diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,5732 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.0,
+  "eval_steps": 500,
+  "global_step": 814,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.002457002457002457,
+      "grad_norm": 12.175590846877855,
+      "learning_rate": 0.0,
+      "loss": 1.2148,
+      "step": 1
+    },
+    {
+      "epoch": 0.004914004914004914,
+      "grad_norm": 13.112846331209713,
+      "learning_rate": 5e-08,
+      "loss": 1.3154,
+      "step": 2
+    },
+    {
+      "epoch": 0.007371007371007371,
+      "grad_norm": 12.768616794783732,
+      "learning_rate": 1e-07,
+      "loss": 1.2865,
+      "step": 3
+    },
+    {
+      "epoch": 0.009828009828009828,
+      "grad_norm": 12.570106090577944,
+      "learning_rate": 1.5e-07,
+      "loss": 1.2459,
+      "step": 4
+    },
+    {
+      "epoch": 0.012285012285012284,
+      "grad_norm": 12.026844749086528,
+      "learning_rate": 2e-07,
+      "loss": 1.2496,
+      "step": 5
+    },
+    {
+      "epoch": 0.014742014742014743,
+      "grad_norm": 12.166619597781125,
+      "learning_rate": 2.5e-07,
+      "loss": 1.2179,
+      "step": 6
+    },
+    {
+      "epoch": 0.0171990171990172,
+      "grad_norm": 12.450604071787023,
+      "learning_rate": 3e-07,
+      "loss": 1.2451,
+      "step": 7
+    },
+    {
+      "epoch": 0.019656019656019656,
+      "grad_norm": 11.75536378463392,
+      "learning_rate": 3.5e-07,
+      "loss": 1.2372,
+      "step": 8
+    },
+    {
+      "epoch": 0.022113022113022112,
+      "grad_norm": 11.7494451205682,
+      "learning_rate": 4e-07,
+      "loss": 1.2174,
+      "step": 9
+    },
+    {
+      "epoch": 0.02457002457002457,
+      "grad_norm": 11.485164591676915,
+      "learning_rate": 4.5e-07,
+      "loss": 1.1803,
+      "step": 10
+    },
+    {
+      "epoch": 0.02702702702702703,
+      "grad_norm": 11.41671410777608,
+      "learning_rate": 5e-07,
+      "loss": 1.1829,
+      "step": 11
+    },
+    {
+      "epoch": 0.029484029484029485,
+      "grad_norm": 10.737969644794076,
+      "learning_rate": 5.5e-07,
+      "loss": 1.1912,
+      "step": 12
+    },
+    {
+      "epoch": 0.03194103194103194,
+      "grad_norm": 12.0315769410197,
+      "learning_rate": 6e-07,
+      "loss": 1.2516,
+      "step": 13
+    },
+    {
+      "epoch": 0.0343980343980344,
+      "grad_norm": 11.703183966857443,
+      "learning_rate": 6.5e-07,
+      "loss": 1.2533,
+      "step": 14
+    },
+    {
+      "epoch": 0.036855036855036855,
+      "grad_norm": 10.75777994461547,
+      "learning_rate": 7e-07,
+      "loss": 1.1763,
+      "step": 15
+    },
+    {
+      "epoch": 0.03931203931203931,
+      "grad_norm": 9.517567846059583,
+      "learning_rate": 7.5e-07,
+      "loss": 1.1474,
+      "step": 16
+    },
+    {
+      "epoch": 0.04176904176904177,
+      "grad_norm": 9.825919996727293,
+      "learning_rate": 8e-07,
+      "loss": 1.1725,
+      "step": 17
+    },
+    {
+      "epoch": 0.044226044226044224,
+      "grad_norm": 9.370499752545927,
+      "learning_rate": 8.499999999999999e-07,
+      "loss": 1.2048,
+      "step": 18
+    },
+    {
+      "epoch": 0.04668304668304668,
+      "grad_norm": 8.582608952468464,
+      "learning_rate": 9e-07,
+      "loss": 1.1428,
+      "step": 19
+    },
+    {
+      "epoch": 0.04914004914004914,
+      "grad_norm": 7.4140901371154,
+      "learning_rate": 9.499999999999999e-07,
+      "loss": 1.1583,
+      "step": 20
+    },
+    {
+      "epoch": 0.051597051597051594,
+      "grad_norm": 6.890519345341131,
+      "learning_rate": 1e-06,
+      "loss": 1.1961,
+      "step": 21
+    },
+    {
+      "epoch": 0.05405405405405406,
+      "grad_norm": 5.47854198998431,
+      "learning_rate": 1.05e-06,
+      "loss": 1.126,
+      "step": 22
+    },
+    {
+      "epoch": 0.056511056511056514,
+      "grad_norm": 5.350914917259245,
+      "learning_rate": 1.1e-06,
+      "loss": 1.1484,
+      "step": 23
+    },
+    {
+      "epoch": 0.05896805896805897,
+      "grad_norm": 4.114918765551221,
+      "learning_rate": 1.1499999999999998e-06,
+      "loss": 1.1328,
+      "step": 24
+    },
+    {
+      "epoch": 0.06142506142506143,
+      "grad_norm": 3.535376327350945,
+      "learning_rate": 1.2e-06,
+      "loss": 1.0889,
+      "step": 25
+    },
+    {
+      "epoch": 0.06388206388206388,
+      "grad_norm": 3.2480152784573586,
+      "learning_rate": 1.2499999999999999e-06,
+      "loss": 1.0437,
+      "step": 26
+    },
+    {
+      "epoch": 0.06633906633906633,
+      "grad_norm": 3.4795119249436293,
+      "learning_rate": 1.3e-06,
+      "loss": 1.0966,
+      "step": 27
+    },
+    {
+      "epoch": 0.0687960687960688,
+      "grad_norm": 2.8566181371095047,
+      "learning_rate": 1.35e-06,
+      "loss": 1.1234,
+      "step": 28
+    },
+    {
+      "epoch": 0.07125307125307126,
+      "grad_norm": 3.6578709990785048,
+      "learning_rate": 1.4e-06,
+      "loss": 1.1265,
+      "step": 29
+    },
+    {
+      "epoch": 0.07371007371007371,
+      "grad_norm": 3.541143081282278,
+      "learning_rate": 1.4499999999999999e-06,
+      "loss": 1.0543,
+      "step": 30
+    },
+    {
+      "epoch": 0.07616707616707617,
+      "grad_norm": 6.531880288467122,
+      "learning_rate": 1.5e-06,
+      "loss": 1.1109,
+      "step": 31
+    },
+    {
+      "epoch": 0.07862407862407862,
+      "grad_norm": 6.4715106690759,
+      "learning_rate": 1.55e-06,
+      "loss": 1.1257,
+      "step": 32
+    },
+    {
+      "epoch": 0.08108108108108109,
+      "grad_norm": 5.399614723171929,
+      "learning_rate": 1.6e-06,
+      "loss": 1.0694,
+      "step": 33
+    },
+    {
+      "epoch": 0.08353808353808354,
+      "grad_norm": 5.447520160063898,
+      "learning_rate": 1.6499999999999999e-06,
+      "loss": 1.047,
+      "step": 34
+    },
+    {
+      "epoch": 0.085995085995086,
+      "grad_norm": 15.759399235218536,
+      "learning_rate": 1.6999999999999998e-06,
+      "loss": 1.0684,
+      "step": 35
+    },
+    {
+      "epoch": 0.08845208845208845,
+      "grad_norm": 6.138108825882396,
+      "learning_rate": 1.75e-06,
+      "loss": 1.0568,
+      "step": 36
+    },
+    {
+      "epoch": 0.09090909090909091,
+      "grad_norm": 28.96303499581736,
+      "learning_rate": 1.8e-06,
+      "loss": 1.0748,
+      "step": 37
+    },
+    {
+      "epoch": 0.09336609336609336,
+      "grad_norm": 3.400779377353331,
+      "learning_rate": 1.85e-06,
+      "loss": 1.0841,
+      "step": 38
+    },
+    {
+      "epoch": 0.09582309582309582,
+      "grad_norm": 3.2088443423156305,
+      "learning_rate": 1.8999999999999998e-06,
+      "loss": 1.0637,
+      "step": 39
+    },
+    {
+      "epoch": 0.09828009828009827,
+      "grad_norm": 3.650839591876822,
+      "learning_rate": 1.95e-06,
+      "loss": 1.065,
+      "step": 40
+    },
+    {
+      "epoch": 0.10073710073710074,
+      "grad_norm": 2.48874031529166,
+      "learning_rate": 2e-06,
+      "loss": 1.0325,
+      "step": 41
+    },
+    {
+      "epoch": 0.10319410319410319,
+      "grad_norm": 2.4764261500155436,
+      "learning_rate": 1.9999917626554467e-06,
+      "loss": 1.0484,
+      "step": 42
+    },
+    {
+      "epoch": 0.10565110565110565,
+      "grad_norm": 2.0092623603336794,
+      "learning_rate": 1.9999670507574944e-06,
+      "loss": 1.0859,
+      "step": 43
+    },
+    {
+      "epoch": 0.10810810810810811,
+      "grad_norm": 1.915125996872455,
+      "learning_rate": 1.9999258647132643e-06,
+      "loss": 1.0423,
+      "step": 44
+    },
+    {
+      "epoch": 0.11056511056511056,
+      "grad_norm": 1.8459238588876408,
+      "learning_rate": 1.9998682052012837e-06,
+      "loss": 1.0686,
+      "step": 45
+    },
+    {
+      "epoch": 0.11302211302211303,
+      "grad_norm": 2.238673168744122,
+      "learning_rate": 1.9997940731714745e-06,
+      "loss": 1.0601,
+      "step": 46
+    },
+    {
+      "epoch": 0.11547911547911548,
+      "grad_norm": 4.306515287574078,
+      "learning_rate": 1.9997034698451393e-06,
+      "loss": 1.0367,
+      "step": 47
+    },
+    {
+      "epoch": 0.11793611793611794,
+      "grad_norm": 5.689973830908464,
+      "learning_rate": 1.9995963967149398e-06,
+      "loss": 1.0542,
+      "step": 48
+    },
+    {
+      "epoch": 0.12039312039312039,
+      "grad_norm": 2.1880376147294522,
+      "learning_rate": 1.999472855544872e-06,
+      "loss": 1.0635,
+      "step": 49
+    },
+    {
+      "epoch": 0.12285012285012285,
+      "grad_norm": 3.433335591384566,
+      "learning_rate": 1.999332848370239e-06,
+      "loss": 1.0323,
+      "step": 50
+    },
+    {
+      "epoch": 0.12530712530712532,
+      "grad_norm": 1.8169785229927564,
+      "learning_rate": 1.9991763774976158e-06,
+      "loss": 1.0218,
+      "step": 51
+    },
+    {
+      "epoch": 0.12776412776412777,
+      "grad_norm": 1.8670733937364128,
+      "learning_rate": 1.9990034455048096e-06,
+      "loss": 1.0405,
+      "step": 52
+    },
+    {
+      "epoch": 0.13022113022113022,
+      "grad_norm": 2.439448369982022,
+      "learning_rate": 1.998814055240823e-06,
+      "loss": 1.0519,
+      "step": 53
+    },
+    {
+      "epoch": 0.13267813267813267,
+      "grad_norm": 1.5940089652939573,
+      "learning_rate": 1.9986082098258006e-06,
+      "loss": 1.0359,
+      "step": 54
+    },
+    {
+      "epoch": 0.13513513513513514,
+      "grad_norm": 1.7404888899145354,
+      "learning_rate": 1.9983859126509824e-06,
+      "loss": 1.0087,
+      "step": 55
+    },
+    {
+      "epoch": 0.1375921375921376,
+      "grad_norm": 2.2476917865688026,
+      "learning_rate": 1.998147167378645e-06,
+      "loss": 1.0138,
+      "step": 56
+    },
+    {
+      "epoch": 0.14004914004914004,
+      "grad_norm": 1.8002247115537675,
+      "learning_rate": 1.9978919779420425e-06,
+      "loss": 1.0488,
+      "step": 57
+    },
+    {
+      "epoch": 0.14250614250614252,
+      "grad_norm": 1.663251752279236,
+      "learning_rate": 1.9976203485453413e-06,
+      "loss": 0.9809,
+      "step": 58
+    },
+    {
+      "epoch": 0.14496314496314497,
+      "grad_norm": 1.7653797682084729,
+      "learning_rate": 1.9973322836635515e-06,
+      "loss": 1.0686,
+      "step": 59
+    },
+    {
+      "epoch": 0.14742014742014742,
+      "grad_norm": 1.5336222427197481,
+      "learning_rate": 1.9970277880424526e-06,
+      "loss": 0.986,
+      "step": 60
+    },
+    {
+      "epoch": 0.14987714987714987,
+      "grad_norm": 1.4421704567479543,
+      "learning_rate": 1.9967068666985146e-06,
+      "loss": 0.9948,
+      "step": 61
+    },
+    {
+      "epoch": 0.15233415233415235,
+      "grad_norm": 1.4042358001427657,
+      "learning_rate": 1.996369524918818e-06,
+      "loss": 0.967,
+      "step": 62
+    },
+    {
+      "epoch": 0.1547911547911548,
+      "grad_norm": 1.4295825578558987,
+      "learning_rate": 1.9960157682609633e-06,
+      "loss": 0.9723,
+      "step": 63
+    },
+    {
+      "epoch": 0.15724815724815724,
+      "grad_norm": 1.6093180146467703,
+      "learning_rate": 1.9956456025529804e-06,
+      "loss": 0.9737,
+      "step": 64
+    },
+    {
+      "epoch": 0.1597051597051597,
+      "grad_norm": 1.7273275051707897,
+      "learning_rate": 1.9952590338932358e-06,
+      "loss": 0.9935,
+      "step": 65
+    },
+    {
+      "epoch": 0.16216216216216217,
+      "grad_norm": 1.647950655242118,
+      "learning_rate": 1.994856068650327e-06,
+      "loss": 1.0509,
+      "step": 66
+    },
+    {
+      "epoch": 0.16461916461916462,
+      "grad_norm": 1.6052330571582885,
+      "learning_rate": 1.994436713462982e-06,
+      "loss": 1.0295,
+      "step": 67
+    },
+    {
+      "epoch": 0.16707616707616707,
+      "grad_norm": 1.694905424784479,
+      "learning_rate": 1.994000975239946e-06,
+      "loss": 1.0256,
+      "step": 68
+    },
+    {
+      "epoch": 0.16953316953316952,
+      "grad_norm": 1.5039835360361733,
+      "learning_rate": 1.9935488611598713e-06,
+      "loss": 1.0347,
+      "step": 69
+    },
+    {
+      "epoch": 0.171990171990172,
+      "grad_norm": 1.734126157202943,
+      "learning_rate": 1.9930803786711967e-06,
+      "loss": 1.023,
+      "step": 70
+    },
+    {
+      "epoch": 0.17444717444717445,
+      "grad_norm": 1.5242931763881968,
+      "learning_rate": 1.9925955354920264e-06,
+      "loss": 1.0406,
+      "step": 71
+    },
+    {
+      "epoch": 0.1769041769041769,
+      "grad_norm": 1.606370733528827,
+      "learning_rate": 1.99209433961e-06,
+      "loss": 1.0222,
+      "step": 72
+    },
+    {
+      "epoch": 0.17936117936117937,
+      "grad_norm": 2.4762050432839318,
+      "learning_rate": 1.991576799282164e-06,
+      "loss": 1.0278,
+      "step": 73
+    },
+    {
+      "epoch": 0.18181818181818182,
+      "grad_norm": 1.4747727870105878,
+      "learning_rate": 1.9910429230348343e-06,
+      "loss": 0.9779,
+      "step": 74
+    },
+    {
+      "epoch": 0.18427518427518427,
+      "grad_norm": 2.0977776864765167,
+      "learning_rate": 1.990492719663457e-06,
+      "loss": 1.0235,
+      "step": 75
+    },
+    {
+      "epoch": 0.18673218673218672,
+      "grad_norm": 1.9728896794221442,
+      "learning_rate": 1.9899261982324607e-06,
+      "loss": 0.9926,
+      "step": 76
+    },
+    {
+      "epoch": 0.1891891891891892,
+      "grad_norm": 1.7736978075188574,
+      "learning_rate": 1.98934336807511e-06,
+      "loss": 1.0354,
+      "step": 77
+    },
+    {
+      "epoch": 0.19164619164619165,
+      "grad_norm": 2.1716440393079464,
+      "learning_rate": 1.988744238793351e-06,
+      "loss": 0.977,
+      "step": 78
+    },
+    {
+      "epoch": 0.1941031941031941,
+      "grad_norm": 1.599345820298087,
+      "learning_rate": 1.9881288202576518e-06,
+      "loss": 0.9685,
+      "step": 79
+    },
+    {
+      "epoch": 0.19656019656019655,
+      "grad_norm": 1.4582630785882413,
+      "learning_rate": 1.9874971226068415e-06,
+      "loss": 0.943,
+      "step": 80
+    },
+    {
+      "epoch": 0.19901719901719903,
+      "grad_norm": 1.4275025088638074,
+      "learning_rate": 1.9868491562479424e-06,
+      "loss": 1.0295,
+      "step": 81
+    },
+    {
+      "epoch": 0.20147420147420148,
+      "grad_norm": 1.609247365977497,
+      "learning_rate": 1.9861849318559994e-06,
+      "loss": 0.9531,
+      "step": 82
+    },
+    {
+      "epoch": 0.20393120393120392,
+      "grad_norm": 1.711398906589491,
+      "learning_rate": 1.9855044603739028e-06,
+      "loss": 1.0302,
+      "step": 83
+    },
+    {
+      "epoch": 0.20638820638820637,
+      "grad_norm": 1.56099328882278,
+      "learning_rate": 1.984807753012208e-06,
+      "loss": 0.9684,
+      "step": 84
+    },
+    {
+      "epoch": 0.20884520884520885,
+      "grad_norm": 1.5428758027146767,
+      "learning_rate": 1.9840948212489527e-06,
+      "loss": 1.0129,
+      "step": 85
+    },
+    {
+      "epoch": 0.2113022113022113,
+      "grad_norm": 1.9669978126079752,
+      "learning_rate": 1.983365676829466e-06,
+      "loss": 0.9976,
+      "step": 86
+    },
+    {
+      "epoch": 0.21375921375921375,
+      "grad_norm": 1.5144216325008544,
+      "learning_rate": 1.9826203317661756e-06,
+      "loss": 0.9967,
+      "step": 87
+    },
+    {
+      "epoch": 0.21621621621621623,
+      "grad_norm": 2.1395022590018367,
+      "learning_rate": 1.9818587983384094e-06,
+      "loss": 0.9693,
+      "step": 88
+    },
+    {
+      "epoch": 0.21867321867321868,
+      "grad_norm": 1.5091268331439933,
+      "learning_rate": 1.981081089092194e-06,
+      "loss": 0.9271,
+      "step": 89
+    },
+    {
+      "epoch": 0.22113022113022113,
+      "grad_norm": 1.8173026339961775,
+      "learning_rate": 1.9802872168400476e-06,
+      "loss": 1.0819,
+      "step": 90
+    },
+    {
+      "epoch": 0.22358722358722358,
+      "grad_norm": 1.5986438761107704,
+      "learning_rate": 1.979477194660769e-06,
+      "loss": 0.9594,
+      "step": 91
+    },
+    {
+      "epoch": 0.22604422604422605,
+      "grad_norm": 1.5027188037219825,
+      "learning_rate": 1.978651035899221e-06,
+      "loss": 0.9964,
+      "step": 92
+    },
+    {
+      "epoch": 0.2285012285012285,
+      "grad_norm": 1.3856569445993616,
+      "learning_rate": 1.977808754166113e-06,
+      "loss": 0.9476,
+      "step": 93
+    },
+    {
+      "epoch": 0.23095823095823095,
+      "grad_norm": 1.445303089718021,
+      "learning_rate": 1.976950363337774e-06,
+      "loss": 0.9664,
+      "step": 94
+    },
+    {
+      "epoch": 0.2334152334152334,
+      "grad_norm": 1.9267867671090184,
+      "learning_rate": 1.9760758775559273e-06,
+      "loss": 1.0446,
+      "step": 95
+    },
+    {
+      "epoch": 0.23587223587223588,
+      "grad_norm": 1.4687411243380588,
+      "learning_rate": 1.975185311227453e-06,
+      "loss": 0.9483,
+      "step": 96
+    },
+    {
+      "epoch": 0.23832923832923833,
+      "grad_norm": 1.5489115105120785,
+      "learning_rate": 1.9742786790241546e-06,
+      "loss": 1.0073,
+      "step": 97
+    },
+    {
+      "epoch": 0.24078624078624078,
+      "grad_norm": 2.5463794252310676,
+      "learning_rate": 1.9733559958825166e-06,
+      "loss": 1.0046,
+      "step": 98
+    },
+    {
+      "epoch": 0.24324324324324326,
+      "grad_norm": 1.56125807056489,
+      "learning_rate": 1.9724172770034566e-06,
+      "loss": 0.971,
+      "step": 99
+    },
+    {
+      "epoch": 0.2457002457002457,
+      "grad_norm": 1.740701138818393,
+      "learning_rate": 1.971462537852076e-06,
+      "loss": 0.9856,
+      "step": 100
+    },
+    {
+      "epoch": 0.24815724815724816,
+      "grad_norm": 1.4470402917114091,
+      "learning_rate": 1.970491794157405e-06,
+      "loss": 0.9897,
+      "step": 101
+    },
+    {
+      "epoch": 0.25061425061425063,
+      "grad_norm": 1.5706400471340238,
+      "learning_rate": 1.9695050619121456e-06,
+      "loss": 0.9328,
+      "step": 102
+    },
+    {
+      "epoch": 0.25307125307125306,
+      "grad_norm": 1.9773708156391945,
+      "learning_rate": 1.9685023573724034e-06,
+      "loss": 0.9719,
+      "step": 103
+    },
+    {
+      "epoch": 0.25552825552825553,
+      "grad_norm": 1.9195746558098163,
+      "learning_rate": 1.9674836970574253e-06,
+      "loss": 0.978,
+      "step": 104
+    },
+    {
+      "epoch": 0.257985257985258,
+      "grad_norm": 1.8027098570962006,
+      "learning_rate": 1.966449097749322e-06,
+      "loss": 1.0023,
+      "step": 105
+    },
+    {
+      "epoch": 0.26044226044226043,
+      "grad_norm": 1.7805523167150648,
+      "learning_rate": 1.965398576492796e-06,
+      "loss": 0.9768,
+      "step": 106
+    },
+    {
+      "epoch": 0.2628992628992629,
+      "grad_norm": 1.9751836869745634,
+      "learning_rate": 1.9643321505948585e-06,
+      "loss": 1.0132,
+      "step": 107
+    },
+    {
+      "epoch": 0.26535626535626533,
+      "grad_norm": 1.6135714113454362,
+      "learning_rate": 1.9632498376245444e-06,
+      "loss": 1.0169,
+      "step": 108
+    },
+    {
+      "epoch": 0.2678132678132678,
+      "grad_norm": 1.6366575829852479,
+      "learning_rate": 1.9621516554126237e-06,
+      "loss": 0.9618,
+      "step": 109
+    },
+    {
+      "epoch": 0.2702702702702703,
+      "grad_norm": 1.4439243574845881,
+      "learning_rate": 1.9610376220513063e-06,
+      "loss": 1.0111,
+      "step": 110
+    },
+    {
+      "epoch": 0.2727272727272727,
+      "grad_norm": 1.574054550140938,
+      "learning_rate": 1.9599077558939466e-06,
+      "loss": 0.9979,
+      "step": 111
+    },
+    {
+      "epoch": 0.2751842751842752,
+      "grad_norm": 1.9768484700868765,
+      "learning_rate": 1.958762075554737e-06,
+      "loss": 1.0266,
+      "step": 112
+    },
+    {
+      "epoch": 0.27764127764127766,
+      "grad_norm": 1.6821288693476542,
+      "learning_rate": 1.957600599908406e-06,
+      "loss": 0.983,
+      "step": 113
+    },
+    {
+      "epoch": 0.2800982800982801,
+      "grad_norm": 1.6250908693533597,
+      "learning_rate": 1.956423348089903e-06,
+      "loss": 0.9735,
+      "step": 114
+    },
+    {
+      "epoch": 0.28255528255528256,
+      "grad_norm": 1.5158186810772534,
+      "learning_rate": 1.9552303394940857e-06,
+      "loss": 0.9767,
+      "step": 115
+    },
+    {
+      "epoch": 0.28501228501228504,
+      "grad_norm": 1.7591914283817052,
+      "learning_rate": 1.9540215937754004e-06,
+      "loss": 0.9515,
+      "step": 116
+    },
+    {
+      "epoch": 0.28746928746928746,
+      "grad_norm": 1.329980995859771,
+      "learning_rate": 1.952797130847557e-06,
+      "loss": 0.944,
+      "step": 117
+    },
+    {
+      "epoch": 0.28992628992628994,
+      "grad_norm": 1.7122579873222803,
+      "learning_rate": 1.951556970883201e-06,
+      "loss": 1.0098,
+      "step": 118
+    },
+    {
+      "epoch": 0.29238329238329236,
+      "grad_norm": 1.8427163819779844,
+      "learning_rate": 1.9503011343135825e-06,
+      "loss": 0.9815,
+      "step": 119
+    },
+    {
+      "epoch": 0.29484029484029484,
+      "grad_norm": 1.3722808372435795,
+      "learning_rate": 1.9490296418282184e-06,
+      "loss": 0.9914,
+      "step": 120
+    },
+    {
+      "epoch": 0.2972972972972973,
+      "grad_norm": 1.2858110734745247,
+      "learning_rate": 1.9477425143745523e-06,
+      "loss": 0.9727,
+      "step": 121
+    },
+    {
+      "epoch": 0.29975429975429974,
+      "grad_norm": 1.3478569354201173,
+      "learning_rate": 1.9464397731576094e-06,
+      "loss": 0.9557,
+      "step": 122
+    },
+    {
+      "epoch": 0.3022113022113022,
+      "grad_norm": 1.534212309472206,
+      "learning_rate": 1.945121439639645e-06,
+      "loss": 0.9995,
+      "step": 123
+    },
+    {
+      "epoch": 0.3046683046683047,
+      "grad_norm": 1.4583668620611645,
+      "learning_rate": 1.943787535539795e-06,
+      "loss": 0.9244,
+      "step": 124
+    },
+    {
+      "epoch": 0.3071253071253071,
+      "grad_norm": 1.3048849492711914,
+      "learning_rate": 1.9424380828337143e-06,
+      "loss": 0.9749,
+      "step": 125
+    },
+    {
+      "epoch": 0.3095823095823096,
+      "grad_norm": 1.9214890498265684,
+      "learning_rate": 1.9410731037532167e-06,
+      "loss": 1.0126,
+      "step": 126
+    },
+    {
+      "epoch": 0.31203931203931207,
+      "grad_norm": 1.492998355230107,
+      "learning_rate": 1.9396926207859082e-06,
+      "loss": 0.9216,
+      "step": 127
+    },
+    {
+      "epoch": 0.3144963144963145,
+      "grad_norm": 1.529088663498555,
+      "learning_rate": 1.9382966566748166e-06,
+      "loss": 1.0266,
+      "step": 128
+    },
+    {
+      "epoch": 0.31695331695331697,
+      "grad_norm": 1.4961622214991601,
+      "learning_rate": 1.9368852344180167e-06,
+      "loss": 0.949,
+      "step": 129
+    },
+    {
+      "epoch": 0.3194103194103194,
+      "grad_norm": 1.5657840764329438,
+      "learning_rate": 1.935458377268251e-06,
+      "loss": 0.9768,
+      "step": 130
+    },
+    {
+      "epoch": 0.32186732186732187,
+      "grad_norm": 1.458088079896784,
+      "learning_rate": 1.934016108732548e-06,
+      "loss": 0.9831,
+      "step": 131
+    },
+    {
+      "epoch": 0.32432432432432434,
+      "grad_norm": 1.5704249898970843,
+      "learning_rate": 1.932558452571833e-06,
+      "loss": 0.9845,
+      "step": 132
+    },
+    {
+      "epoch": 0.32678132678132676,
+      "grad_norm": 1.366875012537187,
+      "learning_rate": 1.9310854328005377e-06,
+      "loss": 0.9751,
+      "step": 133
+    },
+    {
+      "epoch": 0.32923832923832924,
+      "grad_norm": 1.6682519282454646,
+      "learning_rate": 1.9295970736862064e-06,
+      "loss": 0.9278,
+      "step": 134
+    },
+    {
+      "epoch": 0.3316953316953317,
+      "grad_norm": 1.7010058864881312,
+      "learning_rate": 1.928093399749091e-06,
+      "loss": 0.9479,
+      "step": 135
+    },
+    {
+      "epoch": 0.33415233415233414,
+      "grad_norm": 2.774339563092048,
+      "learning_rate": 1.926574435761753e-06,
+      "loss": 0.9216,
+      "step": 136
+    },
+    {
+      "epoch": 0.3366093366093366,
+      "grad_norm": 1.5996469164653206,
+      "learning_rate": 1.925040206748652e-06,
+      "loss": 0.9749,
+      "step": 137
+    },
+    {
+      "epoch": 0.33906633906633904,
+      "grad_norm": 1.6140642370600065,
+      "learning_rate": 1.9234907379857334e-06,
+      "loss": 1.0195,
+      "step": 138
+    },
+    {
+      "epoch": 0.3415233415233415,
+      "grad_norm": 1.6724211149688903,
+      "learning_rate": 1.9219260550000143e-06,
+      "loss": 1.0086,
+      "step": 139
+    },
+    {
+      "epoch": 0.343980343980344,
+      "grad_norm": 1.6697054307200352,
+      "learning_rate": 1.920346183569159e-06,
+      "loss": 1.0177,
+      "step": 140
+    },
+    {
+      "epoch": 0.3464373464373464,
+      "grad_norm": 1.5507362171936665,
+      "learning_rate": 1.91875114972106e-06,
+      "loss": 0.9701,
+      "step": 141
+    },
+    {
+      "epoch": 0.3488943488943489,
+      "grad_norm": 1.572772556059398,
+      "learning_rate": 1.9171409797334026e-06,
+      "loss": 0.9893,
+      "step": 142
+    },
+    {
+      "epoch": 0.35135135135135137,
+      "grad_norm": 1.4851773686144312,
+      "learning_rate": 1.9155157001332373e-06,
+      "loss": 0.9617,
+      "step": 143
+    },
+    {
+      "epoch": 0.3538083538083538,
+      "grad_norm": 1.3935370189220326,
+      "learning_rate": 1.91387533769654e-06,
+      "loss": 0.9647,
+      "step": 144
+    },
+    {
+      "epoch": 0.35626535626535627,
+      "grad_norm": 2.086032038197472,
+      "learning_rate": 1.912219919447772e-06,
+      "loss": 1.0138,
+      "step": 145
+    },
+    {
+      "epoch": 0.35872235872235875,
+      "grad_norm": 1.8988521674993861,
+      "learning_rate": 1.9105494726594343e-06,
+      "loss": 0.9619,
+      "step": 146
+    },
+    {
+      "epoch": 0.36117936117936117,
+      "grad_norm": 1.7707337422933225,
+      "learning_rate": 1.9088640248516184e-06,
+      "loss": 0.9689,
+      "step": 147
+    },
+    {
+      "epoch": 0.36363636363636365,
+      "grad_norm": 10.58019032383696,
+      "learning_rate": 1.9071636037915533e-06,
+      "loss": 0.9886,
+      "step": 148
+    },
+    {
+      "epoch": 0.36609336609336607,
+      "grad_norm": 1.6326349036691947,
+      "learning_rate": 1.9054482374931466e-06,
+      "loss": 0.9928,
+      "step": 149
+    },
+    {
+      "epoch": 0.36855036855036855,
+      "grad_norm": 1.4602979723959129,
+      "learning_rate": 1.9037179542165251e-06,
+      "loss": 0.9451,
+      "step": 150
+    },
+    {
+      "epoch": 0.371007371007371,
+      "grad_norm": 1.5983113709927816,
+      "learning_rate": 1.9019727824675682e-06,
+      "loss": 0.9488,
+      "step": 151
+    },
+    {
+      "epoch": 0.37346437346437344,
+      "grad_norm": 1.394284340663908,
+      "learning_rate": 1.9002127509974373e-06,
+      "loss": 0.9717,
+      "step": 152
+    },
+    {
+      "epoch": 0.3759213759213759,
+      "grad_norm": 1.856689232868256,
+      "learning_rate": 1.8984378888021041e-06,
+      "loss": 0.9286,
+      "step": 153
+    },
+    {
+      "epoch": 0.3783783783783784,
+      "grad_norm": 1.5404154844189102,
+      "learning_rate": 1.8966482251218714e-06,
+      "loss": 0.9625,
+      "step": 154
+    },
+    {
+      "epoch": 0.3808353808353808,
+      "grad_norm": 1.5417688714414741,
+      "learning_rate": 1.8948437894408916e-06,
+      "loss": 0.9314,
+      "step": 155
+    },
+    {
+      "epoch": 0.3832923832923833,
+      "grad_norm": 1.8457706124864284,
+      "learning_rate": 1.8930246114866822e-06,
+      "loss": 1.0147,
+      "step": 156
+    },
+    {
+      "epoch": 0.3857493857493858,
+      "grad_norm": 1.4971057129868792,
+      "learning_rate": 1.8911907212296341e-06,
+      "loss": 0.9309,
+      "step": 157
+    },
+    {
+      "epoch": 0.3882063882063882,
+      "grad_norm": 1.6347670467700182,
+      "learning_rate": 1.8893421488825188e-06,
+      "loss": 0.9621,
+      "step": 158
+    },
+    {
+      "epoch": 0.3906633906633907,
+      "grad_norm": 1.6324227560172353,
+      "learning_rate": 1.8874789248999915e-06,
+      "loss": 0.9772,
+      "step": 159
+    },
+    {
+      "epoch": 0.3931203931203931,
+      "grad_norm": 1.6267750782874413,
+      "learning_rate": 1.885601079978088e-06,
+      "loss": 0.9861,
+      "step": 160
+    },
+    {
+      "epoch": 0.3955773955773956,
+      "grad_norm": 1.4406136327704948,
+      "learning_rate": 1.8837086450537192e-06,
+      "loss": 0.9406,
+      "step": 161
+    },
+    {
+      "epoch": 0.39803439803439805,
+      "grad_norm": 1.5622857080744104,
+      "learning_rate": 1.8818016513041624e-06,
+      "loss": 0.94,
+      "step": 162
+    },
+    {
+      "epoch": 0.4004914004914005,
+      "grad_norm": 1.7312279395333572,
+      "learning_rate": 1.8798801301465468e-06,
+      "loss": 0.96,
+      "step": 163
+    },
+    {
+      "epoch": 0.40294840294840295,
+      "grad_norm": 1.5736688729195003,
+      "learning_rate": 1.877944113237336e-06,
+      "loss": 0.9793,
+      "step": 164
+    },
+    {
+      "epoch": 0.40540540540540543,
+      "grad_norm": 1.501899237192728,
+      "learning_rate": 1.8759936324718066e-06,
+      "loss": 0.9178,
+      "step": 165
+    },
+    {
+      "epoch": 0.40786240786240785,
+      "grad_norm": 1.3483386139829852,
+      "learning_rate": 1.874028719983523e-06,
+      "loss": 0.9593,
+      "step": 166
+    },
+    {
+      "epoch": 0.4103194103194103,
+      "grad_norm": 1.5326296765065217,
+      "learning_rate": 1.8720494081438077e-06,
+      "loss": 0.9794,
+      "step": 167
+    },
+    {
+      "epoch": 0.41277641277641275,
+      "grad_norm": 1.5759774818522712,
+      "learning_rate": 1.870055729561207e-06,
+      "loss": 0.9418,
+      "step": 168
+    },
+    {
+      "epoch": 0.4152334152334152,
+      "grad_norm": 2.475109331033288,
+      "learning_rate": 1.8680477170809572e-06,
+      "loss": 0.9933,
+      "step": 169
+    },
+    {
+      "epoch": 0.4176904176904177,
+      "grad_norm": 1.4136668445270784,
+      "learning_rate": 1.8660254037844386e-06,
+      "loss": 0.9287,
+      "step": 170
+    },
+    {
+      "epoch": 0.4201474201474201,
+      "grad_norm": 1.7485785437438173,
+      "learning_rate": 1.863988822988634e-06,
+      "loss": 0.9749,
+      "step": 171
+    },
+    {
+      "epoch": 0.4226044226044226,
+      "grad_norm": 1.825009421972696,
+      "learning_rate": 1.8619380082455796e-06,
+      "loss": 0.9637,
+      "step": 172
+    },
+    {
+      "epoch": 0.4250614250614251,
+      "grad_norm": 1.6686645342967068,
+      "learning_rate": 1.85987299334181e-06,
+      "loss": 0.9951,
+      "step": 173
+    },
+    {
+      "epoch": 0.4275184275184275,
+      "grad_norm": 1.9731456114498873,
+      "learning_rate": 1.857793812297804e-06,
+      "loss": 0.9426,
+      "step": 174
+    },
+    {
+      "epoch": 0.42997542997543,
+      "grad_norm": 1.5558317172807916,
+      "learning_rate": 1.855700499367423e-06,
+      "loss": 0.9787,
+      "step": 175
+    },
+    {
+      "epoch": 0.43243243243243246,
+      "grad_norm": 2.098470725111635,
+      "learning_rate": 1.8535930890373463e-06,
+      "loss": 0.9936,
+      "step": 176
+    },
+    {
+      "epoch": 0.4348894348894349,
+      "grad_norm": 1.525785332653322,
+      "learning_rate": 1.8514716160265044e-06,
+      "loss": 0.9727,
+      "step": 177
+    },
+    {
+      "epoch": 0.43734643734643736,
+      "grad_norm": 1.4913555879508107,
+      "learning_rate": 1.8493361152855056e-06,
+      "loss": 0.9748,
+      "step": 178
+    },
+    {
+      "epoch": 0.4398034398034398,
+      "grad_norm": 1.340195087338962,
+      "learning_rate": 1.84718662199606e-06,
+      "loss": 0.9706,
+      "step": 179
+    },
+    {
+      "epoch": 0.44226044226044225,
+      "grad_norm": 2.5585128719755867,
+      "learning_rate": 1.8450231715704026e-06,
+      "loss": 0.9467,
+      "step": 180
+    },
+    {
+      "epoch": 0.44471744471744473,
+      "grad_norm": 1.3066387641923265,
+      "learning_rate": 1.8428457996507053e-06,
+      "loss": 0.9194,
+      "step": 181
+    },
+    {
+      "epoch": 0.44717444717444715,
+      "grad_norm": 1.6465015301936612,
+      "learning_rate": 1.8406545421084938e-06,
+      "loss": 1.0045,
+      "step": 182
+    },
+    {
+      "epoch": 0.44963144963144963,
+      "grad_norm": 1.4122218928192685,
+      "learning_rate": 1.8384494350440552e-06,
+      "loss": 0.9596,
+      "step": 183
+    },
+    {
+      "epoch": 0.4520884520884521,
+      "grad_norm": 1.4364915893717223,
+      "learning_rate": 1.8362305147858427e-06,
+      "loss": 0.9375,
+      "step": 184
+    },
+    {
+      "epoch": 0.45454545454545453,
+      "grad_norm": 1.5320964201665865,
+      "learning_rate": 1.833997817889878e-06,
+      "loss": 0.98,
+      "step": 185
+    },
+    {
+      "epoch": 0.457002457002457,
+      "grad_norm": 1.8963928416673594,
+      "learning_rate": 1.8317513811391477e-06,
+      "loss": 0.9499,
+      "step": 186
+    },
+    {
+      "epoch": 0.4594594594594595,
+      "grad_norm": 1.4450513723222074,
+      "learning_rate": 1.8294912415429993e-06,
+      "loss": 0.9387,
+      "step": 187
+    },
+    {
+      "epoch": 0.4619164619164619,
+      "grad_norm": 1.4227834836237356,
+      "learning_rate": 1.8272174363365298e-06,
+      "loss": 0.9856,
+      "step": 188
+    },
+    {
+      "epoch": 0.4643734643734644,
+      "grad_norm": 1.4225544869881186,
+      "learning_rate": 1.8249300029799732e-06,
+      "loss": 0.9201,
+      "step": 189
+    },
+    {
+      "epoch": 0.4668304668304668,
+      "grad_norm": 1.6580547740077536,
+      "learning_rate": 1.822628979158083e-06,
+      "loss": 0.9746,
+      "step": 190
+    },
+    {
+      "epoch": 0.4692874692874693,
+      "grad_norm": 1.6225782675239633,
+      "learning_rate": 1.8203144027795108e-06,
+      "loss": 0.9493,
+      "step": 191
+    },
+    {
+      "epoch": 0.47174447174447176,
+      "grad_norm": 1.8558149480612793,
+      "learning_rate": 1.8179863119761834e-06,
+      "loss": 0.9551,
+      "step": 192
+    },
+    {
+      "epoch": 0.4742014742014742,
+      "grad_norm": 1.5948966144212158,
+      "learning_rate": 1.815644745102673e-06,
+      "loss": 0.9799,
+      "step": 193
+    },
+    {
+      "epoch": 0.47665847665847666,
+      "grad_norm": 1.464419249562776,
+      "learning_rate": 1.8132897407355654e-06,
+      "loss": 0.9634,
+      "step": 194
+    },
+    {
+      "epoch": 0.47911547911547914,
+      "grad_norm": 1.6709063152230734,
+      "learning_rate": 1.8109213376728257e-06,
+      "loss": 0.977,
+      "step": 195
+    },
+    {
+      "epoch": 0.48157248157248156,
+      "grad_norm": 1.4980373204140025,
+      "learning_rate": 1.8085395749331579e-06,
+      "loss": 0.9839,
+      "step": 196
+    },
+    {
+      "epoch": 0.48402948402948404,
+      "grad_norm": 1.4728568491372855,
+      "learning_rate": 1.8061444917553627e-06,
+      "loss": 0.9791,
+      "step": 197
+    },
+    {
+      "epoch": 0.4864864864864865,
+      "grad_norm": 1.3529262817353658,
+      "learning_rate": 1.803736127597691e-06,
+      "loss": 0.9534,
+      "step": 198
+    },
+    {
+      "epoch": 0.48894348894348894,
+      "grad_norm": 1.4726330094781561,
+      "learning_rate": 1.8013145221371935e-06,
+      "loss": 0.9397,
+      "step": 199
+    },
+    {
+      "epoch": 0.4914004914004914,
+      "grad_norm": 1.4468557644109676,
+      "learning_rate": 1.7988797152690668e-06,
+      "loss": 0.8993,
+      "step": 200
+    },
+    {
+      "epoch": 0.49385749385749383,
+      "grad_norm": 1.39576948942371,
+      "learning_rate": 1.796431747105998e-06,
+      "loss": 0.945,
+      "step": 201
+    },
+    {
+      "epoch": 0.4963144963144963,
+      "grad_norm": 1.4248364851440443,
+      "learning_rate": 1.793970657977501e-06,
+      "loss": 0.9669,
+      "step": 202
+    },
+    {
+      "epoch": 0.4987714987714988,
+      "grad_norm": 1.4850571599264226,
+      "learning_rate": 1.791496488429254e-06,
+      "loss": 0.9572,
+      "step": 203
+    },
+    {
+      "epoch": 0.5012285012285013,
+      "grad_norm": 1.4619304758472125,
+      "learning_rate": 1.7890092792224313e-06,
+      "loss": 0.9604,
+      "step": 204
+    },
+    {
+      "epoch": 0.5036855036855037,
+      "grad_norm": 1.6208674632812228,
+      "learning_rate": 1.786509071333031e-06,
+      "loss": 0.9577,
+      "step": 205
+    },
+    {
+      "epoch": 0.5061425061425061,
+      "grad_norm": 1.4904987175582993,
+      "learning_rate": 1.7839959059512013e-06,
+      "loss": 0.9039,
+      "step": 206
+    },
+    {
+      "epoch": 0.5085995085995086,
+      "grad_norm": 1.4636256072722436,
+      "learning_rate": 1.7814698244805601e-06,
+      "loss": 0.9481,
+      "step": 207
+    },
+    {
+      "epoch": 0.5110565110565111,
+      "grad_norm": 1.3573990171789403,
+      "learning_rate": 1.7789308685375146e-06,
+      "loss": 0.9531,
+      "step": 208
+    },
+    {
+      "epoch": 0.5135135135135135,
+      "grad_norm": 1.4176202461890393,
+      "learning_rate": 1.7763790799505745e-06,
+      "loss": 0.959,
+      "step": 209
+    },
+    {
+      "epoch": 0.515970515970516,
+      "grad_norm": 1.5707837841315286,
+      "learning_rate": 1.773814500759663e-06,
+      "loss": 0.9798,
+      "step": 210
+    },
+    {
+      "epoch": 0.5184275184275184,
+      "grad_norm": 1.4706701895149787,
+      "learning_rate": 1.7712371732154256e-06,
+      "loss": 0.9928,
+      "step": 211
+    },
+    {
+      "epoch": 0.5208845208845209,
+      "grad_norm": 2.069991518090132,
+      "learning_rate": 1.7686471397785318e-06,
+      "loss": 1.0153,
+      "step": 212
+    },
+    {
+      "epoch": 0.5233415233415234,
+      "grad_norm": 1.443867260475285,
+      "learning_rate": 1.766044443118978e-06,
+      "loss": 0.9452,
+      "step": 213
+    },
+    {
+      "epoch": 0.5257985257985258,
+      "grad_norm": 1.5151736278472239,
+      "learning_rate": 1.7634291261153817e-06,
+      "loss": 0.9114,
+      "step": 214
+    },
+    {
+      "epoch": 0.5282555282555282,
+      "grad_norm": 1.3847804036493465,
+      "learning_rate": 1.7608012318542777e-06,
+      "loss": 0.9144,
+      "step": 215
+    },
+    {
+      "epoch": 0.5307125307125307,
+      "grad_norm": 1.415085937475802,
+      "learning_rate": 1.7581608036294075e-06,
+      "loss": 0.9726,
+      "step": 216
+    },
+    {
+      "epoch": 0.5331695331695332,
+      "grad_norm": 1.558702474871841,
+      "learning_rate": 1.7555078849410042e-06,
+      "loss": 0.9243,
+      "step": 217
+    },
+    {
+      "epoch": 0.5356265356265356,
+      "grad_norm": 1.826623068304597,
+      "learning_rate": 1.7528425194950793e-06,
+      "loss": 1.0106,
+      "step": 218
+    },
+    {
+      "epoch": 0.538083538083538,
+      "grad_norm": 1.5238015749509666,
+      "learning_rate": 1.7501647512026994e-06,
+      "loss": 1.0093,
+      "step": 219
+    },
+    {
+      "epoch": 0.5405405405405406,
+      "grad_norm": 1.6514187408836531,
+      "learning_rate": 1.7474746241792646e-06,
+      "loss": 0.991,
+      "step": 220
+    },
+    {
+      "epoch": 0.542997542997543,
+      "grad_norm": 1.5401870547381438,
+      "learning_rate": 1.7447721827437819e-06,
+      "loss": 0.8904,
+      "step": 221
+    },
+    {
+      "epoch": 0.5454545454545454,
+      "grad_norm": 1.58085107300811,
+      "learning_rate": 1.742057471418133e-06,
+      "loss": 0.907,
+      "step": 222
+    },
+    {
+      "epoch": 0.547911547911548,
+      "grad_norm": 1.4376088351554372,
+      "learning_rate": 1.7393305349263432e-06,
+      "loss": 0.9369,
+      "step": 223
+    },
+    {
+      "epoch": 0.5503685503685504,
+      "grad_norm": 1.3807320173451696,
+      "learning_rate": 1.7365914181938437e-06,
+      "loss": 0.8945,
+      "step": 224
+    },
+    {
+      "epoch": 0.5528255528255528,
+      "grad_norm": 1.5104969018841128,
+      "learning_rate": 1.7338401663467309e-06,
+      "loss": 0.9537,
+      "step": 225
+    },
+    {
+      "epoch": 0.5552825552825553,
+      "grad_norm": 1.4036248703567802,
+      "learning_rate": 1.7310768247110232e-06,
+      "loss": 0.9112,
+      "step": 226
+    },
+    {
+      "epoch": 0.5577395577395577,
+      "grad_norm": 1.4153113695817048,
+      "learning_rate": 1.7283014388119158e-06,
+      "loss": 0.9285,
+      "step": 227
+    },
+    {
+      "epoch": 0.5601965601965602,
+      "grad_norm": 1.6827361424168579,
+      "learning_rate": 1.7255140543730282e-06,
+      "loss": 0.9368,
+      "step": 228
+    },
+    {
+      "epoch": 0.5626535626535627,
+      "grad_norm": 1.52442672792253,
+      "learning_rate": 1.7227147173156523e-06,
+      "loss": 0.9838,
+      "step": 229
+    },
+    {
+      "epoch": 0.5651105651105651,
+      "grad_norm": 1.8631309960035536,
+      "learning_rate": 1.7199034737579958e-06,
+      "loss": 0.9311,
+      "step": 230
+    },
+    {
+      "epoch": 0.5675675675675675,
+      "grad_norm": 1.38816407656762,
+      "learning_rate": 1.7170803700144224e-06,
+      "loss": 0.945,
+      "step": 231
+    },
+    {
+      "epoch": 0.5700245700245701,
+      "grad_norm": 1.5370320107384547,
+      "learning_rate": 1.7142454525946889e-06,
+      "loss": 0.955,
+      "step": 232
+    },
+    {
+      "epoch": 0.5724815724815725,
+      "grad_norm": 1.5632251869407223,
+      "learning_rate": 1.7113987682031778e-06,
+      "loss": 0.9384,
+      "step": 233
+    },
+    {
+      "epoch": 0.5749385749385749,
+      "grad_norm": 1.602474983632772,
+      "learning_rate": 1.70854036373813e-06,
+      "loss": 0.992,
+      "step": 234
+    },
+    {
+      "epoch": 0.5773955773955773,
+      "grad_norm": 1.3375523516298262,
+      "learning_rate": 1.70567028629087e-06,
+      "loss": 0.941,
+      "step": 235
+    },
+    {
+      "epoch": 0.5798525798525799,
+      "grad_norm": 1.539702405955399,
+      "learning_rate": 1.7027885831450317e-06,
+      "loss": 0.9516,
+      "step": 236
+    },
+    {
+      "epoch": 0.5823095823095823,
+      "grad_norm": 1.6116292160800432,
+      "learning_rate": 1.6998953017757784e-06,
+      "loss": 0.992,
+      "step": 237
+    },
+    {
+      "epoch": 0.5847665847665847,
+      "grad_norm": 2.3507568939715835,
+      "learning_rate": 1.6969904898490213e-06,
+      "loss": 0.9165,
+      "step": 238
+    },
+    {
+      "epoch": 0.5872235872235873,
+      "grad_norm": 2.847402959001731,
+      "learning_rate": 1.694074195220634e-06,
+      "loss": 0.9985,
+      "step": 239
+    },
+    {
+      "epoch": 0.5896805896805897,
+      "grad_norm": 1.5211685594705786,
+      "learning_rate": 1.6911464659356632e-06,
+      "loss": 0.9365,
+      "step": 240
+    },
+    {
+      "epoch": 0.5921375921375921,
+      "grad_norm": 2.034636443003393,
+      "learning_rate": 1.688207350227539e-06,
+      "loss": 0.9094,
+      "step": 241
+    },
+    {
+      "epoch": 0.5945945945945946,
+      "grad_norm": 1.8768465174864484,
+      "learning_rate": 1.6852568965172791e-06,
+      "loss": 0.9898,
+      "step": 242
+    },
+    {
+      "epoch": 0.597051597051597,
+      "grad_norm": 1.7384412195063934,
+      "learning_rate": 1.6822951534126908e-06,
+      "loss": 0.9955,
+      "step": 243
+    },
+    {
+      "epoch": 0.5995085995085995,
+      "grad_norm": 1.4908111245385776,
+      "learning_rate": 1.6793221697075715e-06,
+      "loss": 0.9503,
+      "step": 244
+    },
+    {
+      "epoch": 0.601965601965602,
+      "grad_norm": 1.3104233864174921,
+      "learning_rate": 1.6763379943809026e-06,
+      "loss": 0.9087,
+      "step": 245
+    },
+    {
+      "epoch": 0.6044226044226044,
+      "grad_norm": 1.8378209471753764,
+      "learning_rate": 1.6733426765960456e-06,
+      "loss": 0.9295,
+      "step": 246
+    },
+    {
+      "epoch": 0.6068796068796068,
+      "grad_norm": 1.5513249549838153,
+      "learning_rate": 1.6703362656999299e-06,
+      "loss": 0.9622,
+      "step": 247
+    },
+    {
+      "epoch": 0.6093366093366094,
+      "grad_norm": 1.950619315682982,
+      "learning_rate": 1.6673188112222395e-06,
+      "loss": 0.9114,
+      "step": 248
+    },
+    {
+      "epoch": 0.6117936117936118,
+      "grad_norm": 1.5443899853661804,
+      "learning_rate": 1.6642903628745993e-06,
+      "loss": 0.9385,
+      "step": 249
+    },
+    {
+      "epoch": 0.6142506142506142,
+      "grad_norm": 1.842691343791668,
+      "learning_rate": 1.6612509705497541e-06,
+      "loss": 0.9681,
+      "step": 250
+    },
+    {
+      "epoch": 0.6167076167076168,
+      "grad_norm": 1.4579884817325401,
+      "learning_rate": 1.6582006843207479e-06,
+      "loss": 0.9967,
+      "step": 251
+    },
+    {
+      "epoch": 0.6191646191646192,
+      "grad_norm": 1.602595851064425,
+      "learning_rate": 1.655139554440098e-06,
+      "loss": 0.9674,
+      "step": 252
+    },
+    {
+      "epoch": 0.6216216216216216,
+      "grad_norm": 1.394530716544746,
+      "learning_rate": 1.652067631338967e-06,
+      "loss": 0.907,
+      "step": 253
+    },
+    {
+      "epoch": 0.6240786240786241,
+      "grad_norm": 1.6411505351669575,
+      "learning_rate": 1.6489849656263334e-06,
+      "loss": 0.9602,
+      "step": 254
+    },
+    {
+      "epoch": 0.6265356265356266,
+      "grad_norm": 1.7684281051463475,
+      "learning_rate": 1.6458916080881565e-06,
+      "loss": 1.0141,
+      "step": 255
+    },
+    {
+      "epoch": 0.628992628992629,
+      "grad_norm": 1.771327098736445,
+      "learning_rate": 1.6427876096865393e-06,
+      "loss": 0.9917,
+      "step": 256
+    },
+    {
+      "epoch": 0.6314496314496314,
+      "grad_norm": 1.440535084208374,
+      "learning_rate": 1.6396730215588912e-06,
+      "loss": 0.9326,
+      "step": 257
+    },
+    {
+      "epoch": 0.6339066339066339,
+      "grad_norm": 1.5197592107480333,
+      "learning_rate": 1.6365478950170833e-06,
+      "loss": 0.9475,
+      "step": 258
+    },
+    {
+      "epoch": 0.6363636363636364,
+      "grad_norm": 1.706431633980188,
+      "learning_rate": 1.6334122815466031e-06,
+      "loss": 0.9351,
+      "step": 259
+    },
+    {
+      "epoch": 0.6388206388206388,
+      "grad_norm": 1.323673519490258,
+      "learning_rate": 1.6302662328057085e-06,
+      "loss": 0.911,
+      "step": 260
+    },
+    {
+      "epoch": 0.6412776412776413,
+      "grad_norm": 1.5394197428392233,
+      "learning_rate": 1.627109800624574e-06,
+      "loss": 0.9812,
+      "step": 261
+    },
+    {
+      "epoch": 0.6437346437346437,
+      "grad_norm": 1.4717326057521631,
+      "learning_rate": 1.6239430370044386e-06,
+      "loss": 0.8713,
+      "step": 262
+    },
+    {
+      "epoch": 0.6461916461916462,
+      "grad_norm": 1.5141060178146766,
+      "learning_rate": 1.6207659941167484e-06,
+      "loss": 0.934,
+      "step": 263
+    },
+    {
+      "epoch": 0.6486486486486487,
+      "grad_norm": 1.606823271595554,
+      "learning_rate": 1.617578724302297e-06,
+      "loss": 0.9689,
+      "step": 264
+    },
+    {
+      "epoch": 0.6511056511056511,
+      "grad_norm": 1.436173862475149,
+      "learning_rate": 1.6143812800703643e-06,
+      "loss": 0.9524,
+      "step": 265
+    },
+    {
+      "epoch": 0.6535626535626535,
+      "grad_norm": 1.6794632700616086,
+      "learning_rate": 1.6111737140978491e-06,
+      "loss": 0.9377,
+      "step": 266
+    },
+    {
+      "epoch": 0.6560196560196561,
+      "grad_norm": 1.481170005860829,
+      "learning_rate": 1.6079560792284045e-06,
+      "loss": 0.9718,
+      "step": 267
+    },
+    {
+      "epoch": 0.6584766584766585,
+      "grad_norm": 1.5488863504284036,
+      "learning_rate": 1.604728428471564e-06,
+      "loss": 0.9597,
+      "step": 268
+    },
+    {
+      "epoch": 0.6609336609336609,
+      "grad_norm": 1.5751377061062102,
+      "learning_rate": 1.6014908150018703e-06,
+      "loss": 0.9741,
+      "step": 269
+    },
+    {
+      "epoch": 0.6633906633906634,
+      "grad_norm": 1.8133205336540266,
+      "learning_rate": 1.598243292157999e-06,
+      "loss": 0.9234,
+      "step": 270
+    },
+    {
+      "epoch": 0.6658476658476659,
+      "grad_norm": 2.014347908541205,
+      "learning_rate": 1.5949859134418797e-06,
+      "loss": 0.976,
+      "step": 271
+    },
+    {
+      "epoch": 0.6683046683046683,
+      "grad_norm": 1.3995748329733648,
+      "learning_rate": 1.5917187325178136e-06,
+      "loss": 0.9182,
+      "step": 272
+    },
+    {
+      "epoch": 0.6707616707616708,
+      "grad_norm": 1.536609373517338,
+      "learning_rate": 1.5884418032115906e-06,
+      "loss": 0.8921,
+      "step": 273
+    },
+    {
+      "epoch": 0.6732186732186732,
+      "grad_norm": 1.7529235851864882,
+      "learning_rate": 1.5851551795096026e-06,
+      "loss": 0.9083,
+      "step": 274
+    },
+    {
+      "epoch": 0.6756756756756757,
+      "grad_norm": 1.5263377609284405,
+      "learning_rate": 1.5818589155579529e-06,
+      "loss": 0.921,
+      "step": 275
+    },
+    {
+      "epoch": 0.6781326781326781,
+      "grad_norm": 1.841830762323041,
+      "learning_rate": 1.5785530656615652e-06,
+      "loss": 0.943,
+      "step": 276
+    },
+    {
+      "epoch": 0.6805896805896806,
+      "grad_norm": 1.5135421099319812,
+      "learning_rate": 1.5752376842832898e-06,
+      "loss": 0.9315,
+      "step": 277
+    },
+    {
+      "epoch": 0.683046683046683,
+      "grad_norm": 1.492488292559091,
+      "learning_rate": 1.571912826043003e-06,
+      "loss": 0.891,
+      "step": 278
+    },
+    {
+      "epoch": 0.6855036855036855,
+      "grad_norm": 1.9836354455315945,
+      "learning_rate": 1.5685785457167113e-06,
+      "loss": 0.8706,
+      "step": 279
+    },
+    {
+      "epoch": 0.687960687960688,
+      "grad_norm": 1.5134741374320924,
+      "learning_rate": 1.565234898235646e-06,
+      "loss": 0.9786,
+      "step": 280
+    },
+    {
+      "epoch": 0.6904176904176904,
+      "grad_norm": 1.4859838259301754,
+      "learning_rate": 1.5618819386853603e-06,
+      "loss": 0.9595,
+      "step": 281
+    },
+    {
+      "epoch": 0.6928746928746928,
+      "grad_norm": 1.567447965101693,
+      "learning_rate": 1.5585197223048202e-06,
+      "loss": 0.9373,
+      "step": 282
+    },
+    {
+      "epoch": 0.6953316953316954,
+      "grad_norm": 1.361598883424366,
+      "learning_rate": 1.5551483044854952e-06,
+      "loss": 0.9697,
+      "step": 283
+    },
+    {
+      "epoch": 0.6977886977886978,
+      "grad_norm": 1.6640657059750952,
+      "learning_rate": 1.551767740770446e-06,
+      "loss": 0.9244,
+      "step": 284
+    },
+    {
+      "epoch": 0.7002457002457002,
+      "grad_norm": 1.362536522369026,
+      "learning_rate": 1.5483780868534083e-06,
+      "loss": 0.981,
+      "step": 285
+    },
+    {
+      "epoch": 0.7027027027027027,
+      "grad_norm": 1.6010684699844948,
+      "learning_rate": 1.544979398577877e-06,
+      "loss": 1.0169,
+      "step": 286
+    },
+    {
+      "epoch": 0.7051597051597052,
+      "grad_norm": 1.6122664157541595,
+      "learning_rate": 1.5415717319361845e-06,
+      "loss": 0.9507,
+      "step": 287
+    },
+    {
+      "epoch": 0.7076167076167076,
+      "grad_norm": 1.5044459630841276,
+      "learning_rate": 1.5381551430685796e-06,
+      "loss": 0.9431,
+      "step": 288
+    },
+    {
+      "epoch": 0.7100737100737101,
+      "grad_norm": 1.464490997097453,
+      "learning_rate": 1.5347296882623016e-06,
+      "loss": 0.9726,
+      "step": 289
+    },
+    {
+      "epoch": 0.7125307125307125,
+      "grad_norm": 1.8674476124140849,
+      "learning_rate": 1.5312954239506533e-06,
+      "loss": 0.9406,
+      "step": 290
+    },
+    {
+      "epoch": 0.714987714987715,
+      "grad_norm": 1.3556749625191538,
+      "learning_rate": 1.5278524067120718e-06,
+      "loss": 0.8896,
+      "step": 291
+    },
+    {
+      "epoch": 0.7174447174447175,
+      "grad_norm": 1.4118443972410692,
+      "learning_rate": 1.5244006932691954e-06,
+      "loss": 0.9722,
+      "step": 292
+    },
+    {
+      "epoch": 0.7199017199017199,
+      "grad_norm": 1.6225004312255737,
+      "learning_rate": 1.5209403404879303e-06,
+      "loss": 0.929,
+      "step": 293
+    },
+    {
+      "epoch": 0.7223587223587223,
+      "grad_norm": 1.6022157008964288,
+      "learning_rate": 1.5174714053765123e-06,
+      "loss": 0.9497,
+      "step": 294
+    },
+    {
+      "epoch": 0.7248157248157249,
+      "grad_norm": 1.5891302329335815,
+      "learning_rate": 1.5139939450845699e-06,
+      "loss": 0.973,
+      "step": 295
+    },
+    {
+      "epoch": 0.7272727272727273,
+      "grad_norm": 1.6988844584422054,
+      "learning_rate": 1.5105080169021789e-06,
+      "loss": 0.9184,
+      "step": 296
+    },
+    {
+      "epoch": 0.7297297297297297,
+      "grad_norm": 1.453985001696748,
+      "learning_rate": 1.5070136782589234e-06,
+      "loss": 0.9477,
+      "step": 297
+    },
+    {
+      "epoch": 0.7321867321867321,
+      "grad_norm": 1.604839180071105,
+      "learning_rate": 1.5035109867229457e-06,
+      "loss": 0.9409,
+      "step": 298
+    },
+    {
+      "epoch": 0.7346437346437347,
+      "grad_norm": 1.5773771725228862,
+      "learning_rate": 1.5e-06,
+      "loss": 0.937,
+      "step": 299
+    },
+    {
+      "epoch": 0.7371007371007371,
+      "grad_norm": 1.4272988880705397,
+      "learning_rate": 1.496480775932501e-06,
+      "loss": 0.9481,
+      "step": 300
+    },
+    {
+      "epoch": 0.7395577395577395,
+      "grad_norm": 1.534183187925811,
+      "learning_rate": 1.492953372498571e-06,
+      "loss": 0.9111,
+      "step": 301
+    },
+    {
+      "epoch": 0.742014742014742,
+      "grad_norm": 1.6285985984253928,
+      "learning_rate": 1.4894178478110854e-06,
+      "loss": 0.949,
+      "step": 302
+    },
+    {
+      "epoch": 0.7444717444717445,
+      "grad_norm": 1.521782496026601,
+      "learning_rate": 1.4858742601167138e-06,
+      "loss": 0.9405,
+      "step": 303
+    },
+    {
+      "epoch": 0.7469287469287469,
+      "grad_norm": 1.522930763087887,
+      "learning_rate": 1.4823226677949622e-06,
+      "loss": 0.949,
+      "step": 304
+    },
+    {
+      "epoch": 0.7493857493857494,
+      "grad_norm": 5.354032445168916,
+      "learning_rate": 1.4787631293572092e-06,
+      "loss": 0.9708,
+      "step": 305
+    },
+    {
+      "epoch": 0.7518427518427518,
+      "grad_norm": 1.199402404810751,
+      "learning_rate": 1.4751957034457446e-06,
+      "loss": 0.8895,
+      "step": 306
+    },
+    {
+      "epoch": 0.7542997542997543,
+      "grad_norm": 1.4889622591506924,
+      "learning_rate": 1.4716204488328005e-06,
+      "loss": 0.9258,
+      "step": 307
+    },
+    {
+      "epoch": 0.7567567567567568,
+      "grad_norm": 1.586929861427342,
+      "learning_rate": 1.4680374244195858e-06,
+      "loss": 0.9523,
+      "step": 308
+    },
+    {
+      "epoch": 0.7592137592137592,
+      "grad_norm": 1.7542006852572127,
+      "learning_rate": 1.464446689235314e-06,
+      "loss": 0.9671,
+      "step": 309
+    },
+    {
+      "epoch": 0.7616707616707616,
+      "grad_norm": 1.906886186477315,
+      "learning_rate": 1.4608483024362302e-06,
+      "loss": 0.9479,
+      "step": 310
+    },
+    {
+      "epoch": 0.7641277641277642,
+      "grad_norm": 1.3426576295326547,
+      "learning_rate": 1.4572423233046385e-06,
+      "loss": 0.9245,
+      "step": 311
+    },
+    {
+      "epoch": 0.7665847665847666,
+      "grad_norm": 1.544123812480689,
+      "learning_rate": 1.453628811247924e-06,
+      "loss": 0.9833,
+      "step": 312
+    },
+    {
+      "epoch": 0.769041769041769,
+      "grad_norm": 1.3070207560130556,
+      "learning_rate": 1.4500078257975745e-06,
+      "loss": 0.9518,
+      "step": 313
+    },
+    {
+      "epoch": 0.7714987714987716,
+      "grad_norm": 1.953852840958092,
+      "learning_rate": 1.4463794266081992e-06,
+      "loss": 0.8871,
+      "step": 314
+    },
+    {
+      "epoch": 0.773955773955774,
+      "grad_norm": 1.5767416529599687,
+      "learning_rate": 1.4427436734565473e-06,
+      "loss": 0.9921,
+      "step": 315
+    },
+    {
+      "epoch": 0.7764127764127764,
+      "grad_norm": 1.5600472966901433,
+      "learning_rate": 1.4391006262405212e-06,
+      "loss": 0.9068,
+      "step": 316
+    },
+    {
+      "epoch": 0.7788697788697788,
+      "grad_norm": 1.8524989491255999,
+      "learning_rate": 1.4354503449781913e-06,
+      "loss": 0.8956,
+      "step": 317
+    },
+    {
+      "epoch": 0.7813267813267813,
+      "grad_norm": 1.4701804028885344,
+      "learning_rate": 1.4317928898068066e-06,
+      "loss": 0.9547,
+      "step": 318
+    },
+    {
+      "epoch": 0.7837837837837838,
+      "grad_norm": 1.505304585103799,
+      "learning_rate": 1.4281283209818039e-06,
+      "loss": 0.9214,
+      "step": 319
+    },
+    {
+      "epoch": 0.7862407862407862,
+      "grad_norm": 1.5458422714438642,
+      "learning_rate": 1.424456698875815e-06,
+      "loss": 0.9711,
+      "step": 320
+    },
+    {
+      "epoch": 0.7886977886977887,
+      "grad_norm": 1.8126365786822831,
+      "learning_rate": 1.4207780839776732e-06,
+      "loss": 0.9638,
+      "step": 321
+    },
+    {
+      "epoch": 0.7911547911547911,
+      "grad_norm": 1.3634824425424106,
+      "learning_rate": 1.4170925368914151e-06,
+      "loss": 0.925,
+      "step": 322
+    },
+    {
+      "epoch": 0.7936117936117936,
+      "grad_norm": 1.5523412959213532,
+      "learning_rate": 1.413400118335283e-06,
+      "loss": 0.9681,
+      "step": 323
+    },
+    {
+      "epoch": 0.7960687960687961,
+      "grad_norm": 1.4281815228481956,
+      "learning_rate": 1.4097008891407245e-06,
+      "loss": 0.9469,
+      "step": 324
+    },
+    {
+      "epoch": 0.7985257985257985,
+      "grad_norm": 1.755464439866289,
+      "learning_rate": 1.4059949102513913e-06,
+      "loss": 0.9614,
+      "step": 325
+    },
+    {
+      "epoch": 0.800982800982801,
+      "grad_norm": 1.3755915444522364,
+      "learning_rate": 1.4022822427221323e-06,
+      "loss": 0.9344,
+      "step": 326
+    },
+    {
+      "epoch": 0.8034398034398035,
+      "grad_norm": 1.6217750884199984,
+      "learning_rate": 1.3985629477179915e-06,
+      "loss": 0.9499,
+      "step": 327
+    },
+    {
+      "epoch": 0.8058968058968059,
+      "grad_norm": 1.5965334157340827,
+      "learning_rate": 1.3948370865131978e-06,
+      "loss": 0.9433,
+      "step": 328
+    },
+    {
+      "epoch": 0.8083538083538083,
+      "grad_norm": 1.8371436403022903,
+      "learning_rate": 1.3911047204901558e-06,
+      "loss": 0.9899,
+      "step": 329
+    },
+    {
+      "epoch": 0.8108108108108109,
+      "grad_norm": 1.8839520322523513,
+      "learning_rate": 1.387365911138436e-06,
+      "loss": 0.9664,
+      "step": 330
+    },
+    {
+      "epoch": 0.8132678132678133,
+      "grad_norm": 1.495735998773526,
+      "learning_rate": 1.3836207200537594e-06,
+      "loss": 0.9784,
+      "step": 331
+    },
+    {
+      "epoch": 0.8157248157248157,
+      "grad_norm": 1.4893770976659941,
+      "learning_rate": 1.3798692089369855e-06,
+      "loss": 0.9046,
+      "step": 332
+    },
+    {
+      "epoch": 0.8181818181818182,
+      "grad_norm": 1.5373781226372853,
+      "learning_rate": 1.3761114395930928e-06,
+      "loss": 0.9461,
+      "step": 333
+    },
+    {
+      "epoch": 0.8206388206388207,
+      "grad_norm": 1.497679579290753,
+      "learning_rate": 1.3723474739301634e-06,
+      "loss": 0.9109,
+      "step": 334
+    },
+    {
+      "epoch": 0.8230958230958231,
+      "grad_norm": 1.6333646657141205,
+      "learning_rate": 1.3685773739583618e-06,
+      "loss": 0.893,
+      "step": 335
+    },
+    {
+      "epoch": 0.8255528255528255,
+      "grad_norm": 1.4444606139573457,
+      "learning_rate": 1.3648012017889121e-06,
+      "loss": 0.9616,
+      "step": 336
+    },
+    {
+      "epoch": 0.828009828009828,
+      "grad_norm": 1.633866194720376,
+      "learning_rate": 1.3610190196330775e-06,
+      "loss": 0.9316,
+      "step": 337
+    },
+    {
+      "epoch": 0.8304668304668305,
+      "grad_norm": 1.4965633625059045,
+      "learning_rate": 1.3572308898011327e-06,
+      "loss": 0.8995,
+      "step": 338
+    },
+    {
+      "epoch": 0.8329238329238329,
+      "grad_norm": 1.4593159509482825,
+      "learning_rate": 1.3534368747013394e-06,
+      "loss": 0.9416,
+      "step": 339
+    },
+    {
+      "epoch": 0.8353808353808354,
+      "grad_norm": 1.5060565758495885,
+      "learning_rate": 1.3496370368389165e-06,
+      "loss": 0.9385,
+      "step": 340
+    },
+    {
+      "epoch": 0.8378378378378378,
+      "grad_norm": 1.5228727270675948,
+      "learning_rate": 1.3458314388150113e-06,
+      "loss": 0.9249,
+      "step": 341
+    },
+    {
+      "epoch": 0.8402948402948403,
+      "grad_norm": 1.6137336621038334,
+      "learning_rate": 1.3420201433256689e-06,
+      "loss": 0.9549,
+      "step": 342
+    },
+    {
+      "epoch": 0.8427518427518428,
+      "grad_norm": 1.2744326108891244,
+      "learning_rate": 1.3382032131607966e-06,
+      "loss": 0.9275,
+      "step": 343
+    },
+    {
+      "epoch": 0.8452088452088452,
+      "grad_norm": 1.5480593545772812,
+      "learning_rate": 1.3343807112031326e-06,
+      "loss": 0.9342,
+      "step": 344
+    },
+    {
+      "epoch": 0.8476658476658476,
+      "grad_norm": 1.6690119423341088,
+      "learning_rate": 1.3305527004272087e-06,
+      "loss": 0.9375,
+      "step": 345
+    },
+    {
+      "epoch": 0.8501228501228502,
+      "grad_norm": 1.4907525884430786,
+      "learning_rate": 1.3267192438983117e-06,
+      "loss": 0.9183,
+      "step": 346
+    },
+    {
+      "epoch": 0.8525798525798526,
+      "grad_norm": 1.487251447150466,
+      "learning_rate": 1.3228804047714462e-06,
+      "loss": 0.9572,
+      "step": 347
+    },
+    {
+      "epoch": 0.855036855036855,
+      "grad_norm": 2.022254168777352,
+      "learning_rate": 1.3190362462902936e-06,
+      "loss": 0.9431,
+      "step": 348
+    },
+    {
+      "epoch": 0.8574938574938575,
+      "grad_norm": 1.570248437527878,
+      "learning_rate": 1.31518683178617e-06,
+      "loss": 0.8945,
+      "step": 349
+    },
+    {
+      "epoch": 0.85995085995086,
+      "grad_norm": 1.5017196199325449,
+      "learning_rate": 1.3113322246769816e-06,
+      "loss": 0.8841,
+      "step": 350
+    },
+    {
+      "epoch": 0.8624078624078624,
+      "grad_norm": 1.5003226831400922,
+      "learning_rate": 1.3074724884661831e-06,
+      "loss": 0.9785,
+      "step": 351
+    },
+    {
+      "epoch": 0.8648648648648649,
+      "grad_norm": 1.6237720015743136,
+      "learning_rate": 1.3036076867417286e-06,
+      "loss": 0.9942,
+      "step": 352
+    },
+    {
+      "epoch": 0.8673218673218673,
+      "grad_norm": 1.586236748920305,
+      "learning_rate": 1.299737883175024e-06,
+      "loss": 0.9036,
+      "step": 353
+    },
+    {
+      "epoch": 0.8697788697788698,
+      "grad_norm": 1.6356021082940453,
+      "learning_rate": 1.2958631415198813e-06,
+      "loss": 0.9723,
+      "step": 354
+    },
+    {
+      "epoch": 0.8722358722358723,
+      "grad_norm": 1.5448677133530493,
+      "learning_rate": 1.2919835256114637e-06,
+      "loss": 0.9391,
+      "step": 355
+    },
+    {
+      "epoch": 0.8746928746928747,
+      "grad_norm": 1.6212696290821345,
+      "learning_rate": 1.2880990993652377e-06,
+      "loss": 0.9202,
+      "step": 356
+    },
+    {
+      "epoch": 0.8771498771498771,
+      "grad_norm": 1.5176857112402762,
+      "learning_rate": 1.2842099267759175e-06,
+      "loss": 0.9037,
+      "step": 357
+    },
+    {
+      "epoch": 0.8796068796068796,
+      "grad_norm": 1.3944467856106335,
+      "learning_rate": 1.2803160719164126e-06,
+      "loss": 0.8854,
+      "step": 358
+    },
+    {
+      "epoch": 0.8820638820638821,
+      "grad_norm": 1.4377067419239682,
+      "learning_rate": 1.2764175989367715e-06,
+      "loss": 0.905,
+      "step": 359
+    },
+    {
+      "epoch": 0.8845208845208845,
+      "grad_norm": 1.6123612736739008,
+      "learning_rate": 1.272514572063124e-06,
+      "loss": 0.9881,
+      "step": 360
+    },
+    {
+      "epoch": 0.8869778869778869,
+      "grad_norm": 1.589984246633212,
+      "learning_rate": 1.2686070555966253e-06,
+      "loss": 0.9529,
+      "step": 361
+    },
+    {
+      "epoch": 0.8894348894348895,
+      "grad_norm": 1.4049932596277341,
+      "learning_rate": 1.2646951139123933e-06,
+      "loss": 0.8715,
+      "step": 362
+    },
+    {
+      "epoch": 0.8918918918918919,
+      "grad_norm": 1.7376495814908663,
+      "learning_rate": 1.260778811458452e-06,
+      "loss": 0.9551,
+      "step": 363
+    },
+    {
+      "epoch": 0.8943488943488943,
+      "grad_norm": 1.6356081133105504,
+      "learning_rate": 1.2568582127546661e-06,
+      "loss": 0.9357,
+      "step": 364
+    },
+    {
+      "epoch": 0.8968058968058968,
+      "grad_norm": 1.5620481027616258,
+      "learning_rate": 1.2529333823916806e-06,
+      "loss": 0.9724,
+      "step": 365
+    },
+    {
+      "epoch": 0.8992628992628993,
+      "grad_norm": 1.678396238284589,
+      "learning_rate": 1.2490043850298557e-06,
+      "loss": 0.9156,
+      "step": 366
+    },
+    {
+      "epoch": 0.9017199017199017,
+      "grad_norm": 2.354758429687385,
+      "learning_rate": 1.2450712853982013e-06,
+      "loss": 0.8628,
+      "step": 367
+    },
+    {
+      "epoch": 0.9041769041769042,
+      "grad_norm": 1.600899191967723,
+      "learning_rate": 1.2411341482933108e-06,
+      "loss": 0.9431,
+      "step": 368
+    },
+    {
+      "epoch": 0.9066339066339066,
+      "grad_norm": 1.4058440155556784,
+      "learning_rate": 1.2371930385782943e-06,
+      "loss": 0.9446,
+      "step": 369
+    },
+    {
+      "epoch": 0.9090909090909091,
+      "grad_norm": 1.4613065350370973,
+      "learning_rate": 1.233248021181709e-06,
+      "loss": 0.9164,
+      "step": 370
+    },
+    {
+      "epoch": 0.9115479115479116,
+      "grad_norm": 1.6113469899034865,
+      "learning_rate": 1.22929916109649e-06,
+      "loss": 0.9209,
+      "step": 371
+    },
+    {
+      "epoch": 0.914004914004914,
+      "grad_norm": 1.4540218388463704,
+      "learning_rate": 1.2253465233788794e-06,
+      "loss": 0.9865,
+      "step": 372
+    },
+    {
+      "epoch": 0.9164619164619164,
+      "grad_norm": 1.4002046410996662,
+      "learning_rate": 1.2213901731473553e-06,
+      "loss": 0.9229,
+      "step": 373
+    },
+    {
+      "epoch": 0.918918918918919,
+      "grad_norm": 1.4545211228962136,
+      "learning_rate": 1.217430175581557e-06,
+      "loss": 0.9401,
+      "step": 374
+    },
+    {
+      "epoch": 0.9213759213759214,
+      "grad_norm": 1.4460543335595146,
+      "learning_rate": 1.2134665959212135e-06,
+      "loss": 0.9552,
+      "step": 375
+    },
+    {
+      "epoch": 0.9238329238329238,
+      "grad_norm": 1.5791517432252247,
+      "learning_rate": 1.209499499465068e-06,
+      "loss": 0.9547,
+      "step": 376
+    },
+    {
+      "epoch": 0.9262899262899262,
+      "grad_norm": 1.4523281683604505,
+      "learning_rate": 1.2055289515698006e-06,
+      "loss": 0.9669,
+      "step": 377
+    },
+    {
+      "epoch": 0.9287469287469288,
+      "grad_norm": 1.8584285735562196,
+      "learning_rate": 1.2015550176489537e-06,
+      "loss": 0.9434,
+      "step": 378
+    },
+    {
+      "epoch": 0.9312039312039312,
+      "grad_norm": 1.5086682042761248,
+      "learning_rate": 1.1975777631718531e-06,
+      "loss": 0.9377,
+      "step": 379
+    },
+    {
+      "epoch": 0.9336609336609336,
+      "grad_norm": 1.7455613963553367,
+      "learning_rate": 1.19359725366253e-06,
+      "loss": 0.919,
+      "step": 380
+    },
+    {
+      "epoch": 0.9361179361179361,
+      "grad_norm": 1.5307030954736855,
+      "learning_rate": 1.1896135546986409e-06,
+      "loss": 0.9282,
+      "step": 381
+    },
+    {
+      "epoch": 0.9385749385749386,
+      "grad_norm": 1.5159714020228543,
+      "learning_rate": 1.1856267319103876e-06,
+      "loss": 0.9765,
+      "step": 382
+    },
+    {
+      "epoch": 0.941031941031941,
+      "grad_norm": 1.5839779729377836,
+      "learning_rate": 1.1816368509794364e-06,
+      "loss": 0.9793,
+      "step": 383
+    },
+    {
+      "epoch": 0.9434889434889435,
+      "grad_norm": 1.3472301054596898,
+      "learning_rate": 1.177643977637835e-06,
+      "loss": 0.9189,
+      "step": 384
+    },
+    {
+      "epoch": 0.9459459459459459,
+      "grad_norm": 1.5340710636652473,
+      "learning_rate": 1.1736481776669305e-06,
+      "loss": 0.9663,
+      "step": 385
+    },
+    {
+      "epoch": 0.9484029484029484,
+      "grad_norm": 1.4047565697947708,
+      "learning_rate": 1.1696495168962846e-06,
+      "loss": 0.9207,
+      "step": 386
+    },
+    {
+      "epoch": 0.9508599508599509,
+      "grad_norm": 1.6035760956637677,
+      "learning_rate": 1.165648061202591e-06,
+      "loss": 0.9793,
+      "step": 387
+    },
+    {
+      "epoch": 0.9533169533169533,
+      "grad_norm": 1.4948457768549497,
+      "learning_rate": 1.1616438765085882e-06,
+      "loss": 0.9381,
+      "step": 388
+    },
+    {
+      "epoch": 0.9557739557739557,
+      "grad_norm": 1.5241351339551144,
+      "learning_rate": 1.1576370287819734e-06,
+      "loss": 0.931,
+      "step": 389
+    },
+    {
+      "epoch": 0.9582309582309583,
+      "grad_norm": 1.4995817174577337,
+      "learning_rate": 1.1536275840343183e-06,
+      "loss": 0.9594,
+      "step": 390
+    },
+    {
+      "epoch": 0.9606879606879607,
+      "grad_norm": 1.8781230050093483,
+      "learning_rate": 1.149615608319978e-06,
+      "loss": 0.9536,
+      "step": 391
+    },
+    {
+      "epoch": 0.9631449631449631,
+      "grad_norm": 1.5385341714632899,
+      "learning_rate": 1.145601167735005e-06,
+      "loss": 0.9281,
+      "step": 392
+    },
+    {
+      "epoch": 0.9656019656019657,
+      "grad_norm": 1.6183085918500024,
+      "learning_rate": 1.1415843284160597e-06,
+      "loss": 0.9366,
+      "step": 393
+    },
+    {
+      "epoch": 0.9680589680589681,
+      "grad_norm": 1.5534604794185942,
+      "learning_rate": 1.137565156539322e-06,
+      "loss": 0.9316,
+      "step": 394
+    },
+    {
+      "epoch": 0.9705159705159705,
+      "grad_norm": 1.423811752511575,
+      "learning_rate": 1.133543718319398e-06,
+      "loss": 0.8927,
+      "step": 395
+    },
+    {
+      "epoch": 0.972972972972973,
+      "grad_norm": 1.4578205842089746,
+      "learning_rate": 1.1295200800082325e-06,
+      "loss": 0.951,
+      "step": 396
+    },
+    {
+      "epoch": 0.9754299754299754,
+      "grad_norm": 1.6386231495575438,
+      "learning_rate": 1.125494307894016e-06,
+      "loss": 0.8952,
+      "step": 397
+    },
+    {
+      "epoch": 0.9778869778869779,
+      "grad_norm": 1.5075389802303287,
+      "learning_rate": 1.1214664683000924e-06,
+      "loss": 0.8966,
+      "step": 398
+    },
+    {
+      "epoch": 0.9803439803439803,
+      "grad_norm": 1.631579827773725,
+      "learning_rate": 1.1174366275838662e-06,
+      "loss": 0.9411,
+      "step": 399
+    },
+    {
+      "epoch": 0.9828009828009828,
+      "grad_norm": 1.8249088799770508,
+      "learning_rate": 1.1134048521357116e-06,
+      "loss": 0.9232,
+      "step": 400
+    },
+    {
+      "epoch": 0.9852579852579852,
+      "grad_norm": 1.4813873308103862,
+      "learning_rate": 1.1093712083778746e-06,
+      "loss": 0.9625,
+      "step": 401
+    },
+    {
+      "epoch": 0.9877149877149877,
+      "grad_norm": 1.5996089579510389,
+      "learning_rate": 1.105335762763382e-06,
+      "loss": 0.9592,
+      "step": 402
+    },
+    {
+      "epoch": 0.9901719901719902,
+      "grad_norm": 1.4507338601430246,
+      "learning_rate": 1.1012985817749462e-06,
+      "loss": 0.9558,
+      "step": 403
+    },
+    {
+      "epoch": 0.9926289926289926,
+      "grad_norm": 1.452275573847242,
+      "learning_rate": 1.097259731923869e-06,
+      "loss": 0.9582,
+      "step": 404
+    },
+    {
+      "epoch": 0.995085995085995,
+      "grad_norm": 1.4529835839819856,
+      "learning_rate": 1.0932192797489458e-06,
+      "loss": 0.8952,
+      "step": 405
+    },
+    {
+      "epoch": 0.9975429975429976,
+      "grad_norm": 2.622272111272181,
+      "learning_rate": 1.0891772918153694e-06,
+      "loss": 0.918,
+      "step": 406
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5285112861748582,
+      "learning_rate": 1.0851338347136355e-06,
+      "loss": 0.877,
+      "step": 407
+    },
+    {
+      "epoch": 1.0024570024570025,
+      "grad_norm": 1.4118845101420152,
+      "learning_rate": 1.0810889750584424e-06,
+      "loss": 0.8647,
+      "step": 408
+    },
+    {
+      "epoch": 1.0049140049140048,
+      "grad_norm": 1.6747623259684568,
+      "learning_rate": 1.0770427794875948e-06,
+      "loss": 0.9671,
+      "step": 409
+    },
+    {
+      "epoch": 1.0073710073710074,
+      "grad_norm": 1.307864675513283,
+      "learning_rate": 1.0729953146609075e-06,
+      "loss": 0.8296,
+      "step": 410
+    },
+    {
+      "epoch": 1.00982800982801,
+      "grad_norm": 1.4553500313995769,
+      "learning_rate": 1.0689466472591048e-06,
+      "loss": 0.937,
+      "step": 411
+    },
+    {
+      "epoch": 1.0122850122850122,
+      "grad_norm": 1.4315419855901317,
+      "learning_rate": 1.0648968439827239e-06,
+      "loss": 0.9042,
+      "step": 412
+    },
+    {
+      "epoch": 1.0147420147420148,
+      "grad_norm": 1.5739844573571937,
+      "learning_rate": 1.0608459715510139e-06,
+      "loss": 0.9011,
+      "step": 413
+    },
+    {
+      "epoch": 1.0171990171990173,
+      "grad_norm": 1.437687738713176,
+      "learning_rate": 1.0567940967008395e-06,
+      "loss": 0.9363,
+      "step": 414
+    },
+    {
+      "epoch": 1.0196560196560196,
+      "grad_norm": 1.534448861262034,
+      "learning_rate": 1.052741286185579e-06,
+      "loss": 0.9539,
+      "step": 415
+    },
+    {
+      "epoch": 1.0221130221130221,
+      "grad_norm": 1.644806928510893,
+      "learning_rate": 1.0486876067740252e-06,
+      "loss": 0.9802,
+      "step": 416
+    },
+    {
+      "epoch": 1.0245700245700247,
+      "grad_norm": 1.5874571507432869,
+      "learning_rate": 1.0446331252492864e-06,
+      "loss": 0.8751,
+      "step": 417
+    },
+    {
+      "epoch": 1.027027027027027,
+      "grad_norm": 1.5244813435420985,
+      "learning_rate": 1.0405779084076856e-06,
+      "loss": 0.8844,
+      "step": 418
+    },
+    {
+      "epoch": 1.0294840294840295,
+      "grad_norm": 1.5721376461339462,
+      "learning_rate": 1.036522023057659e-06,
+      "loss": 0.9593,
+      "step": 419
+    },
+    {
+      "epoch": 1.031941031941032,
+      "grad_norm": 3.006639344932307,
+      "learning_rate": 1.0324655360186567e-06,
+      "loss": 0.9055,
+      "step": 420
+    },
+    {
+      "epoch": 1.0343980343980343,
+      "grad_norm": 1.3782225003410025,
+      "learning_rate": 1.0284085141200423e-06,
+      "loss": 0.9183,
+      "step": 421
+    },
+    {
+      "epoch": 1.0368550368550369,
+      "grad_norm": 1.376873801999615,
+      "learning_rate": 1.0243510241999897e-06,
+      "loss": 0.8767,
+      "step": 422
+    },
+    {
+      "epoch": 1.0393120393120394,
+      "grad_norm": 1.394373200963149,
+      "learning_rate": 1.0202931331043839e-06,
+      "loss": 0.8821,
+      "step": 423
+    },
+    {
+      "epoch": 1.0417690417690417,
+      "grad_norm": 1.571320978169866,
+      "learning_rate": 1.0162349076857191e-06,
+      "loss": 0.8959,
+      "step": 424
+    },
+    {
+      "epoch": 1.0442260442260443,
+      "grad_norm": 1.5336886338120022,
+      "learning_rate": 1.0121764148019975e-06,
+      "loss": 0.9375,
+      "step": 425
+    },
+    {
+      "epoch": 1.0466830466830466,
+      "grad_norm": 1.4693963607748404,
+      "learning_rate": 1.0081177213156277e-06,
+      "loss": 0.8909,
+      "step": 426
+    },
+    {
+      "epoch": 1.049140049140049,
+      "grad_norm": 1.5830327984938106,
+      "learning_rate": 1.004058894092323e-06,
+      "loss": 0.9207,
+      "step": 427
+    },
+    {
+      "epoch": 1.0515970515970516,
+      "grad_norm": 1.5170501490428927,
+      "learning_rate": 1e-06,
+      "loss": 0.9036,
+      "step": 428
+    },
+    {
+      "epoch": 1.054054054054054,
+      "grad_norm": 1.5692066826348359,
+      "learning_rate": 9.95941105907677e-07,
+      "loss": 0.9193,
+      "step": 429
+    },
+    {
+      "epoch": 1.0565110565110565,
+      "grad_norm": 1.4357790543992526,
+      "learning_rate": 9.918822786843724e-07,
+      "loss": 0.9124,
+      "step": 430
+    },
+    {
+      "epoch": 1.058968058968059,
+      "grad_norm": 1.423440533928829,
+      "learning_rate": 9.878235851980024e-07,
+      "loss": 0.9146,
+      "step": 431
+    },
+    {
+      "epoch": 1.0614250614250613,
+      "grad_norm": 1.590468077725602,
+      "learning_rate": 9.837650923142808e-07,
+      "loss": 0.8705,
+      "step": 432
+    },
+    {
+      "epoch": 1.0638820638820639,
+      "grad_norm": 1.461231262115798,
+      "learning_rate": 9.797068668956163e-07,
+      "loss": 0.9438,
+      "step": 433
+    },
+    {
+      "epoch": 1.0663390663390664,
+      "grad_norm": 1.42661968764142,
+      "learning_rate": 9.756489758000105e-07,
+      "loss": 0.8747,
+      "step": 434
+    },
+    {
+      "epoch": 1.0687960687960687,
+      "grad_norm": 1.4657641576186988,
+      "learning_rate": 9.715914858799576e-07,
+      "loss": 0.8939,
+      "step": 435
+    },
+    {
+      "epoch": 1.0712530712530712,
+      "grad_norm": 1.5022324634860065,
+      "learning_rate": 9.675344639813434e-07,
+      "loss": 0.922,
+      "step": 436
+    },
+    {
+      "epoch": 1.0737100737100738,
+      "grad_norm": 1.424039088022573,
+      "learning_rate": 9.63477976942341e-07,
+      "loss": 0.953,
+      "step": 437
+    },
+    {
+      "epoch": 1.076167076167076,
+      "grad_norm": 1.5965077678963564,
+      "learning_rate": 9.594220915923147e-07,
+      "loss": 0.9443,
+      "step": 438
+    },
+    {
+      "epoch": 1.0786240786240786,
+      "grad_norm": 1.6371319904729476,
+      "learning_rate": 9.553668747507138e-07,
+      "loss": 0.9228,
+      "step": 439
+    },
+    {
+      "epoch": 1.0810810810810811,
+      "grad_norm": 1.8309314739878833,
+      "learning_rate": 9.51312393225975e-07,
+      "loss": 0.8753,
+      "step": 440
+    },
+    {
+      "epoch": 1.0835380835380835,
+      "grad_norm": 1.39319457517673,
+      "learning_rate": 9.472587138144215e-07,
+      "loss": 0.8939,
+      "step": 441
+    },
+    {
+      "epoch": 1.085995085995086,
+      "grad_norm": 1.5070197645260877,
+      "learning_rate": 9.432059032991607e-07,
+      "loss": 0.9123,
+      "step": 442
+    },
+    {
+      "epoch": 1.0884520884520885,
+      "grad_norm": 1.3753531136081873,
+      "learning_rate": 9.391540284489861e-07,
+      "loss": 0.9328,
+      "step": 443
+    },
+    {
+      "epoch": 1.0909090909090908,
+      "grad_norm": 1.4036108581529319,
+      "learning_rate": 9.351031560172764e-07,
+      "loss": 0.9138,
+      "step": 444
+    },
+    {
+      "epoch": 1.0933660933660934,
+      "grad_norm": 1.460079134056615,
+      "learning_rate": 9.310533527408951e-07,
+      "loss": 0.8736,
+      "step": 445
+    },
+    {
+      "epoch": 1.095823095823096,
+      "grad_norm": 2.1567900139463663,
+      "learning_rate": 9.270046853390924e-07,
+      "loss": 0.9649,
+      "step": 446
+    },
+    {
+      "epoch": 1.0982800982800982,
+      "grad_norm": 1.5086116320926348,
+      "learning_rate": 9.229572205124051e-07,
+      "loss": 0.8839,
+      "step": 447
+    },
+    {
+      "epoch": 1.1007371007371007,
+      "grad_norm": 1.4696338643312472,
+      "learning_rate": 9.189110249415576e-07,
+      "loss": 0.8971,
+      "step": 448
+    },
+    {
+      "epoch": 1.1031941031941033,
+      "grad_norm": 1.5799491675072612,
+      "learning_rate": 9.148661652863642e-07,
+      "loss": 0.8839,
+      "step": 449
+    },
+    {
+      "epoch": 1.1056511056511056,
+      "grad_norm": 1.633966629237029,
+      "learning_rate": 9.108227081846304e-07,
+      "loss": 0.9053,
+      "step": 450
+    },
+    {
+      "epoch": 1.1081081081081081,
+      "grad_norm": 1.5227318714591582,
+      "learning_rate": 9.067807202510542e-07,
+      "loss": 0.8734,
+      "step": 451
+    },
+    {
+      "epoch": 1.1105651105651106,
+      "grad_norm": 1.6390683121943366,
+      "learning_rate": 9.027402680761308e-07,
+      "loss": 0.8389,
+      "step": 452
+    },
+    {
+      "epoch": 1.113022113022113,
+      "grad_norm": 1.8480120093100039,
+      "learning_rate": 8.987014182250537e-07,
+      "loss": 0.8977,
+      "step": 453
+    },
+    {
+      "epoch": 1.1154791154791155,
+      "grad_norm": 1.6432113495263132,
+      "learning_rate": 8.94664237236618e-07,
+      "loss": 0.8924,
+      "step": 454
+    },
+    {
+      "epoch": 1.117936117936118,
+      "grad_norm": 1.7383339935167776,
+      "learning_rate": 8.906287916221258e-07,
+      "loss": 0.9181,
+      "step": 455
+    },
+    {
+      "epoch": 1.1203931203931203,
+      "grad_norm": 1.5888604077241295,
+      "learning_rate": 8.865951478642886e-07,
+      "loss": 0.9276,
+      "step": 456
+    },
+    {
+      "epoch": 1.1228501228501229,
+      "grad_norm": 1.4607976669981695,
+      "learning_rate": 8.825633724161335e-07,
+      "loss": 0.8802,
+      "step": 457
+    },
+    {
+      "epoch": 1.1253071253071254,
+      "grad_norm": 1.1981728862199128,
+      "learning_rate": 8.785335316999078e-07,
+      "loss": 0.9242,
+      "step": 458
+    },
+    {
+      "epoch": 1.1277641277641277,
+      "grad_norm": 1.5537897318688183,
+      "learning_rate": 8.745056921059839e-07,
+      "loss": 0.9001,
+      "step": 459
+    },
+    {
+      "epoch": 1.1302211302211302,
+      "grad_norm": 1.475963647033354,
+      "learning_rate": 8.704799199917672e-07,
+      "loss": 0.8775,
+      "step": 460
+    },
+    {
+      "epoch": 1.1326781326781328,
+      "grad_norm": 1.3982302407588918,
+      "learning_rate": 8.664562816806021e-07,
+      "loss": 0.8828,
+      "step": 461
+    },
+    {
+      "epoch": 1.135135135135135,
+      "grad_norm": 1.4892642936338247,
+      "learning_rate": 8.624348434606781e-07,
+      "loss": 0.9042,
+      "step": 462
+    },
+    {
+      "epoch": 1.1375921375921376,
+      "grad_norm": 1.8141339480004706,
+      "learning_rate": 8.584156715839401e-07,
+      "loss": 0.8748,
+      "step": 463
+    },
+    {
+      "epoch": 1.1400491400491402,
+      "grad_norm": 1.641438804529928,
+      "learning_rate": 8.543988322649953e-07,
+      "loss": 0.9006,
+      "step": 464
+    },
+    {
+      "epoch": 1.1425061425061425,
+      "grad_norm": 1.5215620001272283,
+      "learning_rate": 8.503843916800219e-07,
+      "loss": 0.9418,
+      "step": 465
+    },
+    {
+      "epoch": 1.144963144963145,
+      "grad_norm": 1.5095379213125475,
+      "learning_rate": 8.463724159656815e-07,
+      "loss": 0.9072,
+      "step": 466
+    },
+    {
+      "epoch": 1.1474201474201475,
+      "grad_norm": 1.4101920682297036,
+      "learning_rate": 8.423629712180265e-07,
+      "loss": 0.8929,
+      "step": 467
+    },
+    {
+      "epoch": 1.1498771498771498,
+      "grad_norm": 1.5017183949322925,
+      "learning_rate": 8.38356123491412e-07,
+      "loss": 0.9246,
+      "step": 468
+    },
+    {
+      "epoch": 1.1523341523341524,
+      "grad_norm": 1.4487713518305103,
+      "learning_rate": 8.34351938797409e-07,
+      "loss": 0.9978,
+      "step": 469
+    },
+    {
+      "epoch": 1.154791154791155,
+      "grad_norm": 1.9019229859778328,
+      "learning_rate": 8.303504831037153e-07,
+      "loss": 0.8963,
+      "step": 470
+    },
+    {
+      "epoch": 1.1572481572481572,
+      "grad_norm": 1.5553306743565323,
+      "learning_rate": 8.263518223330696e-07,
+      "loss": 0.9076,
+      "step": 471
+    },
+    {
+      "epoch": 1.1597051597051597,
+      "grad_norm": 1.528301521498108,
+      "learning_rate": 8.22356022362165e-07,
+      "loss": 0.9005,
+      "step": 472
+    },
+    {
+      "epoch": 1.1621621621621623,
+      "grad_norm": 1.4246755498283765,
+      "learning_rate": 8.183631490205635e-07,
+      "loss": 0.8751,
+      "step": 473
+    },
+    {
+      "epoch": 1.1646191646191646,
+      "grad_norm": 1.4837658920113785,
+      "learning_rate": 8.143732680896123e-07,
+      "loss": 0.8911,
+      "step": 474
+    },
+    {
+      "epoch": 1.1670761670761671,
+      "grad_norm": 1.3838573607006248,
+      "learning_rate": 8.103864453013592e-07,
+      "loss": 0.8752,
+      "step": 475
+    },
+    {
+      "epoch": 1.1695331695331694,
+      "grad_norm": 1.3956886722399857,
+      "learning_rate": 8.0640274633747e-07,
+      "loss": 0.907,
+      "step": 476
+    },
+    {
+      "epoch": 1.171990171990172,
+      "grad_norm": 1.4606205140333972,
+      "learning_rate": 8.024222368281468e-07,
+      "loss": 0.964,
+      "step": 477
+    },
+    {
+      "epoch": 1.1744471744471745,
+      "grad_norm": 1.438679071763127,
+      "learning_rate": 7.984449823510466e-07,
+      "loss": 0.8499,
+      "step": 478
+    },
+    {
+      "epoch": 1.1769041769041768,
+      "grad_norm": 1.6564632287349166,
+      "learning_rate": 7.944710484301994e-07,
+      "loss": 0.892,
+      "step": 479
+    },
+    {
+      "epoch": 1.1793611793611793,
+      "grad_norm": 1.5772645911948133,
+      "learning_rate": 7.905005005349319e-07,
+      "loss": 0.8451,
+      "step": 480
+    },
+    {
+      "epoch": 1.1818181818181819,
+      "grad_norm": 1.5408140168729385,
+      "learning_rate": 7.865334040787865e-07,
+      "loss": 0.9295,
+      "step": 481
+    },
+    {
+      "epoch": 1.1842751842751842,
+      "grad_norm": 1.5386445563903186,
+      "learning_rate": 7.825698244184431e-07,
+      "loss": 0.9029,
+      "step": 482
+    },
+    {
+      "epoch": 1.1867321867321867,
+      "grad_norm": 1.3266763848904035,
+      "learning_rate": 7.786098268526447e-07,
+      "loss": 0.8801,
+      "step": 483
+    },
+    {
+      "epoch": 1.1891891891891893,
+      "grad_norm": 1.7962161638699905,
+      "learning_rate": 7.746534766211206e-07,
+      "loss": 0.8578,
+      "step": 484
+    },
+    {
+      "epoch": 1.1916461916461916,
+      "grad_norm": 1.477358511769343,
+      "learning_rate": 7.7070083890351e-07,
+      "loss": 0.9048,
+      "step": 485
+    },
+    {
+      "epoch": 1.194103194103194,
+      "grad_norm": 1.4147723236609893,
+      "learning_rate": 7.667519788182911e-07,
+      "loss": 0.8898,
+      "step": 486
+    },
+    {
+      "epoch": 1.1965601965601966,
+      "grad_norm": 1.55983989969961,
+      "learning_rate": 7.628069614217057e-07,
+      "loss": 0.9509,
+      "step": 487
+    },
+    {
+      "epoch": 1.199017199017199,
+      "grad_norm": 1.4383185383531423,
+      "learning_rate": 7.588658517066892e-07,
+      "loss": 0.9418,
+      "step": 488
+    },
+    {
+      "epoch": 1.2014742014742015,
+      "grad_norm": 1.7624942370774765,
+      "learning_rate": 7.549287146017988e-07,
+      "loss": 0.8313,
+      "step": 489
+    },
+    {
+      "epoch": 1.203931203931204,
+      "grad_norm": 1.67001710532923,
+      "learning_rate": 7.509956149701443e-07,
+      "loss": 0.8549,
+      "step": 490
+    },
+    {
+      "epoch": 1.2063882063882063,
+      "grad_norm": 1.6268288582039303,
+      "learning_rate": 7.470666176083191e-07,
+      "loss": 0.8649,
+      "step": 491
+    },
+    {
+      "epoch": 1.2088452088452089,
+      "grad_norm": 1.4853773676445685,
+      "learning_rate": 7.431417872453339e-07,
+      "loss": 0.8956,
+      "step": 492
+    },
+    {
+      "epoch": 1.2113022113022114,
+      "grad_norm": 1.548578088799907,
+      "learning_rate": 7.39221188541548e-07,
+      "loss": 0.923,
+      "step": 493
+    },
+    {
+      "epoch": 1.2137592137592137,
+      "grad_norm": 1.710249233454194,
+      "learning_rate": 7.353048860876063e-07,
+      "loss": 0.9336,
+      "step": 494
+    },
+    {
+      "epoch": 1.2162162162162162,
+      "grad_norm": 1.5916044293544283,
+      "learning_rate": 7.31392944403375e-07,
+      "loss": 0.9304,
+      "step": 495
+    },
+    {
+      "epoch": 1.2186732186732188,
+      "grad_norm": 1.5308549158931122,
+      "learning_rate": 7.274854279368758e-07,
+      "loss": 0.9349,
+      "step": 496
+    },
+    {
+      "epoch": 1.221130221130221,
+      "grad_norm": 1.5229178156260872,
+      "learning_rate": 7.235824010632283e-07,
+      "loss": 0.9153,
+      "step": 497
+    },
+    {
+      "epoch": 1.2235872235872236,
+      "grad_norm": 1.462343882146908,
+      "learning_rate": 7.196839280835875e-07,
+      "loss": 0.9376,
+      "step": 498
+    },
+    {
+      "epoch": 1.2260442260442261,
+      "grad_norm": 1.5273674514727056,
+      "learning_rate": 7.157900732240827e-07,
+      "loss": 0.9231,
+      "step": 499
+    },
+    {
+      "epoch": 1.2285012285012284,
+      "grad_norm": 1.396230424270659,
+      "learning_rate": 7.119009006347624e-07,
+      "loss": 0.8868,
+      "step": 500
+    },
+    {
+      "epoch": 1.230958230958231,
+      "grad_norm": 1.6364859015395088,
+      "learning_rate": 7.080164743885363e-07,
+      "loss": 0.9287,
+      "step": 501
+    },
+    {
+      "epoch": 1.2334152334152333,
+      "grad_norm": 1.5048783187743708,
+      "learning_rate": 7.041368584801186e-07,
+      "loss": 0.932,
+      "step": 502
+    },
+    {
+      "epoch": 1.2358722358722358,
+      "grad_norm": 1.5929336784240413,
+      "learning_rate": 7.002621168249758e-07,
+      "loss": 0.9302,
+      "step": 503
+    },
+    {
+      "epoch": 1.2383292383292384,
+      "grad_norm": 1.5383384042486663,
+      "learning_rate": 6.963923132582715e-07,
+      "loss": 0.9059,
+      "step": 504
+    },
+    {
+      "epoch": 1.2407862407862407,
+      "grad_norm": 1.6790422515254697,
+      "learning_rate": 6.925275115338167e-07,
+      "loss": 0.8862,
+      "step": 505
+    },
+    {
+      "epoch": 1.2432432432432432,
+      "grad_norm": 1.4423572903428208,
+      "learning_rate": 6.886677753230183e-07,
+      "loss": 0.9004,
+      "step": 506
+    },
+    {
+      "epoch": 1.2457002457002457,
+      "grad_norm": 1.5146851789864428,
+      "learning_rate": 6.848131682138303e-07,
+      "loss": 0.8873,
+      "step": 507
+    },
+    {
+      "epoch": 1.248157248157248,
+      "grad_norm": 1.7060116939979457,
+      "learning_rate": 6.809637537097062e-07,
+      "loss": 0.9047,
+      "step": 508
+    },
+    {
+      "epoch": 1.2506142506142506,
+      "grad_norm": 1.3733517118208705,
+      "learning_rate": 6.77119595228554e-07,
+      "loss": 0.8742,
+      "step": 509
+    },
+    {
+      "epoch": 1.253071253071253,
+      "grad_norm": 1.4283683639472458,
+      "learning_rate": 6.732807561016885e-07,
+      "loss": 0.8697,
+      "step": 510
+    },
+    {
+      "epoch": 1.2555282555282554,
+      "grad_norm": 1.4832890039194944,
+      "learning_rate": 6.694472995727914e-07,
+      "loss": 0.925,
+      "step": 511
+    },
+    {
+      "epoch": 1.257985257985258,
+      "grad_norm": 1.4760580473983451,
+      "learning_rate": 6.656192887968674e-07,
+      "loss": 0.8959,
+      "step": 512
+    },
+    {
+      "epoch": 1.2604422604422605,
+      "grad_norm": 1.4517094927145753,
+      "learning_rate": 6.617967868392036e-07,
+      "loss": 0.8935,
+      "step": 513
+    },
+    {
+      "epoch": 1.2628992628992628,
+      "grad_norm": 1.5125319950260128,
+      "learning_rate": 6.579798566743313e-07,
+      "loss": 0.8556,
+      "step": 514
+    },
+    {
+      "epoch": 1.2653562653562653,
+      "grad_norm": 1.4510038145639286,
+      "learning_rate": 6.541685611849886e-07,
+      "loss": 0.9068,
+      "step": 515
+    },
+    {
+      "epoch": 1.2678132678132679,
+      "grad_norm": 1.786431321409276,
+      "learning_rate": 6.503629631610836e-07,
+      "loss": 0.8951,
+      "step": 516
+    },
+    {
+      "epoch": 1.2702702702702702,
+      "grad_norm": 4.646846867306252,
+      "learning_rate": 6.465631252986608e-07,
+      "loss": 0.8717,
+      "step": 517
+    },
+    {
+      "epoch": 1.2727272727272727,
+      "grad_norm": 1.6325223465886023,
+      "learning_rate": 6.427691101988672e-07,
+      "loss": 0.9277,
+      "step": 518
+    },
+    {
+      "epoch": 1.2751842751842752,
+      "grad_norm": 1.46996914036538,
+      "learning_rate": 6.389809803669225e-07,
+      "loss": 0.8614,
+      "step": 519
+    },
+    {
+      "epoch": 1.2776412776412776,
+      "grad_norm": 1.4901572767377598,
+      "learning_rate": 6.351987982110879e-07,
+      "loss": 0.8791,
+      "step": 520
+    },
+    {
+      "epoch": 1.28009828009828,
+      "grad_norm": 1.3956563694145843,
+      "learning_rate": 6.314226260416383e-07,
+      "loss": 0.8682,
+      "step": 521
+    },
+    {
+      "epoch": 1.2825552825552826,
+      "grad_norm": 1.527881875257898,
+      "learning_rate": 6.276525260698363e-07,
+      "loss": 0.9153,
+      "step": 522
+    },
+    {
+      "epoch": 1.285012285012285,
+      "grad_norm": 1.5192904446038689,
+      "learning_rate": 6.238885604069074e-07,
+      "loss": 0.9263,
+      "step": 523
+    },
+    {
+      "epoch": 1.2874692874692875,
+      "grad_norm": 1.5953405324354368,
+      "learning_rate": 6.201307910630145e-07,
+      "loss": 0.8788,
+      "step": 524
+    },
+    {
+      "epoch": 1.28992628992629,
+      "grad_norm": 1.4403593350780215,
+      "learning_rate": 6.163792799462402e-07,
+      "loss": 0.9414,
+      "step": 525
+    },
+    {
+      "epoch": 1.2923832923832923,
+      "grad_norm": 1.4506987352971705,
+      "learning_rate": 6.126340888615642e-07,
+      "loss": 0.9307,
+      "step": 526
+    },
+    {
+      "epoch": 1.2948402948402948,
+      "grad_norm": 1.610182690312099,
+      "learning_rate": 6.088952795098441e-07,
+      "loss": 0.9168,
+      "step": 527
+    },
+    {
+      "epoch": 1.2972972972972974,
+      "grad_norm": 1.5978713160514366,
+      "learning_rate": 6.051629134868019e-07,
+      "loss": 0.8393,
+      "step": 528
+    },
+    {
+      "epoch": 1.2997542997542997,
+      "grad_norm": 1.3547671161607415,
+      "learning_rate": 6.014370522820085e-07,
+      "loss": 0.8972,
+      "step": 529
+    },
+    {
+      "epoch": 1.3022113022113022,
+      "grad_norm": 1.3090769200596557,
+      "learning_rate": 5.977177572778677e-07,
+      "loss": 0.8823,
+      "step": 530
+    },
+    {
+      "epoch": 1.3046683046683047,
+      "grad_norm": 1.4761807072270576,
+      "learning_rate": 5.940050897486089e-07,
+      "loss": 0.9733,
+      "step": 531
+    },
+    {
+      "epoch": 1.307125307125307,
+      "grad_norm": 1.5243972285729563,
+      "learning_rate": 5.902991108592754e-07,
+      "loss": 0.8859,
+      "step": 532
+    },
+    {
+      "epoch": 1.3095823095823096,
+      "grad_norm": 1.764985461344906,
+      "learning_rate": 5.865998816647171e-07,
+      "loss": 0.8868,
+      "step": 533
+    },
+    {
+      "epoch": 1.3120393120393121,
+      "grad_norm": 1.6317180984567103,
+      "learning_rate": 5.829074631085852e-07,
+      "loss": 0.9361,
+      "step": 534
+    },
+    {
+      "epoch": 1.3144963144963144,
+      "grad_norm": 1.4981668702031359,
+      "learning_rate": 5.792219160223267e-07,
+      "loss": 0.8472,
+      "step": 535
+    },
+    {
+      "epoch": 1.316953316953317,
+      "grad_norm": 2.637743576229916,
+      "learning_rate": 5.755433011241851e-07,
+      "loss": 0.8736,
+      "step": 536
+    },
+    {
+      "epoch": 1.3194103194103195,
+      "grad_norm": 1.3474590182260624,
+      "learning_rate": 5.718716790181965e-07,
+      "loss": 0.8794,
+      "step": 537
+    },
+    {
+      "epoch": 1.3218673218673218,
+      "grad_norm": 1.42787784398219,
+      "learning_rate": 5.682071101931936e-07,
+      "loss": 0.8738,
+      "step": 538
+    },
+    {
+      "epoch": 1.3243243243243243,
+      "grad_norm": 1.6067570011968997,
+      "learning_rate": 5.645496550218089e-07,
+      "loss": 0.8879,
+      "step": 539
+    },
+    {
+      "epoch": 1.3267813267813269,
+      "grad_norm": 1.5914444312640619,
+      "learning_rate": 5.60899373759479e-07,
+      "loss": 0.8754,
+      "step": 540
+    },
+    {
+      "epoch": 1.3292383292383292,
+      "grad_norm": 1.4758896244599573,
+      "learning_rate": 5.572563265434526e-07,
+      "loss": 0.9307,
+      "step": 541
+    },
+    {
+      "epoch": 1.3316953316953317,
+      "grad_norm": 1.2626534960911335,
+      "learning_rate": 5.536205733918007e-07,
+      "loss": 0.843,
+      "step": 542
+    },
+    {
+      "epoch": 1.3341523341523343,
+      "grad_norm": 1.701842268857865,
+      "learning_rate": 5.499921742024257e-07,
+      "loss": 0.9247,
+      "step": 543
+    },
+    {
+      "epoch": 1.3366093366093366,
+      "grad_norm": 1.4205425380782735,
+      "learning_rate": 5.463711887520759e-07,
+      "loss": 0.8628,
+      "step": 544
+    },
+    {
+      "epoch": 1.339066339066339,
+      "grad_norm": 1.9140145931124193,
+      "learning_rate": 5.427576766953614e-07,
+      "loss": 0.9073,
+      "step": 545
+    },
+    {
+      "epoch": 1.3415233415233416,
+      "grad_norm": 1.7824661674136197,
+      "learning_rate": 5.391516975637699e-07,
+      "loss": 0.9133,
+      "step": 546
+    },
+    {
+      "epoch": 1.343980343980344,
+      "grad_norm": 1.5617612083029389,
+      "learning_rate": 5.355533107646858e-07,
+      "loss": 0.9531,
+      "step": 547
+    },
+    {
+      "epoch": 1.3464373464373465,
+      "grad_norm": 1.5245643140200376,
+      "learning_rate": 5.319625755804138e-07,
+      "loss": 0.8915,
+      "step": 548
+    },
+    {
+      "epoch": 1.348894348894349,
+      "grad_norm": 1.7286911090374188,
+      "learning_rate": 5.283795511671994e-07,
+      "loss": 0.863,
+      "step": 549
+    },
+    {
+      "epoch": 1.3513513513513513,
+      "grad_norm": 1.6181660068040473,
+      "learning_rate": 5.248042965542559e-07,
+      "loss": 0.8852,
+      "step": 550
+    },
+    {
+      "epoch": 1.3538083538083538,
+      "grad_norm": 1.5075998723303305,
+      "learning_rate": 5.212368706427912e-07,
+      "loss": 0.8965,
+      "step": 551
+    },
+    {
+      "epoch": 1.3562653562653564,
+      "grad_norm": 1.5100541752180852,
+      "learning_rate": 5.176773322050381e-07,
+      "loss": 0.8829,
+      "step": 552
+    },
+    {
+      "epoch": 1.3587223587223587,
+      "grad_norm": 1.863955084942053,
+      "learning_rate": 5.141257398832862e-07,
+      "loss": 0.9475,
+      "step": 553
+    },
+    {
+      "epoch": 1.3611793611793612,
+      "grad_norm": 1.314918195852255,
+      "learning_rate": 5.105821521889147e-07,
+      "loss": 0.8643,
+      "step": 554
+    },
+    {
+      "epoch": 1.3636363636363638,
+      "grad_norm": 1.8350990255330366,
+      "learning_rate": 5.070466275014287e-07,
+      "loss": 0.9441,
+      "step": 555
+    },
+    {
+      "epoch": 1.366093366093366,
+      "grad_norm": 1.5322707436835765,
+      "learning_rate": 5.03519224067499e-07,
+      "loss": 0.9033,
+      "step": 556
+    },
+    {
+      "epoch": 1.3685503685503686,
+      "grad_norm": 1.5974243335350318,
+      "learning_rate": 5.000000000000002e-07,
+      "loss": 0.8914,
+      "step": 557
+    },
+    {
+      "epoch": 1.3710073710073711,
+      "grad_norm": 1.4435566226659817,
+      "learning_rate": 4.964890132770543e-07,
+      "loss": 0.878,
+      "step": 558
+    },
+    {
+      "epoch": 1.3734643734643734,
+      "grad_norm": 1.8477167694746348,
+      "learning_rate": 4.929863217410767e-07,
+      "loss": 0.8909,
+      "step": 559
+    },
+    {
+      "epoch": 1.375921375921376,
+      "grad_norm": 1.4801649083461632,
+      "learning_rate": 4.894919830978212e-07,
+      "loss": 0.9079,
+      "step": 560
+    },
+    {
+      "epoch": 1.3783783783783785,
+      "grad_norm": 1.6811015996059142,
+      "learning_rate": 4.860060549154301e-07,
+      "loss": 0.815,
+      "step": 561
+    },
+    {
+      "epoch": 1.3808353808353808,
+      "grad_norm": 2.3145109947990155,
+      "learning_rate": 4.825285946234874e-07,
+      "loss": 0.9305,
+      "step": 562
+    },
+    {
+      "epoch": 1.3832923832923834,
+      "grad_norm": 1.4759252526730806,
+      "learning_rate": 4.790596595120698e-07,
+      "loss": 0.8589,
+      "step": 563
+    },
+    {
+      "epoch": 1.3857493857493859,
+      "grad_norm": 1.3378195425536465,
+      "learning_rate": 4.7559930673080475e-07,
+      "loss": 0.8659,
+      "step": 564
+    },
+    {
+      "epoch": 1.3882063882063882,
+      "grad_norm": 1.6393333264623262,
+      "learning_rate": 4.721475932879282e-07,
+      "loss": 0.8618,
+      "step": 565
+    },
+    {
+      "epoch": 1.3906633906633907,
+      "grad_norm": 1.6916138016252256,
+      "learning_rate": 4.6870457604934675e-07,
+      "loss": 0.9028,
+      "step": 566
+    },
+    {
+      "epoch": 1.393120393120393,
+      "grad_norm": 1.4225654726940884,
+      "learning_rate": 4.6527031173769857e-07,
+      "loss": 0.8765,
+      "step": 567
+    },
+    {
+      "epoch": 1.3955773955773956,
+      "grad_norm": 1.342379943145193,
+      "learning_rate": 4.618448569314206e-07,
+      "loss": 0.8855,
+      "step": 568
+    },
+    {
+      "epoch": 1.398034398034398,
+      "grad_norm": 2.442614220259753,
+      "learning_rate": 4.584282680638154e-07,
+      "loss": 0.8804,
+      "step": 569
+    },
+    {
+      "epoch": 1.4004914004914004,
+      "grad_norm": 1.5267874267053692,
+      "learning_rate": 4.550206014221232e-07,
+      "loss": 0.9668,
+      "step": 570
+    },
+    {
+      "epoch": 1.402948402948403,
+      "grad_norm": 1.8692131154426705,
+      "learning_rate": 4.5162191314659194e-07,
+      "loss": 0.8979,
+      "step": 571
+    },
+    {
+      "epoch": 1.4054054054054055,
+      "grad_norm": 2.022804110507779,
+      "learning_rate": 4.48232259229554e-07,
+      "loss": 0.8396,
+      "step": 572
+    },
+    {
+      "epoch": 1.4078624078624078,
+      "grad_norm": 1.444700494646088,
+      "learning_rate": 4.448516955145047e-07,
+      "loss": 0.8798,
+      "step": 573
+    },
+    {
+      "epoch": 1.4103194103194103,
+      "grad_norm": 1.5147425793812193,
+      "learning_rate": 4.414802776951798e-07,
+      "loss": 0.884,
+      "step": 574
+    },
+    {
+      "epoch": 1.4127764127764126,
+      "grad_norm": 1.4905275337775719,
+      "learning_rate": 4.3811806131463946e-07,
+      "loss": 0.8755,
+      "step": 575
+    },
+    {
+      "epoch": 1.4152334152334152,
+      "grad_norm": 1.5203731499114503,
+      "learning_rate": 4.347651017643539e-07,
+      "loss": 0.9089,
+      "step": 576
+    },
+    {
+      "epoch": 1.4176904176904177,
+      "grad_norm": 1.4792627181832774,
+      "learning_rate": 4.314214542832888e-07,
+      "loss": 0.8995,
+      "step": 577
+    },
+    {
+      "epoch": 1.42014742014742,
+      "grad_norm": 1.498952396682308,
+      "learning_rate": 4.280871739569971e-07,
+      "loss": 0.9,
+      "step": 578
+    },
+    {
+      "epoch": 1.4226044226044225,
+      "grad_norm": 1.5993242263021854,
+      "learning_rate": 4.247623157167102e-07,
+      "loss": 0.9004,
+      "step": 579
+    },
+    {
+      "epoch": 1.425061425061425,
+      "grad_norm": 1.2909684524904117,
+      "learning_rate": 4.214469343384346e-07,
+      "loss": 0.9442,
+      "step": 580
+    },
+    {
+      "epoch": 1.4275184275184274,
+      "grad_norm": 1.6601643291459118,
+      "learning_rate": 4.181410844420473e-07,
+      "loss": 0.8924,
+      "step": 581
+    },
+    {
+      "epoch": 1.42997542997543,
+      "grad_norm": 1.5456219866657583,
+      "learning_rate": 4.148448204903977e-07,
+      "loss": 0.8785,
+      "step": 582
+    },
+    {
+      "epoch": 1.4324324324324325,
+      "grad_norm": 2.3413636182110906,
+      "learning_rate": 4.115581967884093e-07,
+      "loss": 0.8867,
+      "step": 583
+    },
+    {
+      "epoch": 1.4348894348894348,
+      "grad_norm": 1.600188568539889,
+      "learning_rate": 4.082812674821865e-07,
+      "loss": 0.9163,
+      "step": 584
+    },
+    {
+      "epoch": 1.4373464373464373,
+      "grad_norm": 1.775885867839352,
+      "learning_rate": 4.0501408655812044e-07,
+      "loss": 0.94,
+      "step": 585
+    },
+    {
+      "epoch": 1.4398034398034398,
+      "grad_norm": 1.4853461496779028,
+      "learning_rate": 4.0175670784200066e-07,
+      "loss": 0.8828,
+      "step": 586
+    },
+    {
+      "epoch": 1.4422604422604421,
+      "grad_norm": 1.3437874918142632,
+      "learning_rate": 3.985091849981297e-07,
+      "loss": 0.906,
+      "step": 587
+    },
+    {
+      "epoch": 1.4447174447174447,
+      "grad_norm": 1.8022847542011342,
+      "learning_rate": 3.952715715284363e-07,
+      "loss": 0.9078,
+      "step": 588
+    },
+    {
+      "epoch": 1.4471744471744472,
+      "grad_norm": 1.4544558952772775,
+      "learning_rate": 3.9204392077159544e-07,
+      "loss": 0.9007,
+      "step": 589
+    },
+    {
+      "epoch": 1.4496314496314495,
+      "grad_norm": 1.5362982163739145,
+      "learning_rate": 3.888262859021507e-07,
+      "loss": 0.9329,
+      "step": 590
+    },
+    {
+      "epoch": 1.452088452088452,
+      "grad_norm": 1.35776640743268,
+      "learning_rate": 3.856187199296358e-07,
+      "loss": 0.9142,
+      "step": 591
+    },
+    {
+      "epoch": 1.4545454545454546,
+      "grad_norm": 1.552024766985016,
+      "learning_rate": 3.8242127569770265e-07,
+      "loss": 0.8624,
+      "step": 592
+    },
+    {
+      "epoch": 1.457002457002457,
+      "grad_norm": 1.6837813084688251,
+      "learning_rate": 3.792340058832515e-07,
+      "loss": 0.8515,
+      "step": 593
+    },
+    {
+      "epoch": 1.4594594594594594,
+      "grad_norm": 1.6200568897908438,
+      "learning_rate": 3.760569629955613e-07,
+      "loss": 0.9182,
+      "step": 594
+    },
+    {
+      "epoch": 1.461916461916462,
+      "grad_norm": 1.3613611600691327,
+      "learning_rate": 3.72890199375426e-07,
+      "loss": 0.9212,
+      "step": 595
+    },
+    {
+      "epoch": 1.4643734643734643,
+      "grad_norm": 1.5082364375002426,
+      "learning_rate": 3.697337671942913e-07,
+      "loss": 0.9084,
+      "step": 596
+    },
+    {
+      "epoch": 1.4668304668304668,
+      "grad_norm": 1.6876950054566262,
+      "learning_rate": 3.6658771845339676e-07,
+      "loss": 0.9065,
+      "step": 597
+    },
+    {
+      "epoch": 1.4692874692874693,
+      "grad_norm": 1.5733845918989662,
+      "learning_rate": 3.634521049829169e-07,
+      "loss": 0.8815,
+      "step": 598
+    },
+    {
+      "epoch": 1.4717444717444716,
+      "grad_norm": 1.8147232098247035,
+      "learning_rate": 3.603269784411089e-07,
+      "loss": 0.9133,
+      "step": 599
+    },
+    {
+      "epoch": 1.4742014742014742,
+      "grad_norm": 1.4595323005431404,
+      "learning_rate": 3.5721239031346063e-07,
+      "loss": 0.8071,
+      "step": 600
+    },
+    {
+      "epoch": 1.4766584766584767,
+      "grad_norm": 2.769727304993137,
+      "learning_rate": 3.541083919118438e-07,
+      "loss": 0.9423,
+      "step": 601
+    },
+    {
+      "epoch": 1.479115479115479,
+      "grad_norm": 1.610693658075675,
+      "learning_rate": 3.5101503437366675e-07,
+      "loss": 0.8946,
+      "step": 602
+    },
+    {
+      "epoch": 1.4815724815724816,
+      "grad_norm": 1.6365761390735136,
+      "learning_rate": 3.479323686610329e-07,
+      "loss": 0.9079,
+      "step": 603
+    },
+    {
+      "epoch": 1.484029484029484,
+      "grad_norm": 1.5083166362769518,
+      "learning_rate": 3.448604455599021e-07,
+      "loss": 0.8975,
+      "step": 604
+    },
+    {
+      "epoch": 1.4864864864864864,
+      "grad_norm": 1.60995686456512,
+      "learning_rate": 3.4179931567925214e-07,
+      "loss": 0.9498,
+      "step": 605
+    },
+    {
+      "epoch": 1.488943488943489,
+      "grad_norm": 1.7185280239304535,
+      "learning_rate": 3.387490294502456e-07,
+      "loss": 0.906,
+      "step": 606
+    },
+    {
+      "epoch": 1.4914004914004915,
+      "grad_norm": 1.321610762107921,
+      "learning_rate": 3.357096371254008e-07,
+      "loss": 0.8499,
+      "step": 607
+    },
+    {
+      "epoch": 1.4938574938574938,
+      "grad_norm": 1.2808013142695502,
+      "learning_rate": 3.3268118877776064e-07,
+      "loss": 0.8546,
+      "step": 608
+    },
+    {
+      "epoch": 1.4963144963144963,
+      "grad_norm": 1.3917333892709016,
+      "learning_rate": 3.296637343000704e-07,
+      "loss": 0.8776,
+      "step": 609
+    },
+    {
+      "epoch": 1.4987714987714988,
+      "grad_norm": 1.9760211104870224,
+      "learning_rate": 3.266573234039541e-07,
+      "loss": 0.8829,
+      "step": 610
+    },
+    {
+      "epoch": 1.5012285012285012,
+      "grad_norm": 1.9349300228421435,
+      "learning_rate": 3.2366200561909717e-07,
+      "loss": 0.9721,
+      "step": 611
+    },
+    {
+      "epoch": 1.5036855036855037,
+      "grad_norm": 1.5825092236415208,
+      "learning_rate": 3.2067783029242865e-07,
+      "loss": 0.8569,
+      "step": 612
+    },
+    {
+      "epoch": 1.5061425061425062,
+      "grad_norm": 1.7071421937295232,
+      "learning_rate": 3.1770484658730887e-07,
+      "loss": 0.9586,
+      "step": 613
+    },
+    {
+      "epoch": 1.5085995085995085,
+      "grad_norm": 1.3114653023022718,
+      "learning_rate": 3.147431034827208e-07,
+      "loss": 0.8879,
+      "step": 614
+    },
+    {
+      "epoch": 1.511056511056511,
+      "grad_norm": 1.3660758453242297,
+      "learning_rate": 3.11792649772461e-07,
+      "loss": 0.9227,
+      "step": 615
+    },
+    {
+      "epoch": 1.5135135135135136,
+      "grad_norm": 1.5443322628991898,
+      "learning_rate": 3.08853534064337e-07,
+      "loss": 0.8756,
+      "step": 616
+    },
+    {
+      "epoch": 1.515970515970516,
+      "grad_norm": 1.6205632640066696,
+      "learning_rate": 3.0592580477936604e-07,
+      "loss": 0.8631,
+      "step": 617
+    },
+    {
+      "epoch": 1.5184275184275184,
+      "grad_norm": 1.6750050731142352,
+      "learning_rate": 3.0300951015097863e-07,
+      "loss": 0.9308,
+      "step": 618
+    },
+    {
+      "epoch": 1.520884520884521,
+      "grad_norm": 1.7541208454820925,
+      "learning_rate": 3.001046982242216e-07,
+      "loss": 0.9097,
+      "step": 619
+    },
+    {
+      "epoch": 1.5233415233415233,
+      "grad_norm": 1.9317379819011606,
+      "learning_rate": 2.972114168549682e-07,
+      "loss": 0.9212,
+      "step": 620
+    },
+    {
+      "epoch": 1.5257985257985258,
+      "grad_norm": 1.2650696951039682,
+      "learning_rate": 2.9432971370913e-07,
+      "loss": 0.8836,
+      "step": 621
+    },
+    {
+      "epoch": 1.5282555282555284,
+      "grad_norm": 1.4960196905274168,
+      "learning_rate": 2.914596362618701e-07,
+      "loss": 0.8946,
+      "step": 622
+    },
+    {
+      "epoch": 1.5307125307125307,
+      "grad_norm": 1.507394873293028,
+      "learning_rate": 2.8860123179682236e-07,
+      "loss": 0.9141,
+      "step": 623
+    },
+    {
+      "epoch": 1.5331695331695332,
+      "grad_norm": 1.3335466599194608,
+      "learning_rate": 2.8575454740531115e-07,
+      "loss": 0.8667,
+      "step": 624
+    },
+    {
+      "epoch": 1.5356265356265357,
+      "grad_norm": 1.522699880941422,
+      "learning_rate": 2.829196299855775e-07,
+      "loss": 0.9901,
+      "step": 625
+    },
+    {
+      "epoch": 1.538083538083538,
+      "grad_norm": 1.4978237170833106,
+      "learning_rate": 2.800965262420043e-07,
+      "loss": 0.8367,
+      "step": 626
+    },
+    {
+      "epoch": 1.5405405405405406,
+      "grad_norm": 1.386657482706535,
+      "learning_rate": 2.7728528268434757e-07,
+      "loss": 0.8411,
+      "step": 627
+    },
+    {
+      "epoch": 1.542997542997543,
+      "grad_norm": 1.6884446408530738,
+      "learning_rate": 2.744859456269718e-07,
+      "loss": 0.942,
+      "step": 628
+    },
+    {
+      "epoch": 1.5454545454545454,
+      "grad_norm": 1.5125614736127029,
+      "learning_rate": 2.716985611880841e-07,
+      "loss": 0.8738,
+      "step": 629
+    },
+    {
+      "epoch": 1.547911547911548,
+      "grad_norm": 1.469704160523891,
+      "learning_rate": 2.689231752889768e-07,
+      "loss": 0.9165,
+      "step": 630
+    },
+    {
+      "epoch": 1.5503685503685505,
+      "grad_norm": 1.4870015368902652,
+      "learning_rate": 2.661598336532692e-07,
+      "loss": 0.8944,
+      "step": 631
+    },
+    {
+      "epoch": 1.5528255528255528,
+      "grad_norm": 1.4322815674497984,
+      "learning_rate": 2.6340858180615645e-07,
+      "loss": 0.8968,
+      "step": 632
+    },
+    {
+      "epoch": 1.5552825552825553,
+      "grad_norm": 1.3623692296259504,
+      "learning_rate": 2.6066946507365696e-07,
+      "loss": 0.8935,
+      "step": 633
+    },
+    {
+      "epoch": 1.5577395577395579,
+      "grad_norm": 1.484336106115407,
+      "learning_rate": 2.5794252858186706e-07,
+      "loss": 0.8993,
+      "step": 634
+    },
+    {
+      "epoch": 1.5601965601965602,
+      "grad_norm": 1.4403746721888513,
+      "learning_rate": 2.552278172562181e-07,
+      "loss": 0.8937,
+      "step": 635
+    },
+    {
+      "epoch": 1.5626535626535627,
+      "grad_norm": 1.8086294010381407,
+      "learning_rate": 2.525253758207353e-07,
+      "loss": 0.939,
+      "step": 636
+    },
+    {
+      "epoch": 1.5651105651105652,
+      "grad_norm": 1.4766926006894956,
+      "learning_rate": 2.4983524879730066e-07,
+      "loss": 0.9023,
+      "step": 637
+    },
+    {
+      "epoch": 1.5675675675675675,
+      "grad_norm": 1.3776357069750294,
+      "learning_rate": 2.471574805049206e-07,
+      "loss": 0.9233,
+      "step": 638
+    },
+    {
+      "epoch": 1.57002457002457,
+      "grad_norm": 1.6402577176804487,
+      "learning_rate": 2.444921150589957e-07,
+      "loss": 0.9319,
+      "step": 639
+    },
+    {
+      "epoch": 1.5724815724815726,
+      "grad_norm": 1.633491894306784,
+      "learning_rate": 2.418391963705927e-07,
+      "loss": 0.8818,
+      "step": 640
+    },
+    {
+      "epoch": 1.574938574938575,
+      "grad_norm": 1.5059260563661474,
+      "learning_rate": 2.391987681457219e-07,
+      "loss": 0.9138,
+      "step": 641
+    },
+    {
+      "epoch": 1.5773955773955772,
+      "grad_norm": 1.474101178051224,
+      "learning_rate": 2.3657087388461815e-07,
+      "loss": 0.8847,
+      "step": 642
+    },
+    {
+      "epoch": 1.57985257985258,
+      "grad_norm": 1.5606071866155247,
+      "learning_rate": 2.339555568810221e-07,
+      "loss": 0.9451,
+      "step": 643
+    },
+    {
+      "epoch": 1.5823095823095823,
+      "grad_norm": 1.5959144837669506,
+      "learning_rate": 2.3135286022146782e-07,
+      "loss": 0.9202,
+      "step": 644
+    },
+    {
+      "epoch": 1.5847665847665846,
+      "grad_norm": 1.4739665598722158,
+      "learning_rate": 2.2876282678457438e-07,
+      "loss": 0.8617,
+      "step": 645
+    },
+    {
+      "epoch": 1.5872235872235874,
+      "grad_norm": 1.324486221068965,
+      "learning_rate": 2.2618549924033704e-07,
+      "loss": 0.8702,
+      "step": 646
+    },
+    {
+      "epoch": 1.5896805896805897,
+      "grad_norm": 1.3762365282216713,
+      "learning_rate": 2.236209200494258e-07,
+      "loss": 0.9137,
+      "step": 647
+    },
+    {
+      "epoch": 1.592137592137592,
+      "grad_norm": 1.5081639071286088,
+      "learning_rate": 2.210691314624854e-07,
+      "loss": 0.8869,
+      "step": 648
+    },
+    {
+      "epoch": 1.5945945945945947,
+      "grad_norm": 1.5818213999124708,
+      "learning_rate": 2.1853017551943987e-07,
+      "loss": 0.9312,
+      "step": 649
+    },
+    {
+      "epoch": 1.597051597051597,
+      "grad_norm": 1.7757893297397498,
+      "learning_rate": 2.160040940487987e-07,
+      "loss": 0.8902,
+      "step": 650
+    },
+    {
+      "epoch": 1.5995085995085994,
+      "grad_norm": 1.3784061956740852,
+      "learning_rate": 2.1349092866696906e-07,
+      "loss": 0.8984,
+      "step": 651
+    },
+    {
+      "epoch": 1.6019656019656021,
+      "grad_norm": 1.4232738419586703,
+      "learning_rate": 2.1099072077756863e-07,
+      "loss": 0.8676,
+      "step": 652
+    },
+    {
+      "epoch": 1.6044226044226044,
+      "grad_norm": 3.799736709775996,
+      "learning_rate": 2.0850351157074597e-07,
+      "loss": 0.9107,
+      "step": 653
+    },
+    {
+      "epoch": 1.6068796068796067,
+      "grad_norm": 2.8003735792292055,
+      "learning_rate": 2.060293420224991e-07,
+      "loss": 0.8836,
+      "step": 654
+    },
+    {
+      "epoch": 1.6093366093366095,
+      "grad_norm": 1.320688739165864,
+      "learning_rate": 2.0356825289400183e-07,
+      "loss": 0.8912,
+      "step": 655
+    },
+    {
+      "epoch": 1.6117936117936118,
+      "grad_norm": 1.9634181459213618,
+      "learning_rate": 2.011202847309329e-07,
+      "loss": 0.9238,
+      "step": 656
+    },
+    {
+      "epoch": 1.6142506142506141,
+      "grad_norm": 1.5624806988822288,
+      "learning_rate": 1.9868547786280665e-07,
+      "loss": 0.8707,
+      "step": 657
+    },
+    {
+      "epoch": 1.6167076167076169,
+      "grad_norm": 1.4107049261598619,
+      "learning_rate": 1.9626387240230891e-07,
+      "loss": 0.868,
+      "step": 658
+    },
+    {
+      "epoch": 1.6191646191646192,
+      "grad_norm": 1.492505632847286,
+      "learning_rate": 1.9385550824463725e-07,
+      "loss": 0.8894,
+      "step": 659
+    },
+    {
+      "epoch": 1.6216216216216215,
+      "grad_norm": 1.4382690869482055,
+      "learning_rate": 1.9146042506684223e-07,
+      "loss": 0.8978,
+      "step": 660
+    },
+    {
+      "epoch": 1.6240786240786242,
+      "grad_norm": 1.627212055293953,
+      "learning_rate": 1.890786623271743e-07,
+      "loss": 0.871,
+      "step": 661
+    },
+    {
+      "epoch": 1.6265356265356266,
+      "grad_norm": 1.6843094898216577,
+      "learning_rate": 1.8671025926443463e-07,
+      "loss": 0.8952,
+      "step": 662
+    },
+    {
+      "epoch": 1.6289926289926289,
+      "grad_norm": 1.6654696363977666,
+      "learning_rate": 1.8435525489732718e-07,
+      "loss": 0.8591,
+      "step": 663
+    },
+    {
+      "epoch": 1.6314496314496314,
+      "grad_norm": 1.5188779880666268,
+      "learning_rate": 1.8201368802381657e-07,
+      "loss": 0.9323,
+      "step": 664
+    },
+    {
+      "epoch": 1.633906633906634,
+      "grad_norm": 1.3399754864662434,
+      "learning_rate": 1.7968559722048903e-07,
+      "loss": 0.8543,
+      "step": 665
+    },
+    {
+      "epoch": 1.6363636363636362,
+      "grad_norm": 1.3715522224138303,
+      "learning_rate": 1.7737102084191703e-07,
+      "loss": 0.918,
+      "step": 666
+    },
+    {
+      "epoch": 1.6388206388206388,
+      "grad_norm": 1.6259531621746015,
+      "learning_rate": 1.7506999702002678e-07,
+      "loss": 0.8694,
+      "step": 667
+    },
+    {
+      "epoch": 1.6412776412776413,
+      "grad_norm": 2.365813450212486,
+      "learning_rate": 1.7278256366347032e-07,
+      "loss": 0.8874,
+      "step": 668
+    },
+    {
+      "epoch": 1.6437346437346436,
+      "grad_norm": 1.7862394193970157,
+      "learning_rate": 1.7050875845700064e-07,
+      "loss": 0.8803,
+      "step": 669
+    },
+    {
+      "epoch": 1.6461916461916462,
+      "grad_norm": 1.557725205467593,
+      "learning_rate": 1.682486188608523e-07,
+      "loss": 0.8842,
+      "step": 670
+    },
+    {
+      "epoch": 1.6486486486486487,
+      "grad_norm": 1.6097866192991963,
+      "learning_rate": 1.6600218211012218e-07,
+      "loss": 0.9428,
+      "step": 671
+    },
+    {
+      "epoch": 1.651105651105651,
+      "grad_norm": 1.4644367524635458,
+      "learning_rate": 1.6376948521415711e-07,
+      "loss": 0.9066,
+      "step": 672
+    },
+    {
+      "epoch": 1.6535626535626535,
+      "grad_norm": 1.6645946570893102,
+      "learning_rate": 1.6155056495594465e-07,
+      "loss": 0.905,
+      "step": 673
+    },
+    {
+      "epoch": 1.656019656019656,
+      "grad_norm": 1.3397486338053894,
+      "learning_rate": 1.5934545789150622e-07,
+      "loss": 0.8785,
+      "step": 674
+    },
+    {
+      "epoch": 1.6584766584766584,
+      "grad_norm": 1.3685689789923863,
+      "learning_rate": 1.5715420034929473e-07,
+      "loss": 0.9179,
+      "step": 675
+    },
+    {
+      "epoch": 1.660933660933661,
+      "grad_norm": 1.5445547208764443,
+      "learning_rate": 1.5497682842959747e-07,
+      "loss": 0.9187,
+      "step": 676
+    },
+    {
+      "epoch": 1.6633906633906634,
+      "grad_norm": 1.7915250554930906,
+      "learning_rate": 1.5281337800393968e-07,
+      "loss": 0.8996,
+      "step": 677
+    },
+    {
+      "epoch": 1.6658476658476657,
+      "grad_norm": 1.425206424154818,
+      "learning_rate": 1.5066388471449453e-07,
+      "loss": 0.9244,
+      "step": 678
+    },
+    {
+      "epoch": 1.6683046683046683,
+      "grad_norm": 1.3711937617299463,
+      "learning_rate": 1.4852838397349542e-07,
+      "loss": 0.9239,
+      "step": 679
+    },
+    {
+      "epoch": 1.6707616707616708,
+      "grad_norm": 1.4206125252604587,
+      "learning_rate": 1.4640691096265357e-07,
+      "loss": 0.9127,
+      "step": 680
+    },
+    {
+      "epoch": 1.6732186732186731,
+      "grad_norm": 1.5824939121397916,
+      "learning_rate": 1.4429950063257712e-07,
+      "loss": 0.9243,
+      "step": 681
+    },
+    {
+      "epoch": 1.6756756756756757,
+      "grad_norm": 1.5523209337418313,
+      "learning_rate": 1.4220618770219604e-07,
+      "loss": 0.8743,
+      "step": 682
+    },
+    {
+      "epoch": 1.6781326781326782,
+      "grad_norm": 2.449668270011937,
+      "learning_rate": 1.4012700665818988e-07,
+      "loss": 0.9022,
+      "step": 683
+    },
+    {
+      "epoch": 1.6805896805896805,
+      "grad_norm": 2.797795087084986,
+      "learning_rate": 1.3806199175442046e-07,
+      "loss": 0.9127,
+      "step": 684
+    },
+    {
+      "epoch": 1.683046683046683,
+      "grad_norm": 1.594299579779607,
+      "learning_rate": 1.3601117701136588e-07,
+      "loss": 0.8958,
+      "step": 685
+    },
+    {
+      "epoch": 1.6855036855036856,
+      "grad_norm": 1.3744841196118298,
+      "learning_rate": 1.3397459621556128e-07,
+      "loss": 0.8847,
+      "step": 686
+    },
+    {
+      "epoch": 1.6879606879606879,
+      "grad_norm": 1.4967014666487075,
+      "learning_rate": 1.319522829190427e-07,
+      "loss": 0.9375,
+      "step": 687
+    },
+    {
+      "epoch": 1.6904176904176904,
+      "grad_norm": 1.257226409390929,
+      "learning_rate": 1.2994427043879275e-07,
+      "loss": 0.87,
+      "step": 688
+    },
+    {
+      "epoch": 1.692874692874693,
+      "grad_norm": 1.5489462493350459,
+      "learning_rate": 1.279505918561923e-07,
+      "loss": 0.8997,
+      "step": 689
+    },
+    {
+      "epoch": 1.6953316953316953,
+      "grad_norm": 1.478167106209186,
+      "learning_rate": 1.2597128001647695e-07,
+      "loss": 0.88,
+      "step": 690
+    },
+    {
+      "epoch": 1.6977886977886978,
+      "grad_norm": 1.530749437315356,
+      "learning_rate": 1.2400636752819337e-07,
+      "loss": 0.9155,
+      "step": 691
+    },
+    {
+      "epoch": 1.7002457002457003,
+      "grad_norm": 1.312749085771371,
+      "learning_rate": 1.2205588676266386e-07,
+      "loss": 0.9148,
+      "step": 692
+    },
+    {
+      "epoch": 1.7027027027027026,
+      "grad_norm": 1.2944478484019626,
+      "learning_rate": 1.2011986985345313e-07,
+      "loss": 0.9421,
+      "step": 693
+    },
+    {
+      "epoch": 1.7051597051597052,
+      "grad_norm": 1.9652193892846588,
+      "learning_rate": 1.1819834869583767e-07,
+      "loss": 0.8827,
+      "step": 694
+    },
+    {
+      "epoch": 1.7076167076167077,
+      "grad_norm": 1.4459180494396398,
+      "learning_rate": 1.1629135494628095e-07,
+      "loss": 0.9373,
+      "step": 695
+    },
+    {
+      "epoch": 1.71007371007371,
+      "grad_norm": 1.4835720268158477,
+      "learning_rate": 1.1439892002191231e-07,
+      "loss": 0.8919,
+      "step": 696
+    },
+    {
+      "epoch": 1.7125307125307125,
+      "grad_norm": 1.8195889649889634,
+      "learning_rate": 1.1252107510000841e-07,
+      "loss": 0.8234,
+      "step": 697
+    },
+    {
+      "epoch": 1.714987714987715,
+      "grad_norm": 1.3885033939390314,
+      "learning_rate": 1.1065785111748117e-07,
+      "loss": 0.8727,
+      "step": 698
+    },
+    {
+      "epoch": 1.7174447174447174,
+      "grad_norm": 1.8847414217621874,
+      "learning_rate": 1.0880927877036606e-07,
+      "loss": 0.9219,
+      "step": 699
+    },
+    {
+      "epoch": 1.71990171990172,
+      "grad_norm": 1.4570728152695678,
+      "learning_rate": 1.0697538851331767e-07,
+      "loss": 0.8921,
+      "step": 700
+    },
+    {
+      "epoch": 1.7223587223587224,
+      "grad_norm": 1.6297111465619418,
+      "learning_rate": 1.0515621055910817e-07,
+      "loss": 0.8523,
+      "step": 701
+    },
+    {
+      "epoch": 1.7248157248157248,
+      "grad_norm": 1.3305859076255666,
+      "learning_rate": 1.0335177487812862e-07,
+      "loss": 0.8873,
+      "step": 702
+    },
+    {
+      "epoch": 1.7272727272727273,
+      "grad_norm": 1.4781364488325777,
+      "learning_rate": 1.0156211119789581e-07,
+      "loss": 0.9512,
+      "step": 703
+    },
+    {
+      "epoch": 1.7297297297297298,
+      "grad_norm": 1.3860356549035335,
+      "learning_rate": 9.978724900256264e-08,
+      "loss": 0.816,
+      "step": 704
+    },
+    {
+      "epoch": 1.7321867321867321,
+      "grad_norm": 1.8635382664585312,
+      "learning_rate": 9.802721753243193e-08,
+      "loss": 0.8981,
+      "step": 705
+    },
+    {
+      "epoch": 1.7346437346437347,
+      "grad_norm": 1.7250194554756548,
+      "learning_rate": 9.628204578347466e-08,
+      "loss": 0.8479,
+      "step": 706
+    },
+    {
+      "epoch": 1.7371007371007372,
+      "grad_norm": 1.697724825854203,
+      "learning_rate": 9.455176250685337e-08,
+      "loss": 0.8986,
+      "step": 707
+    },
+    {
+      "epoch": 1.7395577395577395,
+      "grad_norm": 1.6181977387870101,
+      "learning_rate": 9.283639620844686e-08,
+      "loss": 0.8641,
+      "step": 708
+    },
+    {
+      "epoch": 1.742014742014742,
+      "grad_norm": 1.318237926606282,
+      "learning_rate": 9.113597514838134e-08,
+      "loss": 0.9309,
+      "step": 709
+    },
+    {
+      "epoch": 1.7444717444717446,
+      "grad_norm": 2.254843209174964,
+      "learning_rate": 8.94505273405658e-08,
+      "loss": 0.94,
+      "step": 710
+    },
+    {
+      "epoch": 1.746928746928747,
+      "grad_norm": 1.362243307286928,
+      "learning_rate": 8.778008055222807e-08,
+      "loss": 0.8757,
+      "step": 711
+    },
+    {
+      "epoch": 1.7493857493857494,
+      "grad_norm": 1.6314398152119172,
+      "learning_rate": 8.612466230346016e-08,
+      "loss": 0.8814,
+      "step": 712
+    },
+    {
+      "epoch": 1.751842751842752,
+      "grad_norm": 1.754218537191555,
+      "learning_rate": 8.448429986676298e-08,
+      "loss": 0.9166,
+      "step": 713
+    },
+    {
+      "epoch": 1.7542997542997543,
+      "grad_norm": 1.4911185442079653,
+      "learning_rate": 8.285902026659741e-08,
+      "loss": 0.9213,
+      "step": 714
+    },
+    {
+      "epoch": 1.7567567567567568,
+      "grad_norm": 1.4620109120308156,
+      "learning_rate": 8.124885027894013e-08,
+      "loss": 0.9285,
+      "step": 715
+    },
+    {
+      "epoch": 1.7592137592137593,
+      "grad_norm": 1.5197712931486254,
+      "learning_rate": 7.965381643084068e-08,
+      "loss": 0.8578,
+      "step": 716
+    },
+    {
+      "epoch": 1.7616707616707616,
+      "grad_norm": 1.3554180543190764,
+      "learning_rate": 7.807394499998576e-08,
+      "loss": 0.8945,
+      "step": 717
+    },
+    {
+      "epoch": 1.7641277641277642,
+      "grad_norm": 1.7074271145927862,
+      "learning_rate": 7.650926201426634e-08,
+      "loss": 0.9014,
+      "step": 718
+    },
+    {
+      "epoch": 1.7665847665847667,
+      "grad_norm": 1.7305821717380618,
+      "learning_rate": 7.495979325134805e-08,
+      "loss": 0.9427,
+      "step": 719
+    },
+    {
+      "epoch": 1.769041769041769,
+      "grad_norm": 1.6549796883627756,
+      "learning_rate": 7.342556423824675e-08,
+      "loss": 0.8929,
+      "step": 720
+    },
+    {
+      "epoch": 1.7714987714987716,
+      "grad_norm": 1.693588021942915,
+      "learning_rate": 7.19066002509089e-08,
+      "loss": 0.9149,
+      "step": 721
+    },
+    {
+      "epoch": 1.773955773955774,
+      "grad_norm": 1.3883677751773678,
+      "learning_rate": 7.040292631379386e-08,
+      "loss": 0.9379,
+      "step": 722
+    },
+    {
+      "epoch": 1.7764127764127764,
+      "grad_norm": 1.415233971009828,
+      "learning_rate": 6.891456719946187e-08,
+      "loss": 0.8786,
+      "step": 723
+    },
+    {
+      "epoch": 1.7788697788697787,
+      "grad_norm": 1.3887984611992066,
+      "learning_rate": 6.744154742816721e-08,
+      "loss": 0.8809,
+      "step": 724
+    },
+    {
+      "epoch": 1.7813267813267815,
+      "grad_norm": 1.4599586632095534,
+      "learning_rate": 6.598389126745207e-08,
+      "loss": 0.8906,
+      "step": 725
+    },
+    {
+      "epoch": 1.7837837837837838,
+      "grad_norm": 1.3618867212694985,
+      "learning_rate": 6.454162273174879e-08,
+      "loss": 0.8875,
+      "step": 726
+    },
+    {
+      "epoch": 1.786240786240786,
+      "grad_norm": 1.616703409825558,
+      "learning_rate": 6.311476558198337e-08,
+      "loss": 0.9186,
+      "step": 727
+    },
+    {
+      "epoch": 1.7886977886977888,
+      "grad_norm": 1.5643574404054077,
+      "learning_rate": 6.170334332518323e-08,
+      "loss": 0.8191,
+      "step": 728
+    },
+    {
+      "epoch": 1.7911547911547911,
+      "grad_norm": 1.6547417394218757,
+      "learning_rate": 6.030737921409168e-08,
+      "loss": 0.9015,
+      "step": 729
+    },
+    {
+      "epoch": 1.7936117936117935,
+      "grad_norm": 1.3258614962104087,
+      "learning_rate": 5.892689624678326e-08,
+      "loss": 0.9003,
+      "step": 730
+    },
+    {
+      "epoch": 1.7960687960687962,
+      "grad_norm": 1.4886792779962632,
+      "learning_rate": 5.756191716628556e-08,
+      "loss": 0.9197,
+      "step": 731
+    },
+    {
+      "epoch": 1.7985257985257985,
+      "grad_norm": 1.4228121623738594,
+      "learning_rate": 5.621246446020489e-08,
+      "loss": 0.9151,
+      "step": 732
+    },
+    {
+      "epoch": 1.8009828009828008,
+      "grad_norm": 1.4694783098206843,
+      "learning_rate": 5.487856036035488e-08,
+      "loss": 0.9478,
+      "step": 733
+    },
+    {
+      "epoch": 1.8034398034398036,
+      "grad_norm": 1.6835863076974373,
+      "learning_rate": 5.356022684239059e-08,
+      "loss": 0.8852,
+      "step": 734
+    },
+    {
+      "epoch": 1.805896805896806,
+      "grad_norm": 1.468678481999939,
+      "learning_rate": 5.22574856254474e-08,
+      "loss": 0.9227,
+      "step": 735
+    },
+    {
+      "epoch": 1.8083538083538082,
+      "grad_norm": 1.664873807912385,
+      "learning_rate": 5.0970358171781525e-08,
+      "loss": 0.9709,
+      "step": 736
+    },
+    {
+      "epoch": 1.810810810810811,
+      "grad_norm": 1.4869146515566096,
+      "learning_rate": 4.969886568641757e-08,
+      "loss": 0.8627,
+      "step": 737
+    },
+    {
+      "epoch": 1.8132678132678133,
+      "grad_norm": 1.5302504751076207,
+      "learning_rate": 4.844302911679909e-08,
+      "loss": 0.8694,
+      "step": 738
+    },
+    {
+      "epoch": 1.8157248157248156,
+      "grad_norm": 1.3703447874730004,
+      "learning_rate": 4.720286915244309e-08,
+      "loss": 0.8388,
+      "step": 739
+    },
+    {
+      "epoch": 1.8181818181818183,
+      "grad_norm": 1.8909215620115458,
+      "learning_rate": 4.597840622459936e-08,
+      "loss": 0.8697,
+      "step": 740
+    },
+    {
+      "epoch": 1.8206388206388207,
+      "grad_norm": 1.8396393900066417,
+      "learning_rate": 4.476966050591413e-08,
+      "loss": 0.9238,
+      "step": 741
+    },
+    {
+      "epoch": 1.823095823095823,
+      "grad_norm": 1.4266543724530107,
+      "learning_rate": 4.3576651910097076e-08,
+      "loss": 0.8901,
+      "step": 742
+    },
+    {
+      "epoch": 1.8255528255528255,
+      "grad_norm": 1.3873776301248721,
+      "learning_rate": 4.239940009159415e-08,
+      "loss": 0.904,
+      "step": 743
+    },
+    {
+      "epoch": 1.828009828009828,
+      "grad_norm": 1.554746782909149,
+      "learning_rate": 4.1237924445262864e-08,
+      "loss": 0.8605,
+      "step": 744
+    },
+    {
+      "epoch": 1.8304668304668303,
+      "grad_norm": 1.5101859068421868,
+      "learning_rate": 4.0092244106053386e-08,
+      "loss": 0.8981,
+      "step": 745
+    },
+    {
+      "epoch": 1.8329238329238329,
+      "grad_norm": 1.559988660277713,
+      "learning_rate": 3.896237794869339e-08,
+      "loss": 0.8617,
+      "step": 746
+    },
+    {
+      "epoch": 1.8353808353808354,
+      "grad_norm": 1.7088654168690152,
+      "learning_rate": 3.7848344587376294e-08,
+      "loss": 0.8696,
+      "step": 747
+    },
+    {
+      "epoch": 1.8378378378378377,
+      "grad_norm": 1.6151695281653022,
+      "learning_rate": 3.675016237545536e-08,
+      "loss": 0.9057,
+      "step": 748
+    },
+    {
+      "epoch": 1.8402948402948403,
+      "grad_norm": 1.6260854298806076,
+      "learning_rate": 3.566784940514145e-08,
+      "loss": 0.8715,
+      "step": 749
+    },
+    {
+      "epoch": 1.8427518427518428,
+      "grad_norm": 1.5483874580947579,
+      "learning_rate": 3.460142350720396e-08,
+      "loss": 0.8928,
+      "step": 750
+    },
+    {
+      "epoch": 1.845208845208845,
+      "grad_norm": 1.3856178647115844,
+      "learning_rate": 3.3550902250677914e-08,
+      "loss": 0.8898,
+      "step": 751
+    },
+    {
+      "epoch": 1.8476658476658476,
+      "grad_norm": 1.6217388770422025,
+      "learning_rate": 3.251630294257479e-08,
+      "loss": 0.8887,
+      "step": 752
+    },
+    {
+      "epoch": 1.8501228501228502,
+      "grad_norm": 1.7461700477068736,
+      "learning_rate": 3.1497642627596245e-08,
+      "loss": 0.9144,
+      "step": 753
+    },
+    {
+      "epoch": 1.8525798525798525,
+      "grad_norm": 1.3204466111990083,
+      "learning_rate": 3.0494938087854306e-08,
+      "loss": 0.9046,
+      "step": 754
+    },
+    {
+      "epoch": 1.855036855036855,
+      "grad_norm": 1.5311966806895796,
+      "learning_rate": 2.9508205842594724e-08,
+      "loss": 0.9614,
+      "step": 755
+    },
+    {
+      "epoch": 1.8574938574938575,
+      "grad_norm": 1.5618296845089266,
+      "learning_rate": 2.8537462147924184e-08,
+      "loss": 0.9499,
+      "step": 756
+    },
+    {
+      "epoch": 1.8599508599508598,
+      "grad_norm": 1.506551170833026,
+      "learning_rate": 2.7582722996543405e-08,
+      "loss": 0.8963,
+      "step": 757
+    },
+    {
+      "epoch": 1.8624078624078624,
+      "grad_norm": 1.6976856223987622,
+      "learning_rate": 2.6644004117483355e-08,
+      "loss": 0.938,
+      "step": 758
+    },
+    {
+      "epoch": 1.864864864864865,
+      "grad_norm": 1.7152873001996989,
+      "learning_rate": 2.572132097584523e-08,
+      "loss": 0.9302,
+      "step": 759
+    },
+    {
+      "epoch": 1.8673218673218672,
+      "grad_norm": 1.4747780984909165,
+      "learning_rate": 2.481468877254722e-08,
+      "loss": 0.8927,
+      "step": 760
+    },
+    {
+      "epoch": 1.8697788697788698,
+      "grad_norm": 1.6455964184060006,
+      "learning_rate": 2.3924122444072936e-08,
+      "loss": 0.92,
+      "step": 761
+    },
+    {
+      "epoch": 1.8722358722358723,
+      "grad_norm": 1.4135913992521028,
+      "learning_rate": 2.3049636662225702e-08,
+      "loss": 0.9095,
+      "step": 762
+    },
+    {
+      "epoch": 1.8746928746928746,
+      "grad_norm": 1.3283327210815499,
+      "learning_rate": 2.2191245833886985e-08,
+      "loss": 0.923,
+      "step": 763
+    },
+    {
+      "epoch": 1.8771498771498771,
+      "grad_norm": 2.1206540317643316,
+      "learning_rate": 2.134896410077891e-08,
+      "loss": 0.8759,
+      "step": 764
+    },
+    {
+      "epoch": 1.8796068796068797,
+      "grad_norm": 1.4686301829717907,
+      "learning_rate": 2.052280533923101e-08,
+      "loss": 0.91,
+      "step": 765
+    },
+    {
+      "epoch": 1.882063882063882,
+      "grad_norm": 2.2833412802518835,
+      "learning_rate": 1.9712783159952063e-08,
+      "loss": 0.8681,
+      "step": 766
+    },
+    {
+      "epoch": 1.8845208845208845,
+      "grad_norm": 1.5664189962805568,
+      "learning_rate": 1.8918910907805728e-08,
+      "loss": 0.8958,
+      "step": 767
+    },
+    {
+      "epoch": 1.886977886977887,
+      "grad_norm": 1.3597898472460088,
+      "learning_rate": 1.81412016615905e-08,
+      "loss": 0.8846,
+      "step": 768
+    },
+    {
+      "epoch": 1.8894348894348894,
+      "grad_norm": 1.794391013663922,
+      "learning_rate": 1.737966823382442e-08,
+      "loss": 0.8704,
+      "step": 769
+    },
+    {
+      "epoch": 1.8918918918918919,
+      "grad_norm": 1.3876663428773923,
+      "learning_rate": 1.6634323170533926e-08,
+      "loss": 0.8826,
+      "step": 770
+    },
+    {
+      "epoch": 1.8943488943488944,
+      "grad_norm": 1.964992745747424,
+      "learning_rate": 1.5905178751047134e-08,
+      "loss": 0.9233,
+      "step": 771
+    },
+    {
+      "epoch": 1.8968058968058967,
+      "grad_norm": 1.5474022069405389,
+      "learning_rate": 1.519224698779198e-08,
+      "loss": 0.9574,
+      "step": 772
+    },
+    {
+      "epoch": 1.8992628992628993,
+      "grad_norm": 1.367308795693324,
+      "learning_rate": 1.4495539626097287e-08,
+      "loss": 0.8783,
+      "step": 773
+    },
+    {
+      "epoch": 1.9017199017199018,
+      "grad_norm": 1.4446894160381547,
+      "learning_rate": 1.3815068144000352e-08,
+      "loss": 0.9278,
+      "step": 774
+    },
+    {
+      "epoch": 1.904176904176904,
+      "grad_norm": 1.4909159417858395,
+      "learning_rate": 1.3150843752057438e-08,
+      "loss": 0.8681,
+      "step": 775
+    },
+    {
+      "epoch": 1.9066339066339066,
+      "grad_norm": 1.358589936876708,
+      "learning_rate": 1.2502877393158585e-08,
+      "loss": 0.8662,
+      "step": 776
+    },
+    {
+      "epoch": 1.9090909090909092,
+      "grad_norm": 1.4371079365419295,
+      "learning_rate": 1.1871179742348413e-08,
+      "loss": 0.8844,
+      "step": 777
+    },
+    {
+      "epoch": 1.9115479115479115,
+      "grad_norm": 1.5058059815383584,
+      "learning_rate": 1.1255761206649061e-08,
+      "loss": 0.8885,
+      "step": 778
+    },
+    {
+      "epoch": 1.914004914004914,
+      "grad_norm": 1.4300739849238244,
+      "learning_rate": 1.0656631924889748e-08,
+      "loss": 0.8997,
+      "step": 779
+    },
+    {
+      "epoch": 1.9164619164619165,
+      "grad_norm": 1.5658510974559385,
+      "learning_rate": 1.0073801767539247e-08,
+      "loss": 0.8983,
+      "step": 780
+    },
+    {
+      "epoch": 1.9189189189189189,
+      "grad_norm": 1.4965427543973944,
+      "learning_rate": 9.507280336542911e-09,
+      "loss": 0.8676,
+      "step": 781
+    },
+    {
+      "epoch": 1.9213759213759214,
+      "grad_norm": 1.36581135793563,
+      "learning_rate": 8.957076965165234e-09,
+      "loss": 0.8792,
+      "step": 782
+    },
+    {
+      "epoch": 1.923832923832924,
+      "grad_norm": 1.5156402618603095,
+      "learning_rate": 8.423200717835976e-09,
+      "loss": 0.8626,
+      "step": 783
+    },
+    {
+      "epoch": 1.9262899262899262,
+      "grad_norm": 1.3074563333752685,
+      "learning_rate": 7.905660390000069e-09,
+      "loss": 0.9315,
+      "step": 784
+    },
+    {
+      "epoch": 1.9287469287469288,
+      "grad_norm": 1.7609221141327198,
+      "learning_rate": 7.404464507973607e-09,
+      "loss": 0.8862,
+      "step": 785
+    },
+    {
+      "epoch": 1.9312039312039313,
+      "grad_norm": 1.5426817574716647,
+      "learning_rate": 6.919621328802972e-09,
+      "loss": 0.9543,
+      "step": 786
+    },
+    {
+      "epoch": 1.9336609336609336,
+      "grad_norm": 2.3064282757005663,
+      "learning_rate": 6.4511388401286e-09,
+      "loss": 0.9583,
+      "step": 787
+    },
+    {
+      "epoch": 1.9361179361179361,
+      "grad_norm": 1.67978354162131,
+      "learning_rate": 5.999024760054094e-09,
+      "loss": 0.9364,
+      "step": 788
+    },
+    {
+      "epoch": 1.9385749385749387,
+      "grad_norm": 2.269285024101559,
+      "learning_rate": 5.563286537018319e-09,
+      "loss": 0.858,
+      "step": 789
+    },
+    {
+      "epoch": 1.941031941031941,
+      "grad_norm": 1.7349413650625205,
+      "learning_rate": 5.143931349672837e-09,
+      "loss": 0.9019,
+      "step": 790
+    },
+    {
+      "epoch": 1.9434889434889435,
+      "grad_norm": 1.652477600280537,
+      "learning_rate": 4.7409661067642215e-09,
+      "loss": 0.9325,
+      "step": 791
+    },
+    {
+      "epoch": 1.945945945945946,
+      "grad_norm": 1.354819301610776,
+      "learning_rate": 4.354397447019375e-09,
+      "loss": 0.9047,
+      "step": 792
+    },
+    {
+      "epoch": 1.9484029484029484,
+      "grad_norm": 1.615707852796759,
+      "learning_rate": 3.9842317390369474e-09,
+      "loss": 0.9033,
+      "step": 793
+    },
+    {
+      "epoch": 1.950859950859951,
+      "grad_norm": 1.4784764684056093,
+      "learning_rate": 3.630475081181861e-09,
+      "loss": 0.8419,
+      "step": 794
+    },
+    {
+      "epoch": 1.9533169533169534,
+      "grad_norm": 1.316270670020737,
+      "learning_rate": 3.2931333014850626e-09,
+      "loss": 0.8478,
+      "step": 795
+    },
+    {
+      "epoch": 1.9557739557739557,
+      "grad_norm": 1.4172812525176979,
+      "learning_rate": 2.972211957547377e-09,
+      "loss": 0.8208,
+      "step": 796
+    },
+    {
+      "epoch": 1.9582309582309583,
+      "grad_norm": 1.6364175608068232,
+      "learning_rate": 2.667716336448356e-09,
+      "loss": 0.9062,
+      "step": 797
+    },
+    {
+      "epoch": 1.9606879606879608,
+      "grad_norm": 1.6122614160580477,
+      "learning_rate": 2.379651454658571e-09,
+      "loss": 0.8939,
+      "step": 798
+    },
+    {
+      "epoch": 1.9631449631449631,
+      "grad_norm": 1.5173118240679087,
+      "learning_rate": 2.108022057957348e-09,
+      "loss": 0.8899,
+      "step": 799
+    },
+    {
+      "epoch": 1.9656019656019657,
+      "grad_norm": 1.4032962894445904,
+      "learning_rate": 1.8528326213548273e-09,
+      "loss": 0.9269,
+      "step": 800
+    },
+    {
+      "epoch": 1.9680589680589682,
+      "grad_norm": 1.491232169954221,
+      "learning_rate": 1.6140873490172457e-09,
+      "loss": 0.887,
+      "step": 801
+    },
+    {
+      "epoch": 1.9705159705159705,
+      "grad_norm": 1.452706323624227,
+      "learning_rate": 1.3917901741989923e-09,
+      "loss": 0.9033,
+      "step": 802
+    },
+    {
+      "epoch": 1.972972972972973,
+      "grad_norm": 2.9039670137491176,
+      "learning_rate": 1.1859447591769934e-09,
+      "loss": 0.8912,
+      "step": 803
+    },
+    {
+      "epoch": 1.9754299754299756,
+      "grad_norm": 1.426897627539779,
+      "learning_rate": 9.96554495190205e-10,
+      "loss": 0.8626,
+      "step": 804
+    },
+    {
+      "epoch": 1.9778869778869779,
+      "grad_norm": 1.333240805775527,
+      "learning_rate": 8.236225023844357e-10,
+      "loss": 0.8773,
+      "step": 805
+    },
+    {
+      "epoch": 1.9803439803439802,
+      "grad_norm": 1.3907079742810688,
+      "learning_rate": 6.671516297606094e-10,
+      "loss": 0.9252,
+      "step": 806
+    },
+    {
+      "epoch": 1.982800982800983,
+      "grad_norm": 1.698786893046404,
+      "learning_rate": 5.271444551276927e-10,
+      "loss": 0.9504,
+      "step": 807
+    },
+    {
+      "epoch": 1.9852579852579852,
+      "grad_norm": 1.8549325727891992,
+      "learning_rate": 4.0360328506017226e-10,
+      "loss": 0.8953,
+      "step": 808
+    },
+    {
+      "epoch": 1.9877149877149876,
+      "grad_norm": 1.4404540052198962,
+      "learning_rate": 2.965301548606414e-10,
+      "loss": 0.8946,
+      "step": 809
+    },
+    {
+      "epoch": 1.9901719901719903,
+      "grad_norm": 1.2420916620768656,
+      "learning_rate": 2.0592682852549338e-10,
+      "loss": 0.8543,
+      "step": 810
+    },
+    {
+      "epoch": 1.9926289926289926,
+      "grad_norm": 1.5902830913290869,
+      "learning_rate": 1.317947987163892e-10,
+      "loss": 0.9287,
+      "step": 811
+    },
+    {
+      "epoch": 1.995085995085995,
+      "grad_norm": 1.7356553731894724,
+      "learning_rate": 7.41352867354994e-11,
+      "loss": 0.9318,
+      "step": 812
+    },
+    {
+      "epoch": 1.9975429975429977,
+      "grad_norm": 1.3632994827873626,
+      "learning_rate": 3.294924250540898e-11,
+      "loss": 0.9022,
+      "step": 813
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 1.5450655192747615,
+      "learning_rate": 8.237344553241321e-12,
+      "loss": 0.9407,
+      "step": 814
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 814,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 2,
+  "save_steps": 407,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 8.602630032143155e+17,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}