605781189996e-06,
+      "loss": 0.7913,
+      "step": 6364
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.789903155601039,
+      "learning_rate": 4.9152821891286404e-06,
+      "loss": 0.3982,
+      "step": 6365
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.6060905109286803,
+      "learning_rate": 4.913958603005636e-06,
+      "loss": 0.6695,
+      "step": 6366
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.7958090281529833,
+      "learning_rate": 4.91263502291376e-06,
+      "loss": 0.5765,
+      "step": 6367
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.5029185894999735,
+      "learning_rate": 4.911311448945787e-06,
+      "loss": 0.9681,
+      "step": 6368
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.238696692533975,
+      "learning_rate": 4.9099878811944965e-06,
+      "loss": 0.7806,
+      "step": 6369
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.8919735811733625,
+      "learning_rate": 4.908664319752663e-06,
+      "loss": 0.9318,
+      "step": 6370
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.736754873965216,
+      "learning_rate": 4.9073407647130625e-06,
+      "loss": 0.6905,
+      "step": 6371
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.136277462289465,
+      "learning_rate": 4.906017216168471e-06,
+      "loss": 1.0667,
+      "step": 6372
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.2902996833226683,
+      "learning_rate": 4.904693674211667e-06,
+      "loss": 0.6025,
+      "step": 6373
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.8055175213525025,
+      "learning_rate": 4.9033701389354185e-06,
+      "loss": 0.7545,
+      "step": 6374
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9379557571515578,
+      "learning_rate": 4.902046610432504e-06,
+      "loss": 0.3936,
+      "step": 6375
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 5.6127646401142055,
+      "learning_rate": 4.900723088795695e-06,
+      "loss": 0.9573,
+      "step": 6376
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.634383651385539,
+      "learning_rate": 4.899399574117766e-06,
+      "loss": 0.928,
+      "step": 6377
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.3841366044644765,
+      "learning_rate": 4.898076066491488e-06,
+      "loss": 0.8414,
+      "step": 6378
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.8380715593656936,
+      "learning_rate": 4.896752566009637e-06,
+      "loss": 0.6047,
+      "step": 6379
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.3478182534713485,
+      "learning_rate": 4.8954290727649785e-06,
+      "loss": 0.3525,
+      "step": 6380
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.2704191305723413,
+      "learning_rate": 4.894105586850288e-06,
+      "loss": 0.5744,
+      "step": 6381
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 5.296254206143929,
+      "learning_rate": 4.892782108358335e-06,
+      "loss": 0.9147,
+      "step": 6382
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.577524099642258,
+      "learning_rate": 4.891458637381891e-06,
+      "loss": 0.6596,
+      "step": 6383
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.5392351497191736,
+      "learning_rate": 4.8901351740137235e-06,
+      "loss": 0.7526,
+      "step": 6384
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.0395114611346887,
+      "learning_rate": 4.888811718346602e-06,
+      "loss": 0.7119,
+      "step": 6385
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.111179056014712,
+      "learning_rate": 4.887488270473294e-06,
+      "loss": 0.6011,
+      "step": 6386
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.5648992088187637,
+      "learning_rate": 4.886164830486569e-06,
+      "loss": 0.5581,
+      "step": 6387
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.3273150749747025,
+      "learning_rate": 4.8848413984791935e-06,
+      "loss": 0.8686,
+      "step": 6388
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.942505465751809,
+      "learning_rate": 4.883517974543935e-06,
+      "loss": 0.9702,
+      "step": 6389
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.8595179418584245,
+      "learning_rate": 4.882194558773562e-06,
+      "loss": 0.9561,
+      "step": 6390
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.280663704527302,
+      "learning_rate": 4.8808711512608355e-06,
+      "loss": 0.8193,
+      "step": 6391
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.2210982354831654,
+      "learning_rate": 4.879547752098524e-06,
+      "loss": 0.5106,
+      "step": 6392
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.337642173676028,
+      "learning_rate": 4.878224361379389e-06,
+      "loss": 1.0615,
+      "step": 6393
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.14583018619186,
+      "learning_rate": 4.8769009791961975e-06,
+      "loss": 0.7129,
+      "step": 6394
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.1424088613650816,
+      "learning_rate": 4.875577605641711e-06,
+      "loss": 0.3761,
+      "step": 6395
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.8310450197505426,
+      "learning_rate": 4.8742542408086955e-06,
+      "loss": 0.8652,
+      "step": 6396
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.2018217096329407,
+      "learning_rate": 4.8729308847899075e-06,
+      "loss": 0.5413,
+      "step": 6397
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.362855687172706,
+      "learning_rate": 4.8716075376781115e-06,
+      "loss": 0.5127,
+      "step": 6398
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.967190852418591,
+      "learning_rate": 4.87028419956607e-06,
+      "loss": 0.5075,
+      "step": 6399
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.1975202187908796,
+      "learning_rate": 4.86896087054654e-06,
+      "loss": 0.5307,
+      "step": 6400
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 5.1581629182550195,
+      "learning_rate": 4.867637550712283e-06,
+      "loss": 0.9677,
+      "step": 6401
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.6676487285144606,
+      "learning_rate": 4.86631424015606e-06,
+      "loss": 0.6706,
+      "step": 6402
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.6524160507678043,
+      "learning_rate": 4.864990938970624e-06,
+      "loss": 0.6049,
+      "step": 6403
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.650624751612126,
+      "learning_rate": 4.863667647248737e-06,
+      "loss": 1.1215,
+      "step": 6404
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.530953242466552,
+      "learning_rate": 4.862344365083154e-06,
+      "loss": 0.9914,
+      "step": 6405
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.1659566501334755,
+      "learning_rate": 4.861021092566633e-06,
+      "loss": 0.5622,
+      "step": 6406
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.536544577037245,
+      "learning_rate": 4.859697829791927e-06,
+      "loss": 1.2066,
+      "step": 6407
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.33958469940971,
+      "learning_rate": 4.858374576851795e-06,
+      "loss": 1.2222,
+      "step": 6408
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.381614360695345,
+      "learning_rate": 4.857051333838987e-06,
+      "loss": 0.65,
+      "step": 6409
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.2292333130820845,
+      "learning_rate": 4.855728100846258e-06,
+      "loss": 0.6844,
+      "step": 6410
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.292140738399272,
+      "learning_rate": 4.854404877966361e-06,
+      "loss": 0.9661,
+      "step": 6411
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.2999953285623556,
+      "learning_rate": 4.8530816652920485e-06,
+      "loss": 0.5868,
+      "step": 6412
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.0015489970748885,
+      "learning_rate": 4.851758462916075e-06,
+      "loss": 0.7151,
+      "step": 6413
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9300620487820964,
+      "learning_rate": 4.850435270931184e-06,
+      "loss": 0.1667,
+      "step": 6414
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.440839701309359,
+      "learning_rate": 4.849112089430133e-06,
+      "loss": 0.3568,
+      "step": 6415
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.295708322856452,
+      "learning_rate": 4.847788918505665e-06,
+      "loss": 0.8336,
+      "step": 6416
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.7894794538723233,
+      "learning_rate": 4.846465758250532e-06,
+      "loss": 0.8585,
+      "step": 6417
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0371011365815022,
+      "learning_rate": 4.845142608757481e-06,
+      "loss": 0.1409,
+      "step": 6418
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.118848364575265,
+      "learning_rate": 4.843819470119262e-06,
+      "loss": 0.608,
+      "step": 6419
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.2868040020579445,
+      "learning_rate": 4.842496342428616e-06,
+      "loss": 0.471,
+      "step": 6420
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.479520859133651,
+      "learning_rate": 4.84117322577829e-06,
+      "loss": 0.8966,
+      "step": 6421
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.594747610138365,
+      "learning_rate": 4.839850120261032e-06,
+      "loss": 1.0153,
+      "step": 6422
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.33839452827434,
+      "learning_rate": 4.838527025969582e-06,
+      "loss": 0.8278,
+      "step": 6423
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.867393160179802,
+      "learning_rate": 4.837203942996687e-06,
+      "loss": 0.7852,
+      "step": 6424
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.9408893881153704,
+      "learning_rate": 4.8358808714350856e-06,
+      "loss": 0.3435,
+      "step": 6425
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.157718397068329,
+      "learning_rate": 4.834557811377519e-06,
+      "loss": 0.3427,
+      "step": 6426
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.9450794345286235,
+      "learning_rate": 4.833234762916731e-06,
+      "loss": 0.8995,
+      "step": 6427
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.727524322944186,
+      "learning_rate": 4.831911726145461e-06,
+      "loss": 1.1968,
+      "step": 6428
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.605545754356394,
+      "learning_rate": 4.830588701156448e-06,
+      "loss": 0.5252,
+      "step": 6429
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.202165484465828,
+      "learning_rate": 4.829265688042429e-06,
+      "loss": 0.7731,
+      "step": 6430
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.30692738832285,
+      "learning_rate": 4.827942686896143e-06,
+      "loss": 0.9813,
+      "step": 6431
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.3485257582972343,
+      "learning_rate": 4.8266196978103245e-06,
+      "loss": 0.5746,
+      "step": 6432
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9948209222241482,
+      "learning_rate": 4.825296720877711e-06,
+      "loss": 0.4288,
+      "step": 6433
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.359141276541377,
+      "learning_rate": 4.823973756191037e-06,
+      "loss": 0.4782,
+      "step": 6434
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.3460161213266657,
+      "learning_rate": 4.822650803843037e-06,
+      "loss": 0.8516,
+      "step": 6435
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.9047246603491295,
+      "learning_rate": 4.821327863926445e-06,
+      "loss": 0.8164,
+      "step": 6436
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.389933865662566,
+      "learning_rate": 4.8200049365339905e-06,
+      "loss": 0.9688,
+      "step": 6437
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.90959305816749,
+      "learning_rate": 4.818682021758407e-06,
+      "loss": 1.114,
+      "step": 6438
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.746758804316483,
+      "learning_rate": 4.817359119692424e-06,
+      "loss": 1.0432,
+      "step": 6439
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.4099016585770023,
+      "learning_rate": 4.816036230428773e-06,
+      "loss": 0.32,
+      "step": 6440
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.881278816114833,
+      "learning_rate": 4.814713354060181e-06,
+      "loss": 1.0872,
+      "step": 6441
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.622871480181095,
+      "learning_rate": 4.8133904906793776e-06,
+      "loss": 0.9624,
+      "step": 6442
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.9472587561883765,
+      "learning_rate": 4.8120676403790875e-06,
+      "loss": 1.0216,
+      "step": 6443
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.361458801690792,
+      "learning_rate": 4.8107448032520376e-06,
+      "loss": 0.9053,
+      "step": 6444
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.2503964795396,
+      "learning_rate": 4.809421979390954e-06,
+      "loss": 0.943,
+      "step": 6445
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.839225942265717,
+      "learning_rate": 4.8080991688885606e-06,
+      "loss": 0.6044,
+      "step": 6446
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.701583715591647,
+      "learning_rate": 4.806776371837581e-06,
+      "loss": 0.4485,
+      "step": 6447
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.4206292275300285,
+      "learning_rate": 4.805453588330735e-06,
+      "loss": 0.366,
+      "step": 6448
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.9656245324891324,
+      "learning_rate": 4.804130818460746e-06,
+      "loss": 0.8929,
+      "step": 6449
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2925316239131028,
+      "learning_rate": 4.802808062320334e-06,
+      "loss": 0.181,
+      "step": 6450
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.1121094496228907,
+      "learning_rate": 4.801485320002219e-06,
+      "loss": 0.9133,
+      "step": 6451
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.748963455706853,
+      "learning_rate": 4.8001625915991205e-06,
+      "loss": 1.0718,
+      "step": 6452
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.125724567974992,
+      "learning_rate": 4.798839877203754e-06,
+      "loss": 1.165,
+      "step": 6453
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.5048961563875642,
+      "learning_rate": 4.7975171769088366e-06,
+      "loss": 0.3907,
+      "step": 6454
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.241688967850177,
+      "learning_rate": 4.7961944908070835e-06,
+      "loss": 1.272,
+      "step": 6455
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.530270446774279,
+      "learning_rate": 4.7948718189912095e-06,
+      "loss": 0.7915,
+      "step": 6456
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.395649687464648,
+      "learning_rate": 4.793549161553927e-06,
+      "loss": 1.0097,
+      "step": 6457
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.084798442453911,
+      "learning_rate": 4.792226518587952e-06,
+      "loss": 0.6873,
+      "step": 6458
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.339275731224011,
+      "learning_rate": 4.7909038901859945e-06,
+      "loss": 0.4039,
+      "step": 6459
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.936160088236716,
+      "learning_rate": 4.789581276440762e-06,
+      "loss": 0.7648,
+      "step": 6460
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.014448484107769,
+      "learning_rate": 4.788258677444967e-06,
+      "loss": 0.9322,
+      "step": 6461
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.740184648909575,
+      "learning_rate": 4.786936093291318e-06,
+      "loss": 0.7014,
+      "step": 6462
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.564045541750739,
+      "learning_rate": 4.78561352407252e-06,
+      "loss": 0.4463,
+      "step": 6463
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.101364495462342,
+      "learning_rate": 4.784290969881284e-06,
+      "loss": 0.5848,
+      "step": 6464
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.295584779861668,
+      "learning_rate": 4.78296843081031e-06,
+      "loss": 0.9924,
+      "step": 6465
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.119756801063344,
+      "learning_rate": 4.781645906952304e-06,
+      "loss": 0.4947,
+      "step": 6466
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.2737077807255406,
+      "learning_rate": 4.78032339839997e-06,
+      "loss": 0.6644,
+      "step": 6467
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.500344929627298,
+      "learning_rate": 4.779000905246009e-06,
+      "loss": 0.5359,
+      "step": 6468
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.267609818219153,
+      "learning_rate": 4.777678427583124e-06,
+      "loss": 0.5198,
+      "step": 6469
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.657578711382833,
+      "learning_rate": 4.776355965504015e-06,
+      "loss": 0.5331,
+      "step": 6470
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.994108345839051,
+      "learning_rate": 4.775033519101378e-06,
+      "loss": 0.9264,
+      "step": 6471
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.8699942361158346,
+      "learning_rate": 4.773711088467912e-06,
+      "loss": 0.7568,
+      "step": 6472
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.175424319748729,
+      "learning_rate": 4.772388673696314e-06,
+      "loss": 1.2296,
+      "step": 6473
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.502369361692485,
+      "learning_rate": 4.771066274879279e-06,
+      "loss": 0.9849,
+      "step": 6474
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.651851635025855,
+      "learning_rate": 4.769743892109502e-06,
+      "loss": 0.5328,
+      "step": 6475
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.544804100513222,
+      "learning_rate": 4.768421525479677e-06,
+      "loss": 1.0894,
+      "step": 6476
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.1437570454179444,
+      "learning_rate": 4.767099175082495e-06,
+      "loss": 0.7833,
+      "step": 6477
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.020674057653643,
+      "learning_rate": 4.765776841010647e-06,
+      "loss": 0.7353,
+      "step": 6478
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.518597556006481,
+      "learning_rate": 4.764454523356823e-06,
+      "loss": 1.1217,
+      "step": 6479
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.13953384114246,
+      "learning_rate": 4.763132222213711e-06,
+      "loss": 0.3948,
+      "step": 6480
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.9619726314103363,
+      "learning_rate": 4.761809937673999e-06,
+      "loss": 0.6049,
+      "step": 6481
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.2256903440489735,
+      "learning_rate": 4.760487669830377e-06,
+      "loss": 0.4989,
+      "step": 6482
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.624195051629224,
+      "learning_rate": 4.7591654187755245e-06,
+      "loss": 0.4404,
+      "step": 6483
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.8447451022651813,
+      "learning_rate": 4.757843184602128e-06,
+      "loss": 0.9409,
+      "step": 6484
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.210561404766787,
+      "learning_rate": 4.756520967402871e-06,
+      "loss": 1.0253,
+      "step": 6485
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.2958280305185474,
+      "learning_rate": 4.755198767270434e-06,
+      "loss": 0.6959,
+      "step": 6486
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.8705923794882873,
+      "learning_rate": 4.7538765842975e-06,
+      "loss": 0.9916,
+      "step": 6487
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.17973960051939,
+      "learning_rate": 4.752554418576744e-06,
+      "loss": 0.8638,
+      "step": 6488
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.8058675230780707,
+      "learning_rate": 4.7512322702008475e-06,
+      "loss": 0.9882,
+      "step": 6489
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.584219461622293,
+      "learning_rate": 4.749910139262485e-06,
+      "loss": 0.459,
+      "step": 6490
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.5689914149607276,
+      "learning_rate": 4.748588025854334e-06,
+      "loss": 0.9858,
+      "step": 6491
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6919630395843155,
+      "learning_rate": 4.747265930069069e-06,
+      "loss": 0.2479,
+      "step": 6492
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.534827490112639,
+      "learning_rate": 4.745943851999362e-06,
+      "loss": 0.8208,
+      "step": 6493
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.060101817436625,
+      "learning_rate": 4.744621791737886e-06,
+      "loss": 1.1179,
+      "step": 6494
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.358457354866757,
+      "learning_rate": 4.74329974937731e-06,
+      "loss": 1.2671,
+      "step": 6495
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.4693194824047855,
+      "learning_rate": 4.741977725010304e-06,
+      "loss": 0.6998,
+      "step": 6496
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.503786475228983,
+      "learning_rate": 4.740655718729537e-06,
+      "loss": 0.9596,
+      "step": 6497
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.5345182678278477,
+      "learning_rate": 4.739333730627674e-06,
+      "loss": 0.966,
+      "step": 6498
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.3117682019112844,
+      "learning_rate": 4.7380117607973855e-06,
+      "loss": 0.5517,
+      "step": 6499
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.778259677896063,
+      "learning_rate": 4.73668980933133e-06,
+      "loss": 0.5979,
+      "step": 6500
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.104106387260651,
+      "learning_rate": 4.735367876322174e-06,
+      "loss": 0.6847,
+      "step": 6501
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.7767560467196026,
+      "learning_rate": 4.734045961862577e-06,
+      "loss": 1.1592,
+      "step": 6502
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.879394886737245,
+      "learning_rate": 4.732724066045201e-06,
+      "loss": 1.032,
+      "step": 6503
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.215580320281627,
+      "learning_rate": 4.731402188962706e-06,
+      "loss": 0.7044,
+      "step": 6504
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.941829061345157,
+      "learning_rate": 4.730080330707748e-06,
+      "loss": 0.7978,
+      "step": 6505
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.2538614465290667,
+      "learning_rate": 4.728758491372983e-06,
+      "loss": 0.5861,
+      "step": 6506
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.061368353748383,
+      "learning_rate": 4.727436671051068e-06,
+      "loss": 0.9944,
+      "step": 6507
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.395148902636811,
+      "learning_rate": 4.726114869834656e-06,
+      "loss": 0.4854,
+      "step": 6508
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.547254005610118,
+      "learning_rate": 4.7247930878164e-06,
+      "loss": 0.892,
+      "step": 6509
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.354189589986255,
+      "learning_rate": 4.723471325088953e-06,
+      "loss": 0.3938,
+      "step": 6510
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.010307150721494,
+      "learning_rate": 4.722149581744959e-06,
+      "loss": 0.4741,
+      "step": 6511
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.220441302810838,
+      "learning_rate": 4.720827857877071e-06,
+      "loss": 0.5945,
+      "step": 6512
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.697154306835642,
+      "learning_rate": 4.719506153577935e-06,
+      "loss": 0.6007,
+      "step": 6513
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.4702241418488216,
+      "learning_rate": 4.718184468940197e-06,
+      "loss": 0.9425,
+      "step": 6514
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.4005861961412664,
+      "learning_rate": 4.716862804056503e-06,
+      "loss": 0.7388,
+      "step": 6515
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.177697256568991,
+      "learning_rate": 4.7155411590194935e-06,
+      "loss": 0.6258,
+      "step": 6516
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.5965537600315365,
+      "learning_rate": 4.714219533921811e-06,
+      "loss": 0.3597,
+      "step": 6517
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.821519678293872,
+      "learning_rate": 4.712897928856095e-06,
+      "loss": 0.6991,
+      "step": 6518
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.183071395135402,
+      "learning_rate": 4.711576343914984e-06,
+      "loss": 0.4475,
+      "step": 6519
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.3883270732955357,
+      "learning_rate": 4.710254779191116e-06,
+      "loss": 0.4484,
+      "step": 6520
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.2618431776639802,
+      "learning_rate": 4.7089332347771265e-06,
+      "loss": 0.4983,
+      "step": 6521
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8306121244153692,
+      "learning_rate": 4.707611710765654e-06,
+      "loss": 0.3587,
+      "step": 6522
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.443170147328888,
+      "learning_rate": 4.706290207249325e-06,
+      "loss": 1.4299,
+      "step": 6523
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.4569262766400835,
+      "learning_rate": 4.704968724320775e-06,
+      "loss": 0.7632,
+      "step": 6524
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.247821939618158,
+      "learning_rate": 4.703647262072634e-06,
+      "loss": 0.353,
+      "step": 6525
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.1100807442195255,
+      "learning_rate": 4.702325820597528e-06,
+      "loss": 0.5021,
+      "step": 6526
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.355309216517904,
+      "learning_rate": 4.70100439998809e-06,
+      "loss": 0.8495,
+      "step": 6527
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.0247338209320747,
+      "learning_rate": 4.699683000336941e-06,
+      "loss": 0.7517,
+      "step": 6528
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.236116874762068,
+      "learning_rate": 4.698361621736705e-06,
+      "loss": 1.0296,
+      "step": 6529
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.955894405304145,
+      "learning_rate": 4.6970402642800075e-06,
+      "loss": 0.3625,
+      "step": 6530
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.878574051928466,
+      "learning_rate": 4.695718928059469e-06,
+      "loss": 0.506,
+      "step": 6531
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.8906841416371174,
+      "learning_rate": 4.694397613167709e-06,
+      "loss": 0.6025,
+      "step": 6532
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.6763792751475264,
+      "learning_rate": 4.69307631969735e-06,
+      "loss": 0.6466,
+      "step": 6533
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.0301141217686656,
+      "learning_rate": 4.691755047741001e-06,
+      "loss": 0.6439,
+      "step": 6534
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.391320284304078,
+      "learning_rate": 4.690433797391282e-06,
+      "loss": 0.8834,
+      "step": 6535
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.0619941823911168,
+      "learning_rate": 4.689112568740807e-06,
+      "loss": 0.3854,
+      "step": 6536
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.758029545952147,
+      "learning_rate": 4.687791361882188e-06,
+      "loss": 1.0864,
+      "step": 6537
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.363849172380941,
+      "learning_rate": 4.6864701769080364e-06,
+      "loss": 1.0227,
+      "step": 6538
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.9221309019995574,
+      "learning_rate": 4.685149013910962e-06,
+      "loss": 0.3992,
+      "step": 6539
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.769912945729529,
+      "learning_rate": 4.683827872983571e-06,
+      "loss": 0.7576,
+      "step": 6540
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.032375109346851,
+      "learning_rate": 4.682506754218469e-06,
+      "loss": 0.7517,
+      "step": 6541
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.84124810640443,
+      "learning_rate": 4.681185657708261e-06,
+      "loss": 0.3858,
+      "step": 6542
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.26168632843942,
+      "learning_rate": 4.679864583545552e-06,
+      "loss": 0.4775,
+      "step": 6543
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.778346260856777,
+      "learning_rate": 4.678543531822944e-06,
+      "loss": 0.433,
+      "step": 6544
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.001229869305055,
+      "learning_rate": 4.677222502633033e-06,
+      "loss": 0.6559,
+      "step": 6545
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.434317474416247,
+      "learning_rate": 4.67590149606842e-06,
+      "loss": 0.5421,
+      "step": 6546
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.331987535791356,
+      "learning_rate": 4.674580512221703e-06,
+      "loss": 0.6373,
+      "step": 6547
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.1064710446924804,
+      "learning_rate": 4.673259551185475e-06,
+      "loss": 0.6903,
+      "step": 6548
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.482780717530685,
+      "learning_rate": 4.67193861305233e-06,
+      "loss": 0.7344,
+      "step": 6549
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.1185714991572953,
+      "learning_rate": 4.670617697914863e-06,
+      "loss": 0.3799,
+      "step": 6550
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.986888769209969,
+      "learning_rate": 4.66929680586566e-06,
+      "loss": 0.7964,
+      "step": 6551
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.1396347568967338,
+      "learning_rate": 4.667975936997311e-06,
+      "loss": 0.3424,
+      "step": 6552
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.826732914643811,
+      "learning_rate": 4.666655091402404e-06,
+      "loss": 0.8784,
+      "step": 6553
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.2725881096348632,
+      "learning_rate": 4.665334269173526e-06,
+      "loss": 0.8182,
+      "step": 6554
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.082508256581759,
+      "learning_rate": 4.664013470403258e-06,
+      "loss": 1.0632,
+      "step": 6555
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.5090485390867383,
+      "learning_rate": 4.662692695184184e-06,
+      "loss": 0.5078,
+      "step": 6556
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.547058739510734,
+      "learning_rate": 4.661371943608884e-06,
+      "loss": 0.5732,
+      "step": 6557
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.6117772081754747,
+      "learning_rate": 4.660051215769937e-06,
+      "loss": 0.5969,
+      "step": 6558
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.578479510510918,
+      "learning_rate": 4.65873051175992e-06,
+      "loss": 1.3698,
+      "step": 6559
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.660950646030756,
+      "learning_rate": 4.6574098316714086e-06,
+      "loss": 1.2019,
+      "step": 6560
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.683994687913521,
+      "learning_rate": 4.656089175596978e-06,
+      "loss": 0.8219,
+      "step": 6561
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.582453556871559,
+      "learning_rate": 4.6547685436292e-06,
+      "loss": 1.3325,
+      "step": 6562
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.104896611457791,
+      "learning_rate": 4.653447935860642e-06,
+      "loss": 0.4239,
+      "step": 6563
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.6478755011837922,
+      "learning_rate": 4.652127352383879e-06,
+      "loss": 0.3726,
+      "step": 6564
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.3335035105942667,
+      "learning_rate": 4.650806793291472e-06,
+      "loss": 0.3934,
+      "step": 6565
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.783339180629704,
+      "learning_rate": 4.649486258675989e-06,
+      "loss": 1.1521,
+      "step": 6566
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4381816934589031,
+      "learning_rate": 4.6481657486299965e-06,
+      "loss": 0.2135,
+      "step": 6567
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.2725112225425272,
+      "learning_rate": 4.646845263246052e-06,
+      "loss": 0.8093,
+      "step": 6568
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.744853814936621,
+      "learning_rate": 4.645524802616717e-06,
+      "loss": 0.3797,
+      "step": 6569
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.3123920628629215,
+      "learning_rate": 4.644204366834551e-06,
+      "loss": 0.7096,
+      "step": 6570
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.841646773780519,
+      "learning_rate": 4.6428839559921115e-06,
+      "loss": 0.9956,
+      "step": 6571
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.0695846330423233,
+      "learning_rate": 4.6415635701819515e-06,
+      "loss": 0.5472,
+      "step": 6572
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.7488795781565565,
+      "learning_rate": 4.640243209496627e-06,
+      "loss": 0.8055,
+      "step": 6573
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.29002043121706,
+      "learning_rate": 4.638922874028686e-06,
+      "loss": 1.171,
+      "step": 6574
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.652437636545116,
+      "learning_rate": 4.63760256387068e-06,
+      "loss": 0.746,
+      "step": 6575
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.036073634733401,
+      "learning_rate": 4.636282279115157e-06,
+      "loss": 0.3928,
+      "step": 6576
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.9691247579664775,
+      "learning_rate": 4.634962019854664e-06,
+      "loss": 0.64,
+      "step": 6577
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.546770284448663,
+      "learning_rate": 4.633641786181746e-06,
+      "loss": 1.0569,
+      "step": 6578
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.7128870863572634,
+      "learning_rate": 4.632321578188943e-06,
+      "loss": 0.5789,
+      "step": 6579
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.9631752562069495,
+      "learning_rate": 4.6310013959687985e-06,
+      "loss": 0.7038,
+      "step": 6580
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.112639351331203,
+      "learning_rate": 4.629681239613848e-06,
+      "loss": 0.499,
+      "step": 6581
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.10163436923461,
+      "learning_rate": 4.628361109216633e-06,
+      "loss": 0.5696,
+      "step": 6582
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.5160555089852474,
+      "learning_rate": 4.627041004869684e-06,
+      "loss": 0.7325,
+      "step": 6583
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.279189696864294,
+      "learning_rate": 4.625720926665542e-06,
+      "loss": 0.8291,
+      "step": 6584
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.2144240576872196,
+      "learning_rate": 4.624400874696731e-06,
+      "loss": 0.4094,
+      "step": 6585
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.3151730118113907,
+      "learning_rate": 4.623080849055784e-06,
+      "loss": 0.4029,
+      "step": 6586
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.492740953954248,
+      "learning_rate": 4.6217608498352305e-06,
+      "loss": 0.8886,
+      "step": 6587
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.2091740236208612,
+      "learning_rate": 4.620440877127594e-06,
+      "loss": 0.5968,
+      "step": 6588
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.130614751465802,
+      "learning_rate": 4.619120931025401e-06,
+      "loss": 0.9419,
+      "step": 6589
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.9276537385180665,
+      "learning_rate": 4.617801011621175e-06,
+      "loss": 1.1146,
+      "step": 6590
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.517852696399206,
+      "learning_rate": 4.6164811190074314e-06,
+      "loss": 0.7668,
+      "step": 6591
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.9414529003498098,
+      "learning_rate": 4.615161253276693e-06,
+      "loss": 0.7907,
+      "step": 6592
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.829034509793269,
+      "learning_rate": 4.613841414521477e-06,
+      "loss": 0.7181,
+      "step": 6593
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.935798880433848,
+      "learning_rate": 4.612521602834297e-06,
+      "loss": 1.1686,
+      "step": 6594
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.9536139118306277,
+      "learning_rate": 4.611201818307666e-06,
+      "loss": 1.3524,
+      "step": 6595
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.098698829021364,
+      "learning_rate": 4.609882061034097e-06,
+      "loss": 0.9503,
+      "step": 6596
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.207210196248606,
+      "learning_rate": 4.608562331106096e-06,
+      "loss": 0.9931,
+      "step": 6597
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.2699870899783843,
+      "learning_rate": 4.607242628616171e-06,
+      "loss": 0.6659,
+      "step": 6598
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.2458183261833398,
+      "learning_rate": 4.605922953656829e-06,
+      "loss": 0.8099,
+      "step": 6599
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.6046779267216835,
+      "learning_rate": 4.604603306320574e-06,
+      "loss": 0.418,
+      "step": 6600
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.6416739769033932,
+      "learning_rate": 4.603283686699907e-06,
+      "loss": 0.5294,
+      "step": 6601
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.029470485869123,
+      "learning_rate": 4.601964094887327e-06,
+      "loss": 0.392,
+      "step": 6602
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.069810495616436,
+      "learning_rate": 4.600644530975331e-06,
+      "loss": 0.3439,
+      "step": 6603
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.1883788923176763,
+      "learning_rate": 4.599324995056415e-06,
+      "loss": 0.5165,
+      "step": 6604
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.3009145317121376,
+      "learning_rate": 4.598005487223073e-06,
+      "loss": 0.3408,
+      "step": 6605
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.1050672621062265,
+      "learning_rate": 4.596686007567797e-06,
+      "loss": 1.0344,
+      "step": 6606
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.8337463877223708,
+      "learning_rate": 4.595366556183079e-06,
+      "loss": 0.6772,
+      "step": 6607
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.48425966090482,
+      "learning_rate": 4.5940471331614014e-06,
+      "loss": 0.879,
+      "step": 6608
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.4774955498265934,
+      "learning_rate": 4.592727738595254e-06,
+      "loss": 0.6558,
+      "step": 6609
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.3091363448551796,
+      "learning_rate": 4.59140837257712e-06,
+      "loss": 0.668,
+      "step": 6610
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.8952445582123945,
+      "learning_rate": 4.59008903519948e-06,
+      "loss": 0.9435,
+      "step": 6611
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.7482706881409085,
+      "learning_rate": 4.588769726554814e-06,
+      "loss": 1.3747,
+      "step": 6612
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.4056673960688952,
+      "learning_rate": 4.587450446735604e-06,
+      "loss": 0.5709,
+      "step": 6613
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.7836948198696514,
+      "learning_rate": 4.586131195834319e-06,
+      "loss": 0.8782,
+      "step": 6614
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.9011139557679586,
+      "learning_rate": 4.584811973943437e-06,
+      "loss": 0.4426,
+      "step": 6615
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.6092322214967996,
+      "learning_rate": 4.583492781155428e-06,
+      "loss": 0.5468,
+      "step": 6616
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.779247375258173,
+      "learning_rate": 4.582173617562764e-06,
+      "loss": 0.7156,
+      "step": 6617
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.4377675092726236,
+      "learning_rate": 4.5808544832579095e-06,
+      "loss": 0.6088,
+      "step": 6618
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.2807342570697653,
+      "learning_rate": 4.579535378333334e-06,
+      "loss": 0.31,
+      "step": 6619
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.3071067500491274,
+      "learning_rate": 4.578216302881497e-06,
+      "loss": 0.3559,
+      "step": 6620
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.017469989892803,
+      "learning_rate": 4.576897256994861e-06,
+      "loss": 0.5663,
+      "step": 6621
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.223549494510722,
+      "learning_rate": 4.575578240765885e-06,
+      "loss": 0.5989,
+      "step": 6622
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 6.180017186692998,
+      "learning_rate": 4.574259254287028e-06,
+      "loss": 1.4431,
+      "step": 6623
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.595082314447604,
+      "learning_rate": 4.572940297650747e-06,
+      "loss": 1.1177,
+      "step": 6624
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3635989281444825,
+      "learning_rate": 4.57162137094949e-06,
+      "loss": 0.1789,
+      "step": 6625
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.05576306019229,
+      "learning_rate": 4.570302474275712e-06,
+      "loss": 0.3513,
+      "step": 6626
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.255732174985103,
+      "learning_rate": 4.568983607721859e-06,
+      "loss": 0.7319,
+      "step": 6627
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.9994180654619003,
+      "learning_rate": 4.567664771380379e-06,
+      "loss": 0.5239,
+      "step": 6628
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.9709190789507844,
+      "learning_rate": 4.566345965343718e-06,
+      "loss": 0.6596,
+      "step": 6629
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.4005590334116964,
+      "learning_rate": 4.5650271897043195e-06,
+      "loss": 0.6441,
+      "step": 6630
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.121247443116029,
+      "learning_rate": 4.563708444554619e-06,
+      "loss": 0.928,
+      "step": 6631
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.1053500780565817,
+      "learning_rate": 4.562389729987059e-06,
+      "loss": 0.6048,
+      "step": 6632
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.3267540639919053,
+      "learning_rate": 4.561071046094075e-06,
+      "loss": 0.3739,
+      "step": 6633
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.113156196181085,
+      "learning_rate": 4.5597523929680986e-06,
+      "loss": 0.9381,
+      "step": 6634
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.388165656572617,
+      "learning_rate": 4.558433770701565e-06,
+      "loss": 0.5184,
+      "step": 6635
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.799676204479062,
+      "learning_rate": 4.557115179386903e-06,
+      "loss": 0.6542,
+      "step": 6636
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.259239487215517,
+      "learning_rate": 4.555796619116538e-06,
+      "loss": 0.9032,
+      "step": 6637
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.6370771361107526,
+      "learning_rate": 4.554478089982897e-06,
+      "loss": 1.0361,
+      "step": 6638
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.756742079319636,
+      "learning_rate": 4.553159592078403e-06,
+      "loss": 0.6837,
+      "step": 6639
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.149914278587901,
+      "learning_rate": 4.551841125495477e-06,
+      "loss": 0.8688,
+      "step": 6640
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.347737195032477,
+      "learning_rate": 4.550522690326538e-06,
+      "loss": 0.7162,
+      "step": 6641
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.1569733317515665,
+      "learning_rate": 4.5492042866640045e-06,
+      "loss": 0.6957,
+      "step": 6642
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.681665935187307,
+      "learning_rate": 4.547885914600285e-06,
+      "loss": 1.0,
+      "step": 6643
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.004368873239083,
+      "learning_rate": 4.546567574227796e-06,
+      "loss": 1.0405,
+      "step": 6644
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.6669501577059718,
+      "learning_rate": 4.545249265638947e-06,
+      "loss": 0.4618,
+      "step": 6645
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.91651972542203,
+      "learning_rate": 4.543930988926145e-06,
+      "loss": 0.9557,
+      "step": 6646
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.1338220022417818,
+      "learning_rate": 4.542612744181799e-06,
+      "loss": 0.5671,
+      "step": 6647
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.158874878095333,
+      "learning_rate": 4.541294531498306e-06,
+      "loss": 1.2718,
+      "step": 6648
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.720407098990491,
+      "learning_rate": 4.539976350968071e-06,
+      "loss": 0.9709,
+      "step": 6649
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.2245280801283176,
+      "learning_rate": 4.53865820268349e-06,
+      "loss": 0.4954,
+      "step": 6650
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.147232656477744,
+      "learning_rate": 4.537340086736963e-06,
+      "loss": 0.4342,
+      "step": 6651
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.721820243847922,
+      "learning_rate": 4.536022003220882e-06,
+      "loss": 0.9516,
+      "step": 6652
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 6.140103242181031,
+      "learning_rate": 4.534703952227641e-06,
+      "loss": 1.5263,
+      "step": 6653
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8124082668299026,
+      "learning_rate": 4.533385933849626e-06,
+      "loss": 0.3272,
+      "step": 6654
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.421852785391417,
+      "learning_rate": 4.532067948179227e-06,
+      "loss": 0.9541,
+      "step": 6655
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.7558805640610675,
+      "learning_rate": 4.530749995308827e-06,
+      "loss": 0.8192,
+      "step": 6656
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.297025578531439,
+      "learning_rate": 4.529432075330811e-06,
+      "loss": 0.7935,
+      "step": 6657
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.023969964098497,
+      "learning_rate": 4.528114188337559e-06,
+      "loss": 1.112,
+      "step": 6658
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.27069614561819,
+      "learning_rate": 4.52679633442145e-06,
+      "loss": 0.7922,
+      "step": 6659
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.3307633982499767,
+      "learning_rate": 4.525478513674857e-06,
+      "loss": 0.3234,
+      "step": 6660
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.15814670184417,
+      "learning_rate": 4.5241607261901545e-06,
+      "loss": 0.3978,
+      "step": 6661
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.2839993498903466,
+      "learning_rate": 4.522842972059715e-06,
+      "loss": 0.46,
+      "step": 6662
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.085813041441138,
+      "learning_rate": 4.5215252513759065e-06,
+      "loss": 0.9939,
+      "step": 6663
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.624647901844173,
+      "learning_rate": 4.520207564231096e-06,
+      "loss": 0.7483,
+      "step": 6664
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.423834373085487,
+      "learning_rate": 4.518889910717646e-06,
+      "loss": 0.892,
+      "step": 6665
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3592898624579992,
+      "learning_rate": 4.517572290927922e-06,
+      "loss": 0.1814,
+      "step": 6666
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.094091936733789,
+      "learning_rate": 4.516254704954279e-06,
+      "loss": 0.9623,
+      "step": 6667
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.380531451903595,
+      "learning_rate": 4.514937152889074e-06,
+      "loss": 0.9385,
+      "step": 6668
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.548515060595646,
+      "learning_rate": 4.5136196348246655e-06,
+      "loss": 0.2365,
+      "step": 6669
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.871283572351933,
+      "learning_rate": 4.5123021508534055e-06,
+      "loss": 1.5944,
+      "step": 6670
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.586024416815253,
+      "learning_rate": 4.510984701067641e-06,
+      "loss": 0.6512,
+      "step": 6671
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.684192797698732,
+      "learning_rate": 4.509667285559719e-06,
+      "loss": 0.8768,
+      "step": 6672
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.4344216835852714,
+      "learning_rate": 4.5083499044219896e-06,
+      "loss": 0.9899,
+      "step": 6673
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.5196756901261192,
+      "learning_rate": 4.50703255774679e-06,
+      "loss": 0.8292,
+      "step": 6674
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8425163445693746,
+      "learning_rate": 4.505715245626462e-06,
+      "loss": 0.2697,
+      "step": 6675
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.7258897847745307,
+      "learning_rate": 4.5043979681533475e-06,
+      "loss": 0.5433,
+      "step": 6676
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.688133066641052,
+      "learning_rate": 4.5030807254197755e-06,
+      "loss": 1.1534,
+      "step": 6677
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.01387823574584,
+      "learning_rate": 4.501763517518082e-06,
+      "loss": 0.9374,
+      "step": 6678
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.690055779675321,
+      "learning_rate": 4.500446344540598e-06,
+      "loss": 1.1436,
+      "step": 6679
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.7781687517608864,
+      "learning_rate": 4.499129206579653e-06,
+      "loss": 0.5423,
+      "step": 6680
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.301595233545369,
+      "learning_rate": 4.497812103727569e-06,
+      "loss": 0.8588,
+      "step": 6681
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.7899519352888555,
+      "learning_rate": 4.496495036076673e-06,
+      "loss": 0.6174,
+      "step": 6682
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.606542284391494,
+      "learning_rate": 4.4951780037192814e-06,
+      "loss": 1.2281,
+      "step": 6683
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.4470797538474462,
+      "learning_rate": 4.4938610067477155e-06,
+      "loss": 0.5242,
+      "step": 6684
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.6842774274565326,
+      "learning_rate": 4.4925440452542905e-06,
+      "loss": 0.7994,
+      "step": 6685
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.313781227321232,
+      "learning_rate": 4.491227119331319e-06,
+      "loss": 0.7083,
+      "step": 6686
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.5080033230773133,
+      "learning_rate": 4.489910229071113e-06,
+      "loss": 0.7683,
+      "step": 6687
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.298710508526099,
+      "learning_rate": 4.488593374565979e-06,
+      "loss": 0.8617,
+      "step": 6688
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.9292771100654753,
+      "learning_rate": 4.487276555908225e-06,
+      "loss": 0.342,
+      "step": 6689
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.569969537758229,
+      "learning_rate": 4.4859597731901525e-06,
+      "loss": 0.8084,
+      "step": 6690
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.8138563785328468,
+      "learning_rate": 4.484643026504063e-06,
+      "loss": 0.8283,
+      "step": 6691
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.7839647659751714,
+      "learning_rate": 4.483326315942253e-06,
+      "loss": 0.5061,
+      "step": 6692
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.8229391413619958,
+      "learning_rate": 4.482009641597023e-06,
+      "loss": 0.8583,
+      "step": 6693
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5440870104220314,
+      "learning_rate": 4.48069300356066e-06,
+      "loss": 0.1245,
+      "step": 6694
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.9891717366161417,
+      "learning_rate": 4.479376401925457e-06,
+      "loss": 0.6948,
+      "step": 6695
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.258479099635694,
+      "learning_rate": 4.4780598367837045e-06,
+      "loss": 1.1285,
+      "step": 6696
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.046156263768596,
+      "learning_rate": 4.476743308227685e-06,
+      "loss": 0.3902,
+      "step": 6697
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.5052250731095094,
+      "learning_rate": 4.475426816349682e-06,
+      "loss": 0.4513,
+      "step": 6698
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.004491542181198,
+      "learning_rate": 4.4741103612419785e-06,
+      "loss": 1.2451,
+      "step": 6699
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.030912891854148,
+      "learning_rate": 4.472793942996848e-06,
+      "loss": 0.7664,
+      "step": 6700
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.0108380696713715,
+      "learning_rate": 4.471477561706567e-06,
+      "loss": 0.7946,
+      "step": 6701
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.347382738183261,
+      "learning_rate": 4.470161217463409e-06,
+      "loss": 0.7192,
+      "step": 6702
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.086513506690113,
+      "learning_rate": 4.468844910359645e-06,
+      "loss": 0.9626,
+      "step": 6703
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.7659728831313295,
+      "learning_rate": 4.467528640487541e-06,
+      "loss": 0.6089,
+      "step": 6704
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.9804538854072966,
+      "learning_rate": 4.466212407939362e-06,
+      "loss": 0.9291,
+      "step": 6705
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.299522483564469,
+      "learning_rate": 4.464896212807369e-06,
+      "loss": 0.2137,
+      "step": 6706
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.5323175506029703,
+      "learning_rate": 4.463580055183824e-06,
+      "loss": 0.8715,
+      "step": 6707
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7972727230387882,
+      "learning_rate": 4.462263935160982e-06,
+      "loss": 0.4094,
+      "step": 6708
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.017408192743553,
+      "learning_rate": 4.460947852831097e-06,
+      "loss": 0.8983,
+      "step": 6709
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.943276224413567,
+      "learning_rate": 4.459631808286424e-06,
+      "loss": 0.5783,
+      "step": 6710
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.3353441614559265,
+      "learning_rate": 4.458315801619208e-06,
+      "loss": 0.37,
+      "step": 6711
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.189435494692342,
+      "learning_rate": 4.456999832921697e-06,
+      "loss": 1.2434,
+      "step": 6712
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.3544064433359244,
+      "learning_rate": 4.455683902286134e-06,
+      "loss": 0.9277,
+      "step": 6713
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.212854139587158,
+      "learning_rate": 4.454368009804761e-06,
+      "loss": 0.913,
+      "step": 6714
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.389668485127944,
+      "learning_rate": 4.453052155569816e-06,
+      "loss": 0.9254,
+      "step": 6715
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.041303875057696,
+      "learning_rate": 4.451736339673536e-06,
+      "loss": 0.5614,
+      "step": 6716
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.9140860955081225,
+      "learning_rate": 4.450420562208151e-06,
+      "loss": 0.62,
+      "step": 6717
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.6971464338906084,
+      "learning_rate": 4.449104823265893e-06,
+      "loss": 0.8889,
+      "step": 6718
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.4099851679890616,
+      "learning_rate": 4.4477891229389895e-06,
+      "loss": 0.6155,
+      "step": 6719
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.341321347347397,
+      "learning_rate": 4.446473461319664e-06,
+      "loss": 1.0529,
+      "step": 6720
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.3881489681389207,
+      "learning_rate": 4.445157838500141e-06,
+      "loss": 0.4215,
+      "step": 6721
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9211033591281976,
+      "learning_rate": 4.44384225457264e-06,
+      "loss": 0.4279,
+      "step": 6722
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.70667426714046,
+      "learning_rate": 4.442526709629376e-06,
+      "loss": 0.4008,
+      "step": 6723
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.1879485480740595,
+      "learning_rate": 4.441211203762562e-06,
+      "loss": 0.7089,
+      "step": 6724
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.074765884183015,
+      "learning_rate": 4.439895737064411e-06,
+      "loss": 0.935,
+      "step": 6725
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.171599425732302,
+      "learning_rate": 4.438580309627132e-06,
+      "loss": 1.009,
+      "step": 6726
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.8459991702099563,
+      "learning_rate": 4.437264921542931e-06,
+      "loss": 0.7308,
+      "step": 6727
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.9044927686931064,
+      "learning_rate": 4.4359495729040095e-06,
+      "loss": 0.8916,
+      "step": 6728
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.7245129839287126,
+      "learning_rate": 4.434634263802567e-06,
+      "loss": 0.484,
+      "step": 6729
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.745818385243437,
+      "learning_rate": 4.433318994330802e-06,
+      "loss": 1.1118,
+      "step": 6730
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.9149526073559415,
+      "learning_rate": 4.43200376458091e-06,
+      "loss": 0.7711,
+      "step": 6731
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.0550435934262605,
+      "learning_rate": 4.430688574645081e-06,
+      "loss": 0.9976,
+      "step": 6732
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.442141227850007,
+      "learning_rate": 4.429373424615509e-06,
+      "loss": 1.1589,
+      "step": 6733
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.4050159968489413,
+      "learning_rate": 4.428058314584373e-06,
+      "loss": 0.7095,
+      "step": 6734
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.473274924192312,
+      "learning_rate": 4.426743244643862e-06,
+      "loss": 1.6246,
+      "step": 6735
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.516636084525008,
+      "learning_rate": 4.425428214886153e-06,
+      "loss": 1.0359,
+      "step": 6736
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.0959314160256275,
+      "learning_rate": 4.424113225403425e-06,
+      "loss": 1.5183,
+      "step": 6737
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.9595000335141557,
+      "learning_rate": 4.422798276287855e-06,
+      "loss": 0.7305,
+      "step": 6738
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.95466063717298,
+      "learning_rate": 4.421483367631616e-06,
+      "loss": 0.6199,
+      "step": 6739
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.506402893224004,
+      "learning_rate": 4.420168499526872e-06,
+      "loss": 1.046,
+      "step": 6740
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.537825050212767,
+      "learning_rate": 4.418853672065794e-06,
+      "loss": 1.3168,
+      "step": 6741
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.8395774802404476,
+      "learning_rate": 4.4175388853405445e-06,
+      "loss": 0.8424,
+      "step": 6742
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.11543314203391,
+      "learning_rate": 4.4162241394432834e-06,
+      "loss": 1.1579,
+      "step": 6743
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.842197311328773,
+      "learning_rate": 4.414909434466172e-06,
+      "loss": 1.0507,
+      "step": 6744
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.917293590730472,
+      "learning_rate": 4.41359477050136e-06,
+      "loss": 0.6057,
+      "step": 6745
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.337147638873848,
+      "learning_rate": 4.412280147641003e-06,
+      "loss": 1.1647,
+      "step": 6746
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.5815547564880315,
+      "learning_rate": 4.410965565977251e-06,
+      "loss": 0.5415,
+      "step": 6747
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.4075359973084156,
+      "learning_rate": 4.409651025602248e-06,
+      "loss": 0.8486,
+      "step": 6748
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.8845777983966663,
+      "learning_rate": 4.40833652660814e-06,
+      "loss": 1.0452,
+      "step": 6749
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.680515425628574,
+      "learning_rate": 4.407022069087067e-06,
+      "loss": 0.7546,
+      "step": 6750
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.7372695630352295,
+      "learning_rate": 4.405707653131166e-06,
+      "loss": 0.2813,
+      "step": 6751
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.607317444290473,
+      "learning_rate": 4.404393278832572e-06,
+      "loss": 0.5272,
+      "step": 6752
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.780700197590161,
+      "learning_rate": 4.403078946283416e-06,
+      "loss": 1.0148,
+      "step": 6753
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.169287698761378,
+      "learning_rate": 4.401764655575828e-06,
+      "loss": 1.033,
+      "step": 6754
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.5874236085275633,
+      "learning_rate": 4.400450406801935e-06,
+      "loss": 0.8101,
+      "step": 6755
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.411895553826789,
+      "learning_rate": 4.39913620005386e-06,
+      "loss": 0.5623,
+      "step": 6756
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.1675888394351133,
+      "learning_rate": 4.3978220354237215e-06,
+      "loss": 0.6919,
+      "step": 6757
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.7486331496396943,
+      "learning_rate": 4.396507913003638e-06,
+      "loss": 0.3992,
+      "step": 6758
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.34685623752115,
+      "learning_rate": 4.395193832885723e-06,
+      "loss": 0.7302,
+      "step": 6759
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.252400234021796,
+      "learning_rate": 4.393879795162088e-06,
+      "loss": 0.9949,
+      "step": 6760
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.978397095087008,
+      "learning_rate": 4.392565799924841e-06,
+      "loss": 0.7132,
+      "step": 6761
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.466544139260184,
+      "learning_rate": 4.391251847266091e-06,
+      "loss": 0.7536,
+      "step": 6762
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6048107934800868,
+      "learning_rate": 4.389937937277934e-06,
+      "loss": 0.3515,
+      "step": 6763
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.7738442378916877,
+      "learning_rate": 4.388624070052473e-06,
+      "loss": 0.7473,
+      "step": 6764
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.139769946318573,
+      "learning_rate": 4.387310245681805e-06,
+      "loss": 0.6443,
+      "step": 6765
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.117095300850483,
+      "learning_rate": 4.385996464258021e-06,
+      "loss": 0.7387,
+      "step": 6766
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.228842653356213,
+      "learning_rate": 4.384682725873215e-06,
+      "loss": 1.046,
+      "step": 6767
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.4775465019126184,
+      "learning_rate": 4.383369030619471e-06,
+      "loss": 0.7722,
+      "step": 6768
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.1534807109644785,
+      "learning_rate": 4.382055378588873e-06,
+      "loss": 0.5986,
+      "step": 6769
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.5132800394612183,
+      "learning_rate": 4.380741769873504e-06,
+      "loss": 0.5259,
+      "step": 6770
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.468051046678788,
+      "learning_rate": 4.379428204565442e-06,
+      "loss": 1.0535,
+      "step": 6771
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.447498264682462,
+      "learning_rate": 4.378114682756764e-06,
+      "loss": 0.6449,
+      "step": 6772
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.113338299264036,
+      "learning_rate": 4.3768012045395395e-06,
+      "loss": 0.4076,
+      "step": 6773
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.2080865823927436,
+      "learning_rate": 4.375487770005837e-06,
+      "loss": 0.5992,
+      "step": 6774
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.74180423498849,
+      "learning_rate": 4.374174379247726e-06,
+      "loss": 0.5277,
+      "step": 6775
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.7026056758885635,
+      "learning_rate": 4.372861032357265e-06,
+      "loss": 0.5905,
+      "step": 6776
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.380073082270201,
+      "learning_rate": 4.371547729426517e-06,
+      "loss": 1.1251,
+      "step": 6777
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.17091378817613,
+      "learning_rate": 4.370234470547538e-06,
+      "loss": 0.9241,
+      "step": 6778
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.807829620202627,
+      "learning_rate": 4.3689212558123846e-06,
+      "loss": 0.9373,
+      "step": 6779
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.167161769805908,
+      "learning_rate": 4.367608085313102e-06,
+      "loss": 0.9693,
+      "step": 6780
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.0216638364730692,
+      "learning_rate": 4.36629495914174e-06,
+      "loss": 0.467,
+      "step": 6781
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.565712444915998,
+      "learning_rate": 4.364981877390345e-06,
+      "loss": 0.8293,
+      "step": 6782
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.587781430703773,
+      "learning_rate": 4.363668840150956e-06,
+      "loss": 0.8427,
+      "step": 6783
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.102385554796526,
+      "learning_rate": 4.362355847515614e-06,
+      "loss": 0.6292,
+      "step": 6784
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.8173956924667825,
+      "learning_rate": 4.36104289957635e-06,
+      "loss": 0.5476,
+      "step": 6785
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.8232940754466638,
+      "learning_rate": 4.359729996425198e-06,
+      "loss": 0.9364,
+      "step": 6786
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.1896318124601213,
+      "learning_rate": 4.358417138154186e-06,
+      "loss": 0.274,
+      "step": 6787
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.882635089458406,
+      "learning_rate": 4.357104324855342e-06,
+      "loss": 0.6152,
+      "step": 6788
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.9231177114410576,
+      "learning_rate": 4.355791556620686e-06,
+      "loss": 0.6004,
+      "step": 6789
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.074891731812101,
+      "learning_rate": 4.35447883354224e-06,
+      "loss": 0.306,
+      "step": 6790
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.845996157750704,
+      "learning_rate": 4.353166155712018e-06,
+      "loss": 1.2494,
+      "step": 6791
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.174181340098687,
+      "learning_rate": 4.351853523222032e-06,
+      "loss": 0.7239,
+      "step": 6792
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.133578831771037,
+      "learning_rate": 4.350540936164293e-06,
+      "loss": 1.085,
+      "step": 6793
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.31131228805281,
+      "learning_rate": 4.349228394630808e-06,
+      "loss": 0.84,
+      "step": 6794
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.599116948148214,
+      "learning_rate": 4.347915898713581e-06,
+      "loss": 0.6953,
+      "step": 6795
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.989679839429888,
+      "learning_rate": 4.346603448504614e-06,
+      "loss": 1.3605,
+      "step": 6796
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.9755423751338297,
+      "learning_rate": 4.345291044095898e-06,
+      "loss": 0.581,
+      "step": 6797
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.652500888585743,
+      "learning_rate": 4.343978685579433e-06,
+      "loss": 1.1582,
+      "step": 6798
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.579899946824376,
+      "learning_rate": 4.342666373047207e-06,
+      "loss": 0.8965,
+      "step": 6799
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.180070198560346,
+      "learning_rate": 4.341354106591205e-06,
+      "loss": 1.0277,
+      "step": 6800
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.1973548176785074,
+      "learning_rate": 4.340041886303415e-06,
+      "loss": 0.7563,
+      "step": 6801
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.585822185199269,
+      "learning_rate": 4.338729712275818e-06,
+      "loss": 0.4064,
+      "step": 6802
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.914939939545324,
+      "learning_rate": 4.337417584600389e-06,
+      "loss": 0.586,
+      "step": 6803
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.983914984009011,
+      "learning_rate": 4.336105503369104e-06,
+      "loss": 1.0038,
+      "step": 6804
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.850325783736151,
+      "learning_rate": 4.334793468673935e-06,
+      "loss": 0.913,
+      "step": 6805
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.1456873238432825,
+      "learning_rate": 4.333481480606847e-06,
+      "loss": 0.3679,
+      "step": 6806
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.485251348069992,
+      "learning_rate": 4.332169539259809e-06,
+      "loss": 0.8325,
+      "step": 6807
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.295788439384921,
+      "learning_rate": 4.330857644724778e-06,
+      "loss": 0.2682,
+      "step": 6808
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.598856570533403,
+      "learning_rate": 4.329545797093713e-06,
+      "loss": 0.287,
+      "step": 6809
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.7782431799103495,
+      "learning_rate": 4.3282339964585705e-06,
+      "loss": 0.3696,
+      "step": 6810
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.723726663255735,
+      "learning_rate": 4.326922242911302e-06,
+      "loss": 0.9142,
+      "step": 6811
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.014153303089965,
+      "learning_rate": 4.325610536543855e-06,
+      "loss": 1.1812,
+      "step": 6812
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.004017551772608,
+      "learning_rate": 4.324298877448176e-06,
+      "loss": 0.98,
+      "step": 6813
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.3388195160299645,
+      "learning_rate": 4.3229872657162034e-06,
+      "loss": 0.3867,
+      "step": 6814
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.417142475877144,
+      "learning_rate": 4.3216757014398755e-06,
+      "loss": 0.9917,
+      "step": 6815
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.5876620978645044,
+      "learning_rate": 4.320364184711131e-06,
+      "loss": 0.3205,
+      "step": 6816
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.843499879562751,
+      "learning_rate": 4.319052715621898e-06,
+      "loss": 0.7618,
+      "step": 6817
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.722391630062092,
+      "learning_rate": 4.317741294264106e-06,
+      "loss": 0.8201,
+      "step": 6818
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.7302836216449453,
+      "learning_rate": 4.3164299207296824e-06,
+      "loss": 0.6738,
+      "step": 6819
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.365487702840651,
+      "learning_rate": 4.315118595110545e-06,
+      "loss": 0.4978,
+      "step": 6820
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.5141520687520944,
+      "learning_rate": 4.313807317498614e-06,
+      "loss": 0.608,
+      "step": 6821
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.628472427927134,
+      "learning_rate": 4.312496087985802e-06,
+      "loss": 1.4771,
+      "step": 6822
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.272710188604787,
+      "learning_rate": 4.3111849066640234e-06,
+      "loss": 0.6231,
+      "step": 6823
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.9297442148002624,
+      "learning_rate": 4.309873773625187e-06,
+      "loss": 0.8096,
+      "step": 6824
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0858669492068054,
+      "learning_rate": 4.308562688961193e-06,
+      "loss": 0.448,
+      "step": 6825
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.9865323854928296,
+      "learning_rate": 4.3072516527639456e-06,
+      "loss": 0.874,
+      "step": 6826
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.3713817084126427,
+      "learning_rate": 4.305940665125342e-06,
+      "loss": 0.5426,
+      "step": 6827
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.9105183931123766,
+      "learning_rate": 4.304629726137279e-06,
+      "loss": 0.7191,
+      "step": 6828
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.5341255524071977,
+      "learning_rate": 4.303318835891645e-06,
+      "loss": 0.7492,
+      "step": 6829
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3118491259015854,
+      "learning_rate": 4.302007994480331e-06,
+      "loss": 0.1992,
+      "step": 6830
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.6402227805381098,
+      "learning_rate": 4.300697201995216e-06,
+      "loss": 0.8198,
+      "step": 6831
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.861804706465785,
+      "learning_rate": 4.299386458528184e-06,
+      "loss": 0.5432,
+      "step": 6832
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.6059084259740115,
+      "learning_rate": 4.298075764171112e-06,
+      "loss": 0.581,
+      "step": 6833
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.4516925600421895,
+      "learning_rate": 4.2967651190158745e-06,
+      "loss": 0.5308,
+      "step": 6834
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.462719347637248,
+      "learning_rate": 4.295454523154342e-06,
+      "loss": 0.7217,
+      "step": 6835
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.8513362667586333,
+      "learning_rate": 4.294143976678382e-06,
+      "loss": 0.4031,
+      "step": 6836
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.175222583089542,
+      "learning_rate": 4.292833479679857e-06,
+      "loss": 1.0811,
+      "step": 6837
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.269094356348749,
+      "learning_rate": 4.291523032250627e-06,
+      "loss": 1.0881,
+      "step": 6838
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.5117461525024996,
+      "learning_rate": 4.290212634482549e-06,
+      "loss": 0.3385,
+      "step": 6839
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.76109821823398,
+      "learning_rate": 4.2889022864674755e-06,
+      "loss": 1.2529,
+      "step": 6840
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.260502714074846,
+      "learning_rate": 4.287591988297257e-06,
+      "loss": 0.9448,
+      "step": 6841
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.6874940118687105,
+      "learning_rate": 4.286281740063743e-06,
+      "loss": 0.8363,
+      "step": 6842
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.32013293811396,
+      "learning_rate": 4.28497154185877e-06,
+      "loss": 0.3973,
+      "step": 6843
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.7190494425481817,
+      "learning_rate": 4.283661393774181e-06,
+      "loss": 0.8758,
+      "step": 6844
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.4261964662966773,
+      "learning_rate": 4.28235129590181e-06,
+      "loss": 1.0139,
+      "step": 6845
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.030158882763015,
+      "learning_rate": 4.28104124833349e-06,
+      "loss": 0.8224,
+      "step": 6846
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.473704631683547,
+      "learning_rate": 4.279731251161051e-06,
+      "loss": 0.5438,
+      "step": 6847
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.1995239724376523,
+      "learning_rate": 4.278421304476316e-06,
+      "loss": 0.574,
+      "step": 6848
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.054270936042657,
+      "learning_rate": 4.277111408371106e-06,
+      "loss": 0.704,
+      "step": 6849
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.0846166069226713,
+      "learning_rate": 4.27580156293724e-06,
+      "loss": 1.0241,
+      "step": 6850
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.395305755295111,
+      "learning_rate": 4.274491768266535e-06,
+      "loss": 1.1476,
+      "step": 6851
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.7818834248217987,
+      "learning_rate": 4.273182024450799e-06,
+      "loss": 0.856,
+      "step": 6852
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.8201990121495544,
+      "learning_rate": 4.271872331581841e-06,
+      "loss": 0.4176,
+      "step": 6853
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.316887402577849,
+      "learning_rate": 4.270562689751461e-06,
+      "loss": 0.7332,
+      "step": 6854
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.173018717309746,
+      "learning_rate": 4.2692530990514625e-06,
+      "loss": 0.5993,
+      "step": 6855
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.8334914810271368,
+      "learning_rate": 4.267943559573642e-06,
+      "loss": 0.7014,
+      "step": 6856
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.8593540784128573,
+      "learning_rate": 4.2666340714097915e-06,
+      "loss": 0.6737,
+      "step": 6857
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.774865952838886,
+      "learning_rate": 4.265324634651703e-06,
+      "loss": 0.4685,
+      "step": 6858
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.985799646092007,
+      "learning_rate": 4.264015249391159e-06,
+      "loss": 0.4988,
+      "step": 6859
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.204955750493154,
+      "learning_rate": 4.2627059157199435e-06,
+      "loss": 0.6423,
+      "step": 6860
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.36506220921179,
+      "learning_rate": 4.261396633729834e-06,
+      "loss": 1.126,
+      "step": 6861
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.383543259218011,
+      "learning_rate": 4.260087403512605e-06,
+      "loss": 0.7018,
+      "step": 6862
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.8830217325549654,
+      "learning_rate": 4.25877822516003e-06,
+      "loss": 0.6265,
+      "step": 6863
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.628668987731295,
+      "learning_rate": 4.2574690987638745e-06,
+      "loss": 0.8785,
+      "step": 6864
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.219833557255795,
+      "learning_rate": 4.2561600244159066e-06,
+      "loss": 0.7595,
+      "step": 6865
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.6218901336310694,
+      "learning_rate": 4.254851002207882e-06,
+      "loss": 0.5658,
+      "step": 6866
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.121018907344332,
+      "learning_rate": 4.253542032231559e-06,
+      "loss": 0.8594,
+      "step": 6867
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.638940263029625,
+      "learning_rate": 4.252233114578691e-06,
+      "loss": 0.5599,
+      "step": 6868
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.9511356064965457,
+      "learning_rate": 4.250924249341028e-06,
+      "loss": 0.7547,
+      "step": 6869
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.170116867479131,
+      "learning_rate": 4.249615436610316e-06,
+      "loss": 1.0425,
+      "step": 6870
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.177039311445076,
+      "learning_rate": 4.248306676478295e-06,
+      "loss": 0.7487,
+      "step": 6871
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.817298145174552,
+      "learning_rate": 4.246997969036703e-06,
+      "loss": 0.3305,
+      "step": 6872
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.12940330283581,
+      "learning_rate": 4.245689314377277e-06,
+      "loss": 0.6051,
+      "step": 6873
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.131631486801754,
+      "learning_rate": 4.244380712591749e-06,
+      "loss": 0.4731,
+      "step": 6874
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.4150405961163743,
+      "learning_rate": 4.243072163771843e-06,
+      "loss": 0.3221,
+      "step": 6875
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.999646529642497,
+      "learning_rate": 4.241763668009286e-06,
+      "loss": 0.9084,
+      "step": 6876
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.358170987405386,
+      "learning_rate": 4.240455225395796e-06,
+      "loss": 0.7714,
+      "step": 6877
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.5208093951344193,
+      "learning_rate": 4.239146836023087e-06,
+      "loss": 0.3809,
+      "step": 6878
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.927345876190294,
+      "learning_rate": 4.237838499982874e-06,
+      "loss": 1.2175,
+      "step": 6879
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.817855918603794,
+      "learning_rate": 4.236530217366865e-06,
+      "loss": 1.1805,
+      "step": 6880
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.6204962988593254,
+      "learning_rate": 4.235221988266766e-06,
+      "loss": 0.547,
+      "step": 6881
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.289319688627827,
+      "learning_rate": 4.233913812774278e-06,
+      "loss": 0.899,
+      "step": 6882
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.9054539055753694,
+      "learning_rate": 4.232605690981096e-06,
+      "loss": 0.9904,
+      "step": 6883
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.786545958744005,
+      "learning_rate": 4.231297622978917e-06,
+      "loss": 0.9344,
+      "step": 6884
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.10221020007079,
+      "learning_rate": 4.229989608859428e-06,
+      "loss": 1.1117,
+      "step": 6885
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.2718030494194648,
+      "learning_rate": 4.228681648714317e-06,
+      "loss": 0.4801,
+      "step": 6886
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.260089357399006,
+      "learning_rate": 4.2273737426352665e-06,
+      "loss": 0.851,
+      "step": 6887
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.5932974222685434,
+      "learning_rate": 4.226065890713953e-06,
+      "loss": 0.781,
+      "step": 6888
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.915015621717631,
+      "learning_rate": 4.224758093042052e-06,
+      "loss": 0.6147,
+      "step": 6889
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.5451163006412014,
+      "learning_rate": 4.223450349711235e-06,
+      "loss": 0.732,
+      "step": 6890
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.196982720790917,
+      "learning_rate": 4.222142660813169e-06,
+      "loss": 1.0102,
+      "step": 6891
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.8062168968693655,
+      "learning_rate": 4.220835026439517e-06,
+      "loss": 0.7792,
+      "step": 6892
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.01107608177263,
+      "learning_rate": 4.219527446681941e-06,
+      "loss": 0.9339,
+      "step": 6893
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.502385615477792,
+      "learning_rate": 4.218219921632093e-06,
+      "loss": 0.9728,
+      "step": 6894
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.103095125130039,
+      "learning_rate": 4.2169124513816245e-06,
+      "loss": 0.9507,
+      "step": 6895
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.0188942558864795,
+      "learning_rate": 4.2156050360221855e-06,
+      "loss": 0.9009,
+      "step": 6896
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.223842439731416,
+      "learning_rate": 4.21429767564542e-06,
+      "loss": 0.9399,
+      "step": 6897
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.502148157216099,
+      "learning_rate": 4.21299037034297e-06,
+      "loss": 0.6637,
+      "step": 6898
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.084211832079616,
+      "learning_rate": 4.211683120206469e-06,
+      "loss": 1.052,
+      "step": 6899
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.043921168635646,
+      "learning_rate": 4.21037592532755e-06,
+      "loss": 0.3334,
+      "step": 6900
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.333588962768206,
+      "learning_rate": 4.209068785797842e-06,
+      "loss": 0.6464,
+      "step": 6901
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.1643242914476115,
+      "learning_rate": 4.20776170170897e-06,
+      "loss": 1.1004,
+      "step": 6902
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.581833199513512,
+      "learning_rate": 4.2064546731525545e-06,
+      "loss": 0.766,
+      "step": 6903
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.16260367601079,
+      "learning_rate": 4.205147700220214e-06,
+      "loss": 0.6797,
+      "step": 6904
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.0079859962392805,
+      "learning_rate": 4.203840783003561e-06,
+      "loss": 0.8811,
+      "step": 6905
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.7685315304775404,
+      "learning_rate": 4.202533921594203e-06,
+      "loss": 0.5685,
+      "step": 6906
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.8447103263989066,
+      "learning_rate": 4.201227116083747e-06,
+      "loss": 0.8439,
+      "step": 6907
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3527188800025902,
+      "learning_rate": 4.199920366563793e-06,
+      "loss": 0.1809,
+      "step": 6908
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.6407532064114183,
+      "learning_rate": 4.19861367312594e-06,
+      "loss": 0.4898,
+      "step": 6909
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.9041728475041895,
+      "learning_rate": 4.197307035861783e-06,
+      "loss": 0.7451,
+      "step": 6910
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.247228990935387,
+      "learning_rate": 4.196000454862907e-06,
+      "loss": 0.598,
+      "step": 6911
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7470822297571256,
+      "learning_rate": 4.194693930220899e-06,
+      "loss": 0.3764,
+      "step": 6912
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.8398047777086575,
+      "learning_rate": 4.193387462027343e-06,
+      "loss": 0.8945,
+      "step": 6913
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.47561128374342,
+      "learning_rate": 4.1920810503738165e-06,
+      "loss": 0.6011,
+      "step": 6914
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.356979311294287,
+      "learning_rate": 4.190774695351891e-06,
+      "loss": 0.6791,
+      "step": 6915
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6033166663574987,
+      "learning_rate": 4.18946839705314e-06,
+      "loss": 0.3343,
+      "step": 6916
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.180106896231907,
+      "learning_rate": 4.188162155569124e-06,
+      "loss": 0.9223,
+      "step": 6917
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.3692075779863893,
+      "learning_rate": 4.186855970991409e-06,
+      "loss": 0.5588,
+      "step": 6918
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.250783680882565,
+      "learning_rate": 4.18554984341155e-06,
+      "loss": 0.2383,
+      "step": 6919
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.012464368258494,
+      "learning_rate": 4.184243772921104e-06,
+      "loss": 0.8274,
+      "step": 6920
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.615573314560767,
+      "learning_rate": 4.18293775961162e-06,
+      "loss": 1.0733,
+      "step": 6921
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.063651310014804,
+      "learning_rate": 4.181631803574643e-06,
+      "loss": 0.7932,
+      "step": 6922
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.69790970871073,
+      "learning_rate": 4.180325904901715e-06,
+      "loss": 0.9528,
+      "step": 6923
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.9197981238712365,
+      "learning_rate": 4.179020063684373e-06,
+      "loss": 0.6375,
+      "step": 6924
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.274650253765275,
+      "learning_rate": 4.177714280014151e-06,
+      "loss": 0.8576,
+      "step": 6925
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.550335842713816,
+      "learning_rate": 4.176408553982581e-06,
+      "loss": 1.2013,
+      "step": 6926
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.934264354807582,
+      "learning_rate": 4.175102885681187e-06,
+      "loss": 0.4646,
+      "step": 6927
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.6260526181564816,
+      "learning_rate": 4.17379727520149e-06,
+      "loss": 0.834,
+      "step": 6928
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.768560143083979,
+      "learning_rate": 4.1724917226350084e-06,
+      "loss": 1.3616,
+      "step": 6929
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.8626606591188324,
+      "learning_rate": 4.171186228073256e-06,
+      "loss": 0.8391,
+      "step": 6930
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.188468437834785,
+      "learning_rate": 4.169880791607741e-06,
+      "loss": 1.2928,
+      "step": 6931
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.109587675048897,
+      "learning_rate": 4.168575413329971e-06,
+      "loss": 0.6591,
+      "step": 6932
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.8197203035624305,
+      "learning_rate": 4.167270093331447e-06,
+      "loss": 0.6995,
+      "step": 6933
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.8125429356691902,
+      "learning_rate": 4.165964831703663e-06,
+      "loss": 0.4956,
+      "step": 6934
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.289835103679579,
+      "learning_rate": 4.164659628538116e-06,
+      "loss": 1.5974,
+      "step": 6935
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.8978445321359225,
+      "learning_rate": 4.163354483926292e-06,
+      "loss": 0.5473,
+      "step": 6936
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.442305017248432,
+      "learning_rate": 4.1620493979596795e-06,
+      "loss": 0.3806,
+      "step": 6937
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.3640488602564815,
+      "learning_rate": 4.160744370729757e-06,
+      "loss": 0.9416,
+      "step": 6938
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.333629385556718,
+      "learning_rate": 4.159439402328003e-06,
+      "loss": 0.8327,
+      "step": 6939
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.066121830157949,
+      "learning_rate": 4.158134492845886e-06,
+      "loss": 0.4121,
+      "step": 6940
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.269603193669817,
+      "learning_rate": 4.1568296423748785e-06,
+      "loss": 0.6631,
+      "step": 6941
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0426816193321784,
+      "learning_rate": 4.155524851006444e-06,
+      "loss": 0.1427,
+      "step": 6942
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.587387024359568,
+      "learning_rate": 4.154220118832041e-06,
+      "loss": 0.9754,
+      "step": 6943
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2798933960380874,
+      "learning_rate": 4.1529154459431285e-06,
+      "loss": 0.2721,
+      "step": 6944
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.342283592432994,
+      "learning_rate": 4.151610832431156e-06,
+      "loss": 1.3131,
+      "step": 6945
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.365002485112433,
+      "learning_rate": 4.150306278387573e-06,
+      "loss": 1.095,
+      "step": 6946
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.6816522520739703,
+      "learning_rate": 4.14900178390382e-06,
+      "loss": 0.4537,
+      "step": 6947
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.7082743424961486,
+      "learning_rate": 4.147697349071339e-06,
+      "loss": 0.6203,
+      "step": 6948
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.7347242164312022,
+      "learning_rate": 4.146392973981564e-06,
+      "loss": 0.6242,
+      "step": 6949
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.870684993167908,
+      "learning_rate": 4.14508865872593e-06,
+      "loss": 0.9103,
+      "step": 6950
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.7617659166446287,
+      "learning_rate": 4.143784403395858e-06,
+      "loss": 0.5403,
+      "step": 6951
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.47068227471356,
+      "learning_rate": 4.142480208082771e-06,
+      "loss": 0.9692,
+      "step": 6952
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.2552461800628345,
+      "learning_rate": 4.141176072878093e-06,
+      "loss": 1.0206,
+      "step": 6953
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.822356209708549,
+      "learning_rate": 4.1398719978732324e-06,
+      "loss": 1.2444,
+      "step": 6954
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.07159222227165,
+      "learning_rate": 4.138567983159601e-06,
+      "loss": 0.7202,
+      "step": 6955
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.20873396021626,
+      "learning_rate": 4.137264028828609e-06,
+      "loss": 0.6758,
+      "step": 6956
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.56375038944988,
+      "learning_rate": 4.1359601349716504e-06,
+      "loss": 1.035,
+      "step": 6957
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.0266586675579226,
+      "learning_rate": 4.134656301680126e-06,
+      "loss": 0.8107,
+      "step": 6958
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.6120283958518833,
+      "learning_rate": 4.133352529045429e-06,
+      "loss": 0.6101,
+      "step": 6959
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.9478032953611444,
+      "learning_rate": 4.13204881715895e-06,
+      "loss": 0.8561,
+      "step": 6960
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.541264148590845,
+      "learning_rate": 4.130745166112069e-06,
+      "loss": 0.9062,
+      "step": 6961
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.446684652833886,
+      "learning_rate": 4.129441575996172e-06,
+      "loss": 0.7517,
+      "step": 6962
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.053587922886659,
+      "learning_rate": 4.128138046902629e-06,
+      "loss": 0.9202,
+      "step": 6963
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.1761630302225887,
+      "learning_rate": 4.126834578922816e-06,
+      "loss": 0.836,
+      "step": 6964
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.8479761947356645,
+      "learning_rate": 4.1255311721480975e-06,
+      "loss": 1.0355,
+      "step": 6965
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.74405953449905,
+      "learning_rate": 4.124227826669839e-06,
+      "loss": 0.6668,
+      "step": 6966
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.427789690531807,
+      "learning_rate": 4.1229245425794004e-06,
+      "loss": 0.4646,
+      "step": 6967
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.196806288674009,
+      "learning_rate": 4.121621319968131e-06,
+      "loss": 1.1004,
+      "step": 6968
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.8513313981058475,
+      "learning_rate": 4.120318158927387e-06,
+      "loss": 0.7007,
+      "step": 6969
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.613814344524273,
+      "learning_rate": 4.11901505954851e-06,
+      "loss": 0.6972,
+      "step": 6970
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.802800757327425,
+      "learning_rate": 4.117712021922843e-06,
+      "loss": 0.502,
+      "step": 6971
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.7788402929949156,
+      "learning_rate": 4.116409046141725e-06,
+      "loss": 0.8423,
+      "step": 6972
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.9727719345173997,
+      "learning_rate": 4.115106132296488e-06,
+      "loss": 1.042,
+      "step": 6973
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.137717977015083,
+      "learning_rate": 4.113803280478458e-06,
+      "loss": 0.3862,
+      "step": 6974
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.174766112087384,
+      "learning_rate": 4.112500490778962e-06,
+      "loss": 0.6982,
+      "step": 6975
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.699326857290857,
+      "learning_rate": 4.1111977632893195e-06,
+      "loss": 0.847,
+      "step": 6976
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.016945658519239,
+      "learning_rate": 4.109895098100845e-06,
+      "loss": 0.8596,
+      "step": 6977
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.494855177875274,
+      "learning_rate": 4.108592495304851e-06,
+      "loss": 1.012,
+      "step": 6978
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.9046747025960418,
+      "learning_rate": 4.107289954992646e-06,
+      "loss": 0.5896,
+      "step": 6979
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.042097661435592,
+      "learning_rate": 4.1059874772555265e-06,
+      "loss": 0.8972,
+      "step": 6980
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.9732746412026763,
+      "learning_rate": 4.104685062184795e-06,
+      "loss": 0.8474,
+      "step": 6981
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.6544356724749,
+      "learning_rate": 4.103382709871744e-06,
+      "loss": 1.3651,
+      "step": 6982
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.6851193126074655,
+      "learning_rate": 4.102080420407662e-06,
+      "loss": 0.6084,
+      "step": 6983
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.122859640156609,
+      "learning_rate": 4.100778193883838e-06,
+      "loss": 1.1105,
+      "step": 6984
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.0783988319102744,
+      "learning_rate": 4.099476030391548e-06,
+      "loss": 0.4396,
+      "step": 6985
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.685476561129729,
+      "learning_rate": 4.098173930022069e-06,
+      "loss": 1.0255,
+      "step": 6986
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.006861256635884,
+      "learning_rate": 4.096871892866672e-06,
+      "loss": 1.0051,
+      "step": 6987
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.209919906489276,
+      "learning_rate": 4.095569919016624e-06,
+      "loss": 0.5489,
+      "step": 6988
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.427039188500597,
+      "learning_rate": 4.0942680085631896e-06,
+      "loss": 0.844,
+      "step": 6989
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.688646828015336,
+      "learning_rate": 4.092966161597628e-06,
+      "loss": 1.3874,
+      "step": 6990
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.9209706639075392,
+      "learning_rate": 4.0916643782111885e-06,
+      "loss": 0.7576,
+      "step": 6991
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.602547173452972,
+      "learning_rate": 4.0903626584951235e-06,
+      "loss": 0.7423,
+      "step": 6992
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.218153447306228,
+      "learning_rate": 4.089061002540678e-06,
+      "loss": 0.1575,
+      "step": 6993
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.162180261822791,
+      "learning_rate": 4.087759410439091e-06,
+      "loss": 0.8636,
+      "step": 6994
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.211893565610061,
+      "learning_rate": 4.0864578822815996e-06,
+      "loss": 0.5571,
+      "step": 6995
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.93994923532354,
+      "learning_rate": 4.085156418159436e-06,
+      "loss": 0.255,
+      "step": 6996
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.078531851812658,
+      "learning_rate": 4.083855018163825e-06,
+      "loss": 0.7768,
+      "step": 6997
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.6106329231752055,
+      "learning_rate": 4.0825536823859895e-06,
+      "loss": 0.8278,
+      "step": 6998
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.9039369701823583,
+      "learning_rate": 4.081252410917148e-06,
+      "loss": 0.705,
+      "step": 6999
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.937544893436525,
+      "learning_rate": 4.079951203848515e-06,
+      "loss": 0.8428,
+      "step": 7000
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.265860819313912,
+      "learning_rate": 4.078650061271298e-06,
+      "loss": 1.3106,
+      "step": 7001
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.076213834629379,
+      "learning_rate": 4.077348983276705e-06,
+      "loss": 0.1628,
+      "step": 7002
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.9747004961131625,
+      "learning_rate": 4.0760479699559295e-06,
+      "loss": 1.2271,
+      "step": 7003
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.229486804362292,
+      "learning_rate": 4.074747021400171e-06,
+      "loss": 0.5791,
+      "step": 7004
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.696071073445764,
+      "learning_rate": 4.073446137700619e-06,
+      "loss": 0.9656,
+      "step": 7005
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.023258765103961,
+      "learning_rate": 4.072145318948461e-06,
+      "loss": 0.9494,
+      "step": 7006
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.3544085903462877,
+      "learning_rate": 4.0708445652348795e-06,
+      "loss": 0.3698,
+      "step": 7007
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.5439805146070493,
+      "learning_rate": 4.069543876651048e-06,
+      "loss": 0.4269,
+      "step": 7008
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.0477291576943464,
+      "learning_rate": 4.068243253288143e-06,
+      "loss": 0.491,
+      "step": 7009
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.424843315953494,
+      "learning_rate": 4.0669426952373305e-06,
+      "loss": 0.8975,
+      "step": 7010
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.902101146987203,
+      "learning_rate": 4.065642202589774e-06,
+      "loss": 1.0385,
+      "step": 7011
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.566245633850951,
+      "learning_rate": 4.064341775436632e-06,
+      "loss": 0.6051,
+      "step": 7012
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.341015013701762,
+      "learning_rate": 4.063041413869062e-06,
+      "loss": 0.9417,
+      "step": 7013
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.06159040114982,
+      "learning_rate": 4.061741117978209e-06,
+      "loss": 0.7986,
+      "step": 7014
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.947224703436263,
+      "learning_rate": 4.06044088785522e-06,
+      "loss": 1.0728,
+      "step": 7015
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.263300477571825,
+      "learning_rate": 4.059140723591238e-06,
+      "loss": 0.6382,
+      "step": 7016
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.3658893841201705,
+      "learning_rate": 4.057840625277395e-06,
+      "loss": 0.8812,
+      "step": 7017
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.202865379750988,
+      "learning_rate": 4.056540593004823e-06,
+      "loss": 0.484,
+      "step": 7018
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.352692183555134,
+      "learning_rate": 4.0552406268646524e-06,
+      "loss": 0.6808,
+      "step": 7019
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.1367246178274346,
+      "learning_rate": 4.053940726948001e-06,
+      "loss": 0.6093,
+      "step": 7020
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.8307141323573917,
+      "learning_rate": 4.052640893345986e-06,
+      "loss": 0.3585,
+      "step": 7021
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.9444468720987205,
+      "learning_rate": 4.051341126149722e-06,
+      "loss": 0.4466,
+      "step": 7022
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.089568245849912,
+      "learning_rate": 4.0500414254503174e-06,
+      "loss": 0.1585,
+      "step": 7023
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.5099561653851397,
+      "learning_rate": 4.048741791338874e-06,
+      "loss": 0.7526,
+      "step": 7024
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.843930801874921,
+      "learning_rate": 4.047442223906493e-06,
+      "loss": 0.7592,
+      "step": 7025
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.210986763199321,
+      "learning_rate": 4.046142723244264e-06,
+      "loss": 0.5776,
+      "step": 7026
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.040105628903596,
+      "learning_rate": 4.044843289443279e-06,
+      "loss": 1.2542,
+      "step": 7027
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.132099269451299,
+      "learning_rate": 4.043543922594623e-06,
+      "loss": 0.8952,
+      "step": 7028
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 6.111833355656927,
+      "learning_rate": 4.042244622789376e-06,
+      "loss": 1.1106,
+      "step": 7029
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.988197594451673,
+      "learning_rate": 4.040945390118614e-06,
+      "loss": 0.6804,
+      "step": 7030
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.509084378635519,
+      "learning_rate": 4.039646224673404e-06,
+      "loss": 1.1252,
+      "step": 7031
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.7563337982585474,
+      "learning_rate": 4.038347126544816e-06,
+      "loss": 0.9299,
+      "step": 7032
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.396042004521007,
+      "learning_rate": 4.037048095823907e-06,
+      "loss": 0.8917,
+      "step": 7033
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.6091353439783886,
+      "learning_rate": 4.035749132601738e-06,
+      "loss": 0.8339,
+      "step": 7034
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.324665985027785,
+      "learning_rate": 4.034450236969357e-06,
+      "loss": 0.8793,
+      "step": 7035
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.889280898988449,
+      "learning_rate": 4.033151409017814e-06,
+      "loss": 0.6727,
+      "step": 7036
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.705194395612646,
+      "learning_rate": 4.031852648838148e-06,
+      "loss": 1.1622,
+      "step": 7037
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.899128719302082,
+      "learning_rate": 4.030553956521397e-06,
+      "loss": 0.9026,
+      "step": 7038
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.4238661464589204,
+      "learning_rate": 4.029255332158597e-06,
+      "loss": 0.641,
+      "step": 7039
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.179772961300419,
+      "learning_rate": 4.0279567758407715e-06,
+      "loss": 0.738,
+      "step": 7040
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.810268008598644,
+      "learning_rate": 4.026658287658947e-06,
+      "loss": 0.7484,
+      "step": 7041
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.924214179817267,
+      "learning_rate": 4.025359867704141e-06,
+      "loss": 0.8885,
+      "step": 7042
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.578335547143393,
+      "learning_rate": 4.024061516067365e-06,
+      "loss": 1.2875,
+      "step": 7043
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0325952968801584,
+      "learning_rate": 4.02276323283963e-06,
+      "loss": 0.4974,
+      "step": 7044
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 6.395851544449063,
+      "learning_rate": 4.021465018111939e-06,
+      "loss": 1.2998,
+      "step": 7045
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.706246330303951,
+      "learning_rate": 4.020166871975293e-06,
+      "loss": 0.8945,
+      "step": 7046
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.104683337642188,
+      "learning_rate": 4.0188687945206846e-06,
+      "loss": 1.0408,
+      "step": 7047
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8628401219992465,
+      "learning_rate": 4.0175707858391035e-06,
+      "loss": 0.4056,
+      "step": 7048
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.183498230905607,
+      "learning_rate": 4.016272846021534e-06,
+      "loss": 0.9974,
+      "step": 7049
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.068505142474466,
+      "learning_rate": 4.014974975158958e-06,
+      "loss": 0.8056,
+      "step": 7050
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.3480882349045986,
+      "learning_rate": 4.013677173342348e-06,
+      "loss": 0.6373,
+      "step": 7051
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.9158743043304876,
+      "learning_rate": 4.012379440662676e-06,
+      "loss": 0.7279,
+      "step": 7052
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.9453187149906,
+      "learning_rate": 4.011081777210909e-06,
+      "loss": 1.505,
+      "step": 7053
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6622951104295558,
+      "learning_rate": 4.009784183078004e-06,
+      "loss": 0.3531,
+      "step": 7054
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.9462489165130714,
+      "learning_rate": 4.008486658354919e-06,
+      "loss": 0.4783,
+      "step": 7055
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.435618489485107,
+      "learning_rate": 4.007189203132603e-06,
+      "loss": 0.9431,
+      "step": 7056
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.505739082053664,
+      "learning_rate": 4.005891817502004e-06,
+      "loss": 1.2068,
+      "step": 7057
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.567645285380558,
+      "learning_rate": 4.004594501554061e-06,
+      "loss": 0.8364,
+      "step": 7058
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.8855141545565215,
+      "learning_rate": 4.003297255379715e-06,
+      "loss": 0.6913,
+      "step": 7059
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.4254509751515205,
+      "learning_rate": 4.0020000790698895e-06,
+      "loss": 0.5555,
+      "step": 7060
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.424551143787788,
+      "learning_rate": 4.0007029727155165e-06,
+      "loss": 0.5029,
+      "step": 7061
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8812165633556992,
+      "learning_rate": 3.999405936407517e-06,
+      "loss": 0.3657,
+      "step": 7062
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.6522064217633807,
+      "learning_rate": 3.998108970236807e-06,
+      "loss": 0.6894,
+      "step": 7063
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.6574313938311116,
+      "learning_rate": 3.9968120742942965e-06,
+      "loss": 1.0746,
+      "step": 7064
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.8574976031253674,
+      "learning_rate": 3.995515248670896e-06,
+      "loss": 0.8057,
+      "step": 7065
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.902393834271197,
+      "learning_rate": 3.994218493457503e-06,
+      "loss": 0.8585,
+      "step": 7066
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7013551577078005,
+      "learning_rate": 3.992921808745016e-06,
+      "loss": 0.3055,
+      "step": 7067
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.1942799397653476,
+      "learning_rate": 3.991625194624328e-06,
+      "loss": 0.7487,
+      "step": 7068
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.478872092128973,
+      "learning_rate": 3.990328651186326e-06,
+      "loss": 0.2751,
+      "step": 7069
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.540871815467057,
+      "learning_rate": 3.989032178521892e-06,
+      "loss": 0.8078,
+      "step": 7070
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.353383629970022,
+      "learning_rate": 3.987735776721902e-06,
+      "loss": 1.0104,
+      "step": 7071
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.733842174247616,
+      "learning_rate": 3.9864394458772275e-06,
+      "loss": 0.9171,
+      "step": 7072
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.91998825679797,
+      "learning_rate": 3.9851431860787376e-06,
+      "loss": 0.9968,
+      "step": 7073
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.401450605560654,
+      "learning_rate": 3.983846997417293e-06,
+      "loss": 0.5211,
+      "step": 7074
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.6189428504676515,
+      "learning_rate": 3.982550879983752e-06,
+      "loss": 0.9904,
+      "step": 7075
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.908054204358741,
+      "learning_rate": 3.981254833868968e-06,
+      "loss": 0.5373,
+      "step": 7076
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.439929750668236,
+      "learning_rate": 3.979958859163785e-06,
+      "loss": 1.0696,
+      "step": 7077
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.936184099302287,
+      "learning_rate": 3.978662955959047e-06,
+      "loss": 0.5715,
+      "step": 7078
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.8749727121931197,
+      "learning_rate": 3.977367124345591e-06,
+      "loss": 0.9434,
+      "step": 7079
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.5969238585770107,
+      "learning_rate": 3.976071364414248e-06,
+      "loss": 0.661,
+      "step": 7080
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.761432159512601,
+      "learning_rate": 3.974775676255847e-06,
+      "loss": 0.8004,
+      "step": 7081
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.105157516199045,
+      "learning_rate": 3.973480059961211e-06,
+      "loss": 1.0297,
+      "step": 7082
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.2195043443638456,
+      "learning_rate": 3.9721845156211535e-06,
+      "loss": 0.6143,
+      "step": 7083
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.3895344869403745,
+      "learning_rate": 3.970889043326488e-06,
+      "loss": 0.9823,
+      "step": 7084
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.4684250527361247,
+      "learning_rate": 3.969593643168022e-06,
+      "loss": 0.4505,
+      "step": 7085
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.388774080285407,
+      "learning_rate": 3.968298315236558e-06,
+      "loss": 0.7791,
+      "step": 7086
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.0887367584929875,
+      "learning_rate": 3.967003059622893e-06,
+      "loss": 0.6912,
+      "step": 7087
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.66271295613892,
+      "learning_rate": 3.965707876417818e-06,
+      "loss": 1.1753,
+      "step": 7088
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.455896539629951,
+      "learning_rate": 3.964412765712118e-06,
+      "loss": 0.811,
+      "step": 7089
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.664757526270815,
+      "learning_rate": 3.963117727596576e-06,
+      "loss": 0.674,
+      "step": 7090
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.662817213759074,
+      "learning_rate": 3.961822762161969e-06,
+      "loss": 0.6596,
+      "step": 7091
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.843816071724626,
+      "learning_rate": 3.960527869499068e-06,
+      "loss": 0.868,
+      "step": 7092
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.971949880221828,
+      "learning_rate": 3.959233049698642e-06,
+      "loss": 0.5862,
+      "step": 7093
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.403184894041416,
+      "learning_rate": 3.957938302851447e-06,
+      "loss": 0.8985,
+      "step": 7094
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.151657750427979,
+      "learning_rate": 3.956643629048244e-06,
+      "loss": 0.3323,
+      "step": 7095
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.326153968617495,
+      "learning_rate": 3.95534902837978e-06,
+      "loss": 0.893,
+      "step": 7096
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.0435757373843595,
+      "learning_rate": 3.954054500936803e-06,
+      "loss": 0.8832,
+      "step": 7097
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.846159514245381,
+      "learning_rate": 3.952760046810054e-06,
+      "loss": 0.4438,
+      "step": 7098
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.182475428073442,
+      "learning_rate": 3.951465666090269e-06,
+      "loss": 0.975,
+      "step": 7099
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.230426826177786,
+      "learning_rate": 3.950171358868177e-06,
+      "loss": 1.4132,
+      "step": 7100
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.8291254257521725,
+      "learning_rate": 3.948877125234502e-06,
+      "loss": 0.6377,
+      "step": 7101
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.216084276924933,
+      "learning_rate": 3.947582965279969e-06,
+      "loss": 0.6111,
+      "step": 7102
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6706962900628781,
+      "learning_rate": 3.9462888790952885e-06,
+      "loss": 0.2765,
+      "step": 7103
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.138040175328101,
+      "learning_rate": 3.944994866771171e-06,
+      "loss": 0.339,
+      "step": 7104
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.604511535772637,
+      "learning_rate": 3.943700928398325e-06,
+      "loss": 0.7229,
+      "step": 7105
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.5932734228415804,
+      "learning_rate": 3.942407064067444e-06,
+      "loss": 0.5461,
+      "step": 7106
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.790858424347929,
+      "learning_rate": 3.941113273869226e-06,
+      "loss": 0.7984,
+      "step": 7107
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.991797473632268,
+      "learning_rate": 3.939819557894358e-06,
+      "loss": 0.7425,
+      "step": 7108
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.2829158230152045,
+      "learning_rate": 3.938525916233527e-06,
+      "loss": 1.1413,
+      "step": 7109
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.27835612607791,
+      "learning_rate": 3.93723234897741e-06,
+      "loss": 0.3811,
+      "step": 7110
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.0617761846025,
+      "learning_rate": 3.93593885621668e-06,
+      "loss": 0.6642,
+      "step": 7111
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.211114673808272,
+      "learning_rate": 3.934645438042004e-06,
+      "loss": 1.2995,
+      "step": 7112
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.1915693087058945,
+      "learning_rate": 3.933352094544045e-06,
+      "loss": 0.7407,
+      "step": 7113
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.701250053167851,
+      "learning_rate": 3.932058825813464e-06,
+      "loss": 1.1612,
+      "step": 7114
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.970892171852544,
+      "learning_rate": 3.930765631940911e-06,
+      "loss": 0.8714,
+      "step": 7115
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.527673185339682,
+      "learning_rate": 3.929472513017036e-06,
+      "loss": 0.653,
+      "step": 7116
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.623928334585889,
+      "learning_rate": 3.928179469132477e-06,
+      "loss": 0.8409,
+      "step": 7117
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.8302572266053065,
+      "learning_rate": 3.926886500377874e-06,
+      "loss": 1.0464,
+      "step": 7118
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.6459442820124863,
+      "learning_rate": 3.925593606843856e-06,
+      "loss": 0.8235,
+      "step": 7119
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.4826241592611815,
+      "learning_rate": 3.924300788621049e-06,
+      "loss": 0.6283,
+      "step": 7120
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.9437138488386823,
+      "learning_rate": 3.923008045800077e-06,
+      "loss": 0.461,
+      "step": 7121
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.932708836589702,
+      "learning_rate": 3.921715378471555e-06,
+      "loss": 0.9496,
+      "step": 7122
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.194910499503078,
+      "learning_rate": 3.92042278672609e-06,
+      "loss": 0.6686,
+      "step": 7123
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.1726502179347773,
+      "learning_rate": 3.91913027065429e-06,
+      "loss": 0.5938,
+      "step": 7124
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8286749940870537,
+      "learning_rate": 3.917837830346754e-06,
+      "loss": 0.4451,
+      "step": 7125
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.945794551030212,
+      "learning_rate": 3.916545465894077e-06,
+      "loss": 1.0091,
+      "step": 7126
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.53195597074014,
+      "learning_rate": 3.915253177386849e-06,
+      "loss": 0.6858,
+      "step": 7127
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.440713321478638,
+      "learning_rate": 3.91396096491565e-06,
+      "loss": 0.4462,
+      "step": 7128
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.8823848511405292,
+      "learning_rate": 3.912668828571061e-06,
+      "loss": 0.9204,
+      "step": 7129
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.9367897793515216,
+      "learning_rate": 3.9113767684436555e-06,
+      "loss": 0.73,
+      "step": 7130
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.190586554285153,
+      "learning_rate": 3.910084784624001e-06,
+      "loss": 0.9722,
+      "step": 7131
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.9779115794191515,
+      "learning_rate": 3.90879287720266e-06,
+      "loss": 1.2202,
+      "step": 7132
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.5050948355733933,
+      "learning_rate": 3.907501046270189e-06,
+      "loss": 0.4069,
+      "step": 7133
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.415859284537811,
+      "learning_rate": 3.906209291917141e-06,
+      "loss": 0.8027,
+      "step": 7134
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.9358216022893835,
+      "learning_rate": 3.904917614234061e-06,
+      "loss": 1.2471,
+      "step": 7135
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.9747677120767912,
+      "learning_rate": 3.903626013311489e-06,
+      "loss": 0.7411,
+      "step": 7136
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.988415913963125,
+      "learning_rate": 3.902334489239963e-06,
+      "loss": 0.7372,
+      "step": 7137
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.738282110028436,
+      "learning_rate": 3.901043042110012e-06,
+      "loss": 0.8475,
+      "step": 7138
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.100754157933643,
+      "learning_rate": 3.899751672012163e-06,
+      "loss": 0.8092,
+      "step": 7139
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.5068835161335983,
+      "learning_rate": 3.898460379036931e-06,
+      "loss": 0.7991,
+      "step": 7140
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.933771498849496,
+      "learning_rate": 3.897169163274835e-06,
+      "loss": 0.8676,
+      "step": 7141
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.602783614402582,
+      "learning_rate": 3.895878024816378e-06,
+      "loss": 0.7007,
+      "step": 7142
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.797454968338988,
+      "learning_rate": 3.894586963752068e-06,
+      "loss": 0.5681,
+      "step": 7143
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.1546536385429227,
+      "learning_rate": 3.893295980172401e-06,
+      "loss": 0.7844,
+      "step": 7144
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.454972069711611,
+      "learning_rate": 3.892005074167871e-06,
+      "loss": 0.5962,
+      "step": 7145
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.4058536007061235,
+      "learning_rate": 3.890714245828961e-06,
+      "loss": 0.7945,
+      "step": 7146
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.3281395348638245,
+      "learning_rate": 3.889423495246155e-06,
+      "loss": 0.8408,
+      "step": 7147
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.1678889532437804,
+      "learning_rate": 3.88813282250993e-06,
+      "loss": 0.2202,
+      "step": 7148
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.6827832514151853,
+      "learning_rate": 3.8868422277107536e-06,
+      "loss": 0.7201,
+      "step": 7149
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.427866915998655,
+      "learning_rate": 3.885551710939095e-06,
+      "loss": 0.5189,
+      "step": 7150
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.738846069484021,
+      "learning_rate": 3.884261272285409e-06,
+      "loss": 0.8858,
+      "step": 7151
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.843459236720858,
+      "learning_rate": 3.8829709118401525e-06,
+      "loss": 0.4664,
+      "step": 7152
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.153689597283749,
+      "learning_rate": 3.881680629693774e-06,
+      "loss": 0.4918,
+      "step": 7153
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.088250182020076,
+      "learning_rate": 3.8803904259367156e-06,
+      "loss": 0.6217,
+      "step": 7154
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.021547163364894,
+      "learning_rate": 3.879100300659417e-06,
+      "loss": 1.1175,
+      "step": 7155
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.582998666770339,
+      "learning_rate": 3.87781025395231e-06,
+      "loss": 1.0289,
+      "step": 7156
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.112295157664457,
+      "learning_rate": 3.87652028590582e-06,
+      "loss": 0.6874,
+      "step": 7157
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.8179712349560764,
+      "learning_rate": 3.875230396610367e-06,
+      "loss": 0.3803,
+      "step": 7158
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9175547172248338,
+      "learning_rate": 3.873940586156368e-06,
+      "loss": 0.1297,
+      "step": 7159
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.2277994379466444,
+      "learning_rate": 3.8726508546342346e-06,
+      "loss": 0.2989,
+      "step": 7160
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.338092565175346,
+      "learning_rate": 3.8713612021343695e-06,
+      "loss": 0.7526,
+      "step": 7161
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.7247563648076056,
+      "learning_rate": 3.870071628747174e-06,
+      "loss": 0.7655,
+      "step": 7162
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.352743472567859,
+      "learning_rate": 3.868782134563038e-06,
+      "loss": 1.3943,
+      "step": 7163
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.3405860367231086,
+      "learning_rate": 3.867492719672352e-06,
+      "loss": 0.4234,
+      "step": 7164
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.123619589734145,
+      "learning_rate": 3.866203384165497e-06,
+      "loss": 0.9548,
+      "step": 7165
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3149482566165072,
+      "learning_rate": 3.86491412813285e-06,
+      "loss": 0.1749,
+      "step": 7166
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0442256070497082,
+      "learning_rate": 3.863624951664785e-06,
+      "loss": 0.3793,
+      "step": 7167
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.1812285417600368,
+      "learning_rate": 3.862335854851664e-06,
+      "loss": 0.7994,
+      "step": 7168
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.998063794562161,
+      "learning_rate": 3.861046837783847e-06,
+      "loss": 0.5162,
+      "step": 7169
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6757687591265147,
+      "learning_rate": 3.859757900551691e-06,
+      "loss": 0.2664,
+      "step": 7170
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.706957271055265,
+      "learning_rate": 3.8584690432455456e-06,
+      "loss": 1.1055,
+      "step": 7171
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.327806797529627,
+      "learning_rate": 3.85718026595575e-06,
+      "loss": 0.6231,
+      "step": 7172
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.6424505231979754,
+      "learning_rate": 3.855891568772646e-06,
+      "loss": 0.8499,
+      "step": 7173
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8250076554024681,
+      "learning_rate": 3.854602951786562e-06,
+      "loss": 0.1412,
+      "step": 7174
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.9142664387470116,
+      "learning_rate": 3.8533144150878275e-06,
+      "loss": 0.6215,
+      "step": 7175
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.347449923340581,
+      "learning_rate": 3.8520259587667605e-06,
+      "loss": 1.1855,
+      "step": 7176
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.15536023408874,
+      "learning_rate": 3.850737582913679e-06,
+      "loss": 0.7286,
+      "step": 7177
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.8471299159732375,
+      "learning_rate": 3.849449287618892e-06,
+      "loss": 0.9961,
+      "step": 7178
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.839480131416237,
+      "learning_rate": 3.848161072972702e-06,
+      "loss": 1.4947,
+      "step": 7179
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.7682019393758934,
+      "learning_rate": 3.846872939065409e-06,
+      "loss": 0.669,
+      "step": 7180
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.5419341999668967,
+      "learning_rate": 3.8455848859873035e-06,
+      "loss": 0.7145,
+      "step": 7181
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.1367184258575165,
+      "learning_rate": 3.8442969138286726e-06,
+      "loss": 0.6236,
+      "step": 7182
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.862127518357182,
+      "learning_rate": 3.843009022679799e-06,
+      "loss": 1.1014,
+      "step": 7183
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.6911367705490776,
+      "learning_rate": 3.841721212630958e-06,
+      "loss": 0.5209,
+      "step": 7184
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.800166453494574,
+      "learning_rate": 3.8404334837724205e-06,
+      "loss": 0.5094,
+      "step": 7185
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.127442610276738,
+      "learning_rate": 3.8391458361944475e-06,
+      "loss": 0.2214,
+      "step": 7186
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.3877133824762238,
+      "learning_rate": 3.837858269987299e-06,
+      "loss": 0.8476,
+      "step": 7187
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.600332999607474,
+      "learning_rate": 3.836570785241231e-06,
+      "loss": 0.7541,
+      "step": 7188
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.847101954500791,
+      "learning_rate": 3.835283382046484e-06,
+      "loss": 0.6709,
+      "step": 7189
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.822372588945832,
+      "learning_rate": 3.833996060493307e-06,
+      "loss": 1.0825,
+      "step": 7190
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.36848710279312,
+      "learning_rate": 3.832708820671928e-06,
+      "loss": 1.3174,
+      "step": 7191
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.109358838896932,
+      "learning_rate": 3.831421662672582e-06,
+      "loss": 1.0806,
+      "step": 7192
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.408353703345709,
+      "learning_rate": 3.830134586585491e-06,
+      "loss": 0.8491,
+      "step": 7193
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.424293970670651,
+      "learning_rate": 3.828847592500875e-06,
+      "loss": 1.0811,
+      "step": 7194
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.882556551822584,
+      "learning_rate": 3.827560680508946e-06,
+      "loss": 0.8556,
+      "step": 7195
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.2613405765315,
+      "learning_rate": 3.826273850699912e-06,
+      "loss": 0.8493,
+      "step": 7196
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.55056975151563,
+      "learning_rate": 3.824987103163972e-06,
+      "loss": 0.5399,
+      "step": 7197
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.130506068996539,
+      "learning_rate": 3.823700437991321e-06,
+      "loss": 1.0217,
+      "step": 7198
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.646202870650748,
+      "learning_rate": 3.822413855272151e-06,
+      "loss": 0.7791,
+      "step": 7199
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.460344371701214,
+      "learning_rate": 3.821127355096645e-06,
+      "loss": 0.6711,
+      "step": 7200
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.8134014500209763,
+      "learning_rate": 3.81984093755498e-06,
+      "loss": 0.518,
+      "step": 7201
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.253356241796577,
+      "learning_rate": 3.8185546027373325e-06,
+      "loss": 0.807,
+      "step": 7202
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.7265383373452403,
+      "learning_rate": 3.817268350733862e-06,
+      "loss": 0.677,
+      "step": 7203
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.600616746223601,
+      "learning_rate": 3.815982181634735e-06,
+      "loss": 0.6427,
+      "step": 7204
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.1665978297552964,
+      "learning_rate": 3.814696095530103e-06,
+      "loss": 0.3392,
+      "step": 7205
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.16134808595628,
+      "learning_rate": 3.813410092510116e-06,
+      "loss": 1.0075,
+      "step": 7206
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1975198351294647,
+      "learning_rate": 3.8121241726649195e-06,
+      "loss": 0.2097,
+      "step": 7207
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.6734953521052582,
+      "learning_rate": 3.8108383360846467e-06,
+      "loss": 0.6527,
+      "step": 7208
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.394152793853723,
+      "learning_rate": 3.809552582859432e-06,
+      "loss": 0.9742,
+      "step": 7209
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.63757007236222,
+      "learning_rate": 3.8082669130793998e-06,
+      "loss": 0.4478,
+      "step": 7210
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.644899905912618,
+      "learning_rate": 3.8069813268346717e-06,
+      "loss": 0.3406,
+      "step": 7211
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.91484487437852,
+      "learning_rate": 3.8056958242153598e-06,
+      "loss": 0.7422,
+      "step": 7212
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.109537964323625,
+      "learning_rate": 3.804410405311575e-06,
+      "loss": 0.6795,
+      "step": 7213
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.625631627386881,
+      "learning_rate": 3.8031250702134148e-06,
+      "loss": 0.9754,
+      "step": 7214
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.138406520861325,
+      "learning_rate": 3.801839819010979e-06,
+      "loss": 0.575,
+      "step": 7215
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.9259807505491606,
+      "learning_rate": 3.800554651794357e-06,
+      "loss": 0.9691,
+      "step": 7216
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.172195251924218,
+      "learning_rate": 3.7992695686536345e-06,
+      "loss": 1.3919,
+      "step": 7217
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.575012156658763,
+      "learning_rate": 3.7979845696788903e-06,
+      "loss": 0.3482,
+      "step": 7218
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.898379099253242,
+      "learning_rate": 3.7966996549601968e-06,
+      "loss": 0.6983,
+      "step": 7219
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.230484488211065,
+      "learning_rate": 3.795414824587621e-06,
+      "loss": 0.7996,
+      "step": 7220
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.50312455766705,
+      "learning_rate": 3.794130078651222e-06,
+      "loss": 0.8794,
+      "step": 7221
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9917695078674407,
+      "learning_rate": 3.7928454172410565e-06,
+      "loss": 0.397,
+      "step": 7222
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.7089917338476215,
+      "learning_rate": 3.7915608404471738e-06,
+      "loss": 0.5728,
+      "step": 7223
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.7570602077823247,
+      "learning_rate": 3.7902763483596173e-06,
+      "loss": 0.6975,
+      "step": 7224
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.485543040083705,
+      "learning_rate": 3.7889919410684262e-06,
+      "loss": 0.6952,
+      "step": 7225
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.7751991440072854,
+      "learning_rate": 3.7877076186636275e-06,
+      "loss": 0.5311,
+      "step": 7226
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.946904427500743,
+      "learning_rate": 3.7864233812352497e-06,
+      "loss": 0.5584,
+      "step": 7227
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.3960138117057532,
+      "learning_rate": 3.78513922887331e-06,
+      "loss": 0.7777,
+      "step": 7228
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.8388110904642248,
+      "learning_rate": 3.783855161667824e-06,
+      "loss": 0.9353,
+      "step": 7229
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.208385056053049,
+      "learning_rate": 3.7825711797088e-06,
+      "loss": 0.7865,
+      "step": 7230
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.654728205483348,
+      "learning_rate": 3.7812872830862363e-06,
+      "loss": 1.0619,
+      "step": 7231
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.943147912524849,
+      "learning_rate": 3.78000347189013e-06,
+      "loss": 0.9678,
+      "step": 7232
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6386007331503682,
+      "learning_rate": 3.778719746210471e-06,
+      "loss": 0.2088,
+      "step": 7233
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.974326653202862,
+      "learning_rate": 3.777436106137244e-06,
+      "loss": 0.8962,
+      "step": 7234
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.247202635701191,
+      "learning_rate": 3.7761525517604237e-06,
+      "loss": 0.918,
+      "step": 7235
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.920840009975349,
+      "learning_rate": 3.7748690831699858e-06,
+      "loss": 0.6603,
+      "step": 7236
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.024972776872732,
+      "learning_rate": 3.7735857004558913e-06,
+      "loss": 0.3891,
+      "step": 7237
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5436469189015576,
+      "learning_rate": 3.772302403708102e-06,
+      "loss": 0.2604,
+      "step": 7238
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.024412906807474,
+      "learning_rate": 3.7710191930165705e-06,
+      "loss": 0.8262,
+      "step": 7239
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 6.155739632668728,
+      "learning_rate": 3.769736068471246e-06,
+      "loss": 1.5707,
+      "step": 7240
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.7645413009827906,
+      "learning_rate": 3.7684530301620693e-06,
+      "loss": 0.3664,
+      "step": 7241
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.8908487487689944,
+      "learning_rate": 3.7671700781789753e-06,
+      "loss": 0.7639,
+      "step": 7242
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.865174935670953,
+      "learning_rate": 3.7658872126118945e-06,
+      "loss": 0.6395,
+      "step": 7243
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.682703256402187,
+      "learning_rate": 3.7646044335507474e-06,
+      "loss": 1.2663,
+      "step": 7244
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.220821504408669,
+      "learning_rate": 3.7633217410854534e-06,
+      "loss": 0.4474,
+      "step": 7245
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.811613429874075,
+      "learning_rate": 3.7620391353059232e-06,
+      "loss": 0.9988,
+      "step": 7246
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.060604468939529,
+      "learning_rate": 3.760756616302064e-06,
+      "loss": 0.7735,
+      "step": 7247
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.596254542379706,
+      "learning_rate": 3.759474184163771e-06,
+      "loss": 0.7325,
+      "step": 7248
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.334206059149519,
+      "learning_rate": 3.7581918389809384e-06,
+      "loss": 0.7668,
+      "step": 7249
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.1112879445181103,
+      "learning_rate": 3.756909580843455e-06,
+      "loss": 0.6829,
+      "step": 7250
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.016626681173837,
+      "learning_rate": 3.7556274098411993e-06,
+      "loss": 0.9484,
+      "step": 7251
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.9305174867410395,
+      "learning_rate": 3.754345326064046e-06,
+      "loss": 0.8028,
+      "step": 7252
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.047174790150774,
+      "learning_rate": 3.7530633296018664e-06,
+      "loss": 0.8824,
+      "step": 7253
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.7846759776594423,
+      "learning_rate": 3.7517814205445187e-06,
+      "loss": 0.5059,
+      "step": 7254
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.173612013099389,
+      "learning_rate": 3.7504995989818615e-06,
+      "loss": 1.1138,
+      "step": 7255
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5987499473447744,
+      "learning_rate": 3.749217865003744e-06,
+      "loss": 0.179,
+      "step": 7256
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.470430144097223,
+      "learning_rate": 3.747936218700012e-06,
+      "loss": 0.9139,
+      "step": 7257
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.7070475368116615,
+      "learning_rate": 3.7466546601605012e-06,
+      "loss": 0.6481,
+      "step": 7258
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.589366742179254,
+      "learning_rate": 3.745373189475046e-06,
+      "loss": 1.0235,
+      "step": 7259
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.8524322868345173,
+      "learning_rate": 3.744091806733468e-06,
+      "loss": 0.8042,
+      "step": 7260
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.538043867165417,
+      "learning_rate": 3.742810512025589e-06,
+      "loss": 1.122,
+      "step": 7261
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.42242285248518,
+      "learning_rate": 3.7415293054412216e-06,
+      "loss": 1.1899,
+      "step": 7262
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.037165970787136,
+      "learning_rate": 3.7402481870701722e-06,
+      "loss": 1.1023,
+      "step": 7263
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.7586712690069137,
+      "learning_rate": 3.7389671570022445e-06,
+      "loss": 0.6374,
+      "step": 7264
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.649014060677766,
+      "learning_rate": 3.7376862153272307e-06,
+      "loss": 0.6938,
+      "step": 7265
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.3655819352546277,
+      "learning_rate": 3.7364053621349193e-06,
+      "loss": 0.8945,
+      "step": 7266
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0661749900771826,
+      "learning_rate": 3.7351245975150924e-06,
+      "loss": 0.3209,
+      "step": 7267
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.109721130348778,
+      "learning_rate": 3.733843921557526e-06,
+      "loss": 0.8675,
+      "step": 7268
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.4110127821534957,
+      "learning_rate": 3.7325633343519907e-06,
+      "loss": 0.4387,
+      "step": 7269
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3270126397093516,
+      "learning_rate": 3.731282835988252e-06,
+      "loss": 0.1857,
+      "step": 7270
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.1982952028614506,
+      "learning_rate": 3.7300024265560623e-06,
+      "loss": 0.6263,
+      "step": 7271
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.054983253419086,
+      "learning_rate": 3.7287221061451763e-06,
+      "loss": 0.6723,
+      "step": 7272
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.006387213108985,
+      "learning_rate": 3.7274418748453378e-06,
+      "loss": 0.5593,
+      "step": 7273
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.177843696204133,
+      "learning_rate": 3.7261617327462857e-06,
+      "loss": 0.4144,
+      "step": 7274
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.154411830595848,
+      "learning_rate": 3.7248816799377517e-06,
+      "loss": 0.2859,
+      "step": 7275
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.7085682586083197,
+      "learning_rate": 3.723601716509465e-06,
+      "loss": 0.7336,
+      "step": 7276
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.532344507880409,
+      "learning_rate": 3.7223218425511416e-06,
+      "loss": 0.743,
+      "step": 7277
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.1145090200213295,
+      "learning_rate": 3.721042058152496e-06,
+      "loss": 1.0092,
+      "step": 7278
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.9824707968840727,
+      "learning_rate": 3.719762363403236e-06,
+      "loss": 0.6897,
+      "step": 7279
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.93318616633958,
+      "learning_rate": 3.718482758393064e-06,
+      "loss": 1.0741,
+      "step": 7280
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.873862285417129,
+      "learning_rate": 3.7172032432116724e-06,
+      "loss": 0.6303,
+      "step": 7281
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.028763878606627,
+      "learning_rate": 3.715923817948752e-06,
+      "loss": 0.952,
+      "step": 7282
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.664191244431255,
+      "learning_rate": 3.7146444826939828e-06,
+      "loss": 0.8097,
+      "step": 7283
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.243476704374192,
+      "learning_rate": 3.7133652375370404e-06,
+      "loss": 0.9209,
+      "step": 7284
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.81678470925151,
+      "learning_rate": 3.7120860825675965e-06,
+      "loss": 0.5091,
+      "step": 7285
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.085855589423921,
+      "learning_rate": 3.710807017875312e-06,
+      "loss": 0.775,
+      "step": 7286
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.6442545721216475,
+      "learning_rate": 3.7095280435498476e-06,
+      "loss": 1.1102,
+      "step": 7287
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.050918337247399,
+      "learning_rate": 3.708249159680849e-06,
+      "loss": 0.4903,
+      "step": 7288
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.5453549381261493,
+      "learning_rate": 3.7069703663579626e-06,
+      "loss": 0.5086,
+      "step": 7289
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.55975029920748,
+      "learning_rate": 3.7056916636708275e-06,
+      "loss": 0.4368,
+      "step": 7290
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.9211923351279503,
+      "learning_rate": 3.7044130517090725e-06,
+      "loss": 0.7221,
+      "step": 7291
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.8695486692822603,
+      "learning_rate": 3.7031345305623247e-06,
+      "loss": 0.5645,
+      "step": 7292
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.0322526274600685,
+      "learning_rate": 3.701856100320205e-06,
+      "loss": 0.993,
+      "step": 7293
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.397881077392223,
+      "learning_rate": 3.700577761072319e-06,
+      "loss": 0.9325,
+      "step": 7294
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.109022269602481,
+      "learning_rate": 3.6992995129082787e-06,
+      "loss": 0.9902,
+      "step": 7295
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.0430714662205114,
+      "learning_rate": 3.6980213559176806e-06,
+      "loss": 0.5342,
+      "step": 7296
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.979238248294013,
+      "learning_rate": 3.6967432901901214e-06,
+      "loss": 0.8024,
+      "step": 7297
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.357106852408541,
+      "learning_rate": 3.695465315815184e-06,
+      "loss": 0.6754,
+      "step": 7298
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 6.536203245639044,
+      "learning_rate": 3.6941874328824528e-06,
+      "loss": 1.4261,
+      "step": 7299
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.114188179870397,
+      "learning_rate": 3.692909641481498e-06,
+      "loss": 1.0592,
+      "step": 7300
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.7910836878874328,
+      "learning_rate": 3.691631941701889e-06,
+      "loss": 0.7355,
+      "step": 7301
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.512324231847214,
+      "learning_rate": 3.690354333633186e-06,
+      "loss": 0.7389,
+      "step": 7302
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.08733216422255,
+      "learning_rate": 3.689076817364945e-06,
+      "loss": 0.6672,
+      "step": 7303
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.5066420531243563,
+      "learning_rate": 3.6877993929867146e-06,
+      "loss": 0.5613,
+      "step": 7304
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.030634115273332,
+      "learning_rate": 3.6865220605880363e-06,
+      "loss": 1.0986,
+      "step": 7305
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.683643009920432,
+      "learning_rate": 3.6852448202584457e-06,
+      "loss": 0.9423,
+      "step": 7306
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.59612249188272,
+      "learning_rate": 3.6839676720874695e-06,
+      "loss": 0.4942,
+      "step": 7307
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.3689628530978375,
+      "learning_rate": 3.6826906161646325e-06,
+      "loss": 0.9352,
+      "step": 7308
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.9925971887147798,
+      "learning_rate": 3.681413652579451e-06,
+      "loss": 0.4234,
+      "step": 7309
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.259614527439117,
+      "learning_rate": 3.680136781421435e-06,
+      "loss": 0.3901,
+      "step": 7310
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.3481061105666785,
+      "learning_rate": 3.6788600027800847e-06,
+      "loss": 0.8297,
+      "step": 7311
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.272554803335286,
+      "learning_rate": 3.677583316744899e-06,
+      "loss": 1.3087,
+      "step": 7312
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.2133791927491147,
+      "learning_rate": 3.6763067234053686e-06,
+      "loss": 0.3896,
+      "step": 7313
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.4051697569394843,
+      "learning_rate": 3.6750302228509747e-06,
+      "loss": 0.6691,
+      "step": 7314
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 6.793644442129009,
+      "learning_rate": 3.6737538151711965e-06,
+      "loss": 1.5565,
+      "step": 7315
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.1024584202356036,
+      "learning_rate": 3.6724775004555056e-06,
+      "loss": 0.4786,
+      "step": 7316
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.2672856545005318,
+      "learning_rate": 3.6712012787933627e-06,
+      "loss": 0.8188,
+      "step": 7317
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.534730237228595,
+      "learning_rate": 3.669925150274227e-06,
+      "loss": 0.7064,
+      "step": 7318
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.176559579252146,
+      "learning_rate": 3.66864911498755e-06,
+      "loss": 1.1786,
+      "step": 7319
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.01451385356136,
+      "learning_rate": 3.667373173022777e-06,
+      "loss": 0.6224,
+      "step": 7320
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.702301279569575,
+      "learning_rate": 3.6660973244693443e-06,
+      "loss": 0.4707,
+      "step": 7321
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.6518809111910415,
+      "learning_rate": 3.6648215694166854e-06,
+      "loss": 0.6508,
+      "step": 7322
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.837301252009652,
+      "learning_rate": 3.663545907954222e-06,
+      "loss": 1.0315,
+      "step": 7323
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.467377058353138,
+      "learning_rate": 3.662270340171374e-06,
+      "loss": 0.9296,
+      "step": 7324
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.037556790617754,
+      "learning_rate": 3.660994866157553e-06,
+      "loss": 0.4212,
+      "step": 7325
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.7411227792089052,
+      "learning_rate": 3.659719486002165e-06,
+      "loss": 0.5611,
+      "step": 7326
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.029110782388751,
+      "learning_rate": 3.65844419979461e-06,
+      "loss": 0.7441,
+      "step": 7327
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.598556112905251,
+      "learning_rate": 3.6571690076242762e-06,
+      "loss": 0.5019,
+      "step": 7328
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.103631475632426,
+      "learning_rate": 3.6558939095805524e-06,
+      "loss": 0.669,
+      "step": 7329
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.278943860233236,
+      "learning_rate": 3.654618905752814e-06,
+      "loss": 0.792,
+      "step": 7330
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.379718370226008,
+      "learning_rate": 3.6533439962304363e-06,
+      "loss": 0.6466,
+      "step": 7331
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.2511593392125406,
+      "learning_rate": 3.6520691811027833e-06,
+      "loss": 0.6463,
+      "step": 7332
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.9418092034160632,
+      "learning_rate": 3.650794460459216e-06,
+      "loss": 0.7541,
+      "step": 7333
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.3988686548257045,
+      "learning_rate": 3.6495198343890834e-06,
+      "loss": 0.5703,
+      "step": 7334
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.596962690304631,
+      "learning_rate": 3.6482453029817335e-06,
+      "loss": 0.3518,
+      "step": 7335
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.3513090857832393,
+      "learning_rate": 3.6469708663265058e-06,
+      "loss": 0.3199,
+      "step": 7336
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.101769394719204,
+      "learning_rate": 3.645696524512731e-06,
+      "loss": 0.713,
+      "step": 7337
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.2251311010046115,
+      "learning_rate": 3.6444222776297356e-06,
+      "loss": 0.7979,
+      "step": 7338
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6622082215963914,
+      "learning_rate": 3.6431481257668417e-06,
+      "loss": 0.2782,
+      "step": 7339
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.8688816543762545,
+      "learning_rate": 3.641874069013357e-06,
+      "loss": 0.4553,
+      "step": 7340
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.884857490036489,
+      "learning_rate": 3.640600107458589e-06,
+      "loss": 0.6496,
+      "step": 7341
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.5034422972171426,
+      "learning_rate": 3.639326241191837e-06,
+      "loss": 0.4517,
+      "step": 7342
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.427570284980379,
+      "learning_rate": 3.6380524703023955e-06,
+      "loss": 1.0919,
+      "step": 7343
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.414248128192576,
+      "learning_rate": 3.636778794879548e-06,
+      "loss": 0.6897,
+      "step": 7344
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.009681478728431,
+      "learning_rate": 3.6355052150125756e-06,
+      "loss": 0.8974,
+      "step": 7345
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.60090679232242,
+      "learning_rate": 3.6342317307907476e-06,
+      "loss": 0.7755,
+      "step": 7346
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.080563646163504,
+      "learning_rate": 3.632958342303331e-06,
+      "loss": 1.1131,
+      "step": 7347
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.6892240634984996,
+      "learning_rate": 3.6316850496395863e-06,
+      "loss": 0.7426,
+      "step": 7348
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.709703790317675,
+      "learning_rate": 3.630411852888763e-06,
+      "loss": 0.776,
+      "step": 7349
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.7744602064414847,
+      "learning_rate": 3.6291387521401116e-06,
+      "loss": 0.3538,
+      "step": 7350
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.682406043099396,
+      "learning_rate": 3.6278657474828655e-06,
+      "loss": 1.0515,
+      "step": 7351
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.959454819170579,
+      "learning_rate": 3.62659283900626e-06,
+      "loss": 0.477,
+      "step": 7352
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.224783213547431,
+      "learning_rate": 3.625320026799518e-06,
+      "loss": 0.7098,
+      "step": 7353
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.6252261600463815,
+      "learning_rate": 3.6240473109518595e-06,
+      "loss": 0.3263,
+      "step": 7354
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.377159100802485,
+      "learning_rate": 3.6227746915524964e-06,
+      "loss": 1.1255,
+      "step": 7355
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.271968038367136,
+      "learning_rate": 3.621502168690636e-06,
+      "loss": 0.6547,
+      "step": 7356
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.449237596352372,
+      "learning_rate": 3.6202297424554723e-06,
+      "loss": 0.7299,
+      "step": 7357
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6632260784943322,
+      "learning_rate": 3.618957412936199e-06,
+      "loss": 0.23,
+      "step": 7358
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.7017260069150733,
+      "learning_rate": 3.6176851802220015e-06,
+      "loss": 0.4136,
+      "step": 7359
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.1140772619513952,
+      "learning_rate": 3.6164130444020557e-06,
+      "loss": 0.6248,
+      "step": 7360
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.261273199477408,
+      "learning_rate": 3.6151410055655346e-06,
+      "loss": 0.9102,
+      "step": 7361
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.254035443186992,
+      "learning_rate": 3.613869063801604e-06,
+      "loss": 0.5402,
+      "step": 7362
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.826738840401367,
+      "learning_rate": 3.6125972191994167e-06,
+      "loss": 0.8809,
+      "step": 7363
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.627710868115313,
+      "learning_rate": 3.611325471848127e-06,
+      "loss": 1.0343,
+      "step": 7364
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.942724274888949,
+      "learning_rate": 3.6100538218368788e-06,
+      "loss": 1.0607,
+      "step": 7365
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.6652416919731916,
+      "learning_rate": 3.608782269254809e-06,
+      "loss": 0.7414,
+      "step": 7366
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 6.567687807888815,
+      "learning_rate": 3.6075108141910477e-06,
+      "loss": 1.2659,
+      "step": 7367
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.565997631667151,
+      "learning_rate": 3.606239456734718e-06,
+      "loss": 1.0212,
+      "step": 7368
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.7521879553116104,
+      "learning_rate": 3.604968196974936e-06,
+      "loss": 0.4801,
+      "step": 7369
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.227513652989928,
+      "learning_rate": 3.6036970350008117e-06,
+      "loss": 0.8487,
+      "step": 7370
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.8028013106733445,
+      "learning_rate": 3.6024259709014485e-06,
+      "loss": 0.7161,
+      "step": 7371
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.561632017753492,
+      "learning_rate": 3.601155004765943e-06,
+      "loss": 0.9349,
+      "step": 7372
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.8234288647297845,
+      "learning_rate": 3.599884136683386e-06,
+      "loss": 0.6112,
+      "step": 7373
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.1618931501398477,
+      "learning_rate": 3.5986133667428552e-06,
+      "loss": 0.29,
+      "step": 7374
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.341504592456218,
+      "learning_rate": 3.59734269503343e-06,
+      "loss": 0.9025,
+      "step": 7375
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.567997868746615,
+      "learning_rate": 3.596072121644176e-06,
+      "loss": 1.6976,
+      "step": 7376
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.8884371462644887,
+      "learning_rate": 3.5948016466641565e-06,
+      "loss": 0.9424,
+      "step": 7377
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.3051617998588774,
+      "learning_rate": 3.593531270182426e-06,
+      "loss": 0.6257,
+      "step": 7378
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.1393146329806285,
+      "learning_rate": 3.5922609922880347e-06,
+      "loss": 0.5962,
+      "step": 7379
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.380352244126183,
+      "learning_rate": 3.5909908130700196e-06,
+      "loss": 0.8863,
+      "step": 7380
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.7033105237332506,
+      "learning_rate": 3.5897207326174162e-06,
+      "loss": 0.7133,
+      "step": 7381
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.520638027193085,
+      "learning_rate": 3.5884507510192524e-06,
+      "loss": 0.9772,
+      "step": 7382
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 6.268886224791158,
+      "learning_rate": 3.5871808683645475e-06,
+      "loss": 1.332,
+      "step": 7383
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.086688366684082,
+      "learning_rate": 3.585911084742315e-06,
+      "loss": 0.7312,
+      "step": 7384
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.756162826036348,
+      "learning_rate": 3.584641400241563e-06,
+      "loss": 1.1717,
+      "step": 7385
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.053332697897676,
+      "learning_rate": 3.5833718149512874e-06,
+      "loss": 0.8567,
+      "step": 7386
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.551526636895038,
+      "learning_rate": 3.582102328960483e-06,
+      "loss": 0.7164,
+      "step": 7387
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.299671804905078,
+      "learning_rate": 3.580832942358134e-06,
+      "loss": 0.7368,
+      "step": 7388
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5397836578283757,
+      "learning_rate": 3.5795636552332203e-06,
+      "loss": 0.1627,
+      "step": 7389
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.319864688208959,
+      "learning_rate": 3.5782944676747135e-06,
+      "loss": 0.3598,
+      "step": 7390
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.855430350446534,
+      "learning_rate": 3.5770253797715747e-06,
+      "loss": 0.8059,
+      "step": 7391
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.4382552861938604,
+      "learning_rate": 3.5757563916127665e-06,
+      "loss": 0.848,
+      "step": 7392
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.002761732304879,
+      "learning_rate": 3.574487503287235e-06,
+      "loss": 0.9404,
+      "step": 7393
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.8270837023387534,
+      "learning_rate": 3.5732187148839257e-06,
+      "loss": 0.5003,
+      "step": 7394
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.440678145686728,
+      "learning_rate": 3.571950026491776e-06,
+      "loss": 1.0051,
+      "step": 7395
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.452605562559225,
+      "learning_rate": 3.5706814381997157e-06,
+      "loss": 0.6493,
+      "step": 7396
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.001457635681139,
+      "learning_rate": 3.5694129500966645e-06,
+      "loss": 0.7764,
+      "step": 7397
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.664373470123157,
+      "learning_rate": 3.5681445622715396e-06,
+      "loss": 0.5855,
+      "step": 7398
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.8412498830805792,
+      "learning_rate": 3.56687627481325e-06,
+      "loss": 0.6159,
+      "step": 7399
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.8788036431099524,
+      "learning_rate": 3.5656080878106957e-06,
+      "loss": 0.6772,
+      "step": 7400
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.203657324782098,
+      "learning_rate": 3.5643400013527723e-06,
+      "loss": 1.1349,
+      "step": 7401
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.3686955244781855,
+      "learning_rate": 3.5630720155283686e-06,
+      "loss": 0.5233,
+      "step": 7402
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3251001828269917,
+      "learning_rate": 3.561804130426361e-06,
+      "loss": 0.2024,
+      "step": 7403
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.5517867637136455,
+      "learning_rate": 3.560536346135625e-06,
+      "loss": 0.4885,
+      "step": 7404
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.65419707229043,
+      "learning_rate": 3.559268662745027e-06,
+      "loss": 0.8909,
+      "step": 7405
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.0702596670441102,
+      "learning_rate": 3.5580010803434254e-06,
+      "loss": 0.6835,
+      "step": 7406
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.9999027181662297,
+      "learning_rate": 3.5567335990196725e-06,
+      "loss": 0.4085,
+      "step": 7407
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.6525097086960625,
+      "learning_rate": 3.5554662188626147e-06,
+      "loss": 0.5371,
+      "step": 7408
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.987470859525362,
+      "learning_rate": 3.5541989399610866e-06,
+      "loss": 0.1876,
+      "step": 7409
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.367384021483217,
+      "learning_rate": 3.5529317624039205e-06,
+      "loss": 0.4064,
+      "step": 7410
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.4214064440204153,
+      "learning_rate": 3.5516646862799404e-06,
+      "loss": 0.414,
+      "step": 7411
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4276191576511317,
+      "learning_rate": 3.5503977116779624e-06,
+      "loss": 0.2154,
+      "step": 7412
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.0645046994455445,
+      "learning_rate": 3.5491308386867983e-06,
+      "loss": 0.9574,
+      "step": 7413
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.187906423787895,
+      "learning_rate": 3.5478640673952456e-06,
+      "loss": 0.3924,
+      "step": 7414
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.386770954240138,
+      "learning_rate": 3.5465973978921042e-06,
+      "loss": 0.9238,
+      "step": 7415
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.7501762987264438,
+      "learning_rate": 3.545330830266158e-06,
+      "loss": 0.5472,
+      "step": 7416
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.476586699802735,
+      "learning_rate": 3.54406436460619e-06,
+      "loss": 0.3458,
+      "step": 7417
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.781993328385959,
+      "learning_rate": 3.5427980010009746e-06,
+      "loss": 0.713,
+      "step": 7418
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.432373455583828,
+      "learning_rate": 3.541531739539279e-06,
+      "loss": 0.9409,
+      "step": 7419
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.6775218937185987,
+      "learning_rate": 3.540265580309859e-06,
+      "loss": 0.8305,
+      "step": 7420
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.84216649186372,
+      "learning_rate": 3.538999523401469e-06,
+      "loss": 0.723,
+      "step": 7421
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.6949795452225156,
+      "learning_rate": 3.5377335689028556e-06,
+      "loss": 0.5788,
+      "step": 7422
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.958017014079015,
+      "learning_rate": 3.536467716902754e-06,
+      "loss": 1.0746,
+      "step": 7423
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.173675695552111,
+      "learning_rate": 3.5352019674898956e-06,
+      "loss": 0.1809,
+      "step": 7424
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.5554370486237157,
+      "learning_rate": 3.533936320753007e-06,
+      "loss": 0.786,
+      "step": 7425
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.5327408285560624,
+      "learning_rate": 3.5326707767808e-06,
+      "loss": 0.5137,
+      "step": 7426
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.520860779424227,
+      "learning_rate": 3.5314053356619852e-06,
+      "loss": 1.2974,
+      "step": 7427
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.882673289960211,
+      "learning_rate": 3.5301399974852656e-06,
+      "loss": 0.9882,
+      "step": 7428
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.430372009046639,
+      "learning_rate": 3.528874762339336e-06,
+      "loss": 0.6721,
+      "step": 7429
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.8215136183775615,
+      "learning_rate": 3.5276096303128837e-06,
+      "loss": 0.5899,
+      "step": 7430
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.9066179246724344,
+      "learning_rate": 3.526344601494588e-06,
+      "loss": 0.8211,
+      "step": 7431
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.62528976844446,
+      "learning_rate": 3.525079675973121e-06,
+      "loss": 0.6035,
+      "step": 7432
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.377924604021585,
+      "learning_rate": 3.5238148538371506e-06,
+      "loss": 0.9099,
+      "step": 7433
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.741153913875198,
+      "learning_rate": 3.5225501351753346e-06,
+      "loss": 0.5172,
+      "step": 7434
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.934678384231139,
+      "learning_rate": 3.5212855200763237e-06,
+      "loss": 0.9718,
+      "step": 7435
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.491095188918565,
+      "learning_rate": 3.5200210086287646e-06,
+      "loss": 1.1595,
+      "step": 7436
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.229394049045571,
+      "learning_rate": 3.51875660092129e-06,
+      "loss": 1.2175,
+      "step": 7437
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1439429695257528,
+      "learning_rate": 3.5174922970425317e-06,
+      "loss": 0.1642,
+      "step": 7438
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.755070223779588,
+      "learning_rate": 3.5162280970811115e-06,
+      "loss": 0.3445,
+      "step": 7439
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.578523707958233,
+      "learning_rate": 3.5149640011256438e-06,
+      "loss": 0.5381,
+      "step": 7440
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.715438769454317,
+      "learning_rate": 3.5137000092647366e-06,
+      "loss": 0.8441,
+      "step": 7441
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.880341497534685,
+      "learning_rate": 3.512436121586993e-06,
+      "loss": 0.5466,
+      "step": 7442
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.493807636256478,
+      "learning_rate": 3.5111723381810005e-06,
+      "loss": 0.9816,
+      "step": 7443
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8737191790534322,
+      "learning_rate": 3.509908659135348e-06,
+      "loss": 0.4204,
+      "step": 7444
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.901209827546255,
+      "learning_rate": 3.5086450845386145e-06,
+      "loss": 0.5692,
+      "step": 7445
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.1984061838214535,
+      "learning_rate": 3.5073816144793695e-06,
+      "loss": 0.6392,
+      "step": 7446
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 6.097471089521399,
+      "learning_rate": 3.5061182490461775e-06,
+      "loss": 1.7763,
+      "step": 7447
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.848732075469702,
+      "learning_rate": 3.5048549883275962e-06,
+      "loss": 0.867,
+      "step": 7448
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.370455763998571,
+      "learning_rate": 3.503591832412172e-06,
+      "loss": 1.0785,
+      "step": 7449
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.194037593480006,
+      "learning_rate": 3.5023287813884476e-06,
+      "loss": 1.0598,
+      "step": 7450
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.4061901521566673,
+      "learning_rate": 3.5010658353449576e-06,
+      "loss": 0.9436,
+      "step": 7451
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.807685581479717,
+      "learning_rate": 3.4998029943702305e-06,
+      "loss": 0.7093,
+      "step": 7452
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.000582686325802,
+      "learning_rate": 3.498540258552785e-06,
+      "loss": 0.7585,
+      "step": 7453
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.807383909205339,
+      "learning_rate": 3.497277627981132e-06,
+      "loss": 0.7692,
+      "step": 7454
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.908451982034939,
+      "learning_rate": 3.496015102743777e-06,
+      "loss": 0.7361,
+      "step": 7455
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.320412268232292,
+      "learning_rate": 3.4947526829292177e-06,
+      "loss": 0.9214,
+      "step": 7456
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.731564301514826,
+      "learning_rate": 3.4934903686259445e-06,
+      "loss": 0.8081,
+      "step": 7457
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.433002279457299,
+      "learning_rate": 3.4922281599224404e-06,
+      "loss": 0.515,
+      "step": 7458
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.935834415095173,
+      "learning_rate": 3.4909660569071823e-06,
+      "loss": 1.3236,
+      "step": 7459
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.0732773788320507,
+      "learning_rate": 3.4897040596686345e-06,
+      "loss": 0.7249,
+      "step": 7460
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.9806807216526545,
+      "learning_rate": 3.4884421682952596e-06,
+      "loss": 0.5422,
+      "step": 7461
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.475308889726864,
+      "learning_rate": 3.4871803828755102e-06,
+      "loss": 0.8781,
+      "step": 7462
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.0605076079031805,
+      "learning_rate": 3.4859187034978315e-06,
+      "loss": 0.8476,
+      "step": 7463
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.7484248994060683,
+      "learning_rate": 3.4846571302506624e-06,
+      "loss": 0.5632,
+      "step": 7464
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.288806473145816,
+      "learning_rate": 3.4833956632224364e-06,
+      "loss": 0.6753,
+      "step": 7465
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7944051302550061,
+      "learning_rate": 3.482134302501572e-06,
+      "loss": 0.3302,
+      "step": 7466
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.089755641139934,
+      "learning_rate": 3.480873048176486e-06,
+      "loss": 0.8287,
+      "step": 7467
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.201852703903197,
+      "learning_rate": 3.47961190033559e-06,
+      "loss": 0.5108,
+      "step": 7468
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.298715241648617,
+      "learning_rate": 3.478350859067282e-06,
+      "loss": 1.2063,
+      "step": 7469
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.5977723557337526,
+      "learning_rate": 3.477089924459959e-06,
+      "loss": 0.5909,
+      "step": 7470
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.519710106347817,
+      "learning_rate": 3.475829096602002e-06,
+      "loss": 0.8826,
+      "step": 7471
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.543261817142673,
+      "learning_rate": 3.4745683755817917e-06,
+      "loss": 0.6256,
+      "step": 7472
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.2491847333109622,
+      "learning_rate": 3.4733077614877003e-06,
+      "loss": 0.7993,
+      "step": 7473
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.2529523946365866,
+      "learning_rate": 3.472047254408091e-06,
+      "loss": 0.9358,
+      "step": 7474
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.239751609942161,
+      "learning_rate": 3.4707868544313196e-06,
+      "loss": 0.9912,
+      "step": 7475
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.688500912592552,
+      "learning_rate": 3.469526561645735e-06,
+      "loss": 0.8995,
+      "step": 7476
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.623117656250236,
+      "learning_rate": 3.4682663761396773e-06,
+      "loss": 0.8724,
+      "step": 7477
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.9838805874911793,
+      "learning_rate": 3.4670062980014795e-06,
+      "loss": 0.9372,
+      "step": 7478
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.245121124665357,
+      "learning_rate": 3.465746327319469e-06,
+      "loss": 1.4008,
+      "step": 7479
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.453068472780212,
+      "learning_rate": 3.4644864641819635e-06,
+      "loss": 1.1487,
+      "step": 7480
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5552816233293498,
+      "learning_rate": 3.463226708677275e-06,
+      "loss": 0.3086,
+      "step": 7481
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.190192010096277,
+      "learning_rate": 3.4619670608937074e-06,
+      "loss": 1.0783,
+      "step": 7482
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.978473378064943,
+      "learning_rate": 3.460707520919554e-06,
+      "loss": 1.2116,
+      "step": 7483
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.974293050204034,
+      "learning_rate": 3.4594480888431046e-06,
+      "loss": 0.9243,
+      "step": 7484
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.470372717599689,
+      "learning_rate": 3.4581887647526393e-06,
+      "loss": 0.7718,
+      "step": 7485
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.1744644582526265,
+      "learning_rate": 3.456929548736431e-06,
+      "loss": 0.8462,
+      "step": 7486
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.8233489871206396,
+      "learning_rate": 3.455670440882746e-06,
+      "loss": 0.5939,
+      "step": 7487
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.1547175216729215,
+      "learning_rate": 3.4544114412798447e-06,
+      "loss": 0.5356,
+      "step": 7488
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.0007252413208176,
+      "learning_rate": 3.4531525500159724e-06,
+      "loss": 0.6602,
+      "step": 7489
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.97396173912392,
+      "learning_rate": 3.451893767179375e-06,
+      "loss": 0.4509,
+      "step": 7490
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.641610811802933,
+      "learning_rate": 3.4506350928582878e-06,
+      "loss": 1.0176,
+      "step": 7491
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.148555565788872,
+      "learning_rate": 3.449376527140936e-06,
+      "loss": 0.8331,
+      "step": 7492
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.6461205122890075,
+      "learning_rate": 3.4481180701155435e-06,
+      "loss": 0.7634,
+      "step": 7493
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.602022670249473,
+      "learning_rate": 3.4468597218703203e-06,
+      "loss": 0.6074,
+      "step": 7494
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.776633567319659,
+      "learning_rate": 3.44560148249347e-06,
+      "loss": 1.1803,
+      "step": 7495
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.1504778031021816,
+      "learning_rate": 3.4443433520731908e-06,
+      "loss": 0.5468,
+      "step": 7496
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.802423189114528,
+      "learning_rate": 3.443085330697673e-06,
+      "loss": 0.7454,
+      "step": 7497
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.1758347245807057,
+      "learning_rate": 3.441827418455098e-06,
+      "loss": 0.7281,
+      "step": 7498
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.841314650957273,
+      "learning_rate": 3.44056961543364e-06,
+      "loss": 0.6456,
+      "step": 7499
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.301548555979058,
+      "learning_rate": 3.4393119217214643e-06,
+      "loss": 1.1869,
+      "step": 7500
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.4204896042815256,
+      "learning_rate": 3.438054337406732e-06,
+      "loss": 0.551,
+      "step": 7501
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.2855473071915275,
+      "learning_rate": 3.4367968625775923e-06,
+      "loss": 0.793,
+      "step": 7502
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.264205516013644,
+      "learning_rate": 3.435539497322189e-06,
+      "loss": 0.9784,
+      "step": 7503
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.189963158184375,
+      "learning_rate": 3.4342822417286586e-06,
+      "loss": 0.9427,
+      "step": 7504
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.31453330800363,
+      "learning_rate": 3.433025095885131e-06,
+      "loss": 0.922,
+      "step": 7505
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.0832892899562063,
+      "learning_rate": 3.4317680598797227e-06,
+      "loss": 0.4771,
+      "step": 7506
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.9636349208458657,
+      "learning_rate": 3.4305111338005483e-06,
+      "loss": 0.931,
+      "step": 7507
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.7181152910442608,
+      "learning_rate": 3.429254317735714e-06,
+      "loss": 0.4166,
+      "step": 7508
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.991948585753611,
+      "learning_rate": 3.4279976117733148e-06,
+      "loss": 1.1894,
+      "step": 7509
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.5817551031001433,
+      "learning_rate": 3.426741016001444e-06,
+      "loss": 0.5615,
+      "step": 7510
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.5481343706597617,
+      "learning_rate": 3.4254845305081796e-06,
+      "loss": 0.5636,
+      "step": 7511
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2740628976561783,
+      "learning_rate": 3.4242281553815963e-06,
+      "loss": 0.2066,
+      "step": 7512
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.854818236466965,
+      "learning_rate": 3.422971890709762e-06,
+      "loss": 1.035,
+      "step": 7513
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9314099983526627,
+      "learning_rate": 3.4217157365807352e-06,
+      "loss": 0.435,
+      "step": 7514
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.1837886110940596,
+      "learning_rate": 3.4204596930825674e-06,
+      "loss": 0.547,
+      "step": 7515
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.7329810664237026,
+      "learning_rate": 3.419203760303301e-06,
+      "loss": 0.7076,
+      "step": 7516
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.96816950610144,
+      "learning_rate": 3.417947938330971e-06,
+      "loss": 0.6511,
+      "step": 7517
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.106187483351225,
+      "learning_rate": 3.416692227253604e-06,
+      "loss": 0.6246,
+      "step": 7518
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.4304170799997538,
+      "learning_rate": 3.4154366271592222e-06,
+      "loss": 0.2782,
+      "step": 7519
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.7879286041422535,
+      "learning_rate": 3.4141811381358364e-06,
+      "loss": 1.486,
+      "step": 7520
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.3751720368744564,
+      "learning_rate": 3.4129257602714514e-06,
+      "loss": 0.6315,
+      "step": 7521
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.683938628475828,
+      "learning_rate": 3.4116704936540656e-06,
+      "loss": 0.7881,
+      "step": 7522
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.1457861012787807,
+      "learning_rate": 3.4104153383716644e-06,
+      "loss": 0.5855,
+      "step": 7523
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.8336870542509214,
+      "learning_rate": 3.4091602945122305e-06,
+      "loss": 0.5423,
+      "step": 7524
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.984032714223586,
+      "learning_rate": 3.4079053621637346e-06,
+      "loss": 0.7277,
+      "step": 7525
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.337185833290463,
+      "learning_rate": 3.4066505414141453e-06,
+      "loss": 1.0906,
+      "step": 7526
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.3899786036667607,
+      "learning_rate": 3.4053958323514185e-06,
+      "loss": 0.3782,
+      "step": 7527
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.017552575570357,
+      "learning_rate": 3.404141235063506e-06,
+      "loss": 0.377,
+      "step": 7528
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.178231000840538,
+      "learning_rate": 3.4028867496383454e-06,
+      "loss": 0.6578,
+      "step": 7529
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.787115596496799,
+      "learning_rate": 3.4016323761638737e-06,
+      "loss": 0.8246,
+      "step": 7530
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.1835348520912077,
+      "learning_rate": 3.400378114728017e-06,
+      "loss": 0.4493,
+      "step": 7531
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.0554416692959383,
+      "learning_rate": 3.399123965418692e-06,
+      "loss": 0.7489,
+      "step": 7532
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.752411187164596,
+      "learning_rate": 3.3978699283238117e-06,
+      "loss": 0.7197,
+      "step": 7533
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.4033966339527217,
+      "learning_rate": 3.396616003531275e-06,
+      "loss": 0.4122,
+      "step": 7534
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9913020044422074,
+      "learning_rate": 3.3953621911289784e-06,
+      "loss": 0.3394,
+      "step": 7535
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.767458762737945,
+      "learning_rate": 3.3941084912048094e-06,
+      "loss": 1.2963,
+      "step": 7536
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.405917623453279,
+      "learning_rate": 3.392854903846645e-06,
+      "loss": 0.8285,
+      "step": 7537
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.751814868947692,
+      "learning_rate": 3.39160142914236e-06,
+      "loss": 0.8188,
+      "step": 7538
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.227358384881683,
+      "learning_rate": 3.3903480671798145e-06,
+      "loss": 1.3352,
+      "step": 7539
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.904301672916866,
+      "learning_rate": 3.389094818046864e-06,
+      "loss": 0.5046,
+      "step": 7540
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.168266207804141,
+      "learning_rate": 3.3878416818313555e-06,
+      "loss": 0.9672,
+      "step": 7541
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.752994067863077,
+      "learning_rate": 3.3865886586211285e-06,
+      "loss": 0.6885,
+      "step": 7542
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.105697614699745,
+      "learning_rate": 3.385335748504015e-06,
+      "loss": 0.9677,
+      "step": 7543
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.653123686297841,
+      "learning_rate": 3.3840829515678386e-06,
+      "loss": 0.7451,
+      "step": 7544
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8318854848639077,
+      "learning_rate": 3.382830267900417e-06,
+      "loss": 0.3442,
+      "step": 7545
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.125126939314708,
+      "learning_rate": 3.381577697589554e-06,
+      "loss": 1.1589,
+      "step": 7546
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.436915972131372,
+      "learning_rate": 3.380325240723051e-06,
+      "loss": 0.5953,
+      "step": 7547
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.305223721072696,
+      "learning_rate": 3.3790728973886994e-06,
+      "loss": 0.8719,
+      "step": 7548
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.119188080328333,
+      "learning_rate": 3.377820667674283e-06,
+      "loss": 0.8093,
+      "step": 7549
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.495245249192985,
+      "learning_rate": 3.3765685516675805e-06,
+      "loss": 0.7286,
+      "step": 7550
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.3455194062619977,
+      "learning_rate": 3.3753165494563554e-06,
+      "loss": 0.7161,
+      "step": 7551
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.926704792998524,
+      "learning_rate": 3.3740646611283687e-06,
+      "loss": 0.9703,
+      "step": 7552
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.408141112156821,
+      "learning_rate": 3.3728128867713743e-06,
+      "loss": 0.8305,
+      "step": 7553
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.8160753357356905,
+      "learning_rate": 3.3715612264731155e-06,
+      "loss": 1.3461,
+      "step": 7554
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.103611211003639,
+      "learning_rate": 3.3703096803213263e-06,
+      "loss": 1.0364,
+      "step": 7555
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.048244769376617,
+      "learning_rate": 3.3690582484037393e-06,
+      "loss": 0.8223,
+      "step": 7556
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.33130685895782,
+      "learning_rate": 3.367806930808068e-06,
+      "loss": 0.9185,
+      "step": 7557
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.541873494179832,
+      "learning_rate": 3.366555727622028e-06,
+      "loss": 0.6951,
+      "step": 7558
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.521856016581727,
+      "learning_rate": 3.365304638933322e-06,
+      "loss": 0.2948,
+      "step": 7559
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.633589064668164,
+      "learning_rate": 3.3640536648296473e-06,
+      "loss": 1.1136,
+      "step": 7560
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.977653755333384,
+      "learning_rate": 3.362802805398692e-06,
+      "loss": 0.4834,
+      "step": 7561
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.62936734920231,
+      "learning_rate": 3.361552060728135e-06,
+      "loss": 0.9729,
+      "step": 7562
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.7147232100280645,
+      "learning_rate": 3.360301430905648e-06,
+      "loss": 1.0255,
+      "step": 7563
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.422361900806105,
+      "learning_rate": 3.3590509160188935e-06,
+      "loss": 0.7291,
+      "step": 7564
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.136258418736888,
+      "learning_rate": 3.3578005161555284e-06,
+      "loss": 0.6473,
+      "step": 7565
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.260504074934148,
+      "learning_rate": 3.3565502314032006e-06,
+      "loss": 1.5827,
+      "step": 7566
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.064556639812398,
+      "learning_rate": 3.355300061849549e-06,
+      "loss": 0.7217,
+      "step": 7567
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.6113118562312545,
+      "learning_rate": 3.354050007582207e-06,
+      "loss": 0.5611,
+      "step": 7568
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.3387791111899765,
+      "learning_rate": 3.3528000686887946e-06,
+      "loss": 0.6625,
+      "step": 7569
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.446240931359145,
+      "learning_rate": 3.3515502452569293e-06,
+      "loss": 0.7795,
+      "step": 7570
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9553336893169155,
+      "learning_rate": 3.350300537374217e-06,
+      "loss": 0.3551,
+      "step": 7571
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.8888905259600195,
+      "learning_rate": 3.3490509451282582e-06,
+      "loss": 0.6538,
+      "step": 7572
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.248790654389525,
+      "learning_rate": 3.3478014686066448e-06,
+      "loss": 0.8886,
+      "step": 7573
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.075163711164652,
+      "learning_rate": 3.3465521078969552e-06,
+      "loss": 0.4495,
+      "step": 7574
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.875334794198349,
+      "learning_rate": 3.345302863086768e-06,
+      "loss": 0.6683,
+      "step": 7575
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.0236885132985227,
+      "learning_rate": 3.3440537342636483e-06,
+      "loss": 0.4801,
+      "step": 7576
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.222717550677892,
+      "learning_rate": 3.3428047215151566e-06,
+      "loss": 1.1999,
+      "step": 7577
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.616336755588772,
+      "learning_rate": 3.3415558249288404e-06,
+      "loss": 0.9043,
+      "step": 7578
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.039906669102835,
+      "learning_rate": 3.340307044592245e-06,
+      "loss": 1.0051,
+      "step": 7579
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.1416453078958284,
+      "learning_rate": 3.3390583805929016e-06,
+      "loss": 0.5973,
+      "step": 7580
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.480569071267736,
+      "learning_rate": 3.3378098330183366e-06,
+      "loss": 0.6455,
+      "step": 7581
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.8108546519330186,
+      "learning_rate": 3.336561401956069e-06,
+      "loss": 0.568,
+      "step": 7582
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.7071940548697038,
+      "learning_rate": 3.3353130874936074e-06,
+      "loss": 1.0497,
+      "step": 7583
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.649638239315057,
+      "learning_rate": 3.3340648897184546e-06,
+      "loss": 0.6986,
+      "step": 7584
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.0271069934779264,
+      "learning_rate": 3.3328168087181036e-06,
+      "loss": 0.744,
+      "step": 7585
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.646900193708152,
+      "learning_rate": 3.3315688445800376e-06,
+      "loss": 0.9999,
+      "step": 7586
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.4062328965541,
+      "learning_rate": 3.330320997391734e-06,
+      "loss": 1.0649,
+      "step": 7587
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.968267573998116,
+      "learning_rate": 3.329073267240662e-06,
+      "loss": 0.3551,
+      "step": 7588
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.2727032722386675,
+      "learning_rate": 3.3278256542142818e-06,
+      "loss": 0.5636,
+      "step": 7589
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.485911056813762,
+      "learning_rate": 3.326578158400049e-06,
+      "loss": 0.3504,
+      "step": 7590
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.810087278594246,
+      "learning_rate": 3.325330779885401e-06,
+      "loss": 0.8325,
+      "step": 7591
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.5155577878663395,
+      "learning_rate": 3.324083518757778e-06,
+      "loss": 0.7222,
+      "step": 7592
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.226243750137132,
+      "learning_rate": 3.322836375104608e-06,
+      "loss": 0.8349,
+      "step": 7593
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.579907909260929,
+      "learning_rate": 3.3215893490133076e-06,
+      "loss": 0.8511,
+      "step": 7594
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.839432439727437,
+      "learning_rate": 3.320342440571289e-06,
+      "loss": 0.4744,
+      "step": 7595
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.348776624156341,
+      "learning_rate": 3.319095649865958e-06,
+      "loss": 1.0481,
+      "step": 7596
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.9393705216616834,
+      "learning_rate": 3.3178489769847046e-06,
+      "loss": 0.7586,
+      "step": 7597
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.455380150704934,
+      "learning_rate": 3.3166024220149173e-06,
+      "loss": 0.5484,
+      "step": 7598
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.055728788130167,
+      "learning_rate": 3.3153559850439737e-06,
+      "loss": 1.0842,
+      "step": 7599
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.7015896855427592,
+      "learning_rate": 3.3141096661592455e-06,
+      "loss": 0.5129,
+      "step": 7600
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.337951165458357,
+      "learning_rate": 3.3128634654480906e-06,
+      "loss": 0.9046,
+      "step": 7601
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.379378514832218,
+      "learning_rate": 3.3116173829978666e-06,
+      "loss": 0.8637,
+      "step": 7602
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.7526183219145404,
+      "learning_rate": 3.3103714188959156e-06,
+      "loss": 0.6631,
+      "step": 7603
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.788574484025864,
+      "learning_rate": 3.3091255732295736e-06,
+      "loss": 0.9459,
+      "step": 7604
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.6683996965418135,
+      "learning_rate": 3.3078798460861704e-06,
+      "loss": 0.5804,
+      "step": 7605
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.916929575285865,
+      "learning_rate": 3.306634237553026e-06,
+      "loss": 0.9159,
+      "step": 7606
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.091661616270836,
+      "learning_rate": 3.305388747717453e-06,
+      "loss": 0.7485,
+      "step": 7607
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.253125457820861,
+      "learning_rate": 3.3041433766667535e-06,
+      "loss": 0.5195,
+      "step": 7608
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.9477457676776226,
+      "learning_rate": 3.302898124488222e-06,
+      "loss": 0.8162,
+      "step": 7609
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.8355738202745426,
+      "learning_rate": 3.3016529912691476e-06,
+      "loss": 0.9991,
+      "step": 7610
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.310163980369584,
+      "learning_rate": 3.3004079770968055e-06,
+      "loss": 0.6773,
+      "step": 7611
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.6459081353394605,
+      "learning_rate": 3.299163082058468e-06,
+      "loss": 0.5445,
+      "step": 7612
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.2563150817337876,
+      "learning_rate": 3.297918306241399e-06,
+      "loss": 0.745,
+      "step": 7613
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.496579311994659,
+      "learning_rate": 3.2966736497328463e-06,
+      "loss": 0.3865,
+      "step": 7614
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.477069154674302,
+      "learning_rate": 3.2954291126200577e-06,
+      "loss": 0.8421,
+      "step": 7615
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.45013467679716,
+      "learning_rate": 3.294184694990271e-06,
+      "loss": 1.1175,
+      "step": 7616
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.9947369308971163,
+      "learning_rate": 3.2929403969307137e-06,
+      "loss": 0.7213,
+      "step": 7617
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.387494139290766,
+      "learning_rate": 3.291696218528605e-06,
+      "loss": 1.2012,
+      "step": 7618
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.169223623551574,
+      "learning_rate": 3.290452159871158e-06,
+      "loss": 1.0328,
+      "step": 7619
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.866520823535412,
+      "learning_rate": 3.289208221045573e-06,
+      "loss": 0.5849,
+      "step": 7620
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9425109571369379,
+      "learning_rate": 3.2879644021390468e-06,
+      "loss": 0.3425,
+      "step": 7621
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.468745940756694,
+      "learning_rate": 3.286720703238765e-06,
+      "loss": 0.7145,
+      "step": 7622
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.913128123597565,
+      "learning_rate": 3.2854771244319052e-06,
+      "loss": 0.7324,
+      "step": 7623
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.211279066072754,
+      "learning_rate": 3.2842336658056383e-06,
+      "loss": 0.6421,
+      "step": 7624
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.2310589376655137,
+      "learning_rate": 3.2829903274471253e-06,
+      "loss": 0.5612,
+      "step": 7625
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.2109699398316724,
+      "learning_rate": 3.281747109443517e-06,
+      "loss": 1.3728,
+      "step": 7626
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.4741011651583875,
+      "learning_rate": 3.2805040118819574e-06,
+      "loss": 0.9257,
+      "step": 7627
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.96943882155435,
+      "learning_rate": 3.279261034849584e-06,
+      "loss": 0.7778,
+      "step": 7628
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.230742837335636,
+      "learning_rate": 3.278018178433523e-06,
+      "loss": 0.5448,
+      "step": 7629
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.6382394352891763,
+      "learning_rate": 3.276775442720896e-06,
+      "loss": 0.7383,
+      "step": 7630
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.943406264757444,
+      "learning_rate": 3.2755328277988084e-06,
+      "loss": 1.0892,
+      "step": 7631
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.364702175304437,
+      "learning_rate": 3.274290333754365e-06,
+      "loss": 1.0101,
+      "step": 7632
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.5276837531138274,
+      "learning_rate": 3.2730479606746594e-06,
+      "loss": 0.5491,
+      "step": 7633
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.3067665969011113,
+      "learning_rate": 3.271805708646776e-06,
+      "loss": 0.7095,
+      "step": 7634
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.819460150118888,
+      "learning_rate": 3.2705635777577904e-06,
+      "loss": 0.5795,
+      "step": 7635
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.61967776942314,
+      "learning_rate": 3.2693215680947737e-06,
+      "loss": 0.6914,
+      "step": 7636
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.6878796081754066,
+      "learning_rate": 3.268079679744781e-06,
+      "loss": 0.5596,
+      "step": 7637
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0924427201373974,
+      "learning_rate": 3.2668379127948656e-06,
+      "loss": 0.3269,
+      "step": 7638
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.884712377878976,
+      "learning_rate": 3.26559626733207e-06,
+      "loss": 0.8669,
+      "step": 7639
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.37203318965418,
+      "learning_rate": 3.264354743443429e-06,
+      "loss": 0.8481,
+      "step": 7640
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.399144750533389,
+      "learning_rate": 3.2631133412159656e-06,
+      "loss": 0.8149,
+      "step": 7641
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.265972866984133,
+      "learning_rate": 3.2618720607367e-06,
+      "loss": 0.6379,
+      "step": 7642
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.488346454035597,
+      "learning_rate": 3.2606309020926364e-06,
+      "loss": 1.2761,
+      "step": 7643
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.0438213532347667,
+      "learning_rate": 3.2593898653707773e-06,
+      "loss": 0.6676,
+      "step": 7644
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.061186580438532,
+      "learning_rate": 3.2581489506581134e-06,
+      "loss": 1.1957,
+      "step": 7645
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.092072730761351,
+      "learning_rate": 3.2569081580416273e-06,
+      "loss": 0.562,
+      "step": 7646
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.5801424147558554,
+      "learning_rate": 3.2556674876082937e-06,
+      "loss": 0.527,
+      "step": 7647
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.061391757672997,
+      "learning_rate": 3.254426939445079e-06,
+      "loss": 0.7016,
+      "step": 7648
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.898385610961387,
+      "learning_rate": 3.2531865136389383e-06,
+      "loss": 0.7654,
+      "step": 7649
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.1875028295410495,
+      "learning_rate": 3.251946210276821e-06,
+      "loss": 1.0348,
+      "step": 7650
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.034755909633507,
+      "learning_rate": 3.2507060294456653e-06,
+      "loss": 0.5311,
+      "step": 7651
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.705049010180583,
+      "learning_rate": 3.249465971232405e-06,
+      "loss": 0.5876,
+      "step": 7652
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.3703704410665343,
+      "learning_rate": 3.248226035723963e-06,
+      "loss": 0.4805,
+      "step": 7653
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7354395150446371,
+      "learning_rate": 3.2469862230072507e-06,
+      "loss": 0.2601,
+      "step": 7654
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.281087596216457,
+      "learning_rate": 3.245746533169175e-06,
+      "loss": 0.8449,
+      "step": 7655
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.8102663809612016,
+      "learning_rate": 3.244506966296633e-06,
+      "loss": 0.9758,
+      "step": 7656
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0974875755818427,
+      "learning_rate": 3.2432675224765133e-06,
+      "loss": 0.2948,
+      "step": 7657
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.934298887973439,
+      "learning_rate": 3.242028201795694e-06,
+      "loss": 0.8,
+      "step": 7658
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.9409185012799575,
+      "learning_rate": 3.240789004341049e-06,
+      "loss": 0.846,
+      "step": 7659
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.956801842865033,
+      "learning_rate": 3.2395499301994366e-06,
+      "loss": 0.272,
+      "step": 7660
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.100245111197015,
+      "learning_rate": 3.238310979457713e-06,
+      "loss": 0.5009,
+      "step": 7661
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.908941130615876,
+      "learning_rate": 3.2370721522027226e-06,
+      "loss": 0.8223,
+      "step": 7662
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.062481040532317,
+      "learning_rate": 3.235833448521303e-06,
+      "loss": 0.5634,
+      "step": 7663
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.42252314474069,
+      "learning_rate": 3.2345948685002796e-06,
+      "loss": 1.076,
+      "step": 7664
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.611775133421519,
+      "learning_rate": 3.2333564122264755e-06,
+      "loss": 0.5999,
+      "step": 7665
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.5563105544668465,
+      "learning_rate": 3.2321180797866962e-06,
+      "loss": 0.7487,
+      "step": 7666
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 6.936191030571098,
+      "learning_rate": 3.2308798712677456e-06,
+      "loss": 1.237,
+      "step": 7667
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.145405830437945,
+      "learning_rate": 3.2296417867564166e-06,
+      "loss": 0.5232,
+      "step": 7668
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.453801351461035,
+      "learning_rate": 3.2284038263394946e-06,
+      "loss": 0.4514,
+      "step": 7669
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.517791485006161,
+      "learning_rate": 3.2271659901037555e-06,
+      "loss": 0.6854,
+      "step": 7670
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.6261401874406745,
+      "learning_rate": 3.2259282781359634e-06,
+      "loss": 0.763,
+      "step": 7671
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.000540569718012,
+      "learning_rate": 3.224690690522879e-06,
+      "loss": 0.693,
+      "step": 7672
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.120716139586882,
+      "learning_rate": 3.22345322735125e-06,
+      "loss": 1.1604,
+      "step": 7673
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.383118903590275,
+      "learning_rate": 3.2222158887078187e-06,
+      "loss": 0.2229,
+      "step": 7674
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.976290591456961,
+      "learning_rate": 3.2209786746793163e-06,
+      "loss": 0.9455,
+      "step": 7675
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5314161645389535,
+      "learning_rate": 3.219741585352469e-06,
+      "loss": 0.2351,
+      "step": 7676
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.191337605130976,
+      "learning_rate": 3.218504620813986e-06,
+      "loss": 0.442,
+      "step": 7677
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.561383212432562,
+      "learning_rate": 3.2172677811505766e-06,
+      "loss": 0.5061,
+      "step": 7678
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.391575981241691,
+      "learning_rate": 3.216031066448938e-06,
+      "loss": 0.6082,
+      "step": 7679
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.234858555107159,
+      "learning_rate": 3.2147944767957565e-06,
+      "loss": 1.1577,
+      "step": 7680
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6846154268784963,
+      "learning_rate": 3.213558012277713e-06,
+      "loss": 0.3,
+      "step": 7681
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.810587396979411,
+      "learning_rate": 3.212321672981481e-06,
+      "loss": 1.0758,
+      "step": 7682
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.22152061579591,
+      "learning_rate": 3.2110854589937166e-06,
+      "loss": 0.259,
+      "step": 7683
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.48217190544856,
+      "learning_rate": 3.2098493704010768e-06,
+      "loss": 0.741,
+      "step": 7684
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.013228565717424,
+      "learning_rate": 3.208613407290206e-06,
+      "loss": 0.9747,
+      "step": 7685
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.6159848966723724,
+      "learning_rate": 3.2073775697477393e-06,
+      "loss": 0.8999,
+      "step": 7686
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.7380397280342086,
+      "learning_rate": 3.2061418578603028e-06,
+      "loss": 0.5477,
+      "step": 7687
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.539704720378294,
+      "learning_rate": 3.2049062717145168e-06,
+      "loss": 0.4176,
+      "step": 7688
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.274538445434723,
+      "learning_rate": 3.203670811396987e-06,
+      "loss": 0.975,
+      "step": 7689
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.67172399250756,
+      "learning_rate": 3.2024354769943163e-06,
+      "loss": 0.9177,
+      "step": 7690
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.475114485121228,
+      "learning_rate": 3.2012002685930947e-06,
+      "loss": 0.8625,
+      "step": 7691
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.3914588561677315,
+      "learning_rate": 3.1999651862799063e-06,
+      "loss": 0.5683,
+      "step": 7692
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.081062531847189,
+      "learning_rate": 3.198730230141327e-06,
+      "loss": 0.4778,
+      "step": 7693
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.118738109466712,
+      "learning_rate": 3.197495400263917e-06,
+      "loss": 0.9559,
+      "step": 7694
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.8691710480218955,
+      "learning_rate": 3.1962606967342356e-06,
+      "loss": 0.7172,
+      "step": 7695
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.794101928588226,
+      "learning_rate": 3.1950261196388287e-06,
+      "loss": 0.974,
+      "step": 7696
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.493952052033231,
+      "learning_rate": 3.1937916690642356e-06,
+      "loss": 0.9769,
+      "step": 7697
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.172623602492804,
+      "learning_rate": 3.192557345096986e-06,
+      "loss": 0.4682,
+      "step": 7698
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.5391023927997174,
+      "learning_rate": 3.191323147823602e-06,
+      "loss": 0.3686,
+      "step": 7699
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.2908197921338616,
+      "learning_rate": 3.1900890773305926e-06,
+      "loss": 0.613,
+      "step": 7700
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.957444384242329,
+      "learning_rate": 3.1888551337044615e-06,
+      "loss": 0.1528,
+      "step": 7701
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.736174740514147,
+      "learning_rate": 3.1876213170317048e-06,
+      "loss": 0.2971,
+      "step": 7702
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.9588233427147275,
+      "learning_rate": 3.186387627398805e-06,
+      "loss": 0.7388,
+      "step": 7703
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.5151779126495444,
+      "learning_rate": 3.1851540648922398e-06,
+      "loss": 0.5993,
+      "step": 7704
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.3530189237922574,
+      "learning_rate": 3.1839206295984786e-06,
+      "loss": 0.3692,
+      "step": 7705
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.095440207688822,
+      "learning_rate": 3.1826873216039757e-06,
+      "loss": 0.3849,
+      "step": 7706
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.292342872790543,
+      "learning_rate": 3.181454140995182e-06,
+      "loss": 0.6714,
+      "step": 7707
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.1720375945195083,
+      "learning_rate": 3.1802210878585395e-06,
+      "loss": 0.5537,
+      "step": 7708
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.183990698007337,
+      "learning_rate": 3.1789881622804797e-06,
+      "loss": 0.8163,
+      "step": 7709
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.5311093273622,
+      "learning_rate": 3.1777553643474247e-06,
+      "loss": 0.6283,
+      "step": 7710
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.442566058851378,
+      "learning_rate": 3.1765226941457866e-06,
+      "loss": 0.8416,
+      "step": 7711
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.1627473864282765,
+      "learning_rate": 3.1752901517619733e-06,
+      "loss": 0.6516,
+      "step": 7712
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.5284453351576794,
+      "learning_rate": 3.1740577372823785e-06,
+      "loss": 0.6561,
+      "step": 7713
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.3797594442551833,
+      "learning_rate": 3.1728254507933892e-06,
+      "loss": 0.5099,
+      "step": 7714
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.576113825725778,
+      "learning_rate": 3.1715932923813843e-06,
+      "loss": 1.1334,
+      "step": 7715
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.286629317778869,
+      "learning_rate": 3.170361262132734e-06,
+      "loss": 0.845,
+      "step": 7716
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.009390554053943,
+      "learning_rate": 3.1691293601337953e-06,
+      "loss": 0.8019,
+      "step": 7717
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.370703724590661,
+      "learning_rate": 3.16789758647092e-06,
+      "loss": 0.7708,
+      "step": 7718
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.200191529868398,
+      "learning_rate": 3.166665941230451e-06,
+      "loss": 1.1286,
+      "step": 7719
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.135015759148786,
+      "learning_rate": 3.1654344244987213e-06,
+      "loss": 1.1156,
+      "step": 7720
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.4909081881591635,
+      "learning_rate": 3.1642030363620534e-06,
+      "loss": 0.5092,
+      "step": 7721
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.492886760434505,
+      "learning_rate": 3.1629717769067654e-06,
+      "loss": 1.2737,
+      "step": 7722
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.733333203752356,
+      "learning_rate": 3.161740646219159e-06,
+      "loss": 0.7632,
+      "step": 7723
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.695872923314846,
+      "learning_rate": 3.1605096443855333e-06,
+      "loss": 0.853,
+      "step": 7724
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.2332939949985526,
+      "learning_rate": 3.159278771492176e-06,
+      "loss": 0.4684,
+      "step": 7725
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.055708716202285,
+      "learning_rate": 3.1580480276253665e-06,
+      "loss": 1.2795,
+      "step": 7726
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.895087699370973,
+      "learning_rate": 3.1568174128713738e-06,
+      "loss": 0.5558,
+      "step": 7727
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.45024502271313,
+      "learning_rate": 3.15558692731646e-06,
+      "loss": 0.7325,
+      "step": 7728
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.3016898771741525,
+      "learning_rate": 3.1543565710468743e-06,
+      "loss": 0.4324,
+      "step": 7729
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.6489388052277896,
+      "learning_rate": 3.1531263441488607e-06,
+      "loss": 0.7163,
+      "step": 7730
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.193545400934648,
+      "learning_rate": 3.1518962467086527e-06,
+      "loss": 1.1451,
+      "step": 7731
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.479125338088091,
+      "learning_rate": 3.150666278812475e-06,
+      "loss": 1.0333,
+      "step": 7732
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.0994094237191834,
+      "learning_rate": 3.149436440546545e-06,
+      "loss": 0.7642,
+      "step": 7733
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.9580754736867,
+      "learning_rate": 3.1482067319970642e-06,
+      "loss": 1.2643,
+      "step": 7734
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.753015333630123,
+      "learning_rate": 3.1469771532502336e-06,
+      "loss": 0.8311,
+      "step": 7735
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.4290167032293177,
+      "learning_rate": 3.145747704392239e-06,
+      "loss": 0.8285,
+      "step": 7736
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.3129908290943595,
+      "learning_rate": 3.144518385509261e-06,
+      "loss": 0.5139,
+      "step": 7737
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.672563867413507,
+      "learning_rate": 3.143289196687469e-06,
+      "loss": 1.1612,
+      "step": 7738
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 6.207846649305042,
+      "learning_rate": 3.142060138013026e-06,
+      "loss": 1.1907,
+      "step": 7739
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.780734791010286,
+      "learning_rate": 3.1408312095720794e-06,
+      "loss": 0.508,
+      "step": 7740
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.380561111288435,
+      "learning_rate": 3.139602411450774e-06,
+      "loss": 0.6048,
+      "step": 7741
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.586262330486288,
+      "learning_rate": 3.138373743735244e-06,
+      "loss": 0.4451,
+      "step": 7742
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.8144155997734766,
+      "learning_rate": 3.1371452065116116e-06,
+      "loss": 0.476,
+      "step": 7743
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.15381259807162,
+      "learning_rate": 3.1359167998659933e-06,
+      "loss": 0.752,
+      "step": 7744
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.023697205289873,
+      "learning_rate": 3.134688523884497e-06,
+      "loss": 0.9533,
+      "step": 7745
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.9939570834213014,
+      "learning_rate": 3.1334603786532147e-06,
+      "loss": 1.0225,
+      "step": 7746
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9617810999492078,
+      "learning_rate": 3.1322323642582374e-06,
+      "loss": 0.4235,
+      "step": 7747
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.096482581655145,
+      "learning_rate": 3.131004480785642e-06,
+      "loss": 0.7987,
+      "step": 7748
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.7038563822744615,
+      "learning_rate": 3.1297767283214998e-06,
+      "loss": 0.9277,
+      "step": 7749
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.025907679321576,
+      "learning_rate": 3.1285491069518705e-06,
+      "loss": 0.5833,
+      "step": 7750
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.7108010699875966,
+      "learning_rate": 3.127321616762803e-06,
+      "loss": 0.6736,
+      "step": 7751
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.533908261684587,
+      "learning_rate": 3.1260942578403395e-06,
+      "loss": 0.9325,
+      "step": 7752
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.5851016859848137,
+      "learning_rate": 3.1248670302705143e-06,
+      "loss": 0.614,
+      "step": 7753
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.291879126802277,
+      "learning_rate": 3.1236399341393486e-06,
+      "loss": 0.8511,
+      "step": 7754
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.166298720756765,
+      "learning_rate": 3.122412969532858e-06,
+      "loss": 0.413,
+      "step": 7755
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.7107254456828316,
+      "learning_rate": 3.121186136537049e-06,
+      "loss": 0.2174,
+      "step": 7756
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.405980873573781,
+      "learning_rate": 3.119959435237913e-06,
+      "loss": 0.4214,
+      "step": 7757
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.297894197921948,
+      "learning_rate": 3.11873286572144e-06,
+      "loss": 1.178,
+      "step": 7758
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4532413310352328,
+      "learning_rate": 3.1175064280736044e-06,
+      "loss": 0.2372,
+      "step": 7759
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.050043333659418,
+      "learning_rate": 3.1162801223803756e-06,
+      "loss": 0.8248,
+      "step": 7760
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.781108835281411,
+      "learning_rate": 3.1150539487277125e-06,
+      "loss": 1.0726,
+      "step": 7761
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.1836495197151167,
+      "learning_rate": 3.1138279072015666e-06,
+      "loss": 0.8683,
+      "step": 7762
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.032521770289908,
+      "learning_rate": 3.112601997887873e-06,
+      "loss": 0.7148,
+      "step": 7763
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.2801631855651623,
+      "learning_rate": 3.111376220872565e-06,
+      "loss": 0.567,
+      "step": 7764
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9504549299211302,
+      "learning_rate": 3.1101505762415668e-06,
+      "loss": 0.3778,
+      "step": 7765
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.444505613029491,
+      "learning_rate": 3.1089250640807865e-06,
+      "loss": 1.1581,
+      "step": 7766
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.876983373389087,
+      "learning_rate": 3.10769968447613e-06,
+      "loss": 0.9147,
+      "step": 7767
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.9787989385327642,
+      "learning_rate": 3.106474437513492e-06,
+      "loss": 0.5907,
+      "step": 7768
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.775997969258128,
+      "learning_rate": 3.1052493232787533e-06,
+      "loss": 0.9217,
+      "step": 7769
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.468930199595321,
+      "learning_rate": 3.104024341857791e-06,
+      "loss": 0.9899,
+      "step": 7770
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.359483541220007,
+      "learning_rate": 3.1027994933364715e-06,
+      "loss": 0.9143,
+      "step": 7771
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.138079374083297,
+      "learning_rate": 3.101574777800651e-06,
+      "loss": 0.5309,
+      "step": 7772
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.473590000027721,
+      "learning_rate": 3.100350195336177e-06,
+      "loss": 0.7457,
+      "step": 7773
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.1993030318420637,
+      "learning_rate": 3.099125746028887e-06,
+      "loss": 0.3841,
+      "step": 7774
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.940398928712938,
+      "learning_rate": 3.0979014299646088e-06,
+      "loss": 0.4528,
+      "step": 7775
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1397946602839084,
+      "learning_rate": 3.0966772472291623e-06,
+      "loss": 0.1403,
+      "step": 7776
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.985916223616124,
+      "learning_rate": 3.0954531979083575e-06,
+      "loss": 0.7692,
+      "step": 7777
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.298616697163275,
+      "learning_rate": 3.094229282087995e-06,
+      "loss": 1.0578,
+      "step": 7778
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.393406375595787,
+      "learning_rate": 3.0930054998538672e-06,
+      "loss": 0.7686,
+      "step": 7779
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.8192625380118126,
+      "learning_rate": 3.091781851291753e-06,
+      "loss": 0.5322,
+      "step": 7780
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.437817787193546,
+      "learning_rate": 3.0905583364874282e-06,
+      "loss": 1.3211,
+      "step": 7781
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.235151842126615,
+      "learning_rate": 3.0893349555266517e-06,
+      "loss": 1.1217,
+      "step": 7782
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.0124361892118876,
+      "learning_rate": 3.088111708495181e-06,
+      "loss": 0.5699,
+      "step": 7783
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.278913011180968,
+      "learning_rate": 3.0868885954787577e-06,
+      "loss": 0.6601,
+      "step": 7784
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.3234343380067917,
+      "learning_rate": 3.0856656165631204e-06,
+      "loss": 0.4015,
+      "step": 7785
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.225939320324209,
+      "learning_rate": 3.084442771833991e-06,
+      "loss": 0.4539,
+      "step": 7786
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.555604538407607,
+      "learning_rate": 3.0832200613770857e-06,
+      "loss": 0.5867,
+      "step": 7787
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.8838307134062102,
+      "learning_rate": 3.081997485278113e-06,
+      "loss": 0.3765,
+      "step": 7788
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.6069289209766198,
+      "learning_rate": 3.0807750436227695e-06,
+      "loss": 0.7507,
+      "step": 7789
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.254856263935714,
+      "learning_rate": 3.079552736496745e-06,
+      "loss": 0.3731,
+      "step": 7790
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.8397696175368514,
+      "learning_rate": 3.0783305639857132e-06,
+      "loss": 1.0205,
+      "step": 7791
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 6.492414526493002,
+      "learning_rate": 3.077108526175345e-06,
+      "loss": 1.1767,
+      "step": 7792
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.120289953683984,
+      "learning_rate": 3.075886623151302e-06,
+      "loss": 0.856,
+      "step": 7793
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.1113371636214087,
+      "learning_rate": 3.074664854999232e-06,
+      "loss": 0.6099,
+      "step": 7794
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.380585449987341,
+      "learning_rate": 3.0734432218047783e-06,
+      "loss": 0.6947,
+      "step": 7795
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.683271562483727,
+      "learning_rate": 3.07222172365357e-06,
+      "loss": 0.6046,
+      "step": 7796
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.127171353450437,
+      "learning_rate": 3.0710003606312292e-06,
+      "loss": 0.6197,
+      "step": 7797
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.421346023231678,
+      "learning_rate": 3.069779132823367e-06,
+      "loss": 0.9477,
+      "step": 7798
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.286625588838192,
+      "learning_rate": 3.068558040315588e-06,
+      "loss": 1.1859,
+      "step": 7799
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.930364597159306,
+      "learning_rate": 3.0673370831934833e-06,
+      "loss": 0.6226,
+      "step": 7800
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.8869195032304655,
+      "learning_rate": 3.066116261542639e-06,
+      "loss": 0.5016,
+      "step": 7801
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.3190297058512295,
+      "learning_rate": 3.064895575448631e-06,
+      "loss": 0.9508,
+      "step": 7802
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.201891244728968,
+      "learning_rate": 3.0636750249970184e-06,
+      "loss": 0.6445,
+      "step": 7803
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9073656928455973,
+      "learning_rate": 3.062454610273361e-06,
+      "loss": 0.3428,
+      "step": 7804
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.1274350126950825,
+      "learning_rate": 3.061234331363203e-06,
+      "loss": 0.5776,
+      "step": 7805
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.757643213513995,
+      "learning_rate": 3.0600141883520796e-06,
+      "loss": 0.8018,
+      "step": 7806
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.816896954660318,
+      "learning_rate": 3.0587941813255196e-06,
+      "loss": 0.6786,
+      "step": 7807
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.991835587302752,
+      "learning_rate": 3.0575743103690408e-06,
+      "loss": 1.0708,
+      "step": 7808
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.723297776792715,
+      "learning_rate": 3.056354575568148e-06,
+      "loss": 1.0296,
+      "step": 7809
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 6.138433420633675,
+      "learning_rate": 3.05513497700834e-06,
+      "loss": 1.1589,
+      "step": 7810
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.7835659165003084,
+      "learning_rate": 3.0539155147751074e-06,
+      "loss": 0.3927,
+      "step": 7811
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.000530651589436,
+      "learning_rate": 3.0526961889539265e-06,
+      "loss": 0.642,
+      "step": 7812
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.289083728058848,
+      "learning_rate": 3.0514769996302696e-06,
+      "loss": 1.0834,
+      "step": 7813
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.179387690922487,
+      "learning_rate": 3.050257946889594e-06,
+      "loss": 0.7185,
+      "step": 7814
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.614143591546221,
+      "learning_rate": 3.049039030817351e-06,
+      "loss": 0.636,
+      "step": 7815
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0175919934837028,
+      "learning_rate": 3.0478202514989813e-06,
+      "loss": 0.3539,
+      "step": 7816
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.425152565734742,
+      "learning_rate": 3.046601609019916e-06,
+      "loss": 1.0657,
+      "step": 7817
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7070207030060107,
+      "learning_rate": 3.0453831034655766e-06,
+      "loss": 0.2859,
+      "step": 7818
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.551122283558936,
+      "learning_rate": 3.0441647349213764e-06,
+      "loss": 1.3125,
+      "step": 7819
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.221502422957849,
+      "learning_rate": 3.042946503472716e-06,
+      "loss": 0.9485,
+      "step": 7820
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.348367201920117,
+      "learning_rate": 3.041728409204988e-06,
+      "loss": 0.3886,
+      "step": 7821
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.970306040011278,
+      "learning_rate": 3.040510452203576e-06,
+      "loss": 1.1752,
+      "step": 7822
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.290881887789476,
+      "learning_rate": 3.039292632553853e-06,
+      "loss": 0.9586,
+      "step": 7823
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.038877725597535,
+      "learning_rate": 3.038074950341184e-06,
+      "loss": 0.799,
+      "step": 7824
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.033598891945941,
+      "learning_rate": 3.036857405650925e-06,
+      "loss": 0.1455,
+      "step": 7825
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.3779552426693624,
+      "learning_rate": 3.0356399985684153e-06,
+      "loss": 0.7625,
+      "step": 7826
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.673095220192702,
+      "learning_rate": 3.0344227291789928e-06,
+      "loss": 0.9734,
+      "step": 7827
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.6096029345475955,
+      "learning_rate": 3.033205597567984e-06,
+      "loss": 0.4377,
+      "step": 7828
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.324901287473498,
+      "learning_rate": 3.0319886038207023e-06,
+      "loss": 0.9661,
+      "step": 7829
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.727965111416931,
+      "learning_rate": 3.0307717480224572e-06,
+      "loss": 0.3719,
+      "step": 7830
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.456816251346994,
+      "learning_rate": 3.02955503025854e-06,
+      "loss": 0.8164,
+      "step": 7831
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.2486302261587734,
+      "learning_rate": 3.0283384506142397e-06,
+      "loss": 0.572,
+      "step": 7832
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0676076427367587,
+      "learning_rate": 3.027122009174834e-06,
+      "loss": 0.1451,
+      "step": 7833
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.957496653333703,
+      "learning_rate": 3.0259057060255887e-06,
+      "loss": 0.7304,
+      "step": 7834
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.362065522951962,
+      "learning_rate": 3.024689541251763e-06,
+      "loss": 0.9303,
+      "step": 7835
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.788726112158268,
+      "learning_rate": 3.023473514938604e-06,
+      "loss": 1.4238,
+      "step": 7836
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.3604292834068765,
+      "learning_rate": 3.02225762717135e-06,
+      "loss": 0.5941,
+      "step": 7837
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.26210210267571,
+      "learning_rate": 3.021041878035228e-06,
+      "loss": 0.8583,
+      "step": 7838
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.0775307873462916,
+      "learning_rate": 3.0198262676154583e-06,
+      "loss": 0.573,
+      "step": 7839
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0254216863292878,
+      "learning_rate": 3.018610795997249e-06,
+      "loss": 0.4853,
+      "step": 7840
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.7059526815793205,
+      "learning_rate": 3.0173954632657996e-06,
+      "loss": 0.3955,
+      "step": 7841
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.240168978264894,
+      "learning_rate": 3.0161802695063024e-06,
+      "loss": 0.8435,
+      "step": 7842
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.0858055095332695,
+      "learning_rate": 3.014965214803933e-06,
+      "loss": 1.332,
+      "step": 7843
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.2462707758145446,
+      "learning_rate": 3.013750299243864e-06,
+      "loss": 0.5558,
+      "step": 7844
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.650193179594442,
+      "learning_rate": 3.0125355229112536e-06,
+      "loss": 1.0065,
+      "step": 7845
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.7676518700399604,
+      "learning_rate": 3.0113208858912533e-06,
+      "loss": 1.1567,
+      "step": 7846
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9609440709283663,
+      "learning_rate": 3.0101063882690046e-06,
+      "loss": 0.3932,
+      "step": 7847
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.373511093110997,
+      "learning_rate": 3.00889203012964e-06,
+      "loss": 0.6438,
+      "step": 7848
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.3333735327867577,
+      "learning_rate": 3.007677811558276e-06,
+      "loss": 0.3598,
+      "step": 7849
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.5025719551068635,
+      "learning_rate": 3.006463732640028e-06,
+      "loss": 0.8174,
+      "step": 7850
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.132660480178307,
+      "learning_rate": 3.0052497934599966e-06,
+      "loss": 0.771,
+      "step": 7851
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.51776375842111,
+      "learning_rate": 3.0040359941032727e-06,
+      "loss": 1.3559,
+      "step": 7852
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.977461470056396,
+      "learning_rate": 3.0028223346549413e-06,
+      "loss": 0.6852,
+      "step": 7853
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.09231298367694,
+      "learning_rate": 3.00160881520007e-06,
+      "loss": 0.5279,
+      "step": 7854
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.180758071905827,
+      "learning_rate": 3.000395435823724e-06,
+      "loss": 0.3719,
+      "step": 7855
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.916943911528527,
+      "learning_rate": 2.9991821966109558e-06,
+      "loss": 0.4142,
+      "step": 7856
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.163191128066598,
+      "learning_rate": 2.9979690976468083e-06,
+      "loss": 1.2684,
+      "step": 7857
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.612341928275986,
+      "learning_rate": 2.9967561390163148e-06,
+      "loss": 1.0136,
+      "step": 7858
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9913312384897335,
+      "learning_rate": 2.9955433208044983e-06,
+      "loss": 0.2951,
+      "step": 7859
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.2750060420121265,
+      "learning_rate": 2.994330643096371e-06,
+      "loss": 0.6136,
+      "step": 7860
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.1301120622750145,
+      "learning_rate": 2.993118105976936e-06,
+      "loss": 0.3387,
+      "step": 7861
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.1105291461179583,
+      "learning_rate": 2.9919057095311874e-06,
+      "loss": 0.6402,
+      "step": 7862
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.6144575613625025,
+      "learning_rate": 2.99069345384411e-06,
+      "loss": 0.6512,
+      "step": 7863
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.065422619060882,
+      "learning_rate": 2.9894813390006773e-06,
+      "loss": 0.6632,
+      "step": 7864
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.9735292436274023,
+      "learning_rate": 2.988269365085854e-06,
+      "loss": 0.7177,
+      "step": 7865
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.2401111070258053,
+      "learning_rate": 2.9870575321845916e-06,
+      "loss": 0.5627,
+      "step": 7866
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.697499884583697,
+      "learning_rate": 2.985845840381837e-06,
+      "loss": 1.6977,
+      "step": 7867
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.468404310057439,
+      "learning_rate": 2.9846342897625215e-06,
+      "loss": 1.0907,
+      "step": 7868
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.3973242492973816,
+      "learning_rate": 2.983422880411572e-06,
+      "loss": 0.5834,
+      "step": 7869
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.437769930198844,
+      "learning_rate": 2.9822116124139045e-06,
+      "loss": 0.9437,
+      "step": 7870
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 6.109339166692898,
+      "learning_rate": 2.9810004858544194e-06,
+      "loss": 1.3969,
+      "step": 7871
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.323342271677376,
+      "learning_rate": 2.9797895008180135e-06,
+      "loss": 0.3465,
+      "step": 7872
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.320943319032637,
+      "learning_rate": 2.9785786573895713e-06,
+      "loss": 0.9443,
+      "step": 7873
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.368512254444603,
+      "learning_rate": 2.9773679556539696e-06,
+      "loss": 1.0702,
+      "step": 7874
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.100965499627617,
+      "learning_rate": 2.9761573956960706e-06,
+      "loss": 0.424,
+      "step": 7875
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.743011989969749,
+      "learning_rate": 2.9749469776007324e-06,
+      "loss": 0.9294,
+      "step": 7876
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.097132767859566,
+      "learning_rate": 2.973736701452795e-06,
+      "loss": 1.146,
+      "step": 7877
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.7628372451622654,
+      "learning_rate": 2.9725265673370973e-06,
+      "loss": 0.8206,
+      "step": 7878
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.387763699050225,
+      "learning_rate": 2.971316575338464e-06,
+      "loss": 1.0001,
+      "step": 7879
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.202384180316912,
+      "learning_rate": 2.9701067255417092e-06,
+      "loss": 0.9383,
+      "step": 7880
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.36968601017581,
+      "learning_rate": 2.96889701803164e-06,
+      "loss": 0.4805,
+      "step": 7881
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.9970162535064464,
+      "learning_rate": 2.967687452893051e-06,
+      "loss": 0.4226,
+      "step": 7882
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.5169857613363527,
+      "learning_rate": 2.9664780302107266e-06,
+      "loss": 0.9481,
+      "step": 7883
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.560675619392749,
+      "learning_rate": 2.965268750069441e-06,
+      "loss": 0.5625,
+      "step": 7884
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.6906020056209887,
+      "learning_rate": 2.964059612553961e-06,
+      "loss": 0.5511,
+      "step": 7885
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.3707827503539995,
+      "learning_rate": 2.962850617749042e-06,
+      "loss": 0.9854,
+      "step": 7886
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.169952648374127,
+      "learning_rate": 2.961641765739429e-06,
+      "loss": 0.6199,
+      "step": 7887
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.658972696363917,
+      "learning_rate": 2.9604330566098588e-06,
+      "loss": 1.0271,
+      "step": 7888
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.5783860166921713,
+      "learning_rate": 2.9592244904450536e-06,
+      "loss": 0.6164,
+      "step": 7889
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.6210062336360935,
+      "learning_rate": 2.9580160673297307e-06,
+      "loss": 0.4776,
+      "step": 7890
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.656576655258052,
+      "learning_rate": 2.956807787348594e-06,
+      "loss": 1.2903,
+      "step": 7891
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.450910711609922,
+      "learning_rate": 2.9555996505863394e-06,
+      "loss": 0.779,
+      "step": 7892
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.602170961498176,
+      "learning_rate": 2.954391657127654e-06,
+      "loss": 0.9086,
+      "step": 7893
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.2815459596395264,
+      "learning_rate": 2.9531838070572084e-06,
+      "loss": 1.129,
+      "step": 7894
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.8505282976510635,
+      "learning_rate": 2.9519761004596708e-06,
+      "loss": 0.8004,
+      "step": 7895
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.8476698443942707,
+      "learning_rate": 2.9507685374196954e-06,
+      "loss": 0.7421,
+      "step": 7896
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.2051223760366647,
+      "learning_rate": 2.9495611180219287e-06,
+      "loss": 0.4676,
+      "step": 7897
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.059341674196565,
+      "learning_rate": 2.948353842351002e-06,
+      "loss": 0.7832,
+      "step": 7898
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.376156592708591,
+      "learning_rate": 2.947146710491545e-06,
+      "loss": 0.9234,
+      "step": 7899
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.525761659169275,
+      "learning_rate": 2.9459397225281673e-06,
+      "loss": 0.8849,
+      "step": 7900
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.880761815644048,
+      "learning_rate": 2.9447328785454752e-06,
+      "loss": 0.9877,
+      "step": 7901
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.1428597004178456,
+      "learning_rate": 2.9435261786280645e-06,
+      "loss": 0.509,
+      "step": 7902
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.394815004925326,
+      "learning_rate": 2.942319622860519e-06,
+      "loss": 0.363,
+      "step": 7903
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.4582482783216015,
+      "learning_rate": 2.9411132113274132e-06,
+      "loss": 0.8826,
+      "step": 7904
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.753784692097969,
+      "learning_rate": 2.9399069441133116e-06,
+      "loss": 0.5,
+      "step": 7905
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.0731841609712225,
+      "learning_rate": 2.9387008213027675e-06,
+      "loss": 0.6792,
+      "step": 7906
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.9450796550415745,
+      "learning_rate": 2.937494842980324e-06,
+      "loss": 0.7015,
+      "step": 7907
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1214717249721415,
+      "learning_rate": 2.9362890092305158e-06,
+      "loss": 0.1444,
+      "step": 7908
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.938371891994917,
+      "learning_rate": 2.935083320137867e-06,
+      "loss": 0.9034,
+      "step": 7909
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.9877577235309896,
+      "learning_rate": 2.9338777757868923e-06,
+      "loss": 0.8414,
+      "step": 7910
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.6328141567571284,
+      "learning_rate": 2.9326723762620924e-06,
+      "loss": 0.4078,
+      "step": 7911
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.990145345446904,
+      "learning_rate": 2.931467121647962e-06,
+      "loss": 0.7359,
+      "step": 7912
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.807798979886192,
+      "learning_rate": 2.930262012028984e-06,
+      "loss": 0.8907,
+      "step": 7913
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.8619260709390737,
+      "learning_rate": 2.929057047489632e-06,
+      "loss": 0.7976,
+      "step": 7914
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.245285559043202,
+      "learning_rate": 2.9278522281143667e-06,
+      "loss": 0.681,
+      "step": 7915
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.452101166075296,
+      "learning_rate": 2.9266475539876447e-06,
+      "loss": 0.5975,
+      "step": 7916
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.729503388721803,
+      "learning_rate": 2.9254430251939046e-06,
+      "loss": 1.127,
+      "step": 7917
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.718684875541518,
+      "learning_rate": 2.9242386418175793e-06,
+      "loss": 0.9964,
+      "step": 7918
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.6566710096133983,
+      "learning_rate": 2.9230344039430913e-06,
+      "loss": 0.8345,
+      "step": 7919
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0860241466730605,
+      "learning_rate": 2.921830311654853e-06,
+      "loss": 0.332,
+      "step": 7920
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.050894560130508,
+      "learning_rate": 2.9206263650372668e-06,
+      "loss": 0.5202,
+      "step": 7921
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.6740888860709515,
+      "learning_rate": 2.919422564174722e-06,
+      "loss": 0.7684,
+      "step": 7922
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.539719103044569,
+      "learning_rate": 2.9182189091516017e-06,
+      "loss": 1.2597,
+      "step": 7923
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.534927915504683,
+      "learning_rate": 2.9170154000522744e-06,
+      "loss": 0.3663,
+      "step": 7924
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.5698272821477635,
+      "learning_rate": 2.915812036961103e-06,
+      "loss": 1.0274,
+      "step": 7925
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.478769023476825,
+      "learning_rate": 2.914608819962437e-06,
+      "loss": 0.6595,
+      "step": 7926
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.068271121802564,
+      "learning_rate": 2.9134057491406163e-06,
+      "loss": 0.662,
+      "step": 7927
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 6.19661409649374,
+      "learning_rate": 2.912202824579975e-06,
+      "loss": 1.6517,
+      "step": 7928
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.4816319057866587,
+      "learning_rate": 2.911000046364827e-06,
+      "loss": 0.4724,
+      "step": 7929
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.149261359307092,
+      "learning_rate": 2.9097974145794843e-06,
+      "loss": 0.6429,
+      "step": 7930
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.424747973768166,
+      "learning_rate": 2.908594929308246e-06,
+      "loss": 0.45,
+      "step": 7931
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.8112624800692414,
+      "learning_rate": 2.907392590635404e-06,
+      "loss": 0.3496,
+      "step": 7932
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.0302301469921433,
+      "learning_rate": 2.9061903986452323e-06,
+      "loss": 0.6052,
+      "step": 7933
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.9909120510204357,
+      "learning_rate": 2.904988353422003e-06,
+      "loss": 0.6163,
+      "step": 7934
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.088865606963298,
+      "learning_rate": 2.9037864550499704e-06,
+      "loss": 0.9871,
+      "step": 7935
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.7197420214941594,
+      "learning_rate": 2.902584703613385e-06,
+      "loss": 1.0631,
+      "step": 7936
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.4037252525132544,
+      "learning_rate": 2.9013830991964838e-06,
+      "loss": 0.5726,
+      "step": 7937
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.86857257087212,
+      "learning_rate": 2.900181641883494e-06,
+      "loss": 1.2704,
+      "step": 7938
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.9548905996573103,
+      "learning_rate": 2.8989803317586353e-06,
+      "loss": 0.4708,
+      "step": 7939
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.606607452653022,
+      "learning_rate": 2.8977791689061087e-06,
+      "loss": 0.7239,
+      "step": 7940
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.19923764561386,
+      "learning_rate": 2.8965781534101132e-06,
+      "loss": 1.0878,
+      "step": 7941
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.4083004770293286,
+      "learning_rate": 2.895377285354836e-06,
+      "loss": 0.5902,
+      "step": 7942
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.7862425947863994,
+      "learning_rate": 2.8941765648244513e-06,
+      "loss": 0.7949,
+      "step": 7943
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.819961564630001,
+      "learning_rate": 2.892975991903125e-06,
+      "loss": 1.3285,
+      "step": 7944
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.505179163497323,
+      "learning_rate": 2.891775566675014e-06,
+      "loss": 1.2131,
+      "step": 7945
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.800564826454519,
+      "learning_rate": 2.8905752892242587e-06,
+      "loss": 1.0754,
+      "step": 7946
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.6659245553942985,
+      "learning_rate": 2.889375159634995e-06,
+      "loss": 0.2595,
+      "step": 7947
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.867679753509104,
+      "learning_rate": 2.8881751779913498e-06,
+      "loss": 0.8299,
+      "step": 7948
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.1805636058524747,
+      "learning_rate": 2.886975344377432e-06,
+      "loss": 0.7716,
+      "step": 7949
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.9457803081893306,
+      "learning_rate": 2.8857756588773457e-06,
+      "loss": 0.673,
+      "step": 7950
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.854688609728728,
+      "learning_rate": 2.884576121575187e-06,
+      "loss": 0.8473,
+      "step": 7951
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.55813344986194,
+      "learning_rate": 2.8833767325550345e-06,
+      "loss": 0.8929,
+      "step": 7952
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.062536876817694,
+      "learning_rate": 2.8821774919009605e-06,
+      "loss": 1.2235,
+      "step": 7953
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.91308594382761,
+      "learning_rate": 2.8809783996970274e-06,
+      "loss": 0.9671,
+      "step": 7954
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.705628943410485,
+      "learning_rate": 2.8797794560272875e-06,
+      "loss": 1.1877,
+      "step": 7955
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.095020489564123,
+      "learning_rate": 2.8785806609757815e-06,
+      "loss": 0.8418,
+      "step": 7956
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.12676106657002,
+      "learning_rate": 2.8773820146265375e-06,
+      "loss": 1.25,
+      "step": 7957
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.389999860397088,
+      "learning_rate": 2.8761835170635765e-06,
+      "loss": 0.8575,
+      "step": 7958
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.4720159253229315,
+      "learning_rate": 2.8749851683709072e-06,
+      "loss": 0.6071,
+      "step": 7959
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.980189054467754,
+      "learning_rate": 2.8737869686325304e-06,
+      "loss": 0.4888,
+      "step": 7960
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.750422703747019,
+      "learning_rate": 2.872588917932434e-06,
+      "loss": 0.8012,
+      "step": 7961
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.50027378894247,
+      "learning_rate": 2.871391016354597e-06,
+      "loss": 0.9064,
+      "step": 7962
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.299368076828362,
+      "learning_rate": 2.8701932639829846e-06,
+      "loss": 0.5323,
+      "step": 7963
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.9160283537957787,
+      "learning_rate": 2.868995660901557e-06,
+      "loss": 1.0933,
+      "step": 7964
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.6159973575585713,
+      "learning_rate": 2.867798207194258e-06,
+      "loss": 0.4785,
+      "step": 7965
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.854649565232945,
+      "learning_rate": 2.866600902945025e-06,
+      "loss": 0.2629,
+      "step": 7966
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.442187897459728,
+      "learning_rate": 2.865403748237784e-06,
+      "loss": 0.6322,
+      "step": 7967
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.0672308602635106,
+      "learning_rate": 2.864206743156453e-06,
+      "loss": 0.5079,
+      "step": 7968
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.447223811256291,
+      "learning_rate": 2.8630098877849322e-06,
+      "loss": 0.6896,
+      "step": 7969
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.772683191649797,
+      "learning_rate": 2.861813182207117e-06,
+      "loss": 1.1045,
+      "step": 7970
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.90869334797886,
+      "learning_rate": 2.8606166265068935e-06,
+      "loss": 1.2193,
+      "step": 7971
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.66447128756878,
+      "learning_rate": 2.8594202207681333e-06,
+      "loss": 0.9918,
+      "step": 7972
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.1709142790837173,
+      "learning_rate": 2.8582239650747024e-06,
+      "loss": 0.3579,
+      "step": 7973
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.542718894700696,
+      "learning_rate": 2.8570278595104478e-06,
+      "loss": 0.4313,
+      "step": 7974
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.470565752435574,
+      "learning_rate": 2.855831904159214e-06,
+      "loss": 1.3507,
+      "step": 7975
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.017187077630708,
+      "learning_rate": 2.8546360991048325e-06,
+      "loss": 0.4235,
+      "step": 7976
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.411440865234151,
+      "learning_rate": 2.8534404444311235e-06,
+      "loss": 0.6309,
+      "step": 7977
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.200528672997894,
+      "learning_rate": 2.8522449402218984e-06,
+      "loss": 1.0034,
+      "step": 7978
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.373192243595398,
+      "learning_rate": 2.8510495865609573e-06,
+      "loss": 0.6623,
+      "step": 7979
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.968360904913381,
+      "learning_rate": 2.8498543835320856e-06,
+      "loss": 1.2938,
+      "step": 7980
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.224467989703306,
+      "learning_rate": 2.8486593312190668e-06,
+      "loss": 0.5887,
+      "step": 7981
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.654242604119693,
+      "learning_rate": 2.8474644297056643e-06,
+      "loss": 0.4886,
+      "step": 7982
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.7602370743626894,
+      "learning_rate": 2.8462696790756362e-06,
+      "loss": 0.6524,
+      "step": 7983
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.902133042146245,
+      "learning_rate": 2.845075079412731e-06,
+      "loss": 0.9172,
+      "step": 7984
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.455886719214672,
+      "learning_rate": 2.8438806308006874e-06,
+      "loss": 1.2014,
+      "step": 7985
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.8729653038084955,
+      "learning_rate": 2.842686333323226e-06,
+      "loss": 0.5934,
+      "step": 7986
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.189609924684086,
+      "learning_rate": 2.841492187064063e-06,
+      "loss": 0.3625,
+      "step": 7987
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.3526536316346363,
+      "learning_rate": 2.8402981921069044e-06,
+      "loss": 0.4967,
+      "step": 7988
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.623895560107847,
+      "learning_rate": 2.8391043485354436e-06,
+      "loss": 0.5459,
+      "step": 7989
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.154512611660718,
+      "learning_rate": 2.8379106564333637e-06,
+      "loss": 0.4984,
+      "step": 7990
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.224795514043236,
+      "learning_rate": 2.8367171158843386e-06,
+      "loss": 0.1345,
+      "step": 7991
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.1610715773663984,
+      "learning_rate": 2.835523726972028e-06,
+      "loss": 0.6905,
+      "step": 7992
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.3696818259978287,
+      "learning_rate": 2.834330489780084e-06,
+      "loss": 0.7145,
+      "step": 7993
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.2641499570271626,
+      "learning_rate": 2.8331374043921472e-06,
+      "loss": 0.7458,
+      "step": 7994
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.941652968343014,
+      "learning_rate": 2.831944470891851e-06,
+      "loss": 1.0364,
+      "step": 7995
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.5782653321947455,
+      "learning_rate": 2.8307516893628097e-06,
+      "loss": 1.2017,
+      "step": 7996
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.768539344559609,
+      "learning_rate": 2.8295590598886356e-06,
+      "loss": 0.9792,
+      "step": 7997
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.882226709861375,
+      "learning_rate": 2.828366582552924e-06,
+      "loss": 0.7185,
+      "step": 7998
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.30969061938397,
+      "learning_rate": 2.827174257439265e-06,
+      "loss": 0.8603,
+      "step": 7999
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.865148421794005,
+      "learning_rate": 2.8259820846312326e-06,
+      "loss": 0.6394,
+      "step": 8000
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.216494413050757,
+      "learning_rate": 2.824790064212396e-06,
+      "loss": 0.4633,
+      "step": 8001
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.186243312675972,
+      "learning_rate": 2.8235981962663107e-06,
+      "loss": 0.5483,
+      "step": 8002
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.440202878375771,
+      "learning_rate": 2.8224064808765182e-06,
+      "loss": 0.9243,
+      "step": 8003
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.79307107805714,
+      "learning_rate": 2.8212149181265547e-06,
+      "loss": 1.046,
+      "step": 8004
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.234343446088011,
+      "learning_rate": 2.820023508099944e-06,
+      "loss": 0.5725,
+      "step": 8005
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.7712916691735443,
+      "learning_rate": 2.8188322508801967e-06,
+      "loss": 0.9586,
+      "step": 8006
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.110195433030602,
+      "learning_rate": 2.817641146550817e-06,
+      "loss": 0.8975,
+      "step": 8007
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.564456190651159,
+      "learning_rate": 2.8164501951952973e-06,
+      "loss": 0.7762,
+      "step": 8008
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.222844540856964,
+      "learning_rate": 2.8152593968971143e-06,
+      "loss": 0.7783,
+      "step": 8009
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.2829279961044495,
+      "learning_rate": 2.814068751739739e-06,
+      "loss": 0.8104,
+      "step": 8010
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.6984117927456697,
+      "learning_rate": 2.8128782598066327e-06,
+      "loss": 0.4785,
+      "step": 8011
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.0633281365883747,
+      "learning_rate": 2.8116879211812407e-06,
+      "loss": 0.7631,
+      "step": 8012
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.632535475645752,
+      "learning_rate": 2.810497735947003e-06,
+      "loss": 0.5679,
+      "step": 8013
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.8235228300560817,
+      "learning_rate": 2.809307704187344e-06,
+      "loss": 0.3668,
+      "step": 8014
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.161518638795988,
+      "learning_rate": 2.8081178259856813e-06,
+      "loss": 1.2164,
+      "step": 8015
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.788168760095346,
+      "learning_rate": 2.806928101425419e-06,
+      "loss": 0.748,
+      "step": 8016
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.595830211979231,
+      "learning_rate": 2.8057385305899533e-06,
+      "loss": 0.7062,
+      "step": 8017
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.2491092651746767,
+      "learning_rate": 2.804549113562667e-06,
+      "loss": 0.7872,
+      "step": 8018
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.167759757082825,
+      "learning_rate": 2.803359850426935e-06,
+      "loss": 0.296,
+      "step": 8019
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.6383704732156836,
+      "learning_rate": 2.802170741266116e-06,
+      "loss": 0.8412,
+      "step": 8020
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.4452165564431327,
+      "learning_rate": 2.8009817861635622e-06,
+      "loss": 0.4422,
+      "step": 8021
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.406342192462903,
+      "learning_rate": 2.7997929852026164e-06,
+      "loss": 1.0347,
+      "step": 8022
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.505017668203887,
+      "learning_rate": 2.7986043384666055e-06,
+      "loss": 0.4232,
+      "step": 8023
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.8697517932358667,
+      "learning_rate": 2.797415846038851e-06,
+      "loss": 0.4334,
+      "step": 8024
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.7275342691111857,
+      "learning_rate": 2.7962275080026612e-06,
+      "loss": 0.6212,
+      "step": 8025
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.908361652822283,
+      "learning_rate": 2.795039324441331e-06,
+      "loss": 0.8908,
+      "step": 8026
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.727079111363858,
+      "learning_rate": 2.7938512954381503e-06,
+      "loss": 1.0258,
+      "step": 8027
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.144923134552206,
+      "learning_rate": 2.79266342107639e-06,
+      "loss": 0.6403,
+      "step": 8028
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.8843594757838744,
+      "learning_rate": 2.791475701439317e-06,
+      "loss": 0.3594,
+      "step": 8029
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.49582032249646,
+      "learning_rate": 2.790288136610187e-06,
+      "loss": 1.0168,
+      "step": 8030
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.0150535405758636,
+      "learning_rate": 2.7891007266722435e-06,
+      "loss": 0.7466,
+      "step": 8031
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.821039224362037,
+      "learning_rate": 2.787913471708715e-06,
+      "loss": 0.7878,
+      "step": 8032
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.5964597099929008,
+      "learning_rate": 2.7867263718028246e-06,
+      "loss": 0.8349,
+      "step": 8033
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.215248501926799,
+      "learning_rate": 2.7855394270377843e-06,
+      "loss": 0.7862,
+      "step": 8034
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.023413160171227,
+      "learning_rate": 2.784352637496792e-06,
+      "loss": 1.2161,
+      "step": 8035
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.616951739827295,
+      "learning_rate": 2.7831660032630405e-06,
+      "loss": 0.7639,
+      "step": 8036
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.343166757027107,
+      "learning_rate": 2.781979524419701e-06,
+      "loss": 0.9321,
+      "step": 8037
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.542165060427963,
+      "learning_rate": 2.780793201049945e-06,
+      "loss": 0.6174,
+      "step": 8038
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.00732678304251,
+      "learning_rate": 2.7796070332369274e-06,
+      "loss": 1.2496,
+      "step": 8039
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.386606258896781,
+      "learning_rate": 2.7784210210637937e-06,
+      "loss": 1.5798,
+      "step": 8040
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.456299499160534,
+      "learning_rate": 2.7772351646136795e-06,
+      "loss": 0.6512,
+      "step": 8041
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.750388118818402,
+      "learning_rate": 2.776049463969705e-06,
+      "loss": 1.2437,
+      "step": 8042
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.1606977279478032,
+      "learning_rate": 2.7748639192149863e-06,
+      "loss": 0.3467,
+      "step": 8043
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8466243448013244,
+      "learning_rate": 2.7736785304326217e-06,
+      "loss": 0.3452,
+      "step": 8044
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.14618056656971,
+      "learning_rate": 2.772493297705703e-06,
+      "loss": 0.7901,
+      "step": 8045
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.886642021842415,
+      "learning_rate": 2.771308221117309e-06,
+      "loss": 1.0205,
+      "step": 8046
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.977361372003828,
+      "learning_rate": 2.7701233007505104e-06,
+      "loss": 0.6876,
+      "step": 8047
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.668384948430321,
+      "learning_rate": 2.7689385366883654e-06,
+      "loss": 0.3277,
+      "step": 8048
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.426996210231428,
+      "learning_rate": 2.7677539290139177e-06,
+      "loss": 1.0532,
+      "step": 8049
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.458992091406994,
+      "learning_rate": 2.766569477810205e-06,
+      "loss": 0.9114,
+      "step": 8050
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.174900594278151,
+      "learning_rate": 2.7653851831602514e-06,
+      "loss": 0.7051,
+      "step": 8051
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.8040453815644266,
+      "learning_rate": 2.764201045147071e-06,
+      "loss": 0.8327,
+      "step": 8052
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.293419797368659,
+      "learning_rate": 2.76301706385367e-06,
+      "loss": 1.0003,
+      "step": 8053
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.393621564583542,
+      "learning_rate": 2.7618332393630353e-06,
+      "loss": 0.6693,
+      "step": 8054
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.725986573141455,
+      "learning_rate": 2.7606495717581498e-06,
+      "loss": 0.7114,
+      "step": 8055
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.669401274710586,
+      "learning_rate": 2.7594660611219838e-06,
+      "loss": 1.4499,
+      "step": 8056
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.61047399172766,
+      "learning_rate": 2.7582827075374987e-06,
+      "loss": 0.9271,
+      "step": 8057
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.410563440010257,
+      "learning_rate": 2.7570995110876364e-06,
+      "loss": 1.7803,
+      "step": 8058
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.106155063712242,
+      "learning_rate": 2.75591647185534e-06,
+      "loss": 0.3749,
+      "step": 8059
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.6487211232728236,
+      "learning_rate": 2.7547335899235304e-06,
+      "loss": 0.6993,
+      "step": 8060
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.1968521220751,
+      "learning_rate": 2.7535508653751252e-06,
+      "loss": 0.5003,
+      "step": 8061
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.1849582457043737,
+      "learning_rate": 2.7523682982930278e-06,
+      "loss": 0.4475,
+      "step": 8062
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.046657932174976,
+      "learning_rate": 2.7511858887601304e-06,
+      "loss": 0.7241,
+      "step": 8063
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.054131823721945,
+      "learning_rate": 2.7500036368593153e-06,
+      "loss": 1.1251,
+      "step": 8064
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.0832769446657418,
+      "learning_rate": 2.7488215426734554e-06,
+      "loss": 0.6398,
+      "step": 8065
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.3105219144880476,
+      "learning_rate": 2.7476396062854065e-06,
+      "loss": 0.5005,
+      "step": 8066
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.8711836856018884,
+      "learning_rate": 2.7464578277780187e-06,
+      "loss": 0.5241,
+      "step": 8067
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.941614254848235,
+      "learning_rate": 2.74527620723413e-06,
+      "loss": 0.7623,
+      "step": 8068
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.47330668948666,
+      "learning_rate": 2.7440947447365664e-06,
+      "loss": 0.9966,
+      "step": 8069
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.6291212023548796,
+      "learning_rate": 2.7429134403681435e-06,
+      "loss": 0.5474,
+      "step": 8070
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.157885256218511,
+      "learning_rate": 2.741732294211667e-06,
+      "loss": 1.1754,
+      "step": 8071
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.6000847904713003,
+      "learning_rate": 2.740551306349927e-06,
+      "loss": 0.9091,
+      "step": 8072
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8847891185750434,
+      "learning_rate": 2.739370476865707e-06,
+      "loss": 0.1354,
+      "step": 8073
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.179574267260189,
+      "learning_rate": 2.738189805841781e-06,
+      "loss": 0.4367,
+      "step": 8074
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.54789867852281,
+      "learning_rate": 2.7370092933609037e-06,
+      "loss": 1.2261,
+      "step": 8075
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.0687699554112644,
+      "learning_rate": 2.7358289395058284e-06,
+      "loss": 0.5557,
+      "step": 8076
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.7615195258812313,
+      "learning_rate": 2.7346487443592888e-06,
+      "loss": 0.8794,
+      "step": 8077
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.876920388057226,
+      "learning_rate": 2.7334687080040134e-06,
+      "loss": 1.0369,
+      "step": 8078
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.344281397286893,
+      "learning_rate": 2.732288830522718e-06,
+      "loss": 0.9742,
+      "step": 8079
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.4912933701112414,
+      "learning_rate": 2.731109111998106e-06,
+      "loss": 0.6292,
+      "step": 8080
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.110955242715626,
+      "learning_rate": 2.729929552512871e-06,
+      "loss": 1.0262,
+      "step": 8081
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.7977212752673126,
+      "learning_rate": 2.7287501521496966e-06,
+      "loss": 0.7897,
+      "step": 8082
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.5688064240040456,
+      "learning_rate": 2.7275709109912506e-06,
+      "loss": 0.8668,
+      "step": 8083
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.3072778372473155,
+      "learning_rate": 2.726391829120194e-06,
+      "loss": 0.579,
+      "step": 8084
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.019471690935791,
+      "learning_rate": 2.7252129066191758e-06,
+      "loss": 0.6334,
+      "step": 8085
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1563780617412596,
+      "learning_rate": 2.7240341435708316e-06,
+      "loss": 0.1597,
+      "step": 8086
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.7132075006676404,
+      "learning_rate": 2.7228555400577904e-06,
+      "loss": 0.5548,
+      "step": 8087
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.412152437555208,
+      "learning_rate": 2.7216770961626672e-06,
+      "loss": 0.8515,
+      "step": 8088
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.37663916419196,
+      "learning_rate": 2.720498811968062e-06,
+      "loss": 0.6956,
+      "step": 8089
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.9745511389014134,
+      "learning_rate": 2.7193206875565715e-06,
+      "loss": 0.764,
+      "step": 8090
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.1630440192612665,
+      "learning_rate": 2.7181427230107738e-06,
+      "loss": 0.8398,
+      "step": 8091
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.285493124951717,
+      "learning_rate": 2.7169649184132403e-06,
+      "loss": 0.1758,
+      "step": 8092
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.303442121768274,
+      "learning_rate": 2.7157872738465317e-06,
+      "loss": 0.4663,
+      "step": 8093
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.4441901971587376,
+      "learning_rate": 2.714609789393193e-06,
+      "loss": 0.5931,
+      "step": 8094
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.705057991518503,
+      "learning_rate": 2.7134324651357625e-06,
+      "loss": 0.9862,
+      "step": 8095
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.168949543449048,
+      "learning_rate": 2.7122553011567636e-06,
+      "loss": 0.5788,
+      "step": 8096
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.597665183472927,
+      "learning_rate": 2.711078297538713e-06,
+      "loss": 0.8572,
+      "step": 8097
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.510929055765163,
+      "learning_rate": 2.7099014543641116e-06,
+      "loss": 0.8463,
+      "step": 8098
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.4589615554959883,
+      "learning_rate": 2.708724771715454e-06,
+      "loss": 0.3028,
+      "step": 8099
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1060126440401574,
+      "learning_rate": 2.707548249675216e-06,
+      "loss": 0.155,
+      "step": 8100
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.7680275308789666,
+      "learning_rate": 2.7063718883258694e-06,
+      "loss": 0.4032,
+      "step": 8101
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.683328697382099,
+      "learning_rate": 2.7051956877498707e-06,
+      "loss": 0.2433,
+      "step": 8102
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.528927015598015,
+      "learning_rate": 2.7040196480296677e-06,
+      "loss": 0.6996,
+      "step": 8103
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9414316471303987,
+      "learning_rate": 2.702843769247698e-06,
+      "loss": 0.2005,
+      "step": 8104
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.1785227594638092,
+      "learning_rate": 2.7016680514863796e-06,
+      "loss": 0.3247,
+      "step": 8105
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.559538548866543,
+      "learning_rate": 2.700492494828131e-06,
+      "loss": 0.8933,
+      "step": 8106
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.3675280676526933,
+      "learning_rate": 2.699317099355349e-06,
+      "loss": 0.5503,
+      "step": 8107
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.46769595659413,
+      "learning_rate": 2.6981418651504256e-06,
+      "loss": 0.6654,
+      "step": 8108
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3525915329662892,
+      "learning_rate": 2.69696679229574e-06,
+      "loss": 0.1927,
+      "step": 8109
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.8039875762312847,
+      "learning_rate": 2.6957918808736593e-06,
+      "loss": 0.7026,
+      "step": 8110
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.980857406008913,
+      "learning_rate": 2.6946171309665413e-06,
+      "loss": 0.7007,
+      "step": 8111
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.9225626645846536,
+      "learning_rate": 2.693442542656728e-06,
+      "loss": 0.6474,
+      "step": 8112
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.054818429499179,
+      "learning_rate": 2.692268116026554e-06,
+      "loss": 1.148,
+      "step": 8113
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.626438590520175,
+      "learning_rate": 2.6910938511583424e-06,
+      "loss": 0.7244,
+      "step": 8114
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.309158002175827,
+      "learning_rate": 2.689919748134403e-06,
+      "loss": 1.201,
+      "step": 8115
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.8702321853862456,
+      "learning_rate": 2.6887458070370374e-06,
+      "loss": 0.726,
+      "step": 8116
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.6861780001976747,
+      "learning_rate": 2.6875720279485305e-06,
+      "loss": 0.8203,
+      "step": 8117
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.616881851475864,
+      "learning_rate": 2.686398410951161e-06,
+      "loss": 0.5582,
+      "step": 8118
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.726228925977059,
+      "learning_rate": 2.685224956127194e-06,
+      "loss": 0.4601,
+      "step": 8119
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.5189700331903744,
+      "learning_rate": 2.684051663558884e-06,
+      "loss": 0.7785,
+      "step": 8120
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.1994821182522797,
+      "learning_rate": 2.6828785333284736e-06,
+      "loss": 0.5161,
+      "step": 8121
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.233334262619315,
+      "learning_rate": 2.6817055655181947e-06,
+      "loss": 0.1662,
+      "step": 8122
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.166189713882294,
+      "learning_rate": 2.6805327602102647e-06,
+      "loss": 1.0081,
+      "step": 8123
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.427378529153368,
+      "learning_rate": 2.6793601174868934e-06,
+      "loss": 0.8678,
+      "step": 8124
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.995741956743334,
+      "learning_rate": 2.678187637430279e-06,
+      "loss": 0.6496,
+      "step": 8125
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.462789535181405,
+      "learning_rate": 2.677015320122607e-06,
+      "loss": 0.4996,
+      "step": 8126
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.286132345182877,
+      "learning_rate": 2.6758431656460503e-06,
+      "loss": 1.2132,
+      "step": 8127
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.582607649074299,
+      "learning_rate": 2.6746711740827757e-06,
+      "loss": 0.9266,
+      "step": 8128
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.279407521043276,
+      "learning_rate": 2.6734993455149295e-06,
+      "loss": 0.2849,
+      "step": 8129
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.293848497282886,
+      "learning_rate": 2.6723276800246544e-06,
+      "loss": 0.7391,
+      "step": 8130
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.5846162274642515,
+      "learning_rate": 2.671156177694079e-06,
+      "loss": 0.8377,
+      "step": 8131
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.437207797111822,
+      "learning_rate": 2.6699848386053208e-06,
+      "loss": 1.0667,
+      "step": 8132
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.6627445024534078,
+      "learning_rate": 2.6688136628404858e-06,
+      "loss": 0.9905,
+      "step": 8133
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.202182827673672,
+      "learning_rate": 2.6676426504816666e-06,
+      "loss": 0.5172,
+      "step": 8134
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.655701150778445,
+      "learning_rate": 2.666471801610947e-06,
+      "loss": 0.7972,
+      "step": 8135
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.265227474862339,
+      "learning_rate": 2.665301116310401e-06,
+      "loss": 1.0035,
+      "step": 8136
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.2479087704532916,
+      "learning_rate": 2.664130594662083e-06,
+      "loss": 0.3387,
+      "step": 8137
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.905996753323291,
+      "learning_rate": 2.6629602367480456e-06,
+      "loss": 1.3322,
+      "step": 8138
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.489019225321688,
+      "learning_rate": 2.6617900426503267e-06,
+      "loss": 0.4995,
+      "step": 8139
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.9753243013472632,
+      "learning_rate": 2.6606200124509474e-06,
+      "loss": 0.9357,
+      "step": 8140
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.2508496008821526,
+      "learning_rate": 2.6594501462319243e-06,
+      "loss": 0.4631,
+      "step": 8141
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.6037522115894522,
+      "learning_rate": 2.658280444075261e-06,
+      "loss": 0.8874,
+      "step": 8142
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.9317164559359346,
+      "learning_rate": 2.657110906062946e-06,
+      "loss": 0.6426,
+      "step": 8143
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 6.017729186475728,
+      "learning_rate": 2.6559415322769604e-06,
+      "loss": 1.4262,
+      "step": 8144
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.8748166235299153,
+      "learning_rate": 2.654772322799274e-06,
+      "loss": 0.509,
+      "step": 8145
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.122497071149187,
+      "learning_rate": 2.65360327771184e-06,
+      "loss": 0.7123,
+      "step": 8146
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6791406594043676,
+      "learning_rate": 2.6524343970966036e-06,
+      "loss": 0.1717,
+      "step": 8147
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7615678659571652,
+      "learning_rate": 2.6512656810355e-06,
+      "loss": 0.3311,
+      "step": 8148
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.697669974612375,
+      "learning_rate": 2.6500971296104506e-06,
+      "loss": 0.84,
+      "step": 8149
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.7332577255496306,
+      "learning_rate": 2.648928742903367e-06,
+      "loss": 0.5106,
+      "step": 8150
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.330105445552753,
+      "learning_rate": 2.6477605209961453e-06,
+      "loss": 0.7977,
+      "step": 8151
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.589037364480364,
+      "learning_rate": 2.6465924639706753e-06,
+      "loss": 1.2426,
+      "step": 8152
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6558531337873237,
+      "learning_rate": 2.64542457190883e-06,
+      "loss": 0.1925,
+      "step": 8153
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.077278142721715,
+      "learning_rate": 2.6442568448924754e-06,
+      "loss": 0.4792,
+      "step": 8154
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.43354957949788,
+      "learning_rate": 2.6430892830034634e-06,
+      "loss": 0.8561,
+      "step": 8155
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.8153529256250165,
+      "learning_rate": 2.6419218863236374e-06,
+      "loss": 0.4848,
+      "step": 8156
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.555375178166355,
+      "learning_rate": 2.640754654934823e-06,
+      "loss": 0.4334,
+      "step": 8157
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.515588811786006,
+      "learning_rate": 2.6395875889188393e-06,
+      "loss": 1.1924,
+      "step": 8158
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.011717132887081,
+      "learning_rate": 2.638420688357493e-06,
+      "loss": 0.5709,
+      "step": 8159
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.4878893153708304,
+      "learning_rate": 2.6372539533325793e-06,
+      "loss": 0.546,
+      "step": 8160
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.902388199732581,
+      "learning_rate": 2.6360873839258804e-06,
+      "loss": 1.0235,
+      "step": 8161
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.194007613848852,
+      "learning_rate": 2.63492098021917e-06,
+      "loss": 0.7716,
+      "step": 8162
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.464355787218569,
+      "learning_rate": 2.6337547422942046e-06,
+      "loss": 0.4862,
+      "step": 8163
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.964077826087806,
+      "learning_rate": 2.6325886702327335e-06,
+      "loss": 1.1743,
+      "step": 8164
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.992992442416339,
+      "learning_rate": 2.6314227641164936e-06,
+      "loss": 1.2173,
+      "step": 8165
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.2534889852906685,
+      "learning_rate": 2.6302570240272118e-06,
+      "loss": 1.0937,
+      "step": 8166
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.036733498750401,
+      "learning_rate": 2.629091450046598e-06,
+      "loss": 0.3887,
+      "step": 8167
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.7368829772271814,
+      "learning_rate": 2.6279260422563567e-06,
+      "loss": 0.7874,
+      "step": 8168
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.411251795911712,
+      "learning_rate": 2.6267608007381745e-06,
+      "loss": 0.4679,
+      "step": 8169
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.112337241361299,
+      "learning_rate": 2.625595725573732e-06,
+      "loss": 0.5023,
+      "step": 8170
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.904542380357021,
+      "learning_rate": 2.6244308168446958e-06,
+      "loss": 0.6493,
+      "step": 8171
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.8234626865452395,
+      "learning_rate": 2.623266074632721e-06,
+      "loss": 0.5458,
+      "step": 8172
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.07468524187268,
+      "learning_rate": 2.622101499019453e-06,
+      "loss": 1.0779,
+      "step": 8173
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.548773180308887,
+      "learning_rate": 2.6209370900865183e-06,
+      "loss": 0.4463,
+      "step": 8174
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.957265224808206,
+      "learning_rate": 2.6197728479155403e-06,
+      "loss": 0.79,
+      "step": 8175
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9613969150310653,
+      "learning_rate": 2.618608772588127e-06,
+      "loss": 0.3682,
+      "step": 8176
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.452334592025262,
+      "learning_rate": 2.6174448641858744e-06,
+      "loss": 0.4594,
+      "step": 8177
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.301706376374276,
+      "learning_rate": 2.6162811227903683e-06,
+      "loss": 0.8597,
+      "step": 8178
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.402394813245053,
+      "learning_rate": 2.6151175484831835e-06,
+      "loss": 1.2955,
+      "step": 8179
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.030558375461062,
+      "learning_rate": 2.613954141345878e-06,
+      "loss": 0.63,
+      "step": 8180
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.090141052480777,
+      "learning_rate": 2.612790901460003e-06,
+      "loss": 0.8791,
+      "step": 8181
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.8795763058173005,
+      "learning_rate": 2.6116278289070963e-06,
+      "loss": 0.6026,
+      "step": 8182
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.105086730112901,
+      "learning_rate": 2.6104649237686864e-06,
+      "loss": 1.0118,
+      "step": 8183
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.711704806278075,
+      "learning_rate": 2.609302186126284e-06,
+      "loss": 0.324,
+      "step": 8184
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.544883390181114,
+      "learning_rate": 2.6081396160613957e-06,
+      "loss": 1.4639,
+      "step": 8185
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.220550620883011,
+      "learning_rate": 2.60697721365551e-06,
+      "loss": 0.5899,
+      "step": 8186
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.0355312131320495,
+      "learning_rate": 2.6058149789901066e-06,
+      "loss": 0.2735,
+      "step": 8187
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.8453055950699087,
+      "learning_rate": 2.6046529121466537e-06,
+      "loss": 0.6375,
+      "step": 8188
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.551515254155337,
+      "learning_rate": 2.6034910132066066e-06,
+      "loss": 0.675,
+      "step": 8189
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.613874741705259,
+      "learning_rate": 2.60232928225141e-06,
+      "loss": 0.5582,
+      "step": 8190
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.874853243116013,
+      "learning_rate": 2.6011677193624984e-06,
+      "loss": 1.1525,
+      "step": 8191
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.102358534071514,
+      "learning_rate": 2.6000063246212882e-06,
+      "loss": 0.68,
+      "step": 8192
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.009976735808661,
+      "learning_rate": 2.598845098109189e-06,
+      "loss": 1.2542,
+      "step": 8193
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.4149214680031434,
+      "learning_rate": 2.5976840399075987e-06,
+      "loss": 0.8226,
+      "step": 8194
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.9917850748077006,
+      "learning_rate": 2.5965231500979026e-06,
+      "loss": 0.5335,
+      "step": 8195
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.042737043474195,
+      "learning_rate": 2.595362428761476e-06,
+      "loss": 0.8692,
+      "step": 8196
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.079629908302849,
+      "learning_rate": 2.5942018759796756e-06,
+      "loss": 1.0781,
+      "step": 8197
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 6.623417473132643,
+      "learning_rate": 2.5930414918338542e-06,
+      "loss": 1.107,
+      "step": 8198
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.6628269852169915,
+      "learning_rate": 2.59188127640535e-06,
+      "loss": 1.1345,
+      "step": 8199
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.9365790682542925,
+      "learning_rate": 2.590721229775487e-06,
+      "loss": 0.7339,
+      "step": 8200
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.549101199921481,
+      "learning_rate": 2.589561352025581e-06,
+      "loss": 0.5565,
+      "step": 8201
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.8907557400194692,
+      "learning_rate": 2.5884016432369352e-06,
+      "loss": 0.827,
+      "step": 8202
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.923777107347025,
+      "learning_rate": 2.587242103490837e-06,
+      "loss": 0.958,
+      "step": 8203
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.0709012492772443,
+      "learning_rate": 2.586082732868567e-06,
+      "loss": 0.5223,
+      "step": 8204
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.845501551249635,
+      "learning_rate": 2.5849235314513923e-06,
+      "loss": 0.9612,
+      "step": 8205
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.085875781413251,
+      "learning_rate": 2.583764499320567e-06,
+      "loss": 0.64,
+      "step": 8206
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.994995047525389,
+      "learning_rate": 2.5826056365573356e-06,
+      "loss": 0.421,
+      "step": 8207
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.005993934062253,
+      "learning_rate": 2.58144694324293e-06,
+      "loss": 0.7486,
+      "step": 8208
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.7070171019910383,
+      "learning_rate": 2.5802884194585664e-06,
+      "loss": 0.4059,
+      "step": 8209
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.6311262154020905,
+      "learning_rate": 2.5791300652854536e-06,
+      "loss": 0.421,
+      "step": 8210
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.126244190997835,
+      "learning_rate": 2.5779718808047882e-06,
+      "loss": 0.8678,
+      "step": 8211
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.054898585744836,
+      "learning_rate": 2.576813866097753e-06,
+      "loss": 0.7063,
+      "step": 8212
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.269210773538143,
+      "learning_rate": 2.5756560212455216e-06,
+      "loss": 1.0215,
+      "step": 8213
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.403454994734814,
+      "learning_rate": 2.5744983463292504e-06,
+      "loss": 0.9039,
+      "step": 8214
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.943293988287031,
+      "learning_rate": 2.5733408414300914e-06,
+      "loss": 0.5114,
+      "step": 8215
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.0435797640128968,
+      "learning_rate": 2.5721835066291767e-06,
+      "loss": 0.2742,
+      "step": 8216
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.8488265573426546,
+      "learning_rate": 2.571026342007632e-06,
+      "loss": 0.9684,
+      "step": 8217
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.298377465252303,
+      "learning_rate": 2.5698693476465704e-06,
+      "loss": 0.3515,
+      "step": 8218
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.391490206326064,
+      "learning_rate": 2.568712523627093e-06,
+      "loss": 0.6765,
+      "step": 8219
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.595083739065445,
+      "learning_rate": 2.567555870030285e-06,
+      "loss": 1.0814,
+      "step": 8220
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.22699420619472,
+      "learning_rate": 2.5663993869372244e-06,
+      "loss": 0.8324,
+      "step": 8221
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.8398058751542443,
+      "learning_rate": 2.565243074428976e-06,
+      "loss": 0.7015,
+      "step": 8222
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.710784195102657,
+      "learning_rate": 2.5640869325865912e-06,
+      "loss": 0.4557,
+      "step": 8223
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.9712039669774226,
+      "learning_rate": 2.5629309614911123e-06,
+      "loss": 0.6181,
+      "step": 8224
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.573667204632561,
+      "learning_rate": 2.561775161223568e-06,
+      "loss": 1.0692,
+      "step": 8225
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.3745376434268177,
+      "learning_rate": 2.560619531864972e-06,
+      "loss": 0.6221,
+      "step": 8226
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.8847605733345594,
+      "learning_rate": 2.5594640734963306e-06,
+      "loss": 0.7908,
+      "step": 8227
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.16579864794284,
+      "learning_rate": 2.5583087861986365e-06,
+      "loss": 0.7152,
+      "step": 8228
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.332205382754697,
+      "learning_rate": 2.557153670052872e-06,
+      "loss": 0.1777,
+      "step": 8229
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.4438633903948213,
+      "learning_rate": 2.5559987251400024e-06,
+      "loss": 0.5551,
+      "step": 8230
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.765037699508793,
+      "learning_rate": 2.554843951540987e-06,
+      "loss": 0.1958,
+      "step": 8231
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.59795873464682,
+      "learning_rate": 2.553689349336769e-06,
+      "loss": 0.8692,
+      "step": 8232
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.002247911399786,
+      "learning_rate": 2.5525349186082793e-06,
+      "loss": 0.8691,
+      "step": 8233
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.9778972950174603,
+      "learning_rate": 2.551380659436441e-06,
+      "loss": 0.5852,
+      "step": 8234
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.141822415205915,
+      "learning_rate": 2.550226571902162e-06,
+      "loss": 0.4145,
+      "step": 8235
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.8667126450551237,
+      "learning_rate": 2.549072656086341e-06,
+      "loss": 0.6026,
+      "step": 8236
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.996300879810245,
+      "learning_rate": 2.547918912069859e-06,
+      "loss": 0.9623,
+      "step": 8237
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.6904477810105543,
+      "learning_rate": 2.546765339933589e-06,
+      "loss": 0.7617,
+      "step": 8238
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.581868726425581,
+      "learning_rate": 2.5456119397583923e-06,
+      "loss": 0.853,
+      "step": 8239
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 6.0197529089066615,
+      "learning_rate": 2.544458711625117e-06,
+      "loss": 1.4534,
+      "step": 8240
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.83866500070881,
+      "learning_rate": 2.5433056556145996e-06,
+      "loss": 0.9129,
+      "step": 8241
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.852122294853424,
+      "learning_rate": 2.5421527718076657e-06,
+      "loss": 0.7918,
+      "step": 8242
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.895027520673276,
+      "learning_rate": 2.541000060285125e-06,
+      "loss": 0.7276,
+      "step": 8243
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.2427021641886995,
+      "learning_rate": 2.539847521127777e-06,
+      "loss": 0.5534,
+      "step": 8244
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.890394714776982,
+      "learning_rate": 2.538695154416414e-06,
+      "loss": 0.2905,
+      "step": 8245
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.208652062064113,
+      "learning_rate": 2.537542960231807e-06,
+      "loss": 1.1838,
+      "step": 8246
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.599313100308787,
+      "learning_rate": 2.536390938654722e-06,
+      "loss": 1.379,
+      "step": 8247
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.5905100343439336,
+      "learning_rate": 2.535239089765912e-06,
+      "loss": 0.6203,
+      "step": 8248
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.955544128632737,
+      "learning_rate": 2.5340874136461138e-06,
+      "loss": 0.8615,
+      "step": 8249
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.1298616750007944,
+      "learning_rate": 2.5329359103760555e-06,
+      "loss": 0.5261,
+      "step": 8250
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.328799199804836,
+      "learning_rate": 2.5317845800364538e-06,
+      "loss": 1.198,
+      "step": 8251
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3281557786608587,
+      "learning_rate": 2.530633422708011e-06,
+      "loss": 0.1665,
+      "step": 8252
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.7688107112634257,
+      "learning_rate": 2.529482438471421e-06,
+      "loss": 0.6611,
+      "step": 8253
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 6.398786157877805,
+      "learning_rate": 2.5283316274073577e-06,
+      "loss": 1.1927,
+      "step": 8254
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.653815342421929,
+      "learning_rate": 2.527180989596491e-06,
+      "loss": 1.0186,
+      "step": 8255
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.075590632175973,
+      "learning_rate": 2.526030525119475e-06,
+      "loss": 0.6921,
+      "step": 8256
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.855568441119522,
+      "learning_rate": 2.524880234056952e-06,
+      "loss": 1.2602,
+      "step": 8257
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.687147851686595,
+      "learning_rate": 2.5237301164895538e-06,
+      "loss": 0.6533,
+      "step": 8258
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.554934045264841,
+      "learning_rate": 2.5225801724978994e-06,
+      "loss": 0.7608,
+      "step": 8259
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.047030105038087,
+      "learning_rate": 2.5214304021625906e-06,
+      "loss": 0.8476,
+      "step": 8260
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.319276269283469,
+      "learning_rate": 2.5202808055642264e-06,
+      "loss": 0.2347,
+      "step": 8261
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.030062028737353,
+      "learning_rate": 2.5191313827833834e-06,
+      "loss": 0.6413,
+      "step": 8262
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.86433153661269,
+      "learning_rate": 2.517982133900634e-06,
+      "loss": 0.6122,
+      "step": 8263
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.360572337027347,
+      "learning_rate": 2.5168330589965356e-06,
+      "loss": 0.8149,
+      "step": 8264
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.517862354534247,
+      "learning_rate": 2.5156841581516344e-06,
+      "loss": 0.7741,
+      "step": 8265
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.048209300032961,
+      "learning_rate": 2.5145354314464606e-06,
+      "loss": 1.0705,
+      "step": 8266
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.9624911465341786,
+      "learning_rate": 2.5133868789615357e-06,
+      "loss": 1.0833,
+      "step": 8267
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.4665013356509604,
+      "learning_rate": 2.5122385007773685e-06,
+      "loss": 0.337,
+      "step": 8268
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.2481453941790934,
+      "learning_rate": 2.5110902969744567e-06,
+      "loss": 0.4088,
+      "step": 8269
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.853603954957759,
+      "learning_rate": 2.5099422676332825e-06,
+      "loss": 0.3283,
+      "step": 8270
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.159783771639473,
+      "learning_rate": 2.508794412834321e-06,
+      "loss": 1.1678,
+      "step": 8271
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.484698844000725,
+      "learning_rate": 2.507646732658027e-06,
+      "loss": 0.9776,
+      "step": 8272
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.201518136585807,
+      "learning_rate": 2.5064992271848504e-06,
+      "loss": 0.5837,
+      "step": 8273
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.9111991079769775,
+      "learning_rate": 2.505351896495226e-06,
+      "loss": 0.7384,
+      "step": 8274
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.718743486607333,
+      "learning_rate": 2.504204740669579e-06,
+      "loss": 0.7954,
+      "step": 8275
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.008261239270777,
+      "learning_rate": 2.5030577597883166e-06,
+      "loss": 0.6152,
+      "step": 8276
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 7.10932136638636,
+      "learning_rate": 2.5019109539318374e-06,
+      "loss": 1.3207,
+      "step": 8277
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.8281334738196775,
+      "learning_rate": 2.5007643231805316e-06,
+      "loss": 0.5473,
+      "step": 8278
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.88466449570412,
+      "learning_rate": 2.499617867614768e-06,
+      "loss": 0.5475,
+      "step": 8279
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.86738131822349,
+      "learning_rate": 2.4984715873149097e-06,
+      "loss": 0.9485,
+      "step": 8280
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.210014976007869,
+      "learning_rate": 2.497325482361307e-06,
+      "loss": 0.8218,
+      "step": 8281
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.485578773064971,
+      "learning_rate": 2.4961795528342977e-06,
+      "loss": 0.6658,
+      "step": 8282
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.424818209340035,
+      "learning_rate": 2.495033798814203e-06,
+      "loss": 1.1699,
+      "step": 8283
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.1328923324244364,
+      "learning_rate": 2.493888220381338e-06,
+      "loss": 0.3901,
+      "step": 8284
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.107847943479869,
+      "learning_rate": 2.492742817616002e-06,
+      "loss": 1.0158,
+      "step": 8285
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.08409785382572,
+      "learning_rate": 2.4915975905984825e-06,
+      "loss": 0.5241,
+      "step": 8286
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.605708211617868,
+      "learning_rate": 2.490452539409055e-06,
+      "loss": 0.6845,
+      "step": 8287
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.7541933540948054,
+      "learning_rate": 2.4893076641279857e-06,
+      "loss": 0.7459,
+      "step": 8288
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.325334689993187,
+      "learning_rate": 2.4881629648355197e-06,
+      "loss": 0.7375,
+      "step": 8289
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.571401693590738,
+      "learning_rate": 2.487018441611899e-06,
+      "loss": 1.1225,
+      "step": 8290
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.972920543433598,
+      "learning_rate": 2.485874094537349e-06,
+      "loss": 1.2411,
+      "step": 8291
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.8754292288507934,
+      "learning_rate": 2.484729923692085e-06,
+      "loss": 0.6286,
+      "step": 8292
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3322231779253642,
+      "learning_rate": 2.4835859291563054e-06,
+      "loss": 0.2827,
+      "step": 8293
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.1227954724856115,
+      "learning_rate": 2.4824421110102022e-06,
+      "loss": 0.7501,
+      "step": 8294
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.5513355488999294,
+      "learning_rate": 2.481298469333949e-06,
+      "loss": 0.5861,
+      "step": 8295
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.4934041611030273,
+      "learning_rate": 2.4801550042077118e-06,
+      "loss": 0.8104,
+      "step": 8296
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.914640469913183,
+      "learning_rate": 2.4790117157116417e-06,
+      "loss": 1.4283,
+      "step": 8297
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.321212069371395,
+      "learning_rate": 2.477868603925879e-06,
+      "loss": 0.7995,
+      "step": 8298
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.9954868489166917,
+      "learning_rate": 2.4767256689305537e-06,
+      "loss": 0.8571,
+      "step": 8299
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.999989814980641,
+      "learning_rate": 2.475582910805775e-06,
+      "loss": 0.6888,
+      "step": 8300
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.4585531621945518,
+      "learning_rate": 2.474440329631648e-06,
+      "loss": 0.7443,
+      "step": 8301
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.4130655659528393,
+      "learning_rate": 2.473297925488263e-06,
+      "loss": 0.8947,
+      "step": 8302
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 7.689007018264186,
+      "learning_rate": 2.4721556984556968e-06,
+      "loss": 1.1999,
+      "step": 8303
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.470032069770875,
+      "learning_rate": 2.471013648614015e-06,
+      "loss": 1.0237,
+      "step": 8304
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.3840771568384413,
+      "learning_rate": 2.4698717760432723e-06,
+      "loss": 0.3854,
+      "step": 8305
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.5984289964252936,
+      "learning_rate": 2.468730080823505e-06,
+      "loss": 0.5884,
+      "step": 8306
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.580650107432294,
+      "learning_rate": 2.4675885630347423e-06,
+      "loss": 0.4019,
+      "step": 8307
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.814392672951054,
+      "learning_rate": 2.466447222757003e-06,
+      "loss": 1.1511,
+      "step": 8308
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.8923529634070535,
+      "learning_rate": 2.465306060070285e-06,
+      "loss": 0.7966,
+      "step": 8309
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.927392623000533,
+      "learning_rate": 2.4641650750545816e-06,
+      "loss": 0.4876,
+      "step": 8310
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.713209401249808,
+      "learning_rate": 2.4630242677898718e-06,
+      "loss": 0.5281,
+      "step": 8311
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.069931845338178,
+      "learning_rate": 2.461883638356118e-06,
+      "loss": 1.0768,
+      "step": 8312
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.043007353870255,
+      "learning_rate": 2.4607431868332756e-06,
+      "loss": 0.7361,
+      "step": 8313
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.543555884846014,
+      "learning_rate": 2.4596029133012845e-06,
+      "loss": 0.5821,
+      "step": 8314
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.6930194005161634,
+      "learning_rate": 2.4584628178400737e-06,
+      "loss": 0.4279,
+      "step": 8315
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.3378103611275263,
+      "learning_rate": 2.45732290052956e-06,
+      "loss": 0.591,
+      "step": 8316
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.6515986206695095,
+      "learning_rate": 2.456183161449644e-06,
+      "loss": 0.696,
+      "step": 8317
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.368809991491645,
+      "learning_rate": 2.455043600680217e-06,
+      "loss": 0.237,
+      "step": 8318
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.278257713890669,
+      "learning_rate": 2.4539042183011585e-06,
+      "loss": 1.0608,
+      "step": 8319
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.390456938320225,
+      "learning_rate": 2.4527650143923334e-06,
+      "loss": 0.5158,
+      "step": 8320
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.320303803659349,
+      "learning_rate": 2.4516259890335947e-06,
+      "loss": 0.322,
+      "step": 8321
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.9903038975532277,
+      "learning_rate": 2.450487142304786e-06,
+      "loss": 0.8123,
+      "step": 8322
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.1891243338684485,
+      "learning_rate": 2.4493484742857316e-06,
+      "loss": 0.4653,
+      "step": 8323
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.5908340128294984,
+      "learning_rate": 2.4482099850562496e-06,
+      "loss": 0.6931,
+      "step": 8324
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.3660223577689057,
+      "learning_rate": 2.447071674696141e-06,
+      "loss": 0.4745,
+      "step": 8325
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.8100675807839512,
+      "learning_rate": 2.4459335432851977e-06,
+      "loss": 0.9717,
+      "step": 8326
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.339816506479468,
+      "learning_rate": 2.4447955909031973e-06,
+      "loss": 1.1936,
+      "step": 8327
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.270165734853613,
+      "learning_rate": 2.443657817629908e-06,
+      "loss": 1.266,
+      "step": 8328
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1624670879267303,
+      "learning_rate": 2.442520223545078e-06,
+      "loss": 0.1972,
+      "step": 8329
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.0678233836269317,
+      "learning_rate": 2.4413828087284504e-06,
+      "loss": 0.2616,
+      "step": 8330
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.583266213174848,
+      "learning_rate": 2.440245573259753e-06,
+      "loss": 0.9658,
+      "step": 8331
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.120895013381585,
+      "learning_rate": 2.4391085172187005e-06,
+      "loss": 0.6695,
+      "step": 8332
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.680942057669382,
+      "learning_rate": 2.437971640684998e-06,
+      "loss": 0.5309,
+      "step": 8333
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.200500902511551,
+      "learning_rate": 2.4368349437383314e-06,
+      "loss": 1.583,
+      "step": 8334
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.202678182159394,
+      "learning_rate": 2.4356984264583806e-06,
+      "loss": 0.8447,
+      "step": 8335
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.9494206946055037,
+      "learning_rate": 2.43456208892481e-06,
+      "loss": 0.5021,
+      "step": 8336
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2127148178284841,
+      "learning_rate": 2.433425931217272e-06,
+      "loss": 0.1644,
+      "step": 8337
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.0608201347588615,
+      "learning_rate": 2.4322899534154085e-06,
+      "loss": 0.9022,
+      "step": 8338
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.657162564141952,
+      "learning_rate": 2.4311541555988433e-06,
+      "loss": 0.8073,
+      "step": 8339
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.011262415440675,
+      "learning_rate": 2.430018537847193e-06,
+      "loss": 0.4338,
+      "step": 8340
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.61578686555714,
+      "learning_rate": 2.4288831002400574e-06,
+      "loss": 1.118,
+      "step": 8341
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.3072127189866585,
+      "learning_rate": 2.427747842857027e-06,
+      "loss": 0.9043,
+      "step": 8342
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.891045801602025,
+      "learning_rate": 2.4266127657776777e-06,
+      "loss": 0.6071,
+      "step": 8343
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.199359033982037,
+      "learning_rate": 2.4254778690815743e-06,
+      "loss": 0.8963,
+      "step": 8344
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.821403285387405,
+      "learning_rate": 2.42434315284827e-06,
+      "loss": 0.7843,
+      "step": 8345
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.576793763179655,
+      "learning_rate": 2.4232086171572993e-06,
+      "loss": 0.3895,
+      "step": 8346
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.1108494663998565,
+      "learning_rate": 2.4220742620881906e-06,
+      "loss": 0.9387,
+      "step": 8347
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.403348312465952,
+      "learning_rate": 2.420940087720457e-06,
+      "loss": 0.6194,
+      "step": 8348
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.4689509241681535,
+      "learning_rate": 2.4198060941335987e-06,
+      "loss": 0.5313,
+      "step": 8349
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.4975703045065374,
+      "learning_rate": 2.4186722814071043e-06,
+      "loss": 0.5038,
+      "step": 8350
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4524371048512315,
+      "learning_rate": 2.4175386496204513e-06,
+      "loss": 0.3769,
+      "step": 8351
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.33869852814967,
+      "learning_rate": 2.416405198853098e-06,
+      "loss": 0.9853,
+      "step": 8352
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.479560559718521,
+      "learning_rate": 2.415271929184496e-06,
+      "loss": 0.3689,
+      "step": 8353
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.584706802703382,
+      "learning_rate": 2.4141388406940852e-06,
+      "loss": 0.5739,
+      "step": 8354
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.314608982252052,
+      "learning_rate": 2.413005933461286e-06,
+      "loss": 0.6743,
+      "step": 8355
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.9993676635570075,
+      "learning_rate": 2.4118732075655144e-06,
+      "loss": 0.7207,
+      "step": 8356
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.742565487556265,
+      "learning_rate": 2.410740663086165e-06,
+      "loss": 0.722,
+      "step": 8357
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.9454738545193697,
+      "learning_rate": 2.409608300102627e-06,
+      "loss": 0.6875,
+      "step": 8358
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9855888175733343,
+      "learning_rate": 2.4084761186942734e-06,
+      "loss": 0.3532,
+      "step": 8359
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.011871695896224,
+      "learning_rate": 2.4073441189404657e-06,
+      "loss": 0.5849,
+      "step": 8360
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.309430061951973,
+      "learning_rate": 2.4062123009205525e-06,
+      "loss": 0.5575,
+      "step": 8361
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.5880549181074146,
+      "learning_rate": 2.40508066471387e-06,
+      "loss": 0.3133,
+      "step": 8362
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.3269320174595007,
+      "learning_rate": 2.403949210399738e-06,
+      "loss": 0.7437,
+      "step": 8363
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.732629529579054,
+      "learning_rate": 2.4028179380574684e-06,
+      "loss": 0.7449,
+      "step": 8364
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.433683946189046,
+      "learning_rate": 2.4016868477663586e-06,
+      "loss": 0.3557,
+      "step": 8365
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.778321708777118,
+      "learning_rate": 2.400555939605693e-06,
+      "loss": 0.8214,
+      "step": 8366
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.4557327901594586,
+      "learning_rate": 2.3994252136547426e-06,
+      "loss": 0.6301,
+      "step": 8367
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.922938761690179,
+      "learning_rate": 2.398294669992769e-06,
+      "loss": 1.1046,
+      "step": 8368
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.402524175059253,
+      "learning_rate": 2.397164308699014e-06,
+      "loss": 0.6317,
+      "step": 8369
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.8519980969718537,
+      "learning_rate": 2.396034129852716e-06,
+      "loss": 0.9654,
+      "step": 8370
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.287095492080058,
+      "learning_rate": 2.3949041335330914e-06,
+      "loss": 0.5373,
+      "step": 8371
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.4762121347365547,
+      "learning_rate": 2.3937743198193493e-06,
+      "loss": 0.3418,
+      "step": 8372
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.290139824225999,
+      "learning_rate": 2.3926446887906867e-06,
+      "loss": 0.6132,
+      "step": 8373
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.6184754356381212,
+      "learning_rate": 2.3915152405262824e-06,
+      "loss": 0.5967,
+      "step": 8374
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.263744342832122,
+      "learning_rate": 2.390385975105308e-06,
+      "loss": 0.6789,
+      "step": 8375
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.8193430996779605,
+      "learning_rate": 2.3892568926069186e-06,
+      "loss": 0.8379,
+      "step": 8376
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.210336423817697,
+      "learning_rate": 2.3881279931102602e-06,
+      "loss": 0.5075,
+      "step": 8377
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.762238365586828,
+      "learning_rate": 2.386999276694462e-06,
+      "loss": 1.466,
+      "step": 8378
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.179692768554767,
+      "learning_rate": 2.3858707434386447e-06,
+      "loss": 1.0722,
+      "step": 8379
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3857918978716115,
+      "learning_rate": 2.3847423934219094e-06,
+      "loss": 0.2436,
+      "step": 8380
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.555509662623433,
+      "learning_rate": 2.383614226723351e-06,
+      "loss": 0.7331,
+      "step": 8381
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.9332803872256683,
+      "learning_rate": 2.382486243422049e-06,
+      "loss": 0.5333,
+      "step": 8382
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.0154124204619657,
+      "learning_rate": 2.381358443597069e-06,
+      "loss": 0.5797,
+      "step": 8383
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.465284294569563,
+      "learning_rate": 2.3802308273274682e-06,
+      "loss": 0.7175,
+      "step": 8384
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.100309840782991,
+      "learning_rate": 2.379103394692284e-06,
+      "loss": 0.4776,
+      "step": 8385
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.2263436218775166,
+      "learning_rate": 2.3779761457705443e-06,
+      "loss": 0.635,
+      "step": 8386
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.996779188540018,
+      "learning_rate": 2.376849080641268e-06,
+      "loss": 0.578,
+      "step": 8387
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.8123089400031707,
+      "learning_rate": 2.375722199383454e-06,
+      "loss": 0.5775,
+      "step": 8388
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.928601130154547,
+      "learning_rate": 2.374595502076092e-06,
+      "loss": 0.6623,
+      "step": 8389
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.828229489008046,
+      "learning_rate": 2.37346898879816e-06,
+      "loss": 0.3584,
+      "step": 8390
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 6.349399585739406,
+      "learning_rate": 2.372342659628623e-06,
+      "loss": 1.202,
+      "step": 8391
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.303782777829738,
+      "learning_rate": 2.371216514646428e-06,
+      "loss": 0.6508,
+      "step": 8392
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.7366792714066372,
+      "learning_rate": 2.3700905539305147e-06,
+      "loss": 0.6159,
+      "step": 8393
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.456788330418907,
+      "learning_rate": 2.3689647775598084e-06,
+      "loss": 0.9043,
+      "step": 8394
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.601094095713435,
+      "learning_rate": 2.3678391856132203e-06,
+      "loss": 0.8011,
+      "step": 8395
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.441912768334063,
+      "learning_rate": 2.366713778169653e-06,
+      "loss": 1.1761,
+      "step": 8396
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.56436476136244,
+      "learning_rate": 2.365588555307987e-06,
+      "loss": 1.0629,
+      "step": 8397
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.026481870763221,
+      "learning_rate": 2.364463517107099e-06,
+      "loss": 0.7102,
+      "step": 8398
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.1710057109812704,
+      "learning_rate": 2.363338663645848e-06,
+      "loss": 0.5157,
+      "step": 8399
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.685673201066071,
+      "learning_rate": 2.362213995003082e-06,
+      "loss": 0.6991,
+      "step": 8400
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.694032386329679,
+      "learning_rate": 2.3610895112576372e-06,
+      "loss": 0.7543,
+      "step": 8401
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.799422054752237,
+      "learning_rate": 2.359965212488331e-06,
+      "loss": 0.7221,
+      "step": 8402
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.511747862169461,
+      "learning_rate": 2.3588410987739763e-06,
+      "loss": 0.6048,
+      "step": 8403
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.7636626894282883,
+      "learning_rate": 2.3577171701933638e-06,
+      "loss": 0.517,
+      "step": 8404
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.0276506358452417,
+      "learning_rate": 2.3565934268252787e-06,
+      "loss": 0.5485,
+      "step": 8405
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.492551641659161,
+      "learning_rate": 2.35546986874849e-06,
+      "loss": 1.0428,
+      "step": 8406
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.7812895303791634,
+      "learning_rate": 2.354346496041755e-06,
+      "loss": 0.5173,
+      "step": 8407
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.335214488177956,
+      "learning_rate": 2.353223308783818e-06,
+      "loss": 1.0085,
+      "step": 8408
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.818332111917985,
+      "learning_rate": 2.3521003070534065e-06,
+      "loss": 0.9629,
+      "step": 8409
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.515678266661508,
+      "learning_rate": 2.35097749092924e-06,
+      "loss": 0.9099,
+      "step": 8410
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.35139416211975,
+      "learning_rate": 2.349854860490023e-06,
+      "loss": 1.1736,
+      "step": 8411
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.887872785684858,
+      "learning_rate": 2.3487324158144463e-06,
+      "loss": 0.8037,
+      "step": 8412
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.827747082554949,
+      "learning_rate": 2.347610156981191e-06,
+      "loss": 1.2159,
+      "step": 8413
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.086398257738493,
+      "learning_rate": 2.3464880840689187e-06,
+      "loss": 0.692,
+      "step": 8414
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.767076008447161,
+      "learning_rate": 2.345366197156283e-06,
+      "loss": 0.7561,
+      "step": 8415
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.47313845423047,
+      "learning_rate": 2.3442444963219246e-06,
+      "loss": 0.6902,
+      "step": 8416
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.653822729716385,
+      "learning_rate": 2.3431229816444704e-06,
+      "loss": 1.0065,
+      "step": 8417
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.483539108307993,
+      "learning_rate": 2.3420016532025312e-06,
+      "loss": 1.0273,
+      "step": 8418
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.6041543053271745,
+      "learning_rate": 2.34088051107471e-06,
+      "loss": 0.7899,
+      "step": 8419
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.3378761203948524,
+      "learning_rate": 2.3397595553395903e-06,
+      "loss": 0.5621,
+      "step": 8420
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.6709496478302652,
+      "learning_rate": 2.3386387860757487e-06,
+      "loss": 0.4824,
+      "step": 8421
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.699658076987211,
+      "learning_rate": 2.337518203361746e-06,
+      "loss": 0.8225,
+      "step": 8422
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.009785215675382,
+      "learning_rate": 2.33639780727613e-06,
+      "loss": 1.0486,
+      "step": 8423
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.150383172250053,
+      "learning_rate": 2.3352775978974355e-06,
+      "loss": 0.9872,
+      "step": 8424
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.66804436038213,
+      "learning_rate": 2.334157575304186e-06,
+      "loss": 0.5993,
+      "step": 8425
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.642409328570551,
+      "learning_rate": 2.3330377395748878e-06,
+      "loss": 0.5431,
+      "step": 8426
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.604401620902554,
+      "learning_rate": 2.331918090788037e-06,
+      "loss": 0.9804,
+      "step": 8427
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.9247541734623117,
+      "learning_rate": 2.3307986290221162e-06,
+      "loss": 0.8486,
+      "step": 8428
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.126764203684734,
+      "learning_rate": 2.329679354355595e-06,
+      "loss": 0.5038,
+      "step": 8429
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.196092083621251,
+      "learning_rate": 2.32856026686693e-06,
+      "loss": 0.882,
+      "step": 8430
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.9754652441676246,
+      "learning_rate": 2.3274413666345665e-06,
+      "loss": 0.2893,
+      "step": 8431
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.1589207779363084,
+      "learning_rate": 2.32632265373693e-06,
+      "loss": 0.4983,
+      "step": 8432
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.651300803151679,
+      "learning_rate": 2.325204128252441e-06,
+      "loss": 0.8112,
+      "step": 8433
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.402444537328381,
+      "learning_rate": 2.3240857902595002e-06,
+      "loss": 0.4586,
+      "step": 8434
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.410746347629286,
+      "learning_rate": 2.3229676398365e-06,
+      "loss": 1.0668,
+      "step": 8435
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.030115535323992,
+      "learning_rate": 2.32184967706182e-06,
+      "loss": 0.9424,
+      "step": 8436
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.7400689708001287,
+      "learning_rate": 2.3207319020138197e-06,
+      "loss": 0.5923,
+      "step": 8437
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.384599097966671,
+      "learning_rate": 2.319614314770853e-06,
+      "loss": 0.8099,
+      "step": 8438
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.172882423162879,
+      "learning_rate": 2.3184969154112585e-06,
+      "loss": 0.4573,
+      "step": 8439
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.6910280218831772,
+      "learning_rate": 2.3173797040133595e-06,
+      "loss": 0.4993,
+      "step": 8440
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.786782603564133,
+      "learning_rate": 2.3162626806554687e-06,
+      "loss": 0.746,
+      "step": 8441
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.511630429143293,
+      "learning_rate": 2.3151458454158867e-06,
+      "loss": 0.656,
+      "step": 8442
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.738657960078739,
+      "learning_rate": 2.3140291983728936e-06,
+      "loss": 0.9872,
+      "step": 8443
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.782023982418744,
+      "learning_rate": 2.312912739604765e-06,
+      "loss": 0.3056,
+      "step": 8444
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.234880690556221,
+      "learning_rate": 2.311796469189759e-06,
+      "loss": 0.469,
+      "step": 8445
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.070585301608283,
+      "learning_rate": 2.310680387206121e-06,
+      "loss": 0.3329,
+      "step": 8446
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.966473252835259,
+      "learning_rate": 2.309564493732086e-06,
+      "loss": 1.1336,
+      "step": 8447
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.5164535722361734,
+      "learning_rate": 2.3084487888458697e-06,
+      "loss": 0.5251,
+      "step": 8448
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.947278176717033,
+      "learning_rate": 2.3073332726256807e-06,
+      "loss": 1.2562,
+      "step": 8449
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.970368894447125,
+      "learning_rate": 2.3062179451497095e-06,
+      "loss": 0.9964,
+      "step": 8450
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.754486466883372,
+      "learning_rate": 2.305102806496137e-06,
+      "loss": 1.2073,
+      "step": 8451
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.720551063877111,
+      "learning_rate": 2.303987856743129e-06,
+      "loss": 0.5452,
+      "step": 8452
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.287684194272693,
+      "learning_rate": 2.3028730959688417e-06,
+      "loss": 0.7301,
+      "step": 8453
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.782316811576312,
+      "learning_rate": 2.30175852425141e-06,
+      "loss": 0.9407,
+      "step": 8454
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.008913663948316,
+      "learning_rate": 2.3006441416689633e-06,
+      "loss": 0.3661,
+      "step": 8455
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.318025086232537,
+      "learning_rate": 2.2995299482996146e-06,
+      "loss": 0.6932,
+      "step": 8456
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.502260191231009,
+      "learning_rate": 2.2984159442214637e-06,
+      "loss": 0.625,
+      "step": 8457
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.8011882692816124,
+      "learning_rate": 2.2973021295125985e-06,
+      "loss": 0.607,
+      "step": 8458
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.626678714352333,
+      "learning_rate": 2.2961885042510927e-06,
+      "loss": 1.0223,
+      "step": 8459
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.755927228048318,
+      "learning_rate": 2.2950750685150045e-06,
+      "loss": 0.8516,
+      "step": 8460
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.896744191915756,
+      "learning_rate": 2.293961822382382e-06,
+      "loss": 0.6381,
+      "step": 8461
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.604139263103291,
+      "learning_rate": 2.2928487659312594e-06,
+      "loss": 0.5576,
+      "step": 8462
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.739531463436227,
+      "learning_rate": 2.291735899239658e-06,
+      "loss": 0.9477,
+      "step": 8463
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.074405910792501,
+      "learning_rate": 2.2906232223855824e-06,
+      "loss": 0.7331,
+      "step": 8464
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.6067943074762,
+      "learning_rate": 2.289510735447029e-06,
+      "loss": 0.8256,
+      "step": 8465
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.256104320093683,
+      "learning_rate": 2.288398438501976e-06,
+      "loss": 1.0537,
+      "step": 8466
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.9481643168142435,
+      "learning_rate": 2.2872863316283906e-06,
+      "loss": 0.5068,
+      "step": 8467
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.7919291921357177,
+      "learning_rate": 2.2861744149042275e-06,
+      "loss": 0.7032,
+      "step": 8468
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.4314916854245245,
+      "learning_rate": 2.285062688407428e-06,
+      "loss": 0.4225,
+      "step": 8469
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.039513891459796,
+      "learning_rate": 2.283951152215918e-06,
+      "loss": 0.3852,
+      "step": 8470
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.839655352129458,
+      "learning_rate": 2.282839806407614e-06,
+      "loss": 0.5821,
+      "step": 8471
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.067054657714496,
+      "learning_rate": 2.2817286510604125e-06,
+      "loss": 0.7038,
+      "step": 8472
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.5436743447807757,
+      "learning_rate": 2.280617686252203e-06,
+      "loss": 0.7465,
+      "step": 8473
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.627480036880374,
+      "learning_rate": 2.279506912060859e-06,
+      "loss": 0.5592,
+      "step": 8474
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.5964562108123093,
+      "learning_rate": 2.2783963285642403e-06,
+      "loss": 0.6405,
+      "step": 8475
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.097271718922358,
+      "learning_rate": 2.2772859358401962e-06,
+      "loss": 1.1977,
+      "step": 8476
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.390702689594146,
+      "learning_rate": 2.2761757339665576e-06,
+      "loss": 0.6897,
+      "step": 8477
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.4958214177554683,
+      "learning_rate": 2.2750657230211452e-06,
+      "loss": 0.8866,
+      "step": 8478
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.150754538967326,
+      "learning_rate": 2.2739559030817687e-06,
+      "loss": 0.811,
+      "step": 8479
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.187699471097217,
+      "learning_rate": 2.272846274226218e-06,
+      "loss": 1.2004,
+      "step": 8480
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.077357679192823,
+      "learning_rate": 2.2717368365322747e-06,
+      "loss": 0.3078,
+      "step": 8481
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.514300456812758,
+      "learning_rate": 2.2706275900777075e-06,
+      "loss": 0.936,
+      "step": 8482
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.7356245830003685,
+      "learning_rate": 2.2695185349402664e-06,
+      "loss": 0.9696,
+      "step": 8483
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.098094550770222,
+      "learning_rate": 2.2684096711976926e-06,
+      "loss": 0.8091,
+      "step": 8484
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.791551196522658,
+      "learning_rate": 2.2673009989277136e-06,
+      "loss": 0.6636,
+      "step": 8485
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.060922227655955,
+      "learning_rate": 2.266192518208041e-06,
+      "loss": 0.5593,
+      "step": 8486
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.8403922186175374,
+      "learning_rate": 2.2650842291163755e-06,
+      "loss": 0.7785,
+      "step": 8487
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.6763855490718926,
+      "learning_rate": 2.2639761317304047e-06,
+      "loss": 0.4132,
+      "step": 8488
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.600935939884642,
+      "learning_rate": 2.262868226127799e-06,
+      "loss": 1.0364,
+      "step": 8489
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.982098126888399,
+      "learning_rate": 2.261760512386218e-06,
+      "loss": 0.9001,
+      "step": 8490
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.4049285672036613,
+      "learning_rate": 2.260652990583308e-06,
+      "loss": 0.4007,
+      "step": 8491
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9316133492785437,
+      "learning_rate": 2.259545660796702e-06,
+      "loss": 0.4114,
+      "step": 8492
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.183810875278495,
+      "learning_rate": 2.2584385231040202e-06,
+      "loss": 0.5691,
+      "step": 8493
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.435946658191863,
+      "learning_rate": 2.2573315775828655e-06,
+      "loss": 0.9812,
+      "step": 8494
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.831926589661101,
+      "learning_rate": 2.2562248243108305e-06,
+      "loss": 0.8021,
+      "step": 8495
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.4545849790556784,
+      "learning_rate": 2.255118263365496e-06,
+      "loss": 0.7481,
+      "step": 8496
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.1328991820797314,
+      "learning_rate": 2.254011894824424e-06,
+      "loss": 0.5929,
+      "step": 8497
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.2939912257282327,
+      "learning_rate": 2.2529057187651675e-06,
+      "loss": 0.7333,
+      "step": 8498
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.835804670368205,
+      "learning_rate": 2.2517997352652663e-06,
+      "loss": 1.0314,
+      "step": 8499
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.942105641728355,
+      "learning_rate": 2.2506939444022423e-06,
+      "loss": 1.1971,
+      "step": 8500
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.051373609684334,
+      "learning_rate": 2.249588346253607e-06,
+      "loss": 0.9748,
+      "step": 8501
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.5037761957391838,
+      "learning_rate": 2.2484829408968593e-06,
+      "loss": 0.3577,
+      "step": 8502
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.378439313967647,
+      "learning_rate": 2.247377728409483e-06,
+      "loss": 0.9042,
+      "step": 8503
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.074400812627169,
+      "learning_rate": 2.246272708868948e-06,
+      "loss": 0.647,
+      "step": 8504
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.423967160613771,
+      "learning_rate": 2.245167882352714e-06,
+      "loss": 0.513,
+      "step": 8505
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.1295114828424264,
+      "learning_rate": 2.24406324893822e-06,
+      "loss": 0.6978,
+      "step": 8506
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.43587445844676,
+      "learning_rate": 2.2429588087028993e-06,
+      "loss": 0.6342,
+      "step": 8507
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.0636716675478004,
+      "learning_rate": 2.2418545617241665e-06,
+      "loss": 0.5572,
+      "step": 8508
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.740701769536277,
+      "learning_rate": 2.2407505080794257e-06,
+      "loss": 1.0533,
+      "step": 8509
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.052240351497302,
+      "learning_rate": 2.239646647846068e-06,
+      "loss": 0.6739,
+      "step": 8510
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.4173514237523195,
+      "learning_rate": 2.2385429811014654e-06,
+      "loss": 0.7716,
+      "step": 8511
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.8918025523546453,
+      "learning_rate": 2.2374395079229837e-06,
+      "loss": 0.7011,
+      "step": 8512
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.4332955544209383,
+      "learning_rate": 2.236336228387968e-06,
+      "loss": 0.5024,
+      "step": 8513
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.512813254636162,
+      "learning_rate": 2.235233142573755e-06,
+      "loss": 0.8418,
+      "step": 8514
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.401873909504076,
+      "learning_rate": 2.2341302505576663e-06,
+      "loss": 0.5164,
+      "step": 8515
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.8503631153374727,
+      "learning_rate": 2.233027552417012e-06,
+      "loss": 0.3192,
+      "step": 8516
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.240333467526782,
+      "learning_rate": 2.2319250482290826e-06,
+      "loss": 0.7147,
+      "step": 8517
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.160602406918777,
+      "learning_rate": 2.2308227380711605e-06,
+      "loss": 0.9787,
+      "step": 8518
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.500392377011563,
+      "learning_rate": 2.229720622020513e-06,
+      "loss": 1.2103,
+      "step": 8519
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.6966145907371657,
+      "learning_rate": 2.2286187001543936e-06,
+      "loss": 0.6251,
+      "step": 8520
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.258155138649007,
+      "learning_rate": 2.2275169725500424e-06,
+      "loss": 0.471,
+      "step": 8521
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.217948330809415,
+      "learning_rate": 2.226415439284687e-06,
+      "loss": 1.0303,
+      "step": 8522
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.27214185715932,
+      "learning_rate": 2.2253141004355367e-06,
+      "loss": 1.1734,
+      "step": 8523
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.254554209891187,
+      "learning_rate": 2.2242129560797933e-06,
+      "loss": 0.5383,
+      "step": 8524
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.3820214523269243,
+      "learning_rate": 2.2231120062946405e-06,
+      "loss": 0.5971,
+      "step": 8525
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.635763140895818,
+      "learning_rate": 2.2220112511572533e-06,
+      "loss": 0.6923,
+      "step": 8526
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.1862781005744125,
+      "learning_rate": 2.2209106907447853e-06,
+      "loss": 0.8966,
+      "step": 8527
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.845679683065618,
+      "learning_rate": 2.2198103251343856e-06,
+      "loss": 1.4883,
+      "step": 8528
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.715864144147195,
+      "learning_rate": 2.2187101544031807e-06,
+      "loss": 0.3922,
+      "step": 8529
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.7776553147772858,
+      "learning_rate": 2.2176101786282893e-06,
+      "loss": 0.994,
+      "step": 8530
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.880908010643266,
+      "learning_rate": 2.216510397886816e-06,
+      "loss": 0.57,
+      "step": 8531
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.80791939885431,
+      "learning_rate": 2.21541081225585e-06,
+      "loss": 0.5118,
+      "step": 8532
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.443131290143332,
+      "learning_rate": 2.214311421812467e-06,
+      "loss": 1.2995,
+      "step": 8533
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.320553504408982,
+      "learning_rate": 2.2132122266337326e-06,
+      "loss": 1.0689,
+      "step": 8534
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.136281392446387,
+      "learning_rate": 2.2121132267966907e-06,
+      "loss": 0.6286,
+      "step": 8535
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.513037998640322,
+      "learning_rate": 2.211014422378378e-06,
+      "loss": 0.8584,
+      "step": 8536
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.308923289937153,
+      "learning_rate": 2.2099158134558175e-06,
+      "loss": 0.9038,
+      "step": 8537
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.8295110576552265,
+      "learning_rate": 2.2088174001060154e-06,
+      "loss": 0.6404,
+      "step": 8538
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.0955056702896093,
+      "learning_rate": 2.2077191824059685e-06,
+      "loss": 0.2939,
+      "step": 8539
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.883527411650362,
+      "learning_rate": 2.2066211604326533e-06,
+      "loss": 1.0983,
+      "step": 8540
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.306017936599753,
+      "learning_rate": 2.2055233342630372e-06,
+      "loss": 0.7268,
+      "step": 8541
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.035332985908827,
+      "learning_rate": 2.204425703974076e-06,
+      "loss": 0.5606,
+      "step": 8542
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.3892802423350403,
+      "learning_rate": 2.2033282696427046e-06,
+      "loss": 0.5014,
+      "step": 8543
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.318912314596869,
+      "learning_rate": 2.2022310313458506e-06,
+      "loss": 0.9304,
+      "step": 8544
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.10509095348596,
+      "learning_rate": 2.201133989160427e-06,
+      "loss": 0.8174,
+      "step": 8545
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.525349725727181,
+      "learning_rate": 2.200037143163328e-06,
+      "loss": 0.6178,
+      "step": 8546
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.3115879393343763,
+      "learning_rate": 2.19894049343144e-06,
+      "loss": 0.7725,
+      "step": 8547
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.3625063552278305,
+      "learning_rate": 2.1978440400416334e-06,
+      "loss": 0.5882,
+      "step": 8548
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.257631859288316,
+      "learning_rate": 2.1967477830707644e-06,
+      "loss": 0.4958,
+      "step": 8549
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.1902414133312607,
+      "learning_rate": 2.195651722595676e-06,
+      "loss": 0.4216,
+      "step": 8550
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.37316639130754,
+      "learning_rate": 2.1945558586931994e-06,
+      "loss": 0.8959,
+      "step": 8551
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.787099690497614,
+      "learning_rate": 2.1934601914401454e-06,
+      "loss": 0.4605,
+      "step": 8552
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.080559088720089,
+      "learning_rate": 2.1923647209133182e-06,
+      "loss": 1.084,
+      "step": 8553
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.811972776542819,
+      "learning_rate": 2.1912694471895053e-06,
+      "loss": 1.3616,
+      "step": 8554
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.9511910177455705,
+      "learning_rate": 2.1901743703454804e-06,
+      "loss": 0.7359,
+      "step": 8555
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.457071580024648,
+      "learning_rate": 2.1890794904580054e-06,
+      "loss": 0.2913,
+      "step": 8556
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 6.205523943575567,
+      "learning_rate": 2.187984807603823e-06,
+      "loss": 0.9544,
+      "step": 8557
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.9482181278145814,
+      "learning_rate": 2.18689032185967e-06,
+      "loss": 0.4539,
+      "step": 8558
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.634360609800492,
+      "learning_rate": 2.1857960333022605e-06,
+      "loss": 0.7313,
+      "step": 8559
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.2260508333855866,
+      "learning_rate": 2.1847019420083014e-06,
+      "loss": 0.7196,
+      "step": 8560
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.2922376954795105,
+      "learning_rate": 2.1836080480544847e-06,
+      "loss": 1.0641,
+      "step": 8561
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.930928441831621,
+      "learning_rate": 2.182514351517488e-06,
+      "loss": 0.9905,
+      "step": 8562
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.876421767273682,
+      "learning_rate": 2.1814208524739723e-06,
+      "loss": 1.2121,
+      "step": 8563
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.4454066148723985,
+      "learning_rate": 2.1803275510005876e-06,
+      "loss": 0.9244,
+      "step": 8564
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.20130506058506,
+      "learning_rate": 2.1792344471739708e-06,
+      "loss": 0.4858,
+      "step": 8565
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.7595268897264678,
+      "learning_rate": 2.178141541070743e-06,
+      "loss": 0.5205,
+      "step": 8566
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.694964593190989,
+      "learning_rate": 2.177048832767513e-06,
+      "loss": 0.8947,
+      "step": 8567
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.112295196327415,
+      "learning_rate": 2.1759563223408754e-06,
+      "loss": 1.0051,
+      "step": 8568
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.065277813148473,
+      "learning_rate": 2.174864009867408e-06,
+      "loss": 0.3969,
+      "step": 8569
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.397957121363607,
+      "learning_rate": 2.173771895423678e-06,
+      "loss": 1.0383,
+      "step": 8570
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.3720889266708243,
+      "learning_rate": 2.1726799790862384e-06,
+      "loss": 0.5068,
+      "step": 8571
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.143698773057102,
+      "learning_rate": 2.17158826093163e-06,
+      "loss": 0.1789,
+      "step": 8572
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.673864608620568,
+      "learning_rate": 2.170496741036373e-06,
+      "loss": 1.0408,
+      "step": 8573
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.5684299500075745,
+      "learning_rate": 2.1694054194769827e-06,
+      "loss": 0.3384,
+      "step": 8574
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.626045869543701,
+      "learning_rate": 2.1683142963299513e-06,
+      "loss": 0.9893,
+      "step": 8575
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.8760043551483974,
+      "learning_rate": 2.1672233716717644e-06,
+      "loss": 0.7937,
+      "step": 8576
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.966860606458185,
+      "learning_rate": 2.166132645578891e-06,
+      "loss": 1.2491,
+      "step": 8577
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.7472715777648595,
+      "learning_rate": 2.165042118127786e-06,
+      "loss": 0.5268,
+      "step": 8578
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.278882749479071,
+      "learning_rate": 2.1639517893948926e-06,
+      "loss": 1.8395,
+      "step": 8579
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.4903472991909483,
+      "learning_rate": 2.162861659456634e-06,
+      "loss": 0.6644,
+      "step": 8580
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.1172260213709335,
+      "learning_rate": 2.161771728389427e-06,
+      "loss": 0.3231,
+      "step": 8581
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.556372738087387,
+      "learning_rate": 2.1606819962696684e-06,
+      "loss": 0.7001,
+      "step": 8582
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.651860880286598,
+      "learning_rate": 2.159592463173746e-06,
+      "loss": 0.7687,
+      "step": 8583
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.119051509401893,
+      "learning_rate": 2.1585031291780302e-06,
+      "loss": 0.4643,
+      "step": 8584
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.4095767126412575,
+      "learning_rate": 2.1574139943588807e-06,
+      "loss": 0.7777,
+      "step": 8585
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.109160471129687,
+      "learning_rate": 2.156325058792637e-06,
+      "loss": 0.4316,
+      "step": 8586
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.725993390309628,
+      "learning_rate": 2.1552363225556316e-06,
+      "loss": 0.9986,
+      "step": 8587
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.528936179212356,
+      "learning_rate": 2.154147785724181e-06,
+      "loss": 0.6542,
+      "step": 8588
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.940343252172774,
+      "learning_rate": 2.153059448374584e-06,
+      "loss": 0.5965,
+      "step": 8589
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.614609999316483,
+      "learning_rate": 2.151971310583129e-06,
+      "loss": 0.5886,
+      "step": 8590
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.928864239744701,
+      "learning_rate": 2.150883372426093e-06,
+      "loss": 0.7739,
+      "step": 8591
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.89137734702575,
+      "learning_rate": 2.149795633979731e-06,
+      "loss": 0.7796,
+      "step": 8592
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.166206859131393,
+      "learning_rate": 2.1487080953202912e-06,
+      "loss": 0.6869,
+      "step": 8593
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.716597618134034,
+      "learning_rate": 2.147620756524004e-06,
+      "loss": 0.5369,
+      "step": 8594
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.704012501500497,
+      "learning_rate": 2.1465336176670893e-06,
+      "loss": 0.804,
+      "step": 8595
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.8171949102624865,
+      "learning_rate": 2.145446678825751e-06,
+      "loss": 0.8484,
+      "step": 8596
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.5497017487230424,
+      "learning_rate": 2.144359940076176e-06,
+      "loss": 0.7923,
+      "step": 8597
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.219090379332379,
+      "learning_rate": 2.1432734014945417e-06,
+      "loss": 0.6083,
+      "step": 8598
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.0059478905592725,
+      "learning_rate": 2.1421870631570083e-06,
+      "loss": 0.8369,
+      "step": 8599
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.82593726098331,
+      "learning_rate": 2.1411009251397257e-06,
+      "loss": 0.7378,
+      "step": 8600
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.2473968851947275,
+      "learning_rate": 2.140014987518826e-06,
+      "loss": 0.8269,
+      "step": 8601
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.875191924636826,
+      "learning_rate": 2.13892925037043e-06,
+      "loss": 0.694,
+      "step": 8602
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.177039573557034,
+      "learning_rate": 2.1378437137706413e-06,
+      "loss": 0.4679,
+      "step": 8603
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.005304744198496,
+      "learning_rate": 2.136758377795552e-06,
+      "loss": 0.8098,
+      "step": 8604
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.245893912971883,
+      "learning_rate": 2.1356732425212406e-06,
+      "loss": 1.0079,
+      "step": 8605
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.2508512678952,
+      "learning_rate": 2.1345883080237684e-06,
+      "loss": 0.7945,
+      "step": 8606
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.246922879484139,
+      "learning_rate": 2.133503574379185e-06,
+      "loss": 0.465,
+      "step": 8607
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.7239696989996944,
+      "learning_rate": 2.1324190416635275e-06,
+      "loss": 0.6172,
+      "step": 8608
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.058882977329405,
+      "learning_rate": 2.131334709952814e-06,
+      "loss": 0.8411,
+      "step": 8609
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 6.055853011000352,
+      "learning_rate": 2.1302505793230534e-06,
+      "loss": 1.2077,
+      "step": 8610
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.238300224717042,
+      "learning_rate": 2.129166649850237e-06,
+      "loss": 0.5948,
+      "step": 8611
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.8779211339993243,
+      "learning_rate": 2.128082921610345e-06,
+      "loss": 0.3973,
+      "step": 8612
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.3118160894168165,
+      "learning_rate": 2.1269993946793414e-06,
+      "loss": 0.6056,
+      "step": 8613
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.916006967340428,
+      "learning_rate": 2.1259160691331794e-06,
+      "loss": 0.5635,
+      "step": 8614
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.3416903112367136,
+      "learning_rate": 2.1248329450477904e-06,
+      "loss": 0.8378,
+      "step": 8615
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.3951543935934,
+      "learning_rate": 2.1237500224990994e-06,
+      "loss": 0.5687,
+      "step": 8616
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.875284908893516,
+      "learning_rate": 2.122667301563014e-06,
+      "loss": 0.8145,
+      "step": 8617
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.8377233678186777,
+      "learning_rate": 2.121584782315429e-06,
+      "loss": 0.3168,
+      "step": 8618
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.5394341484148724,
+      "learning_rate": 2.1205024648322254e-06,
+      "loss": 0.5774,
+      "step": 8619
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.066940806716527,
+      "learning_rate": 2.1194203491892657e-06,
+      "loss": 0.6829,
+      "step": 8620
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.0735806068333074,
+      "learning_rate": 2.1183384354624053e-06,
+      "loss": 0.5733,
+      "step": 8621
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.357778850248498,
+      "learning_rate": 2.117256723727477e-06,
+      "loss": 0.5016,
+      "step": 8622
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.96812977430919,
+      "learning_rate": 2.1161752140603077e-06,
+      "loss": 1.1085,
+      "step": 8623
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.0458089164222693,
+      "learning_rate": 2.1150939065367042e-06,
+      "loss": 0.8446,
+      "step": 8624
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.750234262423207,
+      "learning_rate": 2.114012801232465e-06,
+      "loss": 0.7071,
+      "step": 8625
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.686081707197256,
+      "learning_rate": 2.1129318982233673e-06,
+      "loss": 0.7895,
+      "step": 8626
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.589439961308198,
+      "learning_rate": 2.1118511975851786e-06,
+      "loss": 0.8295,
+      "step": 8627
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.381214266267079,
+      "learning_rate": 2.1107706993936517e-06,
+      "loss": 1.0563,
+      "step": 8628
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.7484594029638356,
+      "learning_rate": 2.109690403724525e-06,
+      "loss": 0.7031,
+      "step": 8629
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.878346462910791,
+      "learning_rate": 2.1086103106535214e-06,
+      "loss": 0.5772,
+      "step": 8630
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.6799841612222877,
+      "learning_rate": 2.1075304202563545e-06,
+      "loss": 0.4902,
+      "step": 8631
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.3065349872163217,
+      "learning_rate": 2.106450732608715e-06,
+      "loss": 0.5272,
+      "step": 8632
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.312670386870131,
+      "learning_rate": 2.105371247786286e-06,
+      "loss": 0.6805,
+      "step": 8633
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.36764524941005,
+      "learning_rate": 2.1042919658647354e-06,
+      "loss": 0.6347,
+      "step": 8634
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0413073985924002,
+      "learning_rate": 2.1032128869197177e-06,
+      "loss": 0.14,
+      "step": 8635
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.05262620919622,
+      "learning_rate": 2.102134011026868e-06,
+      "loss": 1.2925,
+      "step": 8636
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.59306747824626,
+      "learning_rate": 2.1010553382618137e-06,
+      "loss": 0.6344,
+      "step": 8637
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.6625150732439735,
+      "learning_rate": 2.099976868700163e-06,
+      "loss": 1.0499,
+      "step": 8638
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.1795958342969723,
+      "learning_rate": 2.0988986024175124e-06,
+      "loss": 0.5804,
+      "step": 8639
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.363379536622431,
+      "learning_rate": 2.097820539489444e-06,
+      "loss": 1.1703,
+      "step": 8640
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.0096364022705266,
+      "learning_rate": 2.096742679991526e-06,
+      "loss": 0.7554,
+      "step": 8641
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.455715040513336,
+      "learning_rate": 2.0956650239993125e-06,
+      "loss": 0.8611,
+      "step": 8642
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.35198840252164,
+      "learning_rate": 2.0945875715883395e-06,
+      "loss": 0.1979,
+      "step": 8643
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.093523913388497,
+      "learning_rate": 2.0935103228341334e-06,
+      "loss": 0.7536,
+      "step": 8644
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.354438838369588,
+      "learning_rate": 2.092433277812204e-06,
+      "loss": 0.3875,
+      "step": 8645
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.097243062777404,
+      "learning_rate": 2.091356436598049e-06,
+      "loss": 0.555,
+      "step": 8646
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.154097071235236,
+      "learning_rate": 2.0902797992671485e-06,
+      "loss": 0.6429,
+      "step": 8647
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.794055154080764,
+      "learning_rate": 2.0892033658949734e-06,
+      "loss": 1.0275,
+      "step": 8648
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.1779714128589682,
+      "learning_rate": 2.088127136556972e-06,
+      "loss": 0.8594,
+      "step": 8649
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.656850932064562,
+      "learning_rate": 2.087051111328586e-06,
+      "loss": 0.3271,
+      "step": 8650
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.103798675756965,
+      "learning_rate": 2.0859752902852425e-06,
+      "loss": 0.6619,
+      "step": 8651
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.213857090780219,
+      "learning_rate": 2.084899673502347e-06,
+      "loss": 0.4175,
+      "step": 8652
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.799983173929279,
+      "learning_rate": 2.0838242610552974e-06,
+      "loss": 0.3823,
+      "step": 8653
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.879700441990109,
+      "learning_rate": 2.082749053019478e-06,
+      "loss": 0.217,
+      "step": 8654
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.2683120918792414,
+      "learning_rate": 2.081674049470252e-06,
+      "loss": 0.2636,
+      "step": 8655
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.2220936779577714,
+      "learning_rate": 2.080599250482975e-06,
+      "loss": 0.6549,
+      "step": 8656
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.4491755965852686,
+      "learning_rate": 2.0795246561329853e-06,
+      "loss": 0.7324,
+      "step": 8657
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.355232166552895,
+      "learning_rate": 2.078450266495607e-06,
+      "loss": 1.1377,
+      "step": 8658
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.7165940900652568,
+      "learning_rate": 2.077376081646152e-06,
+      "loss": 0.4958,
+      "step": 8659
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.523644529564438,
+      "learning_rate": 2.0763021016599126e-06,
+      "loss": 0.8052,
+      "step": 8660
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.658025097991605,
+      "learning_rate": 2.075228326612172e-06,
+      "loss": 0.6866,
+      "step": 8661
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.67246750953226,
+      "learning_rate": 2.074154756578197e-06,
+      "loss": 1.0907,
+      "step": 8662
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.4601611546585773,
+      "learning_rate": 2.0730813916332406e-06,
+      "loss": 0.4387,
+      "step": 8663
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.437873537315912,
+      "learning_rate": 2.0720082318525405e-06,
+      "loss": 1.1664,
+      "step": 8664
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.804471556316076,
+      "learning_rate": 2.070935277311322e-06,
+      "loss": 1.148,
+      "step": 8665
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.3163855581907216,
+      "learning_rate": 2.0698625280847917e-06,
+      "loss": 0.3145,
+      "step": 8666
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.303632904484169,
+      "learning_rate": 2.0687899842481486e-06,
+      "loss": 1.0898,
+      "step": 8667
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.116890375410322,
+      "learning_rate": 2.0677176458765686e-06,
+      "loss": 0.9163,
+      "step": 8668
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.764200063004666,
+      "learning_rate": 2.06664551304522e-06,
+      "loss": 0.4481,
+      "step": 8669
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.14798796756444,
+      "learning_rate": 2.0655735858292554e-06,
+      "loss": 0.4932,
+      "step": 8670
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 7.033735479867001,
+      "learning_rate": 2.0645018643038132e-06,
+      "loss": 1.2896,
+      "step": 8671
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.9940314904487093,
+      "learning_rate": 2.0634303485440133e-06,
+      "loss": 0.7398,
+      "step": 8672
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.5589449579387455,
+      "learning_rate": 2.0623590386249665e-06,
+      "loss": 0.3774,
+      "step": 8673
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.08627383768658,
+      "learning_rate": 2.0612879346217655e-06,
+      "loss": 1.1062,
+      "step": 8674
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.217803109246731,
+      "learning_rate": 2.0602170366094916e-06,
+      "loss": 0.7815,
+      "step": 8675
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.417057549848958,
+      "learning_rate": 2.059146344663211e-06,
+      "loss": 0.6506,
+      "step": 8676
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.1929784125578,
+      "learning_rate": 2.0580758588579712e-06,
+      "loss": 1.117,
+      "step": 8677
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.326294409846503,
+      "learning_rate": 2.057005579268811e-06,
+      "loss": 0.432,
+      "step": 8678
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.8798573621145542,
+      "learning_rate": 2.055935505970751e-06,
+      "loss": 0.5126,
+      "step": 8679
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.567847232897861,
+      "learning_rate": 2.0548656390388e-06,
+      "loss": 0.579,
+      "step": 8680
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 6.5674547113868655,
+      "learning_rate": 2.0537959785479517e-06,
+      "loss": 1.1413,
+      "step": 8681
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.061943075033028,
+      "learning_rate": 2.052726524573182e-06,
+      "loss": 1.0565,
+      "step": 8682
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.504154197985221,
+      "learning_rate": 2.0516572771894577e-06,
+      "loss": 0.7166,
+      "step": 8683
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.522642446726566,
+      "learning_rate": 2.0505882364717254e-06,
+      "loss": 0.9991,
+      "step": 8684
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.329741163430596,
+      "learning_rate": 2.049519402494922e-06,
+      "loss": 0.8487,
+      "step": 8685
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.969283958897433,
+      "learning_rate": 2.048450775333968e-06,
+      "loss": 0.7496,
+      "step": 8686
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.608275774674261,
+      "learning_rate": 2.0473823550637694e-06,
+      "loss": 1.44,
+      "step": 8687
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.140075850975716,
+      "learning_rate": 2.04631414175922e-06,
+      "loss": 0.7435,
+      "step": 8688
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.7000203667267555,
+      "learning_rate": 2.045246135495192e-06,
+      "loss": 0.7559,
+      "step": 8689
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.45115455835716,
+      "learning_rate": 2.0441783363465517e-06,
+      "loss": 1.0826,
+      "step": 8690
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.183676669097272,
+      "learning_rate": 2.043110744388146e-06,
+      "loss": 0.8767,
+      "step": 8691
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.424393510654391,
+      "learning_rate": 2.042043359694808e-06,
+      "loss": 0.876,
+      "step": 8692
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.366657168875679,
+      "learning_rate": 2.0409761823413583e-06,
+      "loss": 0.5922,
+      "step": 8693
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.664891129493486,
+      "learning_rate": 2.039909212402602e-06,
+      "loss": 0.9714,
+      "step": 8694
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.585462203946919,
+      "learning_rate": 2.038842449953326e-06,
+      "loss": 0.7847,
+      "step": 8695
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.767976704969888,
+      "learning_rate": 2.037775895068307e-06,
+      "loss": 1.002,
+      "step": 8696
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.176040644881519,
+      "learning_rate": 2.0367095478223076e-06,
+      "loss": 0.6875,
+      "step": 8697
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.6062123742969523,
+      "learning_rate": 2.035643408290071e-06,
+      "loss": 0.3599,
+      "step": 8698
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.103258869530679,
+      "learning_rate": 2.034577476546331e-06,
+      "loss": 0.6648,
+      "step": 8699
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.065527963652389,
+      "learning_rate": 2.033511752665806e-06,
+      "loss": 1.0606,
+      "step": 8700
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.89701876913943,
+      "learning_rate": 2.0324462367231953e-06,
+      "loss": 0.9328,
+      "step": 8701
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.282913577894604,
+      "learning_rate": 2.031380928793188e-06,
+      "loss": 1.1449,
+      "step": 8702
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.093768652859012,
+      "learning_rate": 2.0303158289504583e-06,
+      "loss": 0.5894,
+      "step": 8703
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.5779395972063974,
+      "learning_rate": 2.0292509372696652e-06,
+      "loss": 0.5017,
+      "step": 8704
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.4919811415807835,
+      "learning_rate": 2.028186253825454e-06,
+      "loss": 0.5298,
+      "step": 8705
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.553309452899958,
+      "learning_rate": 2.027121778692451e-06,
+      "loss": 0.6475,
+      "step": 8706
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.8409447438725413,
+      "learning_rate": 2.026057511945274e-06,
+      "loss": 0.5548,
+      "step": 8707
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.901430709220455,
+      "learning_rate": 2.0249934536585223e-06,
+      "loss": 1.1477,
+      "step": 8708
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.586139984985449,
+      "learning_rate": 2.023929603906783e-06,
+      "loss": 1.0993,
+      "step": 8709
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.95086982419721,
+      "learning_rate": 2.0228659627646257e-06,
+      "loss": 1.1894,
+      "step": 8710
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.8891510456627842,
+      "learning_rate": 2.02180253030661e-06,
+      "loss": 0.5169,
+      "step": 8711
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.984012817968478,
+      "learning_rate": 2.020739306607274e-06,
+      "loss": 0.7946,
+      "step": 8712
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.3204634258703845,
+      "learning_rate": 2.0196762917411466e-06,
+      "loss": 1.0686,
+      "step": 8713
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.26589177246894,
+      "learning_rate": 2.018613485782743e-06,
+      "loss": 0.9131,
+      "step": 8714
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.39516109327882,
+      "learning_rate": 2.0175508888065563e-06,
+      "loss": 0.7271,
+      "step": 8715
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.154450605412165,
+      "learning_rate": 2.0164885008870755e-06,
+      "loss": 0.7811,
+      "step": 8716
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.72573839412881,
+      "learning_rate": 2.0154263220987642e-06,
+      "loss": 0.7865,
+      "step": 8717
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.212752989508091,
+      "learning_rate": 2.014364352516079e-06,
+      "loss": 0.3108,
+      "step": 8718
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.321230885821851,
+      "learning_rate": 2.013302592213459e-06,
+      "loss": 0.8986,
+      "step": 8719
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.524244018089491,
+      "learning_rate": 2.0122410412653294e-06,
+      "loss": 0.5415,
+      "step": 8720
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.248701771187896,
+      "learning_rate": 2.0111796997460997e-06,
+      "loss": 1.0274,
+      "step": 8721
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.8017975856320105,
+      "learning_rate": 2.010118567730167e-06,
+      "loss": 0.7047,
+      "step": 8722
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.6514091243692848,
+      "learning_rate": 2.0090576452919095e-06,
+      "loss": 0.5102,
+      "step": 8723
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.198097127062312,
+      "learning_rate": 2.0079969325056947e-06,
+      "loss": 0.9692,
+      "step": 8724
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.892927176413908,
+      "learning_rate": 2.006936429445873e-06,
+      "loss": 0.4655,
+      "step": 8725
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.656582107206415,
+      "learning_rate": 2.005876136186782e-06,
+      "loss": 1.165,
+      "step": 8726
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7809918320027742,
+      "learning_rate": 2.0048160528027438e-06,
+      "loss": 0.4113,
+      "step": 8727
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.950507539669394,
+      "learning_rate": 2.003756179368067e-06,
+      "loss": 1.0302,
+      "step": 8728
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.698170971533982,
+      "learning_rate": 2.00269651595704e-06,
+      "loss": 0.5472,
+      "step": 8729
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.94494712874036,
+      "learning_rate": 2.0016370626439454e-06,
+      "loss": 0.7062,
+      "step": 8730
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.128425975096584,
+      "learning_rate": 2.000577819503041e-06,
+      "loss": 0.5132,
+      "step": 8731
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.225973200281443,
+      "learning_rate": 1.9995187866085786e-06,
+      "loss": 0.4715,
+      "step": 8732
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.4534934469698375,
+      "learning_rate": 1.998459964034791e-06,
+      "loss": 0.5686,
+      "step": 8733
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.407785171635663,
+      "learning_rate": 1.9974013518558993e-06,
+      "loss": 0.615,
+      "step": 8734
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.2206323327255437,
+      "learning_rate": 1.996342950146103e-06,
+      "loss": 0.4733,
+      "step": 8735
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.6897605026812093,
+      "learning_rate": 1.995284758979594e-06,
+      "loss": 0.3708,
+      "step": 8736
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.7424303998891046,
+      "learning_rate": 1.9942267784305475e-06,
+      "loss": 0.6507,
+      "step": 8737
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.741602852999754,
+      "learning_rate": 1.9931690085731225e-06,
+      "loss": 0.5439,
+      "step": 8738
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.083622347875406,
+      "learning_rate": 1.9921114494814657e-06,
+      "loss": 0.8254,
+      "step": 8739
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.4516868864992776,
+      "learning_rate": 1.991054101229704e-06,
+      "loss": 0.5842,
+      "step": 8740
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.7441013518235735,
+      "learning_rate": 1.9899969638919554e-06,
+      "loss": 0.2529,
+      "step": 8741
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.86480139464636,
+      "learning_rate": 1.9889400375423196e-06,
+      "loss": 0.9784,
+      "step": 8742
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.910379725924154,
+      "learning_rate": 1.987883322254883e-06,
+      "loss": 0.7598,
+      "step": 8743
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.881084209076858,
+      "learning_rate": 1.9868268181037186e-06,
+      "loss": 0.6126,
+      "step": 8744
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.692959313734585,
+      "learning_rate": 1.9857705251628796e-06,
+      "loss": 0.8387,
+      "step": 8745
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.09280388117034,
+      "learning_rate": 1.98471444350641e-06,
+      "loss": 0.764,
+      "step": 8746
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.425010778688737,
+      "learning_rate": 1.9836585732083334e-06,
+      "loss": 0.9115,
+      "step": 8747
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6966358191954547,
+      "learning_rate": 1.982602914342664e-06,
+      "loss": 0.3736,
+      "step": 8748
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.9070587122298943,
+      "learning_rate": 1.9815474669833985e-06,
+      "loss": 0.7047,
+      "step": 8749
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.9935935008073384,
+      "learning_rate": 1.9804922312045193e-06,
+      "loss": 0.3165,
+      "step": 8750
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.733577568711758,
+      "learning_rate": 1.9794372070799955e-06,
+      "loss": 0.7414,
+      "step": 8751
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.562508336536102,
+      "learning_rate": 1.978382394683776e-06,
+      "loss": 0.8799,
+      "step": 8752
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.489307752013866,
+      "learning_rate": 1.9773277940898007e-06,
+      "loss": 0.9381,
+      "step": 8753
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.075482943719641,
+      "learning_rate": 1.9762734053719923e-06,
+      "loss": 0.9186,
+      "step": 8754
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.720913936362509,
+      "learning_rate": 1.975219228604259e-06,
+      "loss": 1.1617,
+      "step": 8755
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.289513198260462,
+      "learning_rate": 1.9741652638604952e-06,
+      "loss": 0.9923,
+      "step": 8756
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.238267357501408,
+      "learning_rate": 1.9731115112145765e-06,
+      "loss": 1.1233,
+      "step": 8757
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.7919538037923735,
+      "learning_rate": 1.9720579707403677e-06,
+      "loss": 1.1525,
+      "step": 8758
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.342877462199351,
+      "learning_rate": 1.9710046425117175e-06,
+      "loss": 0.6129,
+      "step": 8759
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.035221423131665,
+      "learning_rate": 1.9699515266024614e-06,
+      "loss": 0.9605,
+      "step": 8760
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.81820395917385,
+      "learning_rate": 1.968898623086415e-06,
+      "loss": 1.3903,
+      "step": 8761
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.519836837579079,
+      "learning_rate": 1.967845932037385e-06,
+      "loss": 0.8829,
+      "step": 8762
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.4543995131134193,
+      "learning_rate": 1.966793453529158e-06,
+      "loss": 0.7544,
+      "step": 8763
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9214886394824024,
+      "learning_rate": 1.9657411876355086e-06,
+      "loss": 0.2512,
+      "step": 8764
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.9448899170277545,
+      "learning_rate": 1.9646891344301972e-06,
+      "loss": 0.9378,
+      "step": 8765
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.2417350766028163,
+      "learning_rate": 1.9636372939869677e-06,
+      "loss": 0.4702,
+      "step": 8766
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.483562613077437,
+      "learning_rate": 1.9625856663795495e-06,
+      "loss": 1.1118,
+      "step": 8767
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.59874945966871,
+      "learning_rate": 1.9615342516816595e-06,
+      "loss": 0.5948,
+      "step": 8768
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.248160818585475,
+      "learning_rate": 1.9604830499669927e-06,
+      "loss": 1.2771,
+      "step": 8769
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.0465365777007856,
+      "learning_rate": 1.959432061309236e-06,
+      "loss": 0.6711,
+      "step": 8770
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.113986014902147,
+      "learning_rate": 1.9583812857820595e-06,
+      "loss": 0.8002,
+      "step": 8771
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.5081778890051227,
+      "learning_rate": 1.9573307234591177e-06,
+      "loss": 0.4122,
+      "step": 8772
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.12267022546443,
+      "learning_rate": 1.956280374414051e-06,
+      "loss": 0.1497,
+      "step": 8773
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.569631610718901,
+      "learning_rate": 1.9552302387204847e-06,
+      "loss": 0.4071,
+      "step": 8774
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.095527411898248,
+      "learning_rate": 1.9541803164520264e-06,
+      "loss": 0.6379,
+      "step": 8775
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.0478573325617098,
+      "learning_rate": 1.9531306076822738e-06,
+      "loss": 0.723,
+      "step": 8776
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.275622302591984,
+      "learning_rate": 1.952081112484804e-06,
+      "loss": 0.5652,
+      "step": 8777
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.985750730985547,
+      "learning_rate": 1.951031830933184e-06,
+      "loss": 0.5842,
+      "step": 8778
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.4450346943209933,
+      "learning_rate": 1.9499827631009644e-06,
+      "loss": 0.4955,
+      "step": 8779
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.2878789755454196,
+      "learning_rate": 1.948933909061678e-06,
+      "loss": 0.4982,
+      "step": 8780
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.9108891559935954,
+      "learning_rate": 1.9478852688888467e-06,
+      "loss": 0.4906,
+      "step": 8781
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.1479115211412,
+      "learning_rate": 1.946836842655975e-06,
+      "loss": 0.2868,
+      "step": 8782
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.778667943474665,
+      "learning_rate": 1.9457886304365533e-06,
+      "loss": 0.7543,
+      "step": 8783
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.356567930598804,
+      "learning_rate": 1.9447406323040562e-06,
+      "loss": 0.7398,
+      "step": 8784
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.540616076380403,
+      "learning_rate": 1.9436928483319467e-06,
+      "loss": 0.7212,
+      "step": 8785
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.11103211585045,
+      "learning_rate": 1.942645278593665e-06,
+      "loss": 0.8136,
+      "step": 8786
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.5630316440371184,
+      "learning_rate": 1.9415979231626443e-06,
+      "loss": 0.6716,
+      "step": 8787
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.1048582060855607,
+      "learning_rate": 1.940550782112299e-06,
+      "loss": 0.7179,
+      "step": 8788
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.122237785674999,
+      "learning_rate": 1.9395038555160285e-06,
+      "loss": 0.7376,
+      "step": 8789
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.965063265640028,
+      "learning_rate": 1.93845714344722e-06,
+      "loss": 0.3742,
+      "step": 8790
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.4587164331302445,
+      "learning_rate": 1.9374106459792406e-06,
+      "loss": 0.8145,
+      "step": 8791
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.4020388320522774,
+      "learning_rate": 1.9363643631854483e-06,
+      "loss": 0.6054,
+      "step": 8792
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.382568314360286,
+      "learning_rate": 1.9353182951391793e-06,
+      "loss": 0.8411,
+      "step": 8793
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.2978674612559438,
+      "learning_rate": 1.93427244191376e-06,
+      "loss": 0.5739,
+      "step": 8794
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.3205900275988025,
+      "learning_rate": 1.9332268035825006e-06,
+      "loss": 0.4264,
+      "step": 8795
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.372332606382506,
+      "learning_rate": 1.9321813802186972e-06,
+      "loss": 1.0462,
+      "step": 8796
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.77716927624133,
+      "learning_rate": 1.931136171895627e-06,
+      "loss": 1.1175,
+      "step": 8797
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.4186599091483756,
+      "learning_rate": 1.9300911786865544e-06,
+      "loss": 0.6359,
+      "step": 8798
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.371971117921947,
+      "learning_rate": 1.92904640066473e-06,
+      "loss": 1.0489,
+      "step": 8799
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.125585132588605,
+      "learning_rate": 1.9280018379033884e-06,
+      "loss": 0.8945,
+      "step": 8800
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.69143808714922,
+      "learning_rate": 1.926957490475748e-06,
+      "loss": 0.5875,
+      "step": 8801
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.550472159142619,
+      "learning_rate": 1.925913358455016e-06,
+      "loss": 0.8768,
+      "step": 8802
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.636115589262085,
+      "learning_rate": 1.9248694419143776e-06,
+      "loss": 0.741,
+      "step": 8803
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.041781653491779,
+      "learning_rate": 1.923825740927008e-06,
+      "loss": 0.507,
+      "step": 8804
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.4530873656903327,
+      "learning_rate": 1.922782255566066e-06,
+      "loss": 0.7788,
+      "step": 8805
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.45092229345249,
+      "learning_rate": 1.921738985904696e-06,
+      "loss": 0.9431,
+      "step": 8806
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.465998075197892,
+      "learning_rate": 1.9206959320160286e-06,
+      "loss": 1.1356,
+      "step": 8807
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.004891858662268,
+      "learning_rate": 1.9196530939731727e-06,
+      "loss": 1.044,
+      "step": 8808
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.106135688800069,
+      "learning_rate": 1.9186104718492315e-06,
+      "loss": 1.1551,
+      "step": 8809
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 7.0940084033406805,
+      "learning_rate": 1.917568065717284e-06,
+      "loss": 1.6003,
+      "step": 8810
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.167711096702554,
+      "learning_rate": 1.9165258756504003e-06,
+      "loss": 0.9334,
+      "step": 8811
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.4616970813945356,
+      "learning_rate": 1.9154839017216336e-06,
+      "loss": 0.4413,
+      "step": 8812
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.3795637550189124,
+      "learning_rate": 1.914442144004021e-06,
+      "loss": 0.3302,
+      "step": 8813
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.510062021432281,
+      "learning_rate": 1.913400602570588e-06,
+      "loss": 0.8562,
+      "step": 8814
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.57704329704265,
+      "learning_rate": 1.9123592774943383e-06,
+      "loss": 0.6834,
+      "step": 8815
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.4098023225599374,
+      "learning_rate": 1.911318168848265e-06,
+      "loss": 0.5838,
+      "step": 8816
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.032834390685839,
+      "learning_rate": 1.9102772767053467e-06,
+      "loss": 0.9763,
+      "step": 8817
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.883315323065235,
+      "learning_rate": 1.909236601138545e-06,
+      "loss": 0.8635,
+      "step": 8818
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.9993275066765177,
+      "learning_rate": 1.908196142220808e-06,
+      "loss": 0.3683,
+      "step": 8819
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.358880564580176,
+      "learning_rate": 1.9071559000250633e-06,
+      "loss": 0.8507,
+      "step": 8820
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.209064044779201,
+      "learning_rate": 1.906115874624231e-06,
+      "loss": 0.548,
+      "step": 8821
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.7333119421762047,
+      "learning_rate": 1.905076066091211e-06,
+      "loss": 0.6069,
+      "step": 8822
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.60706101308045,
+      "learning_rate": 1.904036474498891e-06,
+      "loss": 0.6224,
+      "step": 8823
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.546860315142606,
+      "learning_rate": 1.9029970999201387e-06,
+      "loss": 0.8429,
+      "step": 8824
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.8835937404593306,
+      "learning_rate": 1.9019579424278133e-06,
+      "loss": 0.6663,
+      "step": 8825
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.967240733651116,
+      "learning_rate": 1.900919002094752e-06,
+      "loss": 0.4519,
+      "step": 8826
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.90936097733532,
+      "learning_rate": 1.8998802789937815e-06,
+      "loss": 0.5688,
+      "step": 8827
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.5351784039807725,
+      "learning_rate": 1.898841773197711e-06,
+      "loss": 0.5924,
+      "step": 8828
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.772796655438025,
+      "learning_rate": 1.8978034847793364e-06,
+      "loss": 0.2402,
+      "step": 8829
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.2992077415812853,
+      "learning_rate": 1.8967654138114366e-06,
+      "loss": 0.5012,
+      "step": 8830
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.9005704020665144,
+      "learning_rate": 1.895727560366778e-06,
+      "loss": 0.4381,
+      "step": 8831
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.394153183451265,
+      "learning_rate": 1.8946899245181056e-06,
+      "loss": 0.7508,
+      "step": 8832
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.9745202934259174,
+      "learning_rate": 1.893652506338155e-06,
+      "loss": 0.6123,
+      "step": 8833
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.3302935435473335,
+      "learning_rate": 1.892615305899645e-06,
+      "loss": 1.2427,
+      "step": 8834
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.408078375774764,
+      "learning_rate": 1.8915783232752788e-06,
+      "loss": 0.557,
+      "step": 8835
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.289875698847576,
+      "learning_rate": 1.8905415585377458e-06,
+      "loss": 0.7968,
+      "step": 8836
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.004776925266335,
+      "learning_rate": 1.8895050117597152e-06,
+      "loss": 1.0907,
+      "step": 8837
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.6106359538734822,
+      "learning_rate": 1.8884686830138465e-06,
+      "loss": 0.7138,
+      "step": 8838
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.4332554602896295,
+      "learning_rate": 1.8874325723727831e-06,
+      "loss": 0.9122,
+      "step": 8839
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.869664636302528,
+      "learning_rate": 1.8863966799091492e-06,
+      "loss": 0.7228,
+      "step": 8840
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.979299170617078,
+      "learning_rate": 1.885361005695558e-06,
+      "loss": 0.653,
+      "step": 8841
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.048090927646827,
+      "learning_rate": 1.8843255498046065e-06,
+      "loss": 0.6582,
+      "step": 8842
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.7555027623016293,
+      "learning_rate": 1.8832903123088725e-06,
+      "loss": 0.7517,
+      "step": 8843
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.254266635051504,
+      "learning_rate": 1.882255293280924e-06,
+      "loss": 0.9816,
+      "step": 8844
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.299368885423742,
+      "learning_rate": 1.8812204927933108e-06,
+      "loss": 0.7299,
+      "step": 8845
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.2000511906844875,
+      "learning_rate": 1.8801859109185682e-06,
+      "loss": 0.5642,
+      "step": 8846
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.618016929122934,
+      "learning_rate": 1.879151547729216e-06,
+      "loss": 0.7709,
+      "step": 8847
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.2500007244119935,
+      "learning_rate": 1.87811740329776e-06,
+      "loss": 1.1265,
+      "step": 8848
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.7510578134177135,
+      "learning_rate": 1.8770834776966855e-06,
+      "loss": 0.682,
+      "step": 8849
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.617896038858609,
+      "learning_rate": 1.8760497709984683e-06,
+      "loss": 0.4952,
+      "step": 8850
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.9926633184568883,
+      "learning_rate": 1.8750162832755669e-06,
+      "loss": 0.6637,
+      "step": 8851
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.8388230579044857,
+      "learning_rate": 1.873983014600424e-06,
+      "loss": 0.4729,
+      "step": 8852
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.664516812490497,
+      "learning_rate": 1.8729499650454691e-06,
+      "loss": 0.9084,
+      "step": 8853
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.0321434265716807,
+      "learning_rate": 1.8719171346831106e-06,
+      "loss": 0.656,
+      "step": 8854
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.8632003393609278,
+      "learning_rate": 1.8708845235857498e-06,
+      "loss": 0.6406,
+      "step": 8855
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.9673550677903306,
+      "learning_rate": 1.8698521318257635e-06,
+      "loss": 0.5056,
+      "step": 8856
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.9194441145925065,
+      "learning_rate": 1.8688199594755208e-06,
+      "loss": 0.853,
+      "step": 8857
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.0424142971342847,
+      "learning_rate": 1.8677880066073718e-06,
+      "loss": 0.5882,
+      "step": 8858
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.4533604590139966,
+      "learning_rate": 1.866756273293654e-06,
+      "loss": 0.6266,
+      "step": 8859
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1540424959588524,
+      "learning_rate": 1.8657247596066834e-06,
+      "loss": 0.1113,
+      "step": 8860
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.937330664401966,
+      "learning_rate": 1.8646934656187671e-06,
+      "loss": 0.5732,
+      "step": 8861
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.884979876373652,
+      "learning_rate": 1.863662391402194e-06,
+      "loss": 0.6385,
+      "step": 8862
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.175052027188275,
+      "learning_rate": 1.862631537029238e-06,
+      "loss": 0.9419,
+      "step": 8863
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.946973992101543,
+      "learning_rate": 1.8616009025721572e-06,
+      "loss": 0.8249,
+      "step": 8864
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.1444148879339875,
+      "learning_rate": 1.860570488103196e-06,
+      "loss": 0.4323,
+      "step": 8865
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.590713657852092,
+      "learning_rate": 1.85954029369458e-06,
+      "loss": 0.7788,
+      "step": 8866
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.320357196050719,
+      "learning_rate": 1.858510319418521e-06,
+      "loss": 0.1759,
+      "step": 8867
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.33373464431892,
+      "learning_rate": 1.8574805653472178e-06,
+      "loss": 0.7957,
+      "step": 8868
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.852412842501961,
+      "learning_rate": 1.8564510315528517e-06,
+      "loss": 0.6909,
+      "step": 8869
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0505451876407883,
+      "learning_rate": 1.8554217181075862e-06,
+      "loss": 0.3888,
+      "step": 8870
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.758074011530697,
+      "learning_rate": 1.8543926250835749e-06,
+      "loss": 0.7105,
+      "step": 8871
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.977808151424315,
+      "learning_rate": 1.8533637525529485e-06,
+      "loss": 1.0953,
+      "step": 8872
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.328100722581435,
+      "learning_rate": 1.8523351005878293e-06,
+      "loss": 1.1907,
+      "step": 8873
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.510754729106961,
+      "learning_rate": 1.8513066692603204e-06,
+      "loss": 0.4291,
+      "step": 8874
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.8439070213468214,
+      "learning_rate": 1.8502784586425116e-06,
+      "loss": 0.5486,
+      "step": 8875
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.4516253149991885,
+      "learning_rate": 1.849250468806476e-06,
+      "loss": 0.8449,
+      "step": 8876
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.592603633187335,
+      "learning_rate": 1.8482226998242692e-06,
+      "loss": 0.8437,
+      "step": 8877
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.0296036545205443,
+      "learning_rate": 1.8471951517679348e-06,
+      "loss": 0.4635,
+      "step": 8878
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8423645483049949,
+      "learning_rate": 1.846167824709499e-06,
+      "loss": 0.2186,
+      "step": 8879
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.848669674508421,
+      "learning_rate": 1.845140718720973e-06,
+      "loss": 1.1526,
+      "step": 8880
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.037018099967917,
+      "learning_rate": 1.844113833874353e-06,
+      "loss": 0.8278,
+      "step": 8881
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.3710378600880455,
+      "learning_rate": 1.8430871702416198e-06,
+      "loss": 0.8264,
+      "step": 8882
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.301263634150496,
+      "learning_rate": 1.8420607278947362e-06,
+      "loss": 0.791,
+      "step": 8883
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0243568108169714,
+      "learning_rate": 1.8410345069056517e-06,
+      "loss": 0.3202,
+      "step": 8884
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.9360226309033695,
+      "learning_rate": 1.840008507346302e-06,
+      "loss": 0.4884,
+      "step": 8885
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.49497390112113,
+      "learning_rate": 1.838982729288602e-06,
+      "loss": 0.9013,
+      "step": 8886
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.954276049008714,
+      "learning_rate": 1.8379571728044559e-06,
+      "loss": 0.6735,
+      "step": 8887
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.947073723266408,
+      "learning_rate": 1.8369318379657526e-06,
+      "loss": 0.4891,
+      "step": 8888
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.085480599916667,
+      "learning_rate": 1.8359067248443602e-06,
+      "loss": 0.3471,
+      "step": 8889
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.0392365817723133,
+      "learning_rate": 1.8348818335121355e-06,
+      "loss": 0.6283,
+      "step": 8890
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.6103342856232805,
+      "learning_rate": 1.8338571640409203e-06,
+      "loss": 0.9192,
+      "step": 8891
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.587916812618091,
+      "learning_rate": 1.8328327165025384e-06,
+      "loss": 0.7832,
+      "step": 8892
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.671478660931107,
+      "learning_rate": 1.8318084909687995e-06,
+      "loss": 0.8664,
+      "step": 8893
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.428929659668447,
+      "learning_rate": 1.8307844875114993e-06,
+      "loss": 0.3476,
+      "step": 8894
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.1948707359979056,
+      "learning_rate": 1.8297607062024125e-06,
+      "loss": 0.717,
+      "step": 8895
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.149110950463409,
+      "learning_rate": 1.828737147113303e-06,
+      "loss": 0.7536,
+      "step": 8896
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.8963364537499205,
+      "learning_rate": 1.827713810315918e-06,
+      "loss": 0.4259,
+      "step": 8897
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.3710584652754862,
+      "learning_rate": 1.8266906958819892e-06,
+      "loss": 0.5665,
+      "step": 8898
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.246601455389945,
+      "learning_rate": 1.8256678038832342e-06,
+      "loss": 0.7157,
+      "step": 8899
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.7252475510620577,
+      "learning_rate": 1.8246451343913497e-06,
+      "loss": 0.4024,
+      "step": 8900
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.2154462309831513,
+      "learning_rate": 1.8236226874780233e-06,
+      "loss": 0.3761,
+      "step": 8901
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.463399088426534,
+      "learning_rate": 1.822600463214922e-06,
+      "loss": 0.7889,
+      "step": 8902
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.617610782224787,
+      "learning_rate": 1.8215784616736993e-06,
+      "loss": 1.2148,
+      "step": 8903
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.454984366704652,
+      "learning_rate": 1.8205566829259942e-06,
+      "loss": 0.813,
+      "step": 8904
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.205670702293653,
+      "learning_rate": 1.8195351270434303e-06,
+      "loss": 1.1278,
+      "step": 8905
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.285411479543468,
+      "learning_rate": 1.8185137940976111e-06,
+      "loss": 0.5392,
+      "step": 8906
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.7137924721743736,
+      "learning_rate": 1.8174926841601294e-06,
+      "loss": 0.6425,
+      "step": 8907
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.725552627995343,
+      "learning_rate": 1.816471797302559e-06,
+      "loss": 1.1603,
+      "step": 8908
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.415861053345675,
+      "learning_rate": 1.8154511335964619e-06,
+      "loss": 0.5348,
+      "step": 8909
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.1246213158237968,
+      "learning_rate": 1.8144306931133809e-06,
+      "loss": 0.2999,
+      "step": 8910
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.7435419776111294,
+      "learning_rate": 1.8134104759248461e-06,
+      "loss": 0.369,
+      "step": 8911
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.9848728694727753,
+      "learning_rate": 1.8123904821023675e-06,
+      "loss": 0.8606,
+      "step": 8912
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.2749438745423083,
+      "learning_rate": 1.8113707117174433e-06,
+      "loss": 0.6244,
+      "step": 8913
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.182642877270864,
+      "learning_rate": 1.8103511648415556e-06,
+      "loss": 1.3483,
+      "step": 8914
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.0969152987592725,
+      "learning_rate": 1.8093318415461698e-06,
+      "loss": 0.8401,
+      "step": 8915
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.307504013633494,
+      "learning_rate": 1.8083127419027375e-06,
+      "loss": 1.1576,
+      "step": 8916
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.888400058596468,
+      "learning_rate": 1.80729386598269e-06,
+      "loss": 1.08,
+      "step": 8917
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.491074432520151,
+      "learning_rate": 1.8062752138574497e-06,
+      "loss": 1.6848,
+      "step": 8918
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.445135276899701,
+      "learning_rate": 1.805256785598416e-06,
+      "loss": 0.8892,
+      "step": 8919
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.234531444079373,
+      "learning_rate": 1.804238581276978e-06,
+      "loss": 0.8545,
+      "step": 8920
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.143084059745439,
+      "learning_rate": 1.8032206009645077e-06,
+      "loss": 0.7199,
+      "step": 8921
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.042827522492736,
+      "learning_rate": 1.8022028447323619e-06,
+      "loss": 0.6258,
+      "step": 8922
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.826884504598852,
+      "learning_rate": 1.8011853126518786e-06,
+      "loss": 0.5258,
+      "step": 8923
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.623346055293667,
+      "learning_rate": 1.8001680047943836e-06,
+      "loss": 1.0562,
+      "step": 8924
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.9842091905897403,
+      "learning_rate": 1.7991509212311858e-06,
+      "loss": 0.219,
+      "step": 8925
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.48664500069927,
+      "learning_rate": 1.798134062033578e-06,
+      "loss": 0.7726,
+      "step": 8926
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.1405945631380128,
+      "learning_rate": 1.7971174272728381e-06,
+      "loss": 0.3926,
+      "step": 8927
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.0999316996457,
+      "learning_rate": 1.7961010170202293e-06,
+      "loss": 0.9603,
+      "step": 8928
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.7755183810759707,
+      "learning_rate": 1.7950848313469944e-06,
+      "loss": 0.7153,
+      "step": 8929
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.424073871148489,
+      "learning_rate": 1.7940688703243641e-06,
+      "loss": 0.4287,
+      "step": 8930
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.195063690892858,
+      "learning_rate": 1.7930531340235546e-06,
+      "loss": 0.3746,
+      "step": 8931
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.073234184739771,
+      "learning_rate": 1.7920376225157648e-06,
+      "loss": 0.801,
+      "step": 8932
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.9950993521702953,
+      "learning_rate": 1.7910223358721751e-06,
+      "loss": 1.2285,
+      "step": 8933
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.8398240441967366,
+      "learning_rate": 1.7900072741639557e-06,
+      "loss": 0.4967,
+      "step": 8934
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.9808438863197506,
+      "learning_rate": 1.7889924374622552e-06,
+      "loss": 0.7187,
+      "step": 8935
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.3805642712079598,
+      "learning_rate": 1.7879778258382103e-06,
+      "loss": 0.7522,
+      "step": 8936
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.326232272044018,
+      "learning_rate": 1.786963439362941e-06,
+      "loss": 1.0033,
+      "step": 8937
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0476375259141353,
+      "learning_rate": 1.7859492781075511e-06,
+      "loss": 0.3842,
+      "step": 8938
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.872031597441457,
+      "learning_rate": 1.7849353421431316e-06,
+      "loss": 0.7099,
+      "step": 8939
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.912564714887852,
+      "learning_rate": 1.7839216315407498e-06,
+      "loss": 0.8623,
+      "step": 8940
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.953381229946058,
+      "learning_rate": 1.782908146371466e-06,
+      "loss": 0.7803,
+      "step": 8941
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.460200700473453,
+      "learning_rate": 1.7818948867063201e-06,
+      "loss": 0.7107,
+      "step": 8942
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.176029544680021,
+      "learning_rate": 1.780881852616338e-06,
+      "loss": 0.9858,
+      "step": 8943
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.753092611167436,
+      "learning_rate": 1.7798690441725275e-06,
+      "loss": 0.3707,
+      "step": 8944
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.948615487037133,
+      "learning_rate": 1.7788564614458853e-06,
+      "loss": 0.9159,
+      "step": 8945
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.8964014081785283,
+      "learning_rate": 1.7778441045073846e-06,
+      "loss": 0.5617,
+      "step": 8946
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.4158829649665705,
+      "learning_rate": 1.7768319734279894e-06,
+      "loss": 0.5745,
+      "step": 8947
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.9495655938363736,
+      "learning_rate": 1.775820068278647e-06,
+      "loss": 0.8235,
+      "step": 8948
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.3442309344872614,
+      "learning_rate": 1.7748083891302847e-06,
+      "loss": 0.2646,
+      "step": 8949
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.7840108594247086,
+      "learning_rate": 1.7737969360538187e-06,
+      "loss": 0.5071,
+      "step": 8950
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.488174423512263,
+      "learning_rate": 1.7727857091201477e-06,
+      "loss": 0.986,
+      "step": 8951
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.4000460221903386,
+      "learning_rate": 1.771774708400153e-06,
+      "loss": 0.6213,
+      "step": 8952
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.7478045033289273,
+      "learning_rate": 1.7707639339647015e-06,
+      "loss": 0.8226,
+      "step": 8953
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.02873919600009,
+      "learning_rate": 1.7697533858846444e-06,
+      "loss": 0.6703,
+      "step": 8954
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.461477367730992,
+      "learning_rate": 1.7687430642308167e-06,
+      "loss": 0.5836,
+      "step": 8955
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.9165398389129913,
+      "learning_rate": 1.7677329690740397e-06,
+      "loss": 0.7629,
+      "step": 8956
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.06870891643253,
+      "learning_rate": 1.7667231004851132e-06,
+      "loss": 0.8262,
+      "step": 8957
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.8948921154450318,
+      "learning_rate": 1.7657134585348257e-06,
+      "loss": 0.3847,
+      "step": 8958
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.278096040896495,
+      "learning_rate": 1.7647040432939494e-06,
+      "loss": 1.0448,
+      "step": 8959
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.350081697197041,
+      "learning_rate": 1.7636948548332394e-06,
+      "loss": 0.447,
+      "step": 8960
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.743675239557907,
+      "learning_rate": 1.762685893223436e-06,
+      "loss": 0.8835,
+      "step": 8961
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.1219408922093415,
+      "learning_rate": 1.7616771585352638e-06,
+      "loss": 0.6029,
+      "step": 8962
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.932424093664737,
+      "learning_rate": 1.7606686508394278e-06,
+      "loss": 0.4565,
+      "step": 8963
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.350284755062462,
+      "learning_rate": 1.759660370206624e-06,
+      "loss": 0.8739,
+      "step": 8964
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8392207856522242,
+      "learning_rate": 1.7586523167075243e-06,
+      "loss": 0.3384,
+      "step": 8965
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.047881033882093,
+      "learning_rate": 1.7576444904127909e-06,
+      "loss": 0.3368,
+      "step": 8966
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.5023817371920622,
+      "learning_rate": 1.7566368913930677e-06,
+      "loss": 0.5931,
+      "step": 8967
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.0312836485738424,
+      "learning_rate": 1.7556295197189849e-06,
+      "loss": 0.4801,
+      "step": 8968
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.760566696496184,
+      "learning_rate": 1.754622375461152e-06,
+      "loss": 0.8811,
+      "step": 8969
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.2345251307797587,
+      "learning_rate": 1.753615458690166e-06,
+      "loss": 0.5812,
+      "step": 8970
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 6.045578240916088,
+      "learning_rate": 1.7526087694766086e-06,
+      "loss": 1.1997,
+      "step": 8971
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.481839428642073,
+      "learning_rate": 1.7516023078910438e-06,
+      "loss": 0.9402,
+      "step": 8972
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.04590137193262,
+      "learning_rate": 1.7505960740040196e-06,
+      "loss": 0.7429,
+      "step": 8973
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2468634629602602,
+      "learning_rate": 1.749590067886071e-06,
+      "loss": 0.1524,
+      "step": 8974
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.637452769824487,
+      "learning_rate": 1.7485842896077116e-06,
+      "loss": 0.8302,
+      "step": 8975
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.7895625674392375,
+      "learning_rate": 1.7475787392394427e-06,
+      "loss": 0.6474,
+      "step": 8976
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.601124604564186,
+      "learning_rate": 1.7465734168517501e-06,
+      "loss": 0.2956,
+      "step": 8977
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.4072381897503887,
+      "learning_rate": 1.7455683225151037e-06,
+      "loss": 0.9623,
+      "step": 8978
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.9706115422857633,
+      "learning_rate": 1.7445634562999526e-06,
+      "loss": 0.6604,
+      "step": 8979
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.308798596479263,
+      "learning_rate": 1.7435588182767371e-06,
+      "loss": 0.5646,
+      "step": 8980
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.188892167150614,
+      "learning_rate": 1.7425544085158747e-06,
+      "loss": 0.3848,
+      "step": 8981
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.398272377558863,
+      "learning_rate": 1.741550227087772e-06,
+      "loss": 0.703,
+      "step": 8982
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.811819056468078,
+      "learning_rate": 1.7405462740628177e-06,
+      "loss": 1.0272,
+      "step": 8983
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.4139978187096194,
+      "learning_rate": 1.7395425495113838e-06,
+      "loss": 0.3504,
+      "step": 8984
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.703841869876057,
+      "learning_rate": 1.7385390535038299e-06,
+      "loss": 0.5092,
+      "step": 8985
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.341951956627426,
+      "learning_rate": 1.7375357861104924e-06,
+      "loss": 0.9809,
+      "step": 8986
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.687216573165631,
+      "learning_rate": 1.7365327474016979e-06,
+      "loss": 0.5291,
+      "step": 8987
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.2601727656097803,
+      "learning_rate": 1.7355299374477558e-06,
+      "loss": 0.6321,
+      "step": 8988
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.022852657978894,
+      "learning_rate": 1.7345273563189575e-06,
+      "loss": 0.9331,
+      "step": 8989
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.4543615795409,
+      "learning_rate": 1.7335250040855805e-06,
+      "loss": 0.5367,
+      "step": 8990
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.405105561209966,
+      "learning_rate": 1.7325228808178862e-06,
+      "loss": 0.9613,
+      "step": 8991
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.899093810149087,
+      "learning_rate": 1.7315209865861165e-06,
+      "loss": 0.7388,
+      "step": 8992
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.012980749679616,
+      "learning_rate": 1.730519321460501e-06,
+      "loss": 0.9296,
+      "step": 8993
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.868724219782896,
+      "learning_rate": 1.7295178855112537e-06,
+      "loss": 1.0129,
+      "step": 8994
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.921950322463082,
+      "learning_rate": 1.7285166788085683e-06,
+      "loss": 0.7011,
+      "step": 8995
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.467594249950734,
+      "learning_rate": 1.7275157014226274e-06,
+      "loss": 0.6228,
+      "step": 8996
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.208537067500457,
+      "learning_rate": 1.7265149534235925e-06,
+      "loss": 0.8779,
+      "step": 8997
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.587388261769327,
+      "learning_rate": 1.7255144348816134e-06,
+      "loss": 0.7871,
+      "step": 8998
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.725501053167962,
+      "learning_rate": 1.7245141458668213e-06,
+      "loss": 0.6805,
+      "step": 8999
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.741787604017313,
+      "learning_rate": 1.7235140864493327e-06,
+      "loss": 1.0489,
+      "step": 9000
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.2071984650447534,
+      "learning_rate": 1.7225142566992476e-06,
+      "loss": 0.8655,
+      "step": 9001
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.985035641680996,
+      "learning_rate": 1.7215146566866508e-06,
+      "loss": 0.8706,
+      "step": 9002
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.47775958261579,
+      "learning_rate": 1.7205152864816071e-06,
+      "loss": 0.828,
+      "step": 9003
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.7970700653958174,
+      "learning_rate": 1.7195161461541692e-06,
+      "loss": 0.3933,
+      "step": 9004
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.013903958013123,
+      "learning_rate": 1.7185172357743729e-06,
+      "loss": 0.7536,
+      "step": 9005
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.2442239307638188,
+      "learning_rate": 1.7175185554122375e-06,
+      "loss": 0.6861,
+      "step": 9006
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.2467894939410833,
+      "learning_rate": 1.7165201051377657e-06,
+      "loss": 0.3923,
+      "step": 9007
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.5533531296390075,
+      "learning_rate": 1.7155218850209465e-06,
+      "loss": 0.92,
+      "step": 9008
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.097363179597376,
+      "learning_rate": 1.7145238951317473e-06,
+      "loss": 0.688,
+      "step": 9009
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6374955130177864,
+      "learning_rate": 1.7135261355401246e-06,
+      "loss": 0.2497,
+      "step": 9010
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.751931732962066,
+      "learning_rate": 1.712528606316019e-06,
+      "loss": 1.0268,
+      "step": 9011
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3351512488767607,
+      "learning_rate": 1.7115313075293488e-06,
+      "loss": 0.2148,
+      "step": 9012
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.28499191945056,
+      "learning_rate": 1.710534239250023e-06,
+      "loss": 0.6547,
+      "step": 9013
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.7785812157892913,
+      "learning_rate": 1.7095374015479326e-06,
+      "loss": 0.6309,
+      "step": 9014
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.166814066008341,
+      "learning_rate": 1.7085407944929488e-06,
+      "loss": 0.8211,
+      "step": 9015
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.734193097427948,
+      "learning_rate": 1.7075444181549305e-06,
+      "loss": 0.92,
+      "step": 9016
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.9215893062088756,
+      "learning_rate": 1.7065482726037196e-06,
+      "loss": 0.4628,
+      "step": 9017
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 6.093119807275033,
+      "learning_rate": 1.7055523579091422e-06,
+      "loss": 1.0065,
+      "step": 9018
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.323869904369412,
+      "learning_rate": 1.704556674141008e-06,
+      "loss": 0.7332,
+      "step": 9019
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5696067383535437,
+      "learning_rate": 1.7035612213691083e-06,
+      "loss": 0.3485,
+      "step": 9020
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.530204008821355,
+      "learning_rate": 1.7025659996632198e-06,
+      "loss": 0.4463,
+      "step": 9021
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.2497473771022154,
+      "learning_rate": 1.7015710090931047e-06,
+      "loss": 0.5451,
+      "step": 9022
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9079734640877182,
+      "learning_rate": 1.7005762497285078e-06,
+      "loss": 0.3221,
+      "step": 9023
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.956229764069996,
+      "learning_rate": 1.6995817216391559e-06,
+      "loss": 0.7575,
+      "step": 9024
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.8665180173730653,
+      "learning_rate": 1.698587424894763e-06,
+      "loss": 0.8057,
+      "step": 9025
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.7188981255366276,
+      "learning_rate": 1.6975933595650229e-06,
+      "loss": 0.6104,
+      "step": 9026
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9620310746669432,
+      "learning_rate": 1.6965995257196177e-06,
+      "loss": 0.3023,
+      "step": 9027
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.13494334777588,
+      "learning_rate": 1.6956059234282079e-06,
+      "loss": 0.4358,
+      "step": 9028
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.4396347140855688,
+      "learning_rate": 1.6946125527604419e-06,
+      "loss": 0.4067,
+      "step": 9029
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.7671237129759287,
+      "learning_rate": 1.6936194137859508e-06,
+      "loss": 0.7057,
+      "step": 9030
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.9593379420039243,
+      "learning_rate": 1.6926265065743507e-06,
+      "loss": 0.47,
+      "step": 9031
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.659273991675345,
+      "learning_rate": 1.6916338311952373e-06,
+      "loss": 0.6953,
+      "step": 9032
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.132124704254727,
+      "learning_rate": 1.6906413877181948e-06,
+      "loss": 0.7218,
+      "step": 9033
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.13717100686943,
+      "learning_rate": 1.6896491762127882e-06,
+      "loss": 0.6954,
+      "step": 9034
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.821745647616559,
+      "learning_rate": 1.6886571967485677e-06,
+      "loss": 0.7649,
+      "step": 9035
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.6486101795414934,
+      "learning_rate": 1.6876654493950666e-06,
+      "loss": 0.7128,
+      "step": 9036
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.829340506404127,
+      "learning_rate": 1.6866739342218042e-06,
+      "loss": 0.7762,
+      "step": 9037
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.418065515804254,
+      "learning_rate": 1.6856826512982772e-06,
+      "loss": 1.0962,
+      "step": 9038
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.282175126871541,
+      "learning_rate": 1.6846916006939724e-06,
+      "loss": 0.1717,
+      "step": 9039
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.689916488916796,
+      "learning_rate": 1.6837007824783586e-06,
+      "loss": 0.6698,
+      "step": 9040
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.710826508344211,
+      "learning_rate": 1.6827101967208887e-06,
+      "loss": 1.0579,
+      "step": 9041
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.8698950145246562,
+      "learning_rate": 1.6817198434909954e-06,
+      "loss": 0.7306,
+      "step": 9042
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.822068675912592,
+      "learning_rate": 1.6807297228581016e-06,
+      "loss": 0.9391,
+      "step": 9043
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.440054907992409,
+      "learning_rate": 1.6797398348916073e-06,
+      "loss": 0.8555,
+      "step": 9044
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.6654135564767585,
+      "learning_rate": 1.6787501796609001e-06,
+      "loss": 0.3411,
+      "step": 9045
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.0646110765873495,
+      "learning_rate": 1.6777607572353516e-06,
+      "loss": 0.3515,
+      "step": 9046
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.809911256440361,
+      "learning_rate": 1.676771567684316e-06,
+      "loss": 0.9619,
+      "step": 9047
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.248089390765532,
+      "learning_rate": 1.675782611077132e-06,
+      "loss": 0.9298,
+      "step": 9048
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.046653300658506,
+      "learning_rate": 1.6747938874831182e-06,
+      "loss": 1.0072,
+      "step": 9049
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.9210787418348643,
+      "learning_rate": 1.6738053969715818e-06,
+      "loss": 0.6771,
+      "step": 9050
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.7392379153531135,
+      "learning_rate": 1.672817139611811e-06,
+      "loss": 1.1881,
+      "step": 9051
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.729634737365383,
+      "learning_rate": 1.6718291154730792e-06,
+      "loss": 0.8296,
+      "step": 9052
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 7.292886378742193,
+      "learning_rate": 1.6708413246246418e-06,
+      "loss": 1.7863,
+      "step": 9053
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.091853622035211,
+      "learning_rate": 1.6698537671357406e-06,
+      "loss": 0.5248,
+      "step": 9054
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.034897188619175,
+      "learning_rate": 1.6688664430755964e-06,
+      "loss": 0.3249,
+      "step": 9055
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.212051184388152,
+      "learning_rate": 1.6678793525134167e-06,
+      "loss": 0.5452,
+      "step": 9056
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.587073637312639,
+      "learning_rate": 1.6668924955183952e-06,
+      "loss": 1.2495,
+      "step": 9057
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.4326452585038636,
+      "learning_rate": 1.665905872159702e-06,
+      "loss": 0.6913,
+      "step": 9058
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.64344670227028,
+      "learning_rate": 1.6649194825064991e-06,
+      "loss": 1.0694,
+      "step": 9059
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.987460457256301,
+      "learning_rate": 1.6639333266279244e-06,
+      "loss": 0.6845,
+      "step": 9060
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.392028734809809,
+      "learning_rate": 1.662947404593105e-06,
+      "loss": 0.7027,
+      "step": 9061
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.08408362057497,
+      "learning_rate": 1.6619617164711493e-06,
+      "loss": 0.6966,
+      "step": 9062
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.3109251101137485,
+      "learning_rate": 1.6609762623311504e-06,
+      "loss": 0.7128,
+      "step": 9063
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.243086851835061,
+      "learning_rate": 1.659991042242184e-06,
+      "loss": 1.1725,
+      "step": 9064
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.204599777941349,
+      "learning_rate": 1.6590060562733111e-06,
+      "loss": 0.5851,
+      "step": 9065
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.9517819510821637,
+      "learning_rate": 1.6580213044935723e-06,
+      "loss": 0.5402,
+      "step": 9066
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.053050927242831,
+      "learning_rate": 1.6570367869719955e-06,
+      "loss": 0.5968,
+      "step": 9067
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.239009288234037,
+      "learning_rate": 1.656052503777591e-06,
+      "loss": 0.7625,
+      "step": 9068
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.87110788114826,
+      "learning_rate": 1.6550684549793539e-06,
+      "loss": 0.5221,
+      "step": 9069
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.766413422379634,
+      "learning_rate": 1.6540846406462602e-06,
+      "loss": 0.4895,
+      "step": 9070
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.844991604320573,
+      "learning_rate": 1.6531010608472736e-06,
+      "loss": 1.3181,
+      "step": 9071
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.687473886208534,
+      "learning_rate": 1.6521177156513351e-06,
+      "loss": 0.9449,
+      "step": 9072
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.255207412004284,
+      "learning_rate": 1.6511346051273768e-06,
+      "loss": 0.3351,
+      "step": 9073
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.047676731344719,
+      "learning_rate": 1.6501517293443064e-06,
+      "loss": 0.8867,
+      "step": 9074
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.065878701202547,
+      "learning_rate": 1.6491690883710209e-06,
+      "loss": 0.6451,
+      "step": 9075
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.5443967112541754,
+      "learning_rate": 1.6481866822763997e-06,
+      "loss": 0.4899,
+      "step": 9076
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.9996127676580624,
+      "learning_rate": 1.6472045111293072e-06,
+      "loss": 0.6043,
+      "step": 9077
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.0559692397812004,
+      "learning_rate": 1.6462225749985845e-06,
+      "loss": 0.8296,
+      "step": 9078
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.266144230809973,
+      "learning_rate": 1.645240873953064e-06,
+      "loss": 0.8095,
+      "step": 9079
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.841030181663696,
+      "learning_rate": 1.6442594080615581e-06,
+      "loss": 0.6333,
+      "step": 9080
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.7709648439262544,
+      "learning_rate": 1.6432781773928636e-06,
+      "loss": 0.8523,
+      "step": 9081
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.394573663876398,
+      "learning_rate": 1.6422971820157623e-06,
+      "loss": 0.791,
+      "step": 9082
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.9444633502066395,
+      "learning_rate": 1.6413164219990136e-06,
+      "loss": 0.8428,
+      "step": 9083
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9488231025805063,
+      "learning_rate": 1.640335897411367e-06,
+      "loss": 0.3488,
+      "step": 9084
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9429043504182903,
+      "learning_rate": 1.6393556083215528e-06,
+      "loss": 0.3321,
+      "step": 9085
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.459417834868013,
+      "learning_rate": 1.6383755547982844e-06,
+      "loss": 0.5809,
+      "step": 9086
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.792858688546865,
+      "learning_rate": 1.6373957369102616e-06,
+      "loss": 0.8116,
+      "step": 9087
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.1192732924982023,
+      "learning_rate": 1.636416154726162e-06,
+      "loss": 0.4053,
+      "step": 9088
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.621300593467316,
+      "learning_rate": 1.6354368083146532e-06,
+      "loss": 0.574,
+      "step": 9089
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.7491013378790123,
+      "learning_rate": 1.63445769774438e-06,
+      "loss": 0.5252,
+      "step": 9090
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.417582731462544,
+      "learning_rate": 1.6334788230839753e-06,
+      "loss": 0.9197,
+      "step": 9091
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.077667303709065,
+      "learning_rate": 1.6325001844020538e-06,
+      "loss": 1.2808,
+      "step": 9092
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.9993091285089273,
+      "learning_rate": 1.6315217817672142e-06,
+      "loss": 0.558,
+      "step": 9093
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.99726956482455,
+      "learning_rate": 1.6305436152480392e-06,
+      "loss": 0.8917,
+      "step": 9094
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.970521738608599,
+      "learning_rate": 1.6295656849130914e-06,
+      "loss": 0.3274,
+      "step": 9095
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.221441371096554,
+      "learning_rate": 1.628587990830921e-06,
+      "loss": 0.6971,
+      "step": 9096
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.870364589941543,
+      "learning_rate": 1.6276105330700599e-06,
+      "loss": 0.5622,
+      "step": 9097
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.1773326866114715,
+      "learning_rate": 1.6266333116990242e-06,
+      "loss": 0.6384,
+      "step": 9098
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.2829276113017065,
+      "learning_rate": 1.6256563267863135e-06,
+      "loss": 0.348,
+      "step": 9099
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.420048364811174,
+      "learning_rate": 1.6246795784004076e-06,
+      "loss": 0.5856,
+      "step": 9100
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.7505987876192113,
+      "learning_rate": 1.6237030666097736e-06,
+      "loss": 0.7085,
+      "step": 9101
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.8063053381010703,
+      "learning_rate": 1.6227267914828615e-06,
+      "loss": 1.0319,
+      "step": 9102
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.804431995865752,
+      "learning_rate": 1.6217507530881048e-06,
+      "loss": 0.5913,
+      "step": 9103
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.687685126129331,
+      "learning_rate": 1.6207749514939164e-06,
+      "loss": 0.7277,
+      "step": 9104
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.062244202441285,
+      "learning_rate": 1.6197993867686973e-06,
+      "loss": 0.7456,
+      "step": 9105
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.3369190904442183,
+      "learning_rate": 1.6188240589808325e-06,
+      "loss": 0.4212,
+      "step": 9106
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.1709524686338915,
+      "learning_rate": 1.6178489681986842e-06,
+      "loss": 0.9092,
+      "step": 9107
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.3744359543461533,
+      "learning_rate": 1.616874114490604e-06,
+      "loss": 0.6864,
+      "step": 9108
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.58724493813051,
+      "learning_rate": 1.6158994979249255e-06,
+      "loss": 0.4519,
+      "step": 9109
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.6247163893013767,
+      "learning_rate": 1.6149251185699643e-06,
+      "loss": 0.237,
+      "step": 9110
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.723515891720296,
+      "learning_rate": 1.613950976494022e-06,
+      "loss": 0.655,
+      "step": 9111
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.9683408496964163,
+      "learning_rate": 1.6129770717653781e-06,
+      "loss": 0.4724,
+      "step": 9112
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.158516849075152,
+      "learning_rate": 1.6120034044523015e-06,
+      "loss": 0.9067,
+      "step": 9113
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.8198622674689116,
+      "learning_rate": 1.6110299746230419e-06,
+      "loss": 1.0292,
+      "step": 9114
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.7482896303080935,
+      "learning_rate": 1.6100567823458319e-06,
+      "loss": 0.5969,
+      "step": 9115
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.44067565129291,
+      "learning_rate": 1.6090838276888882e-06,
+      "loss": 0.9371,
+      "step": 9116
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.450233594803811,
+      "learning_rate": 1.6081111107204127e-06,
+      "loss": 0.4965,
+      "step": 9117
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.5899593889008092,
+      "learning_rate": 1.6071386315085851e-06,
+      "loss": 0.5021,
+      "step": 9118
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9734439489250912,
+      "learning_rate": 1.606166390121574e-06,
+      "loss": 0.3827,
+      "step": 9119
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.34945668400547,
+      "learning_rate": 1.60519438662753e-06,
+      "loss": 0.4631,
+      "step": 9120
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.239324985671788,
+      "learning_rate": 1.6042226210945838e-06,
+      "loss": 0.5508,
+      "step": 9121
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.7334163443070567,
+      "learning_rate": 1.6032510935908551e-06,
+      "loss": 0.5746,
+      "step": 9122
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.6674181829994215,
+      "learning_rate": 1.6022798041844407e-06,
+      "loss": 0.7384,
+      "step": 9123
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.929516148507866,
+      "learning_rate": 1.6013087529434247e-06,
+      "loss": 1.1649,
+      "step": 9124
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.765302163788124,
+      "learning_rate": 1.6003379399358742e-06,
+      "loss": 0.3319,
+      "step": 9125
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.751803140552971,
+      "learning_rate": 1.5993673652298386e-06,
+      "loss": 0.8956,
+      "step": 9126
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.303559736955207,
+      "learning_rate": 1.5983970288933509e-06,
+      "loss": 1.1144,
+      "step": 9127
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.0912639283894245,
+      "learning_rate": 1.5974269309944296e-06,
+      "loss": 0.6042,
+      "step": 9128
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.236076518778019,
+      "learning_rate": 1.5964570716010708e-06,
+      "loss": 0.7398,
+      "step": 9129
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.891067061234153,
+      "learning_rate": 1.595487450781259e-06,
+      "loss": 0.8603,
+      "step": 9130
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.186974074359808,
+      "learning_rate": 1.5945180686029598e-06,
+      "loss": 0.8253,
+      "step": 9131
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.4969354792471243,
+      "learning_rate": 1.593548925134124e-06,
+      "loss": 0.3732,
+      "step": 9132
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.246678029315557,
+      "learning_rate": 1.5925800204426833e-06,
+      "loss": 0.9658,
+      "step": 9133
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.004737255489296,
+      "learning_rate": 1.5916113545965562e-06,
+      "loss": 0.6139,
+      "step": 9134
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.4217889762322553,
+      "learning_rate": 1.5906429276636376e-06,
+      "loss": 0.3275,
+      "step": 9135
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.798173938948157,
+      "learning_rate": 1.589674739711814e-06,
+      "loss": 0.9051,
+      "step": 9136
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.621642259837425,
+      "learning_rate": 1.5887067908089472e-06,
+      "loss": 0.3025,
+      "step": 9137
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.7705583656212167,
+      "learning_rate": 1.5877390810228888e-06,
+      "loss": 0.5332,
+      "step": 9138
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1907773558344406,
+      "learning_rate": 1.5867716104214725e-06,
+      "loss": 0.1737,
+      "step": 9139
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.354378021368676,
+      "learning_rate": 1.5858043790725096e-06,
+      "loss": 0.6442,
+      "step": 9140
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.819461979920989,
+      "learning_rate": 1.5848373870438016e-06,
+      "loss": 0.2835,
+      "step": 9141
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.5942531966284172,
+      "learning_rate": 1.58387063440313e-06,
+      "loss": 0.4373,
+      "step": 9142
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.81101872617202,
+      "learning_rate": 1.58290412121826e-06,
+      "loss": 0.8678,
+      "step": 9143
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.625409574628458,
+      "learning_rate": 1.5819378475569396e-06,
+      "loss": 0.6234,
+      "step": 9144
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.659023434089425,
+      "learning_rate": 1.5809718134869024e-06,
+      "loss": 0.7718,
+      "step": 9145
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.435425429651767,
+      "learning_rate": 1.5800060190758592e-06,
+      "loss": 0.9598,
+      "step": 9146
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.1947682057148135,
+      "learning_rate": 1.5790404643915108e-06,
+      "loss": 0.3132,
+      "step": 9147
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.460252436859948,
+      "learning_rate": 1.5780751495015379e-06,
+      "loss": 0.5249,
+      "step": 9148
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.794571630750243,
+      "learning_rate": 1.5771100744736039e-06,
+      "loss": 0.7865,
+      "step": 9149
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.601783932317756,
+      "learning_rate": 1.5761452393753596e-06,
+      "loss": 0.5184,
+      "step": 9150
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.265490982404797,
+      "learning_rate": 1.5751806442744315e-06,
+      "loss": 1.082,
+      "step": 9151
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 6.026421806841435,
+      "learning_rate": 1.5742162892384372e-06,
+      "loss": 1.2085,
+      "step": 9152
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.885104571684317,
+      "learning_rate": 1.57325217433497e-06,
+      "loss": 0.9064,
+      "step": 9153
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.5114201265328964,
+      "learning_rate": 1.5722882996316125e-06,
+      "loss": 0.4936,
+      "step": 9154
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.8762422337914098,
+      "learning_rate": 1.5713246651959275e-06,
+      "loss": 0.735,
+      "step": 9155
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.451621781366753,
+      "learning_rate": 1.570361271095462e-06,
+      "loss": 1.0967,
+      "step": 9156
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.57069746566801,
+      "learning_rate": 1.5693981173977468e-06,
+      "loss": 0.6759,
+      "step": 9157
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.530243213795102,
+      "learning_rate": 1.568435204170292e-06,
+      "loss": 0.8214,
+      "step": 9158
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.486699756452967,
+      "learning_rate": 1.5674725314805955e-06,
+      "loss": 1.3609,
+      "step": 9159
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.165946902347504,
+      "learning_rate": 1.5665100993961358e-06,
+      "loss": 0.9671,
+      "step": 9160
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.834807390806405,
+      "learning_rate": 1.565547907984376e-06,
+      "loss": 1.0393,
+      "step": 9161
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.4886776264223816,
+      "learning_rate": 1.564585957312762e-06,
+      "loss": 1.4576,
+      "step": 9162
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.392651734475036,
+      "learning_rate": 1.5636242474487207e-06,
+      "loss": 0.5364,
+      "step": 9163
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.270991036499043,
+      "learning_rate": 1.5626627784596638e-06,
+      "loss": 1.3793,
+      "step": 9164
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.2656349137900924,
+      "learning_rate": 1.5617015504129867e-06,
+      "loss": 0.3967,
+      "step": 9165
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.42978305101965,
+      "learning_rate": 1.560740563376069e-06,
+      "loss": 0.7694,
+      "step": 9166
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 6.110916215287073,
+      "learning_rate": 1.5597798174162693e-06,
+      "loss": 0.8729,
+      "step": 9167
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.143223280894732,
+      "learning_rate": 1.5588193126009332e-06,
+      "loss": 0.587,
+      "step": 9168
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.703192537288563,
+      "learning_rate": 1.557859048997386e-06,
+      "loss": 1.0459,
+      "step": 9169
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.171765841636195,
+      "learning_rate": 1.5568990266729394e-06,
+      "loss": 0.4245,
+      "step": 9170
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.923832562043842,
+      "learning_rate": 1.5559392456948863e-06,
+      "loss": 1.0203,
+      "step": 9171
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.2433702013284265,
+      "learning_rate": 1.5549797061305039e-06,
+      "loss": 0.7897,
+      "step": 9172
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.151892778553473,
+      "learning_rate": 1.5540204080470512e-06,
+      "loss": 1.1794,
+      "step": 9173
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.868456145903217,
+      "learning_rate": 1.5530613515117721e-06,
+      "loss": 0.4656,
+      "step": 9174
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.6448226324506305,
+      "learning_rate": 1.5521025365918895e-06,
+      "loss": 0.4984,
+      "step": 9175
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.23419139243431,
+      "learning_rate": 1.5511439633546143e-06,
+      "loss": 0.5805,
+      "step": 9176
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.0798081955034866,
+      "learning_rate": 1.5501856318671376e-06,
+      "loss": 0.6739,
+      "step": 9177
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.303341304803665,
+      "learning_rate": 1.5492275421966346e-06,
+      "loss": 0.9649,
+      "step": 9178
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.2423816550283253,
+      "learning_rate": 1.5482696944102643e-06,
+      "loss": 0.7856,
+      "step": 9179
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.566822051382423,
+      "learning_rate": 1.5473120885751652e-06,
+      "loss": 0.7349,
+      "step": 9180
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.187762857595358,
+      "learning_rate": 1.5463547247584621e-06,
+      "loss": 0.6959,
+      "step": 9181
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.9031055664051526,
+      "learning_rate": 1.5453976030272645e-06,
+      "loss": 0.7711,
+      "step": 9182
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.1449051622626314,
+      "learning_rate": 1.5444407234486585e-06,
+      "loss": 0.5387,
+      "step": 9183
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.46496397396142,
+      "learning_rate": 1.5434840860897194e-06,
+      "loss": 1.0059,
+      "step": 9184
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.073251346906123,
+      "learning_rate": 1.5425276910175046e-06,
+      "loss": 0.9449,
+      "step": 9185
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.155958714244286,
+      "learning_rate": 1.5415715382990504e-06,
+      "loss": 0.427,
+      "step": 9186
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.203303357750263,
+      "learning_rate": 1.54061562800138e-06,
+      "loss": 0.3593,
+      "step": 9187
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.2183995353328285,
+      "learning_rate": 1.5396599601914986e-06,
+      "loss": 0.5619,
+      "step": 9188
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.199122158242305,
+      "learning_rate": 1.5387045349363948e-06,
+      "loss": 0.6847,
+      "step": 9189
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.7695706975246277,
+      "learning_rate": 1.537749352303039e-06,
+      "loss": 0.7021,
+      "step": 9190
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.6356070296975225,
+      "learning_rate": 1.5367944123583884e-06,
+      "loss": 0.466,
+      "step": 9191
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.199184907548104,
+      "learning_rate": 1.535839715169375e-06,
+      "loss": 0.7087,
+      "step": 9192
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.531727412732683,
+      "learning_rate": 1.5348852608029218e-06,
+      "loss": 0.841,
+      "step": 9193
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.345802142439073,
+      "learning_rate": 1.5339310493259318e-06,
+      "loss": 0.5765,
+      "step": 9194
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.1223332428130117,
+      "learning_rate": 1.5329770808052908e-06,
+      "loss": 0.603,
+      "step": 9195
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.8725909950301562,
+      "learning_rate": 1.5320233553078694e-06,
+      "loss": 0.2948,
+      "step": 9196
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.6803796890064913,
+      "learning_rate": 1.5310698729005163e-06,
+      "loss": 0.3243,
+      "step": 9197
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.310145402547282,
+      "learning_rate": 1.5301166336500701e-06,
+      "loss": 0.7332,
+      "step": 9198
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.252048882124109,
+      "learning_rate": 1.5291636376233453e-06,
+      "loss": 0.5481,
+      "step": 9199
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.9315753599609495,
+      "learning_rate": 1.5282108848871445e-06,
+      "loss": 0.614,
+      "step": 9200
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.032979560741048,
+      "learning_rate": 1.5272583755082516e-06,
+      "loss": 0.7907,
+      "step": 9201
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2516871167184147,
+      "learning_rate": 1.5263061095534343e-06,
+      "loss": 0.1925,
+      "step": 9202
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.881963340392816,
+      "learning_rate": 1.5253540870894395e-06,
+      "loss": 0.7426,
+      "step": 9203
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.977219673175557,
+      "learning_rate": 1.5244023081830018e-06,
+      "loss": 0.7551,
+      "step": 9204
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.988781431535191,
+      "learning_rate": 1.5234507729008363e-06,
+      "loss": 0.4727,
+      "step": 9205
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.142425806367947,
+      "learning_rate": 1.5224994813096417e-06,
+      "loss": 0.6021,
+      "step": 9206
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.4882982513011305,
+      "learning_rate": 1.5215484334760988e-06,
+      "loss": 0.9421,
+      "step": 9207
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.227574799732071,
+      "learning_rate": 1.5205976294668745e-06,
+      "loss": 0.8865,
+      "step": 9208
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.294329145390786,
+      "learning_rate": 1.5196470693486127e-06,
+      "loss": 0.3568,
+      "step": 9209
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.3143291348855994,
+      "learning_rate": 1.5186967531879443e-06,
+      "loss": 0.3798,
+      "step": 9210
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 6.1684971311613435,
+      "learning_rate": 1.517746681051483e-06,
+      "loss": 1.1759,
+      "step": 9211
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.807693670268817,
+      "learning_rate": 1.5167968530058263e-06,
+      "loss": 1.0803,
+      "step": 9212
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.06193247946862,
+      "learning_rate": 1.5158472691175491e-06,
+      "loss": 0.9869,
+      "step": 9213
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.603334917657383,
+      "learning_rate": 1.5148979294532157e-06,
+      "loss": 0.8715,
+      "step": 9214
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.506355503448254,
+      "learning_rate": 1.5139488340793718e-06,
+      "loss": 0.8412,
+      "step": 9215
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.3900375758386714,
+      "learning_rate": 1.512999983062542e-06,
+      "loss": 0.4941,
+      "step": 9216
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.615903306345567,
+      "learning_rate": 1.5120513764692373e-06,
+      "loss": 0.6802,
+      "step": 9217
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.350467604248198,
+      "learning_rate": 1.5111030143659516e-06,
+      "loss": 0.5037,
+      "step": 9218
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.471453585618678,
+      "learning_rate": 1.5101548968191626e-06,
+      "loss": 0.6427,
+      "step": 9219
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.454923873453871,
+      "learning_rate": 1.509207023895326e-06,
+      "loss": 1.0145,
+      "step": 9220
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.56444356398917,
+      "learning_rate": 1.5082593956608848e-06,
+      "loss": 0.3001,
+      "step": 9221
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.721004611228858,
+      "learning_rate": 1.5073120121822642e-06,
+      "loss": 0.9185,
+      "step": 9222
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0366899485017664,
+      "learning_rate": 1.5063648735258713e-06,
+      "loss": 0.3418,
+      "step": 9223
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.1813444894396907,
+      "learning_rate": 1.5054179797580959e-06,
+      "loss": 0.3646,
+      "step": 9224
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.625454279907865,
+      "learning_rate": 1.5044713309453135e-06,
+      "loss": 0.9152,
+      "step": 9225
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.708317774378133,
+      "learning_rate": 1.5035249271538766e-06,
+      "loss": 0.5606,
+      "step": 9226
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.811622971525766,
+      "learning_rate": 1.5025787684501259e-06,
+      "loss": 1.402,
+      "step": 9227
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.1435508201419733,
+      "learning_rate": 1.5016328549003822e-06,
+      "loss": 0.5919,
+      "step": 9228
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.422794846371174,
+      "learning_rate": 1.5006871865709527e-06,
+      "loss": 0.9311,
+      "step": 9229
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.5595881510978926,
+      "learning_rate": 1.4997417635281204e-06,
+      "loss": 0.5799,
+      "step": 9230
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.0844228003472995,
+      "learning_rate": 1.4987965858381587e-06,
+      "loss": 0.4529,
+      "step": 9231
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.2559439791737455,
+      "learning_rate": 1.4978516535673176e-06,
+      "loss": 0.4913,
+      "step": 9232
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.7266898885025044,
+      "learning_rate": 1.4969069667818342e-06,
+      "loss": 1.016,
+      "step": 9233
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6652470891476823,
+      "learning_rate": 1.495962525547927e-06,
+      "loss": 0.3189,
+      "step": 9234
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.8723780905297405,
+      "learning_rate": 1.4950183299317972e-06,
+      "loss": 0.5337,
+      "step": 9235
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.299827963645423,
+      "learning_rate": 1.4940743799996282e-06,
+      "loss": 0.8833,
+      "step": 9236
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4783504472332694,
+      "learning_rate": 1.4931306758175896e-06,
+      "loss": 0.1693,
+      "step": 9237
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9569216401301468,
+      "learning_rate": 1.4921872174518264e-06,
+      "loss": 0.3565,
+      "step": 9238
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.9978458167392135,
+      "learning_rate": 1.491244004968474e-06,
+      "loss": 0.9464,
+      "step": 9239
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.694128103080308,
+      "learning_rate": 1.4903010384336465e-06,
+      "loss": 0.4128,
+      "step": 9240
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.4706648723140625,
+      "learning_rate": 1.4893583179134414e-06,
+      "loss": 0.568,
+      "step": 9241
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.648688088377756,
+      "learning_rate": 1.488415843473942e-06,
+      "loss": 1.097,
+      "step": 9242
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.266772854792492,
+      "learning_rate": 1.4874736151812075e-06,
+      "loss": 1.2822,
+      "step": 9243
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.411746530543385,
+      "learning_rate": 1.4865316331012862e-06,
+      "loss": 0.8683,
+      "step": 9244
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.772661039761835,
+      "learning_rate": 1.4855898973002087e-06,
+      "loss": 0.5959,
+      "step": 9245
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.814719328692366,
+      "learning_rate": 1.484648407843982e-06,
+      "loss": 0.5103,
+      "step": 9246
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.507064544108054,
+      "learning_rate": 1.483707164798604e-06,
+      "loss": 0.6051,
+      "step": 9247
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.202915654401037,
+      "learning_rate": 1.4827661682300521e-06,
+      "loss": 1.0418,
+      "step": 9248
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.469431937363606,
+      "learning_rate": 1.4818254182042834e-06,
+      "loss": 0.6728,
+      "step": 9249
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.4053054748930736,
+      "learning_rate": 1.4808849147872417e-06,
+      "loss": 0.4348,
+      "step": 9250
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.7957141907992247,
+      "learning_rate": 1.4799446580448517e-06,
+      "loss": 0.597,
+      "step": 9251
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.380756083812477,
+      "learning_rate": 1.4790046480430226e-06,
+      "loss": 0.7895,
+      "step": 9252
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.542878811181817,
+      "learning_rate": 1.4780648848476436e-06,
+      "loss": 0.7149,
+      "step": 9253
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.6472417985590964,
+      "learning_rate": 1.4771253685245907e-06,
+      "loss": 0.3477,
+      "step": 9254
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.8230679422187395,
+      "learning_rate": 1.476186099139716e-06,
+      "loss": 0.4883,
+      "step": 9255
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.998509200929273,
+      "learning_rate": 1.47524707675886e-06,
+      "loss": 0.4907,
+      "step": 9256
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.727298671205657,
+      "learning_rate": 1.4743083014478443e-06,
+      "loss": 1.0115,
+      "step": 9257
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.258019229535348,
+      "learning_rate": 1.4733697732724728e-06,
+      "loss": 0.8451,
+      "step": 9258
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.0894962139530087,
+      "learning_rate": 1.472431492298534e-06,
+      "loss": 0.7153,
+      "step": 9259
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.140265764565909,
+      "learning_rate": 1.4714934585917933e-06,
+      "loss": 0.5697,
+      "step": 9260
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.008713959072589,
+      "learning_rate": 1.4705556722180075e-06,
+      "loss": 0.6259,
+      "step": 9261
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.148159226128686,
+      "learning_rate": 1.4696181332429065e-06,
+      "loss": 1.1498,
+      "step": 9262
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.23851846708833,
+      "learning_rate": 1.4686808417322107e-06,
+      "loss": 0.845,
+      "step": 9263
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.2197524105620805,
+      "learning_rate": 1.4677437977516197e-06,
+      "loss": 0.5212,
+      "step": 9264
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.469659448279783,
+      "learning_rate": 1.4668070013668173e-06,
+      "loss": 0.9465,
+      "step": 9265
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.8821891357359815,
+      "learning_rate": 1.465870452643466e-06,
+      "loss": 0.8122,
+      "step": 9266
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.459784063330109,
+      "learning_rate": 1.464934151647215e-06,
+      "loss": 0.9195,
+      "step": 9267
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.3430373100392607,
+      "learning_rate": 1.4639980984436957e-06,
+      "loss": 0.5985,
+      "step": 9268
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.594680733843876,
+      "learning_rate": 1.463062293098521e-06,
+      "loss": 0.5093,
+      "step": 9269
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.152942092158016,
+      "learning_rate": 1.4621267356772867e-06,
+      "loss": 0.7401,
+      "step": 9270
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.643754547322705,
+      "learning_rate": 1.461191426245573e-06,
+      "loss": 0.7576,
+      "step": 9271
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4539657131529113,
+      "learning_rate": 1.4602563648689378e-06,
+      "loss": 0.1889,
+      "step": 9272
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 6.884785121400726,
+      "learning_rate": 1.459321551612926e-06,
+      "loss": 1.0614,
+      "step": 9273
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.28683900158588,
+      "learning_rate": 1.458386986543065e-06,
+      "loss": 0.3655,
+      "step": 9274
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.380924904423539,
+      "learning_rate": 1.4574526697248643e-06,
+      "loss": 0.8172,
+      "step": 9275
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.336139561522197,
+      "learning_rate": 1.4565186012238126e-06,
+      "loss": 0.4901,
+      "step": 9276
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.045062288953997,
+      "learning_rate": 1.4555847811053875e-06,
+      "loss": 0.9607,
+      "step": 9277
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.236958797578773,
+      "learning_rate": 1.4546512094350424e-06,
+      "loss": 0.7349,
+      "step": 9278
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.9752286428550136,
+      "learning_rate": 1.4537178862782175e-06,
+      "loss": 0.7118,
+      "step": 9279
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.75898198448468,
+      "learning_rate": 1.4527848117003357e-06,
+      "loss": 1.1713,
+      "step": 9280
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.520741778373616,
+      "learning_rate": 1.4518519857668012e-06,
+      "loss": 0.8197,
+      "step": 9281
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.377246753506351,
+      "learning_rate": 1.4509194085430024e-06,
+      "loss": 0.7396,
+      "step": 9282
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.599893595646473,
+      "learning_rate": 1.4499870800943055e-06,
+      "loss": 0.4715,
+      "step": 9283
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.1112529821371355,
+      "learning_rate": 1.4490550004860655e-06,
+      "loss": 1.0465,
+      "step": 9284
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.127834502096852,
+      "learning_rate": 1.4481231697836152e-06,
+      "loss": 0.6399,
+      "step": 9285
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5725140266229545,
+      "learning_rate": 1.447191588052273e-06,
+      "loss": 0.1532,
+      "step": 9286
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.859033873850176,
+      "learning_rate": 1.446260255357339e-06,
+      "loss": 1.1814,
+      "step": 9287
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.951917050503969,
+      "learning_rate": 1.4453291717640966e-06,
+      "loss": 0.2644,
+      "step": 9288
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.300478741110975,
+      "learning_rate": 1.4443983373378078e-06,
+      "loss": 0.4078,
+      "step": 9289
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.5342038298764753,
+      "learning_rate": 1.4434677521437213e-06,
+      "loss": 0.4432,
+      "step": 9290
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.663422152939918,
+      "learning_rate": 1.442537416247069e-06,
+      "loss": 0.9114,
+      "step": 9291
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.431581004164061,
+      "learning_rate": 1.44160732971306e-06,
+      "loss": 0.7277,
+      "step": 9292
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.913245516633208,
+      "learning_rate": 1.4406774926068912e-06,
+      "loss": 1.0555,
+      "step": 9293
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.899105723703307,
+      "learning_rate": 1.4397479049937413e-06,
+      "loss": 0.6237,
+      "step": 9294
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.7982967597812856,
+      "learning_rate": 1.4388185669387678e-06,
+      "loss": 0.79,
+      "step": 9295
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.478167377412812,
+      "learning_rate": 1.437889478507114e-06,
+      "loss": 0.8015,
+      "step": 9296
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.683330777135807,
+      "learning_rate": 1.4369606397639058e-06,
+      "loss": 0.8996,
+      "step": 9297
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0947825710626913,
+      "learning_rate": 1.4360320507742503e-06,
+      "loss": 0.3233,
+      "step": 9298
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.9859594539972947,
+      "learning_rate": 1.4351037116032391e-06,
+      "loss": 0.8938,
+      "step": 9299
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.531021143798802,
+      "learning_rate": 1.4341756223159414e-06,
+      "loss": 0.2781,
+      "step": 9300
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.355153049487024,
+      "learning_rate": 1.4332477829774144e-06,
+      "loss": 0.3262,
+      "step": 9301
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.000948490375274,
+      "learning_rate": 1.432320193652695e-06,
+      "loss": 0.7266,
+      "step": 9302
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.299111288992087,
+      "learning_rate": 1.4313928544068033e-06,
+      "loss": 0.7665,
+      "step": 9303
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.5906372535420803,
+      "learning_rate": 1.430465765304742e-06,
+      "loss": 0.7717,
+      "step": 9304
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.570204564671903,
+      "learning_rate": 1.429538926411498e-06,
+      "loss": 0.8141,
+      "step": 9305
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.044477247782365,
+      "learning_rate": 1.4286123377920342e-06,
+      "loss": 1.0105,
+      "step": 9306
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.814453038391482,
+      "learning_rate": 1.4276859995113047e-06,
+      "loss": 1.3584,
+      "step": 9307
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.5677901090269843,
+      "learning_rate": 1.4267599116342384e-06,
+      "loss": 0.7221,
+      "step": 9308
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.9836186775557545,
+      "learning_rate": 1.4258340742257516e-06,
+      "loss": 0.5854,
+      "step": 9309
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.789849731644247,
+      "learning_rate": 1.4249084873507412e-06,
+      "loss": 1.1888,
+      "step": 9310
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.761888111510434,
+      "learning_rate": 1.423983151074088e-06,
+      "loss": 0.5422,
+      "step": 9311
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.060486241860016,
+      "learning_rate": 1.4230580654606523e-06,
+      "loss": 1.1927,
+      "step": 9312
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.9496662520122205,
+      "learning_rate": 1.422133230575279e-06,
+      "loss": 1.3018,
+      "step": 9313
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.520326361871921,
+      "learning_rate": 1.4212086464827957e-06,
+      "loss": 0.9233,
+      "step": 9314
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 6.280643679190604,
+      "learning_rate": 1.420284313248011e-06,
+      "loss": 1.4922,
+      "step": 9315
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.251869664324042,
+      "learning_rate": 1.419360230935717e-06,
+      "loss": 0.6898,
+      "step": 9316
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.65519446900931,
+      "learning_rate": 1.4184363996106888e-06,
+      "loss": 1.1632,
+      "step": 9317
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.138380698480845,
+      "learning_rate": 1.417512819337681e-06,
+      "loss": 0.9813,
+      "step": 9318
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.404739869469795,
+      "learning_rate": 1.4165894901814337e-06,
+      "loss": 0.8594,
+      "step": 9319
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.8786235118438714,
+      "learning_rate": 1.4156664122066678e-06,
+      "loss": 0.4921,
+      "step": 9320
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.155279089720612,
+      "learning_rate": 1.414743585478089e-06,
+      "loss": 1.0607,
+      "step": 9321
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.7988268307826685,
+      "learning_rate": 1.41382101006038e-06,
+      "loss": 0.3708,
+      "step": 9322
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0743694371725223,
+      "learning_rate": 1.412898686018211e-06,
+      "loss": 0.1034,
+      "step": 9323
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.3399958967063275,
+      "learning_rate": 1.411976613416235e-06,
+      "loss": 0.3909,
+      "step": 9324
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.74287426460022,
+      "learning_rate": 1.4110547923190816e-06,
+      "loss": 0.962,
+      "step": 9325
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.6131435057312045,
+      "learning_rate": 1.4101332227913677e-06,
+      "loss": 0.7313,
+      "step": 9326
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.610089113972206,
+      "learning_rate": 1.409211904897692e-06,
+      "loss": 0.8497,
+      "step": 9327
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.298292588477315,
+      "learning_rate": 1.4082908387026362e-06,
+      "loss": 0.9706,
+      "step": 9328
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.271103752048412,
+      "learning_rate": 1.40737002427076e-06,
+      "loss": 1.0153,
+      "step": 9329
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.1823134658759185,
+      "learning_rate": 1.4064494616666096e-06,
+      "loss": 0.4829,
+      "step": 9330
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.532047882558935,
+      "learning_rate": 1.405529150954713e-06,
+      "loss": 0.2629,
+      "step": 9331
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.042392721562254,
+      "learning_rate": 1.4046090921995798e-06,
+      "loss": 0.6231,
+      "step": 9332
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.070483187386631,
+      "learning_rate": 1.4036892854657019e-06,
+      "loss": 0.7259,
+      "step": 9333
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.1113878681312666,
+      "learning_rate": 1.4027697308175554e-06,
+      "loss": 0.4745,
+      "step": 9334
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.709747769472877,
+      "learning_rate": 1.4018504283195938e-06,
+      "loss": 1.1229,
+      "step": 9335
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9946288181004355,
+      "learning_rate": 1.4009313780362582e-06,
+      "loss": 0.3971,
+      "step": 9336
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.925474563921446,
+      "learning_rate": 1.4000125800319702e-06,
+      "loss": 0.9125,
+      "step": 9337
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.7557709471588145,
+      "learning_rate": 1.399094034371134e-06,
+      "loss": 0.5966,
+      "step": 9338
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.1324359061760685,
+      "learning_rate": 1.398175741118134e-06,
+      "loss": 0.8441,
+      "step": 9339
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.233867463157931,
+      "learning_rate": 1.3972577003373406e-06,
+      "loss": 0.6437,
+      "step": 9340
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.818645870947484,
+      "learning_rate": 1.3963399120931014e-06,
+      "loss": 0.5644,
+      "step": 9341
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.272791124987817,
+      "learning_rate": 1.395422376449751e-06,
+      "loss": 0.9272,
+      "step": 9342
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.315108114203056,
+      "learning_rate": 1.3945050934716054e-06,
+      "loss": 0.9911,
+      "step": 9343
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.992811384045854,
+      "learning_rate": 1.3935880632229614e-06,
+      "loss": 1.4808,
+      "step": 9344
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.115425303865779,
+      "learning_rate": 1.3926712857681002e-06,
+      "loss": 1.2953,
+      "step": 9345
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7229788244527938,
+      "learning_rate": 1.3917547611712818e-06,
+      "loss": 0.1135,
+      "step": 9346
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.296847285183059,
+      "learning_rate": 1.3908384894967514e-06,
+      "loss": 0.4233,
+      "step": 9347
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.8014559482617996,
+      "learning_rate": 1.3899224708087356e-06,
+      "loss": 0.7937,
+      "step": 9348
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.5560120219233435,
+      "learning_rate": 1.3890067051714435e-06,
+      "loss": 0.5396,
+      "step": 9349
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.245313196667449,
+      "learning_rate": 1.3880911926490658e-06,
+      "loss": 0.4887,
+      "step": 9350
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.983509551692768,
+      "learning_rate": 1.3871759333057783e-06,
+      "loss": 0.7939,
+      "step": 9351
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.5050686399806574,
+      "learning_rate": 1.3862609272057337e-06,
+      "loss": 0.7862,
+      "step": 9352
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.121032184419712,
+      "learning_rate": 1.3853461744130703e-06,
+      "loss": 0.4938,
+      "step": 9353
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.9763811056618734,
+      "learning_rate": 1.3844316749919113e-06,
+      "loss": 1.0544,
+      "step": 9354
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.587943342699081,
+      "learning_rate": 1.3835174290063553e-06,
+      "loss": 1.1476,
+      "step": 9355
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.1870902498506797,
+      "learning_rate": 1.3826034365204876e-06,
+      "loss": 0.6916,
+      "step": 9356
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.174244088434585,
+      "learning_rate": 1.3816896975983784e-06,
+      "loss": 0.7123,
+      "step": 9357
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.739528189373238,
+      "learning_rate": 1.380776212304073e-06,
+      "loss": 0.7124,
+      "step": 9358
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.591639188404987,
+      "learning_rate": 1.379862980701604e-06,
+      "loss": 0.5921,
+      "step": 9359
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.73035731361834,
+      "learning_rate": 1.378950002854985e-06,
+      "loss": 0.6594,
+      "step": 9360
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.607527610861995,
+      "learning_rate": 1.378037278828212e-06,
+      "loss": 0.8706,
+      "step": 9361
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.870706821427959,
+      "learning_rate": 1.3771248086852646e-06,
+      "loss": 1.1595,
+      "step": 9362
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.951566000798771,
+      "learning_rate": 1.3762125924900998e-06,
+      "loss": 1.2864,
+      "step": 9363
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0097441103001916,
+      "learning_rate": 1.3753006303066612e-06,
+      "loss": 0.1232,
+      "step": 9364
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.441261479938119,
+      "learning_rate": 1.3743889221988744e-06,
+      "loss": 0.895,
+      "step": 9365
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.9012789096869036,
+      "learning_rate": 1.3734774682306446e-06,
+      "loss": 0.515,
+      "step": 9366
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.148929030640898,
+      "learning_rate": 1.372566268465862e-06,
+      "loss": 0.7201,
+      "step": 9367
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.385343386382513,
+      "learning_rate": 1.3716553229683989e-06,
+      "loss": 0.726,
+      "step": 9368
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.111237908195454,
+      "learning_rate": 1.3707446318021051e-06,
+      "loss": 0.8248,
+      "step": 9369
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.449998081835322,
+      "learning_rate": 1.3698341950308198e-06,
+      "loss": 0.7886,
+      "step": 9370
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.300561376325745,
+      "learning_rate": 1.3689240127183572e-06,
+      "loss": 0.8067,
+      "step": 9371
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.250771097655397,
+      "learning_rate": 1.3680140849285196e-06,
+      "loss": 0.779,
+      "step": 9372
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5043779709019471,
+      "learning_rate": 1.3671044117250875e-06,
+      "loss": 0.2278,
+      "step": 9373
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.6883155849475115,
+      "learning_rate": 1.366194993171827e-06,
+      "loss": 0.4736,
+      "step": 9374
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.793560869027931,
+      "learning_rate": 1.3652858293324823e-06,
+      "loss": 0.9813,
+      "step": 9375
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.328732050350097,
+      "learning_rate": 1.3643769202707824e-06,
+      "loss": 0.2956,
+      "step": 9376
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.89171571364176,
+      "learning_rate": 1.3634682660504379e-06,
+      "loss": 1.0947,
+      "step": 9377
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.4013676591947335,
+      "learning_rate": 1.362559866735142e-06,
+      "loss": 0.7116,
+      "step": 9378
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 6.549481225715749,
+      "learning_rate": 1.3616517223885707e-06,
+      "loss": 0.8947,
+      "step": 9379
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.999966242113525,
+      "learning_rate": 1.3607438330743778e-06,
+      "loss": 0.8584,
+      "step": 9380
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.9561831618363765,
+      "learning_rate": 1.3598361988562037e-06,
+      "loss": 0.8152,
+      "step": 9381
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.874434061208297,
+      "learning_rate": 1.3589288197976707e-06,
+      "loss": 0.6809,
+      "step": 9382
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.074947674701273,
+      "learning_rate": 1.358021695962381e-06,
+      "loss": 0.7936,
+      "step": 9383
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.760480366466265,
+      "learning_rate": 1.3571148274139223e-06,
+      "loss": 0.7204,
+      "step": 9384
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.5202573924317875,
+      "learning_rate": 1.356208214215859e-06,
+      "loss": 0.8679,
+      "step": 9385
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.7399295576934057,
+      "learning_rate": 1.3553018564317432e-06,
+      "loss": 0.6583,
+      "step": 9386
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.5029310094502546,
+      "learning_rate": 1.354395754125104e-06,
+      "loss": 0.4584,
+      "step": 9387
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.787278806573176,
+      "learning_rate": 1.3534899073594566e-06,
+      "loss": 0.9349,
+      "step": 9388
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.8050894166241926,
+      "learning_rate": 1.352584316198297e-06,
+      "loss": 0.7285,
+      "step": 9389
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.330536045395092,
+      "learning_rate": 1.351678980705104e-06,
+      "loss": 1.2079,
+      "step": 9390
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.253642168070775,
+      "learning_rate": 1.3507739009433374e-06,
+      "loss": 0.6539,
+      "step": 9391
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 6.463133979464778,
+      "learning_rate": 1.3498690769764378e-06,
+      "loss": 1.4632,
+      "step": 9392
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.254093796324293,
+      "learning_rate": 1.3489645088678305e-06,
+      "loss": 1.1103,
+      "step": 9393
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.596333662761472,
+      "learning_rate": 1.348060196680922e-06,
+      "loss": 0.5275,
+      "step": 9394
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.251869759713248,
+      "learning_rate": 1.3471561404791e-06,
+      "loss": 0.6351,
+      "step": 9395
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.355014066464097,
+      "learning_rate": 1.3462523403257355e-06,
+      "loss": 0.3062,
+      "step": 9396
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.033922825683495,
+      "learning_rate": 1.3453487962841821e-06,
+      "loss": 0.5902,
+      "step": 9397
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.126571279478452,
+      "learning_rate": 1.3444455084177716e-06,
+      "loss": 0.6138,
+      "step": 9398
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7703313801155975,
+      "learning_rate": 1.343542476789822e-06,
+      "loss": 0.3165,
+      "step": 9399
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.172617171443144,
+      "learning_rate": 1.3426397014636334e-06,
+      "loss": 0.4856,
+      "step": 9400
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.597938554244449,
+      "learning_rate": 1.3417371825024832e-06,
+      "loss": 0.9556,
+      "step": 9401
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.4172447302014675,
+      "learning_rate": 1.3408349199696374e-06,
+      "loss": 0.9496,
+      "step": 9402
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.725445692558316,
+      "learning_rate": 1.3399329139283375e-06,
+      "loss": 1.241,
+      "step": 9403
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.7472402363942745,
+      "learning_rate": 1.3390311644418113e-06,
+      "loss": 1.1577,
+      "step": 9404
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.658514626642248,
+      "learning_rate": 1.3381296715732678e-06,
+      "loss": 0.215,
+      "step": 9405
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.9845902825761854,
+      "learning_rate": 1.3372284353858983e-06,
+      "loss": 0.5422,
+      "step": 9406
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.8423579845115987,
+      "learning_rate": 1.3363274559428747e-06,
+      "loss": 0.7291,
+      "step": 9407
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.410067348783659,
+      "learning_rate": 1.335426733307354e-06,
+      "loss": 0.7367,
+      "step": 9408
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.537893849743915,
+      "learning_rate": 1.3345262675424691e-06,
+      "loss": 0.8426,
+      "step": 9409
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.876409687775638,
+      "learning_rate": 1.333626058711341e-06,
+      "loss": 0.679,
+      "step": 9410
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.163415705623005,
+      "learning_rate": 1.3327261068770698e-06,
+      "loss": 1.0723,
+      "step": 9411
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.045484247563682,
+      "learning_rate": 1.331826412102738e-06,
+      "loss": 0.8043,
+      "step": 9412
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.400894042145848,
+      "learning_rate": 1.3309269744514114e-06,
+      "loss": 0.8605,
+      "step": 9413
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.41461886481942,
+      "learning_rate": 1.3300277939861372e-06,
+      "loss": 0.4978,
+      "step": 9414
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.885506898324602,
+      "learning_rate": 1.3291288707699417e-06,
+      "loss": 0.615,
+      "step": 9415
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.452241834588947,
+      "learning_rate": 1.328230204865838e-06,
+      "loss": 1.2617,
+      "step": 9416
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.098897694714867,
+      "learning_rate": 1.327331796336816e-06,
+      "loss": 1.0752,
+      "step": 9417
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.66589764611867,
+      "learning_rate": 1.3264336452458514e-06,
+      "loss": 0.9541,
+      "step": 9418
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.275492019155368,
+      "learning_rate": 1.3255357516559025e-06,
+      "loss": 0.4241,
+      "step": 9419
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.9066721077298308,
+      "learning_rate": 1.3246381156299048e-06,
+      "loss": 0.886,
+      "step": 9420
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.1401612817685454,
+      "learning_rate": 1.3237407372307792e-06,
+      "loss": 0.3662,
+      "step": 9421
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.326368524937841,
+      "learning_rate": 1.3228436165214298e-06,
+      "loss": 0.7366,
+      "step": 9422
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.3468037609646273,
+      "learning_rate": 1.321946753564739e-06,
+      "loss": 0.489,
+      "step": 9423
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.2502527737540756,
+      "learning_rate": 1.3210501484235744e-06,
+      "loss": 0.5933,
+      "step": 9424
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.2470551800294674,
+      "learning_rate": 1.3201538011607845e-06,
+      "loss": 0.3139,
+      "step": 9425
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.489138232154664,
+      "learning_rate": 1.3192577118391975e-06,
+      "loss": 0.7102,
+      "step": 9426
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8914064841933165,
+      "learning_rate": 1.318361880521626e-06,
+      "loss": 0.4059,
+      "step": 9427
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.063844683388602,
+      "learning_rate": 1.3174663072708637e-06,
+      "loss": 0.6773,
+      "step": 9428
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.50563997940386,
+      "learning_rate": 1.3165709921496873e-06,
+      "loss": 0.714,
+      "step": 9429
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.1772613011563724,
+      "learning_rate": 1.3156759352208554e-06,
+      "loss": 0.5923,
+      "step": 9430
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.2247310279851917,
+      "learning_rate": 1.3147811365471048e-06,
+      "loss": 0.3034,
+      "step": 9431
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.25694837834173,
+      "learning_rate": 1.3138865961911585e-06,
+      "loss": 0.8065,
+      "step": 9432
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.787558627022996,
+      "learning_rate": 1.312992314215721e-06,
+      "loss": 0.9318,
+      "step": 9433
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.661175799036978,
+      "learning_rate": 1.3120982906834745e-06,
+      "loss": 0.7151,
+      "step": 9434
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.371334083380921,
+      "learning_rate": 1.3112045256570888e-06,
+      "loss": 1.0989,
+      "step": 9435
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.424428779312583,
+      "learning_rate": 1.3103110191992118e-06,
+      "loss": 1.0371,
+      "step": 9436
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.7596546812463707,
+      "learning_rate": 1.3094177713724765e-06,
+      "loss": 0.7684,
+      "step": 9437
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.099396301324731,
+      "learning_rate": 1.308524782239492e-06,
+      "loss": 0.8312,
+      "step": 9438
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.6834372052861317,
+      "learning_rate": 1.3076320518628554e-06,
+      "loss": 0.5639,
+      "step": 9439
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2043445807513642,
+      "learning_rate": 1.3067395803051425e-06,
+      "loss": 0.2032,
+      "step": 9440
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.290269026486594,
+      "learning_rate": 1.3058473676289118e-06,
+      "loss": 0.7477,
+      "step": 9441
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.87168361128787,
+      "learning_rate": 1.3049554138967052e-06,
+      "loss": 1.0051,
+      "step": 9442
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.492575685994819,
+      "learning_rate": 1.304063719171042e-06,
+      "loss": 0.1826,
+      "step": 9443
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.6077122922831446,
+      "learning_rate": 1.3031722835144266e-06,
+      "loss": 0.4407,
+      "step": 9444
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.58056453786561,
+      "learning_rate": 1.302281106989346e-06,
+      "loss": 0.5929,
+      "step": 9445
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.254577200822444,
+      "learning_rate": 1.3013901896582677e-06,
+      "loss": 0.8297,
+      "step": 9446
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.348001334484949,
+      "learning_rate": 1.3004995315836417e-06,
+      "loss": 1.0947,
+      "step": 9447
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1117699308226239,
+      "learning_rate": 1.2996091328278965e-06,
+      "loss": 0.1159,
+      "step": 9448
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.1536886378875844,
+      "learning_rate": 1.2987189934534488e-06,
+      "loss": 0.4287,
+      "step": 9449
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.998286991655185,
+      "learning_rate": 1.29782911352269e-06,
+      "loss": 0.7968,
+      "step": 9450
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.112248998883585,
+      "learning_rate": 1.2969394930979984e-06,
+      "loss": 1.1064,
+      "step": 9451
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.889365596754016,
+      "learning_rate": 1.2960501322417323e-06,
+      "loss": 0.4504,
+      "step": 9452
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.3994968910567405,
+      "learning_rate": 1.2951610310162326e-06,
+      "loss": 0.8025,
+      "step": 9453
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.7173719745240255,
+      "learning_rate": 1.2942721894838227e-06,
+      "loss": 0.6403,
+      "step": 9454
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.8722081746417203,
+      "learning_rate": 1.2933836077068036e-06,
+      "loss": 0.645,
+      "step": 9455
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.660927074059716,
+      "learning_rate": 1.2924952857474622e-06,
+      "loss": 0.9383,
+      "step": 9456
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.21218245222277,
+      "learning_rate": 1.291607223668066e-06,
+      "loss": 0.2733,
+      "step": 9457
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.122593583023933,
+      "learning_rate": 1.2907194215308644e-06,
+      "loss": 0.6487,
+      "step": 9458
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.298242692595961,
+      "learning_rate": 1.2898318793980903e-06,
+      "loss": 0.9055,
+      "step": 9459
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.478740435234801,
+      "learning_rate": 1.288944597331953e-06,
+      "loss": 1.3584,
+      "step": 9460
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.834110280911659,
+      "learning_rate": 1.288057575394649e-06,
+      "loss": 0.8821,
+      "step": 9461
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.4921116431704475,
+      "learning_rate": 1.2871708136483546e-06,
+      "loss": 0.8571,
+      "step": 9462
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.40767678423515,
+      "learning_rate": 1.2862843121552293e-06,
+      "loss": 0.6731,
+      "step": 9463
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.4209369222793753,
+      "learning_rate": 1.28539807097741e-06,
+      "loss": 0.6032,
+      "step": 9464
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.98294933471214,
+      "learning_rate": 1.2845120901770214e-06,
+      "loss": 0.5268,
+      "step": 9465
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.8442712477361876,
+      "learning_rate": 1.2836263698161638e-06,
+      "loss": 0.527,
+      "step": 9466
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.020183840794972,
+      "learning_rate": 1.2827409099569237e-06,
+      "loss": 0.4005,
+      "step": 9467
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.753830099766738,
+      "learning_rate": 1.2818557106613689e-06,
+      "loss": 0.9774,
+      "step": 9468
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.753391100022944,
+      "learning_rate": 1.280970771991547e-06,
+      "loss": 0.6152,
+      "step": 9469
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.5802211426804424,
+      "learning_rate": 1.2800860940094889e-06,
+      "loss": 0.8815,
+      "step": 9470
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.932757418212026,
+      "learning_rate": 1.2792016767772081e-06,
+      "loss": 0.6008,
+      "step": 9471
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.060272391727669,
+      "learning_rate": 1.278317520356695e-06,
+      "loss": 0.7747,
+      "step": 9472
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.391174501253654,
+      "learning_rate": 1.2774336248099272e-06,
+      "loss": 0.5028,
+      "step": 9473
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.10834258261621,
+      "learning_rate": 1.2765499901988616e-06,
+      "loss": 0.4337,
+      "step": 9474
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.6488900764648484,
+      "learning_rate": 1.275666616585437e-06,
+      "loss": 0.8388,
+      "step": 9475
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.0377874146013193,
+      "learning_rate": 1.2747835040315741e-06,
+      "loss": 0.7262,
+      "step": 9476
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.020586427498907,
+      "learning_rate": 1.2739006525991772e-06,
+      "loss": 0.4525,
+      "step": 9477
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.296675081507472,
+      "learning_rate": 1.273018062350127e-06,
+      "loss": 1.4636,
+      "step": 9478
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.586643243293631,
+      "learning_rate": 1.2721357333462918e-06,
+      "loss": 0.5183,
+      "step": 9479
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.8351867377323905,
+      "learning_rate": 1.2712536656495167e-06,
+      "loss": 0.5367,
+      "step": 9480
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.886196342458603,
+      "learning_rate": 1.2703718593216324e-06,
+      "loss": 0.7353,
+      "step": 9481
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.554634043073279,
+      "learning_rate": 1.2694903144244509e-06,
+      "loss": 0.703,
+      "step": 9482
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.4791022781511325,
+      "learning_rate": 1.2686090310197613e-06,
+      "loss": 0.7757,
+      "step": 9483
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.7325192961946816,
+      "learning_rate": 1.2677280091693395e-06,
+      "loss": 0.7382,
+      "step": 9484
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.521604735262762,
+      "learning_rate": 1.2668472489349416e-06,
+      "loss": 1.0818,
+      "step": 9485
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.000064290386803,
+      "learning_rate": 1.2659667503783047e-06,
+      "loss": 0.2897,
+      "step": 9486
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.1188878553316743,
+      "learning_rate": 1.265086513561148e-06,
+      "loss": 0.4462,
+      "step": 9487
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 6.865122818758232,
+      "learning_rate": 1.2642065385451736e-06,
+      "loss": 1.4022,
+      "step": 9488
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 6.057940881244001,
+      "learning_rate": 1.263326825392061e-06,
+      "loss": 1.1208,
+      "step": 9489
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.035237632457478,
+      "learning_rate": 1.2624473741634764e-06,
+      "loss": 0.8167,
+      "step": 9490
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 6.389708678626936,
+      "learning_rate": 1.2615681849210648e-06,
+      "loss": 1.2094,
+      "step": 9491
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.9356074826170615,
+      "learning_rate": 1.260689257726454e-06,
+      "loss": 0.767,
+      "step": 9492
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.229117168492173,
+      "learning_rate": 1.2598105926412536e-06,
+      "loss": 0.5801,
+      "step": 9493
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.9430776553172686,
+      "learning_rate": 1.2589321897270523e-06,
+      "loss": 0.5619,
+      "step": 9494
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.8731757871049557,
+      "learning_rate": 1.2580540490454246e-06,
+      "loss": 0.689,
+      "step": 9495
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.9330079288498623,
+      "learning_rate": 1.2571761706579216e-06,
+      "loss": 0.5508,
+      "step": 9496
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.306828303358857,
+      "learning_rate": 1.2562985546260804e-06,
+      "loss": 0.5306,
+      "step": 9497
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.9891319839435906,
+      "learning_rate": 1.2554212010114176e-06,
+      "loss": 0.784,
+      "step": 9498
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.56118675712117,
+      "learning_rate": 1.2545441098754336e-06,
+      "loss": 0.7236,
+      "step": 9499
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.921038626828936,
+      "learning_rate": 1.2536672812796057e-06,
+      "loss": 0.7664,
+      "step": 9500
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.779705067114796,
+      "learning_rate": 1.252790715285398e-06,
+      "loss": 0.7688,
+      "step": 9501
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.0120798055306865,
+      "learning_rate": 1.2519144119542526e-06,
+      "loss": 0.7212,
+      "step": 9502
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.3584976659903703,
+      "learning_rate": 1.251038371347595e-06,
+      "loss": 0.3485,
+      "step": 9503
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.913187163884777,
+      "learning_rate": 1.2501625935268325e-06,
+      "loss": 1.0688,
+      "step": 9504
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.87134469525131,
+      "learning_rate": 1.2492870785533539e-06,
+      "loss": 0.8905,
+      "step": 9505
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.8669210139522288,
+      "learning_rate": 1.248411826488527e-06,
+      "loss": 0.771,
+      "step": 9506
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.366795829316916,
+      "learning_rate": 1.2475368373937035e-06,
+      "loss": 0.4478,
+      "step": 9507
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.815116094891325,
+      "learning_rate": 1.2466621113302174e-06,
+      "loss": 0.7628,
+      "step": 9508
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.944149278920771,
+      "learning_rate": 1.2457876483593839e-06,
+      "loss": 0.5918,
+      "step": 9509
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.0388836112634285,
+      "learning_rate": 1.2449134485424969e-06,
+      "loss": 0.5214,
+      "step": 9510
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.327183866109166,
+      "learning_rate": 1.244039511940836e-06,
+      "loss": 0.8289,
+      "step": 9511
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.7530568667721487,
+      "learning_rate": 1.2431658386156576e-06,
+      "loss": 0.4686,
+      "step": 9512
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.480329489320254,
+      "learning_rate": 1.2422924286282045e-06,
+      "loss": 0.6048,
+      "step": 9513
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.1567143431506794,
+      "learning_rate": 1.2414192820396987e-06,
+      "loss": 0.4192,
+      "step": 9514
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.82164518572063,
+      "learning_rate": 1.2405463989113437e-06,
+      "loss": 0.8892,
+      "step": 9515
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.676772750999895,
+      "learning_rate": 1.2396737793043246e-06,
+      "loss": 0.6857,
+      "step": 9516
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.92710925047657,
+      "learning_rate": 1.2388014232798102e-06,
+      "loss": 0.8158,
+      "step": 9517
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.815218867450044,
+      "learning_rate": 1.237929330898946e-06,
+      "loss": 0.628,
+      "step": 9518
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5802954859651215,
+      "learning_rate": 1.2370575022228632e-06,
+      "loss": 0.2097,
+      "step": 9519
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.087610106969515,
+      "learning_rate": 1.2361859373126727e-06,
+      "loss": 0.5337,
+      "step": 9520
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.819997553881703,
+      "learning_rate": 1.2353146362294682e-06,
+      "loss": 0.5812,
+      "step": 9521
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.129135410511098,
+      "learning_rate": 1.234443599034325e-06,
+      "loss": 0.9036,
+      "step": 9522
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.448219456451351,
+      "learning_rate": 1.2335728257882962e-06,
+      "loss": 0.5669,
+      "step": 9523
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3452401635169344,
+      "learning_rate": 1.232702316552421e-06,
+      "loss": 0.2998,
+      "step": 9524
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.8183723015943,
+      "learning_rate": 1.2318320713877191e-06,
+      "loss": 0.9775,
+      "step": 9525
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.135468860988995,
+      "learning_rate": 1.230962090355189e-06,
+      "loss": 1.212,
+      "step": 9526
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.61340378905854,
+      "learning_rate": 1.2300923735158133e-06,
+      "loss": 0.8543,
+      "step": 9527
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.5275409270739684,
+      "learning_rate": 1.2292229209305567e-06,
+      "loss": 0.5539,
+      "step": 9528
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.0836387895110375,
+      "learning_rate": 1.2283537326603611e-06,
+      "loss": 0.4498,
+      "step": 9529
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.9828518185080988,
+      "learning_rate": 1.2274848087661555e-06,
+      "loss": 0.4785,
+      "step": 9530
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.4538261017643257,
+      "learning_rate": 1.2266161493088463e-06,
+      "loss": 0.5159,
+      "step": 9531
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.2069474774767754,
+      "learning_rate": 1.225747754349323e-06,
+      "loss": 0.5612,
+      "step": 9532
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.064186344941871,
+      "learning_rate": 1.2248796239484567e-06,
+      "loss": 0.9742,
+      "step": 9533
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.8123722929754025,
+      "learning_rate": 1.2240117581671013e-06,
+      "loss": 0.5447,
+      "step": 9534
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.408214932514735,
+      "learning_rate": 1.2231441570660863e-06,
+      "loss": 0.8532,
+      "step": 9535
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.284974680679537,
+      "learning_rate": 1.222276820706229e-06,
+      "loss": 1.443,
+      "step": 9536
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.978184535275699,
+      "learning_rate": 1.2214097491483262e-06,
+      "loss": 0.438,
+      "step": 9537
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.719536409936797,
+      "learning_rate": 1.2205429424531556e-06,
+      "loss": 0.823,
+      "step": 9538
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.636600975898947,
+      "learning_rate": 1.2196764006814764e-06,
+      "loss": 0.7855,
+      "step": 9539
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.313568214669807,
+      "learning_rate": 1.2188101238940309e-06,
+      "loss": 0.5635,
+      "step": 9540
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.827929767912652,
+      "learning_rate": 1.2179441121515384e-06,
+      "loss": 1.1951,
+      "step": 9541
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.83635248764761,
+      "learning_rate": 1.2170783655147056e-06,
+      "loss": 0.9488,
+      "step": 9542
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.730236472623352,
+      "learning_rate": 1.2162128840442145e-06,
+      "loss": 0.7573,
+      "step": 9543
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.8731020795937416,
+      "learning_rate": 1.2153476678007331e-06,
+      "loss": 0.7127,
+      "step": 9544
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.5919527598477154,
+      "learning_rate": 1.2144827168449109e-06,
+      "loss": 0.7943,
+      "step": 9545
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.010204035251635,
+      "learning_rate": 1.2136180312373742e-06,
+      "loss": 0.898,
+      "step": 9546
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.8695847466899043,
+      "learning_rate": 1.212753611038735e-06,
+      "loss": 0.5351,
+      "step": 9547
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.32016422722352,
+      "learning_rate": 1.2118894563095857e-06,
+      "loss": 1.0733,
+      "step": 9548
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.568638969482633,
+      "learning_rate": 1.2110255671104997e-06,
+      "loss": 0.6358,
+      "step": 9549
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.875271015558787,
+      "learning_rate": 1.2101619435020311e-06,
+      "loss": 0.9305,
+      "step": 9550
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.022347320932427,
+      "learning_rate": 1.2092985855447193e-06,
+      "loss": 0.6068,
+      "step": 9551
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.3295190510012524,
+      "learning_rate": 1.2084354932990772e-06,
+      "loss": 0.3397,
+      "step": 9552
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.2833774245157272,
+      "learning_rate": 1.207572666825606e-06,
+      "loss": 0.6509,
+      "step": 9553
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.484701560564738,
+      "learning_rate": 1.2067101061847869e-06,
+      "loss": 1.1058,
+      "step": 9554
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.690085106932299,
+      "learning_rate": 1.2058478114370803e-06,
+      "loss": 1.0041,
+      "step": 9555
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.456193791753844,
+      "learning_rate": 1.2049857826429317e-06,
+      "loss": 0.712,
+      "step": 9556
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.223199212769577,
+      "learning_rate": 1.2041240198627617e-06,
+      "loss": 0.7078,
+      "step": 9557
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.8785252470799803,
+      "learning_rate": 1.2032625231569805e-06,
+      "loss": 0.5626,
+      "step": 9558
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.4123505230738402,
+      "learning_rate": 1.2024012925859712e-06,
+      "loss": 0.6199,
+      "step": 9559
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.939580602094508,
+      "learning_rate": 1.201540328210104e-06,
+      "loss": 0.5855,
+      "step": 9560
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.135372367422248,
+      "learning_rate": 1.200679630089729e-06,
+      "loss": 0.1252,
+      "step": 9561
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.966481781191274,
+      "learning_rate": 1.1998191982851787e-06,
+      "loss": 0.9955,
+      "step": 9562
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.694849907298931,
+      "learning_rate": 1.1989590328567623e-06,
+      "loss": 0.4365,
+      "step": 9563
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.8035534047839255,
+      "learning_rate": 1.1980991338647757e-06,
+      "loss": 1.099,
+      "step": 9564
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.229682790370691,
+      "learning_rate": 1.1972395013694944e-06,
+      "loss": 0.7544,
+      "step": 9565
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.3258366679237765,
+      "learning_rate": 1.1963801354311738e-06,
+      "loss": 0.7798,
+      "step": 9566
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.395452828167987,
+      "learning_rate": 1.1955210361100521e-06,
+      "loss": 0.8726,
+      "step": 9567
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.194726586130052,
+      "learning_rate": 1.1946622034663507e-06,
+      "loss": 0.873,
+      "step": 9568
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.329469032605557,
+      "learning_rate": 1.1938036375602662e-06,
+      "loss": 0.4734,
+      "step": 9569
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.477324477489711,
+      "learning_rate": 1.1929453384519818e-06,
+      "loss": 0.8422,
+      "step": 9570
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.86605519298845,
+      "learning_rate": 1.1920873062016613e-06,
+      "loss": 0.7282,
+      "step": 9571
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8644296246619503,
+      "learning_rate": 1.1912295408694496e-06,
+      "loss": 0.3902,
+      "step": 9572
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.126756509218199,
+      "learning_rate": 1.1903720425154702e-06,
+      "loss": 0.3178,
+      "step": 9573
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.1785592662935267,
+      "learning_rate": 1.1895148111998323e-06,
+      "loss": 0.2606,
+      "step": 9574
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.8275021525166313,
+      "learning_rate": 1.1886578469826215e-06,
+      "loss": 0.5906,
+      "step": 9575
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.148790268356531,
+      "learning_rate": 1.1878011499239083e-06,
+      "loss": 0.8455,
+      "step": 9576
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.830681000824894,
+      "learning_rate": 1.1869447200837443e-06,
+      "loss": 0.701,
+      "step": 9577
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.3820428705803245,
+      "learning_rate": 1.1860885575221603e-06,
+      "loss": 0.8088,
+      "step": 9578
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.3999850220494645,
+      "learning_rate": 1.1852326622991712e-06,
+      "loss": 0.2845,
+      "step": 9579
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.5818051654165215,
+      "learning_rate": 1.1843770344747712e-06,
+      "loss": 0.8162,
+      "step": 9580
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.3386044856674983,
+      "learning_rate": 1.183521674108934e-06,
+      "loss": 0.4882,
+      "step": 9581
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.6539031385252825,
+      "learning_rate": 1.1826665812616183e-06,
+      "loss": 1.1817,
+      "step": 9582
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.833873751597924,
+      "learning_rate": 1.1818117559927622e-06,
+      "loss": 0.5528,
+      "step": 9583
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.237939933708527,
+      "learning_rate": 1.1809571983622846e-06,
+      "loss": 0.7856,
+      "step": 9584
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.505227537156948,
+      "learning_rate": 1.1801029084300891e-06,
+      "loss": 0.6273,
+      "step": 9585
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.783948077996453,
+      "learning_rate": 1.1792488862560536e-06,
+      "loss": 0.3235,
+      "step": 9586
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 6.02804589596132,
+      "learning_rate": 1.1783951319000437e-06,
+      "loss": 1.1752,
+      "step": 9587
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.4292192697448938,
+      "learning_rate": 1.177541645421904e-06,
+      "loss": 0.3607,
+      "step": 9588
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.202673607295536,
+      "learning_rate": 1.1766884268814587e-06,
+      "loss": 0.4486,
+      "step": 9589
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.989876397149936,
+      "learning_rate": 1.1758354763385154e-06,
+      "loss": 0.812,
+      "step": 9590
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.622833772162186,
+      "learning_rate": 1.174982793852864e-06,
+      "loss": 0.5832,
+      "step": 9591
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.178032237414417,
+      "learning_rate": 1.1741303794842706e-06,
+      "loss": 0.3657,
+      "step": 9592
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.852195921600761,
+      "learning_rate": 1.1732782332924874e-06,
+      "loss": 0.8374,
+      "step": 9593
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.771156324016596,
+      "learning_rate": 1.1724263553372462e-06,
+      "loss": 0.9483,
+      "step": 9594
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.292588240670132,
+      "learning_rate": 1.1715747456782594e-06,
+      "loss": 0.706,
+      "step": 9595
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.5559926633483454,
+      "learning_rate": 1.1707234043752219e-06,
+      "loss": 0.5238,
+      "step": 9596
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.615783434419688,
+      "learning_rate": 1.1698723314878102e-06,
+      "loss": 0.7433,
+      "step": 9597
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.940962409567911,
+      "learning_rate": 1.1690215270756777e-06,
+      "loss": 0.9237,
+      "step": 9598
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.6793765422907363,
+      "learning_rate": 1.168170991198464e-06,
+      "loss": 0.3631,
+      "step": 9599
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.8653290532875517,
+      "learning_rate": 1.1673207239157874e-06,
+      "loss": 0.6689,
+      "step": 9600
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.976737564961005,
+      "learning_rate": 1.1664707252872481e-06,
+      "loss": 0.6642,
+      "step": 9601
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.9914614639763992,
+      "learning_rate": 1.165620995372429e-06,
+      "loss": 0.9346,
+      "step": 9602
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.1113270695900557,
+      "learning_rate": 1.164771534230889e-06,
+      "loss": 0.7333,
+      "step": 9603
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.8905432502915245,
+      "learning_rate": 1.1639223419221756e-06,
+      "loss": 0.8442,
+      "step": 9604
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.450603824396783,
+      "learning_rate": 1.1630734185058096e-06,
+      "loss": 0.5446,
+      "step": 9605
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.014716241071041,
+      "learning_rate": 1.162224764041298e-06,
+      "loss": 0.7966,
+      "step": 9606
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.548753052212068,
+      "learning_rate": 1.1613763785881294e-06,
+      "loss": 0.8736,
+      "step": 9607
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.183787050875972,
+      "learning_rate": 1.1605282622057718e-06,
+      "loss": 0.836,
+      "step": 9608
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.605722267159306,
+      "learning_rate": 1.1596804149536723e-06,
+      "loss": 0.7694,
+      "step": 9609
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.536756127578316,
+      "learning_rate": 1.1588328368912622e-06,
+      "loss": 0.3536,
+      "step": 9610
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.338837264580741,
+      "learning_rate": 1.157985528077954e-06,
+      "loss": 0.7896,
+      "step": 9611
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.002047789875118,
+      "learning_rate": 1.1571384885731395e-06,
+      "loss": 0.8901,
+      "step": 9612
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.805848427243978,
+      "learning_rate": 1.1562917184361926e-06,
+      "loss": 0.3782,
+      "step": 9613
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.156441147164578,
+      "learning_rate": 1.1554452177264703e-06,
+      "loss": 0.7537,
+      "step": 9614
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.1989631116228505,
+      "learning_rate": 1.1545989865033047e-06,
+      "loss": 0.6675,
+      "step": 9615
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.536201511039549,
+      "learning_rate": 1.1537530248260154e-06,
+      "loss": 0.7386,
+      "step": 9616
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.430036824389245,
+      "learning_rate": 1.1529073327538997e-06,
+      "loss": 1.0954,
+      "step": 9617
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.497001618649554,
+      "learning_rate": 1.1520619103462387e-06,
+      "loss": 0.5644,
+      "step": 9618
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.06404922181396,
+      "learning_rate": 1.1512167576622906e-06,
+      "loss": 0.8917,
+      "step": 9619
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.4379380659815215,
+      "learning_rate": 1.150371874761299e-06,
+      "loss": 0.4712,
+      "step": 9620
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.695014470124851,
+      "learning_rate": 1.1495272617024839e-06,
+      "loss": 0.7112,
+      "step": 9621
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.497502645969183,
+      "learning_rate": 1.1486829185450504e-06,
+      "loss": 0.9781,
+      "step": 9622
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.00480616008217,
+      "learning_rate": 1.147838845348183e-06,
+      "loss": 0.8317,
+      "step": 9623
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.394024203815417,
+      "learning_rate": 1.1469950421710486e-06,
+      "loss": 0.744,
+      "step": 9624
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.134570910059159,
+      "learning_rate": 1.1461515090727943e-06,
+      "loss": 0.5166,
+      "step": 9625
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.2031665141097005,
+      "learning_rate": 1.1453082461125465e-06,
+      "loss": 0.7629,
+      "step": 9626
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 6.273922961464934,
+      "learning_rate": 1.1444652533494143e-06,
+      "loss": 1.3982,
+      "step": 9627
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.001798258529542,
+      "learning_rate": 1.1436225308424885e-06,
+      "loss": 0.504,
+      "step": 9628
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.180857843732888,
+      "learning_rate": 1.1427800786508402e-06,
+      "loss": 0.8142,
+      "step": 9629
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.3082931776411604,
+      "learning_rate": 1.141937896833522e-06,
+      "loss": 0.5347,
+      "step": 9630
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.208351716277365,
+      "learning_rate": 1.1410959854495684e-06,
+      "loss": 0.922,
+      "step": 9631
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 7.570266375131698,
+      "learning_rate": 1.1402543445579905e-06,
+      "loss": 0.8222,
+      "step": 9632
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.74728306378593,
+      "learning_rate": 1.1394129742177856e-06,
+      "loss": 0.7746,
+      "step": 9633
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.1241305119409484,
+      "learning_rate": 1.1385718744879298e-06,
+      "loss": 0.5867,
+      "step": 9634
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.49803535696854,
+      "learning_rate": 1.1377310454273821e-06,
+      "loss": 0.5357,
+      "step": 9635
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.578844931449383,
+      "learning_rate": 1.1368904870950782e-06,
+      "loss": 1.0146,
+      "step": 9636
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.642038477169935,
+      "learning_rate": 1.1360501995499396e-06,
+      "loss": 0.5953,
+      "step": 9637
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.7415699599739503,
+      "learning_rate": 1.135210182850865e-06,
+      "loss": 0.7982,
+      "step": 9638
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.355420188822306,
+      "learning_rate": 1.1343704370567371e-06,
+      "loss": 1.0446,
+      "step": 9639
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.726871691223727,
+      "learning_rate": 1.1335309622264184e-06,
+      "loss": 1.3793,
+      "step": 9640
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.050570243004476,
+      "learning_rate": 1.1326917584187518e-06,
+      "loss": 0.5335,
+      "step": 9641
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.4912518680649693,
+      "learning_rate": 1.1318528256925642e-06,
+      "loss": 0.5002,
+      "step": 9642
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.2478760852442887,
+      "learning_rate": 1.131014164106658e-06,
+      "loss": 0.4667,
+      "step": 9643
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.2488126936794814,
+      "learning_rate": 1.1301757737198206e-06,
+      "loss": 0.4027,
+      "step": 9644
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.655049431304579,
+      "learning_rate": 1.1293376545908202e-06,
+      "loss": 0.9878,
+      "step": 9645
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.700968314352387,
+      "learning_rate": 1.1284998067784047e-06,
+      "loss": 0.7646,
+      "step": 9646
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.00981019441697,
+      "learning_rate": 1.1276622303413043e-06,
+      "loss": 0.7577,
+      "step": 9647
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.716290315788904,
+      "learning_rate": 1.1268249253382303e-06,
+      "loss": 0.7271,
+      "step": 9648
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.505095945581661,
+      "learning_rate": 1.1259878918278717e-06,
+      "loss": 0.852,
+      "step": 9649
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.338128790190635,
+      "learning_rate": 1.1251511298689015e-06,
+      "loss": 0.4403,
+      "step": 9650
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.241555005222036,
+      "learning_rate": 1.1243146395199762e-06,
+      "loss": 0.9327,
+      "step": 9651
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.5833214319323226,
+      "learning_rate": 1.1234784208397254e-06,
+      "loss": 0.7184,
+      "step": 9652
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.954642674831938,
+      "learning_rate": 1.122642473886767e-06,
+      "loss": 0.5282,
+      "step": 9653
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.72774250321284,
+      "learning_rate": 1.121806798719698e-06,
+      "loss": 0.9484,
+      "step": 9654
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.9282920460344455,
+      "learning_rate": 1.120971395397093e-06,
+      "loss": 1.0189,
+      "step": 9655
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.462960738027389,
+      "learning_rate": 1.120136263977512e-06,
+      "loss": 0.8744,
+      "step": 9656
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.760146405525162,
+      "learning_rate": 1.1193014045194934e-06,
+      "loss": 0.4105,
+      "step": 9657
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.2457823955428218,
+      "learning_rate": 1.1184668170815572e-06,
+      "loss": 0.4571,
+      "step": 9658
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.166433479661119,
+      "learning_rate": 1.117632501722205e-06,
+      "loss": 0.7967,
+      "step": 9659
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.560238295755665,
+      "learning_rate": 1.1167984584999197e-06,
+      "loss": 0.3852,
+      "step": 9660
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.640499295917284,
+      "learning_rate": 1.1159646874731612e-06,
+      "loss": 0.7281,
+      "step": 9661
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.491885857740138,
+      "learning_rate": 1.1151311887003747e-06,
+      "loss": 0.6245,
+      "step": 9662
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.0728974812486936,
+      "learning_rate": 1.1142979622399853e-06,
+      "loss": 0.4791,
+      "step": 9663
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.048737541139456,
+      "learning_rate": 1.1134650081503978e-06,
+      "loss": 1.1579,
+      "step": 9664
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.080432675687764,
+      "learning_rate": 1.1126323264900002e-06,
+      "loss": 0.9797,
+      "step": 9665
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.7141021438171187,
+      "learning_rate": 1.1117999173171574e-06,
+      "loss": 0.2737,
+      "step": 9666
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.4687110977368856,
+      "learning_rate": 1.1109677806902203e-06,
+      "loss": 0.9324,
+      "step": 9667
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.114467256724678,
+      "learning_rate": 1.1101359166675151e-06,
+      "loss": 0.5055,
+      "step": 9668
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.174587067675621,
+      "learning_rate": 1.1093043253073538e-06,
+      "loss": 0.7105,
+      "step": 9669
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.8009407831622655,
+      "learning_rate": 1.1084730066680267e-06,
+      "loss": 1.0176,
+      "step": 9670
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 6.475202699175426,
+      "learning_rate": 1.107641960807807e-06,
+      "loss": 1.4798,
+      "step": 9671
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.953644598963639,
+      "learning_rate": 1.1068111877849448e-06,
+      "loss": 0.5494,
+      "step": 9672
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.8234011788648514,
+      "learning_rate": 1.1059806876576756e-06,
+      "loss": 0.7237,
+      "step": 9673
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.920376867534809,
+      "learning_rate": 1.1051504604842128e-06,
+      "loss": 0.6737,
+      "step": 9674
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.50548118931816,
+      "learning_rate": 1.104320506322753e-06,
+      "loss": 0.3674,
+      "step": 9675
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.377095839238889,
+      "learning_rate": 1.1034908252314714e-06,
+      "loss": 0.8886,
+      "step": 9676
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.931779261768232,
+      "learning_rate": 1.1026614172685263e-06,
+      "loss": 0.8574,
+      "step": 9677
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.4732582543823733,
+      "learning_rate": 1.1018322824920535e-06,
+      "loss": 0.4539,
+      "step": 9678
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.568627885041344,
+      "learning_rate": 1.1010034209601727e-06,
+      "loss": 0.7445,
+      "step": 9679
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.593252938454156,
+      "learning_rate": 1.1001748327309835e-06,
+      "loss": 0.8384,
+      "step": 9680
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.594480343185368,
+      "learning_rate": 1.0993465178625678e-06,
+      "loss": 1.0378,
+      "step": 9681
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.46661245479926,
+      "learning_rate": 1.0985184764129847e-06,
+      "loss": 0.8683,
+      "step": 9682
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.046667478036286,
+      "learning_rate": 1.0976907084402776e-06,
+      "loss": 0.9989,
+      "step": 9683
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.233168789512369,
+      "learning_rate": 1.0968632140024683e-06,
+      "loss": 0.7017,
+      "step": 9684
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.852547466308722,
+      "learning_rate": 1.096035993157561e-06,
+      "loss": 0.7889,
+      "step": 9685
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.7693518060992535,
+      "learning_rate": 1.09520904596354e-06,
+      "loss": 1.301,
+      "step": 9686
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2474182959207654,
+      "learning_rate": 1.0943823724783719e-06,
+      "loss": 0.1488,
+      "step": 9687
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3677448085476038,
+      "learning_rate": 1.0935559727600032e-06,
+      "loss": 0.2045,
+      "step": 9688
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.6865572644403133,
+      "learning_rate": 1.0927298468663582e-06,
+      "loss": 0.8139,
+      "step": 9689
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.141834242533534,
+      "learning_rate": 1.0919039948553467e-06,
+      "loss": 1.1825,
+      "step": 9690
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.9705959149538557,
+      "learning_rate": 1.0910784167848576e-06,
+      "loss": 0.5733,
+      "step": 9691
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.5182354410933647,
+      "learning_rate": 1.090253112712759e-06,
+      "loss": 0.4627,
+      "step": 9692
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.3906113326699754,
+      "learning_rate": 1.0894280826969022e-06,
+      "loss": 0.6566,
+      "step": 9693
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.356231320302676,
+      "learning_rate": 1.0886033267951196e-06,
+      "loss": 0.7698,
+      "step": 9694
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.742199082679943,
+      "learning_rate": 1.0877788450652199e-06,
+      "loss": 0.8193,
+      "step": 9695
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.7612987299533867,
+      "learning_rate": 1.086954637564997e-06,
+      "loss": 0.8393,
+      "step": 9696
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.8863342177328155,
+      "learning_rate": 1.0861307043522256e-06,
+      "loss": 0.8334,
+      "step": 9697
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.3466674184223657,
+      "learning_rate": 1.085307045484657e-06,
+      "loss": 0.4735,
+      "step": 9698
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.6521837509972768,
+      "learning_rate": 1.0844836610200282e-06,
+      "loss": 0.5738,
+      "step": 9699
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.035367081615931,
+      "learning_rate": 1.0836605510160558e-06,
+      "loss": 0.6813,
+      "step": 9700
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.302242602103993,
+      "learning_rate": 1.0828377155304332e-06,
+      "loss": 1.1911,
+      "step": 9701
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.31959290473899,
+      "learning_rate": 1.082015154620839e-06,
+      "loss": 1.1347,
+      "step": 9702
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 6.198033537390311,
+      "learning_rate": 1.0811928683449318e-06,
+      "loss": 0.9468,
+      "step": 9703
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.581324713236814,
+      "learning_rate": 1.0803708567603493e-06,
+      "loss": 0.756,
+      "step": 9704
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.227710907963709,
+      "learning_rate": 1.0795491199247133e-06,
+      "loss": 0.3648,
+      "step": 9705
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.7858849391276843,
+      "learning_rate": 1.0787276578956207e-06,
+      "loss": 0.6991,
+      "step": 9706
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.430426207389419,
+      "learning_rate": 1.0779064707306536e-06,
+      "loss": 1.1752,
+      "step": 9707
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.0675236089607156,
+      "learning_rate": 1.077085558487374e-06,
+      "loss": 0.5475,
+      "step": 9708
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.074170239303381,
+      "learning_rate": 1.076264921223324e-06,
+      "loss": 0.8862,
+      "step": 9709
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.14174743579678,
+      "learning_rate": 1.0754445589960273e-06,
+      "loss": 1.2746,
+      "step": 9710
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.7197812301940547,
+      "learning_rate": 1.0746244718629883e-06,
+      "loss": 0.6069,
+      "step": 9711
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.275241200789883,
+      "learning_rate": 1.0738046598816891e-06,
+      "loss": 0.6855,
+      "step": 9712
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.9473450395501235,
+      "learning_rate": 1.0729851231095983e-06,
+      "loss": 0.5234,
+      "step": 9713
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.624716296959476,
+      "learning_rate": 1.0721658616041581e-06,
+      "loss": 1.0341,
+      "step": 9714
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.18194587295983,
+      "learning_rate": 1.0713468754227968e-06,
+      "loss": 0.6545,
+      "step": 9715
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.67836566277044,
+      "learning_rate": 1.0705281646229227e-06,
+      "loss": 0.5493,
+      "step": 9716
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.899433654390494,
+      "learning_rate": 1.0697097292619241e-06,
+      "loss": 0.9894,
+      "step": 9717
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.512719535347329,
+      "learning_rate": 1.0688915693971675e-06,
+      "loss": 0.9985,
+      "step": 9718
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.598644989378368,
+      "learning_rate": 1.0680736850860034e-06,
+      "loss": 0.2486,
+      "step": 9719
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.536177573138868,
+      "learning_rate": 1.0672560763857626e-06,
+      "loss": 0.9369,
+      "step": 9720
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.5585393591174235,
+      "learning_rate": 1.066438743353755e-06,
+      "loss": 1.2031,
+      "step": 9721
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.132905482869254,
+      "learning_rate": 1.065621686047274e-06,
+      "loss": 1.0333,
+      "step": 9722
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.5796683304053936,
+      "learning_rate": 1.0648049045235891e-06,
+      "loss": 0.6483,
+      "step": 9723
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.0256589186041563,
+      "learning_rate": 1.0639883988399547e-06,
+      "loss": 0.7105,
+      "step": 9724
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 6.390922431470076,
+      "learning_rate": 1.0631721690536034e-06,
+      "loss": 0.6557,
+      "step": 9725
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.1932750602916995,
+      "learning_rate": 1.0623562152217503e-06,
+      "loss": 0.3335,
+      "step": 9726
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.586248012852279,
+      "learning_rate": 1.0615405374015913e-06,
+      "loss": 0.6929,
+      "step": 9727
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.199738806020164,
+      "learning_rate": 1.060725135650299e-06,
+      "loss": 0.1826,
+      "step": 9728
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.001062988625413,
+      "learning_rate": 1.059910010025032e-06,
+      "loss": 0.5847,
+      "step": 9729
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.126468899802749,
+      "learning_rate": 1.0590951605829247e-06,
+      "loss": 0.9041,
+      "step": 9730
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.114543696558655,
+      "learning_rate": 1.0582805873810959e-06,
+      "loss": 1.0175,
+      "step": 9731
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.226737038244887,
+      "learning_rate": 1.0574662904766432e-06,
+      "loss": 0.9064,
+      "step": 9732
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.348937215859812,
+      "learning_rate": 1.0566522699266457e-06,
+      "loss": 0.398,
+      "step": 9733
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.014006204755118,
+      "learning_rate": 1.0558385257881637e-06,
+      "loss": 0.8706,
+      "step": 9734
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.926980563123289,
+      "learning_rate": 1.0550250581182353e-06,
+      "loss": 0.4947,
+      "step": 9735
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.774309576349225,
+      "learning_rate": 1.054211866973881e-06,
+      "loss": 1.4669,
+      "step": 9736
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 6.424093652109563,
+      "learning_rate": 1.053398952412103e-06,
+      "loss": 1.2049,
+      "step": 9737
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.4743138653694388,
+      "learning_rate": 1.052586314489883e-06,
+      "loss": 0.6989,
+      "step": 9738
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.101626911918696,
+      "learning_rate": 1.051773953264183e-06,
+      "loss": 0.4607,
+      "step": 9739
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.6506214341008953,
+      "learning_rate": 1.0509618687919476e-06,
+      "loss": 1.0164,
+      "step": 9740
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.475059228286483,
+      "learning_rate": 1.0501500611300974e-06,
+      "loss": 0.4186,
+      "step": 9741
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.1277552416511982,
+      "learning_rate": 1.049338530335538e-06,
+      "loss": 0.4358,
+      "step": 9742
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.009423272641236,
+      "learning_rate": 1.0485272764651543e-06,
+      "loss": 0.7617,
+      "step": 9743
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.815574150796619,
+      "learning_rate": 1.0477162995758133e-06,
+      "loss": 0.5747,
+      "step": 9744
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.4430449005631587,
+      "learning_rate": 1.0469055997243578e-06,
+      "loss": 0.3788,
+      "step": 9745
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.962221865232792,
+      "learning_rate": 1.0460951769676175e-06,
+      "loss": 0.353,
+      "step": 9746
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.4239971821571578,
+      "learning_rate": 1.0452850313623958e-06,
+      "loss": 0.573,
+      "step": 9747
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0293272781540777,
+      "learning_rate": 1.0444751629654831e-06,
+      "loss": 0.1605,
+      "step": 9748
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.101613154249267,
+      "learning_rate": 1.0436655718336464e-06,
+      "loss": 0.6731,
+      "step": 9749
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.113365293730688,
+      "learning_rate": 1.0428562580236358e-06,
+      "loss": 0.175,
+      "step": 9750
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.105455382948563,
+      "learning_rate": 1.0420472215921807e-06,
+      "loss": 1.1378,
+      "step": 9751
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.7861055323893917,
+      "learning_rate": 1.0412384625959887e-06,
+      "loss": 1.08,
+      "step": 9752
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.675653711131224,
+      "learning_rate": 1.0404299810917523e-06,
+      "loss": 0.4479,
+      "step": 9753
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.335133964402627,
+      "learning_rate": 1.0396217771361422e-06,
+      "loss": 0.5115,
+      "step": 9754
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.980571756794141,
+      "learning_rate": 1.0388138507858098e-06,
+      "loss": 1.1913,
+      "step": 9755
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.8508596544370906,
+      "learning_rate": 1.0380062020973875e-06,
+      "loss": 0.6614,
+      "step": 9756
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.914488203777597,
+      "learning_rate": 1.037198831127489e-06,
+      "loss": 0.7279,
+      "step": 9757
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.7708154404824783,
+      "learning_rate": 1.036391737932705e-06,
+      "loss": 0.7997,
+      "step": 9758
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.380952909168642,
+      "learning_rate": 1.0355849225696102e-06,
+      "loss": 0.7888,
+      "step": 9759
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.42437832779043,
+      "learning_rate": 1.0347783850947606e-06,
+      "loss": 0.5281,
+      "step": 9760
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.508575678821504,
+      "learning_rate": 1.0339721255646885e-06,
+      "loss": 1.3033,
+      "step": 9761
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.93822300897433,
+      "learning_rate": 1.0331661440359114e-06,
+      "loss": 0.7981,
+      "step": 9762
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.311374690150685,
+      "learning_rate": 1.0323604405649224e-06,
+      "loss": 0.8647,
+      "step": 9763
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.0385444435253683,
+      "learning_rate": 1.0315550152081988e-06,
+      "loss": 0.5561,
+      "step": 9764
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.508625854582256,
+      "learning_rate": 1.0307498680221988e-06,
+      "loss": 0.4562,
+      "step": 9765
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.6243096831556865,
+      "learning_rate": 1.029944999063358e-06,
+      "loss": 0.2955,
+      "step": 9766
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.7360148103370268,
+      "learning_rate": 1.0291404083880957e-06,
+      "loss": 0.3317,
+      "step": 9767
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.045643567688735,
+      "learning_rate": 1.0283360960528104e-06,
+      "loss": 0.9755,
+      "step": 9768
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.892143673884387,
+      "learning_rate": 1.027532062113879e-06,
+      "loss": 0.8521,
+      "step": 9769
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.793368296844829,
+      "learning_rate": 1.0267283066276618e-06,
+      "loss": 1.1566,
+      "step": 9770
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9362162470033908,
+      "learning_rate": 1.0259248296504986e-06,
+      "loss": 0.1881,
+      "step": 9771
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.454044760591829,
+      "learning_rate": 1.025121631238709e-06,
+      "loss": 0.5303,
+      "step": 9772
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.8740149655278215,
+      "learning_rate": 1.0243187114485953e-06,
+      "loss": 0.4281,
+      "step": 9773
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.395325037236486,
+      "learning_rate": 1.0235160703364384e-06,
+      "loss": 0.5552,
+      "step": 9774
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.774354601146491,
+      "learning_rate": 1.022713707958498e-06,
+      "loss": 1.0098,
+      "step": 9775
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.109845239922136,
+      "learning_rate": 1.0219116243710192e-06,
+      "loss": 0.4951,
+      "step": 9776
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.589572101861237,
+      "learning_rate": 1.021109819630221e-06,
+      "loss": 0.9188,
+      "step": 9777
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.2871748538178953,
+      "learning_rate": 1.0203082937923082e-06,
+      "loss": 0.6727,
+      "step": 9778
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.300448811565389,
+      "learning_rate": 1.019507046913465e-06,
+      "loss": 0.7887,
+      "step": 9779
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.7733135121297705,
+      "learning_rate": 1.0187060790498553e-06,
+      "loss": 0.821,
+      "step": 9780
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.4690246160815486,
+      "learning_rate": 1.0179053902576214e-06,
+      "loss": 0.9153,
+      "step": 9781
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.9915957436347544,
+      "learning_rate": 1.01710498059289e-06,
+      "loss": 0.6375,
+      "step": 9782
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.057028414868053,
+      "learning_rate": 1.0163048501117657e-06,
+      "loss": 1.1214,
+      "step": 9783
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.5991348622918498,
+      "learning_rate": 1.0155049988703342e-06,
+      "loss": 0.4839,
+      "step": 9784
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.125793983286334,
+      "learning_rate": 1.014705426924663e-06,
+      "loss": 0.7087,
+      "step": 9785
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.105332917713791,
+      "learning_rate": 1.013906134330796e-06,
+      "loss": 0.5788,
+      "step": 9786
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.982479435308393,
+      "learning_rate": 1.013107121144762e-06,
+      "loss": 1.4878,
+      "step": 9787
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.090355400449551,
+      "learning_rate": 1.012308387422567e-06,
+      "loss": 0.8028,
+      "step": 9788
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.6454026800282806,
+      "learning_rate": 1.0115099332201999e-06,
+      "loss": 0.7259,
+      "step": 9789
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.3975299683156943,
+      "learning_rate": 1.01071175859363e-06,
+      "loss": 0.4905,
+      "step": 9790
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.916271590526557,
+      "learning_rate": 1.0099138635988026e-06,
+      "loss": 0.8442,
+      "step": 9791
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.5728726610791397,
+      "learning_rate": 1.00911624829165e-06,
+      "loss": 0.7227,
+      "step": 9792
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.115132988655575,
+      "learning_rate": 1.008318912728079e-06,
+      "loss": 0.6475,
+      "step": 9793
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.904296494945222,
+      "learning_rate": 1.00752185696398e-06,
+      "loss": 0.5609,
+      "step": 9794
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.899616251799633,
+      "learning_rate": 1.0067250810552236e-06,
+      "loss": 0.8103,
+      "step": 9795
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.8749327110273764,
+      "learning_rate": 1.00592858505766e-06,
+      "loss": 0.5936,
+      "step": 9796
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.029676630552301,
+      "learning_rate": 1.005132369027122e-06,
+      "loss": 0.6193,
+      "step": 9797
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.158077519218264,
+      "learning_rate": 1.0043364330194178e-06,
+      "loss": 0.7681,
+      "step": 9798
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.9746467392254266,
+      "learning_rate": 1.0035407770903405e-06,
+      "loss": 1.0636,
+      "step": 9799
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.37050920443815,
+      "learning_rate": 1.0027454012956617e-06,
+      "loss": 0.4571,
+      "step": 9800
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 6.804320790908561,
+      "learning_rate": 1.0019503056911346e-06,
+      "loss": 1.0382,
+      "step": 9801
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.35103012277236,
+      "learning_rate": 1.0011554903324928e-06,
+      "loss": 1.043,
+      "step": 9802
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.608746779516858,
+      "learning_rate": 1.0003609552754468e-06,
+      "loss": 0.7013,
+      "step": 9803
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.1014704714558783,
+      "learning_rate": 9.995667005756909e-07,
+      "loss": 0.3483,
+      "step": 9804
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 6.257377286497344,
+      "learning_rate": 9.987727262888997e-07,
+      "loss": 0.7886,
+      "step": 9805
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.308433224780674,
+      "learning_rate": 9.979790324707284e-07,
+      "loss": 0.7515,
+      "step": 9806
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.4792480740472165,
+      "learning_rate": 9.971856191768086e-07,
+      "loss": 0.9002,
+      "step": 9807
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.140277972857411,
+      "learning_rate": 9.963924864627578e-07,
+      "loss": 0.9105,
+      "step": 9808
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.914102260867357,
+      "learning_rate": 9.95599634384169e-07,
+      "loss": 0.9481,
+      "step": 9809
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.513118226771093,
+      "learning_rate": 9.948070629966183e-07,
+      "loss": 0.169,
+      "step": 9810
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.7295513628394326,
+      "learning_rate": 9.940147723556614e-07,
+      "loss": 0.6453,
+      "step": 9811
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.5774326972108734,
+      "learning_rate": 9.932227625168356e-07,
+      "loss": 0.5254,
+      "step": 9812
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.602790864794578,
+      "learning_rate": 9.924310335356563e-07,
+      "loss": 1.1936,
+      "step": 9813
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.360797507501289,
+      "learning_rate": 9.91639585467622e-07,
+      "loss": 0.8715,
+      "step": 9814
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.4461678530552238,
+      "learning_rate": 9.908484183682065e-07,
+      "loss": 0.666,
+      "step": 9815
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.760828772809022,
+      "learning_rate": 9.900575322928696e-07,
+      "loss": 0.3425,
+      "step": 9816
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.34303928784837,
+      "learning_rate": 9.892669272970485e-07,
+      "loss": 0.9513,
+      "step": 9817
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.345416542482766,
+      "learning_rate": 9.884766034361604e-07,
+      "loss": 1.0784,
+      "step": 9818
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.706912283533461,
+      "learning_rate": 9.876865607656045e-07,
+      "loss": 0.466,
+      "step": 9819
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.9258263158249984,
+      "learning_rate": 9.868967993407603e-07,
+      "loss": 0.5956,
+      "step": 9820
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.0342284251333296,
+      "learning_rate": 9.86107319216984e-07,
+      "loss": 0.5981,
+      "step": 9821
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.143853090321942,
+      "learning_rate": 9.853181204496176e-07,
+      "loss": 0.7826,
+      "step": 9822
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.106256324176616,
+      "learning_rate": 9.845292030939775e-07,
+      "loss": 0.7144,
+      "step": 9823
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.946979660030996,
+      "learning_rate": 9.837405672053651e-07,
+      "loss": 0.8796,
+      "step": 9824
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.094283608807544,
+      "learning_rate": 9.829522128390611e-07,
+      "loss": 0.678,
+      "step": 9825
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2068649294230744,
+      "learning_rate": 9.821641400503235e-07,
+      "loss": 0.1685,
+      "step": 9826
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.9386882985967446,
+      "learning_rate": 9.813763488943946e-07,
+      "loss": 0.5521,
+      "step": 9827
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.1806468055796757,
+      "learning_rate": 9.80588839426494e-07,
+      "loss": 0.5293,
+      "step": 9828
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.376950259944358,
+      "learning_rate": 9.798016117018233e-07,
+      "loss": 1.2063,
+      "step": 9829
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.6004839039357366,
+      "learning_rate": 9.790146657755633e-07,
+      "loss": 0.7438,
+      "step": 9830
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.17331895871248,
+      "learning_rate": 9.782280017028777e-07,
+      "loss": 0.9275,
+      "step": 9831
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.179139915704291,
+      "learning_rate": 9.774416195389046e-07,
+      "loss": 0.4853,
+      "step": 9832
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.576781743034318,
+      "learning_rate": 9.766555193387683e-07,
+      "loss": 1.141,
+      "step": 9833
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.229859663170041,
+      "learning_rate": 9.7586970115757e-07,
+      "loss": 1.069,
+      "step": 9834
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.163122161832738,
+      "learning_rate": 9.750841650503928e-07,
+      "loss": 1.1175,
+      "step": 9835
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2391700379530222,
+      "learning_rate": 9.742989110723e-07,
+      "loss": 0.1761,
+      "step": 9836
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.460523201070169,
+      "learning_rate": 9.735139392783326e-07,
+      "loss": 0.3726,
+      "step": 9837
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.9994436411899814,
+      "learning_rate": 9.727292497235151e-07,
+      "loss": 0.8943,
+      "step": 9838
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.0603174081227125,
+      "learning_rate": 9.719448424628514e-07,
+      "loss": 0.6165,
+      "step": 9839
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.8683916770844027,
+      "learning_rate": 9.711607175513228e-07,
+      "loss": 0.7775,
+      "step": 9840
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.3962013038998338,
+      "learning_rate": 9.70376875043894e-07,
+      "loss": 0.5398,
+      "step": 9841
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.466230612997388,
+      "learning_rate": 9.695933149955111e-07,
+      "loss": 0.8821,
+      "step": 9842
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.126243935036026,
+      "learning_rate": 9.688100374610953e-07,
+      "loss": 0.5089,
+      "step": 9843
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.075033137777421,
+      "learning_rate": 9.68027042495552e-07,
+      "loss": 0.8537,
+      "step": 9844
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.044256011589565,
+      "learning_rate": 9.672443301537654e-07,
+      "loss": 0.8834,
+      "step": 9845
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.9775589520839407,
+      "learning_rate": 9.664619004906007e-07,
+      "loss": 0.7768,
+      "step": 9846
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.532802096795839,
+      "learning_rate": 9.65679753560903e-07,
+      "loss": 0.7676,
+      "step": 9847
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.9665152868615134,
+      "learning_rate": 9.648978894194983e-07,
+      "loss": 0.7077,
+      "step": 9848
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.497109619755261,
+      "learning_rate": 9.641163081211891e-07,
+      "loss": 0.542,
+      "step": 9849
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.00698599759604,
+      "learning_rate": 9.633350097207628e-07,
+      "loss": 0.8032,
+      "step": 9850
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.3863573071418904,
+      "learning_rate": 9.62553994272985e-07,
+      "loss": 0.5541,
+      "step": 9851
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.247677039720678,
+      "learning_rate": 9.61773261832601e-07,
+      "loss": 0.5179,
+      "step": 9852
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.8418076360091997,
+      "learning_rate": 9.609928124543376e-07,
+      "loss": 0.6889,
+      "step": 9853
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.118237736916841,
+      "learning_rate": 9.602126461929002e-07,
+      "loss": 0.8431,
+      "step": 9854
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.686936764023641,
+      "learning_rate": 9.594327631029753e-07,
+      "loss": 0.2751,
+      "step": 9855
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.640907225065898,
+      "learning_rate": 9.586531632392282e-07,
+      "loss": 0.5965,
+      "step": 9856
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.207380321843211,
+      "learning_rate": 9.578738466563065e-07,
+      "loss": 0.2198,
+      "step": 9857
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.2635923528591935,
+      "learning_rate": 9.570948134088364e-07,
+      "loss": 0.7252,
+      "step": 9858
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.105489324566011,
+      "learning_rate": 9.563160635514252e-07,
+      "loss": 0.6897,
+      "step": 9859
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.64835001657043,
+      "learning_rate": 9.55537597138661e-07,
+      "loss": 0.6154,
+      "step": 9860
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.7978146569040385,
+      "learning_rate": 9.547594142251089e-07,
+      "loss": 0.7363,
+      "step": 9861
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7402329161412262,
+      "learning_rate": 9.539815148653163e-07,
+      "loss": 0.1861,
+      "step": 9862
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.53530084807777,
+      "learning_rate": 9.532038991138115e-07,
+      "loss": 0.863,
+      "step": 9863
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.914928019255762,
+      "learning_rate": 9.524265670251015e-07,
+      "loss": 0.7861,
+      "step": 9864
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.5257229406622033,
+      "learning_rate": 9.516495186536751e-07,
+      "loss": 0.7089,
+      "step": 9865
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.417659463572757,
+      "learning_rate": 9.508727540539981e-07,
+      "loss": 1.1149,
+      "step": 9866
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.709118153703828,
+      "learning_rate": 9.500962732805192e-07,
+      "loss": 0.6207,
+      "step": 9867
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.524006700894052,
+      "learning_rate": 9.493200763876658e-07,
+      "loss": 0.7884,
+      "step": 9868
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.996789862109213,
+      "learning_rate": 9.485441634298482e-07,
+      "loss": 0.5168,
+      "step": 9869
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.0079425193436378,
+      "learning_rate": 9.477685344614517e-07,
+      "loss": 0.5929,
+      "step": 9870
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.3091738066070024,
+      "learning_rate": 9.469931895368462e-07,
+      "loss": 0.5454,
+      "step": 9871
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.8737147416684445,
+      "learning_rate": 9.462181287103783e-07,
+      "loss": 1.1663,
+      "step": 9872
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.910258465431136,
+      "learning_rate": 9.454433520363776e-07,
+      "loss": 0.5832,
+      "step": 9873
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.3697371231984325,
+      "learning_rate": 9.446688595691522e-07,
+      "loss": 0.9605,
+      "step": 9874
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.0701907349081323,
+      "learning_rate": 9.438946513629915e-07,
+      "loss": 0.7593,
+      "step": 9875
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.494410910517011,
+      "learning_rate": 9.431207274721627e-07,
+      "loss": 0.8716,
+      "step": 9876
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.872713129768486,
+      "learning_rate": 9.423470879509172e-07,
+      "loss": 1.1377,
+      "step": 9877
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.8073964996011616,
+      "learning_rate": 9.415737328534802e-07,
+      "loss": 0.5902,
+      "step": 9878
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.960128248000555,
+      "learning_rate": 9.408006622340627e-07,
+      "loss": 0.6662,
+      "step": 9879
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.8411134500064015,
+      "learning_rate": 9.400278761468523e-07,
+      "loss": 0.6525,
+      "step": 9880
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.271183552028433,
+      "learning_rate": 9.392553746460193e-07,
+      "loss": 0.8974,
+      "step": 9881
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.6937126757581447,
+      "learning_rate": 9.384831577857135e-07,
+      "loss": 0.8745,
+      "step": 9882
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.918194333309644,
+      "learning_rate": 9.37711225620061e-07,
+      "loss": 0.6488,
+      "step": 9883
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.768034398152464,
+      "learning_rate": 9.36939578203172e-07,
+      "loss": 0.5795,
+      "step": 9884
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.5521280178173114,
+      "learning_rate": 9.361682155891382e-07,
+      "loss": 0.492,
+      "step": 9885
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.010274233421411,
+      "learning_rate": 9.353971378320248e-07,
+      "loss": 0.9207,
+      "step": 9886
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6367400905396001,
+      "learning_rate": 9.346263449858828e-07,
+      "loss": 0.3406,
+      "step": 9887
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.299231533318858,
+      "learning_rate": 9.338558371047429e-07,
+      "loss": 0.9699,
+      "step": 9888
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.455281392120958,
+      "learning_rate": 9.33085614242612e-07,
+      "loss": 0.479,
+      "step": 9889
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.576101144784835,
+      "learning_rate": 9.323156764534797e-07,
+      "loss": 0.8125,
+      "step": 9890
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 6.001248491978731,
+      "learning_rate": 9.315460237913159e-07,
+      "loss": 0.9924,
+      "step": 9891
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.043072769369696,
+      "learning_rate": 9.3077665631007e-07,
+      "loss": 1.0591,
+      "step": 9892
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.4866963629472107,
+      "learning_rate": 9.300075740636716e-07,
+      "loss": 0.3196,
+      "step": 9893
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.8031433196931106,
+      "learning_rate": 9.292387771060302e-07,
+      "loss": 0.7287,
+      "step": 9894
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.311911836138839,
+      "learning_rate": 9.284702654910338e-07,
+      "loss": 0.5071,
+      "step": 9895
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.4149683062187797,
+      "learning_rate": 9.277020392725522e-07,
+      "loss": 0.3489,
+      "step": 9896
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.1060268580376,
+      "learning_rate": 9.269340985044345e-07,
+      "loss": 0.9197,
+      "step": 9897
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2150900519384455,
+      "learning_rate": 9.261664432405109e-07,
+      "loss": 0.1616,
+      "step": 9898
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.005804801362847,
+      "learning_rate": 9.253990735345914e-07,
+      "loss": 0.6852,
+      "step": 9899
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.974940928049727,
+      "learning_rate": 9.246319894404632e-07,
+      "loss": 0.7249,
+      "step": 9900
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.038730562058925,
+      "learning_rate": 9.238651910118973e-07,
+      "loss": 0.855,
+      "step": 9901
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.3181916794751687,
+      "learning_rate": 9.230986783026413e-07,
+      "loss": 0.4042,
+      "step": 9902
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.576947794174118,
+      "learning_rate": 9.223324513664245e-07,
+      "loss": 0.7459,
+      "step": 9903
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.720413585163589,
+      "learning_rate": 9.215665102569577e-07,
+      "loss": 0.7138,
+      "step": 9904
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.406165342302664,
+      "learning_rate": 9.208008550279296e-07,
+      "loss": 0.5856,
+      "step": 9905
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.613024596190667,
+      "learning_rate": 9.20035485733008e-07,
+      "loss": 0.5525,
+      "step": 9906
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.11050019329382,
+      "learning_rate": 9.192704024258426e-07,
+      "loss": 1.169,
+      "step": 9907
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.645508394267574,
+      "learning_rate": 9.185056051600627e-07,
+      "loss": 0.8694,
+      "step": 9908
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.5518757837879007,
+      "learning_rate": 9.177410939892772e-07,
+      "loss": 0.5554,
+      "step": 9909
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.638157839908345,
+      "learning_rate": 9.169768689670749e-07,
+      "loss": 0.9099,
+      "step": 9910
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.268801656265276,
+      "learning_rate": 9.162129301470258e-07,
+      "loss": 0.8827,
+      "step": 9911
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.9296386685375695,
+      "learning_rate": 9.154492775826762e-07,
+      "loss": 0.8059,
+      "step": 9912
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9105302074156747,
+      "learning_rate": 9.146859113275569e-07,
+      "loss": 0.333,
+      "step": 9913
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.412695156513237,
+      "learning_rate": 9.13922831435175e-07,
+      "loss": 0.6103,
+      "step": 9914
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.5759420864804925,
+      "learning_rate": 9.131600379590222e-07,
+      "loss": 0.7136,
+      "step": 9915
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.7030317891351676,
+      "learning_rate": 9.123975309525629e-07,
+      "loss": 0.7758,
+      "step": 9916
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.236943959535839,
+      "learning_rate": 9.116353104692488e-07,
+      "loss": 0.8042,
+      "step": 9917
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.01181731080616,
+      "learning_rate": 9.10873376562505e-07,
+      "loss": 0.5553,
+      "step": 9918
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.9999144714800794,
+      "learning_rate": 9.10111729285742e-07,
+      "loss": 0.4767,
+      "step": 9919
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.079564436691244,
+      "learning_rate": 9.093503686923477e-07,
+      "loss": 0.7797,
+      "step": 9920
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.242097952262364,
+      "learning_rate": 9.0858929483569e-07,
+      "loss": 0.6149,
+      "step": 9921
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.012721973636521,
+      "learning_rate": 9.078285077691179e-07,
+      "loss": 0.6988,
+      "step": 9922
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9548161095069063,
+      "learning_rate": 9.07068007545957e-07,
+      "loss": 0.2971,
+      "step": 9923
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.129316033445483,
+      "learning_rate": 9.063077942195164e-07,
+      "loss": 0.8209,
+      "step": 9924
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.084635351098784,
+      "learning_rate": 9.055478678430835e-07,
+      "loss": 1.014,
+      "step": 9925
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.8229568048217413,
+      "learning_rate": 9.047882284699255e-07,
+      "loss": 0.555,
+      "step": 9926
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.2574689239458445,
+      "learning_rate": 9.040288761532911e-07,
+      "loss": 0.4482,
+      "step": 9927
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.806892944470939,
+      "learning_rate": 9.032698109464072e-07,
+      "loss": 0.5379,
+      "step": 9928
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.991868694121024,
+      "learning_rate": 9.0251103290248e-07,
+      "loss": 1.1004,
+      "step": 9929
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.024159326541445,
+      "learning_rate": 9.017525420746964e-07,
+      "loss": 0.703,
+      "step": 9930
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.5474833957255796,
+      "learning_rate": 9.009943385162256e-07,
+      "loss": 0.7723,
+      "step": 9931
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.7371136350187237,
+      "learning_rate": 9.002364222802118e-07,
+      "loss": 0.5427,
+      "step": 9932
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.6272837149566324,
+      "learning_rate": 8.994787934197819e-07,
+      "loss": 0.8529,
+      "step": 9933
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.4494997421343547,
+      "learning_rate": 8.987214519880449e-07,
+      "loss": 0.5914,
+      "step": 9934
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.044674802676371,
+      "learning_rate": 8.979643980380837e-07,
+      "loss": 0.4443,
+      "step": 9935
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.450958132122734,
+      "learning_rate": 8.972076316229661e-07,
+      "loss": 0.8514,
+      "step": 9936
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.997254308784755,
+      "learning_rate": 8.964511527957382e-07,
+      "loss": 0.981,
+      "step": 9937
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.8162827170779,
+      "learning_rate": 8.956949616094257e-07,
+      "loss": 0.9472,
+      "step": 9938
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6060584816750842,
+      "learning_rate": 8.949390581170341e-07,
+      "loss": 0.2576,
+      "step": 9939
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.704545234891308,
+      "learning_rate": 8.941834423715512e-07,
+      "loss": 0.5923,
+      "step": 9940
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.3713590366128905,
+      "learning_rate": 8.934281144259388e-07,
+      "loss": 0.5801,
+      "step": 9941
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.4640626424930945,
+      "learning_rate": 8.926730743331436e-07,
+      "loss": 0.4416,
+      "step": 9942
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.604280564084286,
+      "learning_rate": 8.919183221460909e-07,
+      "loss": 0.6132,
+      "step": 9943
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.4174402956829706,
+      "learning_rate": 8.911638579176851e-07,
+      "loss": 0.5593,
+      "step": 9944
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.2855548280997926,
+      "learning_rate": 8.904096817008129e-07,
+      "loss": 1.0127,
+      "step": 9945
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.4040226749869,
+      "learning_rate": 8.896557935483352e-07,
+      "loss": 0.7921,
+      "step": 9946
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.270894329429153,
+      "learning_rate": 8.889021935130987e-07,
+      "loss": 0.8391,
+      "step": 9947
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.358811433320635,
+      "learning_rate": 8.881488816479278e-07,
+      "loss": 1.1271,
+      "step": 9948
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.4049660304861966,
+      "learning_rate": 8.873958580056241e-07,
+      "loss": 0.3162,
+      "step": 9949
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.207338887500905,
+      "learning_rate": 8.866431226389727e-07,
+      "loss": 0.9544,
+      "step": 9950
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.194212944720493,
+      "learning_rate": 8.858906756007385e-07,
+      "loss": 0.7725,
+      "step": 9951
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.024844323097988,
+      "learning_rate": 8.851385169436616e-07,
+      "loss": 0.7606,
+      "step": 9952
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.96861080229688,
+      "learning_rate": 8.843866467204671e-07,
+      "loss": 1.0984,
+      "step": 9953
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.333868653973619,
+      "learning_rate": 8.836350649838576e-07,
+      "loss": 0.3738,
+      "step": 9954
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.347678846049028,
+      "learning_rate": 8.828837717865151e-07,
+      "loss": 0.5327,
+      "step": 9955
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.5626055060594073,
+      "learning_rate": 8.821327671811025e-07,
+      "loss": 0.6182,
+      "step": 9956
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.323903238945544,
+      "learning_rate": 8.813820512202637e-07,
+      "loss": 0.757,
+      "step": 9957
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.207498333234816,
+      "learning_rate": 8.80631623956617e-07,
+      "loss": 0.8589,
+      "step": 9958
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.567352678695993,
+      "learning_rate": 8.798814854427661e-07,
+      "loss": 0.2841,
+      "step": 9959
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.4653789463370224,
+      "learning_rate": 8.791316357312923e-07,
+      "loss": 0.8585,
+      "step": 9960
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.159159522826038,
+      "learning_rate": 8.783820748747568e-07,
+      "loss": 0.6917,
+      "step": 9961
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.236044226099467,
+      "learning_rate": 8.776328029257014e-07,
+      "loss": 0.9786,
+      "step": 9962
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.8408160000979645,
+      "learning_rate": 8.768838199366448e-07,
+      "loss": 0.8945,
+      "step": 9963
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 6.087986763963806,
+      "learning_rate": 8.761351259600904e-07,
+      "loss": 0.8972,
+      "step": 9964
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.2336086813978655,
+      "learning_rate": 8.753867210485145e-07,
+      "loss": 0.4059,
+      "step": 9965
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.946149759152611,
+      "learning_rate": 8.746386052543793e-07,
+      "loss": 0.5339,
+      "step": 9966
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.999607721075151,
+      "learning_rate": 8.738907786301242e-07,
+      "loss": 0.8304,
+      "step": 9967
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.461783597333245,
+      "learning_rate": 8.731432412281705e-07,
+      "loss": 1.3311,
+      "step": 9968
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.048219803158993,
+      "learning_rate": 8.723959931009135e-07,
+      "loss": 0.4584,
+      "step": 9969
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2166370546771277,
+      "learning_rate": 8.716490343007344e-07,
+      "loss": 0.198,
+      "step": 9970
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.636804300884971,
+      "learning_rate": 8.709023648799908e-07,
+      "loss": 0.5062,
+      "step": 9971
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.841325653511893,
+      "learning_rate": 8.701559848910224e-07,
+      "loss": 0.3741,
+      "step": 9972
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.226905068076976,
+      "learning_rate": 8.694098943861457e-07,
+      "loss": 1.0037,
+      "step": 9973
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.09276388083209,
+      "learning_rate": 8.686640934176604e-07,
+      "loss": 0.5478,
+      "step": 9974
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.2628753408620814,
+      "learning_rate": 8.67918582037841e-07,
+      "loss": 0.6117,
+      "step": 9975
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 6.094075317614659,
+      "learning_rate": 8.671733602989463e-07,
+      "loss": 1.5157,
+      "step": 9976
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.9949040744102318,
+      "learning_rate": 8.664284282532132e-07,
+      "loss": 0.4868,
+      "step": 9977
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.580425700118394,
+      "learning_rate": 8.656837859528589e-07,
+      "loss": 0.5457,
+      "step": 9978
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.7619693330016513,
+      "learning_rate": 8.649394334500777e-07,
+      "loss": 0.5573,
+      "step": 9979
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.4998835827530375,
+      "learning_rate": 8.641953707970468e-07,
+      "loss": 0.3458,
+      "step": 9980
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.5884753213983136,
+      "learning_rate": 8.634515980459207e-07,
+      "loss": 0.4565,
+      "step": 9981
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.5308432908564105,
+      "learning_rate": 8.627081152488353e-07,
+      "loss": 0.6351,
+      "step": 9982
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.6583575584188517,
+      "learning_rate": 8.619649224579051e-07,
+      "loss": 0.7003,
+      "step": 9983
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.610901179277053,
+      "learning_rate": 8.612220197252257e-07,
+      "loss": 0.8103,
+      "step": 9984
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.144654102560811,
+      "learning_rate": 8.604794071028716e-07,
+      "loss": 0.6213,
+      "step": 9985
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.693232625300068,
+      "learning_rate": 8.597370846428943e-07,
+      "loss": 1.0916,
+      "step": 9986
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.512234857020179,
+      "learning_rate": 8.58995052397329e-07,
+      "loss": 0.8597,
+      "step": 9987
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.800831230839158,
+      "learning_rate": 8.582533104181889e-07,
+      "loss": 0.3972,
+      "step": 9988
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.9638615202286096,
+      "learning_rate": 8.575118587574666e-07,
+      "loss": 0.4356,
+      "step": 9989
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.119015109367693,
+      "learning_rate": 8.567706974671353e-07,
+      "loss": 1.125,
+      "step": 9990
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.347167414230204,
+      "learning_rate": 8.560298265991473e-07,
+      "loss": 0.9594,
+      "step": 9991
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.638443725545612,
+      "learning_rate": 8.55289246205433e-07,
+      "loss": 1.0653,
+      "step": 9992
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.2924314784327344,
+      "learning_rate": 8.54548956337905e-07,
+      "loss": 0.4807,
+      "step": 9993
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5310620171207978,
+      "learning_rate": 8.538089570484548e-07,
+      "loss": 0.1781,
+      "step": 9994
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.353735075993639,
+      "learning_rate": 8.530692483889514e-07,
+      "loss": 0.1444,
+      "step": 9995
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.664733022307317,
+      "learning_rate": 8.523298304112465e-07,
+      "loss": 0.4313,
+      "step": 9996
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.8533237551889785,
+      "learning_rate": 8.515907031671705e-07,
+      "loss": 0.5627,
+      "step": 9997
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.1254378915318313,
+      "learning_rate": 8.508518667085314e-07,
+      "loss": 0.4626,
+      "step": 9998
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.272195748833848,
+      "learning_rate": 8.501133210871188e-07,
+      "loss": 0.5535,
+      "step": 9999
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.056707904463845,
+      "learning_rate": 8.493750663547024e-07,
+      "loss": 0.6391,
+      "step": 10000
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.571965715095055,
+      "learning_rate": 8.486371025630302e-07,
+      "loss": 0.8858,
+      "step": 10001
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.207486077789943,
+      "learning_rate": 8.478994297638316e-07,
+      "loss": 0.7767,
+      "step": 10002
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9268246492293135,
+      "learning_rate": 8.471620480088117e-07,
+      "loss": 0.2496,
+      "step": 10003
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.2457661811966805,
+      "learning_rate": 8.464249573496591e-07,
+      "loss": 0.7234,
+      "step": 10004
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.185701483125576,
+      "learning_rate": 8.456881578380405e-07,
+      "loss": 0.683,
+      "step": 10005
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 7.215225747900721,
+      "learning_rate": 8.449516495256022e-07,
+      "loss": 1.4064,
+      "step": 10006
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.90089749768446,
+      "learning_rate": 8.442154324639706e-07,
+      "loss": 0.8309,
+      "step": 10007
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.198820210759882,
+      "learning_rate": 8.434795067047524e-07,
+      "loss": 0.3133,
+      "step": 10008
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.781826961822043,
+      "learning_rate": 8.427438722995301e-07,
+      "loss": 0.3979,
+      "step": 10009
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.160628587137114,
+      "learning_rate": 8.420085292998714e-07,
+      "loss": 0.7002,
+      "step": 10010
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.262311883425024,
+      "learning_rate": 8.412734777573178e-07,
+      "loss": 0.7013,
+      "step": 10011
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.31000974324674,
+      "learning_rate": 8.405387177233948e-07,
+      "loss": 0.5945,
+      "step": 10012
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.785005472523334,
+      "learning_rate": 8.398042492496056e-07,
+      "loss": 0.3029,
+      "step": 10013
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.1983697979929,
+      "learning_rate": 8.390700723874346e-07,
+      "loss": 0.7723,
+      "step": 10014
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.1287679927077345,
+      "learning_rate": 8.383361871883417e-07,
+      "loss": 0.4062,
+      "step": 10015
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.16968495234099,
+      "learning_rate": 8.376025937037702e-07,
+      "loss": 0.644,
+      "step": 10016
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.445627009539799,
+      "learning_rate": 8.368692919851424e-07,
+      "loss": 0.8638,
+      "step": 10017
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.5414590746261316,
+      "learning_rate": 8.361362820838593e-07,
+      "loss": 0.492,
+      "step": 10018
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.417243551042646,
+      "learning_rate": 8.354035640513014e-07,
+      "loss": 1.3996,
+      "step": 10019
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.7135312777360125,
+      "learning_rate": 8.346711379388306e-07,
+      "loss": 0.684,
+      "step": 10020
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7078717728104544,
+      "learning_rate": 8.33939003797784e-07,
+      "loss": 0.268,
+      "step": 10021
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.50771291363033,
+      "learning_rate": 8.332071616794829e-07,
+      "loss": 0.6469,
+      "step": 10022
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.7552968502155464,
+      "learning_rate": 8.324756116352256e-07,
+      "loss": 0.8547,
+      "step": 10023
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.2278798354341065,
+      "learning_rate": 8.317443537162922e-07,
+      "loss": 0.9567,
+      "step": 10024
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.301315101902168,
+      "learning_rate": 8.310133879739379e-07,
+      "loss": 1.1694,
+      "step": 10025
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.3715012305367575,
+      "learning_rate": 8.302827144594028e-07,
+      "loss": 0.5722,
+      "step": 10026
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.066136984584118,
+      "learning_rate": 8.295523332239014e-07,
+      "loss": 0.5825,
+      "step": 10027
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.7098705094390225,
+      "learning_rate": 8.288222443186317e-07,
+      "loss": 0.472,
+      "step": 10028
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.432335998365689,
+      "learning_rate": 8.280924477947699e-07,
+      "loss": 0.8808,
+      "step": 10029
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.0500464971568686,
+      "learning_rate": 8.273629437034708e-07,
+      "loss": 0.5363,
+      "step": 10030
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.639329475382795,
+      "learning_rate": 8.266337320958718e-07,
+      "loss": 0.8161,
+      "step": 10031
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.796479380976112,
+      "learning_rate": 8.25904813023084e-07,
+      "loss": 1.5202,
+      "step": 10032
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.730708086836667,
+      "learning_rate": 8.251761865362035e-07,
+      "loss": 0.6162,
+      "step": 10033
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.6349957122901544,
+      "learning_rate": 8.244478526863026e-07,
+      "loss": 0.4139,
+      "step": 10034
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.723412902142466,
+      "learning_rate": 8.23719811524436e-07,
+      "loss": 0.4541,
+      "step": 10035
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.794269298914522,
+      "learning_rate": 8.229920631016353e-07,
+      "loss": 0.684,
+      "step": 10036
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.290127336830074,
+      "learning_rate": 8.222646074689133e-07,
+      "loss": 1.117,
+      "step": 10037
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.5149409100032685,
+      "learning_rate": 8.215374446772595e-07,
+      "loss": 0.9134,
+      "step": 10038
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.9944726149977607,
+      "learning_rate": 8.208105747776468e-07,
+      "loss": 0.6517,
+      "step": 10039
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 6.588783694265223,
+      "learning_rate": 8.200839978210256e-07,
+      "loss": 1.1527,
+      "step": 10040
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.934127704405573,
+      "learning_rate": 8.193577138583242e-07,
+      "loss": 0.7983,
+      "step": 10041
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.9057358963851736,
+      "learning_rate": 8.186317229404523e-07,
+      "loss": 0.4775,
+      "step": 10042
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.819537120002851,
+      "learning_rate": 8.179060251183007e-07,
+      "loss": 1.0176,
+      "step": 10043
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.3721447545738563,
+      "learning_rate": 8.171806204427351e-07,
+      "loss": 0.3996,
+      "step": 10044
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.3969713669358352,
+      "learning_rate": 8.164555089646048e-07,
+      "loss": 0.6868,
+      "step": 10045
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.814354675455581,
+      "learning_rate": 8.157306907347357e-07,
+      "loss": 0.8552,
+      "step": 10046
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.827139451801565,
+      "learning_rate": 8.150061658039354e-07,
+      "loss": 0.7606,
+      "step": 10047
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.661073688282505,
+      "learning_rate": 8.142819342229913e-07,
+      "loss": 0.8016,
+      "step": 10048
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.984591847641335,
+      "learning_rate": 8.135579960426659e-07,
+      "loss": 0.4342,
+      "step": 10049
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.5052051722141777,
+      "learning_rate": 8.12834351313705e-07,
+      "loss": 0.3033,
+      "step": 10050
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.131568483230662,
+      "learning_rate": 8.121110000868343e-07,
+      "loss": 0.9707,
+      "step": 10051
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.872821654104947,
+      "learning_rate": 8.113879424127564e-07,
+      "loss": 0.9356,
+      "step": 10052
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.499040191623966,
+      "learning_rate": 8.106651783421543e-07,
+      "loss": 0.9858,
+      "step": 10053
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.2972781125121213,
+      "learning_rate": 8.099427079256928e-07,
+      "loss": 0.3593,
+      "step": 10054
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.9651309608674,
+      "learning_rate": 8.092205312140111e-07,
+      "loss": 0.9283,
+      "step": 10055
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.2035955637071405,
+      "learning_rate": 8.084986482577323e-07,
+      "loss": 1.4109,
+      "step": 10056
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8922151956035511,
+      "learning_rate": 8.077770591074574e-07,
+      "loss": 0.3467,
+      "step": 10057
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.039653787579526,
+      "learning_rate": 8.070557638137649e-07,
+      "loss": 1.1254,
+      "step": 10058
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.5731288296777097,
+      "learning_rate": 8.063347624272156e-07,
+      "loss": 0.6772,
+      "step": 10059
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.200731054783732,
+      "learning_rate": 8.056140549983499e-07,
+      "loss": 0.544,
+      "step": 10060
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.2915502818258195,
+      "learning_rate": 8.048936415776837e-07,
+      "loss": 0.6498,
+      "step": 10061
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.621791096115158,
+      "learning_rate": 8.041735222157159e-07,
+      "loss": 0.8866,
+      "step": 10062
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.5508307845696594,
+      "learning_rate": 8.034536969629242e-07,
+      "loss": 0.5232,
+      "step": 10063
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1309327404270295,
+      "learning_rate": 8.027341658697646e-07,
+      "loss": 0.1377,
+      "step": 10064
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.6677811482716183,
+      "learning_rate": 8.020149289866746e-07,
+      "loss": 0.8675,
+      "step": 10065
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.676930005706927,
+      "learning_rate": 8.012959863640674e-07,
+      "loss": 1.0361,
+      "step": 10066
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.9075167493127307,
+      "learning_rate": 8.005773380523386e-07,
+      "loss": 0.974,
+      "step": 10067
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.8272851825025835,
+      "learning_rate": 7.998589841018622e-07,
+      "loss": 0.3865,
+      "step": 10068
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.773220858860217,
+      "learning_rate": 7.991409245629922e-07,
+      "loss": 0.6879,
+      "step": 10069
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.78635399173605,
+      "learning_rate": 7.984231594860614e-07,
+      "loss": 0.6683,
+      "step": 10070
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.5489463805363006,
+      "learning_rate": 7.977056889213831e-07,
+      "loss": 0.8355,
+      "step": 10071
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.73856111582991,
+      "learning_rate": 7.969885129192456e-07,
+      "loss": 0.7822,
+      "step": 10072
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.000026482958678,
+      "learning_rate": 7.962716315299235e-07,
+      "loss": 1.1317,
+      "step": 10073
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.8629512984605703,
+      "learning_rate": 7.955550448036642e-07,
+      "loss": 0.4413,
+      "step": 10074
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.980903230806037,
+      "learning_rate": 7.948387527906987e-07,
+      "loss": 0.4457,
+      "step": 10075
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.403371770741651,
+      "learning_rate": 7.941227555412351e-07,
+      "loss": 0.8789,
+      "step": 10076
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.615071009109892,
+      "learning_rate": 7.934070531054638e-07,
+      "loss": 0.5785,
+      "step": 10077
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.102659969057406,
+      "learning_rate": 7.926916455335498e-07,
+      "loss": 0.8829,
+      "step": 10078
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.573077847289734,
+      "learning_rate": 7.919765328756407e-07,
+      "loss": 0.93,
+      "step": 10079
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.136876063217532,
+      "learning_rate": 7.912617151818636e-07,
+      "loss": 0.6846,
+      "step": 10080
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.4749603064475245,
+      "learning_rate": 7.90547192502324e-07,
+      "loss": 0.8403,
+      "step": 10081
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.909745880030514,
+      "learning_rate": 7.898329648871067e-07,
+      "loss": 0.6864,
+      "step": 10082
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5567840556596326,
+      "learning_rate": 7.891190323862762e-07,
+      "loss": 0.1979,
+      "step": 10083
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.4274023280178167,
+      "learning_rate": 7.884053950498754e-07,
+      "loss": 0.3834,
+      "step": 10084
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.737814252469486,
+      "learning_rate": 7.87692052927927e-07,
+      "loss": 0.6148,
+      "step": 10085
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.9986161359124175,
+      "learning_rate": 7.869790060704341e-07,
+      "loss": 1.178,
+      "step": 10086
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.548081185556967,
+      "learning_rate": 7.862662545273786e-07,
+      "loss": 0.7753,
+      "step": 10087
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.760979136167705,
+      "learning_rate": 7.855537983487194e-07,
+      "loss": 0.3284,
+      "step": 10088
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.172216511458501,
+      "learning_rate": 7.848416375843987e-07,
+      "loss": 0.5237,
+      "step": 10089
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.0377431401690775,
+      "learning_rate": 7.841297722843333e-07,
+      "loss": 0.6412,
+      "step": 10090
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.928285806833154,
+      "learning_rate": 7.834182024984238e-07,
+      "loss": 0.5305,
+      "step": 10091
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.4613951992850245,
+      "learning_rate": 7.827069282765475e-07,
+      "loss": 0.3983,
+      "step": 10092
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.262635263729186,
+      "learning_rate": 7.81995949668562e-07,
+      "loss": 0.9262,
+      "step": 10093
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.319007183839746,
+      "learning_rate": 7.812852667243043e-07,
+      "loss": 0.3667,
+      "step": 10094
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.015797994989106,
+      "learning_rate": 7.805748794935886e-07,
+      "loss": 0.9151,
+      "step": 10095
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4061862206066524,
+      "learning_rate": 7.798647880262111e-07,
+      "loss": 0.1664,
+      "step": 10096
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.410668028791237,
+      "learning_rate": 7.791549923719455e-07,
+      "loss": 0.9573,
+      "step": 10097
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.868404391966383,
+      "learning_rate": 7.784454925805457e-07,
+      "loss": 0.7018,
+      "step": 10098
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.381961548937185,
+      "learning_rate": 7.777362887017448e-07,
+      "loss": 0.2947,
+      "step": 10099
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.892050919340202,
+      "learning_rate": 7.770273807852557e-07,
+      "loss": 1.2152,
+      "step": 10100
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 6.19196362056325,
+      "learning_rate": 7.763187688807677e-07,
+      "loss": 0.6899,
+      "step": 10101
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.0527982191823737,
+      "learning_rate": 7.756104530379526e-07,
+      "loss": 0.3678,
+      "step": 10102
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.1701771145805395,
+      "learning_rate": 7.749024333064614e-07,
+      "loss": 0.7823,
+      "step": 10103
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.286789747108113,
+      "learning_rate": 7.74194709735921e-07,
+      "loss": 0.8221,
+      "step": 10104
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.316492683343482,
+      "learning_rate": 7.73487282375941e-07,
+      "loss": 1.1504,
+      "step": 10105
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.624176181992469,
+      "learning_rate": 7.72780151276108e-07,
+      "loss": 0.6505,
+      "step": 10106
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.622089616941887,
+      "learning_rate": 7.720733164859895e-07,
+      "loss": 0.7511,
+      "step": 10107
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.117173719225514,
+      "learning_rate": 7.713667780551315e-07,
+      "loss": 0.5868,
+      "step": 10108
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.066366457292403,
+      "learning_rate": 7.706605360330594e-07,
+      "loss": 0.4654,
+      "step": 10109
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.733998100948551,
+      "learning_rate": 7.699545904692774e-07,
+      "loss": 0.3261,
+      "step": 10110
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.2497827288864802,
+      "learning_rate": 7.692489414132703e-07,
+      "loss": 0.742,
+      "step": 10111
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.47988675054254,
+      "learning_rate": 7.685435889144993e-07,
+      "loss": 0.6233,
+      "step": 10112
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.8554813264771406,
+      "learning_rate": 7.678385330224075e-07,
+      "loss": 0.4469,
+      "step": 10113
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.4381940487933935,
+      "learning_rate": 7.671337737864159e-07,
+      "loss": 0.5083,
+      "step": 10114
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.6863597272472095,
+      "learning_rate": 7.664293112559251e-07,
+      "loss": 1.0119,
+      "step": 10115
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.4769697594820794,
+      "learning_rate": 7.65725145480315e-07,
+      "loss": 0.4439,
+      "step": 10116
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.370028598981,
+      "learning_rate": 7.65021276508946e-07,
+      "loss": 0.8541,
+      "step": 10117
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.4499207348327525,
+      "learning_rate": 7.643177043911538e-07,
+      "loss": 0.5175,
+      "step": 10118
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.1584533314883565,
+      "learning_rate": 7.636144291762576e-07,
+      "loss": 0.2468,
+      "step": 10119
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.018358478001542,
+      "learning_rate": 7.629114509135521e-07,
+      "loss": 0.9447,
+      "step": 10120
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.5912207702822387,
+      "learning_rate": 7.62208769652314e-07,
+      "loss": 0.5151,
+      "step": 10121
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.328108848632602,
+      "learning_rate": 7.615063854417981e-07,
+      "loss": 0.6842,
+      "step": 10122
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.900438538556681,
+      "learning_rate": 7.608042983312397e-07,
+      "loss": 0.7065,
+      "step": 10123
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.201546029218351,
+      "learning_rate": 7.601025083698499e-07,
+      "loss": 0.4837,
+      "step": 10124
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.026673616477278,
+      "learning_rate": 7.594010156068221e-07,
+      "loss": 0.4566,
+      "step": 10125
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.436617140355298,
+      "learning_rate": 7.586998200913282e-07,
+      "loss": 0.7603,
+      "step": 10126
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.8850862955727252,
+      "learning_rate": 7.579989218725187e-07,
+      "loss": 0.8296,
+      "step": 10127
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 6.665295443382982,
+      "learning_rate": 7.572983209995244e-07,
+      "loss": 0.8821,
+      "step": 10128
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.59546360930877,
+      "learning_rate": 7.565980175214526e-07,
+      "loss": 0.885,
+      "step": 10129
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.8774411324788542,
+      "learning_rate": 7.558980114873921e-07,
+      "loss": 0.726,
+      "step": 10130
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.0323948272572405,
+      "learning_rate": 7.551983029464111e-07,
+      "loss": 0.5955,
+      "step": 10131
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.266205587184443,
+      "learning_rate": 7.544988919475555e-07,
+      "loss": 0.9898,
+      "step": 10132
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.56948708053087,
+      "learning_rate": 7.537997785398515e-07,
+      "loss": 0.7395,
+      "step": 10133
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.4088982946641324,
+      "learning_rate": 7.531009627723035e-07,
+      "loss": 0.6821,
+      "step": 10134
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.0972983217595726,
+      "learning_rate": 7.52402444693896e-07,
+      "loss": 0.6063,
+      "step": 10135
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.8439494774168987,
+      "learning_rate": 7.517042243535899e-07,
+      "loss": 0.719,
+      "step": 10136
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.9165481492503744,
+      "learning_rate": 7.5100630180033e-07,
+      "loss": 0.4903,
+      "step": 10137
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8094712506379884,
+      "learning_rate": 7.503086770830359e-07,
+      "loss": 0.1993,
+      "step": 10138
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.732114676057692,
+      "learning_rate": 7.496113502506092e-07,
+      "loss": 1.0937,
+      "step": 10139
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.573506384797262,
+      "learning_rate": 7.489143213519301e-07,
+      "loss": 1.1496,
+      "step": 10140
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.1810527637860355,
+      "learning_rate": 7.482175904358552e-07,
+      "loss": 0.5457,
+      "step": 10141
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.829992228202246,
+      "learning_rate": 7.475211575512231e-07,
+      "loss": 0.7806,
+      "step": 10142
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.4087777526064107,
+      "learning_rate": 7.468250227468515e-07,
+      "loss": 0.5698,
+      "step": 10143
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.443150332879401,
+      "learning_rate": 7.461291860715359e-07,
+      "loss": 0.9275,
+      "step": 10144
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.8076918964860598,
+      "learning_rate": 7.45433647574052e-07,
+      "loss": 0.7731,
+      "step": 10145
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.00916585804603,
+      "learning_rate": 7.447384073031527e-07,
+      "loss": 0.8045,
+      "step": 10146
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 7.543639619527765,
+      "learning_rate": 7.440434653075723e-07,
+      "loss": 1.4757,
+      "step": 10147
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.744778257036902,
+      "learning_rate": 7.43348821636023e-07,
+      "loss": 1.0878,
+      "step": 10148
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.873126512495574,
+      "learning_rate": 7.426544763371974e-07,
+      "loss": 0.5527,
+      "step": 10149
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.5157252702605923,
+      "learning_rate": 7.419604294597632e-07,
+      "loss": 0.6573,
+      "step": 10150
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.6910821262864113,
+      "learning_rate": 7.412666810523727e-07,
+      "loss": 0.5749,
+      "step": 10151
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.0645375077749435,
+      "learning_rate": 7.405732311636543e-07,
+      "loss": 0.2744,
+      "step": 10152
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 6.118859734677217,
+      "learning_rate": 7.398800798422145e-07,
+      "loss": 0.8487,
+      "step": 10153
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.8014057053335506,
+      "learning_rate": 7.391872271366407e-07,
+      "loss": 0.4396,
+      "step": 10154
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.57035437763408,
+      "learning_rate": 7.384946730954995e-07,
+      "loss": 0.745,
+      "step": 10155
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.6010854374742336,
+      "learning_rate": 7.378024177673354e-07,
+      "loss": 0.6538,
+      "step": 10156
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.783812121647489,
+      "learning_rate": 7.371104612006741e-07,
+      "loss": 0.3029,
+      "step": 10157
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.70317558211454,
+      "learning_rate": 7.364188034440161e-07,
+      "loss": 1.002,
+      "step": 10158
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.9305650101655183,
+      "learning_rate": 7.357274445458446e-07,
+      "loss": 0.7544,
+      "step": 10159
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 6.608343086018153,
+      "learning_rate": 7.35036384554621e-07,
+      "loss": 1.4278,
+      "step": 10160
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.571341322709621,
+      "learning_rate": 7.343456235187857e-07,
+      "loss": 0.6947,
+      "step": 10161
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.4511973270380194,
+      "learning_rate": 7.336551614867582e-07,
+      "loss": 0.6418,
+      "step": 10162
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.478077366704615,
+      "learning_rate": 7.329649985069376e-07,
+      "loss": 1.3947,
+      "step": 10163
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.4354541645576178,
+      "learning_rate": 7.322751346276997e-07,
+      "loss": 0.8503,
+      "step": 10164
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.0040327011560812,
+      "learning_rate": 7.315855698974012e-07,
+      "loss": 0.5111,
+      "step": 10165
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.9865675662372406,
+      "learning_rate": 7.308963043643791e-07,
+      "loss": 0.5029,
+      "step": 10166
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.5068836581670593,
+      "learning_rate": 7.302073380769459e-07,
+      "loss": 0.32,
+      "step": 10167
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.108309155183119,
+      "learning_rate": 7.295186710833973e-07,
+      "loss": 0.6038,
+      "step": 10168
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.1470775770097736,
+      "learning_rate": 7.288303034320038e-07,
+      "loss": 0.5335,
+      "step": 10169
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.930333607704734,
+      "learning_rate": 7.281422351710177e-07,
+      "loss": 0.8595,
+      "step": 10170
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1370544711017538,
+      "learning_rate": 7.274544663486694e-07,
+      "loss": 0.1652,
+      "step": 10171
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.136367665352145,
+      "learning_rate": 7.267669970131691e-07,
+      "loss": 0.8134,
+      "step": 10172
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.141838639356749,
+      "learning_rate": 7.260798272127051e-07,
+      "loss": 1.0518,
+      "step": 10173
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.168489051573621,
+      "learning_rate": 7.25392956995446e-07,
+      "loss": 0.8956,
+      "step": 10174
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.033656979157179,
+      "learning_rate": 7.247063864095361e-07,
+      "loss": 0.8774,
+      "step": 10175
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.684176619776644,
+      "learning_rate": 7.240201155031029e-07,
+      "loss": 0.9595,
+      "step": 10176
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.3587537054595167,
+      "learning_rate": 7.233341443242504e-07,
+      "loss": 0.4504,
+      "step": 10177
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.72953453724526,
+      "learning_rate": 7.22648472921062e-07,
+      "loss": 0.8717,
+      "step": 10178
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.2640117138401425,
+      "learning_rate": 7.219631013416007e-07,
+      "loss": 0.8372,
+      "step": 10179
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.031931141346207,
+      "learning_rate": 7.212780296339095e-07,
+      "loss": 0.7958,
+      "step": 10180
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.3418842891165044,
+      "learning_rate": 7.205932578460056e-07,
+      "loss": 0.5695,
+      "step": 10181
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.416966419131427,
+      "learning_rate": 7.199087860258913e-07,
+      "loss": 0.6697,
+      "step": 10182
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.022172971765802,
+      "learning_rate": 7.192246142215437e-07,
+      "loss": 0.4962,
+      "step": 10183
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.3941894220163036,
+      "learning_rate": 7.185407424809199e-07,
+      "loss": 1.0106,
+      "step": 10184
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.8137243834314782,
+      "learning_rate": 7.178571708519589e-07,
+      "loss": 0.5257,
+      "step": 10185
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.4841375497339975,
+      "learning_rate": 7.171738993825728e-07,
+      "loss": 0.9764,
+      "step": 10186
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.495239260489433,
+      "learning_rate": 7.164909281206573e-07,
+      "loss": 0.2904,
+      "step": 10187
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.932504966643108,
+      "learning_rate": 7.158082571140857e-07,
+      "loss": 0.5881,
+      "step": 10188
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.309706622233228,
+      "learning_rate": 7.151258864107107e-07,
+      "loss": 0.8552,
+      "step": 10189
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.151604904407833,
+      "learning_rate": 7.144438160583633e-07,
+      "loss": 1.3396,
+      "step": 10190
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.175446160551388,
+      "learning_rate": 7.137620461048544e-07,
+      "loss": 0.3172,
+      "step": 10191
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.0943987081345203,
+      "learning_rate": 7.130805765979714e-07,
+      "loss": 0.5698,
+      "step": 10192
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.836987678336197,
+      "learning_rate": 7.123994075854834e-07,
+      "loss": 0.5729,
+      "step": 10193
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.072622197125855,
+      "learning_rate": 7.117185391151371e-07,
+      "loss": 0.3823,
+      "step": 10194
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.870931924754225,
+      "learning_rate": 7.110379712346582e-07,
+      "loss": 0.9687,
+      "step": 10195
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.549694135245669,
+      "learning_rate": 7.103577039917536e-07,
+      "loss": 1.1397,
+      "step": 10196
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.443048944288806,
+      "learning_rate": 7.09677737434104e-07,
+      "loss": 0.9514,
+      "step": 10197
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 6.412042572557696,
+      "learning_rate": 7.089980716093741e-07,
+      "loss": 1.004,
+      "step": 10198
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.8298060473757447,
+      "learning_rate": 7.083187065652042e-07,
+      "loss": 0.492,
+      "step": 10199
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.462097342915266,
+      "learning_rate": 7.076396423492154e-07,
+      "loss": 0.5119,
+      "step": 10200
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.4503156311001444,
+      "learning_rate": 7.069608790090077e-07,
+      "loss": 0.4069,
+      "step": 10201
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.514757874641178,
+      "learning_rate": 7.062824165921589e-07,
+      "loss": 0.2277,
+      "step": 10202
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.0935899306461865,
+      "learning_rate": 7.056042551462273e-07,
+      "loss": 0.9142,
+      "step": 10203
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.408676529971462,
+      "learning_rate": 7.049263947187468e-07,
+      "loss": 0.7984,
+      "step": 10204
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.7888943397001684,
+      "learning_rate": 7.042488353572341e-07,
+      "loss": 0.8042,
+      "step": 10205
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.6447457960019145,
+      "learning_rate": 7.035715771091828e-07,
+      "loss": 0.7367,
+      "step": 10206
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.7901059754746016,
+      "learning_rate": 7.028946200220655e-07,
+      "loss": 0.4016,
+      "step": 10207
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.768076715298043,
+      "learning_rate": 7.022179641433357e-07,
+      "loss": 0.3792,
+      "step": 10208
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.9790775227827817,
+      "learning_rate": 7.015416095204214e-07,
+      "loss": 0.2407,
+      "step": 10209
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.779174705300358,
+      "learning_rate": 7.008655562007333e-07,
+      "loss": 0.9499,
+      "step": 10210
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.504210062203104,
+      "learning_rate": 7.001898042316602e-07,
+      "loss": 0.5965,
+      "step": 10211
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.098430800876282,
+      "learning_rate": 6.995143536605698e-07,
+      "loss": 0.8216,
+      "step": 10212
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.722458855400462,
+      "learning_rate": 6.988392045348063e-07,
+      "loss": 0.5898,
+      "step": 10213
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.603204189743334,
+      "learning_rate": 6.981643569016966e-07,
+      "loss": 0.9451,
+      "step": 10214
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.383153672975866,
+      "learning_rate": 6.974898108085431e-07,
+      "loss": 0.7547,
+      "step": 10215
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.777138167338287,
+      "learning_rate": 6.968155663026294e-07,
+      "loss": 1.0101,
+      "step": 10216
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.949733626148483,
+      "learning_rate": 6.961416234312168e-07,
+      "loss": 0.5266,
+      "step": 10217
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.286890743214652,
+      "learning_rate": 6.954679822415461e-07,
+      "loss": 0.624,
+      "step": 10218
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.196197776760363,
+      "learning_rate": 6.947946427808366e-07,
+      "loss": 1.1136,
+      "step": 10219
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.8562787817039075,
+      "learning_rate": 6.941216050962868e-07,
+      "loss": 0.8969,
+      "step": 10220
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.311974724124445,
+      "learning_rate": 6.934488692350727e-07,
+      "loss": 0.8868,
+      "step": 10221
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.961188506136366,
+      "learning_rate": 6.927764352443506e-07,
+      "loss": 0.8205,
+      "step": 10222
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.955038469711517,
+      "learning_rate": 6.921043031712549e-07,
+      "loss": 0.8039,
+      "step": 10223
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.323796643908006,
+      "learning_rate": 6.914324730629002e-07,
+      "loss": 1.027,
+      "step": 10224
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.0023604844410343,
+      "learning_rate": 6.907609449663788e-07,
+      "loss": 0.49,
+      "step": 10225
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.357653845060114,
+      "learning_rate": 6.900897189287603e-07,
+      "loss": 0.5876,
+      "step": 10226
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.68285164443623,
+      "learning_rate": 6.894187949970954e-07,
+      "loss": 0.3064,
+      "step": 10227
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.3820541379404068,
+      "learning_rate": 6.887481732184148e-07,
+      "loss": 0.6108,
+      "step": 10228
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 6.299569677418883,
+      "learning_rate": 6.880778536397237e-07,
+      "loss": 0.8979,
+      "step": 10229
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.970028067271835,
+      "learning_rate": 6.874078363080089e-07,
+      "loss": 0.6938,
+      "step": 10230
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.031576213495916,
+      "learning_rate": 6.867381212702378e-07,
+      "loss": 0.5359,
+      "step": 10231
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.325165945870046,
+      "learning_rate": 6.860687085733519e-07,
+      "loss": 0.1823,
+      "step": 10232
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.633013472211267,
+      "learning_rate": 6.853995982642753e-07,
+      "loss": 0.681,
+      "step": 10233
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.529771631025134,
+      "learning_rate": 6.847307903899091e-07,
+      "loss": 0.7647,
+      "step": 10234
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.125269134331355,
+      "learning_rate": 6.840622849971352e-07,
+      "loss": 0.451,
+      "step": 10235
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1441559641432029,
+      "learning_rate": 6.833940821328117e-07,
+      "loss": 0.1391,
+      "step": 10236
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.361524339778844,
+      "learning_rate": 6.827261818437781e-07,
+      "loss": 0.7674,
+      "step": 10237
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.0626961124778864,
+      "learning_rate": 6.820585841768496e-07,
+      "loss": 0.5342,
+      "step": 10238
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.582327339796292,
+      "learning_rate": 6.813912891788221e-07,
+      "loss": 1.0739,
+      "step": 10239
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.562027936198054,
+      "learning_rate": 6.807242968964711e-07,
+      "loss": 1.1472,
+      "step": 10240
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.9999078913684025,
+      "learning_rate": 6.800576073765486e-07,
+      "loss": 0.7174,
+      "step": 10241
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.317538348890353,
+      "learning_rate": 6.793912206657893e-07,
+      "loss": 0.4795,
+      "step": 10242
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.5309894163865896,
+      "learning_rate": 6.787251368109005e-07,
+      "loss": 0.5938,
+      "step": 10243
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.284058046073281,
+      "learning_rate": 6.780593558585746e-07,
+      "loss": 0.9317,
+      "step": 10244
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.844064243994371,
+      "learning_rate": 6.773938778554773e-07,
+      "loss": 0.8454,
+      "step": 10245
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.5893596866876116,
+      "learning_rate": 6.767287028482577e-07,
+      "loss": 0.5012,
+      "step": 10246
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.3520825892480266,
+      "learning_rate": 6.760638308835404e-07,
+      "loss": 0.4746,
+      "step": 10247
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.393175137395471,
+      "learning_rate": 6.753992620079325e-07,
+      "loss": 0.4163,
+      "step": 10248
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.4533606783094313,
+      "learning_rate": 6.747349962680144e-07,
+      "loss": 0.5819,
+      "step": 10249
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.7237269976253433,
+      "learning_rate": 6.74071033710349e-07,
+      "loss": 0.4897,
+      "step": 10250
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.746376562234642,
+      "learning_rate": 6.734073743814779e-07,
+      "loss": 1.1306,
+      "step": 10251
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.358985476907801,
+      "learning_rate": 6.727440183279205e-07,
+      "loss": 1.1153,
+      "step": 10252
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.0330042057965145,
+      "learning_rate": 6.720809655961752e-07,
+      "loss": 0.2759,
+      "step": 10253
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.387208326336032,
+      "learning_rate": 6.714182162327198e-07,
+      "loss": 0.7345,
+      "step": 10254
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.9235955723959077,
+      "learning_rate": 6.707557702840084e-07,
+      "loss": 0.5651,
+      "step": 10255
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.368011627009563,
+      "learning_rate": 6.700936277964771e-07,
+      "loss": 1.0694,
+      "step": 10256
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.6604373487739097,
+      "learning_rate": 6.694317888165381e-07,
+      "loss": 0.3197,
+      "step": 10257
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.394383222912035,
+      "learning_rate": 6.687702533905855e-07,
+      "loss": 0.7531,
+      "step": 10258
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.02730281209635,
+      "learning_rate": 6.681090215649872e-07,
+      "loss": 0.3874,
+      "step": 10259
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.4602941295006677,
+      "learning_rate": 6.674480933860938e-07,
+      "loss": 0.7086,
+      "step": 10260
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.408342882524193,
+      "learning_rate": 6.667874689002352e-07,
+      "loss": 0.5866,
+      "step": 10261
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.4173230715632665,
+      "learning_rate": 6.661271481537157e-07,
+      "loss": 0.5306,
+      "step": 10262
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.969856473735861,
+      "learning_rate": 6.654671311928218e-07,
+      "loss": 0.9168,
+      "step": 10263
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.6517478940535586,
+      "learning_rate": 6.648074180638181e-07,
+      "loss": 0.501,
+      "step": 10264
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.653153651391398,
+      "learning_rate": 6.641480088129488e-07,
+      "loss": 0.8284,
+      "step": 10265
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.850258416331185,
+      "learning_rate": 6.634889034864334e-07,
+      "loss": 0.8756,
+      "step": 10266
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.985382741642617,
+      "learning_rate": 6.628301021304734e-07,
+      "loss": 0.3819,
+      "step": 10267
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.380448581255921,
+      "learning_rate": 6.621716047912475e-07,
+      "loss": 0.6911,
+      "step": 10268
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.063915509321915,
+      "learning_rate": 6.615134115149135e-07,
+      "loss": 0.839,
+      "step": 10269
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.769133946396099,
+      "learning_rate": 6.608555223476088e-07,
+      "loss": 0.7109,
+      "step": 10270
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.433360575241961,
+      "learning_rate": 6.60197937335449e-07,
+      "loss": 0.3123,
+      "step": 10271
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.813186992499145,
+      "learning_rate": 6.595406565245255e-07,
+      "loss": 0.9826,
+      "step": 10272
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.6466358439004374,
+      "learning_rate": 6.588836799609128e-07,
+      "loss": 0.6288,
+      "step": 10273
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.693817862440512,
+      "learning_rate": 6.582270076906611e-07,
+      "loss": 0.5749,
+      "step": 10274
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 6.070760684402229,
+      "learning_rate": 6.57570639759802e-07,
+      "loss": 1.4687,
+      "step": 10275
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.607951170928873,
+      "learning_rate": 6.569145762143414e-07,
+      "loss": 0.7955,
+      "step": 10276
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.837526405375264,
+      "learning_rate": 6.562588171002688e-07,
+      "loss": 0.8754,
+      "step": 10277
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.214752328843424,
+      "learning_rate": 6.556033624635482e-07,
+      "loss": 0.9623,
+      "step": 10278
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.033828828863102,
+      "learning_rate": 6.549482123501249e-07,
+      "loss": 0.3344,
+      "step": 10279
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3399389391733743,
+      "learning_rate": 6.542933668059226e-07,
+      "loss": 0.1505,
+      "step": 10280
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.133479400402992,
+      "learning_rate": 6.536388258768423e-07,
+      "loss": 0.9193,
+      "step": 10281
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5262125922080394,
+      "learning_rate": 6.529845896087649e-07,
+      "loss": 0.1784,
+      "step": 10282
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.4484274448162457,
+      "learning_rate": 6.523306580475508e-07,
+      "loss": 0.7567,
+      "step": 10283
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.2346819404260065,
+      "learning_rate": 6.516770312390353e-07,
+      "loss": 1.1449,
+      "step": 10284
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.732851697170656,
+      "learning_rate": 6.510237092290361e-07,
+      "loss": 0.6355,
+      "step": 10285
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 6.066102838902472,
+      "learning_rate": 6.503706920633473e-07,
+      "loss": 1.278,
+      "step": 10286
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.29385326911628,
+      "learning_rate": 6.49717979787744e-07,
+      "loss": 1.0592,
+      "step": 10287
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.741642006414545,
+      "learning_rate": 6.490655724479789e-07,
+      "loss": 0.4784,
+      "step": 10288
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.301411693753504,
+      "learning_rate": 6.48413470089781e-07,
+      "loss": 0.4115,
+      "step": 10289
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.9701793371450704,
+      "learning_rate": 6.477616727588604e-07,
+      "loss": 0.593,
+      "step": 10290
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.334020513818201,
+      "learning_rate": 6.471101805009062e-07,
+      "loss": 0.3358,
+      "step": 10291
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.042172667693351,
+      "learning_rate": 6.464589933615839e-07,
+      "loss": 0.5782,
+      "step": 10292
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.898454941842396,
+      "learning_rate": 6.458081113865395e-07,
+      "loss": 0.98,
+      "step": 10293
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.301511079307743,
+      "learning_rate": 6.451575346213979e-07,
+      "loss": 0.6028,
+      "step": 10294
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.127705479254126,
+      "learning_rate": 6.4450726311176e-07,
+      "loss": 0.2934,
+      "step": 10295
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.6716391467874003,
+      "learning_rate": 6.438572969032075e-07,
+      "loss": 0.495,
+      "step": 10296
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.8791457123016366,
+      "learning_rate": 6.432076360413003e-07,
+      "loss": 0.5122,
+      "step": 10297
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.1169069139262344,
+      "learning_rate": 6.425582805715775e-07,
+      "loss": 0.5587,
+      "step": 10298
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.140954051481482,
+      "learning_rate": 6.419092305395552e-07,
+      "loss": 0.3388,
+      "step": 10299
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.4818241671767765,
+      "learning_rate": 6.412604859907306e-07,
+      "loss": 0.46,
+      "step": 10300
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.585994386964461,
+      "learning_rate": 6.406120469705757e-07,
+      "loss": 0.6881,
+      "step": 10301
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.3936003516441566,
+      "learning_rate": 6.399639135245439e-07,
+      "loss": 0.6548,
+      "step": 10302
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.076143681485743,
+      "learning_rate": 6.393160856980668e-07,
+      "loss": 0.6639,
+      "step": 10303
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.277110114326273,
+      "learning_rate": 6.386685635365542e-07,
+      "loss": 0.8738,
+      "step": 10304
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.097791072129815,
+      "learning_rate": 6.380213470853963e-07,
+      "loss": 0.7573,
+      "step": 10305
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.6666051053495963,
+      "learning_rate": 6.37374436389957e-07,
+      "loss": 0.6457,
+      "step": 10306
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 6.189374771386946,
+      "learning_rate": 6.367278314955849e-07,
+      "loss": 1.1032,
+      "step": 10307
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.2975087719293557,
+      "learning_rate": 6.360815324476016e-07,
+      "loss": 0.5535,
+      "step": 10308
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.001540898596129,
+      "learning_rate": 6.354355392913114e-07,
+      "loss": 0.7347,
+      "step": 10309
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.903098744714241,
+      "learning_rate": 6.347898520719947e-07,
+      "loss": 0.7331,
+      "step": 10310
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.7852382356129795,
+      "learning_rate": 6.341444708349131e-07,
+      "loss": 0.9693,
+      "step": 10311
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.788993743466064,
+      "learning_rate": 6.334993956253033e-07,
+      "loss": 0.5701,
+      "step": 10312
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.4718749584589923,
+      "learning_rate": 6.328546264883822e-07,
+      "loss": 0.5135,
+      "step": 10313
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.1507870284409103,
+      "learning_rate": 6.322101634693461e-07,
+      "loss": 0.5019,
+      "step": 10314
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.6874629571572686,
+      "learning_rate": 6.315660066133689e-07,
+      "loss": 0.6376,
+      "step": 10315
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.5487397987975715,
+      "learning_rate": 6.309221559656026e-07,
+      "loss": 0.599,
+      "step": 10316
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.856529420525376,
+      "learning_rate": 6.302786115711806e-07,
+      "loss": 0.6894,
+      "step": 10317
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.266462849628899,
+      "learning_rate": 6.296353734752098e-07,
+      "loss": 0.9058,
+      "step": 10318
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 6.010416616131682,
+      "learning_rate": 6.289924417227789e-07,
+      "loss": 1.1745,
+      "step": 10319
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1058518824775168,
+      "learning_rate": 6.283498163589558e-07,
+      "loss": 0.1765,
+      "step": 10320
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.802275931450062,
+      "learning_rate": 6.277074974287856e-07,
+      "loss": 1.0474,
+      "step": 10321
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.797194600107948,
+      "learning_rate": 6.270654849772906e-07,
+      "loss": 0.8021,
+      "step": 10322
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.4246262562427128,
+      "learning_rate": 6.264237790494754e-07,
+      "loss": 0.6082,
+      "step": 10323
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.4649718930823123,
+      "learning_rate": 6.257823796903178e-07,
+      "loss": 0.6519,
+      "step": 10324
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.05193623676796,
+      "learning_rate": 6.251412869447793e-07,
+      "loss": 0.3848,
+      "step": 10325
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.309915119979786,
+      "learning_rate": 6.24500500857797e-07,
+      "loss": 0.8005,
+      "step": 10326
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.254255006430809,
+      "learning_rate": 6.238600214742868e-07,
+      "loss": 0.58,
+      "step": 10327
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.961040158682524,
+      "learning_rate": 6.232198488391461e-07,
+      "loss": 0.6447,
+      "step": 10328
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 6.040144358348651,
+      "learning_rate": 6.225799829972445e-07,
+      "loss": 1.0723,
+      "step": 10329
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.395942298388709,
+      "learning_rate": 6.219404239934357e-07,
+      "loss": 0.8869,
+      "step": 10330
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.4425011509638686,
+      "learning_rate": 6.213011718725493e-07,
+      "loss": 0.3754,
+      "step": 10331
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.0282056723859623,
+      "learning_rate": 6.206622266793949e-07,
+      "loss": 0.2786,
+      "step": 10332
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.798718887193467,
+      "learning_rate": 6.200235884587596e-07,
+      "loss": 0.5698,
+      "step": 10333
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.575555834030352,
+      "learning_rate": 6.193852572554104e-07,
+      "loss": 0.8362,
+      "step": 10334
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.5107693830811364,
+      "learning_rate": 6.187472331140887e-07,
+      "loss": 0.6514,
+      "step": 10335
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.401100975202584,
+      "learning_rate": 6.181095160795187e-07,
+      "loss": 0.8509,
+      "step": 10336
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.982092836864028,
+      "learning_rate": 6.174721061964029e-07,
+      "loss": 0.6061,
+      "step": 10337
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.069443870464548,
+      "learning_rate": 6.168350035094178e-07,
+      "loss": 0.6044,
+      "step": 10338
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.6613953155838637,
+      "learning_rate": 6.161982080632239e-07,
+      "loss": 0.7821,
+      "step": 10339
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.951373078869165,
+      "learning_rate": 6.155617199024588e-07,
+      "loss": 0.5071,
+      "step": 10340
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.23635692772508,
+      "learning_rate": 6.14925539071734e-07,
+      "loss": 0.7125,
+      "step": 10341
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2432538500796653,
+      "learning_rate": 6.142896656156455e-07,
+      "loss": 0.1723,
+      "step": 10342
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.1860302085802137,
+      "learning_rate": 6.136540995787649e-07,
+      "loss": 0.3348,
+      "step": 10343
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.883965192436654,
+      "learning_rate": 6.13018841005642e-07,
+      "loss": 0.8817,
+      "step": 10344
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.2281268341711797,
+      "learning_rate": 6.12383889940808e-07,
+      "loss": 0.566,
+      "step": 10345
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.594241528342523,
+      "learning_rate": 6.117492464287666e-07,
+      "loss": 0.9008,
+      "step": 10346
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.73225891709093,
+      "learning_rate": 6.111149105140052e-07,
+      "loss": 0.7543,
+      "step": 10347
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.8994344357396984,
+      "learning_rate": 6.104808822409885e-07,
+      "loss": 0.6075,
+      "step": 10348
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.8249853665661506,
+      "learning_rate": 6.098471616541585e-07,
+      "loss": 1.0379,
+      "step": 10349
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.6325995942495455,
+      "learning_rate": 6.092137487979366e-07,
+      "loss": 0.7691,
+      "step": 10350
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.325602957824057,
+      "learning_rate": 6.08580643716723e-07,
+      "loss": 1.0923,
+      "step": 10351
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.8762181860203797,
+      "learning_rate": 6.079478464548938e-07,
+      "loss": 0.8949,
+      "step": 10352
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.082466940308302,
+      "learning_rate": 6.073153570568074e-07,
+      "loss": 1.0848,
+      "step": 10353
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.0532638854790966,
+      "learning_rate": 6.066831755667962e-07,
+      "loss": 1.1865,
+      "step": 10354
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.5533008575699334,
+      "learning_rate": 6.060513020291753e-07,
+      "loss": 0.3113,
+      "step": 10355
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.5437168783894606,
+      "learning_rate": 6.054197364882347e-07,
+      "loss": 0.6903,
+      "step": 10356
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.869798916119071,
+      "learning_rate": 6.047884789882469e-07,
+      "loss": 0.653,
+      "step": 10357
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.146009182397108,
+      "learning_rate": 6.041575295734576e-07,
+      "loss": 1.0095,
+      "step": 10358
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.8993766307489413,
+      "learning_rate": 6.035268882880941e-07,
+      "loss": 0.4933,
+      "step": 10359
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.448991672717264,
+      "learning_rate": 6.028965551763627e-07,
+      "loss": 1.195,
+      "step": 10360
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.678067743838973,
+      "learning_rate": 6.022665302824465e-07,
+      "loss": 0.8999,
+      "step": 10361
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.6260758312453385,
+      "learning_rate": 6.016368136505074e-07,
+      "loss": 0.7788,
+      "step": 10362
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 6.046366637884028,
+      "learning_rate": 6.010074053246872e-07,
+      "loss": 0.8591,
+      "step": 10363
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.3503412128880736,
+      "learning_rate": 6.003783053491025e-07,
+      "loss": 0.5532,
+      "step": 10364
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.203630680515273,
+      "learning_rate": 5.99749513767851e-07,
+      "loss": 0.581,
+      "step": 10365
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.9230485157594908,
+      "learning_rate": 5.99121030625009e-07,
+      "loss": 0.4968,
+      "step": 10366
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.0128915740778766,
+      "learning_rate": 5.984928559646297e-07,
+      "loss": 0.8049,
+      "step": 10367
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.5001324671821212,
+      "learning_rate": 5.97864989830747e-07,
+      "loss": 0.4365,
+      "step": 10368
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.697993203346618,
+      "learning_rate": 5.97237432267369e-07,
+      "loss": 0.9607,
+      "step": 10369
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.456607685228635,
+      "learning_rate": 5.966101833184873e-07,
+      "loss": 0.4056,
+      "step": 10370
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.821843099442181,
+      "learning_rate": 5.959832430280677e-07,
+      "loss": 0.6264,
+      "step": 10371
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.903722547084163,
+      "learning_rate": 5.953566114400555e-07,
+      "loss": 0.65,
+      "step": 10372
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.238758982285471,
+      "learning_rate": 5.947302885983763e-07,
+      "loss": 0.8129,
+      "step": 10373
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.1879731369082,
+      "learning_rate": 5.941042745469333e-07,
+      "loss": 0.9396,
+      "step": 10374
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.6125945881433394,
+      "learning_rate": 5.934785693296046e-07,
+      "loss": 0.6156,
+      "step": 10375
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.0299076473916093,
+      "learning_rate": 5.928531729902509e-07,
+      "loss": 0.638,
+      "step": 10376
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.2482608712189336,
+      "learning_rate": 5.922280855727103e-07,
+      "loss": 0.4839,
+      "step": 10377
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.30757637027783,
+      "learning_rate": 5.916033071207977e-07,
+      "loss": 1.0658,
+      "step": 10378
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.601260972257547,
+      "learning_rate": 5.909788376783083e-07,
+      "loss": 0.7538,
+      "step": 10379
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.367104308567999,
+      "learning_rate": 5.903546772890151e-07,
+      "loss": 0.7123,
+      "step": 10380
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.307815347324862,
+      "learning_rate": 5.897308259966672e-07,
+      "loss": 0.7737,
+      "step": 10381
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.688619950415655,
+      "learning_rate": 5.891072838449946e-07,
+      "loss": 0.9054,
+      "step": 10382
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.309335159982341,
+      "learning_rate": 5.884840508777056e-07,
+      "loss": 0.6679,
+      "step": 10383
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 6.2618730791039985,
+      "learning_rate": 5.878611271384865e-07,
+      "loss": 1.2592,
+      "step": 10384
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.294292166667183,
+      "learning_rate": 5.872385126709995e-07,
+      "loss": 0.1747,
+      "step": 10385
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.068858159170679,
+      "learning_rate": 5.866162075188892e-07,
+      "loss": 0.5161,
+      "step": 10386
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.841731758237636,
+      "learning_rate": 5.859942117257749e-07,
+      "loss": 0.621,
+      "step": 10387
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.14131196717673,
+      "learning_rate": 5.853725253352566e-07,
+      "loss": 0.6049,
+      "step": 10388
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.604489147954973,
+      "learning_rate": 5.847511483909119e-07,
+      "loss": 0.5231,
+      "step": 10389
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.019636781178539,
+      "learning_rate": 5.841300809362959e-07,
+      "loss": 0.1193,
+      "step": 10390
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.7573726079017606,
+      "learning_rate": 5.835093230149447e-07,
+      "loss": 0.6389,
+      "step": 10391
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.6155034308134195,
+      "learning_rate": 5.828888746703687e-07,
+      "loss": 0.9993,
+      "step": 10392
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.5814248275023153,
+      "learning_rate": 5.822687359460588e-07,
+      "loss": 0.352,
+      "step": 10393
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.959016551167589,
+      "learning_rate": 5.816489068854841e-07,
+      "loss": 0.5663,
+      "step": 10394
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.4804949765515043,
+      "learning_rate": 5.810293875320927e-07,
+      "loss": 0.4629,
+      "step": 10395
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.340824583344594,
+      "learning_rate": 5.804101779293098e-07,
+      "loss": 0.9673,
+      "step": 10396
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.491235360132722,
+      "learning_rate": 5.797912781205406e-07,
+      "loss": 0.5077,
+      "step": 10397
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.178766204446659,
+      "learning_rate": 5.791726881491644e-07,
+      "loss": 0.6482,
+      "step": 10398
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.3425427217509744,
+      "learning_rate": 5.785544080585437e-07,
+      "loss": 0.5488,
+      "step": 10399
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.871187595458767,
+      "learning_rate": 5.779364378920177e-07,
+      "loss": 0.8559,
+      "step": 10400
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.938369484063271,
+      "learning_rate": 5.773187776929017e-07,
+      "loss": 0.9178,
+      "step": 10401
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.239734217262983,
+      "learning_rate": 5.767014275044914e-07,
+      "loss": 0.6713,
+      "step": 10402
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.226986391688428,
+      "learning_rate": 5.760843873700622e-07,
+      "loss": 0.7308,
+      "step": 10403
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.060829800690869,
+      "learning_rate": 5.754676573328632e-07,
+      "loss": 0.5538,
+      "step": 10404
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.299010938722108,
+      "learning_rate": 5.74851237436126e-07,
+      "loss": 0.3488,
+      "step": 10405
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.634979349676279,
+      "learning_rate": 5.742351277230584e-07,
+      "loss": 0.6518,
+      "step": 10406
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.758525735469257,
+      "learning_rate": 5.736193282368474e-07,
+      "loss": 1.0361,
+      "step": 10407
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.071298839841249,
+      "learning_rate": 5.730038390206594e-07,
+      "loss": 1.0557,
+      "step": 10408
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.511955381102928,
+      "learning_rate": 5.723886601176343e-07,
+      "loss": 0.8135,
+      "step": 10409
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.4612770054887276,
+      "learning_rate": 5.717737915708954e-07,
+      "loss": 0.6504,
+      "step": 10410
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.622655119430177,
+      "learning_rate": 5.711592334235416e-07,
+      "loss": 0.6042,
+      "step": 10411
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.370232020220817,
+      "learning_rate": 5.705449857186518e-07,
+      "loss": 0.4626,
+      "step": 10412
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.4557763354897095,
+      "learning_rate": 5.699310484992809e-07,
+      "loss": 0.679,
+      "step": 10413
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.834903902822914,
+      "learning_rate": 5.693174218084652e-07,
+      "loss": 1.3057,
+      "step": 10414
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.8354419775478115,
+      "learning_rate": 5.687041056892145e-07,
+      "loss": 0.6512,
+      "step": 10415
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.903765763800176,
+      "learning_rate": 5.680911001845218e-07,
+      "loss": 0.6977,
+      "step": 10416
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.8246310995433146,
+      "learning_rate": 5.674784053373545e-07,
+      "loss": 0.7733,
+      "step": 10417
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.159198786553831,
+      "learning_rate": 5.668660211906607e-07,
+      "loss": 0.939,
+      "step": 10418
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.1686359363891405,
+      "learning_rate": 5.662539477873657e-07,
+      "loss": 0.5318,
+      "step": 10419
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.145125587558952,
+      "learning_rate": 5.656421851703742e-07,
+      "loss": 0.8105,
+      "step": 10420
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.849234211932994,
+      "learning_rate": 5.650307333825661e-07,
+      "loss": 0.3998,
+      "step": 10421
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.3382026475057884,
+      "learning_rate": 5.644195924668028e-07,
+      "loss": 0.4322,
+      "step": 10422
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.753214314856311,
+      "learning_rate": 5.638087624659216e-07,
+      "loss": 0.3664,
+      "step": 10423
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.688637428225327,
+      "learning_rate": 5.631982434227406e-07,
+      "loss": 0.5746,
+      "step": 10424
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.493311422581332,
+      "learning_rate": 5.625880353800545e-07,
+      "loss": 0.6409,
+      "step": 10425
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.5404767230707925,
+      "learning_rate": 5.619781383806345e-07,
+      "loss": 0.7954,
+      "step": 10426
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.6385739608008625,
+      "learning_rate": 5.613685524672318e-07,
+      "loss": 0.7876,
+      "step": 10427
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.242105143127419,
+      "learning_rate": 5.607592776825777e-07,
+      "loss": 0.7932,
+      "step": 10428
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.131498899837967,
+      "learning_rate": 5.601503140693782e-07,
+      "loss": 0.6301,
+      "step": 10429
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.150034675458454,
+      "learning_rate": 5.595416616703203e-07,
+      "loss": 0.688,
+      "step": 10430
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.3659764124021025,
+      "learning_rate": 5.589333205280662e-07,
+      "loss": 0.8211,
+      "step": 10431
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.9082421698650633,
+      "learning_rate": 5.583252906852594e-07,
+      "loss": 0.9047,
+      "step": 10432
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.194612789126027,
+      "learning_rate": 5.577175721845185e-07,
+      "loss": 0.5861,
+      "step": 10433
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.315130226747433,
+      "learning_rate": 5.571101650684435e-07,
+      "loss": 1.2291,
+      "step": 10434
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.723126091057395,
+      "learning_rate": 5.565030693796098e-07,
+      "loss": 0.5973,
+      "step": 10435
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.598692276891783,
+      "learning_rate": 5.558962851605731e-07,
+      "loss": 0.7576,
+      "step": 10436
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.2228214330867964,
+      "learning_rate": 5.552898124538669e-07,
+      "loss": 0.5391,
+      "step": 10437
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.220146649411876,
+      "learning_rate": 5.546836513020004e-07,
+      "loss": 0.6394,
+      "step": 10438
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.742810055429589,
+      "learning_rate": 5.540778017474635e-07,
+      "loss": 0.2506,
+      "step": 10439
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.035624938028897,
+      "learning_rate": 5.534722638327245e-07,
+      "loss": 0.652,
+      "step": 10440
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.690919519405272,
+      "learning_rate": 5.528670376002282e-07,
+      "loss": 0.5183,
+      "step": 10441
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.442735923919764,
+      "learning_rate": 5.522621230923986e-07,
+      "loss": 0.6551,
+      "step": 10442
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.765068024938463,
+      "learning_rate": 5.516575203516389e-07,
+      "loss": 0.839,
+      "step": 10443
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.2785990526068276,
+      "learning_rate": 5.510532294203264e-07,
+      "loss": 0.5858,
+      "step": 10444
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.778841998641732,
+      "learning_rate": 5.504492503408204e-07,
+      "loss": 0.8196,
+      "step": 10445
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.085316672922174,
+      "learning_rate": 5.498455831554589e-07,
+      "loss": 1.2685,
+      "step": 10446
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.967054918121122,
+      "learning_rate": 5.492422279065535e-07,
+      "loss": 0.8716,
+      "step": 10447
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.017089926733228,
+      "learning_rate": 5.486391846363998e-07,
+      "loss": 0.6948,
+      "step": 10448
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.960935176016891,
+      "learning_rate": 5.48036453387265e-07,
+      "loss": 0.5706,
+      "step": 10449
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.29836270974255,
+      "learning_rate": 5.474340342014007e-07,
+      "loss": 0.6272,
+      "step": 10450
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.618055713945523,
+      "learning_rate": 5.468319271210326e-07,
+      "loss": 0.743,
+      "step": 10451
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.777227654941848,
+      "learning_rate": 5.462301321883661e-07,
+      "loss": 0.4506,
+      "step": 10452
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.755869482729697,
+      "learning_rate": 5.456286494455843e-07,
+      "loss": 0.6088,
+      "step": 10453
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.6009071215185906,
+      "learning_rate": 5.450274789348497e-07,
+      "loss": 0.9747,
+      "step": 10454
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.5863105045010215,
+      "learning_rate": 5.444266206983001e-07,
+      "loss": 1.3258,
+      "step": 10455
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.419252475761903,
+      "learning_rate": 5.438260747780532e-07,
+      "loss": 0.7476,
+      "step": 10456
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.419854446947665,
+      "learning_rate": 5.432258412162056e-07,
+      "loss": 1.1689,
+      "step": 10457
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.0074298968463062,
+      "learning_rate": 5.4262592005483e-07,
+      "loss": 0.3632,
+      "step": 10458
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.225202148603655,
+      "learning_rate": 5.420263113359791e-07,
+      "loss": 0.6382,
+      "step": 10459
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.111944379787736,
+      "learning_rate": 5.414270151016843e-07,
+      "loss": 0.5012,
+      "step": 10460
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.6107706317948147,
+      "learning_rate": 5.408280313939502e-07,
+      "loss": 0.4584,
+      "step": 10461
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9131447266796116,
+      "learning_rate": 5.402293602547659e-07,
+      "loss": 0.3841,
+      "step": 10462
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.9771864336352314,
+      "learning_rate": 5.396310017260931e-07,
+      "loss": 0.8553,
+      "step": 10463
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.495631673597813,
+      "learning_rate": 5.390329558498759e-07,
+      "loss": 0.9242,
+      "step": 10464
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.0870467055378406,
+      "learning_rate": 5.384352226680356e-07,
+      "loss": 1.0474,
+      "step": 10465
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.110527224780676,
+      "learning_rate": 5.378378022224679e-07,
+      "loss": 0.8318,
+      "step": 10466
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7325885690721872,
+      "learning_rate": 5.372406945550507e-07,
+      "loss": 0.1975,
+      "step": 10467
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.80887620719976,
+      "learning_rate": 5.366438997076396e-07,
+      "loss": 0.2936,
+      "step": 10468
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.850940394131275,
+      "learning_rate": 5.360474177220659e-07,
+      "loss": 1.162,
+      "step": 10469
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.180895013424875,
+      "learning_rate": 5.354512486401409e-07,
+      "loss": 0.4575,
+      "step": 10470
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.8108880858287897,
+      "learning_rate": 5.348553925036553e-07,
+      "loss": 0.5706,
+      "step": 10471
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8481081046692012,
+      "learning_rate": 5.342598493543727e-07,
+      "loss": 0.3095,
+      "step": 10472
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.895634104931042,
+      "learning_rate": 5.3366461923404e-07,
+      "loss": 0.7126,
+      "step": 10473
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.960115078884639,
+      "learning_rate": 5.330697021843795e-07,
+      "loss": 0.6181,
+      "step": 10474
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.34098382409636,
+      "learning_rate": 5.324750982470933e-07,
+      "loss": 0.5289,
+      "step": 10475
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.844540506448431,
+      "learning_rate": 5.318808074638598e-07,
+      "loss": 0.7852,
+      "step": 10476
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.316453704559759,
+      "learning_rate": 5.312868298763374e-07,
+      "loss": 0.4248,
+      "step": 10477
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.065511357492012,
+      "learning_rate": 5.306931655261588e-07,
+      "loss": 1.1797,
+      "step": 10478
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.088625813835447,
+      "learning_rate": 5.300998144549402e-07,
+      "loss": 0.7932,
+      "step": 10479
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.17514705001973,
+      "learning_rate": 5.29506776704271e-07,
+      "loss": 0.5734,
+      "step": 10480
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.5910957602546696,
+      "learning_rate": 5.289140523157205e-07,
+      "loss": 0.8801,
+      "step": 10481
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.6013748244996906,
+      "learning_rate": 5.283216413308367e-07,
+      "loss": 0.8214,
+      "step": 10482
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.6717264426904643,
+      "learning_rate": 5.277295437911462e-07,
+      "loss": 0.6352,
+      "step": 10483
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.371896629040389,
+      "learning_rate": 5.271377597381505e-07,
+      "loss": 0.5906,
+      "step": 10484
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.368054379749993,
+      "learning_rate": 5.265462892133317e-07,
+      "loss": 0.4887,
+      "step": 10485
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.8464376430535165,
+      "learning_rate": 5.259551322581496e-07,
+      "loss": 0.3931,
+      "step": 10486
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.399816522319251,
+      "learning_rate": 5.253642889140414e-07,
+      "loss": 0.4729,
+      "step": 10487
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.2664370652522123,
+      "learning_rate": 5.247737592224239e-07,
+      "loss": 0.304,
+      "step": 10488
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.5684092464370605,
+      "learning_rate": 5.241835432246888e-07,
+      "loss": 0.4109,
+      "step": 10489
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.106307996943481,
+      "learning_rate": 5.235936409622083e-07,
+      "loss": 0.6459,
+      "step": 10490
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.120738502869185,
+      "learning_rate": 5.230040524763325e-07,
+      "loss": 1.2144,
+      "step": 10491
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.702304509190002,
+      "learning_rate": 5.224147778083882e-07,
+      "loss": 1.2657,
+      "step": 10492
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.913241630622584,
+      "learning_rate": 5.218258169996825e-07,
+      "loss": 1.1311,
+      "step": 10493
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.9066351228096754,
+      "learning_rate": 5.212371700914976e-07,
+      "loss": 0.7973,
+      "step": 10494
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.202969004089788,
+      "learning_rate": 5.206488371250956e-07,
+      "loss": 1.1047,
+      "step": 10495
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.6915185468451486,
+      "learning_rate": 5.200608181417155e-07,
+      "loss": 0.9084,
+      "step": 10496
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.2002742632248755,
+      "learning_rate": 5.194731131825754e-07,
+      "loss": 0.6965,
+      "step": 10497
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.1255230862771204,
+      "learning_rate": 5.188857222888699e-07,
+      "loss": 0.386,
+      "step": 10498
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.113045842082105,
+      "learning_rate": 5.182986455017741e-07,
+      "loss": 0.7443,
+      "step": 10499
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.8650408809820687,
+      "learning_rate": 5.177118828624395e-07,
+      "loss": 0.8432,
+      "step": 10500
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.533967590398182,
+      "learning_rate": 5.171254344119941e-07,
+      "loss": 1.0344,
+      "step": 10501
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.553799879888968,
+      "learning_rate": 5.165393001915464e-07,
+      "loss": 0.7334,
+      "step": 10502
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.568057645612484,
+      "learning_rate": 5.159534802421817e-07,
+      "loss": 0.272,
+      "step": 10503
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.892616461853972,
+      "learning_rate": 5.153679746049628e-07,
+      "loss": 0.2415,
+      "step": 10504
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.2287915215924756,
+      "learning_rate": 5.147827833209334e-07,
+      "loss": 0.2845,
+      "step": 10505
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.98277451964769,
+      "learning_rate": 5.141979064311098e-07,
+      "loss": 0.5627,
+      "step": 10506
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.759412696679251,
+      "learning_rate": 5.136133439764907e-07,
+      "loss": 0.8121,
+      "step": 10507
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.6435535949517086,
+      "learning_rate": 5.130290959980511e-07,
+      "loss": 0.5884,
+      "step": 10508
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.5847272345069072,
+      "learning_rate": 5.12445162536746e-07,
+      "loss": 0.6392,
+      "step": 10509
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.7730194330075157,
+      "learning_rate": 5.11861543633504e-07,
+      "loss": 0.5449,
+      "step": 10510
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.836953255912099,
+      "learning_rate": 5.112782393292359e-07,
+      "loss": 0.8827,
+      "step": 10511
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.7660198150343227,
+      "learning_rate": 5.106952496648276e-07,
+      "loss": 0.5007,
+      "step": 10512
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.33503023950577,
+      "learning_rate": 5.101125746811447e-07,
+      "loss": 0.6088,
+      "step": 10513
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.8980156264743555,
+      "learning_rate": 5.095302144190307e-07,
+      "loss": 0.6212,
+      "step": 10514
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.84448207594496,
+      "learning_rate": 5.089481689193054e-07,
+      "loss": 0.8081,
+      "step": 10515
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.64459833793355,
+      "learning_rate": 5.083664382227688e-07,
+      "loss": 0.8825,
+      "step": 10516
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.043838681594187,
+      "learning_rate": 5.07785022370198e-07,
+      "loss": 0.3449,
+      "step": 10517
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.154060146279045,
+      "learning_rate": 5.072039214023461e-07,
+      "loss": 0.7281,
+      "step": 10518
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.96466531977924,
+      "learning_rate": 5.06623135359946e-07,
+      "loss": 0.8337,
+      "step": 10519
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.676759933945837,
+      "learning_rate": 5.060426642837096e-07,
+      "loss": 0.6837,
+      "step": 10520
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.565229230187216,
+      "learning_rate": 5.054625082143244e-07,
+      "loss": 0.9036,
+      "step": 10521
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.105733286335197,
+      "learning_rate": 5.048826671924573e-07,
+      "loss": 0.8276,
+      "step": 10522
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.453506602357791,
+      "learning_rate": 5.043031412587529e-07,
+      "loss": 0.7909,
+      "step": 10523
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.4796082685018095,
+      "learning_rate": 5.037239304538328e-07,
+      "loss": 0.3556,
+      "step": 10524
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.4196196199598,
+      "learning_rate": 5.031450348182976e-07,
+      "loss": 0.9265,
+      "step": 10525
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.712548641642958,
+      "learning_rate": 5.025664543927239e-07,
+      "loss": 0.3009,
+      "step": 10526
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.949476720099252,
+      "learning_rate": 5.019881892176692e-07,
+      "loss": 1.3988,
+      "step": 10527
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.287437585260928,
+      "learning_rate": 5.014102393336684e-07,
+      "loss": 0.7728,
+      "step": 10528
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.608351739883492,
+      "learning_rate": 5.008326047812306e-07,
+      "loss": 0.7378,
+      "step": 10529
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.24070022130551,
+      "learning_rate": 5.002552856008463e-07,
+      "loss": 1.0254,
+      "step": 10530
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.3390730038185388,
+      "learning_rate": 4.996782818329843e-07,
+      "loss": 0.6737,
+      "step": 10531
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.083408045700113,
+      "learning_rate": 4.991015935180887e-07,
+      "loss": 0.742,
+      "step": 10532
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.174766349101573,
+      "learning_rate": 4.985252206965841e-07,
+      "loss": 0.6072,
+      "step": 10533
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.172485055619851,
+      "learning_rate": 4.979491634088712e-07,
+      "loss": 0.6658,
+      "step": 10534
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.402005108799347,
+      "learning_rate": 4.973734216953285e-07,
+      "loss": 0.8842,
+      "step": 10535
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.90056761758074,
+      "learning_rate": 4.967979955963132e-07,
+      "loss": 1.0879,
+      "step": 10536
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.39097764998649,
+      "learning_rate": 4.962228851521606e-07,
+      "loss": 0.27,
+      "step": 10537
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.060768350538714,
+      "learning_rate": 4.956480904031829e-07,
+      "loss": 0.7793,
+      "step": 10538
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.0608895948984265,
+      "learning_rate": 4.950736113896726e-07,
+      "loss": 0.4767,
+      "step": 10539
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.2562936229286406,
+      "learning_rate": 4.94499448151895e-07,
+      "loss": 0.5844,
+      "step": 10540
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.4369122871795086,
+      "learning_rate": 4.939256007300997e-07,
+      "loss": 0.6791,
+      "step": 10541
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.343019423477376,
+      "learning_rate": 4.933520691645078e-07,
+      "loss": 0.7955,
+      "step": 10542
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.4430208185765965,
+      "learning_rate": 4.927788534953232e-07,
+      "loss": 0.2051,
+      "step": 10543
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.375757953035753,
+      "learning_rate": 4.922059537627249e-07,
+      "loss": 0.8462,
+      "step": 10544
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.787432113919724,
+      "learning_rate": 4.916333700068732e-07,
+      "loss": 0.7809,
+      "step": 10545
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.277283723496338,
+      "learning_rate": 4.910611022679002e-07,
+      "loss": 0.8131,
+      "step": 10546
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.915573451282051,
+      "learning_rate": 4.904891505859211e-07,
+      "loss": 0.7347,
+      "step": 10547
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.869333183646448,
+      "learning_rate": 4.899175150010266e-07,
+      "loss": 1.0938,
+      "step": 10548
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.985754897390461,
+      "learning_rate": 4.893461955532869e-07,
+      "loss": 0.7476,
+      "step": 10549
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.600746051641911,
+      "learning_rate": 4.887751922827483e-07,
+      "loss": 1.5825,
+      "step": 10550
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.604774612678929,
+      "learning_rate": 4.882045052294371e-07,
+      "loss": 0.7399,
+      "step": 10551
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.329949394899275,
+      "learning_rate": 4.876341344333535e-07,
+      "loss": 0.9543,
+      "step": 10552
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.183678652065017,
+      "learning_rate": 4.870640799344789e-07,
+      "loss": 0.6445,
+      "step": 10553
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.619817988896082,
+      "learning_rate": 4.864943417727719e-07,
+      "loss": 0.2499,
+      "step": 10554
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.7996139038893224,
+      "learning_rate": 4.8592491998817e-07,
+      "loss": 0.6501,
+      "step": 10555
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.8476390589206884,
+      "learning_rate": 4.85355814620585e-07,
+      "loss": 0.5333,
+      "step": 10556
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6991815993314117,
+      "learning_rate": 4.847870257099102e-07,
+      "loss": 0.2765,
+      "step": 10557
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.626329793778804,
+      "learning_rate": 4.842185532960142e-07,
+      "loss": 0.7608,
+      "step": 10558
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.239040453232224,
+      "learning_rate": 4.836503974187446e-07,
+      "loss": 0.6292,
+      "step": 10559
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.090350654471295,
+      "learning_rate": 4.830825581179266e-07,
+      "loss": 0.8787,
+      "step": 10560
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.8632163414524663,
+      "learning_rate": 4.825150354333641e-07,
+      "loss": 0.8996,
+      "step": 10561
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.597256036373677,
+      "learning_rate": 4.819478294048368e-07,
+      "loss": 0.8609,
+      "step": 10562
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.8255565498335624,
+      "learning_rate": 4.813809400721053e-07,
+      "loss": 0.7586,
+      "step": 10563
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.747533371702565,
+      "learning_rate": 4.808143674749044e-07,
+      "loss": 1.0456,
+      "step": 10564
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.7978246930803086,
+      "learning_rate": 4.802481116529484e-07,
+      "loss": 0.8228,
+      "step": 10565
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.369936623859059,
+      "learning_rate": 4.796821726459294e-07,
+      "loss": 0.9767,
+      "step": 10566
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.80137353729828,
+      "learning_rate": 4.79116550493518e-07,
+      "loss": 0.759,
+      "step": 10567
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.835421858453713,
+      "learning_rate": 4.785512452353619e-07,
+      "loss": 0.7454,
+      "step": 10568
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.034874998674549,
+      "learning_rate": 4.77986256911086e-07,
+      "loss": 0.3137,
+      "step": 10569
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.184732909644232,
+      "learning_rate": 4.774215855602932e-07,
+      "loss": 0.847,
+      "step": 10570
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.991251299295491,
+      "learning_rate": 4.768572312225645e-07,
+      "loss": 0.5923,
+      "step": 10571
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.1722346498282974,
+      "learning_rate": 4.762931939374604e-07,
+      "loss": 0.5217,
+      "step": 10572
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.769833138335216,
+      "learning_rate": 4.7572947374451563e-07,
+      "loss": 0.6477,
+      "step": 10573
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.4207100410905196,
+      "learning_rate": 4.751660706832456e-07,
+      "loss": 0.5049,
+      "step": 10574
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.179637135441105,
+      "learning_rate": 4.746029847931405e-07,
+      "loss": 0.6807,
+      "step": 10575
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.9932773081312645,
+      "learning_rate": 4.74040216113672e-07,
+      "loss": 0.4292,
+      "step": 10576
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.1567134578312914,
+      "learning_rate": 4.7347776468428755e-07,
+      "loss": 0.2362,
+      "step": 10577
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.808502722109823,
+      "learning_rate": 4.729156305444121e-07,
+      "loss": 0.2996,
+      "step": 10578
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.193178091505855,
+      "learning_rate": 4.7235381373344877e-07,
+      "loss": 0.34,
+      "step": 10579
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.854881487942244,
+      "learning_rate": 4.717923142907799e-07,
+      "loss": 1.1866,
+      "step": 10580
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1034841366520098,
+      "learning_rate": 4.7123113225576223e-07,
+      "loss": 0.1515,
+      "step": 10581
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.9591419888299337,
+      "learning_rate": 4.7067026766773273e-07,
+      "loss": 1.0687,
+      "step": 10582
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.852620985285636,
+      "learning_rate": 4.701097205660055e-07,
+      "loss": 1.2246,
+      "step": 10583
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.440507337223321,
+      "learning_rate": 4.695494909898729e-07,
+      "loss": 0.4791,
+      "step": 10584
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 6.249098578944652,
+      "learning_rate": 4.689895789786059e-07,
+      "loss": 1.2812,
+      "step": 10585
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.578958964412463,
+      "learning_rate": 4.684299845714485e-07,
+      "loss": 0.5956,
+      "step": 10586
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.58440156649555,
+      "learning_rate": 4.6787070780762833e-07,
+      "loss": 0.3894,
+      "step": 10587
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.827490135978755,
+      "learning_rate": 4.6731174872634844e-07,
+      "loss": 1.1186,
+      "step": 10588
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.8791858692725185,
+      "learning_rate": 4.6675310736678746e-07,
+      "loss": 1.1343,
+      "step": 10589
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.45158548480915,
+      "learning_rate": 4.661947837681052e-07,
+      "loss": 0.3851,
+      "step": 10590
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.8440433803230345,
+      "learning_rate": 4.656367779694382e-07,
+      "loss": 1.0207,
+      "step": 10591
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.627636686901322,
+      "learning_rate": 4.650790900098989e-07,
+      "loss": 0.5841,
+      "step": 10592
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 6.053180697309185,
+      "learning_rate": 4.6452171992857895e-07,
+      "loss": 0.9651,
+      "step": 10593
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.9797866509653845,
+      "learning_rate": 4.6396466776454816e-07,
+      "loss": 0.6952,
+      "step": 10594
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.283156809275132,
+      "learning_rate": 4.634079335568531e-07,
+      "loss": 1.3423,
+      "step": 10595
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.098370401479687,
+      "learning_rate": 4.628515173445186e-07,
+      "loss": 0.9491,
+      "step": 10596
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.4882523325858426,
+      "learning_rate": 4.6229541916654797e-07,
+      "loss": 0.6638,
+      "step": 10597
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.21272766290156,
+      "learning_rate": 4.6173963906191945e-07,
+      "loss": 0.6257,
+      "step": 10598
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.578440361226375,
+      "learning_rate": 4.611841770695913e-07,
+      "loss": 0.6912,
+      "step": 10599
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.7406030006374236,
+      "learning_rate": 4.6062903322849963e-07,
+      "loss": 0.5074,
+      "step": 10600
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.971787159460415,
+      "learning_rate": 4.600742075775572e-07,
+      "loss": 0.4895,
+      "step": 10601
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.549420946392138,
+      "learning_rate": 4.5951970015565617e-07,
+      "loss": 0.527,
+      "step": 10602
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.49758116442198,
+      "learning_rate": 4.5896551100166273e-07,
+      "loss": 0.7076,
+      "step": 10603
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.221557375377842,
+      "learning_rate": 4.584116401544253e-07,
+      "loss": 0.8774,
+      "step": 10604
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.704012548444121,
+      "learning_rate": 4.578580876527661e-07,
+      "loss": 0.905,
+      "step": 10605
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.8933981792163435,
+      "learning_rate": 4.573048535354874e-07,
+      "loss": 1.2283,
+      "step": 10606
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.6890289491684216,
+      "learning_rate": 4.5675193784136873e-07,
+      "loss": 0.7169,
+      "step": 10607
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.078110281832308,
+      "learning_rate": 4.5619934060916747e-07,
+      "loss": 0.4362,
+      "step": 10608
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.943365923867175,
+      "learning_rate": 4.556470618776171e-07,
+      "loss": 1.4185,
+      "step": 10609
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.7983566498192602,
+      "learning_rate": 4.5509510168543045e-07,
+      "loss": 0.7491,
+      "step": 10610
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.1864928678226625,
+      "learning_rate": 4.545434600712978e-07,
+      "loss": 0.6091,
+      "step": 10611
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.680425124795627,
+      "learning_rate": 4.5399213707388645e-07,
+      "loss": 1.067,
+      "step": 10612
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.2976887527520815,
+      "learning_rate": 4.5344113273184223e-07,
+      "loss": 0.6837,
+      "step": 10613
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.247737177096619,
+      "learning_rate": 4.5289044708378914e-07,
+      "loss": 0.2153,
+      "step": 10614
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.1577186393115704,
+      "learning_rate": 4.523400801683253e-07,
+      "loss": 0.4013,
+      "step": 10615
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.368416540525366,
+      "learning_rate": 4.517900320240304e-07,
+      "loss": 1.1825,
+      "step": 10616
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.433922677016759,
+      "learning_rate": 4.512403026894607e-07,
+      "loss": 0.6473,
+      "step": 10617
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.099567381738013,
+      "learning_rate": 4.5069089220315e-07,
+      "loss": 1.078,
+      "step": 10618
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.298397952514416,
+      "learning_rate": 4.5014180060360843e-07,
+      "loss": 0.7946,
+      "step": 10619
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.172604711180217,
+      "learning_rate": 4.4959302792932645e-07,
+      "loss": 1.2498,
+      "step": 10620
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.602856346642938,
+      "learning_rate": 4.490445742187688e-07,
+      "loss": 0.6376,
+      "step": 10621
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.904344180262937,
+      "learning_rate": 4.484964395103808e-07,
+      "loss": 0.8726,
+      "step": 10622
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.3263918944209077,
+      "learning_rate": 4.47948623842584e-07,
+      "loss": 0.4186,
+      "step": 10623
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.617491450908338,
+      "learning_rate": 4.4740112725377817e-07,
+      "loss": 0.7986,
+      "step": 10624
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.1027863871710375,
+      "learning_rate": 4.468539497823399e-07,
+      "loss": 0.7861,
+      "step": 10625
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.262894681306143,
+      "learning_rate": 4.4630709146662623e-07,
+      "loss": 0.4552,
+      "step": 10626
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.224403031382618,
+      "learning_rate": 4.4576055234496595e-07,
+      "loss": 0.8116,
+      "step": 10627
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.141197113716972,
+      "learning_rate": 4.4521433245567127e-07,
+      "loss": 0.7435,
+      "step": 10628
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.5200730619412,
+      "learning_rate": 4.446684318370292e-07,
+      "loss": 0.5186,
+      "step": 10629
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.16613430209177,
+      "learning_rate": 4.4412285052730543e-07,
+      "loss": 0.8888,
+      "step": 10630
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 6.131999128338534,
+      "learning_rate": 4.435775885647431e-07,
+      "loss": 0.9811,
+      "step": 10631
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.7427335505554438,
+      "learning_rate": 4.4303264598756167e-07,
+      "loss": 0.6292,
+      "step": 10632
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.312167453246354,
+      "learning_rate": 4.4248802283395953e-07,
+      "loss": 0.7849,
+      "step": 10633
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.724819689411017,
+      "learning_rate": 4.4194371914211385e-07,
+      "loss": 0.2586,
+      "step": 10634
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.132480578860827,
+      "learning_rate": 4.4139973495017584e-07,
+      "loss": 0.7393,
+      "step": 10635
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.937206205763876,
+      "learning_rate": 4.4085607029627717e-07,
+      "loss": 0.9206,
+      "step": 10636
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.1249392973486874,
+      "learning_rate": 4.403127252185274e-07,
+      "loss": 0.618,
+      "step": 10637
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.403196424429233,
+      "learning_rate": 4.397696997550105e-07,
+      "loss": 0.7381,
+      "step": 10638
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.567039543028251,
+      "learning_rate": 4.392269939437921e-07,
+      "loss": 0.5634,
+      "step": 10639
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.854015625729932,
+      "learning_rate": 4.3868460782291235e-07,
+      "loss": 0.743,
+      "step": 10640
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.9822235908521133,
+      "learning_rate": 4.381425414303908e-07,
+      "loss": 0.5683,
+      "step": 10641
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.996851986473055,
+      "learning_rate": 4.376007948042238e-07,
+      "loss": 1.2439,
+      "step": 10642
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.352824545867161,
+      "learning_rate": 4.370593679823865e-07,
+      "loss": 0.3105,
+      "step": 10643
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.6522769651597775,
+      "learning_rate": 4.3651826100282844e-07,
+      "loss": 0.6531,
+      "step": 10644
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.355682252300801,
+      "learning_rate": 4.3597747390348056e-07,
+      "loss": 0.6397,
+      "step": 10645
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.389657952791804,
+      "learning_rate": 4.354370067222485e-07,
+      "loss": 0.6547,
+      "step": 10646
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.5289816701473318,
+      "learning_rate": 4.348968594970171e-07,
+      "loss": 0.3732,
+      "step": 10647
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.168427711599652,
+      "learning_rate": 4.343570322656498e-07,
+      "loss": 0.8713,
+      "step": 10648
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.680600022944302,
+      "learning_rate": 4.3381752506598373e-07,
+      "loss": 1.2249,
+      "step": 10649
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.8633096515418357,
+      "learning_rate": 4.33278337935838e-07,
+      "loss": 0.6042,
+      "step": 10650
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.708536022959924,
+      "learning_rate": 4.3273947091300504e-07,
+      "loss": 0.3868,
+      "step": 10651
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.033168550954183,
+      "learning_rate": 4.322009240352587e-07,
+      "loss": 0.8516,
+      "step": 10652
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.8082117616109707,
+      "learning_rate": 4.316626973403487e-07,
+      "loss": 0.4788,
+      "step": 10653
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.888987878913506,
+      "learning_rate": 4.311247908660027e-07,
+      "loss": 0.6127,
+      "step": 10654
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.776096194500669,
+      "learning_rate": 4.305872046499243e-07,
+      "loss": 0.5333,
+      "step": 10655
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.281179360056835,
+      "learning_rate": 4.300499387297963e-07,
+      "loss": 0.3566,
+      "step": 10656
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.881596429018882,
+      "learning_rate": 4.2951299314327953e-07,
+      "loss": 1.1082,
+      "step": 10657
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.5948203241454566,
+      "learning_rate": 4.2897636792801123e-07,
+      "loss": 0.3076,
+      "step": 10658
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.400210588906877,
+      "learning_rate": 4.2844006312160625e-07,
+      "loss": 0.7595,
+      "step": 10659
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.7261392672939495,
+      "learning_rate": 4.2790407876165783e-07,
+      "loss": 0.9267,
+      "step": 10660
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 6.234313141007303,
+      "learning_rate": 4.2736841488573543e-07,
+      "loss": 1.3824,
+      "step": 10661
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.613363121137143,
+      "learning_rate": 4.268330715313862e-07,
+      "loss": 0.7748,
+      "step": 10662
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.808480011585258,
+      "learning_rate": 4.2629804873613676e-07,
+      "loss": 1.1513,
+      "step": 10663
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.47762144904202,
+      "learning_rate": 4.2576334653749e-07,
+      "loss": 0.9783,
+      "step": 10664
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.3851991281660267,
+      "learning_rate": 4.2522896497292465e-07,
+      "loss": 0.4157,
+      "step": 10665
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.29677868769382,
+      "learning_rate": 4.2469490407990033e-07,
+      "loss": 0.7274,
+      "step": 10666
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.316192431611226,
+      "learning_rate": 4.2416116389585094e-07,
+      "loss": 1.1019,
+      "step": 10667
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.803931542736004,
+      "learning_rate": 4.236277444581893e-07,
+      "loss": 0.6968,
+      "step": 10668
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.8032771517576185,
+      "learning_rate": 4.2309464580430614e-07,
+      "loss": 0.9831,
+      "step": 10669
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.969830532450726,
+      "learning_rate": 4.2256186797156986e-07,
+      "loss": 0.6159,
+      "step": 10670
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.424609582686892,
+      "learning_rate": 4.220294109973266e-07,
+      "loss": 0.5029,
+      "step": 10671
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.8175042161609865,
+      "learning_rate": 4.2149727491889725e-07,
+      "loss": 0.9314,
+      "step": 10672
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.148672746299841,
+      "learning_rate": 4.2096545977358294e-07,
+      "loss": 0.4523,
+      "step": 10673
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.4410525768851987,
+      "learning_rate": 4.2043396559866224e-07,
+      "loss": 0.4707,
+      "step": 10674
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.6024058720906362,
+      "learning_rate": 4.199027924313903e-07,
+      "loss": 0.8373,
+      "step": 10675
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.6521156174179907,
+      "learning_rate": 4.1937194030899966e-07,
+      "loss": 0.6826,
+      "step": 10676
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.334807522534701,
+      "learning_rate": 4.188414092687021e-07,
+      "loss": 1.1498,
+      "step": 10677
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9210667026365889,
+      "learning_rate": 4.183111993476835e-07,
+      "loss": 0.3626,
+      "step": 10678
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.4504094068519215,
+      "learning_rate": 4.177813105831102e-07,
+      "loss": 0.9136,
+      "step": 10679
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.646032517955008,
+      "learning_rate": 4.172517430121248e-07,
+      "loss": 0.3493,
+      "step": 10680
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.264478556092544,
+      "learning_rate": 4.1672249667184974e-07,
+      "loss": 0.4744,
+      "step": 10681
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.7508307297946977,
+      "learning_rate": 4.161935715993798e-07,
+      "loss": 0.7425,
+      "step": 10682
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 6.403066664815627,
+      "learning_rate": 4.1566496783179257e-07,
+      "loss": 1.1371,
+      "step": 10683
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.686267943481072,
+      "learning_rate": 4.1513668540613895e-07,
+      "loss": 1.2371,
+      "step": 10684
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.765613706065596,
+      "learning_rate": 4.1460872435945046e-07,
+      "loss": 0.5829,
+      "step": 10685
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.861893386867951,
+      "learning_rate": 4.1408108472873466e-07,
+      "loss": 1.0981,
+      "step": 10686
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.095293931200413,
+      "learning_rate": 4.1355376655097704e-07,
+      "loss": 0.732,
+      "step": 10687
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.023464390673694,
+      "learning_rate": 4.1302676986314126e-07,
+      "loss": 0.843,
+      "step": 10688
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.7289135220673826,
+      "learning_rate": 4.125000947021651e-07,
+      "loss": 0.5838,
+      "step": 10689
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.054357836962516,
+      "learning_rate": 4.1197374110496736e-07,
+      "loss": 0.2689,
+      "step": 10690
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.368630862806371,
+      "learning_rate": 4.1144770910844287e-07,
+      "loss": 0.8435,
+      "step": 10691
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.67632656360338,
+      "learning_rate": 4.1092199874946505e-07,
+      "loss": 0.3495,
+      "step": 10692
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.8132894463745277,
+      "learning_rate": 4.103966100648832e-07,
+      "loss": 0.7751,
+      "step": 10693
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.240815316954804,
+      "learning_rate": 4.0987154309152624e-07,
+      "loss": 0.5966,
+      "step": 10694
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.4030123337657274,
+      "learning_rate": 4.0934679786619635e-07,
+      "loss": 0.7006,
+      "step": 10695
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.607028351181928,
+      "learning_rate": 4.0882237442567753e-07,
+      "loss": 0.3138,
+      "step": 10696
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.019722337630489,
+      "learning_rate": 4.082982728067303e-07,
+      "loss": 1.0796,
+      "step": 10697
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.26180541436453,
+      "learning_rate": 4.077744930460903e-07,
+      "loss": 0.6425,
+      "step": 10698
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.5925048951248546,
+      "learning_rate": 4.072510351804726e-07,
+      "loss": 0.7575,
+      "step": 10699
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.354897154606622,
+      "learning_rate": 4.0672789924657065e-07,
+      "loss": 1.1631,
+      "step": 10700
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.8872884073162206,
+      "learning_rate": 4.062050852810523e-07,
+      "loss": 0.5028,
+      "step": 10701
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.504835082816432,
+      "learning_rate": 4.056825933205649e-07,
+      "loss": 0.9385,
+      "step": 10702
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.3366722530498745,
+      "learning_rate": 4.051604234017331e-07,
+      "loss": 0.7271,
+      "step": 10703
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.004438000810595,
+      "learning_rate": 4.0463857556115924e-07,
+      "loss": 1.0754,
+      "step": 10704
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.030954108170035,
+      "learning_rate": 4.0411704983542186e-07,
+      "loss": 0.6232,
+      "step": 10705
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.128803075417094,
+      "learning_rate": 4.0359584626107896e-07,
+      "loss": 0.8658,
+      "step": 10706
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.140969447636292,
+      "learning_rate": 4.0307496487466234e-07,
+      "loss": 0.5852,
+      "step": 10707
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.5609521974211,
+      "learning_rate": 4.025544057126851e-07,
+      "loss": 0.6727,
+      "step": 10708
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.963580000879209,
+      "learning_rate": 4.020341688116358e-07,
+      "loss": 0.7234,
+      "step": 10709
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.490147357337444,
+      "learning_rate": 4.0151425420798087e-07,
+      "loss": 0.6812,
+      "step": 10710
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.6926098684383915,
+      "learning_rate": 4.009946619381649e-07,
+      "loss": 1.3689,
+      "step": 10711
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.35012293405951,
+      "learning_rate": 4.004753920386073e-07,
+      "loss": 0.7819,
+      "step": 10712
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.130592052806018,
+      "learning_rate": 3.999564445457088e-07,
+      "loss": 0.7592,
+      "step": 10713
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.838509212592349,
+      "learning_rate": 3.994378194958426e-07,
+      "loss": 0.6659,
+      "step": 10714
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.5971797437510538,
+      "learning_rate": 3.9891951692536403e-07,
+      "loss": 0.4575,
+      "step": 10715
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.79462436955565,
+      "learning_rate": 3.984015368706029e-07,
+      "loss": 0.9588,
+      "step": 10716
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.457717012861435,
+      "learning_rate": 3.978838793678691e-07,
+      "loss": 0.7413,
+      "step": 10717
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.8643353666091755,
+      "learning_rate": 3.9736654445344583e-07,
+      "loss": 0.4063,
+      "step": 10718
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.625635562975463,
+      "learning_rate": 3.968495321635973e-07,
+      "loss": 0.955,
+      "step": 10719
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.033288993894299,
+      "learning_rate": 3.9633284253456306e-07,
+      "loss": 0.7561,
+      "step": 10720
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.926926563325641,
+      "learning_rate": 3.9581647560256175e-07,
+      "loss": 0.9063,
+      "step": 10721
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.649846015446501,
+      "learning_rate": 3.9530043140378783e-07,
+      "loss": 0.6667,
+      "step": 10722
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.924621114151401,
+      "learning_rate": 3.947847099744151e-07,
+      "loss": 0.8411,
+      "step": 10723
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.908430828876094,
+      "learning_rate": 3.942693113505908e-07,
+      "loss": 0.9769,
+      "step": 10724
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9654199707560478,
+      "learning_rate": 3.937542355684443e-07,
+      "loss": 0.5655,
+      "step": 10725
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.885417670420595,
+      "learning_rate": 3.93239482664079e-07,
+      "loss": 0.6008,
+      "step": 10726
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.18806641565675,
+      "learning_rate": 3.9272505267357817e-07,
+      "loss": 0.9083,
+      "step": 10727
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.632116258430988,
+      "learning_rate": 3.9221094563299924e-07,
+      "loss": 0.6983,
+      "step": 10728
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.415239353925382,
+      "learning_rate": 3.91697161578381e-07,
+      "loss": 1.0464,
+      "step": 10729
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.5919235662955313,
+      "learning_rate": 3.911837005457353e-07,
+      "loss": 0.4949,
+      "step": 10730
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.5218364501714263,
+      "learning_rate": 3.906705625710544e-07,
+      "loss": 0.5958,
+      "step": 10731
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.638661032322979,
+      "learning_rate": 3.9015774769030737e-07,
+      "loss": 0.7758,
+      "step": 10732
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.9578276632227913,
+      "learning_rate": 3.8964525593944037e-07,
+      "loss": 0.3478,
+      "step": 10733
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.5455863004408235,
+      "learning_rate": 3.8913308735437695e-07,
+      "loss": 0.6597,
+      "step": 10734
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9097376668172084,
+      "learning_rate": 3.8862124197101723e-07,
+      "loss": 0.3072,
+      "step": 10735
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.6449222677733357,
+      "learning_rate": 3.8810971982523925e-07,
+      "loss": 0.8079,
+      "step": 10736
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.213293657965279,
+      "learning_rate": 3.875985209528993e-07,
+      "loss": 0.6895,
+      "step": 10737
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9313900646944315,
+      "learning_rate": 3.870876453898292e-07,
+      "loss": 0.2604,
+      "step": 10738
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.680761500486827,
+      "learning_rate": 3.8657709317184043e-07,
+      "loss": 0.411,
+      "step": 10739
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.6893590662009825,
+      "learning_rate": 3.8606686433471986e-07,
+      "loss": 0.3308,
+      "step": 10740
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.943425828588057,
+      "learning_rate": 3.855569589142316e-07,
+      "loss": 0.5602,
+      "step": 10741
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.620359041990299,
+      "learning_rate": 3.8504737694611884e-07,
+      "loss": 0.3706,
+      "step": 10742
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.248144268331347,
+      "learning_rate": 3.8453811846610124e-07,
+      "loss": 0.8726,
+      "step": 10743
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.475629628031191,
+      "learning_rate": 3.8402918350987363e-07,
+      "loss": 1.1841,
+      "step": 10744
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.4896175540043695,
+      "learning_rate": 3.8352057211311187e-07,
+      "loss": 0.7999,
+      "step": 10745
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.9714899109623616,
+      "learning_rate": 3.830122843114681e-07,
+      "loss": 0.59,
+      "step": 10746
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.796663888117933,
+      "learning_rate": 3.825043201405687e-07,
+      "loss": 1.0062,
+      "step": 10747
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.150614319743733,
+      "learning_rate": 3.819966796360214e-07,
+      "loss": 0.6221,
+      "step": 10748
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.915798778933123,
+      "learning_rate": 3.8148936283340876e-07,
+      "loss": 0.9069,
+      "step": 10749
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.0373935126550538,
+      "learning_rate": 3.8098236976829237e-07,
+      "loss": 0.5306,
+      "step": 10750
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.5893956529612456,
+      "learning_rate": 3.804757004762105e-07,
+      "loss": 0.5189,
+      "step": 10751
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.03046284830745,
+      "learning_rate": 3.7996935499267753e-07,
+      "loss": 0.198,
+      "step": 10752
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5964226176458247,
+      "learning_rate": 3.7946333335318553e-07,
+      "loss": 0.2255,
+      "step": 10753
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.306032320380534,
+      "learning_rate": 3.7895763559320565e-07,
+      "loss": 0.597,
+      "step": 10754
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.996963558315773,
+      "learning_rate": 3.784522617481845e-07,
+      "loss": 0.1405,
+      "step": 10755
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.958372766379946,
+      "learning_rate": 3.779472118535471e-07,
+      "loss": 0.576,
+      "step": 10756
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.1559251715806473,
+      "learning_rate": 3.7744248594469514e-07,
+      "loss": 0.6124,
+      "step": 10757
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.607760712270118,
+      "learning_rate": 3.7693808405700693e-07,
+      "loss": 0.9607,
+      "step": 10758
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.64135442254822,
+      "learning_rate": 3.764340062258404e-07,
+      "loss": 0.9674,
+      "step": 10759
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.295047258336011,
+      "learning_rate": 3.7593025248652717e-07,
+      "loss": 0.5003,
+      "step": 10760
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.1594550418128007,
+      "learning_rate": 3.754268228743796e-07,
+      "loss": 0.5412,
+      "step": 10761
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.19015187923477,
+      "learning_rate": 3.74923717424685e-07,
+      "loss": 0.5206,
+      "step": 10762
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.49541054523754,
+      "learning_rate": 3.744209361727102e-07,
+      "loss": 0.9129,
+      "step": 10763
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.813429106006615,
+      "learning_rate": 3.7391847915369703e-07,
+      "loss": 0.8258,
+      "step": 10764
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.620999661923272,
+      "learning_rate": 3.7341634640286507e-07,
+      "loss": 0.7093,
+      "step": 10765
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.338545381906777,
+      "learning_rate": 3.729145379554128e-07,
+      "loss": 0.649,
+      "step": 10766
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.463822813724985,
+      "learning_rate": 3.724130538465137e-07,
+      "loss": 0.6253,
+      "step": 10767
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.582230910851982,
+      "learning_rate": 3.7191189411132145e-07,
+      "loss": 0.2331,
+      "step": 10768
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.611934416139467,
+      "learning_rate": 3.7141105878496284e-07,
+      "loss": 1.1343,
+      "step": 10769
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.332748873450337,
+      "learning_rate": 3.709105479025454e-07,
+      "loss": 0.382,
+      "step": 10770
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.8338039845774174,
+      "learning_rate": 3.704103614991528e-07,
+      "loss": 0.8078,
+      "step": 10771
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.871897611029115,
+      "learning_rate": 3.699104996098457e-07,
+      "loss": 1.0775,
+      "step": 10772
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.132853935258872,
+      "learning_rate": 3.694109622696629e-07,
+      "loss": 1.1363,
+      "step": 10773
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.776054690216075,
+      "learning_rate": 3.6891174951361905e-07,
+      "loss": 1.0977,
+      "step": 10774
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.445085974684133,
+      "learning_rate": 3.684128613767063e-07,
+      "loss": 1.2064,
+      "step": 10775
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.76484603760837,
+      "learning_rate": 3.6791429789389657e-07,
+      "loss": 0.5601,
+      "step": 10776
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.996172031233409,
+      "learning_rate": 3.674160591001347e-07,
+      "loss": 0.5389,
+      "step": 10777
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.919193116332945,
+      "learning_rate": 3.6691814503034607e-07,
+      "loss": 0.8366,
+      "step": 10778
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.3764859737831223,
+      "learning_rate": 3.664205557194322e-07,
+      "loss": 0.5064,
+      "step": 10779
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.426508144399529,
+      "learning_rate": 3.6592329120227254e-07,
+      "loss": 0.7038,
+      "step": 10780
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.179334859223737,
+      "learning_rate": 3.654263515137224e-07,
+      "loss": 0.7756,
+      "step": 10781
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.932650197611428,
+      "learning_rate": 3.649297366886145e-07,
+      "loss": 0.5344,
+      "step": 10782
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.2992859885900576,
+      "learning_rate": 3.644334467617605e-07,
+      "loss": 0.4674,
+      "step": 10783
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.9900866233744483,
+      "learning_rate": 3.6393748176794806e-07,
+      "loss": 0.8017,
+      "step": 10784
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.4378118306854746,
+      "learning_rate": 3.6344184174194166e-07,
+      "loss": 0.5481,
+      "step": 10785
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.747101774626812,
+      "learning_rate": 3.6294652671848506e-07,
+      "loss": 0.7064,
+      "step": 10786
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.2071706917138627,
+      "learning_rate": 3.6245153673229506e-07,
+      "loss": 0.473,
+      "step": 10787
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.71469652859469,
+      "learning_rate": 3.6195687181806995e-07,
+      "loss": 0.8903,
+      "step": 10788
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.7849053291905626,
+      "learning_rate": 3.614625320104831e-07,
+      "loss": 0.6381,
+      "step": 10789
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.502042212694051,
+      "learning_rate": 3.609685173441868e-07,
+      "loss": 0.8736,
+      "step": 10790
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.8469244147375665,
+      "learning_rate": 3.604748278538073e-07,
+      "loss": 0.7559,
+      "step": 10791
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.93578626328786,
+      "learning_rate": 3.599814635739518e-07,
+      "loss": 0.6431,
+      "step": 10792
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.71900741012042,
+      "learning_rate": 3.5948842453920164e-07,
+      "loss": 0.9382,
+      "step": 10793
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.776179272353354,
+      "learning_rate": 3.5899571078411743e-07,
+      "loss": 0.3649,
+      "step": 10794
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.3194053237425853,
+      "learning_rate": 3.5850332234323604e-07,
+      "loss": 0.7113,
+      "step": 10795
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.0038277251839287,
+      "learning_rate": 3.580112592510715e-07,
+      "loss": 0.3154,
+      "step": 10796
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.146379388211302,
+      "learning_rate": 3.5751952154211734e-07,
+      "loss": 0.5958,
+      "step": 10797
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.392956378574398,
+      "learning_rate": 3.570281092508393e-07,
+      "loss": 0.4243,
+      "step": 10798
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.6560757309435115,
+      "learning_rate": 3.565370224116843e-07,
+      "loss": 0.83,
+      "step": 10799
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.617587980117154,
+      "learning_rate": 3.560462610590759e-07,
+      "loss": 1.4945,
+      "step": 10800
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.719625405410017,
+      "learning_rate": 3.555558252274144e-07,
+      "loss": 1.2418,
+      "step": 10801
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.920088732077162,
+      "learning_rate": 3.550657149510761e-07,
+      "loss": 0.9742,
+      "step": 10802
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.4276505680763347,
+      "learning_rate": 3.545759302644175e-07,
+      "loss": 0.3711,
+      "step": 10803
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.042615404497343,
+      "learning_rate": 3.540864712017689e-07,
+      "loss": 0.6435,
+      "step": 10804
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.142617960583135,
+      "learning_rate": 3.5359733779743887e-07,
+      "loss": 1.032,
+      "step": 10805
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.206168299136099,
+      "learning_rate": 3.531085300857151e-07,
+      "loss": 1.0329,
+      "step": 10806
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.689641946745283,
+      "learning_rate": 3.526200481008596e-07,
+      "loss": 1.1883,
+      "step": 10807
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.51102607546849,
+      "learning_rate": 3.5213189187711383e-07,
+      "loss": 0.5884,
+      "step": 10808
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.326322472854147,
+      "learning_rate": 3.516440614486943e-07,
+      "loss": 1.3112,
+      "step": 10809
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.142139814049927,
+      "learning_rate": 3.5115655684979653e-07,
+      "loss": 0.6651,
+      "step": 10810
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.5554431087528937,
+      "learning_rate": 3.50669378114592e-07,
+      "loss": 0.8697,
+      "step": 10811
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.002869827342025,
+      "learning_rate": 3.5018252527723005e-07,
+      "loss": 0.7896,
+      "step": 10812
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.445687922168761,
+      "learning_rate": 3.4969599837183677e-07,
+      "loss": 1.2917,
+      "step": 10813
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.044010478873331,
+      "learning_rate": 3.4920979743251704e-07,
+      "loss": 0.9237,
+      "step": 10814
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.9503255148776963,
+      "learning_rate": 3.487239224933492e-07,
+      "loss": 0.5253,
+      "step": 10815
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.5457223215662785,
+      "learning_rate": 3.482383735883921e-07,
+      "loss": 0.8796,
+      "step": 10816
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.289237360154442,
+      "learning_rate": 3.4775315075168014e-07,
+      "loss": 0.9533,
+      "step": 10817
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.7949480360010073,
+      "learning_rate": 3.472682540172262e-07,
+      "loss": 0.8025,
+      "step": 10818
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.30689665215385,
+      "learning_rate": 3.467836834190186e-07,
+      "loss": 0.484,
+      "step": 10819
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.0670269504963104,
+      "learning_rate": 3.462994389910246e-07,
+      "loss": 0.7331,
+      "step": 10820
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.098845678264885,
+      "learning_rate": 3.4581552076718597e-07,
+      "loss": 1.295,
+      "step": 10821
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.553248193694052,
+      "learning_rate": 3.453319287814255e-07,
+      "loss": 0.7411,
+      "step": 10822
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.688140717897219,
+      "learning_rate": 3.4484866306763896e-07,
+      "loss": 1.076,
+      "step": 10823
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.3193966087256697,
+      "learning_rate": 3.4436572365970145e-07,
+      "loss": 0.4711,
+      "step": 10824
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.2865368687166274,
+      "learning_rate": 3.438831105914653e-07,
+      "loss": 0.516,
+      "step": 10825
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.2698605560686738,
+      "learning_rate": 3.4340082389676065e-07,
+      "loss": 0.3486,
+      "step": 10826
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.251032302544349,
+      "learning_rate": 3.429188636093922e-07,
+      "loss": 0.4758,
+      "step": 10827
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.176776415080748,
+      "learning_rate": 3.4243722976314285e-07,
+      "loss": 0.9426,
+      "step": 10828
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.8686624142305717,
+      "learning_rate": 3.4195592239177455e-07,
+      "loss": 0.5923,
+      "step": 10829
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.897109402196155,
+      "learning_rate": 3.4147494152902414e-07,
+      "loss": 0.8303,
+      "step": 10830
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.5272091603729154,
+      "learning_rate": 3.4099428720860693e-07,
+      "loss": 0.4118,
+      "step": 10831
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.1253640658859645,
+      "learning_rate": 3.4051395946421374e-07,
+      "loss": 0.4998,
+      "step": 10832
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 6.410903311671439,
+      "learning_rate": 3.4003395832951315e-07,
+      "loss": 1.0841,
+      "step": 10833
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.405568517002049,
+      "learning_rate": 3.3955428383815267e-07,
+      "loss": 0.825,
+      "step": 10834
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.344647787528466,
+      "learning_rate": 3.3907493602375386e-07,
+      "loss": 0.6518,
+      "step": 10835
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.571118190302611,
+      "learning_rate": 3.385959149199186e-07,
+      "loss": 0.6786,
+      "step": 10836
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.6834737837192835,
+      "learning_rate": 3.3811722056022287e-07,
+      "loss": 0.8869,
+      "step": 10837
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 6.3061940341381275,
+      "learning_rate": 3.3763885297822153e-07,
+      "loss": 1.5261,
+      "step": 10838
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.5161441151787303,
+      "learning_rate": 3.371608122074455e-07,
+      "loss": 0.5645,
+      "step": 10839
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 6.081472187373712,
+      "learning_rate": 3.36683098281404e-07,
+      "loss": 1.2119,
+      "step": 10840
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.171022016722354,
+      "learning_rate": 3.36205711233582e-07,
+      "loss": 1.1501,
+      "step": 10841
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.322989050379168,
+      "learning_rate": 3.3572865109744334e-07,
+      "loss": 0.767,
+      "step": 10842
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.025603277700914,
+      "learning_rate": 3.3525191790642733e-07,
+      "loss": 0.5912,
+      "step": 10843
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.5155571855514176,
+      "learning_rate": 3.347755116939505e-07,
+      "loss": 0.573,
+      "step": 10844
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.9386049896185673,
+      "learning_rate": 3.342994324934068e-07,
+      "loss": 0.5738,
+      "step": 10845
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.938087478700539,
+      "learning_rate": 3.338236803381684e-07,
+      "loss": 1.1193,
+      "step": 10846
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.832406399777737,
+      "learning_rate": 3.3334825526158185e-07,
+      "loss": 1.208,
+      "step": 10847
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.795716617602275,
+      "learning_rate": 3.328731572969746e-07,
+      "loss": 0.4324,
+      "step": 10848
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.7713239105931833,
+      "learning_rate": 3.32398386477647e-07,
+      "loss": 0.4664,
+      "step": 10849
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.978762454868344,
+      "learning_rate": 3.319239428368787e-07,
+      "loss": 1.4731,
+      "step": 10850
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.066659139746784,
+      "learning_rate": 3.3144982640792633e-07,
+      "loss": 0.7836,
+      "step": 10851
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.594254731150115,
+      "learning_rate": 3.309760372240245e-07,
+      "loss": 0.6424,
+      "step": 10852
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.458815645513699,
+      "learning_rate": 3.3050257531838213e-07,
+      "loss": 0.659,
+      "step": 10853
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.606011251269728,
+      "learning_rate": 3.300294407241883e-07,
+      "loss": 0.3979,
+      "step": 10854
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.3601353790940625,
+      "learning_rate": 3.2955663347460586e-07,
+      "loss": 0.2911,
+      "step": 10855
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.2947077288587123,
+      "learning_rate": 3.2908415360277777e-07,
+      "loss": 0.7556,
+      "step": 10856
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.6814920758815863,
+      "learning_rate": 3.2861200114182257e-07,
+      "loss": 0.5605,
+      "step": 10857
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.588671370200977,
+      "learning_rate": 3.2814017612483596e-07,
+      "loss": 0.9264,
+      "step": 10858
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.818531071367023,
+      "learning_rate": 3.276686785848915e-07,
+      "loss": 1.2616,
+      "step": 10859
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.513390776495794,
+      "learning_rate": 3.2719750855503886e-07,
+      "loss": 0.7851,
+      "step": 10860
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.164370528259902,
+      "learning_rate": 3.267266660683044e-07,
+      "loss": 0.4353,
+      "step": 10861
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.7186871412994567,
+      "learning_rate": 3.2625615115769225e-07,
+      "loss": 0.2845,
+      "step": 10862
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.2796635993385106,
+      "learning_rate": 3.257859638561839e-07,
+      "loss": 0.4609,
+      "step": 10863
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.077730283617591,
+      "learning_rate": 3.2531610419673675e-07,
+      "loss": 0.4293,
+      "step": 10864
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.234160766755855,
+      "learning_rate": 3.248465722122868e-07,
+      "loss": 0.3752,
+      "step": 10865
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.300505913411803,
+      "learning_rate": 3.24377367935747e-07,
+      "loss": 0.7588,
+      "step": 10866
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.388942913410447,
+      "learning_rate": 3.2390849140000403e-07,
+      "loss": 0.8793,
+      "step": 10867
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.0979214837702216,
+      "learning_rate": 3.2343994263792586e-07,
+      "loss": 0.5667,
+      "step": 10868
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.97473027058986,
+      "learning_rate": 3.229717216823552e-07,
+      "loss": 1.0228,
+      "step": 10869
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.864386867807811,
+      "learning_rate": 3.2250382856611193e-07,
+      "loss": 0.6682,
+      "step": 10870
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.2722788925574453,
+      "learning_rate": 3.220362633219948e-07,
+      "loss": 0.5731,
+      "step": 10871
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.379496879342514,
+      "learning_rate": 3.2156902598277585e-07,
+      "loss": 0.8015,
+      "step": 10872
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.994419351663586,
+      "learning_rate": 3.2110211658120784e-07,
+      "loss": 0.6576,
+      "step": 10873
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.780488642815016,
+      "learning_rate": 3.206355351500184e-07,
+      "loss": 0.6979,
+      "step": 10874
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.419326211069094,
+      "learning_rate": 3.2016928172191377e-07,
+      "loss": 0.9318,
+      "step": 10875
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.9204127882648647,
+      "learning_rate": 3.1970335632957595e-07,
+      "loss": 0.4459,
+      "step": 10876
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.156120085557064,
+      "learning_rate": 3.1923775900566444e-07,
+      "loss": 0.6487,
+      "step": 10877
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.2345160732754623,
+      "learning_rate": 3.1877248978281484e-07,
+      "loss": 0.5941,
+      "step": 10878
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.4191464683206716,
+      "learning_rate": 3.18307548693641e-07,
+      "loss": 0.3887,
+      "step": 10879
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.448224384755552,
+      "learning_rate": 3.17842935770733e-07,
+      "loss": 0.6698,
+      "step": 10880
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.715905562647058,
+      "learning_rate": 3.173786510466581e-07,
+      "loss": 0.5083,
+      "step": 10881
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.933740117090663,
+      "learning_rate": 3.1691469455396196e-07,
+      "loss": 0.5269,
+      "step": 10882
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.060932225469689,
+      "learning_rate": 3.164510663251641e-07,
+      "loss": 0.5333,
+      "step": 10883
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.980209462101881,
+      "learning_rate": 3.159877663927635e-07,
+      "loss": 1.1751,
+      "step": 10884
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.362802845681076,
+      "learning_rate": 3.15524794789237e-07,
+      "loss": 0.3487,
+      "step": 10885
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.142789023875797,
+      "learning_rate": 3.1506215154703424e-07,
+      "loss": 0.5471,
+      "step": 10886
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.8573988934118852,
+      "learning_rate": 3.145998366985853e-07,
+      "loss": 0.7298,
+      "step": 10887
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.515108176966938,
+      "learning_rate": 3.141378502762982e-07,
+      "loss": 1.0164,
+      "step": 10888
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8633670065312347,
+      "learning_rate": 3.136761923125542e-07,
+      "loss": 0.3277,
+      "step": 10889
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.194830584266059,
+      "learning_rate": 3.1321486283971357e-07,
+      "loss": 0.4989,
+      "step": 10890
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.6276186717088548,
+      "learning_rate": 3.127538618901144e-07,
+      "loss": 0.5732,
+      "step": 10891
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.570089495193147,
+      "learning_rate": 3.122931894960707e-07,
+      "loss": 0.8861,
+      "step": 10892
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.0135959179798055,
+      "learning_rate": 3.118328456898734e-07,
+      "loss": 0.9223,
+      "step": 10893
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.141074146576756,
+      "learning_rate": 3.1137283050379165e-07,
+      "loss": 0.8711,
+      "step": 10894
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.582126875808922,
+      "learning_rate": 3.10913143970068e-07,
+      "loss": 0.4177,
+      "step": 10895
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.718422207878949,
+      "learning_rate": 3.104537861209267e-07,
+      "loss": 1.2523,
+      "step": 10896
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 6.4516793569985955,
+      "learning_rate": 3.0999475698856583e-07,
+      "loss": 1.3102,
+      "step": 10897
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.7206114005265314,
+      "learning_rate": 3.095360566051614e-07,
+      "loss": 0.5341,
+      "step": 10898
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.436185251661134,
+      "learning_rate": 3.090776850028671e-07,
+      "loss": 0.7484,
+      "step": 10899
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 6.066747356792933,
+      "learning_rate": 3.086196422138116e-07,
+      "loss": 1.1466,
+      "step": 10900
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.889368480857732,
+      "learning_rate": 3.0816192827010317e-07,
+      "loss": 0.9629,
+      "step": 10901
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.6207016500856315,
+      "learning_rate": 3.077045432038234e-07,
+      "loss": 0.4401,
+      "step": 10902
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.332940686131082,
+      "learning_rate": 3.0724748704703435e-07,
+      "loss": 1.1868,
+      "step": 10903
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.058409923124676,
+      "learning_rate": 3.0679075983177376e-07,
+      "loss": 0.9681,
+      "step": 10904
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.279635932894341,
+      "learning_rate": 3.063343615900555e-07,
+      "loss": 0.5421,
+      "step": 10905
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.102073653497036,
+      "learning_rate": 3.0587829235387277e-07,
+      "loss": 0.3761,
+      "step": 10906
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.8061269220006597,
+      "learning_rate": 3.0542255215519177e-07,
+      "loss": 0.7479,
+      "step": 10907
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.31739768655144,
+      "learning_rate": 3.0496714102595914e-07,
+      "loss": 1.0301,
+      "step": 10908
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.5912994301111514,
+      "learning_rate": 3.0451205899809764e-07,
+      "loss": 0.5204,
+      "step": 10909
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.0923652953300005,
+      "learning_rate": 3.0405730610350516e-07,
+      "loss": 0.88,
+      "step": 10910
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.2713263066201983,
+      "learning_rate": 3.0360288237406e-07,
+      "loss": 0.6832,
+      "step": 10911
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.892186852611783,
+      "learning_rate": 3.0314878784161284e-07,
+      "loss": 0.8156,
+      "step": 10912
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.36463854352164,
+      "learning_rate": 3.0269502253799485e-07,
+      "loss": 0.7741,
+      "step": 10913
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.5927460345059146,
+      "learning_rate": 3.0224158649501343e-07,
+      "loss": 0.5198,
+      "step": 10914
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.516528093229124,
+      "learning_rate": 3.017884797444526e-07,
+      "loss": 0.5627,
+      "step": 10915
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.563189913634584,
+      "learning_rate": 3.013357023180724e-07,
+      "loss": 0.779,
+      "step": 10916
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.590422596413634,
+      "learning_rate": 3.008832542476109e-07,
+      "loss": 0.6201,
+      "step": 10917
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.860653508579803,
+      "learning_rate": 3.0043113556478207e-07,
+      "loss": 1.1325,
+      "step": 10918
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.0539682300732185,
+      "learning_rate": 2.999793463012779e-07,
+      "loss": 1.1529,
+      "step": 10919
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.861370069172823,
+      "learning_rate": 2.995278864887674e-07,
+      "loss": 1.1975,
+      "step": 10920
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.283954230503348,
+      "learning_rate": 2.990767561588953e-07,
+      "loss": 1.0167,
+      "step": 10921
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.271556079490615,
+      "learning_rate": 2.986259553432841e-07,
+      "loss": 1.0017,
+      "step": 10922
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8648335588731393,
+      "learning_rate": 2.98175484073534e-07,
+      "loss": 0.2049,
+      "step": 10923
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.5805437442033075,
+      "learning_rate": 2.977253423812193e-07,
+      "loss": 0.901,
+      "step": 10924
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.8726092892928534,
+      "learning_rate": 2.9727553029789303e-07,
+      "loss": 0.5756,
+      "step": 10925
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.73696508058446,
+      "learning_rate": 2.9682604785508664e-07,
+      "loss": 0.6246,
+      "step": 10926
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.6695261519349907,
+      "learning_rate": 2.963768950843054e-07,
+      "loss": 0.5638,
+      "step": 10927
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.692648401985274,
+      "learning_rate": 2.9592807201703486e-07,
+      "loss": 1.2495,
+      "step": 10928
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.626581084693792,
+      "learning_rate": 2.9547957868473307e-07,
+      "loss": 0.7701,
+      "step": 10929
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.8172435855293267,
+      "learning_rate": 2.9503141511883884e-07,
+      "loss": 0.5745,
+      "step": 10930
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.630192348382033,
+      "learning_rate": 2.9458358135076693e-07,
+      "loss": 0.4044,
+      "step": 10931
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.808703663511119,
+      "learning_rate": 2.9413607741190733e-07,
+      "loss": 0.9562,
+      "step": 10932
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.929016503818397,
+      "learning_rate": 2.936889033336288e-07,
+      "loss": 0.5762,
+      "step": 10933
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.436744038954424,
+      "learning_rate": 2.9324205914727674e-07,
+      "loss": 0.8342,
+      "step": 10934
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.253131568729143,
+      "learning_rate": 2.9279554488417186e-07,
+      "loss": 0.6823,
+      "step": 10935
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.066108099578079,
+      "learning_rate": 2.9234936057561336e-07,
+      "loss": 0.2603,
+      "step": 10936
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.652093687095228,
+      "learning_rate": 2.919035062528769e-07,
+      "loss": 0.9055,
+      "step": 10937
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.7600104969135906,
+      "learning_rate": 2.914579819472152e-07,
+      "loss": 0.4802,
+      "step": 10938
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.9176102039855842,
+      "learning_rate": 2.910127876898572e-07,
+      "loss": 0.578,
+      "step": 10939
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.72597740012647,
+      "learning_rate": 2.905679235120096e-07,
+      "loss": 0.8006,
+      "step": 10940
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.30238468052772,
+      "learning_rate": 2.9012338944485463e-07,
+      "loss": 0.7967,
+      "step": 10941
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.69012051220324,
+      "learning_rate": 2.89679185519553e-07,
+      "loss": 0.6067,
+      "step": 10942
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.539522560333616,
+      "learning_rate": 2.8923531176724027e-07,
+      "loss": 0.8178,
+      "step": 10943
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.561569584044321,
+      "learning_rate": 2.887917682190311e-07,
+      "loss": 1.2437,
+      "step": 10944
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.096023672049721,
+      "learning_rate": 2.883485549060167e-07,
+      "loss": 0.7392,
+      "step": 10945
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.5030239018798355,
+      "learning_rate": 2.879056718592627e-07,
+      "loss": 0.4574,
+      "step": 10946
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.384378182815733,
+      "learning_rate": 2.8746311910981485e-07,
+      "loss": 0.6167,
+      "step": 10947
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 6.413790629940976,
+      "learning_rate": 2.8702089668869227e-07,
+      "loss": 1.2176,
+      "step": 10948
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.5476452228294972,
+      "learning_rate": 2.86579004626894e-07,
+      "loss": 0.4998,
+      "step": 10949
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.578922777726011,
+      "learning_rate": 2.861374429553948e-07,
+      "loss": 0.8611,
+      "step": 10950
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.9133060061104805,
+      "learning_rate": 2.856962117051465e-07,
+      "loss": 0.9554,
+      "step": 10951
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.709850585262299,
+      "learning_rate": 2.852553109070766e-07,
+      "loss": 0.5331,
+      "step": 10952
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.8510311396142125,
+      "learning_rate": 2.8481474059209033e-07,
+      "loss": 0.5582,
+      "step": 10953
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.785827042551584,
+      "learning_rate": 2.8437450079107034e-07,
+      "loss": 0.6103,
+      "step": 10954
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.681038383589419,
+      "learning_rate": 2.839345915348757e-07,
+      "loss": 1.1143,
+      "step": 10955
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.7286450723732525,
+      "learning_rate": 2.8349501285434123e-07,
+      "loss": 0.9406,
+      "step": 10956
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.4851305243777806,
+      "learning_rate": 2.830557647802812e-07,
+      "loss": 0.3931,
+      "step": 10957
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.76467683896015,
+      "learning_rate": 2.8261684734348316e-07,
+      "loss": 0.9277,
+      "step": 10958
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.7793588360667427,
+      "learning_rate": 2.8217826057471423e-07,
+      "loss": 0.46,
+      "step": 10959
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.756079446742792,
+      "learning_rate": 2.817400045047164e-07,
+      "loss": 0.5393,
+      "step": 10960
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.3470853112431773,
+      "learning_rate": 2.813020791642118e-07,
+      "loss": 0.3717,
+      "step": 10961
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.240363172485346,
+      "learning_rate": 2.808644845838943e-07,
+      "loss": 0.5289,
+      "step": 10962
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.777083156136472,
+      "learning_rate": 2.804272207944397e-07,
+      "loss": 0.877,
+      "step": 10963
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 6.051156731001556,
+      "learning_rate": 2.799902878264965e-07,
+      "loss": 1.5133,
+      "step": 10964
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.240849695462603,
+      "learning_rate": 2.7955368571069284e-07,
+      "loss": 0.2849,
+      "step": 10965
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.070898820430699,
+      "learning_rate": 2.791174144776321e-07,
+      "loss": 0.8403,
+      "step": 10966
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.889092870804847,
+      "learning_rate": 2.7868147415789526e-07,
+      "loss": 0.5334,
+      "step": 10967
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.791923339154662,
+      "learning_rate": 2.782458647820407e-07,
+      "loss": 0.4653,
+      "step": 10968
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.88705712106383,
+      "learning_rate": 2.778105863806013e-07,
+      "loss": 1.1669,
+      "step": 10969
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.831371608629582,
+      "learning_rate": 2.7737563898408814e-07,
+      "loss": 0.3367,
+      "step": 10970
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 6.111330519551319,
+      "learning_rate": 2.769410226229902e-07,
+      "loss": 0.879,
+      "step": 10971
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.635508115964347,
+      "learning_rate": 2.765067373277719e-07,
+      "loss": 0.2942,
+      "step": 10972
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.9739200019422767,
+      "learning_rate": 2.760727831288745e-07,
+      "loss": 0.3165,
+      "step": 10973
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.514726040277123,
+      "learning_rate": 2.756391600567171e-07,
+      "loss": 0.9741,
+      "step": 10974
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.2116695143322636,
+      "learning_rate": 2.7520586814169303e-07,
+      "loss": 0.5593,
+      "step": 10975
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.753408779265016,
+      "learning_rate": 2.7477290741417526e-07,
+      "loss": 1.1868,
+      "step": 10976
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.6749590100957485,
+      "learning_rate": 2.7434027790451346e-07,
+      "loss": 0.6321,
+      "step": 10977
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.9872602389265652,
+      "learning_rate": 2.739079796430316e-07,
+      "loss": 0.6642,
+      "step": 10978
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.466734413380715,
+      "learning_rate": 2.7347601266003165e-07,
+      "loss": 0.8289,
+      "step": 10979
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.257607083269077,
+      "learning_rate": 2.730443769857943e-07,
+      "loss": 0.6408,
+      "step": 10980
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.639939952215677,
+      "learning_rate": 2.726130726505738e-07,
+      "loss": 0.3134,
+      "step": 10981
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.701429366481553,
+      "learning_rate": 2.721820996846031e-07,
+      "loss": 0.5872,
+      "step": 10982
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.505066897522115,
+      "learning_rate": 2.717514581180919e-07,
+      "loss": 0.8732,
+      "step": 10983
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.8617198443788148,
+      "learning_rate": 2.7132114798122557e-07,
+      "loss": 0.8892,
+      "step": 10984
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.358700844128491,
+      "learning_rate": 2.708911693041683e-07,
+      "loss": 1.0066,
+      "step": 10985
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.923681167525303,
+      "learning_rate": 2.7046152211705865e-07,
+      "loss": 0.5607,
+      "step": 10986
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.404828609512687,
+      "learning_rate": 2.7003220645001325e-07,
+      "loss": 0.482,
+      "step": 10987
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.090455246821734,
+      "learning_rate": 2.696032223331252e-07,
+      "loss": 0.5434,
+      "step": 10988
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.77128135135535,
+      "learning_rate": 2.6917456979646426e-07,
+      "loss": 0.2923,
+      "step": 10989
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.6209549505587475,
+      "learning_rate": 2.687462488700776e-07,
+      "loss": 1.2092,
+      "step": 10990
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.6326542842188356,
+      "learning_rate": 2.683182595839889e-07,
+      "loss": 0.7548,
+      "step": 10991
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.3388253743696557,
+      "learning_rate": 2.6789060196819705e-07,
+      "loss": 0.7041,
+      "step": 10992
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.280074955140449,
+      "learning_rate": 2.6746327605268017e-07,
+      "loss": 0.769,
+      "step": 10993
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.300141434768321,
+      "learning_rate": 2.670362818673922e-07,
+      "loss": 0.8973,
+      "step": 10994
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.038387138665519,
+      "learning_rate": 2.666096194422624e-07,
+      "loss": 0.2636,
+      "step": 10995
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.988886270242992,
+      "learning_rate": 2.6618328880719803e-07,
+      "loss": 0.6738,
+      "step": 10996
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.710406907947948,
+      "learning_rate": 2.6575728999208404e-07,
+      "loss": 0.8795,
+      "step": 10997
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.047388030931536,
+      "learning_rate": 2.653316230267805e-07,
+      "loss": 0.933,
+      "step": 10998
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.702221769066459,
+      "learning_rate": 2.649062879411246e-07,
+      "loss": 0.8378,
+      "step": 10999
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.787486009716643,
+      "learning_rate": 2.644812847649303e-07,
+      "loss": 0.5252,
+      "step": 11000
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4915548159213152,
+      "learning_rate": 2.640566135279893e-07,
+      "loss": 0.126,
+      "step": 11001
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.430806272111197,
+      "learning_rate": 2.636322742600689e-07,
+      "loss": 0.9382,
+      "step": 11002
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.450059826246382,
+      "learning_rate": 2.632082669909136e-07,
+      "loss": 0.541,
+      "step": 11003
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9120320385365477,
+      "learning_rate": 2.627845917502442e-07,
+      "loss": 0.3607,
+      "step": 11004
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.437954820256922,
+      "learning_rate": 2.6236124856775793e-07,
+      "loss": 0.7392,
+      "step": 11005
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.016078629052368,
+      "learning_rate": 2.6193823747313e-07,
+      "loss": 0.9399,
+      "step": 11006
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.8336758820553998,
+      "learning_rate": 2.6151555849601107e-07,
+      "loss": 0.5607,
+      "step": 11007
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.533114744036621,
+      "learning_rate": 2.6109321166603087e-07,
+      "loss": 0.9896,
+      "step": 11008
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.655681551168729,
+      "learning_rate": 2.6067119701279175e-07,
+      "loss": 0.8572,
+      "step": 11009
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.83599426448416,
+      "learning_rate": 2.6024951456587677e-07,
+      "loss": 0.9201,
+      "step": 11010
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.348684784011196,
+      "learning_rate": 2.5982816435484283e-07,
+      "loss": 0.3516,
+      "step": 11011
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9607888032500733,
+      "learning_rate": 2.5940714640922516e-07,
+      "loss": 0.2168,
+      "step": 11012
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.1593505187596693,
+      "learning_rate": 2.5898646075853573e-07,
+      "loss": 0.6291,
+      "step": 11013
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.7827509657674843,
+      "learning_rate": 2.5856610743226265e-07,
+      "loss": 0.4637,
+      "step": 11014
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.249129440152548,
+      "learning_rate": 2.5814608645987e-07,
+      "loss": 1.2756,
+      "step": 11015
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.826583744501296,
+      "learning_rate": 2.5772639787080056e-07,
+      "loss": 1.0892,
+      "step": 11016
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.078319979925435,
+      "learning_rate": 2.5730704169447176e-07,
+      "loss": 0.9656,
+      "step": 11017
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.0615261040398014,
+      "learning_rate": 2.5688801796027895e-07,
+      "loss": 0.4221,
+      "step": 11018
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.3212155015481817,
+      "learning_rate": 2.5646932669759427e-07,
+      "loss": 0.3062,
+      "step": 11019
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.790010764557456,
+      "learning_rate": 2.5605096793576646e-07,
+      "loss": 0.4332,
+      "step": 11020
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.15283502705698,
+      "learning_rate": 2.556329417041192e-07,
+      "loss": 0.5554,
+      "step": 11021
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.284318512087921,
+      "learning_rate": 2.552152480319553e-07,
+      "loss": 0.6837,
+      "step": 11022
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.454615123520454,
+      "learning_rate": 2.5479788694855343e-07,
+      "loss": 0.7399,
+      "step": 11023
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.495421924009523,
+      "learning_rate": 2.5438085848316916e-07,
+      "loss": 0.8502,
+      "step": 11024
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.745216338272797,
+      "learning_rate": 2.5396416266503245e-07,
+      "loss": 1.3025,
+      "step": 11025
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4879481153781136,
+      "learning_rate": 2.535477995233543e-07,
+      "loss": 0.1693,
+      "step": 11026
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.3729847640325024,
+      "learning_rate": 2.531317690873181e-07,
+      "loss": 0.889,
+      "step": 11027
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.671859726369521,
+      "learning_rate": 2.52716071386086e-07,
+      "loss": 1.1565,
+      "step": 11028
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.979680538540261,
+      "learning_rate": 2.5230070644879757e-07,
+      "loss": 0.5938,
+      "step": 11029
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.360130147685159,
+      "learning_rate": 2.518856743045672e-07,
+      "loss": 0.8711,
+      "step": 11030
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.407689265554586,
+      "learning_rate": 2.514709749824884e-07,
+      "loss": 1.0209,
+      "step": 11031
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.421968303096306,
+      "learning_rate": 2.510566085116273e-07,
+      "loss": 0.5024,
+      "step": 11032
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1873366866216424,
+      "learning_rate": 2.5064257492103064e-07,
+      "loss": 0.146,
+      "step": 11033
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.834142366561029,
+      "learning_rate": 2.502288742397202e-07,
+      "loss": 1.1664,
+      "step": 11034
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3021514892264503,
+      "learning_rate": 2.4981550649669504e-07,
+      "loss": 0.1722,
+      "step": 11035
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.065157901732676,
+      "learning_rate": 2.4940247172092924e-07,
+      "loss": 0.3421,
+      "step": 11036
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.163428169715771,
+      "learning_rate": 2.489897699413768e-07,
+      "loss": 0.4166,
+      "step": 11037
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.437325471175225,
+      "learning_rate": 2.4857740118696406e-07,
+      "loss": 0.3751,
+      "step": 11038
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.199057988048138,
+      "learning_rate": 2.481653654865973e-07,
+      "loss": 0.713,
+      "step": 11039
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.619639946601736,
+      "learning_rate": 2.47753662869159e-07,
+      "loss": 0.5802,
+      "step": 11040
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.1151249195378137,
+      "learning_rate": 2.473422933635067e-07,
+      "loss": 0.3998,
+      "step": 11041
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.940103140856555,
+      "learning_rate": 2.469312569984755e-07,
+      "loss": 0.8201,
+      "step": 11042
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.039052681644886,
+      "learning_rate": 2.4652055380287866e-07,
+      "loss": 0.6402,
+      "step": 11043
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.097990478563924,
+      "learning_rate": 2.46110183805503e-07,
+      "loss": 0.6293,
+      "step": 11044
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.976807621091998,
+      "learning_rate": 2.45700147035115e-07,
+      "loss": 0.791,
+      "step": 11045
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.9381474601172486,
+      "learning_rate": 2.4529044352045507e-07,
+      "loss": 0.5721,
+      "step": 11046
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.392734855345606,
+      "learning_rate": 2.448810732902429e-07,
+      "loss": 0.2277,
+      "step": 11047
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.073692288450157,
+      "learning_rate": 2.4447203637317396e-07,
+      "loss": 0.5008,
+      "step": 11048
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.0443640424281004,
+      "learning_rate": 2.440633327979186e-07,
+      "loss": 0.1341,
+      "step": 11049
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.94043086673372,
+      "learning_rate": 2.436549625931256e-07,
+      "loss": 0.487,
+      "step": 11050
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.6347325369876495,
+      "learning_rate": 2.432469257874198e-07,
+      "loss": 0.7156,
+      "step": 11051
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.1073434097379553,
+      "learning_rate": 2.4283922240940285e-07,
+      "loss": 0.3269,
+      "step": 11052
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.4455566731482126,
+      "learning_rate": 2.4243185248765347e-07,
+      "loss": 0.4302,
+      "step": 11053
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.0889595023649,
+      "learning_rate": 2.4202481605072715e-07,
+      "loss": 0.5646,
+      "step": 11054
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.6371914942050356,
+      "learning_rate": 2.4161811312715336e-07,
+      "loss": 0.3783,
+      "step": 11055
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.161157934117554,
+      "learning_rate": 2.41211743745442e-07,
+      "loss": 0.4261,
+      "step": 11056
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.3216016149663545,
+      "learning_rate": 2.40805707934077e-07,
+      "loss": 0.6887,
+      "step": 11057
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.645190440247131,
+      "learning_rate": 2.404000057215189e-07,
+      "loss": 0.9987,
+      "step": 11058
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.5485437389266705,
+      "learning_rate": 2.399946371362072e-07,
+      "loss": 0.9675,
+      "step": 11059
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.003720375727015,
+      "learning_rate": 2.3958960220655637e-07,
+      "loss": 0.814,
+      "step": 11060
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.089568719659635,
+      "learning_rate": 2.391849009609559e-07,
+      "loss": 0.7647,
+      "step": 11061
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.2458157362121085,
+      "learning_rate": 2.387805334277754e-07,
+      "loss": 1.0476,
+      "step": 11062
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.2327623500986915,
+      "learning_rate": 2.3837649963535825e-07,
+      "loss": 0.5227,
+      "step": 11063
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.340149049297397,
+      "learning_rate": 2.379727996120257e-07,
+      "loss": 0.7746,
+      "step": 11064
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.503326569977153,
+      "learning_rate": 2.3756943338607564e-07,
+      "loss": 1.101,
+      "step": 11065
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.111674690062485,
+      "learning_rate": 2.3716640098578326e-07,
+      "loss": 1.2175,
+      "step": 11066
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.474486226758203,
+      "learning_rate": 2.3676370243939706e-07,
+      "loss": 0.7457,
+      "step": 11067
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.7496111669335175,
+      "learning_rate": 2.363613377751456e-07,
+      "loss": 0.9001,
+      "step": 11068
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.2549787221233446,
+      "learning_rate": 2.3595930702123292e-07,
+      "loss": 0.7977,
+      "step": 11069
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.210072232193499,
+      "learning_rate": 2.35557610205841e-07,
+      "loss": 0.8362,
+      "step": 11070
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.482328728267237,
+      "learning_rate": 2.351562473571245e-07,
+      "loss": 0.3899,
+      "step": 11071
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.867913330107275,
+      "learning_rate": 2.3475521850321868e-07,
+      "loss": 0.971,
+      "step": 11072
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.6817365818980528,
+      "learning_rate": 2.3435452367223333e-07,
+      "loss": 0.5856,
+      "step": 11073
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.492208935289324,
+      "learning_rate": 2.3395416289225591e-07,
+      "loss": 1.3804,
+      "step": 11074
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.522649897360985,
+      "learning_rate": 2.3355413619134958e-07,
+      "loss": 0.9513,
+      "step": 11075
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.2002375390908013,
+      "learning_rate": 2.3315444359755468e-07,
+      "loss": 0.5962,
+      "step": 11076
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.699768339248374,
+      "learning_rate": 2.3275508513888822e-07,
+      "loss": 0.6855,
+      "step": 11077
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.2946389437483745,
+      "learning_rate": 2.3235606084334285e-07,
+      "loss": 0.5991,
+      "step": 11078
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.008303054571759,
+      "learning_rate": 2.319573707388889e-07,
+      "loss": 0.6266,
+      "step": 11079
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.0139603695193897,
+      "learning_rate": 2.3155901485347242e-07,
+      "loss": 0.5756,
+      "step": 11080
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.228017886089983,
+      "learning_rate": 2.3116099321501716e-07,
+      "loss": 0.441,
+      "step": 11081
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.21420696840283,
+      "learning_rate": 2.3076330585142138e-07,
+      "loss": 1.3786,
+      "step": 11082
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.7161093803831586,
+      "learning_rate": 2.303659527905633e-07,
+      "loss": 0.5781,
+      "step": 11083
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.732562707840122,
+      "learning_rate": 2.2996893406029396e-07,
+      "loss": 0.6167,
+      "step": 11084
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.9199633925383037,
+      "learning_rate": 2.2957224968844227e-07,
+      "loss": 0.7024,
+      "step": 11085
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.175139268446115,
+      "learning_rate": 2.291758997028165e-07,
+      "loss": 0.7179,
+      "step": 11086
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.684719713666393,
+      "learning_rate": 2.2877988413119613e-07,
+      "loss": 0.814,
+      "step": 11087
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.707293402429215,
+      "learning_rate": 2.2838420300134168e-07,
+      "loss": 1.1273,
+      "step": 11088
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.506806733915437,
+      "learning_rate": 2.2798885634098934e-07,
+      "loss": 1.2383,
+      "step": 11089
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.601802804472051,
+      "learning_rate": 2.2759384417784914e-07,
+      "loss": 0.8593,
+      "step": 11090
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.6701148366908787,
+      "learning_rate": 2.2719916653961117e-07,
+      "loss": 0.8575,
+      "step": 11091
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.647881541444441,
+      "learning_rate": 2.268048234539405e-07,
+      "loss": 0.5893,
+      "step": 11092
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.767204519975867,
+      "learning_rate": 2.264108149484784e-07,
+      "loss": 0.9492,
+      "step": 11093
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.01034330234706,
+      "learning_rate": 2.2601714105084438e-07,
+      "loss": 0.874,
+      "step": 11094
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.393774300890451,
+      "learning_rate": 2.256238017886314e-07,
+      "loss": 0.5693,
+      "step": 11095
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.0602238658907375,
+      "learning_rate": 2.2523079718941188e-07,
+      "loss": 0.3612,
+      "step": 11096
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.202968231086185,
+      "learning_rate": 2.2483812728073372e-07,
+      "loss": 0.9142,
+      "step": 11097
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.485032124470473,
+      "learning_rate": 2.2444579209012106e-07,
+      "loss": 0.4293,
+      "step": 11098
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.022784020013259,
+      "learning_rate": 2.2405379164507524e-07,
+      "loss": 0.6828,
+      "step": 11099
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.363452111618262,
+      "learning_rate": 2.2366212597307424e-07,
+      "loss": 0.618,
+      "step": 11100
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.0461915866057945,
+      "learning_rate": 2.2327079510157112e-07,
+      "loss": 0.7344,
+      "step": 11101
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.5256665808456873,
+      "learning_rate": 2.2287979905799672e-07,
+      "loss": 0.4726,
+      "step": 11102
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.923703005863906,
+      "learning_rate": 2.2248913786975857e-07,
+      "loss": 1.1763,
+      "step": 11103
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.6965162168198713,
+      "learning_rate": 2.2209881156423973e-07,
+      "loss": 0.3289,
+      "step": 11104
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.421436369491871,
+      "learning_rate": 2.2170882016880112e-07,
+      "loss": 0.7714,
+      "step": 11105
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.1651362991109435,
+      "learning_rate": 2.213191637107792e-07,
+      "loss": 0.8261,
+      "step": 11106
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.392052582605556,
+      "learning_rate": 2.2092984221748602e-07,
+      "loss": 0.7156,
+      "step": 11107
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.515651379382963,
+      "learning_rate": 2.205408557162131e-07,
+      "loss": 0.7332,
+      "step": 11108
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.996762719331602,
+      "learning_rate": 2.2015220423422523e-07,
+      "loss": 0.6411,
+      "step": 11109
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.481526049280716,
+      "learning_rate": 2.1976388779876623e-07,
+      "loss": 1.0722,
+      "step": 11110
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 6.452038705988483,
+      "learning_rate": 2.1937590643705542e-07,
+      "loss": 1.3209,
+      "step": 11111
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.40830111144125,
+      "learning_rate": 2.1898826017628772e-07,
+      "loss": 0.7558,
+      "step": 11112
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 6.21407611714033,
+      "learning_rate": 2.1860094904363637e-07,
+      "loss": 1.1922,
+      "step": 11113
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.211696547540112,
+      "learning_rate": 2.1821397306624915e-07,
+      "loss": 0.5482,
+      "step": 11114
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.10794839879072,
+      "learning_rate": 2.1782733227125264e-07,
+      "loss": 0.147,
+      "step": 11115
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.539108132473663,
+      "learning_rate": 2.174410266857474e-07,
+      "loss": 1.0284,
+      "step": 11116
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.4530422166033254,
+      "learning_rate": 2.1705505633681346e-07,
+      "loss": 0.4119,
+      "step": 11117
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.8801438058179403,
+      "learning_rate": 2.1666942125150358e-07,
+      "loss": 0.5742,
+      "step": 11118
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.218928014732488,
+      "learning_rate": 2.1628412145685119e-07,
+      "loss": 0.994,
+      "step": 11119
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.6833087536929625,
+      "learning_rate": 2.158991569798624e-07,
+      "loss": 1.198,
+      "step": 11120
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.325366267395837,
+      "learning_rate": 2.1551452784752236e-07,
+      "loss": 1.2089,
+      "step": 11121
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.651134494656022,
+      "learning_rate": 2.1513023408679168e-07,
+      "loss": 1.0882,
+      "step": 11122
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.374675834672166,
+      "learning_rate": 2.1474627572460826e-07,
+      "loss": 0.9013,
+      "step": 11123
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.39461346185137,
+      "learning_rate": 2.1436265278788448e-07,
+      "loss": 0.5845,
+      "step": 11124
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.167430852177149,
+      "learning_rate": 2.1397936530351214e-07,
+      "loss": 0.5196,
+      "step": 11125
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.3743656027738775,
+      "learning_rate": 2.13596413298357e-07,
+      "loss": 0.5903,
+      "step": 11126
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.548528392729398,
+      "learning_rate": 2.1321379679926314e-07,
+      "loss": 0.4787,
+      "step": 11127
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.7005002557902085,
+      "learning_rate": 2.1283151583305073e-07,
+      "loss": 0.4916,
+      "step": 11128
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.631365562756313,
+      "learning_rate": 2.1244957042651394e-07,
+      "loss": 0.5243,
+      "step": 11129
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.457370939567119,
+      "learning_rate": 2.1206796060642742e-07,
+      "loss": 0.8838,
+      "step": 11130
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.947617050297842,
+      "learning_rate": 2.1168668639953925e-07,
+      "loss": 0.6216,
+      "step": 11131
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.449892882699174,
+      "learning_rate": 2.1130574783257585e-07,
+      "loss": 1.1595,
+      "step": 11132
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.1387838929076355,
+      "learning_rate": 2.109251449322397e-07,
+      "loss": 0.5245,
+      "step": 11133
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.3414999678308424,
+      "learning_rate": 2.105448777252078e-07,
+      "loss": 0.3202,
+      "step": 11134
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.925262482104804,
+      "learning_rate": 2.1016494623813722e-07,
+      "loss": 0.9867,
+      "step": 11135
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.278064810880702,
+      "learning_rate": 2.0978535049765769e-07,
+      "loss": 0.5975,
+      "step": 11136
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.7913682482536117,
+      "learning_rate": 2.0940609053037796e-07,
+      "loss": 0.2869,
+      "step": 11137
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.942497206347499,
+      "learning_rate": 2.090271663628829e-07,
+      "loss": 0.8199,
+      "step": 11138
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.14434134047888,
+      "learning_rate": 2.086485780217329e-07,
+      "loss": 1.3063,
+      "step": 11139
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.727023327112773,
+      "learning_rate": 2.0827032553346615e-07,
+      "loss": 1.2092,
+      "step": 11140
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.28386093935721,
+      "learning_rate": 2.0789240892459485e-07,
+      "loss": 0.6101,
+      "step": 11141
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.052194647387861,
+      "learning_rate": 2.0751482822161106e-07,
+      "loss": 0.6826,
+      "step": 11142
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.5490516592790584,
+      "learning_rate": 2.0713758345098033e-07,
+      "loss": 0.3835,
+      "step": 11143
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.908189906576185,
+      "learning_rate": 2.0676067463914651e-07,
+      "loss": 0.5757,
+      "step": 11144
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.867167669543695,
+      "learning_rate": 2.0638410181252898e-07,
+      "loss": 0.5637,
+      "step": 11145
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.697488403153114,
+      "learning_rate": 2.06007864997525e-07,
+      "loss": 1.2564,
+      "step": 11146
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.991525616886259,
+      "learning_rate": 2.0563196422050568e-07,
+      "loss": 0.4381,
+      "step": 11147
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.7813495047843086,
+      "learning_rate": 2.0525639950781996e-07,
+      "loss": 0.3386,
+      "step": 11148
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.695648298957778,
+      "learning_rate": 2.0488117088579506e-07,
+      "loss": 0.5629,
+      "step": 11149
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.03623000154458,
+      "learning_rate": 2.0450627838073056e-07,
+      "loss": 0.6678,
+      "step": 11150
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.7482338355973073,
+      "learning_rate": 2.0413172201890653e-07,
+      "loss": 0.6516,
+      "step": 11151
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.864430039136069,
+      "learning_rate": 2.0375750182657695e-07,
+      "loss": 0.5004,
+      "step": 11152
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.4182944549418455,
+      "learning_rate": 2.0338361782997252e-07,
+      "loss": 0.7622,
+      "step": 11153
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.5796549918494818,
+      "learning_rate": 2.0301007005530226e-07,
+      "loss": 0.4822,
+      "step": 11154
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.101927255290817,
+      "learning_rate": 2.0263685852874915e-07,
+      "loss": 0.7614,
+      "step": 11155
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.936883196766509,
+      "learning_rate": 2.022639832764739e-07,
+      "loss": 1.0994,
+      "step": 11156
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.1251344484501056,
+      "learning_rate": 2.0189144432461449e-07,
+      "loss": 0.5361,
+      "step": 11157
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.220043319689753,
+      "learning_rate": 2.0151924169928228e-07,
+      "loss": 0.8652,
+      "step": 11158
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.371349797847683,
+      "learning_rate": 2.011473754265686e-07,
+      "loss": 0.8478,
+      "step": 11159
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.1755451624852777,
+      "learning_rate": 2.0077584553253927e-07,
+      "loss": 0.5387,
+      "step": 11160
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.700040156838395,
+      "learning_rate": 2.0040465204323678e-07,
+      "loss": 0.9394,
+      "step": 11161
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.7351776612595184,
+      "learning_rate": 2.0003379498468035e-07,
+      "loss": 0.836,
+      "step": 11162
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.607577896180465,
+      "learning_rate": 1.9966327438286582e-07,
+      "loss": 1.0263,
+      "step": 11163
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.4638615943766453,
+      "learning_rate": 1.992930902637641e-07,
+      "loss": 0.5572,
+      "step": 11164
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.442831959285367,
+      "learning_rate": 1.9892324265332442e-07,
+      "loss": 0.5543,
+      "step": 11165
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.233236165568079,
+      "learning_rate": 1.9855373157747048e-07,
+      "loss": 1.2052,
+      "step": 11166
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.152889625127049,
+      "learning_rate": 1.9818455706210438e-07,
+      "loss": 0.446,
+      "step": 11167
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.1104276927708336,
+      "learning_rate": 1.9781571913310317e-07,
+      "loss": 0.1884,
+      "step": 11168
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.0639604217808065,
+      "learning_rate": 1.9744721781632115e-07,
+      "loss": 1.0051,
+      "step": 11169
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.061353204589289,
+      "learning_rate": 1.9707905313758769e-07,
+      "loss": 0.8107,
+      "step": 11170
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.726630206028116,
+      "learning_rate": 1.96711225122711e-07,
+      "loss": 0.8195,
+      "step": 11171
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 6.104973615834729,
+      "learning_rate": 1.9634373379747274e-07,
+      "loss": 1.4508,
+      "step": 11172
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.120619482005494,
+      "learning_rate": 1.9597657918763335e-07,
+      "loss": 0.6486,
+      "step": 11173
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.191333913205158,
+      "learning_rate": 1.9560976131892894e-07,
+      "loss": 1.0356,
+      "step": 11174
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.905687977075724,
+      "learning_rate": 1.9524328021707118e-07,
+      "loss": 0.8676,
+      "step": 11175
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.586290688649648,
+      "learning_rate": 1.9487713590774948e-07,
+      "loss": 0.6796,
+      "step": 11176
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.000787043097494,
+      "learning_rate": 1.9451132841662778e-07,
+      "loss": 0.3549,
+      "step": 11177
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.2012772722664256,
+      "learning_rate": 1.9414585776934892e-07,
+      "loss": 1.1755,
+      "step": 11178
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.905691271358258,
+      "learning_rate": 1.937807239915307e-07,
+      "loss": 0.5006,
+      "step": 11179
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.424924940593239,
+      "learning_rate": 1.9341592710876656e-07,
+      "loss": 0.5779,
+      "step": 11180
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.047146689063044,
+      "learning_rate": 1.9305146714662826e-07,
+      "loss": 0.5132,
+      "step": 11181
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.7016882955252894,
+      "learning_rate": 1.926873441306615e-07,
+      "loss": 0.4587,
+      "step": 11182
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.9093433744808506,
+      "learning_rate": 1.9232355808639025e-07,
+      "loss": 0.6004,
+      "step": 11183
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 6.0015902471677585,
+      "learning_rate": 1.9196010903931417e-07,
+      "loss": 1.1875,
+      "step": 11184
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.9358852462762313,
+      "learning_rate": 1.9159699701491065e-07,
+      "loss": 0.5939,
+      "step": 11185
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.697605370531831,
+      "learning_rate": 1.9123422203863152e-07,
+      "loss": 0.5391,
+      "step": 11186
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.469171962086072,
+      "learning_rate": 1.908717841359048e-07,
+      "loss": 0.428,
+      "step": 11187
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.3907613046127594,
+      "learning_rate": 1.9050968333213683e-07,
+      "loss": 0.4528,
+      "step": 11188
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.5817391032192507,
+      "learning_rate": 1.9014791965270952e-07,
+      "loss": 0.4375,
+      "step": 11189
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.627772604588454,
+      "learning_rate": 1.897864931229798e-07,
+      "loss": 0.8307,
+      "step": 11190
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.789077798942602,
+      "learning_rate": 1.8942540376828355e-07,
+      "loss": 0.4359,
+      "step": 11191
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.7768245176577677,
+      "learning_rate": 1.890646516139305e-07,
+      "loss": 0.4011,
+      "step": 11192
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.210996217553082,
+      "learning_rate": 1.887042366852082e-07,
+      "loss": 0.9132,
+      "step": 11193
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.906255086628593,
+      "learning_rate": 1.8834415900737978e-07,
+      "loss": 0.5761,
+      "step": 11194
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.6180163634599065,
+      "learning_rate": 1.8798441860568616e-07,
+      "loss": 0.6036,
+      "step": 11195
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.0980136267430478,
+      "learning_rate": 1.8762501550534217e-07,
+      "loss": 0.0995,
+      "step": 11196
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.290132282228848,
+      "learning_rate": 1.87265949731541e-07,
+      "loss": 0.6914,
+      "step": 11197
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.136875746607067,
+      "learning_rate": 1.869072213094525e-07,
+      "loss": 0.7159,
+      "step": 11198
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.467381386165635,
+      "learning_rate": 1.865488302642199e-07,
+      "loss": 0.5064,
+      "step": 11199
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.4473827868656355,
+      "learning_rate": 1.8619077662096696e-07,
+      "loss": 0.4848,
+      "step": 11200
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.864204754691415,
+      "learning_rate": 1.8583306040479032e-07,
+      "loss": 0.8725,
+      "step": 11201
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.180014132625099,
+      "learning_rate": 1.854756816407649e-07,
+      "loss": 0.8338,
+      "step": 11202
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.8155883572348497,
+      "learning_rate": 1.851186403539418e-07,
+      "loss": 0.8488,
+      "step": 11203
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.525992468347426,
+      "learning_rate": 1.8476193656934704e-07,
+      "loss": 1.1812,
+      "step": 11204
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.7447516147126025,
+      "learning_rate": 1.844055703119846e-07,
+      "loss": 1.0543,
+      "step": 11205
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.17017342661183,
+      "learning_rate": 1.8404954160683443e-07,
+      "loss": 0.4521,
+      "step": 11206
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.946381787997414,
+      "learning_rate": 1.8369385047885156e-07,
+      "loss": 0.8788,
+      "step": 11207
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.172479909491882,
+      "learning_rate": 1.8333849695297e-07,
+      "loss": 0.7996,
+      "step": 11208
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.2283229855748674,
+      "learning_rate": 1.8298348105409757e-07,
+      "loss": 0.5261,
+      "step": 11209
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.733079662348577,
+      "learning_rate": 1.826288028071188e-07,
+      "loss": 0.8896,
+      "step": 11210
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.139927670971216,
+      "learning_rate": 1.8227446223689605e-07,
+      "loss": 0.9959,
+      "step": 11211
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.325872302072155,
+      "learning_rate": 1.819204593682672e-07,
+      "loss": 0.775,
+      "step": 11212
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.820136054276924,
+      "learning_rate": 1.8156679422604516e-07,
+      "loss": 0.816,
+      "step": 11213
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.790950829563678,
+      "learning_rate": 1.8121346683502183e-07,
+      "loss": 0.6849,
+      "step": 11214
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.5023638899812206,
+      "learning_rate": 1.808604772199618e-07,
+      "loss": 0.6212,
+      "step": 11215
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.478480719963517,
+      "learning_rate": 1.8050782540560974e-07,
+      "loss": 0.6308,
+      "step": 11216
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.570003184022221,
+      "learning_rate": 1.8015551141668474e-07,
+      "loss": 0.5625,
+      "step": 11217
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.8625066042343503,
+      "learning_rate": 1.7980353527788207e-07,
+      "loss": 0.8208,
+      "step": 11218
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.897763261574733,
+      "learning_rate": 1.7945189701387422e-07,
+      "loss": 0.8867,
+      "step": 11219
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.56730900511684,
+      "learning_rate": 1.7910059664930978e-07,
+      "loss": 0.6531,
+      "step": 11220
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.5322507412594497,
+      "learning_rate": 1.7874963420881242e-07,
+      "loss": 0.4657,
+      "step": 11221
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.7200820452997982,
+      "learning_rate": 1.7839900971698355e-07,
+      "loss": 0.7998,
+      "step": 11222
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.13435755876668,
+      "learning_rate": 1.7804872319840017e-07,
+      "loss": 0.7359,
+      "step": 11223
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.2755152186076573,
+      "learning_rate": 1.7769877467761655e-07,
+      "loss": 0.5649,
+      "step": 11224
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.6590296290393876,
+      "learning_rate": 1.7734916417916136e-07,
+      "loss": 0.2297,
+      "step": 11225
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.842949647874636,
+      "learning_rate": 1.7699989172754284e-07,
+      "loss": 0.7091,
+      "step": 11226
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.819659656122972,
+      "learning_rate": 1.7665095734724136e-07,
+      "loss": 0.8342,
+      "step": 11227
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.748007448179637,
+      "learning_rate": 1.763023610627168e-07,
+      "loss": 0.9059,
+      "step": 11228
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.9416185834126987,
+      "learning_rate": 1.7595410289840352e-07,
+      "loss": 0.5522,
+      "step": 11229
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.515053494653996,
+      "learning_rate": 1.756061828787131e-07,
+      "loss": 0.7281,
+      "step": 11230
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.4827523750750853,
+      "learning_rate": 1.7525860102803438e-07,
+      "loss": 0.321,
+      "step": 11231
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.901220700742775,
+      "learning_rate": 1.749113573707295e-07,
+      "loss": 0.874,
+      "step": 11232
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.183832641562579,
+      "learning_rate": 1.7456445193114014e-07,
+      "loss": 0.733,
+      "step": 11233
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.000577475774218,
+      "learning_rate": 1.742178847335818e-07,
+      "loss": 1.3244,
+      "step": 11234
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.83867868502028,
+      "learning_rate": 1.7387165580234789e-07,
+      "loss": 0.7298,
+      "step": 11235
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3992288505410486,
+      "learning_rate": 1.7352576516170784e-07,
+      "loss": 0.1665,
+      "step": 11236
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.221374384301655,
+      "learning_rate": 1.7318021283590724e-07,
+      "loss": 0.9748,
+      "step": 11237
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5943361673092038,
+      "learning_rate": 1.7283499884916677e-07,
+      "loss": 0.3766,
+      "step": 11238
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.624902684414091,
+      "learning_rate": 1.724901232256848e-07,
+      "loss": 1.0204,
+      "step": 11239
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.970727945735099,
+      "learning_rate": 1.7214558598963594e-07,
+      "loss": 0.7984,
+      "step": 11240
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.866006374575356,
+      "learning_rate": 1.7180138716517025e-07,
+      "loss": 0.6781,
+      "step": 11241
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.6831592115954566,
+      "learning_rate": 1.714575267764157e-07,
+      "loss": 0.6031,
+      "step": 11242
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.117946502570576,
+      "learning_rate": 1.7111400484747409e-07,
+      "loss": 0.756,
+      "step": 11243
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.028915158091609,
+      "learning_rate": 1.7077082140242617e-07,
+      "loss": 1.0491,
+      "step": 11244
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.528198610470189,
+      "learning_rate": 1.7042797646532606e-07,
+      "loss": 0.5895,
+      "step": 11245
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.3915221070718706,
+      "learning_rate": 1.700854700602067e-07,
+      "loss": 0.6763,
+      "step": 11246
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.9585794046649387,
+      "learning_rate": 1.697433022110756e-07,
+      "loss": 0.6661,
+      "step": 11247
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.0598631303407915,
+      "learning_rate": 1.69401472941918e-07,
+      "loss": 0.6134,
+      "step": 11248
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.833142545038157,
+      "learning_rate": 1.6905998227669474e-07,
+      "loss": 1.0471,
+      "step": 11249
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.217539587480963,
+      "learning_rate": 1.6871883023934166e-07,
+      "loss": 0.6563,
+      "step": 11250
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.960147177640882,
+      "learning_rate": 1.6837801685377298e-07,
+      "loss": 0.2802,
+      "step": 11251
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.584210166954635,
+      "learning_rate": 1.680375421438779e-07,
+      "loss": 0.9426,
+      "step": 11252
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.839656649825677,
+      "learning_rate": 1.6769740613352237e-07,
+      "loss": 0.4337,
+      "step": 11253
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.337294598224811,
+      "learning_rate": 1.6735760884654894e-07,
+      "loss": 0.5698,
+      "step": 11254
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.090740134791479,
+      "learning_rate": 1.6701815030677525e-07,
+      "loss": 0.9533,
+      "step": 11255
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.086760664500222,
+      "learning_rate": 1.666790305379956e-07,
+      "loss": 0.6691,
+      "step": 11256
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.063399156552799,
+      "learning_rate": 1.6634024956398098e-07,
+      "loss": 0.6545,
+      "step": 11257
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.305498616104555,
+      "learning_rate": 1.660018074084796e-07,
+      "loss": 0.628,
+      "step": 11258
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.9923655462342778,
+      "learning_rate": 1.656637040952136e-07,
+      "loss": 0.555,
+      "step": 11259
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.48463791219742,
+      "learning_rate": 1.653259396478829e-07,
+      "loss": 0.4043,
+      "step": 11260
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.1888494863078725,
+      "learning_rate": 1.6498851409016304e-07,
+      "loss": 0.3287,
+      "step": 11261
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.391289356990704,
+      "learning_rate": 1.6465142744570617e-07,
+      "loss": 0.9179,
+      "step": 11262
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.4228954424793883,
+      "learning_rate": 1.6431467973814064e-07,
+      "loss": 0.2831,
+      "step": 11263
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.3720171548713602,
+      "learning_rate": 1.6397827099107144e-07,
+      "loss": 0.5803,
+      "step": 11264
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.5179120097759555,
+      "learning_rate": 1.6364220122807862e-07,
+      "loss": 0.6053,
+      "step": 11265
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.620685463962939,
+      "learning_rate": 1.6330647047272052e-07,
+      "loss": 0.8843,
+      "step": 11266
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.078360315208572,
+      "learning_rate": 1.6297107874852836e-07,
+      "loss": 0.5724,
+      "step": 11267
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.353974912838414,
+      "learning_rate": 1.626360260790133e-07,
+      "loss": 1.1554,
+      "step": 11268
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.4036611618739023,
+      "learning_rate": 1.6230131248766046e-07,
+      "loss": 0.6715,
+      "step": 11269
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.58811237856854,
+      "learning_rate": 1.6196693799793162e-07,
+      "loss": 0.7753,
+      "step": 11270
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.064100711903729,
+      "learning_rate": 1.616329026332658e-07,
+      "loss": 0.7528,
+      "step": 11271
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.656043693642232,
+      "learning_rate": 1.6129920641707654e-07,
+      "loss": 1.1097,
+      "step": 11272
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.467223107675906,
+      "learning_rate": 1.6096584937275505e-07,
+      "loss": 0.9837,
+      "step": 11273
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.701685744944751,
+      "learning_rate": 1.6063283152366772e-07,
+      "loss": 0.7138,
+      "step": 11274
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.187258886409783,
+      "learning_rate": 1.6030015289315804e-07,
+      "loss": 1.4469,
+      "step": 11275
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.559429823627061,
+      "learning_rate": 1.5996781350454516e-07,
+      "loss": 0.3951,
+      "step": 11276
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.550707903735016,
+      "learning_rate": 1.5963581338112488e-07,
+      "loss": 0.3247,
+      "step": 11277
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.41070075623855,
+      "learning_rate": 1.5930415254616804e-07,
+      "loss": 0.9562,
+      "step": 11278
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.845711815923627,
+      "learning_rate": 1.5897283102292383e-07,
+      "loss": 0.6267,
+      "step": 11279
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.828948205716296,
+      "learning_rate": 1.5864184883461587e-07,
+      "loss": 0.6446,
+      "step": 11280
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.9044405188980735,
+      "learning_rate": 1.583112060044445e-07,
+      "loss": 0.6316,
+      "step": 11281
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.860168447675666,
+      "learning_rate": 1.5798090255558617e-07,
+      "loss": 1.086,
+      "step": 11282
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.8752415390305104,
+      "learning_rate": 1.5765093851119518e-07,
+      "loss": 0.7086,
+      "step": 11283
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.086774830872845,
+      "learning_rate": 1.5732131389439853e-07,
+      "loss": 0.9672,
+      "step": 11284
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.262079064086391,
+      "learning_rate": 1.5699202872830278e-07,
+      "loss": 0.6275,
+      "step": 11285
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.3456980321069585,
+      "learning_rate": 1.5666308303598833e-07,
+      "loss": 1.023,
+      "step": 11286
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.004805392257209,
+      "learning_rate": 1.5633447684051395e-07,
+      "loss": 0.6211,
+      "step": 11287
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.4750971894269,
+      "learning_rate": 1.5600621016491347e-07,
+      "loss": 0.6443,
+      "step": 11288
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.383512278224929,
+      "learning_rate": 1.5567828303219566e-07,
+      "loss": 0.8311,
+      "step": 11289
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.659952133960186,
+      "learning_rate": 1.5535069546534887e-07,
+      "loss": 0.8986,
+      "step": 11290
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6494431478194589,
+      "learning_rate": 1.5502344748733356e-07,
+      "loss": 0.3387,
+      "step": 11291
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.315047979396131,
+      "learning_rate": 1.5469653912108862e-07,
+      "loss": 1.0557,
+      "step": 11292
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.456468791457157,
+      "learning_rate": 1.5436997038953017e-07,
+      "loss": 1.0145,
+      "step": 11293
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.3868376907378184,
+      "learning_rate": 1.5404374131554877e-07,
+      "loss": 0.7768,
+      "step": 11294
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.3987429541251983,
+      "learning_rate": 1.537178519220106e-07,
+      "loss": 0.361,
+      "step": 11295
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.3747056365014325,
+      "learning_rate": 1.5339230223176016e-07,
+      "loss": 0.7798,
+      "step": 11296
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.5850680765242355,
+      "learning_rate": 1.5306709226761696e-07,
+      "loss": 0.9988,
+      "step": 11297
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9340330748299945,
+      "learning_rate": 1.5274222205237664e-07,
+      "loss": 0.2935,
+      "step": 11298
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.923023096241975,
+      "learning_rate": 1.5241769160881104e-07,
+      "loss": 0.8663,
+      "step": 11299
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.431124122208152,
+      "learning_rate": 1.520935009596697e-07,
+      "loss": 0.6967,
+      "step": 11300
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.5747143000254358,
+      "learning_rate": 1.5176965012767443e-07,
+      "loss": 0.7445,
+      "step": 11301
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.122107757528846,
+      "learning_rate": 1.5144613913552765e-07,
+      "loss": 0.7541,
+      "step": 11302
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.611120114866463,
+      "learning_rate": 1.511229680059051e-07,
+      "loss": 0.7325,
+      "step": 11303
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.315470373503458,
+      "learning_rate": 1.5080013676146032e-07,
+      "loss": 0.5616,
+      "step": 11304
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.124800589520938,
+      "learning_rate": 1.5047764542482245e-07,
+      "loss": 1.1372,
+      "step": 11305
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.052256498311689,
+      "learning_rate": 1.501554940185962e-07,
+      "loss": 0.7783,
+      "step": 11306
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.691173523402315,
+      "learning_rate": 1.49833682565364e-07,
+      "loss": 1.0757,
+      "step": 11307
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.582743442181179,
+      "learning_rate": 1.4951221108768177e-07,
+      "loss": 0.8932,
+      "step": 11308
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.08830725166257,
+      "learning_rate": 1.491910796080842e-07,
+      "loss": 1.0209,
+      "step": 11309
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.038731637184449,
+      "learning_rate": 1.488702881490811e-07,
+      "loss": 1.1201,
+      "step": 11310
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.890836273218682,
+      "learning_rate": 1.4854983673315948e-07,
+      "loss": 0.7534,
+      "step": 11311
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.190074553783258,
+      "learning_rate": 1.482297253827797e-07,
+      "loss": 0.4521,
+      "step": 11312
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.42871335374221,
+      "learning_rate": 1.4790995412038156e-07,
+      "loss": 0.5175,
+      "step": 11313
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.36379925264401,
+      "learning_rate": 1.4759052296837884e-07,
+      "loss": 0.6415,
+      "step": 11314
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.0429800495424937,
+      "learning_rate": 1.4727143194916304e-07,
+      "loss": 0.1184,
+      "step": 11315
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.913291195983895,
+      "learning_rate": 1.4695268108510075e-07,
+      "loss": 1.0162,
+      "step": 11316
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2645169967609522,
+      "learning_rate": 1.4663427039853574e-07,
+      "loss": 0.1846,
+      "step": 11317
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.820504603394702,
+      "learning_rate": 1.4631619991178568e-07,
+      "loss": 0.8034,
+      "step": 11318
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 6.132497291593508,
+      "learning_rate": 1.459984696471467e-07,
+      "loss": 1.0389,
+      "step": 11319
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.483890146892965,
+      "learning_rate": 1.4568107962688981e-07,
+      "loss": 0.6468,
+      "step": 11320
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8060879496132378,
+      "learning_rate": 1.4536402987326448e-07,
+      "loss": 0.3804,
+      "step": 11321
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.6983043868534735,
+      "learning_rate": 1.4504732040849234e-07,
+      "loss": 0.6215,
+      "step": 11322
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.8015248517561604,
+      "learning_rate": 1.4473095125477455e-07,
+      "loss": 0.5368,
+      "step": 11323
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.678951740513247,
+      "learning_rate": 1.4441492243428668e-07,
+      "loss": 0.4824,
+      "step": 11324
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.2553897273348658,
+      "learning_rate": 1.4409923396918102e-07,
+      "loss": 0.5046,
+      "step": 11325
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.679285071330336,
+      "learning_rate": 1.4378388588158598e-07,
+      "loss": 1.0004,
+      "step": 11326
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.874660054217869,
+      "learning_rate": 1.4346887819360667e-07,
+      "loss": 0.9498,
+      "step": 11327
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7774450413499203,
+      "learning_rate": 1.4315421092732262e-07,
+      "loss": 0.3228,
+      "step": 11328
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.245778087293293,
+      "learning_rate": 1.4283988410479233e-07,
+      "loss": 0.5838,
+      "step": 11329
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.677879261074963,
+      "learning_rate": 1.4252589774804705e-07,
+      "loss": 0.3299,
+      "step": 11330
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.396066471615801,
+      "learning_rate": 1.4221225187909692e-07,
+      "loss": 0.5794,
+      "step": 11331
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.993010315065761,
+      "learning_rate": 1.4189894651992665e-07,
+      "loss": 0.4717,
+      "step": 11332
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.70289365334675,
+      "learning_rate": 1.415859816924975e-07,
+      "loss": 0.8767,
+      "step": 11333
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.220641939449342,
+      "learning_rate": 1.412733574187475e-07,
+      "loss": 0.6153,
+      "step": 11334
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.365275761926734,
+      "learning_rate": 1.4096107372058966e-07,
+      "loss": 0.5104,
+      "step": 11335
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.9165392789832314,
+      "learning_rate": 1.406491306199137e-07,
+      "loss": 0.5385,
+      "step": 11336
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.824793724718174,
+      "learning_rate": 1.4033752813858603e-07,
+      "loss": 1.0546,
+      "step": 11337
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.780770875995635,
+      "learning_rate": 1.4002626629844862e-07,
+      "loss": 0.5101,
+      "step": 11338
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.5442777056394754,
+      "learning_rate": 1.3971534512131845e-07,
+      "loss": 0.253,
+      "step": 11339
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.6304779254960815,
+      "learning_rate": 1.3940476462899143e-07,
+      "loss": 0.815,
+      "step": 11340
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.6909485546180965,
+      "learning_rate": 1.3909452484323682e-07,
+      "loss": 0.5588,
+      "step": 11341
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.095782462527793,
+      "learning_rate": 1.3878462578580054e-07,
+      "loss": 0.6584,
+      "step": 11342
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.056814596706321,
+      "learning_rate": 1.3847506747840633e-07,
+      "loss": 0.7376,
+      "step": 11343
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.2531875872661526,
+      "learning_rate": 1.3816584994275238e-07,
+      "loss": 0.5637,
+      "step": 11344
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.926482984992203,
+      "learning_rate": 1.378569732005136e-07,
+      "loss": 0.9198,
+      "step": 11345
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.9457916584839743,
+      "learning_rate": 1.3754843727334156e-07,
+      "loss": 0.5255,
+      "step": 11346
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.727290180209972,
+      "learning_rate": 1.3724024218286224e-07,
+      "loss": 0.9668,
+      "step": 11347
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.224952767802356,
+      "learning_rate": 1.3693238795067897e-07,
+      "loss": 0.9269,
+      "step": 11348
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.965815540685722,
+      "learning_rate": 1.3662487459837114e-07,
+      "loss": 1.1514,
+      "step": 11349
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.049303234978047,
+      "learning_rate": 1.3631770214749374e-07,
+      "loss": 0.6255,
+      "step": 11350
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.981080124905114,
+      "learning_rate": 1.3601087061957953e-07,
+      "loss": 0.515,
+      "step": 11351
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.615161473763682,
+      "learning_rate": 1.3570438003613462e-07,
+      "loss": 0.8371,
+      "step": 11352
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.9398318733080595,
+      "learning_rate": 1.3539823041864354e-07,
+      "loss": 1.1855,
+      "step": 11353
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.9646347094591925,
+      "learning_rate": 1.350924217885652e-07,
+      "loss": 0.5909,
+      "step": 11354
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.4578430140887053,
+      "learning_rate": 1.3478695416733577e-07,
+      "loss": 0.5344,
+      "step": 11355
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.1623744287144895,
+      "learning_rate": 1.3448182757636763e-07,
+      "loss": 0.457,
+      "step": 11356
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.4389502492090305,
+      "learning_rate": 1.3417704203704919e-07,
+      "loss": 0.5576,
+      "step": 11357
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.7181312116356975,
+      "learning_rate": 1.3387259757074334e-07,
+      "loss": 0.8019,
+      "step": 11358
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.336195288213495,
+      "learning_rate": 1.335684941987908e-07,
+      "loss": 1.0168,
+      "step": 11359
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.387383483589837,
+      "learning_rate": 1.3326473194250844e-07,
+      "loss": 0.8317,
+      "step": 11360
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.455915677307732,
+      "learning_rate": 1.3296131082318808e-07,
+      "loss": 0.7971,
+      "step": 11361
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.0769604095285712,
+      "learning_rate": 1.3265823086209828e-07,
+      "loss": 0.1388,
+      "step": 11362
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.403256666098837,
+      "learning_rate": 1.3235549208048426e-07,
+      "loss": 0.5191,
+      "step": 11363
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.7045935963524754,
+      "learning_rate": 1.320530944995657e-07,
+      "loss": 0.5484,
+      "step": 11364
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.352491888105958,
+      "learning_rate": 1.3175103814053958e-07,
+      "loss": 0.7698,
+      "step": 11365
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.037086127689683,
+      "learning_rate": 1.3144932302457948e-07,
+      "loss": 0.3168,
+      "step": 11366
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.628660771907785,
+      "learning_rate": 1.3114794917283403e-07,
+      "loss": 0.8921,
+      "step": 11367
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.473819307131876,
+      "learning_rate": 1.3084691660642746e-07,
+      "loss": 0.7624,
+      "step": 11368
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.545272780439525,
+      "learning_rate": 1.3054622534646234e-07,
+      "loss": 0.7565,
+      "step": 11369
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.37639945632583,
+      "learning_rate": 1.3024587541401402e-07,
+      "loss": 0.8412,
+      "step": 11370
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.231299791289676,
+      "learning_rate": 1.2994586683013677e-07,
+      "loss": 0.5228,
+      "step": 11371
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.6247239231922217,
+      "learning_rate": 1.2964619961585934e-07,
+      "loss": 0.6691,
+      "step": 11372
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.335661847549364,
+      "learning_rate": 1.2934687379218769e-07,
+      "loss": 0.7996,
+      "step": 11373
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.194778115338029,
+      "learning_rate": 1.2904788938010392e-07,
+      "loss": 0.8221,
+      "step": 11374
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.091606578487979,
+      "learning_rate": 1.2874924640056352e-07,
+      "loss": 1.1225,
+      "step": 11375
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.524644004317661,
+      "learning_rate": 1.2845094487450193e-07,
+      "loss": 0.9753,
+      "step": 11376
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.182308703363743,
+      "learning_rate": 1.2815298482282746e-07,
+      "loss": 0.9673,
+      "step": 11377
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.513074224594544,
+      "learning_rate": 1.278553662664267e-07,
+      "loss": 1.0608,
+      "step": 11378
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.050545949195318,
+      "learning_rate": 1.2755808922616075e-07,
+      "loss": 0.2652,
+      "step": 11379
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.282441709700939,
+      "learning_rate": 1.2726115372286852e-07,
+      "loss": 0.6766,
+      "step": 11380
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.193279806048703,
+      "learning_rate": 1.2696455977736278e-07,
+      "loss": 1.0562,
+      "step": 11381
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.149582341907641,
+      "learning_rate": 1.266683074104341e-07,
+      "loss": 0.8034,
+      "step": 11382
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.8986898507073433,
+      "learning_rate": 1.2637239664284816e-07,
+      "loss": 0.6925,
+      "step": 11383
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 6.340439187009353,
+      "learning_rate": 1.2607682749534723e-07,
+      "loss": 0.9223,
+      "step": 11384
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.563330446190727,
+      "learning_rate": 1.2578159998864858e-07,
+      "loss": 1.0805,
+      "step": 11385
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.127752602449478,
+      "learning_rate": 1.2548671414344848e-07,
+      "loss": 0.997,
+      "step": 11386
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.686063809926292,
+      "learning_rate": 1.2519216998041483e-07,
+      "loss": 0.5383,
+      "step": 11387
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.025269407670612,
+      "learning_rate": 1.2489796752019446e-07,
+      "loss": 0.286,
+      "step": 11388
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.5751929466185035,
+      "learning_rate": 1.2460410678341027e-07,
+      "loss": 0.5423,
+      "step": 11389
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.870787922141563,
+      "learning_rate": 1.2431058779066086e-07,
+      "loss": 0.7163,
+      "step": 11390
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.8724761807165287,
+      "learning_rate": 1.2401741056252027e-07,
+      "loss": 0.4614,
+      "step": 11391
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.818761266802908,
+      "learning_rate": 1.2372457511953816e-07,
+      "loss": 0.5997,
+      "step": 11392
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.9234466514415307,
+      "learning_rate": 1.2343208148224205e-07,
+      "loss": 0.3844,
+      "step": 11393
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.6717868926200237,
+      "learning_rate": 1.2313992967113442e-07,
+      "loss": 0.4092,
+      "step": 11394
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.543066468599289,
+      "learning_rate": 1.228481197066933e-07,
+      "loss": 0.8673,
+      "step": 11395
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.16694214591345,
+      "learning_rate": 1.2255665160937346e-07,
+      "loss": 0.9808,
+      "step": 11396
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.468145315616178,
+      "learning_rate": 1.2226552539960578e-07,
+      "loss": 0.8954,
+      "step": 11397
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.62108876858957,
+      "learning_rate": 1.2197474109779672e-07,
+      "loss": 0.9778,
+      "step": 11398
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.3769233539605152,
+      "learning_rate": 1.2168429872432941e-07,
+      "loss": 0.5391,
+      "step": 11399
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.030085613305808,
+      "learning_rate": 1.21394198299562e-07,
+      "loss": 0.8454,
+      "step": 11400
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.005327340476278,
+      "learning_rate": 1.2110443984382936e-07,
+      "loss": 0.9019,
+      "step": 11401
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.066588979693674,
+      "learning_rate": 1.2081502337744245e-07,
+      "loss": 0.5903,
+      "step": 11402
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.59288686119471,
+      "learning_rate": 1.2052594892068892e-07,
+      "loss": 1.2391,
+      "step": 11403
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.5518887254683364,
+      "learning_rate": 1.202372164938298e-07,
+      "loss": 0.5666,
+      "step": 11404
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.788599245153345,
+      "learning_rate": 1.1994882611710502e-07,
+      "loss": 0.6974,
+      "step": 11405
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.8459791543792003,
+      "learning_rate": 1.1966077781073006e-07,
+      "loss": 0.8061,
+      "step": 11406
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.2033452368178885,
+      "learning_rate": 1.1937307159489486e-07,
+      "loss": 1.2126,
+      "step": 11407
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.139334628288193,
+      "learning_rate": 1.1908570748976666e-07,
+      "loss": 0.5134,
+      "step": 11408
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.922911461641456,
+      "learning_rate": 1.1879868551548935e-07,
+      "loss": 0.9709,
+      "step": 11409
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.413980700520785,
+      "learning_rate": 1.1851200569218069e-07,
+      "loss": 0.3545,
+      "step": 11410
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.643047829997474,
+      "learning_rate": 1.1822566803993574e-07,
+      "loss": 0.6406,
+      "step": 11411
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.2187121643032532,
+      "learning_rate": 1.1793967257882621e-07,
+      "loss": 0.3273,
+      "step": 11412
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.026110346408181,
+      "learning_rate": 1.1765401932889886e-07,
+      "loss": 0.279,
+      "step": 11413
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.0060657564552162,
+      "learning_rate": 1.1736870831017711e-07,
+      "loss": 0.6003,
+      "step": 11414
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.191971013035739,
+      "learning_rate": 1.1708373954265884e-07,
+      "loss": 0.9175,
+      "step": 11415
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.064330439717445,
+      "learning_rate": 1.1679911304632086e-07,
+      "loss": 0.9524,
+      "step": 11416
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.169885247510807,
+      "learning_rate": 1.1651482884111276e-07,
+      "loss": 0.8036,
+      "step": 11417
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.785332141244321,
+      "learning_rate": 1.1623088694696194e-07,
+      "loss": 1.19,
+      "step": 11418
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.594104038984991,
+      "learning_rate": 1.1594728738377192e-07,
+      "loss": 0.3819,
+      "step": 11419
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.872396175952463,
+      "learning_rate": 1.156640301714218e-07,
+      "loss": 0.5467,
+      "step": 11420
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.905342440994222,
+      "learning_rate": 1.1538111532976626e-07,
+      "loss": 0.7751,
+      "step": 11421
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.729105469841603,
+      "learning_rate": 1.1509854287863609e-07,
+      "loss": 0.5878,
+      "step": 11422
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.2242795366711547,
+      "learning_rate": 1.1481631283783934e-07,
+      "loss": 0.6588,
+      "step": 11423
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9178608870536981,
+      "learning_rate": 1.1453442522715852e-07,
+      "loss": 0.0959,
+      "step": 11424
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.8820096712829804,
+      "learning_rate": 1.1425288006635283e-07,
+      "loss": 0.4315,
+      "step": 11425
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.015369236993359,
+      "learning_rate": 1.1397167737515813e-07,
+      "loss": 0.6537,
+      "step": 11426
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.291026797185897,
+      "learning_rate": 1.1369081717328423e-07,
+      "loss": 0.458,
+      "step": 11427
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.7474722112109133,
+      "learning_rate": 1.1341029948041871e-07,
+      "loss": 0.6803,
+      "step": 11428
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.460239601393617,
+      "learning_rate": 1.1313012431622472e-07,
+      "loss": 0.7746,
+      "step": 11429
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.5032495408235245,
+      "learning_rate": 1.1285029170034156e-07,
+      "loss": 0.6912,
+      "step": 11430
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.1065549064346887,
+      "learning_rate": 1.125708016523841e-07,
+      "loss": 0.3045,
+      "step": 11431
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.201927306332262,
+      "learning_rate": 1.1229165419194332e-07,
+      "loss": 0.663,
+      "step": 11432
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.893359351425234,
+      "learning_rate": 1.1201284933858581e-07,
+      "loss": 0.6112,
+      "step": 11433
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.5457924134993117,
+      "learning_rate": 1.117343871118548e-07,
+      "loss": 0.3572,
+      "step": 11434
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.7618255828224187,
+      "learning_rate": 1.114562675312697e-07,
+      "loss": 0.7144,
+      "step": 11435
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.830967488476228,
+      "learning_rate": 1.1117849061632491e-07,
+      "loss": 0.668,
+      "step": 11436
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.0456683266792077,
+      "learning_rate": 1.1090105638649262e-07,
+      "loss": 0.6332,
+      "step": 11437
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.645145777245306,
+      "learning_rate": 1.1062396486121785e-07,
+      "loss": 0.7712,
+      "step": 11438
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.69933829079648,
+      "learning_rate": 1.1034721605992504e-07,
+      "loss": 1.1358,
+      "step": 11439
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.3023589956484822,
+      "learning_rate": 1.1007081000201203e-07,
+      "loss": 0.8938,
+      "step": 11440
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.714037306552824,
+      "learning_rate": 1.0979474670685441e-07,
+      "loss": 0.9522,
+      "step": 11441
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.8215571782472995,
+      "learning_rate": 1.0951902619380284e-07,
+      "loss": 0.8513,
+      "step": 11442
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.151668741371961,
+      "learning_rate": 1.0924364848218461e-07,
+      "loss": 1.1703,
+      "step": 11443
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.258142399241855,
+      "learning_rate": 1.0896861359130151e-07,
+      "loss": 0.992,
+      "step": 11444
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.842097233666027,
+      "learning_rate": 1.0869392154043256e-07,
+      "loss": 0.6473,
+      "step": 11445
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.282634427069878,
+      "learning_rate": 1.084195723488335e-07,
+      "loss": 1.0108,
+      "step": 11446
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.82352021553961,
+      "learning_rate": 1.0814556603573334e-07,
+      "loss": 0.611,
+      "step": 11447
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.9540453231454236,
+      "learning_rate": 1.0787190262034008e-07,
+      "loss": 0.6282,
+      "step": 11448
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.724155380279264,
+      "learning_rate": 1.075985821218356e-07,
+      "loss": 0.4692,
+      "step": 11449
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.823909371886717,
+      "learning_rate": 1.0732560455937902e-07,
+      "loss": 0.6942,
+      "step": 11450
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.864402366534795,
+      "learning_rate": 1.070529699521039e-07,
+      "loss": 1.1268,
+      "step": 11451
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.665735809875102,
+      "learning_rate": 1.0678067831912164e-07,
+      "loss": 0.933,
+      "step": 11452
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.071968052088595,
+      "learning_rate": 1.0650872967951864e-07,
+      "loss": 0.5692,
+      "step": 11453
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.402671875049084,
+      "learning_rate": 1.0623712405235742e-07,
+      "loss": 0.799,
+      "step": 11454
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.8339504762304664,
+      "learning_rate": 1.0596586145667553e-07,
+      "loss": 0.6069,
+      "step": 11455
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.554616223835177,
+      "learning_rate": 1.0569494191148832e-07,
+      "loss": 1.0409,
+      "step": 11456
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.355035502155545,
+      "learning_rate": 1.0542436543578505e-07,
+      "loss": 0.9743,
+      "step": 11457
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.8526725805057938,
+      "learning_rate": 1.0515413204853276e-07,
+      "loss": 0.2967,
+      "step": 11458
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.924321023359295,
+      "learning_rate": 1.0488424176867351e-07,
+      "loss": 0.5891,
+      "step": 11459
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.8122948981271754,
+      "learning_rate": 1.0461469461512552e-07,
+      "loss": 0.7662,
+      "step": 11460
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.9773162802992377,
+      "learning_rate": 1.0434549060678201e-07,
+      "loss": 0.8828,
+      "step": 11461
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.494031706823871,
+      "learning_rate": 1.0407662976251453e-07,
+      "loss": 1.0307,
+      "step": 11462
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.871334368520525,
+      "learning_rate": 1.0380811210116748e-07,
+      "loss": 0.2681,
+      "step": 11463
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.420993607912421,
+      "learning_rate": 1.0353993764156356e-07,
+      "loss": 0.5039,
+      "step": 11464
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.5214802423323555,
+      "learning_rate": 1.0327210640250051e-07,
+      "loss": 0.5811,
+      "step": 11465
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.1476833037383765,
+      "learning_rate": 1.0300461840275278e-07,
+      "loss": 0.5421,
+      "step": 11466
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.064162042829437,
+      "learning_rate": 1.0273747366106867e-07,
+      "loss": 0.5128,
+      "step": 11467
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.408249378397272,
+      "learning_rate": 1.024706721961749e-07,
+      "loss": 0.5601,
+      "step": 11468
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.885012696896012,
+      "learning_rate": 1.0220421402677261e-07,
+      "loss": 0.5723,
+      "step": 11469
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.891852518029071,
+      "learning_rate": 1.0193809917154018e-07,
+      "loss": 0.7822,
+      "step": 11470
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.022331987363963,
+      "learning_rate": 1.0167232764913104e-07,
+      "loss": 0.9398,
+      "step": 11471
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.880819542465347,
+      "learning_rate": 1.0140689947817305e-07,
+      "loss": 0.6669,
+      "step": 11472
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.514773015344065,
+      "learning_rate": 1.0114181467727302e-07,
+      "loss": 1.0464,
+      "step": 11473
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.204128200139511,
+      "learning_rate": 1.0087707326501218e-07,
+      "loss": 0.5684,
+      "step": 11474
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.7287083542166064,
+      "learning_rate": 1.0061267525994678e-07,
+      "loss": 0.5869,
+      "step": 11475
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.8170199043790864,
+      "learning_rate": 1.0034862068061147e-07,
+      "loss": 0.9989,
+      "step": 11476
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.513320291452303,
+      "learning_rate": 1.0008490954551419e-07,
+      "loss": 0.7419,
+      "step": 11477
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.078026930248082,
+      "learning_rate": 9.982154187314075e-08,
+      "loss": 0.1895,
+      "step": 11478
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.0083071016056,
+      "learning_rate": 9.95585176819508e-08,
+      "loss": 0.5666,
+      "step": 11479
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.944779680595329,
+      "learning_rate": 9.929583699038182e-08,
+      "loss": 0.9947,
+      "step": 11480
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.471368599698787,
+      "learning_rate": 9.90334998168474e-08,
+      "loss": 1.0571,
+      "step": 11481
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.1216598426591595,
+      "learning_rate": 9.877150617973507e-08,
+      "loss": 0.6191,
+      "step": 11482
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.424815174836256,
+      "learning_rate": 9.850985609741015e-08,
+      "loss": 1.1982,
+      "step": 11483
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.558857226478264,
+      "learning_rate": 9.824854958821295e-08,
+      "loss": 0.857,
+      "step": 11484
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.6881372715841327,
+      "learning_rate": 9.798758667045993e-08,
+      "loss": 0.5128,
+      "step": 11485
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.6042232762236632,
+      "learning_rate": 9.772696736244369e-08,
+      "loss": 0.3684,
+      "step": 11486
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.3849127166399193,
+      "learning_rate": 9.746669168243184e-08,
+      "loss": 0.4047,
+      "step": 11487
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.3419144586146996,
+      "learning_rate": 9.720675964866866e-08,
+      "loss": 0.491,
+      "step": 11488
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.7121433656782155,
+      "learning_rate": 9.69471712793757e-08,
+      "loss": 0.6647,
+      "step": 11489
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3240341254558812,
+      "learning_rate": 9.668792659274729e-08,
+      "loss": 0.1673,
+      "step": 11490
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.5540746486450585,
+      "learning_rate": 9.64290256069561e-08,
+      "loss": 0.9346,
+      "step": 11491
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.080260945098309,
+      "learning_rate": 9.617046834014987e-08,
+      "loss": 1.0139,
+      "step": 11492
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.7921434537664847,
+      "learning_rate": 9.591225481045186e-08,
+      "loss": 0.8898,
+      "step": 11493
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4892994629718517,
+      "learning_rate": 9.565438503596258e-08,
+      "loss": 0.3442,
+      "step": 11494
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.137716244372587,
+      "learning_rate": 9.539685903475704e-08,
+      "loss": 0.4682,
+      "step": 11495
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.5793078099482996,
+      "learning_rate": 9.513967682488634e-08,
+      "loss": 0.7187,
+      "step": 11496
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.9457444408370574,
+      "learning_rate": 9.488283842437829e-08,
+      "loss": 0.5016,
+      "step": 11497
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1224986446700729,
+      "learning_rate": 9.462634385123681e-08,
+      "loss": 0.1085,
+      "step": 11498
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.600689552970591,
+      "learning_rate": 9.437019312343976e-08,
+      "loss": 0.7454,
+      "step": 11499
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.659396098124453,
+      "learning_rate": 9.411438625894331e-08,
+      "loss": 0.5225,
+      "step": 11500
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.199140005369275,
+      "learning_rate": 9.385892327567759e-08,
+      "loss": 0.3812,
+      "step": 11501
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.064817368758036,
+      "learning_rate": 9.360380419154935e-08,
+      "loss": 0.8633,
+      "step": 11502
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.621385853262332,
+      "learning_rate": 9.33490290244421e-08,
+      "loss": 0.5887,
+      "step": 11503
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.058381904526087,
+      "learning_rate": 9.309459779221375e-08,
+      "loss": 0.6112,
+      "step": 11504
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.484172345562006,
+      "learning_rate": 9.284051051269949e-08,
+      "loss": 0.7339,
+      "step": 11505
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.215384163118494,
+      "learning_rate": 9.258676720371007e-08,
+      "loss": 1.1403,
+      "step": 11506
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.3297724747788098,
+      "learning_rate": 9.233336788303016e-08,
+      "loss": 0.5412,
+      "step": 11507
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2661471977938814,
+      "learning_rate": 9.208031256842332e-08,
+      "loss": 0.1712,
+      "step": 11508
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.154101220777355,
+      "learning_rate": 9.182760127762757e-08,
+      "loss": 0.7015,
+      "step": 11509
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.023772803963613,
+      "learning_rate": 9.1575234028356e-08,
+      "loss": 0.3518,
+      "step": 11510
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.673827508559293,
+      "learning_rate": 9.132321083829943e-08,
+      "loss": 0.2518,
+      "step": 11511
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.8791526657741016,
+      "learning_rate": 9.107153172512318e-08,
+      "loss": 0.5467,
+      "step": 11512
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.850501346120811,
+      "learning_rate": 9.082019670646813e-08,
+      "loss": 0.6633,
+      "step": 11513
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.531166946857835,
+      "learning_rate": 9.056920579995299e-08,
+      "loss": 0.5177,
+      "step": 11514
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.5638372089779664,
+      "learning_rate": 9.031855902317033e-08,
+      "loss": 0.4288,
+      "step": 11515
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.875937634908744,
+      "learning_rate": 9.006825639368944e-08,
+      "loss": 0.6471,
+      "step": 11516
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.9193632747619445,
+      "learning_rate": 8.981829792905628e-08,
+      "loss": 0.8261,
+      "step": 11517
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.9089533805539123,
+      "learning_rate": 8.956868364679128e-08,
+      "loss": 0.4271,
+      "step": 11518
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.383332574181771,
+      "learning_rate": 8.9319413564391e-08,
+      "loss": 0.324,
+      "step": 11519
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.6635614921681072,
+      "learning_rate": 8.907048769932813e-08,
+      "loss": 0.4885,
+      "step": 11520
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.417415217501595,
+      "learning_rate": 8.882190606905206e-08,
+      "loss": 0.5147,
+      "step": 11521
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.5786572783250215,
+      "learning_rate": 8.85736686909866e-08,
+      "loss": 0.4211,
+      "step": 11522
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.139069378559404,
+      "learning_rate": 8.832577558253285e-08,
+      "loss": 1.0832,
+      "step": 11523
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.3481033051612243,
+      "learning_rate": 8.807822676106637e-08,
+      "loss": 0.4845,
+      "step": 11524
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.4623507690569926,
+      "learning_rate": 8.783102224393992e-08,
+      "loss": 0.8474,
+      "step": 11525
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.897890325096617,
+      "learning_rate": 8.758416204848019e-08,
+      "loss": 0.5634,
+      "step": 11526
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.2330917403536095,
+      "learning_rate": 8.73376461919917e-08,
+      "loss": 0.9752,
+      "step": 11527
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.622719270232019,
+      "learning_rate": 8.709147469175449e-08,
+      "loss": 0.8331,
+      "step": 11528
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.632434371411331,
+      "learning_rate": 8.684564756502423e-08,
+      "loss": 0.8461,
+      "step": 11529
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.221314964205183,
+      "learning_rate": 8.660016482903156e-08,
+      "loss": 0.8041,
+      "step": 11530
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.096728943256637,
+      "learning_rate": 8.635502650098437e-08,
+      "loss": 0.458,
+      "step": 11531
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.889713743014272,
+      "learning_rate": 8.611023259806561e-08,
+      "loss": 0.3912,
+      "step": 11532
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.032860247914431,
+      "learning_rate": 8.586578313743377e-08,
+      "loss": 0.8403,
+      "step": 11533
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.898698151282602,
+      "learning_rate": 8.562167813622457e-08,
+      "loss": 0.4575,
+      "step": 11534
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.2422326007950137,
+      "learning_rate": 8.537791761154823e-08,
+      "loss": 0.4273,
+      "step": 11535
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.608676149082793,
+      "learning_rate": 8.513450158049109e-08,
+      "loss": 1.062,
+      "step": 11536
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.788955881010936,
+      "learning_rate": 8.489143006011613e-08,
+      "loss": 0.7896,
+      "step": 11537
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.747015684652336,
+      "learning_rate": 8.464870306746087e-08,
+      "loss": 0.9884,
+      "step": 11538
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.603052178222852,
+      "learning_rate": 8.440632061954057e-08,
+      "loss": 0.8629,
+      "step": 11539
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.844263423828449,
+      "learning_rate": 8.416428273334387e-08,
+      "loss": 0.8126,
+      "step": 11540
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.903695648370362,
+      "learning_rate": 8.392258942583775e-08,
+      "loss": 0.6307,
+      "step": 11541
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.425499726387105,
+      "learning_rate": 8.368124071396255e-08,
+      "loss": 0.6124,
+      "step": 11542
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.4550308257582176,
+      "learning_rate": 8.344023661463696e-08,
+      "loss": 0.3733,
+      "step": 11543
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.116310760936996,
+      "learning_rate": 8.319957714475357e-08,
+      "loss": 0.6365,
+      "step": 11544
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.5552693050682,
+      "learning_rate": 8.295926232118168e-08,
+      "loss": 0.6594,
+      "step": 11545
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.678799517409882,
+      "learning_rate": 8.271929216076724e-08,
+      "loss": 0.9425,
+      "step": 11546
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.454174229725478,
+      "learning_rate": 8.247966668032958e-08,
+      "loss": 0.7239,
+      "step": 11547
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.417164602813985,
+      "learning_rate": 8.224038589666639e-08,
+      "loss": 0.7123,
+      "step": 11548
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.485645869758046,
+      "learning_rate": 8.200144982654978e-08,
+      "loss": 1.0404,
+      "step": 11549
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.749260992056846,
+      "learning_rate": 8.17628584867286e-08,
+      "loss": 0.7257,
+      "step": 11550
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.735974713852315,
+      "learning_rate": 8.152461189392725e-08,
+      "loss": 0.947,
+      "step": 11551
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.691532152322852,
+      "learning_rate": 8.128671006484457e-08,
+      "loss": 0.4618,
+      "step": 11552
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.606720790843942,
+      "learning_rate": 8.104915301615723e-08,
+      "loss": 0.5511,
+      "step": 11553
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.203064780496871,
+      "learning_rate": 8.081194076451749e-08,
+      "loss": 0.9137,
+      "step": 11554
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.423709468178528,
+      "learning_rate": 8.057507332655201e-08,
+      "loss": 0.5172,
+      "step": 11555
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.544884911549885,
+      "learning_rate": 8.033855071886476e-08,
+      "loss": 0.7933,
+      "step": 11556
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.05296144706269,
+      "learning_rate": 8.01023729580347e-08,
+      "loss": 0.7496,
+      "step": 11557
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.168752164032706,
+      "learning_rate": 7.986654006061633e-08,
+      "loss": 0.6369,
+      "step": 11558
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.6099727114647076,
+      "learning_rate": 7.963105204314092e-08,
+      "loss": 0.9393,
+      "step": 11559
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.972524823079395,
+      "learning_rate": 7.939590892211523e-08,
+      "loss": 0.9579,
+      "step": 11560
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.4525684664562473,
+      "learning_rate": 7.916111071402222e-08,
+      "loss": 0.3673,
+      "step": 11561
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.365456772535376,
+      "learning_rate": 7.89266574353198e-08,
+      "loss": 0.5321,
+      "step": 11562
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.5464746563623795,
+      "learning_rate": 7.869254910244206e-08,
+      "loss": 0.5361,
+      "step": 11563
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.8067799548706502,
+      "learning_rate": 7.845878573179866e-08,
+      "loss": 0.7528,
+      "step": 11564
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.0718976468322223,
+      "learning_rate": 7.822536733977592e-08,
+      "loss": 0.4032,
+      "step": 11565
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.789649920090571,
+      "learning_rate": 7.799229394273522e-08,
+      "loss": 0.8454,
+      "step": 11566
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.544293359733326,
+      "learning_rate": 7.775956555701458e-08,
+      "loss": 0.8429,
+      "step": 11567
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.0705618784929,
+      "learning_rate": 7.752718219892597e-08,
+      "loss": 0.5423,
+      "step": 11568
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.604745100289837,
+      "learning_rate": 7.729514388476023e-08,
+      "loss": 0.7956,
+      "step": 11569
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.7360391468404823,
+      "learning_rate": 7.706345063078047e-08,
+      "loss": 0.8195,
+      "step": 11570
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.623807845515393,
+      "learning_rate": 7.683210245322869e-08,
+      "loss": 1.0799,
+      "step": 11571
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.2628999944535972,
+      "learning_rate": 7.660109936832027e-08,
+      "loss": 0.6359,
+      "step": 11572
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.573771335494906,
+      "learning_rate": 7.63704413922478e-08,
+      "loss": 0.7494,
+      "step": 11573
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.991034643761258,
+      "learning_rate": 7.614012854118058e-08,
+      "loss": 1.2426,
+      "step": 11574
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.16808363897129,
+      "learning_rate": 7.591016083126124e-08,
+      "loss": 1.2851,
+      "step": 11575
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.2343520940224546,
+      "learning_rate": 7.568053827860966e-08,
+      "loss": 0.5812,
+      "step": 11576
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.2577967730041224,
+      "learning_rate": 7.545126089932186e-08,
+      "loss": 0.7634,
+      "step": 11577
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.050302492327812,
+      "learning_rate": 7.522232870946889e-08,
+      "loss": 0.7246,
+      "step": 11578
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.43669760424669,
+      "learning_rate": 7.499374172509789e-08,
+      "loss": 0.8774,
+      "step": 11579
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.640794761815384,
+      "learning_rate": 7.476549996223215e-08,
+      "loss": 1.0219,
+      "step": 11580
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.831791310140802,
+      "learning_rate": 7.453760343686999e-08,
+      "loss": 0.774,
+      "step": 11581
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.7100391569805735,
+      "learning_rate": 7.431005216498588e-08,
+      "loss": 0.8476,
+      "step": 11582
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.514649034623249,
+      "learning_rate": 7.408284616253092e-08,
+      "loss": 0.6735,
+      "step": 11583
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.860514730999645,
+      "learning_rate": 7.385598544543015e-08,
+      "loss": 0.7812,
+      "step": 11584
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.5803739223711872,
+      "learning_rate": 7.362947002958698e-08,
+      "loss": 0.5444,
+      "step": 11585
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.2568546454751908,
+      "learning_rate": 7.340329993087813e-08,
+      "loss": 0.5813,
+      "step": 11586
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.295020809146325,
+      "learning_rate": 7.31774751651576e-08,
+      "loss": 0.9003,
+      "step": 11587
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.081893972177772,
+      "learning_rate": 7.295199574825384e-08,
+      "loss": 0.6387,
+      "step": 11588
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.705099452527816,
+      "learning_rate": 7.272686169597253e-08,
+      "loss": 0.8707,
+      "step": 11589
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.952450169969662,
+      "learning_rate": 7.250207302409496e-08,
+      "loss": 0.6748,
+      "step": 11590
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.098531567953999,
+      "learning_rate": 7.227762974837793e-08,
+      "loss": 0.6623,
+      "step": 11591
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.735206583392966,
+      "learning_rate": 7.205353188455277e-08,
+      "loss": 0.4195,
+      "step": 11592
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.2890640064475805,
+      "learning_rate": 7.182977944832859e-08,
+      "loss": 0.3627,
+      "step": 11593
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.931514297049116,
+      "learning_rate": 7.160637245538949e-08,
+      "loss": 0.9897,
+      "step": 11594
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.9622194419255887,
+      "learning_rate": 7.138331092139573e-08,
+      "loss": 0.506,
+      "step": 11595
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 6.133161143859423,
+      "learning_rate": 7.116059486198201e-08,
+      "loss": 1.2324,
+      "step": 11596
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.989076014377061,
+      "learning_rate": 7.093822429276032e-08,
+      "loss": 0.6505,
+      "step": 11597
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.539630457357993,
+      "learning_rate": 7.07161992293176e-08,
+      "loss": 0.8703,
+      "step": 11598
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.498339770297644,
+      "learning_rate": 7.049451968721699e-08,
+      "loss": 0.7966,
+      "step": 11599
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.006096866758366,
+      "learning_rate": 7.02731856819977e-08,
+      "loss": 0.6534,
+      "step": 11600
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.770977893654591,
+      "learning_rate": 7.0052197229174e-08,
+      "loss": 0.9353,
+      "step": 11601
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.9628131174003864,
+      "learning_rate": 6.983155434423517e-08,
+      "loss": 1.1756,
+      "step": 11602
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.7831215570400714,
+      "learning_rate": 6.961125704264937e-08,
+      "loss": 0.6618,
+      "step": 11603
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.3595549402144105,
+      "learning_rate": 6.93913053398565e-08,
+      "loss": 1.1731,
+      "step": 11604
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.6089979218198756,
+      "learning_rate": 6.917169925127476e-08,
+      "loss": 0.416,
+      "step": 11605
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.289736387729071,
+      "learning_rate": 6.895243879229852e-08,
+      "loss": 0.6777,
+      "step": 11606
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.1473913374249305,
+      "learning_rate": 6.873352397829603e-08,
+      "loss": 0.5736,
+      "step": 11607
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.712640401611258,
+      "learning_rate": 6.851495482461279e-08,
+      "loss": 0.8186,
+      "step": 11608
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.7269873024694307,
+      "learning_rate": 6.829673134656934e-08,
+      "loss": 0.8095,
+      "step": 11609
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.398859974456807,
+      "learning_rate": 6.807885355946176e-08,
+      "loss": 0.7604,
+      "step": 11610
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.468407590378123,
+      "learning_rate": 6.786132147856283e-08,
+      "loss": 0.7929,
+      "step": 11611
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.0647493326111022,
+      "learning_rate": 6.764413511912094e-08,
+      "loss": 0.1609,
+      "step": 11612
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.014610172645361,
+      "learning_rate": 6.742729449635888e-08,
+      "loss": 0.6214,
+      "step": 11613
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.6337823901937814,
+      "learning_rate": 6.721079962547783e-08,
+      "loss": 0.5395,
+      "step": 11614
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.595467025102142,
+      "learning_rate": 6.699465052165122e-08,
+      "loss": 0.6058,
+      "step": 11615
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.3282336878955827,
+      "learning_rate": 6.677884720003136e-08,
+      "loss": 0.2651,
+      "step": 11616
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.320522459061478,
+      "learning_rate": 6.656338967574505e-08,
+      "loss": 1.1039,
+      "step": 11617
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.061575871391227,
+      "learning_rate": 6.634827796389465e-08,
+      "loss": 1.2823,
+      "step": 11618
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.021015682318374,
+      "learning_rate": 6.613351207955865e-08,
+      "loss": 0.6708,
+      "step": 11619
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.8257923699654635,
+      "learning_rate": 6.591909203779167e-08,
+      "loss": 0.7875,
+      "step": 11620
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.35300776701248,
+      "learning_rate": 6.570501785362227e-08,
+      "loss": 0.7842,
+      "step": 11621
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.960849040274898,
+      "learning_rate": 6.54912895420573e-08,
+      "loss": 0.9143,
+      "step": 11622
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.417844368499578,
+      "learning_rate": 6.527790711807813e-08,
+      "loss": 0.9415,
+      "step": 11623
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.634517255764881,
+      "learning_rate": 6.506487059664113e-08,
+      "loss": 0.6651,
+      "step": 11624
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.3349906425666074,
+      "learning_rate": 6.485217999268045e-08,
+      "loss": 0.4669,
+      "step": 11625
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.088055091691627,
+      "learning_rate": 6.463983532110418e-08,
+      "loss": 0.7205,
+      "step": 11626
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 6.189227315555454,
+      "learning_rate": 6.442783659679596e-08,
+      "loss": 0.8922,
+      "step": 11627
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.062688209602089,
+      "learning_rate": 6.421618383461726e-08,
+      "loss": 0.792,
+      "step": 11628
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.0459100063850295,
+      "learning_rate": 6.400487704940284e-08,
+      "loss": 1.041,
+      "step": 11629
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.179069687860232,
+      "learning_rate": 6.379391625596587e-08,
+      "loss": 0.698,
+      "step": 11630
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.734301783196438,
+      "learning_rate": 6.358330146909231e-08,
+      "loss": 0.8751,
+      "step": 11631
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.6330821281949284,
+      "learning_rate": 6.337303270354644e-08,
+      "loss": 0.7615,
+      "step": 11632
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.1857711430187416,
+      "learning_rate": 6.31631099740665e-08,
+      "loss": 0.5557,
+      "step": 11633
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.2242637373736787,
+      "learning_rate": 6.295353329536736e-08,
+      "loss": 0.5299,
+      "step": 11634
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.627338941591819,
+      "learning_rate": 6.274430268213949e-08,
+      "loss": 0.5305,
+      "step": 11635
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.179602518667075,
+      "learning_rate": 6.253541814904895e-08,
+      "loss": 0.5425,
+      "step": 11636
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.042256472962217,
+      "learning_rate": 6.232687971073792e-08,
+      "loss": 0.5536,
+      "step": 11637
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.5314106009173982,
+      "learning_rate": 6.211868738182303e-08,
+      "loss": 0.3046,
+      "step": 11638
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.710831625897087,
+      "learning_rate": 6.191084117689871e-08,
+      "loss": 0.5028,
+      "step": 11639
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.887473282809646,
+      "learning_rate": 6.170334111053444e-08,
+      "loss": 0.3401,
+      "step": 11640
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.086127827375325,
+      "learning_rate": 6.149618719727358e-08,
+      "loss": 0.1172,
+      "step": 11641
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.805108073029417,
+      "learning_rate": 6.128937945163782e-08,
+      "loss": 0.2569,
+      "step": 11642
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.3531768646930997,
+      "learning_rate": 6.108291788812393e-08,
+      "loss": 0.3978,
+      "step": 11643
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.222728997768416,
+      "learning_rate": 6.087680252120254e-08,
+      "loss": 0.7847,
+      "step": 11644
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.961397698122342,
+      "learning_rate": 6.06710333653221e-08,
+      "loss": 1.2055,
+      "step": 11645
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.9685647046638475,
+      "learning_rate": 6.046561043490606e-08,
+      "loss": 0.9056,
+      "step": 11646
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.588238146585753,
+      "learning_rate": 6.026053374435404e-08,
+      "loss": 1.2207,
+      "step": 11647
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.733864306330916,
+      "learning_rate": 6.005580330804117e-08,
+      "loss": 0.4693,
+      "step": 11648
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.8971312445805486,
+      "learning_rate": 5.985141914031767e-08,
+      "loss": 0.2831,
+      "step": 11649
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.5183158215651575,
+      "learning_rate": 5.964738125550984e-08,
+      "loss": 0.6881,
+      "step": 11650
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.10165312450103,
+      "learning_rate": 5.944368966792014e-08,
+      "loss": 0.5499,
+      "step": 11651
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.395994324312094,
+      "learning_rate": 5.924034439182658e-08,
+      "loss": 1.0545,
+      "step": 11652
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.180941467475703,
+      "learning_rate": 5.903734544148221e-08,
+      "loss": 0.3215,
+      "step": 11653
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.1291064968405005,
+      "learning_rate": 5.8834692831117315e-08,
+      "loss": 0.5487,
+      "step": 11654
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.620358573775542,
+      "learning_rate": 5.863238657493608e-08,
+      "loss": 0.8608,
+      "step": 11655
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.6106870421584873,
+      "learning_rate": 5.8430426687119954e-08,
+      "loss": 0.9586,
+      "step": 11656
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.2684985178727715,
+      "learning_rate": 5.822881318182483e-08,
+      "loss": 0.6098,
+      "step": 11657
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.9087540299140833,
+      "learning_rate": 5.802754607318273e-08,
+      "loss": 0.5448,
+      "step": 11658
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.108416999885527,
+      "learning_rate": 5.7826625375302923e-08,
+      "loss": 0.7536,
+      "step": 11659
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.4963580682447812,
+      "learning_rate": 5.762605110226804e-08,
+      "loss": 0.5322,
+      "step": 11660
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.019122073312331,
+      "learning_rate": 5.742582326813795e-08,
+      "loss": 0.6686,
+      "step": 11661
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.5195503585473022,
+      "learning_rate": 5.722594188694697e-08,
+      "loss": 0.7019,
+      "step": 11662
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.1784774680724825,
+      "learning_rate": 5.702640697270667e-08,
+      "loss": 0.6628,
+      "step": 11663
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.7942647579416007,
+      "learning_rate": 5.682721853940365e-08,
+      "loss": 0.6642,
+      "step": 11664
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.492623859467021,
+      "learning_rate": 5.662837660099951e-08,
+      "loss": 0.5757,
+      "step": 11665
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.765473552833394,
+      "learning_rate": 5.642988117143311e-08,
+      "loss": 1.0545,
+      "step": 11666
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.114077177890035,
+      "learning_rate": 5.6231732264616644e-08,
+      "loss": 0.3435,
+      "step": 11667
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.7433908228118784,
+      "learning_rate": 5.603392989444068e-08,
+      "loss": 0.6281,
+      "step": 11668
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.0458702215608127,
+      "learning_rate": 5.583647407477022e-08,
+      "loss": 0.62,
+      "step": 11669
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.270318292611534,
+      "learning_rate": 5.5639364819445875e-08,
+      "loss": 0.8412,
+      "step": 11670
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.779845237480693,
+      "learning_rate": 5.5442602142284364e-08,
+      "loss": 0.4793,
+      "step": 11671
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.0113458173780154,
+      "learning_rate": 5.5246186057076875e-08,
+      "loss": 0.488,
+      "step": 11672
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.3714534494585875,
+      "learning_rate": 5.505011657759296e-08,
+      "loss": 0.5611,
+      "step": 11673
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.923472859924644,
+      "learning_rate": 5.4854393717574396e-08,
+      "loss": 1.0725,
+      "step": 11674
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.6161754209717114,
+      "learning_rate": 5.465901749074243e-08,
+      "loss": 0.5456,
+      "step": 11675
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.247407489721118,
+      "learning_rate": 5.446398791079055e-08,
+      "loss": 0.7884,
+      "step": 11676
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.7660720679954554,
+      "learning_rate": 5.42693049913906e-08,
+      "loss": 0.676,
+      "step": 11677
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.2160897355578095,
+      "learning_rate": 5.407496874618778e-08,
+      "loss": 1.0326,
+      "step": 11678
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.227074451795597,
+      "learning_rate": 5.388097918880564e-08,
+      "loss": 0.986,
+      "step": 11679
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.7225632670772963,
+      "learning_rate": 5.3687336332841065e-08,
+      "loss": 0.458,
+      "step": 11680
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.588303275939599,
+      "learning_rate": 5.3494040191867655e-08,
+      "loss": 1.1258,
+      "step": 11681
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.487444079482097,
+      "learning_rate": 5.3301090779434574e-08,
+      "loss": 0.3763,
+      "step": 11682
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.7603394226626325,
+      "learning_rate": 5.310848810906766e-08,
+      "loss": 0.626,
+      "step": 11683
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.578678130957751,
+      "learning_rate": 5.2916232194266116e-08,
+      "loss": 1.232,
+      "step": 11684
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.811411493552024,
+      "learning_rate": 5.272432304850694e-08,
+      "loss": 0.7073,
+      "step": 11685
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.6471063109711923,
+      "learning_rate": 5.253276068524216e-08,
+      "loss": 0.5287,
+      "step": 11686
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.6723646418446254,
+      "learning_rate": 5.2341545117899353e-08,
+      "loss": 0.9197,
+      "step": 11687
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.1312490454129955,
+      "learning_rate": 5.215067635988169e-08,
+      "loss": 0.8088,
+      "step": 11688
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.303647713081,
+      "learning_rate": 5.1960154424569587e-08,
+      "loss": 0.6594,
+      "step": 11689
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.3245269674291773,
+      "learning_rate": 5.176997932531569e-08,
+      "loss": 0.2501,
+      "step": 11690
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.024797051793267,
+      "learning_rate": 5.158015107545156e-08,
+      "loss": 0.8117,
+      "step": 11691
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.6393877702713913,
+      "learning_rate": 5.139066968828377e-08,
+      "loss": 0.5532,
+      "step": 11692
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.0030678265837416,
+      "learning_rate": 5.120153517709281e-08,
+      "loss": 0.5272,
+      "step": 11693
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3423220512615752,
+      "learning_rate": 5.101274755513808e-08,
+      "loss": 0.1617,
+      "step": 11694
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.293777898690641,
+      "learning_rate": 5.082430683565065e-08,
+      "loss": 0.9318,
+      "step": 11695
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.721454187574026,
+      "learning_rate": 5.063621303184163e-08,
+      "loss": 0.5775,
+      "step": 11696
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.8328649913876216,
+      "learning_rate": 5.044846615689325e-08,
+      "loss": 0.5656,
+      "step": 11697
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.067704682786649,
+      "learning_rate": 5.02610662239672e-08,
+      "loss": 0.6409,
+      "step": 11698
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.045261495652824,
+      "learning_rate": 5.0074013246199096e-08,
+      "loss": 0.6285,
+      "step": 11699
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.760271628485974,
+      "learning_rate": 4.9887307236700654e-08,
+      "loss": 0.3907,
+      "step": 11700
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.3602929536408532,
+      "learning_rate": 4.97009482085592e-08,
+      "loss": 0.6208,
+      "step": 11701
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.8847486528892277,
+      "learning_rate": 4.9514936174837047e-08,
+      "loss": 0.4227,
+      "step": 11702
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.5872727368371575,
+      "learning_rate": 4.932927114857322e-08,
+      "loss": 0.5445,
+      "step": 11703
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.348455025343582,
+      "learning_rate": 4.914395314278231e-08,
+      "loss": 0.5013,
+      "step": 11704
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.005260929206768,
+      "learning_rate": 4.8958982170453915e-08,
+      "loss": 0.6913,
+      "step": 11705
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.484792319652042,
+      "learning_rate": 4.8774358244554346e-08,
+      "loss": 0.4886,
+      "step": 11706
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.283131853262212,
+      "learning_rate": 4.859008137802379e-08,
+      "loss": 0.6121,
+      "step": 11707
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 6.101735286795017,
+      "learning_rate": 4.840615158378026e-08,
+      "loss": 1.1604,
+      "step": 11708
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.927126922279178,
+      "learning_rate": 4.8222568874716216e-08,
+      "loss": 0.6162,
+      "step": 11709
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.701134441874893,
+      "learning_rate": 4.803933326370025e-08,
+      "loss": 1.0281,
+      "step": 11710
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.122986484143584,
+      "learning_rate": 4.7856444763575424e-08,
+      "loss": 1.07,
+      "step": 11711
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.973057498389193,
+      "learning_rate": 4.7673903387162044e-08,
+      "loss": 1.0825,
+      "step": 11712
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 6.210457935937212,
+      "learning_rate": 4.749170914725543e-08,
+      "loss": 1.2721,
+      "step": 11713
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.6817758788117056,
+      "learning_rate": 4.730986205662702e-08,
+      "loss": 0.6935,
+      "step": 11714
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.7642625493959705,
+      "learning_rate": 4.712836212802274e-08,
+      "loss": 0.601,
+      "step": 11715
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.7641600812630798,
+      "learning_rate": 4.694720937416519e-08,
+      "loss": 0.6075,
+      "step": 11716
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.638167474691066,
+      "learning_rate": 4.67664038077531e-08,
+      "loss": 1.1841,
+      "step": 11717
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.255147060846687,
+      "learning_rate": 4.658594544145911e-08,
+      "loss": 0.641,
+      "step": 11718
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.736717762131721,
+      "learning_rate": 4.640583428793255e-08,
+      "loss": 0.5668,
+      "step": 11719
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.955164869337488,
+      "learning_rate": 4.622607035979942e-08,
+      "loss": 0.869,
+      "step": 11720
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.213184542465023,
+      "learning_rate": 4.6046653669659656e-08,
+      "loss": 0.5982,
+      "step": 11721
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.653497997895835,
+      "learning_rate": 4.5867584230089853e-08,
+      "loss": 0.6563,
+      "step": 11722
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.528597158347067,
+      "learning_rate": 4.568886205364276e-08,
+      "loss": 0.7737,
+      "step": 11723
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.3007991599918984,
+      "learning_rate": 4.551048715284445e-08,
+      "loss": 0.6716,
+      "step": 11724
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.869732963197292,
+      "learning_rate": 4.5332459540198825e-08,
+      "loss": 0.5945,
+      "step": 11725
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.985582699442896,
+      "learning_rate": 4.515477922818479e-08,
+      "loss": 0.6312,
+      "step": 11726
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.484632735394568,
+      "learning_rate": 4.497744622925793e-08,
+      "loss": 0.3245,
+      "step": 11727
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.567783803478978,
+      "learning_rate": 4.480046055584775e-08,
+      "loss": 0.9989,
+      "step": 11728
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.243537942446385,
+      "learning_rate": 4.462382222035988e-08,
+      "loss": 0.6423,
+      "step": 11729
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0891058539920115,
+      "learning_rate": 4.444753123517609e-08,
+      "loss": 0.1768,
+      "step": 11730
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.8314988858753942,
+      "learning_rate": 4.427158761265371e-08,
+      "loss": 0.5654,
+      "step": 11731
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.365254800376362,
+      "learning_rate": 4.4095991365125656e-08,
+      "loss": 0.6513,
+      "step": 11732
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.554928906459958,
+      "learning_rate": 4.3920742504900415e-08,
+      "loss": 1.4647,
+      "step": 11733
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.6148808353058794,
+      "learning_rate": 4.3745841044262606e-08,
+      "loss": 0.7798,
+      "step": 11734
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.058028006949942,
+      "learning_rate": 4.357128699547131e-08,
+      "loss": 0.825,
+      "step": 11735
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.582851708848683,
+      "learning_rate": 4.339708037076229e-08,
+      "loss": 1.3073,
+      "step": 11736
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.427789519056039,
+      "learning_rate": 4.3223221182346894e-08,
+      "loss": 0.7024,
+      "step": 11737
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.9148871543389543,
+      "learning_rate": 4.304970944241149e-08,
+      "loss": 0.6003,
+      "step": 11738
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.306396967498024,
+      "learning_rate": 4.2876545163118566e-08,
+      "loss": 0.8819,
+      "step": 11739
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.8677418803361903,
+      "learning_rate": 4.27037283566073e-08,
+      "loss": 0.9604,
+      "step": 11740
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.408385076006431,
+      "learning_rate": 4.253125903498967e-08,
+      "loss": 0.2947,
+      "step": 11741
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.746703569901941,
+      "learning_rate": 4.2359137210356e-08,
+      "loss": 0.769,
+      "step": 11742
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.6779079957653233,
+      "learning_rate": 4.21873628947711e-08,
+      "loss": 0.5306,
+      "step": 11743
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.4562236694218886,
+      "learning_rate": 4.2015936100275324e-08,
+      "loss": 0.9029,
+      "step": 11744
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.228704550219655,
+      "learning_rate": 4.184485683888573e-08,
+      "loss": 0.4066,
+      "step": 11745
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.424287972794271,
+      "learning_rate": 4.167412512259328e-08,
+      "loss": 1.0257,
+      "step": 11746
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.4697683143800604,
+      "learning_rate": 4.150374096336618e-08,
+      "loss": 0.435,
+      "step": 11747
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.440624052189856,
+      "learning_rate": 4.13337043731471e-08,
+      "loss": 0.6654,
+      "step": 11748
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.7522882705179637,
+      "learning_rate": 4.116401536385539e-08,
+      "loss": 0.3438,
+      "step": 11749
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.4657812590362957,
+      "learning_rate": 4.099467394738543e-08,
+      "loss": 0.4506,
+      "step": 11750
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.639754963246002,
+      "learning_rate": 4.0825680135606615e-08,
+      "loss": 0.6189,
+      "step": 11751
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.5670399308099583,
+      "learning_rate": 4.065703394036613e-08,
+      "loss": 0.2676,
+      "step": 11752
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.113158281399538,
+      "learning_rate": 4.048873537348341e-08,
+      "loss": 0.5267,
+      "step": 11753
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.481364819272389,
+      "learning_rate": 4.03207844467568e-08,
+      "loss": 0.9964,
+      "step": 11754
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.975723737572638,
+      "learning_rate": 4.01531811719591e-08,
+      "loss": 0.8212,
+      "step": 11755
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.07136840966976,
+      "learning_rate": 3.998592556083758e-08,
+      "loss": 0.7558,
+      "step": 11756
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.977434357886183,
+      "learning_rate": 3.981901762511675e-08,
+      "loss": 0.6283,
+      "step": 11757
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 6.1372750896314745,
+      "learning_rate": 3.9652457376496146e-08,
+      "loss": 0.9247,
+      "step": 11758
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.6768346894998594,
+      "learning_rate": 3.9486244826650865e-08,
+      "loss": 0.7823,
+      "step": 11759
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.162561476866158,
+      "learning_rate": 3.932037998723104e-08,
+      "loss": 1.0312,
+      "step": 11760
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.231638084068061,
+      "learning_rate": 3.915486286986403e-08,
+      "loss": 0.718,
+      "step": 11761
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.7585378451458094,
+      "learning_rate": 3.898969348615167e-08,
+      "loss": 0.2716,
+      "step": 11762
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.8033127495889065,
+      "learning_rate": 3.8824871847671366e-08,
+      "loss": 0.6429,
+      "step": 11763
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.7423527907068834,
+      "learning_rate": 3.8660397965976094e-08,
+      "loss": 0.4373,
+      "step": 11764
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.069044778469548,
+      "learning_rate": 3.849627185259497e-08,
+      "loss": 0.6415,
+      "step": 11765
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.315625982864917,
+      "learning_rate": 3.833249351903268e-08,
+      "loss": 0.6324,
+      "step": 11766
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.7279005421132565,
+      "learning_rate": 3.816906297676948e-08,
+      "loss": 0.3445,
+      "step": 11767
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.50701363886008,
+      "learning_rate": 3.800598023726121e-08,
+      "loss": 0.6086,
+      "step": 11768
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.296092547140078,
+      "learning_rate": 3.784324531193928e-08,
+      "loss": 0.7383,
+      "step": 11769
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.23251361838104,
+      "learning_rate": 3.768085821221013e-08,
+      "loss": 0.9903,
+      "step": 11770
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.213264983725306,
+      "learning_rate": 3.7518818949456305e-08,
+      "loss": 1.123,
+      "step": 11771
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.368028614421407,
+      "learning_rate": 3.735712753503706e-08,
+      "loss": 0.9072,
+      "step": 11772
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8344672689359063,
+      "learning_rate": 3.719578398028556e-08,
+      "loss": 0.2159,
+      "step": 11773
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.012385244807552,
+      "learning_rate": 3.703478829651164e-08,
+      "loss": 0.5381,
+      "step": 11774
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.864676780705077,
+      "learning_rate": 3.687414049500015e-08,
+      "loss": 0.9533,
+      "step": 11775
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.4633608720572053,
+      "learning_rate": 3.671384058701155e-08,
+      "loss": 0.6767,
+      "step": 11776
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.1948794508399985,
+      "learning_rate": 3.6553888583782395e-08,
+      "loss": 0.7211,
+      "step": 11777
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.889285302320259,
+      "learning_rate": 3.639428449652427e-08,
+      "loss": 0.5263,
+      "step": 11778
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.5537407969712893,
+      "learning_rate": 3.6235028336426004e-08,
+      "loss": 0.2516,
+      "step": 11779
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.2204192477410345,
+      "learning_rate": 3.607612011464923e-08,
+      "loss": 0.9395,
+      "step": 11780
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.598581066594799,
+      "learning_rate": 3.591755984233391e-08,
+      "loss": 0.8617,
+      "step": 11781
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.756522623237823,
+      "learning_rate": 3.57593475305934e-08,
+      "loss": 0.9931,
+      "step": 11782
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.064404256654831,
+      "learning_rate": 3.560148319051826e-08,
+      "loss": 0.97,
+      "step": 11783
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.465086785120551,
+      "learning_rate": 3.5443966833174084e-08,
+      "loss": 0.3152,
+      "step": 11784
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.096692457761413,
+      "learning_rate": 3.528679846960148e-08,
+      "loss": 0.4706,
+      "step": 11785
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.195789170395069,
+      "learning_rate": 3.5129978110818866e-08,
+      "loss": 0.6458,
+      "step": 11786
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.942434317814537,
+      "learning_rate": 3.497350576781688e-08,
+      "loss": 1.2318,
+      "step": 11787
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.358805952965998,
+      "learning_rate": 3.4817381451564546e-08,
+      "loss": 0.2734,
+      "step": 11788
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.925661326114637,
+      "learning_rate": 3.466160517300532e-08,
+      "loss": 0.7324,
+      "step": 11789
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.084076068790197,
+      "learning_rate": 3.450617694305825e-08,
+      "loss": 0.5408,
+      "step": 11790
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.839331507480073,
+      "learning_rate": 3.4351096772617945e-08,
+      "loss": 0.9241,
+      "step": 11791
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.8008598230611255,
+      "learning_rate": 3.4196364672555715e-08,
+      "loss": 0.6092,
+      "step": 11792
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.8795245977568715,
+      "learning_rate": 3.4041980653716777e-08,
+      "loss": 0.4539,
+      "step": 11793
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.536553255378163,
+      "learning_rate": 3.388794472692303e-08,
+      "loss": 0.4606,
+      "step": 11794
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.871365365644232,
+      "learning_rate": 3.3734256902971385e-08,
+      "loss": 0.7989,
+      "step": 11795
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.1551404757420975,
+      "learning_rate": 3.3580917192635454e-08,
+      "loss": 1.1861,
+      "step": 11796
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.287953974934314,
+      "learning_rate": 3.3427925606663856e-08,
+      "loss": 0.8849,
+      "step": 11797
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.5919220106556877,
+      "learning_rate": 3.327528215577913e-08,
+      "loss": 0.5035,
+      "step": 11798
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.393547504328156,
+      "learning_rate": 3.3122986850682713e-08,
+      "loss": 0.8699,
+      "step": 11799
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.847976191247303,
+      "learning_rate": 3.297103970204829e-08,
+      "loss": 1.0686,
+      "step": 11800
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.3101053476366795,
+      "learning_rate": 3.2819440720527894e-08,
+      "loss": 1.2983,
+      "step": 11801
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.348969443173985,
+      "learning_rate": 3.266818991674692e-08,
+      "loss": 1.0874,
+      "step": 11802
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.313600503412035,
+      "learning_rate": 3.251728730130854e-08,
+      "loss": 1.148,
+      "step": 11803
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.772193377266182,
+      "learning_rate": 3.236673288478931e-08,
+      "loss": 0.8395,
+      "step": 11804
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.48591309072955,
+      "learning_rate": 3.221652667774355e-08,
+      "loss": 1.0058,
+      "step": 11805
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.960003907617976,
+      "learning_rate": 3.2066668690698967e-08,
+      "loss": 0.608,
+      "step": 11806
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.563836151855298,
+      "learning_rate": 3.191715893415992e-08,
+      "loss": 0.1871,
+      "step": 11807
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.8170741667588723,
+      "learning_rate": 3.1767997418607474e-08,
+      "loss": 0.5567,
+      "step": 11808
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.9128601595015318,
+      "learning_rate": 3.1619184154496605e-08,
+      "loss": 0.5273,
+      "step": 11809
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.4878065354273864,
+      "learning_rate": 3.1470719152257856e-08,
+      "loss": 0.5814,
+      "step": 11810
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.383885581930084,
+      "learning_rate": 3.132260242229901e-08,
+      "loss": 0.6354,
+      "step": 11811
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.9485465857088755,
+      "learning_rate": 3.117483397500232e-08,
+      "loss": 0.2627,
+      "step": 11812
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.9127081237186885,
+      "learning_rate": 3.1027413820724494e-08,
+      "loss": 0.8579,
+      "step": 11813
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7947974661861645,
+      "learning_rate": 3.0880341969801164e-08,
+      "loss": 0.2722,
+      "step": 11814
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.421281975902339,
+      "learning_rate": 3.073361843253908e-08,
+      "loss": 0.7727,
+      "step": 11815
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.7079889954384253,
+      "learning_rate": 3.058724321922446e-08,
+      "loss": 0.5543,
+      "step": 11816
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.573949392109921,
+      "learning_rate": 3.044121634011687e-08,
+      "loss": 0.893,
+      "step": 11817
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.674279380148187,
+      "learning_rate": 3.029553780545258e-08,
+      "loss": 0.9186,
+      "step": 11818
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.3034868267727884,
+      "learning_rate": 3.015020762544341e-08,
+      "loss": 0.4583,
+      "step": 11819
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.894623775380931,
+      "learning_rate": 3.00052258102751e-08,
+      "loss": 1.0015,
+      "step": 11820
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.603660981984285,
+      "learning_rate": 2.9860592370111186e-08,
+      "loss": 0.361,
+      "step": 11821
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.915259342684215,
+      "learning_rate": 2.9716307315089677e-08,
+      "loss": 0.7958,
+      "step": 11822
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.8168026583412065,
+      "learning_rate": 2.9572370655324146e-08,
+      "loss": 1.1639,
+      "step": 11823
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.374519272465158,
+      "learning_rate": 2.94287824009043e-08,
+      "loss": 0.89,
+      "step": 11824
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.663285053546188,
+      "learning_rate": 2.928554256189431e-08,
+      "loss": 0.5214,
+      "step": 11825
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.9762625834375305,
+      "learning_rate": 2.914265114833614e-08,
+      "loss": 1.0695,
+      "step": 11826
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.332668720167278,
+      "learning_rate": 2.9000108170244013e-08,
+      "loss": 1.2487,
+      "step": 11827
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.408214880538728,
+      "learning_rate": 2.8857913637610478e-08,
+      "loss": 0.7891,
+      "step": 11828
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.41942697624612,
+      "learning_rate": 2.8716067560403128e-08,
+      "loss": 1.0684,
+      "step": 11829
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.8164144094370407,
+      "learning_rate": 2.8574569948564002e-08,
+      "loss": 0.4722,
+      "step": 11830
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.132366332899768,
+      "learning_rate": 2.8433420812011836e-08,
+      "loss": 1.1149,
+      "step": 11831
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.1411737141967455,
+      "learning_rate": 2.829262016064094e-08,
+      "loss": 1.1777,
+      "step": 11832
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.309970324128132,
+      "learning_rate": 2.815216800432008e-08,
+      "loss": 0.6603,
+      "step": 11833
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.351849238006262,
+      "learning_rate": 2.8012064352894718e-08,
+      "loss": 0.8311,
+      "step": 11834
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.1894191347071508,
+      "learning_rate": 2.7872309216185333e-08,
+      "loss": 0.6127,
+      "step": 11835
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.6702761891664815,
+      "learning_rate": 2.7732902603988532e-08,
+      "loss": 0.5173,
+      "step": 11836
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.9034952122764306,
+      "learning_rate": 2.7593844526075943e-08,
+      "loss": 0.349,
+      "step": 11837
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.847761447024961,
+      "learning_rate": 2.7455134992194767e-08,
+      "loss": 0.7418,
+      "step": 11838
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.856028845939515,
+      "learning_rate": 2.7316774012068337e-08,
+      "loss": 0.6664,
+      "step": 11839
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.4415462519630657,
+      "learning_rate": 2.7178761595394455e-08,
+      "loss": 0.6076,
+      "step": 11840
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.7737666124134774,
+      "learning_rate": 2.70410977518476e-08,
+      "loss": 0.4558,
+      "step": 11841
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.920737183462944,
+      "learning_rate": 2.6903782491077278e-08,
+      "loss": 0.7898,
+      "step": 11842
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.068358658790362,
+      "learning_rate": 2.6766815822709124e-08,
+      "loss": 0.7369,
+      "step": 11843
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.147692732508822,
+      "learning_rate": 2.663019775634379e-08,
+      "loss": 0.7402,
+      "step": 11844
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.802750877458447,
+      "learning_rate": 2.6493928301556947e-08,
+      "loss": 0.5201,
+      "step": 11845
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.292352851159922,
+      "learning_rate": 2.635800746790096e-08,
+      "loss": 1.2759,
+      "step": 11846
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.2793226440870615,
+      "learning_rate": 2.62224352649032e-08,
+      "loss": 0.5542,
+      "step": 11847
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.7325544908920576,
+      "learning_rate": 2.6087211702067184e-08,
+      "loss": 0.465,
+      "step": 11848
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.104911052916324,
+      "learning_rate": 2.5952336788871434e-08,
+      "loss": 0.758,
+      "step": 11849
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.1921776183039756,
+      "learning_rate": 2.581781053476895e-08,
+      "loss": 0.3688,
+      "step": 11850
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.523292465051399,
+      "learning_rate": 2.568363294919052e-08,
+      "loss": 1.1649,
+      "step": 11851
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.707201016531321,
+      "learning_rate": 2.5549804041541392e-08,
+      "loss": 0.3458,
+      "step": 11852
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.421203294947658,
+      "learning_rate": 2.5416323821201848e-08,
+      "loss": 0.9934,
+      "step": 11853
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.8331440448136003,
+      "learning_rate": 2.5283192297528846e-08,
+      "loss": 0.5116,
+      "step": 11854
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.4544071763317814,
+      "learning_rate": 2.5150409479853255e-08,
+      "loss": 0.277,
+      "step": 11855
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.6384962008616775,
+      "learning_rate": 2.5017975377483738e-08,
+      "loss": 0.7894,
+      "step": 11856
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.104411309353446,
+      "learning_rate": 2.4885889999703426e-08,
+      "loss": 0.5235,
+      "step": 11857
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.8549440430266007,
+      "learning_rate": 2.4754153355769915e-08,
+      "loss": 0.5376,
+      "step": 11858
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.0554960882424425,
+      "learning_rate": 2.462276545491804e-08,
+      "loss": 0.8901,
+      "step": 11859
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.691121017551824,
+      "learning_rate": 2.4491726306357656e-08,
+      "loss": 0.9459,
+      "step": 11860
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.946935112710814,
+      "learning_rate": 2.4361035919273635e-08,
+      "loss": 0.8441,
+      "step": 11861
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.272086858888599,
+      "learning_rate": 2.423069430282643e-08,
+      "loss": 1.0533,
+      "step": 11862
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.4291959785483113,
+      "learning_rate": 2.4100701466153177e-08,
+      "loss": 0.3649,
+      "step": 11863
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.7287457594289273,
+      "learning_rate": 2.397105741836603e-08,
+      "loss": 0.6391,
+      "step": 11864
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.531476649737334,
+      "learning_rate": 2.384176216855161e-08,
+      "loss": 0.5818,
+      "step": 11865
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.020496239571027,
+      "learning_rate": 2.3712815725773774e-08,
+      "loss": 0.4824,
+      "step": 11866
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.094119632365212,
+      "learning_rate": 2.3584218099070298e-08,
+      "loss": 0.8664,
+      "step": 11867
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.9308390645019133,
+      "learning_rate": 2.345596929745564e-08,
+      "loss": 0.5896,
+      "step": 11868
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.0067100581001664,
+      "learning_rate": 2.3328069329919824e-08,
+      "loss": 0.7413,
+      "step": 11869
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.6962664252409225,
+      "learning_rate": 2.3200518205427346e-08,
+      "loss": 0.6535,
+      "step": 11870
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.581880556701163,
+      "learning_rate": 2.307331593291995e-08,
+      "loss": 0.9012,
+      "step": 11871
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.5539149399642733,
+      "learning_rate": 2.2946462521313274e-08,
+      "loss": 0.6333,
+      "step": 11872
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.5346921718157773,
+      "learning_rate": 2.2819957979499098e-08,
+      "loss": 0.6966,
+      "step": 11873
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.8985294569505085,
+      "learning_rate": 2.2693802316345327e-08,
+      "loss": 1.0231,
+      "step": 11874
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.3983434931825447,
+      "learning_rate": 2.2567995540694888e-08,
+      "loss": 0.3441,
+      "step": 11875
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.1909594209394427,
+      "learning_rate": 2.2442537661365727e-08,
+      "loss": 0.6087,
+      "step": 11876
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.214137835453607,
+      "learning_rate": 2.231742868715303e-08,
+      "loss": 0.7074,
+      "step": 11877
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.373568865436762,
+      "learning_rate": 2.2192668626824788e-08,
+      "loss": 0.9279,
+      "step": 11878
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.858297833004091,
+      "learning_rate": 2.206825748912733e-08,
+      "loss": 0.5545,
+      "step": 11879
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.673507367762111,
+      "learning_rate": 2.194419528278091e-08,
+      "loss": 0.8805,
+      "step": 11880
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 6.24616288686617,
+      "learning_rate": 2.1820482016481902e-08,
+      "loss": 0.9465,
+      "step": 11881
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.86202337498881,
+      "learning_rate": 2.1697117698901704e-08,
+      "loss": 1.1141,
+      "step": 11882
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.3885582585985783,
+      "learning_rate": 2.1574102338688395e-08,
+      "loss": 0.4595,
+      "step": 11883
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.144123928509865,
+      "learning_rate": 2.1451435944464528e-08,
+      "loss": 1.4252,
+      "step": 11884
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.7105695583681575,
+      "learning_rate": 2.1329118524827662e-08,
+      "loss": 0.586,
+      "step": 11885
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.9768130214957003,
+      "learning_rate": 2.120715008835261e-08,
+      "loss": 0.7267,
+      "step": 11886
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.60573974763432,
+      "learning_rate": 2.1085530643588094e-08,
+      "loss": 0.8107,
+      "step": 11887
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.176063171160444,
+      "learning_rate": 2.096426019906006e-08,
+      "loss": 0.7135,
+      "step": 11888
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.689431942878935,
+      "learning_rate": 2.0843338763268382e-08,
+      "loss": 0.6004,
+      "step": 11889
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.014500092531755,
+      "learning_rate": 2.0722766344689617e-08,
+      "loss": 0.7066,
+      "step": 11890
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.82800488707636,
+      "learning_rate": 2.0602542951774774e-08,
+      "loss": 0.6034,
+      "step": 11891
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.390492523335453,
+      "learning_rate": 2.0482668592951004e-08,
+      "loss": 0.8957,
+      "step": 11892
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.226787649241726,
+      "learning_rate": 2.036314327662159e-08,
+      "loss": 0.8212,
+      "step": 11893
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.9940600042133187,
+      "learning_rate": 2.0243967011164267e-08,
+      "loss": 0.37,
+      "step": 11894
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.627394164356139,
+      "learning_rate": 2.0125139804932913e-08,
+      "loss": 0.602,
+      "step": 11895
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.748931715639397,
+      "learning_rate": 2.0006661666256978e-08,
+      "loss": 0.8945,
+      "step": 11896
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.1576488933848617,
+      "learning_rate": 1.988853260344037e-08,
+      "loss": 0.3595,
+      "step": 11897
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.3844736354884932,
+      "learning_rate": 1.97707526247648e-08,
+      "loss": 0.4471,
+      "step": 11898
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.928263498979704,
+      "learning_rate": 1.965332173848533e-08,
+      "loss": 1.2007,
+      "step": 11899
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.678241608241835,
+      "learning_rate": 1.9536239952833712e-08,
+      "loss": 0.2836,
+      "step": 11900
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.065948311688887,
+      "learning_rate": 1.9419507276016158e-08,
+      "loss": 1.2764,
+      "step": 11901
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.294378649306096,
+      "learning_rate": 1.9303123716215565e-08,
+      "loss": 0.9947,
+      "step": 11902
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.096427690802625,
+      "learning_rate": 1.9187089281589853e-08,
+      "loss": 0.6188,
+      "step": 11903
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.080728240667454,
+      "learning_rate": 1.9071403980273075e-08,
+      "loss": 1.1303,
+      "step": 11904
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.347288640554572,
+      "learning_rate": 1.895606782037318e-08,
+      "loss": 0.6736,
+      "step": 11905
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.780075138837884,
+      "learning_rate": 1.8841080809975933e-08,
+      "loss": 1.0143,
+      "step": 11906
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.762028599734933,
+      "learning_rate": 1.872644295714099e-08,
+      "loss": 0.9305,
+      "step": 11907
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.049603974319265,
+      "learning_rate": 1.8612154269903036e-08,
+      "loss": 1.0608,
+      "step": 11908
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.939565444899833,
+      "learning_rate": 1.8498214756274558e-08,
+      "loss": 0.5482,
+      "step": 11909
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.353598562575978,
+      "learning_rate": 1.8384624424241383e-08,
+      "loss": 0.9264,
+      "step": 11910
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.210083125387604,
+      "learning_rate": 1.827138328176603e-08,
+      "loss": 0.6805,
+      "step": 11911
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.789070033516132,
+      "learning_rate": 1.81584913367866e-08,
+      "loss": 0.6795,
+      "step": 11912
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.278633192332255,
+      "learning_rate": 1.8045948597215646e-08,
+      "loss": 1.0238,
+      "step": 11913
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.605182109465508,
+      "learning_rate": 1.793375507094186e-08,
+      "loss": 0.3206,
+      "step": 11914
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.003408022361503,
+      "learning_rate": 1.7821910765830063e-08,
+      "loss": 0.6041,
+      "step": 11915
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.765548228462366,
+      "learning_rate": 1.771041568971954e-08,
+      "loss": 0.4168,
+      "step": 11916
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.456842463420886,
+      "learning_rate": 1.7599269850426258e-08,
+      "loss": 0.8319,
+      "step": 11917
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.731562967711699,
+      "learning_rate": 1.7488473255740657e-08,
+      "loss": 0.653,
+      "step": 11918
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.1815817245847033,
+      "learning_rate": 1.7378025913428743e-08,
+      "loss": 0.6323,
+      "step": 11919
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.659768289388934,
+      "learning_rate": 1.726792783123321e-08,
+      "loss": 0.5598,
+      "step": 11920
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.5609989582963357,
+      "learning_rate": 1.7158179016870668e-08,
+      "loss": 0.5553,
+      "step": 11921
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.873352263562927,
+      "learning_rate": 1.7048779478034404e-08,
+      "loss": 0.8343,
+      "step": 11922
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0772548699872506,
+      "learning_rate": 1.6939729222393286e-08,
+      "loss": 0.279,
+      "step": 11923
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.0376698788409873,
+      "learning_rate": 1.6831028257590087e-08,
+      "loss": 0.6518,
+      "step": 11924
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.657201488738241,
+      "learning_rate": 1.6722676591245378e-08,
+      "loss": 0.4616,
+      "step": 11925
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.8920629960410293,
+      "learning_rate": 1.6614674230953643e-08,
+      "loss": 0.751,
+      "step": 11926
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.7380583054259944,
+      "learning_rate": 1.6507021184285488e-08,
+      "loss": 0.5969,
+      "step": 11927
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.484805408914432,
+      "learning_rate": 1.63997174587871e-08,
+      "loss": 0.6674,
+      "step": 11928
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.708051571600095,
+      "learning_rate": 1.629276306197969e-08,
+      "loss": 0.4126,
+      "step": 11929
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.009169489358039,
+      "learning_rate": 1.6186158001360587e-08,
+      "loss": 0.9243,
+      "step": 11930
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.5079588041639775,
+      "learning_rate": 1.607990228440215e-08,
+      "loss": 0.6558,
+      "step": 11931
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.309796315389396,
+      "learning_rate": 1.597399591855231e-08,
+      "loss": 1.3317,
+      "step": 11932
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.191601799634067,
+      "learning_rate": 1.5868438911234575e-08,
+      "loss": 0.7678,
+      "step": 11933
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.033040535587328,
+      "learning_rate": 1.5763231269848578e-08,
+      "loss": 1.1157,
+      "step": 11934
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.441359331862082,
+      "learning_rate": 1.5658373001768423e-08,
+      "loss": 1.1209,
+      "step": 11935
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9344080602485448,
+      "learning_rate": 1.555386411434434e-08,
+      "loss": 0.1267,
+      "step": 11936
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.224522930216481,
+      "learning_rate": 1.544970461490214e-08,
+      "loss": 0.7561,
+      "step": 11937
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.551766070858134,
+      "learning_rate": 1.53458945107432e-08,
+      "loss": 1.2067,
+      "step": 11938
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.9293824274572704,
+      "learning_rate": 1.5242433809143364e-08,
+      "loss": 0.2524,
+      "step": 11939
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.711783064616934,
+      "learning_rate": 1.5139322517355172e-08,
+      "loss": 0.9472,
+      "step": 11940
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.415126915218515,
+      "learning_rate": 1.5036560642606167e-08,
+      "loss": 0.8046,
+      "step": 11941
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.426248150142898,
+      "learning_rate": 1.493414819210004e-08,
+      "loss": 0.5639,
+      "step": 11942
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.3045330296380113,
+      "learning_rate": 1.4832085173014376e-08,
+      "loss": 0.6792,
+      "step": 11943
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.062424557004166,
+      "learning_rate": 1.4730371592504567e-08,
+      "loss": 0.1663,
+      "step": 11944
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.925225735951944,
+      "learning_rate": 1.4629007457699906e-08,
+      "loss": 0.6722,
+      "step": 11945
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.3487778120695566,
+      "learning_rate": 1.4527992775704713e-08,
+      "loss": 0.8364,
+      "step": 11946
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.472139605321025,
+      "learning_rate": 1.4427327553601101e-08,
+      "loss": 0.8356,
+      "step": 11947
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.172637141606483,
+      "learning_rate": 1.432701179844398e-08,
+      "loss": 0.495,
+      "step": 11948
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.643966694879904,
+      "learning_rate": 1.4227045517266059e-08,
+      "loss": 1.0396,
+      "step": 11949
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.667096334034253,
+      "learning_rate": 1.4127428717073955e-08,
+      "loss": 1.1284,
+      "step": 11950
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.9324324700908826,
+      "learning_rate": 1.4028161404850415e-08,
+      "loss": 0.5657,
+      "step": 11951
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.679804891627546,
+      "learning_rate": 1.3929243587553764e-08,
+      "loss": 0.6958,
+      "step": 11952
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.805587188154656,
+      "learning_rate": 1.3830675272117344e-08,
+      "loss": 0.8619,
+      "step": 11953
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.6120518053429453,
+      "learning_rate": 1.3732456465451182e-08,
+      "loss": 0.3556,
+      "step": 11954
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.866311449363066,
+      "learning_rate": 1.3634587174439218e-08,
+      "loss": 0.8505,
+      "step": 11955
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.881058075921292,
+      "learning_rate": 1.3537067405942072e-08,
+      "loss": 1.0058,
+      "step": 11956
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8398806888597605,
+      "learning_rate": 1.3439897166795945e-08,
+      "loss": 0.318,
+      "step": 11957
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.063571392378166,
+      "learning_rate": 1.3343076463810389e-08,
+      "loss": 0.5925,
+      "step": 11958
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.298045597186789,
+      "learning_rate": 1.3246605303773864e-08,
+      "loss": 0.6629,
+      "step": 11959
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.473056012845691,
+      "learning_rate": 1.3150483693447625e-08,
+      "loss": 1.2219,
+      "step": 11960
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.532689657101384,
+      "learning_rate": 1.3054711639569616e-08,
+      "loss": 1.2582,
+      "step": 11961
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.7416802434167455,
+      "learning_rate": 1.295928914885336e-08,
+      "loss": 0.8745,
+      "step": 11962
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.9350085760977156,
+      "learning_rate": 1.2864216227986837e-08,
+      "loss": 0.6069,
+      "step": 11963
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.849630921846114,
+      "learning_rate": 1.276949288363527e-08,
+      "loss": 1.3068,
+      "step": 11964
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.206955644824944,
+      "learning_rate": 1.267511912243724e-08,
+      "loss": 0.8276,
+      "step": 11965
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.423765874126548,
+      "learning_rate": 1.2581094951008566e-08,
+      "loss": 0.5732,
+      "step": 11966
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.300025809881357,
+      "learning_rate": 1.2487420375939529e-08,
+      "loss": 0.8077,
+      "step": 11967
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.206757278466938,
+      "learning_rate": 1.2394095403797102e-08,
+      "loss": 1.3847,
+      "step": 11968
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.9055794737494822,
+      "learning_rate": 1.2301120041122161e-08,
+      "loss": 0.6262,
+      "step": 11969
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.613323997082934,
+      "learning_rate": 1.2208494294432272e-08,
+      "loss": 0.8649,
+      "step": 11970
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.230109954266526,
+      "learning_rate": 1.2116218170220018e-08,
+      "loss": 0.2765,
+      "step": 11971
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.2792223501490074,
+      "learning_rate": 1.202429167495356e-08,
+      "loss": 0.3214,
+      "step": 11972
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.2794069348955475,
+      "learning_rate": 1.1932714815076075e-08,
+      "loss": 0.7824,
+      "step": 11973
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.515920252114767,
+      "learning_rate": 1.1841487597007983e-08,
+      "loss": 0.7136,
+      "step": 11974
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.47317136412364,
+      "learning_rate": 1.1750610027142506e-08,
+      "loss": 1.1016,
+      "step": 11975
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1656497913356112,
+      "learning_rate": 1.1660082111850101e-08,
+      "loss": 0.1311,
+      "step": 11976
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.428046457044096,
+      "learning_rate": 1.156990385747736e-08,
+      "loss": 0.5926,
+      "step": 11977
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.010542686065303,
+      "learning_rate": 1.1480075270343671e-08,
+      "loss": 0.4662,
+      "step": 11978
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.1502107158997195,
+      "learning_rate": 1.139059635674733e-08,
+      "loss": 0.6396,
+      "step": 11979
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.514111467989537,
+      "learning_rate": 1.1301467122959432e-08,
+      "loss": 0.6339,
+      "step": 11980
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.807595894519607,
+      "learning_rate": 1.1212687575227754e-08,
+      "loss": 0.7422,
+      "step": 11981
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8296572392135173,
+      "learning_rate": 1.1124257719775655e-08,
+      "loss": 0.2994,
+      "step": 11982
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.495040332096613,
+      "learning_rate": 1.1036177562800954e-08,
+      "loss": 0.5596,
+      "step": 11983
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.374313945016398,
+      "learning_rate": 1.0948447110478711e-08,
+      "loss": 0.8237,
+      "step": 11984
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.3603395917008236,
+      "learning_rate": 1.0861066368957341e-08,
+      "loss": 0.5336,
+      "step": 11985
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.797164024142589,
+      "learning_rate": 1.0774035344363054e-08,
+      "loss": 1.0428,
+      "step": 11986
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.798474908659315,
+      "learning_rate": 1.0687354042795417e-08,
+      "loss": 0.5348,
+      "step": 11987
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.653075763774985,
+      "learning_rate": 1.060102247033068e-08,
+      "loss": 0.8898,
+      "step": 11988
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.7221055998469206,
+      "learning_rate": 1.0515040633020112e-08,
+      "loss": 0.8894,
+      "step": 11989
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.239262202588881,
+      "learning_rate": 1.0429408536891117e-08,
+      "loss": 1.143,
+      "step": 11990
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4338166321686046,
+      "learning_rate": 1.0344126187946113e-08,
+      "loss": 0.1791,
+      "step": 11991
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.6841008740597605,
+      "learning_rate": 1.0259193592162541e-08,
+      "loss": 0.5626,
+      "step": 11992
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.378483590578005,
+      "learning_rate": 1.0174610755493974e-08,
+      "loss": 0.4065,
+      "step": 11993
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.2204144413208717,
+      "learning_rate": 1.0090377683869557e-08,
+      "loss": 0.2904,
+      "step": 11994
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.9902759711671125,
+      "learning_rate": 1.0006494383193454e-08,
+      "loss": 1.1846,
+      "step": 11995
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.853469311831628,
+      "learning_rate": 9.92296085934541e-09,
+      "loss": 0.547,
+      "step": 11996
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.9650594451607555,
+      "learning_rate": 9.839777118181293e-09,
+      "loss": 1.2478,
+      "step": 11997
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.338080966871579,
+      "learning_rate": 9.756943165531441e-09,
+      "loss": 0.9709,
+      "step": 11998
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.606574446922868,
+      "learning_rate": 9.67445900720232e-09,
+      "loss": 1.0099,
+      "step": 11999
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.2828035934487065,
+      "learning_rate": 9.592324648975415e-09,
+      "loss": 0.714,
+      "step": 12000
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.215669569185245,
+      "learning_rate": 9.510540096608345e-09,
+      "loss": 1.2572,
+      "step": 12001
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.2673720077363075,
+      "learning_rate": 9.429105355833745e-09,
+      "loss": 0.5131,
+      "step": 12002
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.8685929308413742,
+      "learning_rate": 9.348020432359829e-09,
+      "loss": 0.7717,
+      "step": 12003
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.6612516918354405,
+      "learning_rate": 9.267285331870378e-09,
+      "loss": 0.8762,
+      "step": 12004
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.835309486902613,
+      "learning_rate": 9.186900060024207e-09,
+      "loss": 1.3228,
+      "step": 12005
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.1110886609000055,
+      "learning_rate": 9.106864622456246e-09,
+      "loss": 0.6116,
+      "step": 12006
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.398732853475271,
+      "learning_rate": 9.02717902477701e-09,
+      "loss": 0.7167,
+      "step": 12007
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.905729799829959,
+      "learning_rate": 8.947843272571477e-09,
+      "loss": 1.4806,
+      "step": 12008
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.3584098306988754,
+      "learning_rate": 8.868857371401306e-09,
+      "loss": 0.3806,
+      "step": 12009
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.295833479211921,
+      "learning_rate": 8.790221326802074e-09,
+      "loss": 0.7821,
+      "step": 12010
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.056794579846256,
+      "learning_rate": 8.711935144287142e-09,
+      "loss": 1.2853,
+      "step": 12011
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.741309227622882,
+      "learning_rate": 8.633998829343237e-09,
+      "loss": 1.0193,
+      "step": 12012
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.416962423605944,
+      "learning_rate": 8.55641238743321e-09,
+      "loss": 1.1095,
+      "step": 12013
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.034344568422561,
+      "learning_rate": 8.479175823996044e-09,
+      "loss": 0.5427,
+      "step": 12014
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.6788069740696177,
+      "learning_rate": 8.40228914444574e-09,
+      "loss": 0.6993,
+      "step": 12015
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.450727521986115,
+      "learning_rate": 8.325752354171324e-09,
+      "loss": 0.3584,
+      "step": 12016
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.6095034716190466,
+      "learning_rate": 8.24956545853739e-09,
+      "loss": 0.5702,
+      "step": 12017
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.655746956771417,
+      "learning_rate": 8.173728462885222e-09,
+      "loss": 0.8922,
+      "step": 12018
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.668643716701874,
+      "learning_rate": 8.098241372530013e-09,
+      "loss": 0.5593,
+      "step": 12019
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.4627075621636396,
+      "learning_rate": 8.023104192763642e-09,
+      "loss": 0.7913,
+      "step": 12020
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.3071089435334793,
+      "learning_rate": 7.948316928851896e-09,
+      "loss": 0.5397,
+      "step": 12021
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.7342020527729853,
+      "learning_rate": 7.873879586037803e-09,
+      "loss": 0.7649,
+      "step": 12022
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.573511773337518,
+      "learning_rate": 7.79979216953941e-09,
+      "loss": 1.1509,
+      "step": 12023
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.709589592626596,
+      "learning_rate": 7.726054684549234e-09,
+      "loss": 1.1941,
+      "step": 12024
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.209516616837038,
+      "learning_rate": 7.652667136235914e-09,
+      "loss": 0.3777,
+      "step": 12025
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.3845853169481668,
+      "learning_rate": 7.579629529744225e-09,
+      "loss": 0.2514,
+      "step": 12026
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.636504614812306,
+      "learning_rate": 7.506941870192851e-09,
+      "loss": 0.6073,
+      "step": 12027
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.594125764328588,
+      "learning_rate": 7.434604162678271e-09,
+      "loss": 0.5927,
+      "step": 12028
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.09639929471377,
+      "learning_rate": 7.362616412269763e-09,
+      "loss": 0.8389,
+      "step": 12029
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.47282169012807,
+      "learning_rate": 7.290978624013289e-09,
+      "loss": 0.6142,
+      "step": 12030
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.471227365977711,
+      "learning_rate": 7.2196908029315e-09,
+      "loss": 0.3131,
+      "step": 12031
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.8664298739220238,
+      "learning_rate": 7.148752954020955e-09,
+      "loss": 0.6902,
+      "step": 12032
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.0530976836411043,
+      "learning_rate": 7.07816508225323e-09,
+      "loss": 0.4628,
+      "step": 12033
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.484740333351431,
+      "learning_rate": 7.0079271925771465e-09,
+      "loss": 1.124,
+      "step": 12034
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.338984250732091,
+      "learning_rate": 6.9380392899159875e-09,
+      "loss": 0.9777,
+      "step": 12035
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.42961326301565,
+      "learning_rate": 6.868501379168058e-09,
+      "loss": 0.6223,
+      "step": 12036
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.52526406643918,
+      "learning_rate": 6.799313465208346e-09,
+      "loss": 1.2402,
+      "step": 12037
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.624715613096466,
+      "learning_rate": 6.730475552886306e-09,
+      "loss": 0.6295,
+      "step": 12038
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.777159478225706,
+      "learning_rate": 6.661987647026969e-09,
+      "loss": 0.6173,
+      "step": 12039
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.5679183930589797,
+      "learning_rate": 6.593849752430936e-09,
+      "loss": 0.4285,
+      "step": 12040
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.2585865016766355,
+      "learning_rate": 6.5260618738749445e-09,
+      "loss": 0.7382,
+      "step": 12041
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.215975875018411,
+      "learning_rate": 6.458624016110193e-09,
+      "loss": 0.9977,
+      "step": 12042
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.232271902640726,
+      "learning_rate": 6.391536183864566e-09,
+      "loss": 0.3883,
+      "step": 12043
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.153310257696214,
+      "learning_rate": 6.324798381839303e-09,
+      "loss": 0.7757,
+      "step": 12044
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.174932807336339,
+      "learning_rate": 6.2584106147134395e-09,
+      "loss": 0.3164,
+      "step": 12045
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.3192495238076245,
+      "learning_rate": 6.192372887139919e-09,
+      "loss": 0.3615,
+      "step": 12046
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.130341919061836,
+      "learning_rate": 6.126685203747818e-09,
+      "loss": 0.802,
+      "step": 12047
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.352963632177688,
+      "learning_rate": 6.0613475691417845e-09,
+      "loss": 0.5643,
+      "step": 12048
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.739231463377654,
+      "learning_rate": 5.996359987902045e-09,
+      "loss": 1.1208,
+      "step": 12049
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.948863433441496,
+      "learning_rate": 5.931722464583289e-09,
+      "loss": 1.0449,
+      "step": 12050
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2005746125622068,
+      "learning_rate": 5.8674350037163374e-09,
+      "loss": 0.1879,
+      "step": 12051
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.313853268223101,
+      "learning_rate": 5.803497609807585e-09,
+      "loss": 0.3264,
+      "step": 12052
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.948738610094623,
+      "learning_rate": 5.7399102873390015e-09,
+      "loss": 1.0435,
+      "step": 12053
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.410587302649346,
+      "learning_rate": 5.676673040767578e-09,
+      "loss": 0.6349,
+      "step": 12054
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.5198431546681044,
+      "learning_rate": 5.613785874525879e-09,
+      "loss": 0.8272,
+      "step": 12055
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.5320677154198457,
+      "learning_rate": 5.551248793022601e-09,
+      "loss": 0.707,
+      "step": 12056
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.365378882592431,
+      "learning_rate": 5.48906180064035e-09,
+      "loss": 0.5722,
+      "step": 12057
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.508363246514844,
+      "learning_rate": 5.42722490173897e-09,
+      "loss": 0.6041,
+      "step": 12058
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.7126384252880693,
+      "learning_rate": 5.365738100652773e-09,
+      "loss": 0.7763,
+      "step": 12059
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.957741692123336,
+      "learning_rate": 5.304601401691089e-09,
+      "loss": 0.64,
+      "step": 12060
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.367737865698479,
+      "learning_rate": 5.243814809140491e-09,
+      "loss": 0.8158,
+      "step": 12061
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.098217074943541,
+      "learning_rate": 5.1833783272609016e-09,
+      "loss": 1.0765,
+      "step": 12062
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.065663853354703,
+      "learning_rate": 5.123291960288934e-09,
+      "loss": 0.8675,
+      "step": 12063
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.3357442866701796,
+      "learning_rate": 5.0635557124362185e-09,
+      "loss": 0.6963,
+      "step": 12064
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.910208697172912,
+      "learning_rate": 5.0041695878905175e-09,
+      "loss": 0.9939,
+      "step": 12065
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.52745161151713,
+      "learning_rate": 4.94513359081461e-09,
+      "loss": 0.7236,
+      "step": 12066
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.697380862613846,
+      "learning_rate": 4.886447725345744e-09,
+      "loss": 0.3768,
+      "step": 12067
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.151388769159717,
+      "learning_rate": 4.82811199559785e-09,
+      "loss": 0.8545,
+      "step": 12068
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 7.4866699720702625,
+      "learning_rate": 4.7701264056609905e-09,
+      "loss": 1.1383,
+      "step": 12069
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.836390930087791,
+      "learning_rate": 4.712490959598581e-09,
+      "loss": 0.4125,
+      "step": 12070
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.1532673518670102,
+      "learning_rate": 4.655205661450724e-09,
+      "loss": 0.7576,
+      "step": 12071
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4577000518900594,
+      "learning_rate": 4.5982705152336496e-09,
+      "loss": 0.1866,
+      "step": 12072
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.3940924734291844,
+      "learning_rate": 4.5416855249375e-09,
+      "loss": 0.4667,
+      "step": 12073
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.407259559105291,
+      "learning_rate": 4.485450694528548e-09,
+      "loss": 0.8662,
+      "step": 12074
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.0699053562622516,
+      "learning_rate": 4.42956602794975e-09,
+      "loss": 0.441,
+      "step": 12075
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.38530006511475,
+      "learning_rate": 4.374031529116862e-09,
+      "loss": 0.7571,
+      "step": 12076
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.387214073840642,
+      "learning_rate": 4.318847201923437e-09,
+      "loss": 0.9429,
+      "step": 12077
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.66135024157567,
+      "learning_rate": 4.264013050238047e-09,
+      "loss": 0.801,
+      "step": 12078
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.0667162608964755,
+      "learning_rate": 4.209529077903174e-09,
+      "loss": 0.5635,
+      "step": 12079
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.742693783361723,
+      "learning_rate": 4.155395288739095e-09,
+      "loss": 0.6254,
+      "step": 12080
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.044281661809891,
+      "learning_rate": 4.101611686539442e-09,
+      "loss": 0.6202,
+      "step": 12081
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.0761978720826253,
+      "learning_rate": 4.0481782750745325e-09,
+      "loss": 0.4201,
+      "step": 12082
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 6.880557286207846,
+      "learning_rate": 3.995095058090259e-09,
+      "loss": 0.9233,
+      "step": 12083
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.5296076657355235,
+      "learning_rate": 3.942362039306979e-09,
+      "loss": 0.5142,
+      "step": 12084
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.464720030419047,
+      "learning_rate": 3.889979222421181e-09,
+      "loss": 1.1719,
+      "step": 12085
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.1649747010164475,
+      "learning_rate": 3.837946611104926e-09,
+      "loss": 0.7354,
+      "step": 12086
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.6497758735676253,
+      "learning_rate": 3.786264209004742e-09,
+      "loss": 0.5491,
+      "step": 12087
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 6.349024582490853,
+      "learning_rate": 3.7349320197443974e-09,
+      "loss": 1.4597,
+      "step": 12088
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7430878185833116,
+      "learning_rate": 3.6839500469210145e-09,
+      "loss": 0.3025,
+      "step": 12089
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.420295160858151,
+      "learning_rate": 3.633318294108956e-09,
+      "loss": 0.7601,
+      "step": 12090
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.9294396445199298,
+      "learning_rate": 3.583036764857051e-09,
+      "loss": 0.8451,
+      "step": 12091
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.166301839906999,
+      "learning_rate": 3.533105462689701e-09,
+      "loss": 0.774,
+      "step": 12092
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.400574956124744,
+      "learning_rate": 3.483524391106885e-09,
+      "loss": 0.9232,
+      "step": 12093
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.33060608225683,
+      "learning_rate": 3.434293553584156e-09,
+      "loss": 0.3535,
+      "step": 12094
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7566506342748722,
+      "learning_rate": 3.385412953572087e-09,
+      "loss": 0.2392,
+      "step": 12095
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.964425428089187,
+      "learning_rate": 3.3368825944973813e-09,
+      "loss": 0.6337,
+      "step": 12096
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.5144354601565175,
+      "learning_rate": 3.2887024797617628e-09,
+      "loss": 0.4264,
+      "step": 12097
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.3015414214756476,
+      "learning_rate": 3.2408726127425294e-09,
+      "loss": 0.7685,
+      "step": 12098
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.7046508179692763,
+      "learning_rate": 3.1933929967919996e-09,
+      "loss": 0.6863,
+      "step": 12099
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.050043842814273,
+      "learning_rate": 3.146263635238622e-09,
+      "loss": 0.498,
+      "step": 12100
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.431672901040743,
+      "learning_rate": 3.0994845313853106e-09,
+      "loss": 0.87,
+      "step": 12101
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 6.753347493145347,
+      "learning_rate": 3.0530556885116637e-09,
+      "loss": 1.0715,
+      "step": 12102
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.7226135852669464,
+      "learning_rate": 3.0069771098723e-09,
+      "loss": 0.4195,
+      "step": 12103
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.2460739822213736,
+      "learning_rate": 2.9612487986968587e-09,
+      "loss": 0.8109,
+      "step": 12104
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.845247910493663,
+      "learning_rate": 2.915870758190553e-09,
+      "loss": 0.4189,
+      "step": 12105
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.5210622289945293,
+      "learning_rate": 2.870842991534173e-09,
+      "loss": 0.3681,
+      "step": 12106
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.3116643272956425,
+      "learning_rate": 2.826165501884082e-09,
+      "loss": 0.5471,
+      "step": 12107
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.084274314757648,
+      "learning_rate": 2.7818382923722188e-09,
+      "loss": 0.7128,
+      "step": 12108
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.071137499790908,
+      "learning_rate": 2.737861366105543e-09,
+      "loss": 0.831,
+      "step": 12109
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.2311637576709096,
+      "learning_rate": 2.694234726166589e-09,
+      "loss": 0.8248,
+      "step": 12110
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.6235213770575205,
+      "learning_rate": 2.650958375613466e-09,
+      "loss": 0.7375,
+      "step": 12111
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.0996566294527779,
+      "learning_rate": 2.6080323174798583e-09,
+      "loss": 0.152,
+      "step": 12112
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.692264412669275,
+      "learning_rate": 2.565456554773915e-09,
+      "loss": 0.6277,
+      "step": 12113
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7889331061473897,
+      "learning_rate": 2.5232310904810265e-09,
+      "loss": 0.2284,
+      "step": 12114
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.218512258355676,
+      "learning_rate": 2.4813559275604914e-09,
+      "loss": 0.7349,
+      "step": 12115
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.812025075684706,
+      "learning_rate": 2.439831068947185e-09,
+      "loss": 1.0505,
+      "step": 12116
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.8313037967846735,
+      "learning_rate": 2.3986565175526665e-09,
+      "loss": 0.6363,
+      "step": 12117
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.111197016156909,
+      "learning_rate": 2.357832276262961e-09,
+      "loss": 0.6512,
+      "step": 12118
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.696129752437114,
+      "learning_rate": 2.3173583479391137e-09,
+      "loss": 0.5707,
+      "step": 12119
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.2821224263699524,
+      "learning_rate": 2.2772347354182987e-09,
+      "loss": 0.4001,
+      "step": 12120
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.429749999447198,
+      "learning_rate": 2.2374614415132666e-09,
+      "loss": 1.0118,
+      "step": 12121
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.773818785295129,
+      "learning_rate": 2.198038469011787e-09,
+      "loss": 1.1201,
+      "step": 12122
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1309515054933859,
+      "learning_rate": 2.1589658206772058e-09,
+      "loss": 0.1553,
+      "step": 12123
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.181621946353914,
+      "learning_rate": 2.1202434992484423e-09,
+      "loss": 0.6917,
+      "step": 12124
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.798752576064015,
+      "learning_rate": 2.081871507439992e-09,
+      "loss": 1.1969,
+      "step": 12125
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.898025030454768,
+      "learning_rate": 2.0438498479413705e-09,
+      "loss": 0.5628,
+      "step": 12126
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.041043124480218,
+      "learning_rate": 2.0061785234176677e-09,
+      "loss": 0.6858,
+      "step": 12127
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.066576743807657,
+      "learning_rate": 1.9688575365095497e-09,
+      "loss": 0.8998,
+      "step": 12128
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.3224466226674085,
+      "learning_rate": 1.9318868898327015e-09,
+      "loss": 0.4893,
+      "step": 12129
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.3158964199414225,
+      "learning_rate": 1.8952665859789387e-09,
+      "loss": 1.0153,
+      "step": 12130
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.705210911064366,
+      "learning_rate": 1.8589966275156523e-09,
+      "loss": 0.7379,
+      "step": 12131
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.579219121921822,
+      "learning_rate": 1.8230770169841427e-09,
+      "loss": 0.7464,
+      "step": 12132
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.598790107251219,
+      "learning_rate": 1.787507756903506e-09,
+      "loss": 0.3411,
+      "step": 12133
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.074528102971161,
+      "learning_rate": 1.7522888497656376e-09,
+      "loss": 0.6873,
+      "step": 12134
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.343070224980005,
+      "learning_rate": 1.7174202980402287e-09,
+      "loss": 0.483,
+      "step": 12135
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.61630229921253,
+      "learning_rate": 1.6829021041708805e-09,
+      "loss": 0.8581,
+      "step": 12136
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.0913027516957134,
+      "learning_rate": 1.6487342705773234e-09,
+      "loss": 1.0862,
+      "step": 12137
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.793490104705858,
+      "learning_rate": 1.6149167996548643e-09,
+      "loss": 0.9786,
+      "step": 12138
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.979654398550657,
+      "learning_rate": 1.5814496937732737e-09,
+      "loss": 1.2805,
+      "step": 12139
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 6.044335416386995,
+      "learning_rate": 1.5483329552790082e-09,
+      "loss": 1.0259,
+      "step": 12140
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.232417148180656,
+      "learning_rate": 1.5155665864935442e-09,
+      "loss": 0.9337,
+      "step": 12141
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.1515921609154787,
+      "learning_rate": 1.4831505897128229e-09,
+      "loss": 0.5646,
+      "step": 12142
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.4521180420535695,
+      "learning_rate": 1.4510849672100258e-09,
+      "loss": 0.3731,
+      "step": 12143
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.650689960042224,
+      "learning_rate": 1.4193697212322444e-09,
+      "loss": 0.5591,
+      "step": 12144
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.795854212893908,
+      "learning_rate": 1.3880048540032554e-09,
+      "loss": 0.5542,
+      "step": 12145
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.290467557286918,
+      "learning_rate": 1.3569903677207453e-09,
+      "loss": 1.0103,
+      "step": 12146
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.233485159890757,
+      "learning_rate": 1.3263262645585307e-09,
+      "loss": 0.3186,
+      "step": 12147
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.040748043743469,
+      "learning_rate": 1.2960125466671137e-09,
+      "loss": 0.4931,
+      "step": 12148
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.1337209065940765,
+      "learning_rate": 1.266049216170906e-09,
+      "loss": 0.9532,
+      "step": 12149
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.040915504923403,
+      "learning_rate": 1.2364362751698944e-09,
+      "loss": 0.9432,
+      "step": 12150
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.61108572071553,
+      "learning_rate": 1.2071737257401962e-09,
+      "loss": 0.6336,
+      "step": 12151
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.652113726957336,
+      "learning_rate": 1.1782615699323929e-09,
+      "loss": 0.5952,
+      "step": 12152
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.98305392012479,
+      "learning_rate": 1.149699809773752e-09,
+      "loss": 0.5316,
+      "step": 12153
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.605235076813467,
+      "learning_rate": 1.1214884472660059e-09,
+      "loss": 0.6964,
+      "step": 12154
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.000185190848004,
+      "learning_rate": 1.0936274843864615e-09,
+      "loss": 0.992,
+      "step": 12155
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.933712871831323,
+      "learning_rate": 1.0661169230891111e-09,
+      "loss": 0.8956,
+      "step": 12156
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.976258849223066,
+      "learning_rate": 1.038956765300747e-09,
+      "loss": 1.0503,
+      "step": 12157
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.8920392286889363,
+      "learning_rate": 1.012147012926512e-09,
+      "loss": 0.361,
+      "step": 12158
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9254614358134056,
+      "learning_rate": 9.856876678443484e-10,
+      "loss": 0.1275,
+      "step": 12159
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 6.667265819278876,
+      "learning_rate": 9.595787319105488e-10,
+      "loss": 0.814,
+      "step": 12160
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 6.063537108631886,
+      "learning_rate": 9.338202069536506e-10,
+      "loss": 1.0902,
+      "step": 12161
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.345372693382865,
+      "learning_rate": 9.084120947805419e-10,
+      "loss": 0.9118,
+      "step": 12162
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.53785749891009,
+      "learning_rate": 8.833543971714653e-10,
+      "loss": 0.6818,
+      "step": 12163
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.580761358394674,
+      "learning_rate": 8.586471158827936e-10,
+      "loss": 1.1031,
+      "step": 12164
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.9874127649923854,
+      "learning_rate": 8.342902526470298e-10,
+      "loss": 0.7016,
+      "step": 12165
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.721465501571412,
+      "learning_rate": 8.102838091705867e-10,
+      "loss": 0.4219,
+      "step": 12166
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.6672290208821887,
+      "learning_rate": 7.866277871371175e-10,
+      "loss": 1.019,
+      "step": 12167
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.720348100268207,
+      "learning_rate": 7.633221882041852e-10,
+      "loss": 0.903,
+      "step": 12168
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.33209776877341,
+      "learning_rate": 7.403670140054831e-10,
+      "loss": 0.9556,
+      "step": 12169
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9504423045579515,
+      "learning_rate": 7.177622661508343e-10,
+      "loss": 0.3021,
+      "step": 12170
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 6.979290745231007,
+      "learning_rate": 6.955079462234171e-10,
+      "loss": 1.5872,
+      "step": 12171
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.9990523954197768,
+      "learning_rate": 6.73604055784205e-10,
+      "loss": 0.449,
+      "step": 12172
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.5592680367105105,
+      "learning_rate": 6.520505963680813e-10,
+      "loss": 0.7497,
+      "step": 12173
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.9589191050959562,
+      "learning_rate": 6.308475694860594e-10,
+      "loss": 0.6861,
+      "step": 12174
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.6985531257099646,
+      "learning_rate": 6.099949766241731e-10,
+      "loss": 0.5641,
+      "step": 12175
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 5.523605378907456,
+      "learning_rate": 5.894928192440308e-10,
+      "loss": 0.9872,
+      "step": 12176
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.1178306652394356,
+      "learning_rate": 5.693410987833714e-10,
+      "loss": 0.7559,
+      "step": 12177
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.862073416113283,
+      "learning_rate": 5.495398166538435e-10,
+      "loss": 0.6149,
+      "step": 12178
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 5.5432532289577665,
+      "learning_rate": 5.30088974244336e-10,
+      "loss": 0.9842,
+      "step": 12179
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.801296327518294,
+      "learning_rate": 5.109885729176478e-10,
+      "loss": 0.5074,
+      "step": 12180
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 5.111438497992996,
+      "learning_rate": 4.922386140127078e-10,
+      "loss": 1.0625,
+      "step": 12181
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.462694911081913,
+      "learning_rate": 4.738390988440201e-10,
+      "loss": 0.8326,
+      "step": 12182
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.6998639465625174,
+      "learning_rate": 4.5579002870110854e-10,
+      "loss": 0.6114,
+      "step": 12183
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 6.221654366729723,
+      "learning_rate": 4.380914048490725e-10,
+      "loss": 0.9554,
+      "step": 12184
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3629454830687455,
+      "learning_rate": 4.207432285291413e-10,
+      "loss": 0.1495,
+      "step": 12185
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.8014041633328066,
+      "learning_rate": 4.0374550095645394e-10,
+      "loss": 0.7206,
+      "step": 12186
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.268004847733571,
+      "learning_rate": 3.8709822332339e-10,
+      "loss": 0.6349,
+      "step": 12187
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 5.32831186500912,
+      "learning_rate": 3.708013967956836e-10,
+      "loss": 0.9325,
+      "step": 12188
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.2594682309992455,
+      "learning_rate": 3.5485502251686457e-10,
+      "loss": 0.326,
+      "step": 12189
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.1085646358781025,
+      "learning_rate": 3.3925910160381713e-10,
+      "loss": 0.1365,
+      "step": 12190
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.098294810681433,
+      "learning_rate": 3.2401363515011106e-10,
+      "loss": 0.7592,
+      "step": 12191
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.3957241265330365,
+      "learning_rate": 3.091186242248911e-10,
+      "loss": 0.6154,
+      "step": 12192
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.451104193699946,
+      "learning_rate": 2.9457406987121184e-10,
+      "loss": 0.788,
+      "step": 12193
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.835245723800428,
+      "learning_rate": 2.8037997310936814e-10,
+      "loss": 0.5518,
+      "step": 12194
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.925595085825402,
+      "learning_rate": 2.6653633493411993e-10,
+      "loss": 0.9264,
+      "step": 12195
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 5.9349646188925025,
+      "learning_rate": 2.530431563152469e-10,
+      "loss": 0.8856,
+      "step": 12196
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.2103321274378227,
+      "learning_rate": 2.3990043819976937e-10,
+      "loss": 0.5727,
+      "step": 12197
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 5.350349197288943,
+      "learning_rate": 2.2710818150750713e-10,
+      "loss": 0.981,
+      "step": 12198
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.35746844741046,
+      "learning_rate": 2.1466638713663057e-10,
+      "loss": 0.764,
+      "step": 12199
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.045052932146058,
+      "learning_rate": 2.0257505595810966e-10,
+      "loss": 0.7412,
+      "step": 12200
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.933136007973078,
+      "learning_rate": 1.908341888195997e-10,
+      "loss": 0.3473,
+      "step": 12201
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.204389975478312,
+      "learning_rate": 1.7944378654488616e-10,
+      "loss": 0.9525,
+      "step": 12202
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 5.066516739160856,
+      "learning_rate": 1.6840384993166425e-10,
+      "loss": 0.7245,
+      "step": 12203
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.003356870050669,
+      "learning_rate": 1.5771437975375948e-10,
+      "loss": 0.8505,
+      "step": 12204
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.6243068643927847,
+      "learning_rate": 1.4737537676112745e-10,
+      "loss": 0.561,
+      "step": 12205
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.980907710831678,
+      "learning_rate": 1.373868416776336e-10,
+      "loss": 0.7577,
+      "step": 12206
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.7313240058721955,
+      "learning_rate": 1.277487752043838e-10,
+      "loss": 0.6469,
+      "step": 12207
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.0864475947641568,
+      "learning_rate": 1.1846117801583846e-10,
+      "loss": 0.5824,
+      "step": 12208
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.742322727664848,
+      "learning_rate": 1.0952405076425365e-10,
+      "loss": 0.9012,
+      "step": 12209
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.228041571087677,
+      "learning_rate": 1.0093739407468495e-10,
+      "loss": 0.4204,
+      "step": 12210
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.0323303040850136,
+      "learning_rate": 9.270120855053855e-11,
+      "loss": 0.6893,
+      "step": 12211
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.983008160472688,
+      "learning_rate": 8.48154947680202e-11,
+      "loss": 0.6281,
+      "step": 12212
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.856370516222304,
+      "learning_rate": 7.728025328002098e-11,
+      "loss": 0.6949,
+      "step": 12213
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.861057508428135,
+      "learning_rate": 7.009548461500704e-11,
+      "loss": 0.5709,
+      "step": 12214
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.8994326876419683,
+      "learning_rate": 6.326118927646452e-11,
+      "loss": 0.3508,
+      "step": 12215
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 5.741387773543702,
+      "learning_rate": 5.677736774400977e-11,
+      "loss": 0.845,
+      "step": 12216
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.373884783945141,
+      "learning_rate": 5.0644020471168894e-11,
+      "loss": 0.5915,
+      "step": 12217
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 5.499579017923998,
+      "learning_rate": 4.4861147888708436e-11,
+      "loss": 1.2168,
+      "step": 12218
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.0255508426308597,
+      "learning_rate": 3.942875040130467e-11,
+      "loss": 0.4464,
+      "step": 12219
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 5.220051245827825,
+      "learning_rate": 3.4346828389764106e-11,
+      "loss": 0.8531,
+      "step": 12220
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.085008211334056,
+      "learning_rate": 2.961538221102345e-11,
+      "loss": 0.5081,
+      "step": 12221
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.9561086440767137,
+      "learning_rate": 2.523441219648426e-11,
+      "loss": 0.7413,
+      "step": 12222
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.4472799777661978,
+      "learning_rate": 2.1203918652568102e-11,
+      "loss": 0.5137,
+      "step": 12223
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 5.6581789617909575,
+      "learning_rate": 1.7523901862381844e-11,
+      "loss": 0.9722,
+      "step": 12224
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 5.314810540569065,
+      "learning_rate": 1.4194362084052338e-11,
+      "loss": 0.9416,
+      "step": 12225
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.561922239934296,
+      "learning_rate": 1.121529955017131e-11,
+      "loss": 0.4867,
+      "step": 12226
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.230862525679776,
+      "learning_rate": 8.586714470570912e-12,
+      "loss": 0.717,
+      "step": 12227
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.8264697118005886,
+      "learning_rate": 6.308607028993052e-12,
+      "loss": 0.8789,
+      "step": 12228
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.792263429259282,
+      "learning_rate": 4.3809773847547365e-12,
+      "loss": 0.7892,
+      "step": 12229
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 5.202397881453768,
+      "learning_rate": 2.803825673858285e-12,
+      "loss": 1.2797,
+      "step": 12230
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.416111011046913,
+      "learning_rate": 1.5771520062157764e-12,
+      "loss": 0.3239,
+      "step": 12231
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.859482561826479,
+      "learning_rate": 7.009564673143843e-13,
+      "loss": 1.079,
+      "step": 12232
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.819431466634126,
+      "learning_rate": 1.7523911988170939e-13,
+      "loss": 0.7518,
+      "step": 12233
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 6.337511901978712,
+      "learning_rate": 0.0,
+      "loss": 1.3796,
+      "step": 12234
+    },
+    {
+      "epoch": 1.0,
+      "step": 12234,
+      "total_flos": 1636528356546560.0,
+      "train_loss": 0.783344195976315,
+      "train_runtime": 55679.5489,
+      "train_samples_per_second": 1.758,
+      "train_steps_per_second": 0.22
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 12234,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 1000,
+  "total_flos": 1636528356546560.0,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}