diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,54665 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.9999359426045737,
+  "eval_steps": 500,
+  "global_step": 7805,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.7584967613220215,
+      "learning_rate": 4.2553191489361707e-08,
+      "loss": 0.7626,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.582927227020264,
+      "learning_rate": 8.510638297872341e-08,
+      "loss": 0.951,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.378379821777344,
+      "learning_rate": 1.276595744680851e-07,
+      "loss": 0.9451,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 9.346311569213867,
+      "learning_rate": 1.7021276595744683e-07,
+      "loss": 1.0475,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.422606945037842,
+      "learning_rate": 2.1276595744680852e-07,
+      "loss": 0.8773,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 13.563347816467285,
+      "learning_rate": 2.553191489361702e-07,
+      "loss": 0.8474,
+      "step": 6
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.2637224197387695,
+      "learning_rate": 2.9787234042553196e-07,
+      "loss": 0.9445,
+      "step": 7
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.488187313079834,
+      "learning_rate": 3.4042553191489365e-07,
+      "loss": 1.0331,
+      "step": 8
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 9.318852424621582,
+      "learning_rate": 3.8297872340425535e-07,
+      "loss": 1.0007,
+      "step": 9
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.960872650146484,
+      "learning_rate": 4.2553191489361704e-07,
+      "loss": 0.9706,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.051608562469482,
+      "learning_rate": 4.6808510638297873e-07,
+      "loss": 1.0274,
+      "step": 11
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.044897556304932,
+      "learning_rate": 5.106382978723404e-07,
+      "loss": 0.9102,
+      "step": 12
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.335330009460449,
+      "learning_rate": 5.531914893617021e-07,
+      "loss": 1.0717,
+      "step": 13
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.157832145690918,
+      "learning_rate": 5.957446808510639e-07,
+      "loss": 0.9263,
+      "step": 14
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.903883934020996,
+      "learning_rate": 6.382978723404255e-07,
+      "loss": 1.1029,
+      "step": 15
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.430100679397583,
+      "learning_rate": 6.808510638297873e-07,
+      "loss": 0.8821,
+      "step": 16
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.4080023765563965,
+      "learning_rate": 7.234042553191489e-07,
+      "loss": 0.9419,
+      "step": 17
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.959429979324341,
+      "learning_rate": 7.659574468085107e-07,
+      "loss": 0.922,
+      "step": 18
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.123316764831543,
+      "learning_rate": 8.085106382978725e-07,
+      "loss": 0.7857,
+      "step": 19
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.357064723968506,
+      "learning_rate": 8.510638297872341e-07,
+      "loss": 0.8643,
+      "step": 20
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.7573935985565186,
+      "learning_rate": 8.936170212765959e-07,
+      "loss": 0.9179,
+      "step": 21
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.969594478607178,
+      "learning_rate": 9.361702127659575e-07,
+      "loss": 0.8299,
+      "step": 22
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.6956169605255127,
+      "learning_rate": 9.787234042553193e-07,
+      "loss": 0.7627,
+      "step": 23
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.103651523590088,
+      "learning_rate": 1.0212765957446809e-06,
+      "loss": 0.8709,
+      "step": 24
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.5962374210357666,
+      "learning_rate": 1.0638297872340427e-06,
+      "loss": 0.6788,
+      "step": 25
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.108245849609375,
+      "learning_rate": 1.1063829787234042e-06,
+      "loss": 0.8235,
+      "step": 26
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.832303762435913,
+      "learning_rate": 1.148936170212766e-06,
+      "loss": 0.8102,
+      "step": 27
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.666902780532837,
+      "learning_rate": 1.1914893617021278e-06,
+      "loss": 0.7508,
+      "step": 28
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.241725206375122,
+      "learning_rate": 1.2340425531914894e-06,
+      "loss": 0.8194,
+      "step": 29
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.413132905960083,
+      "learning_rate": 1.276595744680851e-06,
+      "loss": 0.8131,
+      "step": 30
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.7967190742492676,
+      "learning_rate": 1.3191489361702128e-06,
+      "loss": 0.766,
+      "step": 31
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.8691911697387695,
+      "learning_rate": 1.3617021276595746e-06,
+      "loss": 0.8,
+      "step": 32
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.7689558267593384,
+      "learning_rate": 1.4042553191489364e-06,
+      "loss": 0.8018,
+      "step": 33
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.7399364709854126,
+      "learning_rate": 1.4468085106382978e-06,
+      "loss": 0.5927,
+      "step": 34
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.6319252252578735,
+      "learning_rate": 1.4893617021276596e-06,
+      "loss": 0.6501,
+      "step": 35
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.570933222770691,
+      "learning_rate": 1.5319148936170214e-06,
+      "loss": 0.7304,
+      "step": 36
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.6395325660705566,
+      "learning_rate": 1.5744680851063832e-06,
+      "loss": 0.5934,
+      "step": 37
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.7731611728668213,
+      "learning_rate": 1.617021276595745e-06,
+      "loss": 0.775,
+      "step": 38
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.7520402669906616,
+      "learning_rate": 1.6595744680851064e-06,
+      "loss": 0.6107,
+      "step": 39
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.782818078994751,
+      "learning_rate": 1.7021276595744682e-06,
+      "loss": 0.7435,
+      "step": 40
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6297168731689453,
+      "learning_rate": 1.74468085106383e-06,
+      "loss": 0.738,
+      "step": 41
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5183852910995483,
+      "learning_rate": 1.7872340425531918e-06,
+      "loss": 0.7191,
+      "step": 42
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.116650104522705,
+      "learning_rate": 1.8297872340425531e-06,
+      "loss": 0.7662,
+      "step": 43
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5137921571731567,
+      "learning_rate": 1.872340425531915e-06,
+      "loss": 0.7391,
+      "step": 44
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.1828657388687134,
+      "learning_rate": 1.9148936170212767e-06,
+      "loss": 0.7095,
+      "step": 45
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.526063084602356,
+      "learning_rate": 1.9574468085106385e-06,
+      "loss": 0.7774,
+      "step": 46
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8120944499969482,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 0.6718,
+      "step": 47
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4887276887893677,
+      "learning_rate": 2.0425531914893617e-06,
+      "loss": 0.6486,
+      "step": 48
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4536372423171997,
+      "learning_rate": 2.0851063829787235e-06,
+      "loss": 0.7203,
+      "step": 49
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6725234985351562,
+      "learning_rate": 2.1276595744680853e-06,
+      "loss": 0.6735,
+      "step": 50
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.526537537574768,
+      "learning_rate": 2.170212765957447e-06,
+      "loss": 0.6525,
+      "step": 51
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.7510359287261963,
+      "learning_rate": 2.2127659574468085e-06,
+      "loss": 0.5954,
+      "step": 52
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2007803916931152,
+      "learning_rate": 2.2553191489361703e-06,
+      "loss": 0.7316,
+      "step": 53
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4945361614227295,
+      "learning_rate": 2.297872340425532e-06,
+      "loss": 0.7241,
+      "step": 54
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3967207670211792,
+      "learning_rate": 2.340425531914894e-06,
+      "loss": 0.6345,
+      "step": 55
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3292012214660645,
+      "learning_rate": 2.3829787234042557e-06,
+      "loss": 0.7134,
+      "step": 56
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5092567205429077,
+      "learning_rate": 2.425531914893617e-06,
+      "loss": 0.6689,
+      "step": 57
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9081007242202759,
+      "learning_rate": 2.468085106382979e-06,
+      "loss": 0.8589,
+      "step": 58
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3839454650878906,
+      "learning_rate": 2.5106382978723402e-06,
+      "loss": 0.6063,
+      "step": 59
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3715382814407349,
+      "learning_rate": 2.553191489361702e-06,
+      "loss": 0.5894,
+      "step": 60
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2574067115783691,
+      "learning_rate": 2.595744680851064e-06,
+      "loss": 0.6662,
+      "step": 61
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4591063261032104,
+      "learning_rate": 2.6382978723404256e-06,
+      "loss": 0.6227,
+      "step": 62
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.287207841873169,
+      "learning_rate": 2.6808510638297874e-06,
+      "loss": 0.639,
+      "step": 63
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7553712129592896,
+      "learning_rate": 2.7234042553191492e-06,
+      "loss": 0.6952,
+      "step": 64
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2830992937088013,
+      "learning_rate": 2.765957446808511e-06,
+      "loss": 0.6323,
+      "step": 65
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0832946300506592,
+      "learning_rate": 2.808510638297873e-06,
+      "loss": 0.6222,
+      "step": 66
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8043638467788696,
+      "learning_rate": 2.8510638297872346e-06,
+      "loss": 0.6623,
+      "step": 67
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.494361162185669,
+      "learning_rate": 2.8936170212765956e-06,
+      "loss": 0.6664,
+      "step": 68
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4487807750701904,
+      "learning_rate": 2.9361702127659574e-06,
+      "loss": 0.6772,
+      "step": 69
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7595090866088867,
+      "learning_rate": 2.978723404255319e-06,
+      "loss": 0.6911,
+      "step": 70
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4395978450775146,
+      "learning_rate": 3.021276595744681e-06,
+      "loss": 0.6153,
+      "step": 71
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4251700639724731,
+      "learning_rate": 3.0638297872340428e-06,
+      "loss": 0.6041,
+      "step": 72
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3661823272705078,
+      "learning_rate": 3.1063829787234046e-06,
+      "loss": 0.6282,
+      "step": 73
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.562219262123108,
+      "learning_rate": 3.1489361702127664e-06,
+      "loss": 0.5959,
+      "step": 74
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0947787761688232,
+      "learning_rate": 3.191489361702128e-06,
+      "loss": 0.6361,
+      "step": 75
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0909323692321777,
+      "learning_rate": 3.23404255319149e-06,
+      "loss": 0.659,
+      "step": 76
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1692888736724854,
+      "learning_rate": 3.276595744680851e-06,
+      "loss": 0.6357,
+      "step": 77
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6602727174758911,
+      "learning_rate": 3.3191489361702127e-06,
+      "loss": 0.6931,
+      "step": 78
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3999605178833008,
+      "learning_rate": 3.3617021276595745e-06,
+      "loss": 0.6471,
+      "step": 79
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3343745470046997,
+      "learning_rate": 3.4042553191489363e-06,
+      "loss": 0.6078,
+      "step": 80
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4638773202896118,
+      "learning_rate": 3.446808510638298e-06,
+      "loss": 0.5754,
+      "step": 81
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1183791160583496,
+      "learning_rate": 3.48936170212766e-06,
+      "loss": 0.6925,
+      "step": 82
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.213529348373413,
+      "learning_rate": 3.5319148936170217e-06,
+      "loss": 0.6461,
+      "step": 83
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4197733402252197,
+      "learning_rate": 3.5744680851063835e-06,
+      "loss": 0.5335,
+      "step": 84
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.487101435661316,
+      "learning_rate": 3.6170212765957453e-06,
+      "loss": 0.5241,
+      "step": 85
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4877684116363525,
+      "learning_rate": 3.6595744680851063e-06,
+      "loss": 0.7288,
+      "step": 86
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5158016681671143,
+      "learning_rate": 3.702127659574468e-06,
+      "loss": 0.7598,
+      "step": 87
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8782508373260498,
+      "learning_rate": 3.74468085106383e-06,
+      "loss": 0.5951,
+      "step": 88
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9978234767913818,
+      "learning_rate": 3.7872340425531917e-06,
+      "loss": 0.7127,
+      "step": 89
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.292247772216797,
+      "learning_rate": 3.8297872340425535e-06,
+      "loss": 0.5761,
+      "step": 90
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4210480451583862,
+      "learning_rate": 3.872340425531915e-06,
+      "loss": 0.6722,
+      "step": 91
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5236140489578247,
+      "learning_rate": 3.914893617021277e-06,
+      "loss": 0.6568,
+      "step": 92
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.1957457065582275,
+      "learning_rate": 3.957446808510639e-06,
+      "loss": 0.6213,
+      "step": 93
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5347182750701904,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 0.574,
+      "step": 94
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.1585361957550049,
+      "learning_rate": 4.042553191489362e-06,
+      "loss": 0.5121,
+      "step": 95
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6174169778823853,
+      "learning_rate": 4.085106382978723e-06,
+      "loss": 0.6843,
+      "step": 96
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3933274745941162,
+      "learning_rate": 4.127659574468085e-06,
+      "loss": 0.7019,
+      "step": 97
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.139785051345825,
+      "learning_rate": 4.170212765957447e-06,
+      "loss": 0.6227,
+      "step": 98
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.218185544013977,
+      "learning_rate": 4.212765957446809e-06,
+      "loss": 0.6092,
+      "step": 99
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3104459047317505,
+      "learning_rate": 4.255319148936171e-06,
+      "loss": 0.6566,
+      "step": 100
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3607113361358643,
+      "learning_rate": 4.297872340425532e-06,
+      "loss": 0.6997,
+      "step": 101
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6485787630081177,
+      "learning_rate": 4.340425531914894e-06,
+      "loss": 0.7079,
+      "step": 102
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.3833348751068115,
+      "learning_rate": 4.382978723404256e-06,
+      "loss": 0.5843,
+      "step": 103
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7810088396072388,
+      "learning_rate": 4.425531914893617e-06,
+      "loss": 0.6966,
+      "step": 104
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.511935830116272,
+      "learning_rate": 4.468085106382979e-06,
+      "loss": 0.6428,
+      "step": 105
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4718260765075684,
+      "learning_rate": 4.5106382978723406e-06,
+      "loss": 0.6144,
+      "step": 106
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8542895317077637,
+      "learning_rate": 4.553191489361702e-06,
+      "loss": 0.6564,
+      "step": 107
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.34366774559021,
+      "learning_rate": 4.595744680851064e-06,
+      "loss": 0.5284,
+      "step": 108
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.454941987991333,
+      "learning_rate": 4.638297872340426e-06,
+      "loss": 0.8061,
+      "step": 109
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.738389015197754,
+      "learning_rate": 4.680851063829788e-06,
+      "loss": 0.6759,
+      "step": 110
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6170207262039185,
+      "learning_rate": 4.7234042553191496e-06,
+      "loss": 0.6429,
+      "step": 111
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.628777027130127,
+      "learning_rate": 4.765957446808511e-06,
+      "loss": 0.6829,
+      "step": 112
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.252131700515747,
+      "learning_rate": 4.808510638297872e-06,
+      "loss": 0.5218,
+      "step": 113
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4437956809997559,
+      "learning_rate": 4.851063829787234e-06,
+      "loss": 0.667,
+      "step": 114
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.339471817016602,
+      "learning_rate": 4.893617021276596e-06,
+      "loss": 0.668,
+      "step": 115
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4656023979187012,
+      "learning_rate": 4.936170212765958e-06,
+      "loss": 0.6703,
+      "step": 116
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.32915210723877,
+      "learning_rate": 4.9787234042553195e-06,
+      "loss": 0.6228,
+      "step": 117
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.32205069065094,
+      "learning_rate": 5.0212765957446805e-06,
+      "loss": 0.5615,
+      "step": 118
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2250367403030396,
+      "learning_rate": 5.063829787234042e-06,
+      "loss": 0.6659,
+      "step": 119
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2720946073532104,
+      "learning_rate": 5.106382978723404e-06,
+      "loss": 0.5885,
+      "step": 120
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.233543872833252,
+      "learning_rate": 5.148936170212766e-06,
+      "loss": 0.7094,
+      "step": 121
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4377548694610596,
+      "learning_rate": 5.191489361702128e-06,
+      "loss": 0.7314,
+      "step": 122
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2915871143341064,
+      "learning_rate": 5.2340425531914895e-06,
+      "loss": 0.6415,
+      "step": 123
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.056044101715088,
+      "learning_rate": 5.276595744680851e-06,
+      "loss": 0.6211,
+      "step": 124
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.617048740386963,
+      "learning_rate": 5.319148936170213e-06,
+      "loss": 0.6948,
+      "step": 125
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7886021137237549,
+      "learning_rate": 5.361702127659575e-06,
+      "loss": 0.7082,
+      "step": 126
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2329413890838623,
+      "learning_rate": 5.404255319148937e-06,
+      "loss": 0.6325,
+      "step": 127
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.183271050453186,
+      "learning_rate": 5.4468085106382985e-06,
+      "loss": 0.5629,
+      "step": 128
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.68038809299469,
+      "learning_rate": 5.48936170212766e-06,
+      "loss": 0.6629,
+      "step": 129
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1709634065628052,
+      "learning_rate": 5.531914893617022e-06,
+      "loss": 0.7344,
+      "step": 130
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.245133638381958,
+      "learning_rate": 5.574468085106384e-06,
+      "loss": 0.7115,
+      "step": 131
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9844391345977783,
+      "learning_rate": 5.617021276595746e-06,
+      "loss": 0.6059,
+      "step": 132
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.47454571723938,
+      "learning_rate": 5.6595744680851075e-06,
+      "loss": 0.6277,
+      "step": 133
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5599538087844849,
+      "learning_rate": 5.702127659574469e-06,
+      "loss": 0.7082,
+      "step": 134
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6010850667953491,
+      "learning_rate": 5.744680851063831e-06,
+      "loss": 0.6496,
+      "step": 135
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.419260859489441,
+      "learning_rate": 5.787234042553191e-06,
+      "loss": 0.7087,
+      "step": 136
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7564315795898438,
+      "learning_rate": 5.829787234042553e-06,
+      "loss": 0.6404,
+      "step": 137
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2922577857971191,
+      "learning_rate": 5.872340425531915e-06,
+      "loss": 0.7446,
+      "step": 138
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5558539628982544,
+      "learning_rate": 5.9148936170212766e-06,
+      "loss": 0.6687,
+      "step": 139
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5046948194503784,
+      "learning_rate": 5.957446808510638e-06,
+      "loss": 0.661,
+      "step": 140
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.169022560119629,
+      "learning_rate": 6e-06,
+      "loss": 0.6933,
+      "step": 141
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2964138984680176,
+      "learning_rate": 6.042553191489362e-06,
+      "loss": 0.621,
+      "step": 142
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2966022491455078,
+      "learning_rate": 6.085106382978724e-06,
+      "loss": 0.6333,
+      "step": 143
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.7783942222595215,
+      "learning_rate": 6.1276595744680855e-06,
+      "loss": 0.6224,
+      "step": 144
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.098170280456543,
+      "learning_rate": 6.170212765957447e-06,
+      "loss": 0.6892,
+      "step": 145
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.532152533531189,
+      "learning_rate": 6.212765957446809e-06,
+      "loss": 0.6353,
+      "step": 146
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.6122751235961914,
+      "learning_rate": 6.255319148936171e-06,
+      "loss": 0.603,
+      "step": 147
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3150262832641602,
+      "learning_rate": 6.297872340425533e-06,
+      "loss": 0.68,
+      "step": 148
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8408077955245972,
+      "learning_rate": 6.3404255319148945e-06,
+      "loss": 0.7014,
+      "step": 149
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3194700479507446,
+      "learning_rate": 6.382978723404256e-06,
+      "loss": 0.6772,
+      "step": 150
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.445401668548584,
+      "learning_rate": 6.425531914893618e-06,
+      "loss": 0.5895,
+      "step": 151
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.475083827972412,
+      "learning_rate": 6.46808510638298e-06,
+      "loss": 0.6744,
+      "step": 152
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.218212127685547,
+      "learning_rate": 6.510638297872342e-06,
+      "loss": 0.6075,
+      "step": 153
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3714830875396729,
+      "learning_rate": 6.553191489361702e-06,
+      "loss": 0.6051,
+      "step": 154
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9937951564788818,
+      "learning_rate": 6.595744680851064e-06,
+      "loss": 0.7034,
+      "step": 155
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.379773497581482,
+      "learning_rate": 6.6382978723404254e-06,
+      "loss": 0.6591,
+      "step": 156
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3940249681472778,
+      "learning_rate": 6.680851063829787e-06,
+      "loss": 0.6154,
+      "step": 157
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.862964391708374,
+      "learning_rate": 6.723404255319149e-06,
+      "loss": 0.6576,
+      "step": 158
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3830742835998535,
+      "learning_rate": 6.765957446808511e-06,
+      "loss": 0.6628,
+      "step": 159
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.437272548675537,
+      "learning_rate": 6.808510638297873e-06,
+      "loss": 0.6593,
+      "step": 160
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.473165512084961,
+      "learning_rate": 6.8510638297872344e-06,
+      "loss": 0.7078,
+      "step": 161
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.078347086906433,
+      "learning_rate": 6.893617021276596e-06,
+      "loss": 0.5887,
+      "step": 162
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8319272994995117,
+      "learning_rate": 6.936170212765958e-06,
+      "loss": 0.6499,
+      "step": 163
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3383046388626099,
+      "learning_rate": 6.97872340425532e-06,
+      "loss": 0.7129,
+      "step": 164
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6647669076919556,
+      "learning_rate": 7.021276595744682e-06,
+      "loss": 0.5852,
+      "step": 165
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.294133186340332,
+      "learning_rate": 7.0638297872340434e-06,
+      "loss": 0.6741,
+      "step": 166
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3639767169952393,
+      "learning_rate": 7.106382978723405e-06,
+      "loss": 0.5949,
+      "step": 167
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6028077602386475,
+      "learning_rate": 7.148936170212767e-06,
+      "loss": 0.653,
+      "step": 168
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.474517583847046,
+      "learning_rate": 7.191489361702129e-06,
+      "loss": 0.6168,
+      "step": 169
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2433522939682007,
+      "learning_rate": 7.234042553191491e-06,
+      "loss": 0.7361,
+      "step": 170
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.665745735168457,
+      "learning_rate": 7.2765957446808524e-06,
+      "loss": 0.6894,
+      "step": 171
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8916876316070557,
+      "learning_rate": 7.3191489361702125e-06,
+      "loss": 0.6372,
+      "step": 172
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5415910482406616,
+      "learning_rate": 7.361702127659574e-06,
+      "loss": 0.5954,
+      "step": 173
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2413551807403564,
+      "learning_rate": 7.404255319148936e-06,
+      "loss": 0.69,
+      "step": 174
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9613412618637085,
+      "learning_rate": 7.446808510638298e-06,
+      "loss": 0.6342,
+      "step": 175
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9772624969482422,
+      "learning_rate": 7.48936170212766e-06,
+      "loss": 0.6331,
+      "step": 176
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7678852081298828,
+      "learning_rate": 7.5319148936170215e-06,
+      "loss": 0.629,
+      "step": 177
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1412477493286133,
+      "learning_rate": 7.574468085106383e-06,
+      "loss": 0.622,
+      "step": 178
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3353233337402344,
+      "learning_rate": 7.617021276595745e-06,
+      "loss": 0.6866,
+      "step": 179
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1513804197311401,
+      "learning_rate": 7.659574468085107e-06,
+      "loss": 0.6276,
+      "step": 180
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2835663557052612,
+      "learning_rate": 7.702127659574469e-06,
+      "loss": 0.6185,
+      "step": 181
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3194074630737305,
+      "learning_rate": 7.74468085106383e-06,
+      "loss": 0.6201,
+      "step": 182
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5225788354873657,
+      "learning_rate": 7.787234042553192e-06,
+      "loss": 0.5948,
+      "step": 183
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1869926452636719,
+      "learning_rate": 7.829787234042554e-06,
+      "loss": 0.5794,
+      "step": 184
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3368916511535645,
+      "learning_rate": 7.872340425531916e-06,
+      "loss": 0.7929,
+      "step": 185
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3683775663375854,
+      "learning_rate": 7.914893617021278e-06,
+      "loss": 0.7516,
+      "step": 186
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7117352485656738,
+      "learning_rate": 7.95744680851064e-06,
+      "loss": 0.6344,
+      "step": 187
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4180941581726074,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 0.6158,
+      "step": 188
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.504319667816162,
+      "learning_rate": 8.042553191489363e-06,
+      "loss": 0.79,
+      "step": 189
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4916445016860962,
+      "learning_rate": 8.085106382978723e-06,
+      "loss": 0.7096,
+      "step": 190
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3161081075668335,
+      "learning_rate": 8.127659574468085e-06,
+      "loss": 0.6206,
+      "step": 191
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3922781944274902,
+      "learning_rate": 8.170212765957447e-06,
+      "loss": 0.6443,
+      "step": 192
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5196666717529297,
+      "learning_rate": 8.212765957446809e-06,
+      "loss": 0.6352,
+      "step": 193
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4567962884902954,
+      "learning_rate": 8.25531914893617e-06,
+      "loss": 0.5775,
+      "step": 194
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2705228328704834,
+      "learning_rate": 8.297872340425532e-06,
+      "loss": 0.6438,
+      "step": 195
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4779032468795776,
+      "learning_rate": 8.340425531914894e-06,
+      "loss": 0.6037,
+      "step": 196
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4720426797866821,
+      "learning_rate": 8.382978723404256e-06,
+      "loss": 0.594,
+      "step": 197
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6737343072891235,
+      "learning_rate": 8.425531914893618e-06,
+      "loss": 0.6083,
+      "step": 198
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.425614356994629,
+      "learning_rate": 8.46808510638298e-06,
+      "loss": 0.6219,
+      "step": 199
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4513856172561646,
+      "learning_rate": 8.510638297872341e-06,
+      "loss": 0.6389,
+      "step": 200
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7207236289978027,
+      "learning_rate": 8.553191489361703e-06,
+      "loss": 0.5883,
+      "step": 201
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8808568716049194,
+      "learning_rate": 8.595744680851065e-06,
+      "loss": 0.5775,
+      "step": 202
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1202540397644043,
+      "learning_rate": 8.638297872340427e-06,
+      "loss": 0.5386,
+      "step": 203
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2687338590621948,
+      "learning_rate": 8.680851063829788e-06,
+      "loss": 0.6925,
+      "step": 204
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3964853286743164,
+      "learning_rate": 8.72340425531915e-06,
+      "loss": 0.6858,
+      "step": 205
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5686947107315063,
+      "learning_rate": 8.765957446808512e-06,
+      "loss": 0.6917,
+      "step": 206
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.251679539680481,
+      "learning_rate": 8.808510638297874e-06,
+      "loss": 0.7344,
+      "step": 207
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.4568703174591064,
+      "learning_rate": 8.851063829787234e-06,
+      "loss": 0.6002,
+      "step": 208
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5010696649551392,
+      "learning_rate": 8.893617021276596e-06,
+      "loss": 0.6429,
+      "step": 209
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.468229055404663,
+      "learning_rate": 8.936170212765958e-06,
+      "loss": 0.5552,
+      "step": 210
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.107020378112793,
+      "learning_rate": 8.97872340425532e-06,
+      "loss": 0.5931,
+      "step": 211
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6930768489837646,
+      "learning_rate": 9.021276595744681e-06,
+      "loss": 0.6574,
+      "step": 212
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.752554178237915,
+      "learning_rate": 9.063829787234043e-06,
+      "loss": 0.653,
+      "step": 213
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4542021751403809,
+      "learning_rate": 9.106382978723405e-06,
+      "loss": 0.6302,
+      "step": 214
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3564616441726685,
+      "learning_rate": 9.148936170212767e-06,
+      "loss": 0.6823,
+      "step": 215
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2571592330932617,
+      "learning_rate": 9.191489361702128e-06,
+      "loss": 0.659,
+      "step": 216
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4564090967178345,
+      "learning_rate": 9.23404255319149e-06,
+      "loss": 0.6032,
+      "step": 217
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5122485160827637,
+      "learning_rate": 9.276595744680852e-06,
+      "loss": 0.6931,
+      "step": 218
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4033244848251343,
+      "learning_rate": 9.319148936170214e-06,
+      "loss": 0.6148,
+      "step": 219
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2605109214782715,
+      "learning_rate": 9.361702127659576e-06,
+      "loss": 0.6233,
+      "step": 220
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4508942365646362,
+      "learning_rate": 9.404255319148937e-06,
+      "loss": 0.6114,
+      "step": 221
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.8977622985839844,
+      "learning_rate": 9.446808510638299e-06,
+      "loss": 0.5908,
+      "step": 222
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4348738193511963,
+      "learning_rate": 9.489361702127661e-06,
+      "loss": 0.6485,
+      "step": 223
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2617361545562744,
+      "learning_rate": 9.531914893617023e-06,
+      "loss": 0.6773,
+      "step": 224
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2666085958480835,
+      "learning_rate": 9.574468085106385e-06,
+      "loss": 0.5397,
+      "step": 225
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5835410356521606,
+      "learning_rate": 9.617021276595745e-06,
+      "loss": 0.6741,
+      "step": 226
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.776307225227356,
+      "learning_rate": 9.659574468085106e-06,
+      "loss": 0.6239,
+      "step": 227
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3855375051498413,
+      "learning_rate": 9.702127659574468e-06,
+      "loss": 0.648,
+      "step": 228
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.173299789428711,
+      "learning_rate": 9.74468085106383e-06,
+      "loss": 0.6081,
+      "step": 229
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3490828275680542,
+      "learning_rate": 9.787234042553192e-06,
+      "loss": 0.662,
+      "step": 230
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2873908281326294,
+      "learning_rate": 9.829787234042554e-06,
+      "loss": 0.6217,
+      "step": 231
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5484164953231812,
+      "learning_rate": 9.872340425531915e-06,
+      "loss": 0.6764,
+      "step": 232
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4617042541503906,
+      "learning_rate": 9.914893617021277e-06,
+      "loss": 0.7215,
+      "step": 233
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3531588315963745,
+      "learning_rate": 9.957446808510639e-06,
+      "loss": 0.589,
+      "step": 234
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4462039470672607,
+      "learning_rate": 1e-05,
+      "loss": 0.6209,
+      "step": 235
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5413349866867065,
+      "learning_rate": 9.999999569425815e-06,
+      "loss": 0.6695,
+      "step": 236
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.396835207939148,
+      "learning_rate": 9.999998277703333e-06,
+      "loss": 0.5872,
+      "step": 237
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3538330793380737,
+      "learning_rate": 9.999996124832776e-06,
+      "loss": 0.7435,
+      "step": 238
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7953016757965088,
+      "learning_rate": 9.999993110814515e-06,
+      "loss": 0.6416,
+      "step": 239
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1062463521957397,
+      "learning_rate": 9.999989235649068e-06,
+      "loss": 0.717,
+      "step": 240
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3287006616592407,
+      "learning_rate": 9.999984499337105e-06,
+      "loss": 0.61,
+      "step": 241
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.226269006729126,
+      "learning_rate": 9.99997890187944e-06,
+      "loss": 0.6009,
+      "step": 242
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6038551330566406,
+      "learning_rate": 9.99997244327704e-06,
+      "loss": 0.7504,
+      "step": 243
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3797829151153564,
+      "learning_rate": 9.999965123531012e-06,
+      "loss": 0.5251,
+      "step": 244
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4653568267822266,
+      "learning_rate": 9.999956942642622e-06,
+      "loss": 0.6416,
+      "step": 245
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0714013576507568,
+      "learning_rate": 9.999947900613274e-06,
+      "loss": 0.5643,
+      "step": 246
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.486270785331726,
+      "learning_rate": 9.999937997444528e-06,
+      "loss": 0.6592,
+      "step": 247
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3784781694412231,
+      "learning_rate": 9.999927233138092e-06,
+      "loss": 0.722,
+      "step": 248
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3505719900131226,
+      "learning_rate": 9.999915607695814e-06,
+      "loss": 0.6928,
+      "step": 249
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3439764976501465,
+      "learning_rate": 9.999903121119701e-06,
+      "loss": 0.5936,
+      "step": 250
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3594400882720947,
+      "learning_rate": 9.999889773411903e-06,
+      "loss": 0.6679,
+      "step": 251
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1484427452087402,
+      "learning_rate": 9.999875564574717e-06,
+      "loss": 0.6183,
+      "step": 252
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2887957096099854,
+      "learning_rate": 9.999860494610595e-06,
+      "loss": 0.5997,
+      "step": 253
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2470365762710571,
+      "learning_rate": 9.999844563522123e-06,
+      "loss": 0.6587,
+      "step": 254
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.387832522392273,
+      "learning_rate": 9.999827771312053e-06,
+      "loss": 0.596,
+      "step": 255
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2869324684143066,
+      "learning_rate": 9.999810117983275e-06,
+      "loss": 0.6862,
+      "step": 256
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3541619777679443,
+      "learning_rate": 9.99979160353883e-06,
+      "loss": 0.6792,
+      "step": 257
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6047093868255615,
+      "learning_rate": 9.999772227981905e-06,
+      "loss": 0.6403,
+      "step": 258
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1892220973968506,
+      "learning_rate": 9.999751991315838e-06,
+      "loss": 0.6682,
+      "step": 259
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4554706811904907,
+      "learning_rate": 9.999730893544115e-06,
+      "loss": 0.5767,
+      "step": 260
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3690476417541504,
+      "learning_rate": 9.999708934670366e-06,
+      "loss": 0.7132,
+      "step": 261
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6959372758865356,
+      "learning_rate": 9.99968611469838e-06,
+      "loss": 0.6062,
+      "step": 262
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7080563306808472,
+      "learning_rate": 9.99966243363208e-06,
+      "loss": 0.6362,
+      "step": 263
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2266885042190552,
+      "learning_rate": 9.999637891475549e-06,
+      "loss": 0.6942,
+      "step": 264
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2976837158203125,
+      "learning_rate": 9.99961248823301e-06,
+      "loss": 0.7392,
+      "step": 265
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6553566455841064,
+      "learning_rate": 9.999586223908845e-06,
+      "loss": 0.6517,
+      "step": 266
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.199306607246399,
+      "learning_rate": 9.999559098507571e-06,
+      "loss": 0.6287,
+      "step": 267
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3642069101333618,
+      "learning_rate": 9.999531112033863e-06,
+      "loss": 0.5917,
+      "step": 268
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6378116607666016,
+      "learning_rate": 9.99950226449254e-06,
+      "loss": 0.6069,
+      "step": 269
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.4352381229400635,
+      "learning_rate": 9.99947255588857e-06,
+      "loss": 0.6093,
+      "step": 270
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4313955307006836,
+      "learning_rate": 9.999441986227071e-06,
+      "loss": 0.5984,
+      "step": 271
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6048554182052612,
+      "learning_rate": 9.999410555513308e-06,
+      "loss": 0.6339,
+      "step": 272
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5470483303070068,
+      "learning_rate": 9.999378263752691e-06,
+      "loss": 0.6551,
+      "step": 273
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2593071460723877,
+      "learning_rate": 9.999345110950787e-06,
+      "loss": 0.6564,
+      "step": 274
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4662338495254517,
+      "learning_rate": 9.999311097113303e-06,
+      "loss": 0.503,
+      "step": 275
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6436269283294678,
+      "learning_rate": 9.999276222246094e-06,
+      "loss": 0.7033,
+      "step": 276
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.233489990234375,
+      "learning_rate": 9.999240486355173e-06,
+      "loss": 0.6238,
+      "step": 277
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1347347497940063,
+      "learning_rate": 9.999203889446691e-06,
+      "loss": 0.6618,
+      "step": 278
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7216888666152954,
+      "learning_rate": 9.999166431526952e-06,
+      "loss": 0.5455,
+      "step": 279
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1763736009597778,
+      "learning_rate": 9.999128112602406e-06,
+      "loss": 0.5321,
+      "step": 280
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6581827402114868,
+      "learning_rate": 9.999088932679653e-06,
+      "loss": 0.6222,
+      "step": 281
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3219795227050781,
+      "learning_rate": 9.999048891765443e-06,
+      "loss": 0.601,
+      "step": 282
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4004895687103271,
+      "learning_rate": 9.999007989866671e-06,
+      "loss": 0.6141,
+      "step": 283
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.341827630996704,
+      "learning_rate": 9.99896622699038e-06,
+      "loss": 0.6072,
+      "step": 284
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2262059450149536,
+      "learning_rate": 9.998923603143767e-06,
+      "loss": 0.7096,
+      "step": 285
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5562580823898315,
+      "learning_rate": 9.998880118334167e-06,
+      "loss": 0.6483,
+      "step": 286
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8266093730926514,
+      "learning_rate": 9.998835772569075e-06,
+      "loss": 0.5435,
+      "step": 287
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.291782259941101,
+      "learning_rate": 9.998790565856124e-06,
+      "loss": 0.5632,
+      "step": 288
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3557615280151367,
+      "learning_rate": 9.998744498203104e-06,
+      "loss": 0.7585,
+      "step": 289
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.248962640762329,
+      "learning_rate": 9.998697569617947e-06,
+      "loss": 0.6625,
+      "step": 290
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.591097116470337,
+      "learning_rate": 9.998649780108737e-06,
+      "loss": 0.6722,
+      "step": 291
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3779689073562622,
+      "learning_rate": 9.998601129683703e-06,
+      "loss": 0.6342,
+      "step": 292
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.6783649921417236,
+      "learning_rate": 9.998551618351225e-06,
+      "loss": 0.5614,
+      "step": 293
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.789115071296692,
+      "learning_rate": 9.998501246119828e-06,
+      "loss": 0.6905,
+      "step": 294
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.309683084487915,
+      "learning_rate": 9.998450012998192e-06,
+      "loss": 0.619,
+      "step": 295
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9988428354263306,
+      "learning_rate": 9.998397918995138e-06,
+      "loss": 0.5937,
+      "step": 296
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2818232774734497,
+      "learning_rate": 9.998344964119639e-06,
+      "loss": 0.6503,
+      "step": 297
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3331990242004395,
+      "learning_rate": 9.998291148380813e-06,
+      "loss": 0.7047,
+      "step": 298
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3238475322723389,
+      "learning_rate": 9.998236471787933e-06,
+      "loss": 0.758,
+      "step": 299
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.516418218612671,
+      "learning_rate": 9.998180934350413e-06,
+      "loss": 0.6936,
+      "step": 300
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0435734987258911,
+      "learning_rate": 9.998124536077819e-06,
+      "loss": 0.6652,
+      "step": 301
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.999271035194397,
+      "learning_rate": 9.998067276979863e-06,
+      "loss": 0.639,
+      "step": 302
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2757498025894165,
+      "learning_rate": 9.99800915706641e-06,
+      "loss": 0.5991,
+      "step": 303
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3763235807418823,
+      "learning_rate": 9.997950176347469e-06,
+      "loss": 0.6986,
+      "step": 304
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3771135807037354,
+      "learning_rate": 9.997890334833195e-06,
+      "loss": 0.7234,
+      "step": 305
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.662296175956726,
+      "learning_rate": 9.997829632533897e-06,
+      "loss": 0.7195,
+      "step": 306
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1534143686294556,
+      "learning_rate": 9.99776806946003e-06,
+      "loss": 0.6387,
+      "step": 307
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.804293155670166,
+      "learning_rate": 9.997705645622195e-06,
+      "loss": 0.6533,
+      "step": 308
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2307173013687134,
+      "learning_rate": 9.997642361031147e-06,
+      "loss": 0.6438,
+      "step": 309
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4131581783294678,
+      "learning_rate": 9.997578215697782e-06,
+      "loss": 0.5973,
+      "step": 310
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2833446264266968,
+      "learning_rate": 9.997513209633149e-06,
+      "loss": 0.6023,
+      "step": 311
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.623981237411499,
+      "learning_rate": 9.997447342848443e-06,
+      "loss": 0.604,
+      "step": 312
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3072586059570312,
+      "learning_rate": 9.99738061535501e-06,
+      "loss": 0.6669,
+      "step": 313
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3638581037521362,
+      "learning_rate": 9.997313027164342e-06,
+      "loss": 0.7149,
+      "step": 314
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2896875143051147,
+      "learning_rate": 9.997244578288079e-06,
+      "loss": 0.6452,
+      "step": 315
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1254551410675049,
+      "learning_rate": 9.99717526873801e-06,
+      "loss": 0.6556,
+      "step": 316
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.2857415676116943,
+      "learning_rate": 9.997105098526073e-06,
+      "loss": 0.6838,
+      "step": 317
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.3155784606933594,
+      "learning_rate": 9.997034067664352e-06,
+      "loss": 0.6333,
+      "step": 318
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4330530166625977,
+      "learning_rate": 9.996962176165081e-06,
+      "loss": 0.6069,
+      "step": 319
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.828263282775879,
+      "learning_rate": 9.996889424040644e-06,
+      "loss": 0.6589,
+      "step": 320
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.317088007926941,
+      "learning_rate": 9.996815811303566e-06,
+      "loss": 0.6511,
+      "step": 321
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.6714189052581787,
+      "learning_rate": 9.996741337966531e-06,
+      "loss": 0.6292,
+      "step": 322
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.569022297859192,
+      "learning_rate": 9.996666004042364e-06,
+      "loss": 0.6129,
+      "step": 323
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6436303853988647,
+      "learning_rate": 9.996589809544036e-06,
+      "loss": 0.6159,
+      "step": 324
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2534624338150024,
+      "learning_rate": 9.996512754484675e-06,
+      "loss": 0.6133,
+      "step": 325
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.129281759262085,
+      "learning_rate": 9.996434838877549e-06,
+      "loss": 0.6453,
+      "step": 326
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.673301100730896,
+      "learning_rate": 9.996356062736077e-06,
+      "loss": 0.6298,
+      "step": 327
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3405145406723022,
+      "learning_rate": 9.99627642607383e-06,
+      "loss": 0.6878,
+      "step": 328
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0584392547607422,
+      "learning_rate": 9.996195928904522e-06,
+      "loss": 0.6118,
+      "step": 329
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4414960145950317,
+      "learning_rate": 9.996114571242015e-06,
+      "loss": 0.5883,
+      "step": 330
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4856294393539429,
+      "learning_rate": 9.996032353100324e-06,
+      "loss": 0.5985,
+      "step": 331
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4282467365264893,
+      "learning_rate": 9.995949274493608e-06,
+      "loss": 0.6427,
+      "step": 332
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0834569931030273,
+      "learning_rate": 9.995865335436177e-06,
+      "loss": 0.7503,
+      "step": 333
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4855166673660278,
+      "learning_rate": 9.995780535942485e-06,
+      "loss": 0.6853,
+      "step": 334
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0462682247161865,
+      "learning_rate": 9.99569487602714e-06,
+      "loss": 0.6613,
+      "step": 335
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3587714433670044,
+      "learning_rate": 9.995608355704893e-06,
+      "loss": 0.6956,
+      "step": 336
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.249725103378296,
+      "learning_rate": 9.995520974990646e-06,
+      "loss": 0.6199,
+      "step": 337
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6236398220062256,
+      "learning_rate": 9.99543273389945e-06,
+      "loss": 0.6971,
+      "step": 338
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2508240938186646,
+      "learning_rate": 9.995343632446501e-06,
+      "loss": 0.7062,
+      "step": 339
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3989686965942383,
+      "learning_rate": 9.995253670647146e-06,
+      "loss": 0.7234,
+      "step": 340
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2701517343521118,
+      "learning_rate": 9.995162848516878e-06,
+      "loss": 0.6243,
+      "step": 341
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.829824924468994,
+      "learning_rate": 9.995071166071339e-06,
+      "loss": 0.575,
+      "step": 342
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5379947423934937,
+      "learning_rate": 9.994978623326321e-06,
+      "loss": 0.6044,
+      "step": 343
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1864334344863892,
+      "learning_rate": 9.994885220297763e-06,
+      "loss": 0.6259,
+      "step": 344
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7398015260696411,
+      "learning_rate": 9.994790957001748e-06,
+      "loss": 0.7088,
+      "step": 345
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7878941297531128,
+      "learning_rate": 9.994695833454515e-06,
+      "loss": 0.6905,
+      "step": 346
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.316868543624878,
+      "learning_rate": 9.994599849672446e-06,
+      "loss": 0.6109,
+      "step": 347
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9868359565734863,
+      "learning_rate": 9.994503005672072e-06,
+      "loss": 0.6019,
+      "step": 348
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3597651720046997,
+      "learning_rate": 9.994405301470072e-06,
+      "loss": 0.5773,
+      "step": 349
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5507285594940186,
+      "learning_rate": 9.994306737083275e-06,
+      "loss": 0.6289,
+      "step": 350
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1398341655731201,
+      "learning_rate": 9.994207312528655e-06,
+      "loss": 0.6136,
+      "step": 351
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3331668376922607,
+      "learning_rate": 9.994107027823334e-06,
+      "loss": 0.6668,
+      "step": 352
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4200987815856934,
+      "learning_rate": 9.994005882984588e-06,
+      "loss": 0.6294,
+      "step": 353
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.577136754989624,
+      "learning_rate": 9.993903878029838e-06,
+      "loss": 0.607,
+      "step": 354
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2228367328643799,
+      "learning_rate": 9.993801012976647e-06,
+      "loss": 0.6131,
+      "step": 355
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4739247560501099,
+      "learning_rate": 9.993697287842735e-06,
+      "loss": 0.666,
+      "step": 356
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5883527994155884,
+      "learning_rate": 9.993592702645966e-06,
+      "loss": 0.7499,
+      "step": 357
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7352406978607178,
+      "learning_rate": 9.993487257404352e-06,
+      "loss": 0.6809,
+      "step": 358
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7410889863967896,
+      "learning_rate": 9.993380952136057e-06,
+      "loss": 0.6538,
+      "step": 359
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5513193607330322,
+      "learning_rate": 9.993273786859384e-06,
+      "loss": 0.63,
+      "step": 360
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3114519119262695,
+      "learning_rate": 9.993165761592795e-06,
+      "loss": 0.7388,
+      "step": 361
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.255658745765686,
+      "learning_rate": 9.993056876354892e-06,
+      "loss": 0.6261,
+      "step": 362
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.130070686340332,
+      "learning_rate": 9.992947131164432e-06,
+      "loss": 0.6732,
+      "step": 363
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0119282007217407,
+      "learning_rate": 9.992836526040312e-06,
+      "loss": 0.5841,
+      "step": 364
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1341041326522827,
+      "learning_rate": 9.992725061001585e-06,
+      "loss": 0.6315,
+      "step": 365
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1530150175094604,
+      "learning_rate": 9.992612736067446e-06,
+      "loss": 0.5805,
+      "step": 366
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2198214530944824,
+      "learning_rate": 9.992499551257243e-06,
+      "loss": 0.6518,
+      "step": 367
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5688096284866333,
+      "learning_rate": 9.992385506590467e-06,
+      "loss": 0.6451,
+      "step": 368
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.202008605003357,
+      "learning_rate": 9.992270602086764e-06,
+      "loss": 0.7313,
+      "step": 369
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.660534143447876,
+      "learning_rate": 9.992154837765919e-06,
+      "loss": 0.658,
+      "step": 370
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2800931930541992,
+      "learning_rate": 9.992038213647875e-06,
+      "loss": 0.6472,
+      "step": 371
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4618778228759766,
+      "learning_rate": 9.991920729752713e-06,
+      "loss": 0.6719,
+      "step": 372
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3455302715301514,
+      "learning_rate": 9.991802386100672e-06,
+      "loss": 0.638,
+      "step": 373
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2282071113586426,
+      "learning_rate": 9.99168318271213e-06,
+      "loss": 0.7115,
+      "step": 374
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1178747415542603,
+      "learning_rate": 9.991563119607622e-06,
+      "loss": 0.5796,
+      "step": 375
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1349022388458252,
+      "learning_rate": 9.991442196807823e-06,
+      "loss": 0.5702,
+      "step": 376
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.206775426864624,
+      "learning_rate": 9.991320414333559e-06,
+      "loss": 0.6191,
+      "step": 377
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0808825492858887,
+      "learning_rate": 9.991197772205808e-06,
+      "loss": 0.6128,
+      "step": 378
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1514902114868164,
+      "learning_rate": 9.99107427044569e-06,
+      "loss": 0.7401,
+      "step": 379
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1866756677627563,
+      "learning_rate": 9.990949909074477e-06,
+      "loss": 0.6648,
+      "step": 380
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.040029764175415,
+      "learning_rate": 9.990824688113584e-06,
+      "loss": 0.6372,
+      "step": 381
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2645063400268555,
+      "learning_rate": 9.990698607584584e-06,
+      "loss": 0.6317,
+      "step": 382
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.032130479812622,
+      "learning_rate": 9.990571667509187e-06,
+      "loss": 0.6034,
+      "step": 383
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.203916072845459,
+      "learning_rate": 9.990443867909258e-06,
+      "loss": 0.6927,
+      "step": 384
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2320196628570557,
+      "learning_rate": 9.990315208806807e-06,
+      "loss": 0.6196,
+      "step": 385
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.374750018119812,
+      "learning_rate": 9.990185690223995e-06,
+      "loss": 0.6235,
+      "step": 386
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7351794242858887,
+      "learning_rate": 9.990055312183125e-06,
+      "loss": 0.5726,
+      "step": 387
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2510026693344116,
+      "learning_rate": 9.989924074706654e-06,
+      "loss": 0.7773,
+      "step": 388
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2288150787353516,
+      "learning_rate": 9.989791977817187e-06,
+      "loss": 0.6208,
+      "step": 389
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2345741987228394,
+      "learning_rate": 9.989659021537471e-06,
+      "loss": 0.6605,
+      "step": 390
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.257174015045166,
+      "learning_rate": 9.989525205890407e-06,
+      "loss": 0.6108,
+      "step": 391
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.563079595565796,
+      "learning_rate": 9.989390530899044e-06,
+      "loss": 0.5989,
+      "step": 392
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1823904514312744,
+      "learning_rate": 9.989254996586575e-06,
+      "loss": 0.5835,
+      "step": 393
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3369593620300293,
+      "learning_rate": 9.989118602976343e-06,
+      "loss": 0.6985,
+      "step": 394
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.9907262325286865,
+      "learning_rate": 9.98898135009184e-06,
+      "loss": 0.6928,
+      "step": 395
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.113388538360596,
+      "learning_rate": 9.988843237956703e-06,
+      "loss": 0.5624,
+      "step": 396
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2897573709487915,
+      "learning_rate": 9.98870426659472e-06,
+      "loss": 0.6339,
+      "step": 397
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5871529579162598,
+      "learning_rate": 9.988564436029826e-06,
+      "loss": 0.627,
+      "step": 398
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.316611409187317,
+      "learning_rate": 9.988423746286105e-06,
+      "loss": 0.6233,
+      "step": 399
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2401022911071777,
+      "learning_rate": 9.988282197387787e-06,
+      "loss": 0.7542,
+      "step": 400
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3629307746887207,
+      "learning_rate": 9.988139789359252e-06,
+      "loss": 0.6359,
+      "step": 401
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.324845552444458,
+      "learning_rate": 9.987996522225025e-06,
+      "loss": 0.7213,
+      "step": 402
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.809295177459717,
+      "learning_rate": 9.987852396009783e-06,
+      "loss": 0.634,
+      "step": 403
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5875015258789062,
+      "learning_rate": 9.987707410738346e-06,
+      "loss": 0.6653,
+      "step": 404
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2210239171981812,
+      "learning_rate": 9.987561566435688e-06,
+      "loss": 0.7018,
+      "step": 405
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1161646842956543,
+      "learning_rate": 9.987414863126926e-06,
+      "loss": 0.768,
+      "step": 406
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0861924886703491,
+      "learning_rate": 9.987267300837327e-06,
+      "loss": 0.5714,
+      "step": 407
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3647254705429077,
+      "learning_rate": 9.987118879592303e-06,
+      "loss": 0.7023,
+      "step": 408
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4136096239089966,
+      "learning_rate": 9.986969599417422e-06,
+      "loss": 0.6671,
+      "step": 409
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.715958595275879,
+      "learning_rate": 9.98681946033839e-06,
+      "loss": 0.6588,
+      "step": 410
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.387445330619812,
+      "learning_rate": 9.986668462381065e-06,
+      "loss": 0.7723,
+      "step": 411
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0558326244354248,
+      "learning_rate": 9.98651660557146e-06,
+      "loss": 0.6056,
+      "step": 412
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.917008638381958,
+      "learning_rate": 9.98636388993572e-06,
+      "loss": 0.6206,
+      "step": 413
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2889763116836548,
+      "learning_rate": 9.986210315500154e-06,
+      "loss": 0.7506,
+      "step": 414
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4434642791748047,
+      "learning_rate": 9.986055882291208e-06,
+      "loss": 0.5823,
+      "step": 415
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9407073259353638,
+      "learning_rate": 9.985900590335483e-06,
+      "loss": 0.7348,
+      "step": 416
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.259628415107727,
+      "learning_rate": 9.985744439659724e-06,
+      "loss": 0.5983,
+      "step": 417
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.4348812103271484,
+      "learning_rate": 9.985587430290822e-06,
+      "loss": 0.6106,
+      "step": 418
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2677141427993774,
+      "learning_rate": 9.985429562255822e-06,
+      "loss": 0.6385,
+      "step": 419
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1897544860839844,
+      "learning_rate": 9.985270835581914e-06,
+      "loss": 0.7339,
+      "step": 420
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1162315607070923,
+      "learning_rate": 9.985111250296434e-06,
+      "loss": 0.7,
+      "step": 421
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4913219213485718,
+      "learning_rate": 9.984950806426865e-06,
+      "loss": 0.5671,
+      "step": 422
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8393346071243286,
+      "learning_rate": 9.984789504000844e-06,
+      "loss": 0.7219,
+      "step": 423
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3586616516113281,
+      "learning_rate": 9.98462734304615e-06,
+      "loss": 0.6634,
+      "step": 424
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.41817307472229,
+      "learning_rate": 9.984464323590712e-06,
+      "loss": 0.5488,
+      "step": 425
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3771775960922241,
+      "learning_rate": 9.984300445662608e-06,
+      "loss": 0.64,
+      "step": 426
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3530077934265137,
+      "learning_rate": 9.984135709290061e-06,
+      "loss": 0.5721,
+      "step": 427
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2935279607772827,
+      "learning_rate": 9.983970114501447e-06,
+      "loss": 0.7433,
+      "step": 428
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2527360916137695,
+      "learning_rate": 9.983803661325282e-06,
+      "loss": 0.6697,
+      "step": 429
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7714976072311401,
+      "learning_rate": 9.983636349790235e-06,
+      "loss": 0.626,
+      "step": 430
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2465953826904297,
+      "learning_rate": 9.983468179925124e-06,
+      "loss": 0.6589,
+      "step": 431
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5122458934783936,
+      "learning_rate": 9.983299151758912e-06,
+      "loss": 0.5727,
+      "step": 432
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4289777278900146,
+      "learning_rate": 9.98312926532071e-06,
+      "loss": 0.6082,
+      "step": 433
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2806988954544067,
+      "learning_rate": 9.982958520639778e-06,
+      "loss": 0.6652,
+      "step": 434
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4954502582550049,
+      "learning_rate": 9.982786917745523e-06,
+      "loss": 0.6851,
+      "step": 435
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.040862560272217,
+      "learning_rate": 9.982614456667502e-06,
+      "loss": 0.6674,
+      "step": 436
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2818940877914429,
+      "learning_rate": 9.982441137435414e-06,
+      "loss": 0.6397,
+      "step": 437
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2709026336669922,
+      "learning_rate": 9.982266960079113e-06,
+      "loss": 0.6375,
+      "step": 438
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2646985054016113,
+      "learning_rate": 9.982091924628596e-06,
+      "loss": 0.7385,
+      "step": 439
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0162150859832764,
+      "learning_rate": 9.981916031114011e-06,
+      "loss": 0.5205,
+      "step": 440
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4943784475326538,
+      "learning_rate": 9.98173927956565e-06,
+      "loss": 0.6367,
+      "step": 441
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2825793027877808,
+      "learning_rate": 9.981561670013955e-06,
+      "loss": 0.6305,
+      "step": 442
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1641693115234375,
+      "learning_rate": 9.981383202489517e-06,
+      "loss": 0.6917,
+      "step": 443
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4706101417541504,
+      "learning_rate": 9.981203877023074e-06,
+      "loss": 0.6046,
+      "step": 444
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.171685814857483,
+      "learning_rate": 9.981023693645509e-06,
+      "loss": 0.5857,
+      "step": 445
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3791475296020508,
+      "learning_rate": 9.980842652387855e-06,
+      "loss": 0.542,
+      "step": 446
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1576347351074219,
+      "learning_rate": 9.980660753281296e-06,
+      "loss": 0.6192,
+      "step": 447
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4999419450759888,
+      "learning_rate": 9.980477996357154e-06,
+      "loss": 0.6426,
+      "step": 448
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2739951610565186,
+      "learning_rate": 9.980294381646912e-06,
+      "loss": 0.6035,
+      "step": 449
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.450174331665039,
+      "learning_rate": 9.98010990918219e-06,
+      "loss": 0.667,
+      "step": 450
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3379954099655151,
+      "learning_rate": 9.979924578994761e-06,
+      "loss": 0.6993,
+      "step": 451
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5418367385864258,
+      "learning_rate": 9.979738391116543e-06,
+      "loss": 0.5721,
+      "step": 452
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.351285696029663,
+      "learning_rate": 9.979551345579606e-06,
+      "loss": 0.6084,
+      "step": 453
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3503434658050537,
+      "learning_rate": 9.97936344241616e-06,
+      "loss": 0.7051,
+      "step": 454
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6836271286010742,
+      "learning_rate": 9.979174681658575e-06,
+      "loss": 0.6324,
+      "step": 455
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.482871174812317,
+      "learning_rate": 9.978985063339353e-06,
+      "loss": 0.634,
+      "step": 456
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6858736276626587,
+      "learning_rate": 9.978794587491156e-06,
+      "loss": 0.6554,
+      "step": 457
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5458368062973022,
+      "learning_rate": 9.97860325414679e-06,
+      "loss": 0.599,
+      "step": 458
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.42569899559021,
+      "learning_rate": 9.978411063339205e-06,
+      "loss": 0.6472,
+      "step": 459
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1632460355758667,
+      "learning_rate": 9.978218015101508e-06,
+      "loss": 0.651,
+      "step": 460
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2948800325393677,
+      "learning_rate": 9.978024109466942e-06,
+      "loss": 0.5811,
+      "step": 461
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.991281509399414,
+      "learning_rate": 9.977829346468906e-06,
+      "loss": 0.6956,
+      "step": 462
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.260887622833252,
+      "learning_rate": 9.977633726140942e-06,
+      "loss": 0.6101,
+      "step": 463
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1584416627883911,
+      "learning_rate": 9.977437248516744e-06,
+      "loss": 0.5652,
+      "step": 464
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.210117220878601,
+      "learning_rate": 9.977239913630149e-06,
+      "loss": 0.7002,
+      "step": 465
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4283682107925415,
+      "learning_rate": 9.977041721515146e-06,
+      "loss": 0.6748,
+      "step": 466
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.113726258277893,
+      "learning_rate": 9.976842672205868e-06,
+      "loss": 0.6123,
+      "step": 467
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1056057214736938,
+      "learning_rate": 9.976642765736597e-06,
+      "loss": 0.6603,
+      "step": 468
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.267026424407959,
+      "learning_rate": 9.976442002141767e-06,
+      "loss": 0.6771,
+      "step": 469
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5912866592407227,
+      "learning_rate": 9.976240381455949e-06,
+      "loss": 0.6103,
+      "step": 470
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5797346830368042,
+      "learning_rate": 9.976037903713872e-06,
+      "loss": 0.6114,
+      "step": 471
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.223134160041809,
+      "learning_rate": 9.975834568950406e-06,
+      "loss": 0.6158,
+      "step": 472
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.211577296257019,
+      "learning_rate": 9.975630377200575e-06,
+      "loss": 0.6709,
+      "step": 473
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0716959238052368,
+      "learning_rate": 9.975425328499546e-06,
+      "loss": 0.6498,
+      "step": 474
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.189151644706726,
+      "learning_rate": 9.975219422882631e-06,
+      "loss": 0.6125,
+      "step": 475
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2349804639816284,
+      "learning_rate": 9.975012660385298e-06,
+      "loss": 0.6524,
+      "step": 476
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7489399909973145,
+      "learning_rate": 9.974805041043154e-06,
+      "loss": 0.7575,
+      "step": 477
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7448736429214478,
+      "learning_rate": 9.974596564891958e-06,
+      "loss": 0.6242,
+      "step": 478
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4200482368469238,
+      "learning_rate": 9.974387231967618e-06,
+      "loss": 0.7357,
+      "step": 479
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3394062519073486,
+      "learning_rate": 9.974177042306184e-06,
+      "loss": 0.7014,
+      "step": 480
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8164784908294678,
+      "learning_rate": 9.973965995943857e-06,
+      "loss": 0.745,
+      "step": 481
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.347047209739685,
+      "learning_rate": 9.973754092916989e-06,
+      "loss": 0.7665,
+      "step": 482
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2569183111190796,
+      "learning_rate": 9.973541333262073e-06,
+      "loss": 0.604,
+      "step": 483
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4116610288619995,
+      "learning_rate": 9.973327717015753e-06,
+      "loss": 0.6642,
+      "step": 484
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4871090650558472,
+      "learning_rate": 9.97311324421482e-06,
+      "loss": 0.6648,
+      "step": 485
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.491804599761963,
+      "learning_rate": 9.972897914896215e-06,
+      "loss": 0.7223,
+      "step": 486
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3295527696609497,
+      "learning_rate": 9.97268172909702e-06,
+      "loss": 0.6121,
+      "step": 487
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0643301010131836,
+      "learning_rate": 9.972464686854473e-06,
+      "loss": 0.6677,
+      "step": 488
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3908393383026123,
+      "learning_rate": 9.97224678820595e-06,
+      "loss": 0.616,
+      "step": 489
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.06988525390625,
+      "learning_rate": 9.972028033188984e-06,
+      "loss": 0.6551,
+      "step": 490
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0374271869659424,
+      "learning_rate": 9.97180842184125e-06,
+      "loss": 0.7074,
+      "step": 491
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0564719438552856,
+      "learning_rate": 9.97158795420057e-06,
+      "loss": 0.6798,
+      "step": 492
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.26225745677948,
+      "learning_rate": 9.971366630304917e-06,
+      "loss": 0.6457,
+      "step": 493
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0807775259017944,
+      "learning_rate": 9.971144450192408e-06,
+      "loss": 0.588,
+      "step": 494
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1355947256088257,
+      "learning_rate": 9.97092141390131e-06,
+      "loss": 0.7856,
+      "step": 495
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3302438259124756,
+      "learning_rate": 9.970697521470036e-06,
+      "loss": 0.614,
+      "step": 496
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4943525791168213,
+      "learning_rate": 9.970472772937148e-06,
+      "loss": 0.6594,
+      "step": 497
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7916415929794312,
+      "learning_rate": 9.970247168341352e-06,
+      "loss": 0.68,
+      "step": 498
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2984795570373535,
+      "learning_rate": 9.970020707721505e-06,
+      "loss": 0.676,
+      "step": 499
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.271872878074646,
+      "learning_rate": 9.969793391116613e-06,
+      "loss": 0.7291,
+      "step": 500
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3915470838546753,
+      "learning_rate": 9.969565218565823e-06,
+      "loss": 0.6427,
+      "step": 501
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4289871454238892,
+      "learning_rate": 9.969336190108434e-06,
+      "loss": 0.683,
+      "step": 502
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4833756685256958,
+      "learning_rate": 9.969106305783894e-06,
+      "loss": 0.6101,
+      "step": 503
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1719868183135986,
+      "learning_rate": 9.96887556563179e-06,
+      "loss": 0.611,
+      "step": 504
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2994431257247925,
+      "learning_rate": 9.968643969691869e-06,
+      "loss": 0.6425,
+      "step": 505
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4861727952957153,
+      "learning_rate": 9.968411518004013e-06,
+      "loss": 0.6118,
+      "step": 506
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2328635454177856,
+      "learning_rate": 9.968178210608261e-06,
+      "loss": 0.5606,
+      "step": 507
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6703741550445557,
+      "learning_rate": 9.967944047544795e-06,
+      "loss": 0.6405,
+      "step": 508
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5043929815292358,
+      "learning_rate": 9.967709028853943e-06,
+      "loss": 0.6459,
+      "step": 509
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.377638339996338,
+      "learning_rate": 9.967473154576185e-06,
+      "loss": 0.6761,
+      "step": 510
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2160335779190063,
+      "learning_rate": 9.96723642475214e-06,
+      "loss": 0.627,
+      "step": 511
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4559895992279053,
+      "learning_rate": 9.966998839422587e-06,
+      "loss": 0.6755,
+      "step": 512
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4201894998550415,
+      "learning_rate": 9.966760398628441e-06,
+      "loss": 0.6708,
+      "step": 513
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3131368160247803,
+      "learning_rate": 9.966521102410769e-06,
+      "loss": 0.6067,
+      "step": 514
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.503257393836975,
+      "learning_rate": 9.966280950810787e-06,
+      "loss": 0.6305,
+      "step": 515
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0093584060668945,
+      "learning_rate": 9.966039943869853e-06,
+      "loss": 0.5931,
+      "step": 516
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2581803798675537,
+      "learning_rate": 9.965798081629478e-06,
+      "loss": 0.5784,
+      "step": 517
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0922319889068604,
+      "learning_rate": 9.965555364131316e-06,
+      "loss": 0.6432,
+      "step": 518
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2308220863342285,
+      "learning_rate": 9.965311791417173e-06,
+      "loss": 0.6207,
+      "step": 519
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.4796323776245117,
+      "learning_rate": 9.965067363528996e-06,
+      "loss": 0.697,
+      "step": 520
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.5409436225891113,
+      "learning_rate": 9.964822080508884e-06,
+      "loss": 0.5826,
+      "step": 521
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.550527811050415,
+      "learning_rate": 9.964575942399085e-06,
+      "loss": 0.6782,
+      "step": 522
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4845041036605835,
+      "learning_rate": 9.964328949241987e-06,
+      "loss": 0.6793,
+      "step": 523
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7661019563674927,
+      "learning_rate": 9.964081101080133e-06,
+      "loss": 0.6368,
+      "step": 524
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.430008053779602,
+      "learning_rate": 9.963832397956206e-06,
+      "loss": 0.6865,
+      "step": 525
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7027018070220947,
+      "learning_rate": 9.963582839913042e-06,
+      "loss": 0.6343,
+      "step": 526
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6025042533874512,
+      "learning_rate": 9.963332426993623e-06,
+      "loss": 0.6292,
+      "step": 527
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.252992033958435,
+      "learning_rate": 9.963081159241077e-06,
+      "loss": 0.682,
+      "step": 528
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.218106746673584,
+      "learning_rate": 9.96282903669868e-06,
+      "loss": 0.5634,
+      "step": 529
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6767704486846924,
+      "learning_rate": 9.962576059409854e-06,
+      "loss": 0.6234,
+      "step": 530
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.111166000366211,
+      "learning_rate": 9.96232222741817e-06,
+      "loss": 0.6472,
+      "step": 531
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2849923372268677,
+      "learning_rate": 9.962067540767345e-06,
+      "loss": 0.5941,
+      "step": 532
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.2195000648498535,
+      "learning_rate": 9.961811999501245e-06,
+      "loss": 0.5667,
+      "step": 533
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.106995701789856,
+      "learning_rate": 9.96155560366388e-06,
+      "loss": 0.6523,
+      "step": 534
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.545809268951416,
+      "learning_rate": 9.96129835329941e-06,
+      "loss": 0.5382,
+      "step": 535
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.297607183456421,
+      "learning_rate": 9.96104024845214e-06,
+      "loss": 0.545,
+      "step": 536
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8389314413070679,
+      "learning_rate": 9.960781289166524e-06,
+      "loss": 0.7501,
+      "step": 537
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.588292121887207,
+      "learning_rate": 9.960521475487164e-06,
+      "loss": 0.5668,
+      "step": 538
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2533057928085327,
+      "learning_rate": 9.960260807458806e-06,
+      "loss": 0.6534,
+      "step": 539
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1930246353149414,
+      "learning_rate": 9.959999285126344e-06,
+      "loss": 0.615,
+      "step": 540
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2591867446899414,
+      "learning_rate": 9.959736908534821e-06,
+      "loss": 0.676,
+      "step": 541
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2490293979644775,
+      "learning_rate": 9.959473677729426e-06,
+      "loss": 0.7228,
+      "step": 542
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9905379414558411,
+      "learning_rate": 9.959209592755495e-06,
+      "loss": 0.6434,
+      "step": 543
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.484781265258789,
+      "learning_rate": 9.958944653658513e-06,
+      "loss": 0.7403,
+      "step": 544
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3250967264175415,
+      "learning_rate": 9.958678860484106e-06,
+      "loss": 0.7106,
+      "step": 545
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2482023239135742,
+      "learning_rate": 9.958412213278057e-06,
+      "loss": 0.6753,
+      "step": 546
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0016635656356812,
+      "learning_rate": 9.958144712086286e-06,
+      "loss": 0.6783,
+      "step": 547
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.143125057220459,
+      "learning_rate": 9.957876356954867e-06,
+      "loss": 0.7045,
+      "step": 548
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.036476969718933,
+      "learning_rate": 9.957607147930017e-06,
+      "loss": 0.6173,
+      "step": 549
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2345410585403442,
+      "learning_rate": 9.957337085058103e-06,
+      "loss": 0.6915,
+      "step": 550
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4582602977752686,
+      "learning_rate": 9.957066168385639e-06,
+      "loss": 0.6074,
+      "step": 551
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2883025407791138,
+      "learning_rate": 9.956794397959283e-06,
+      "loss": 0.6289,
+      "step": 552
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3238660097122192,
+      "learning_rate": 9.956521773825842e-06,
+      "loss": 0.6089,
+      "step": 553
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.081383228302002,
+      "learning_rate": 9.956248296032271e-06,
+      "loss": 0.6748,
+      "step": 554
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2733744382858276,
+      "learning_rate": 9.955973964625671e-06,
+      "loss": 0.6841,
+      "step": 555
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1720517873764038,
+      "learning_rate": 9.95569877965329e-06,
+      "loss": 0.6472,
+      "step": 556
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.179827094078064,
+      "learning_rate": 9.955422741162521e-06,
+      "loss": 0.6,
+      "step": 557
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3065040111541748,
+      "learning_rate": 9.95514584920091e-06,
+      "loss": 0.6316,
+      "step": 558
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4125103950500488,
+      "learning_rate": 9.95486810381614e-06,
+      "loss": 0.6008,
+      "step": 559
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5217939615249634,
+      "learning_rate": 9.954589505056054e-06,
+      "loss": 0.704,
+      "step": 560
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4721038341522217,
+      "learning_rate": 9.95431005296863e-06,
+      "loss": 0.6193,
+      "step": 561
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0634781122207642,
+      "learning_rate": 9.954029747601999e-06,
+      "loss": 0.5784,
+      "step": 562
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4376946687698364,
+      "learning_rate": 9.95374858900444e-06,
+      "loss": 0.6244,
+      "step": 563
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3898831605911255,
+      "learning_rate": 9.953466577224374e-06,
+      "loss": 0.6728,
+      "step": 564
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.651649832725525,
+      "learning_rate": 9.953183712310373e-06,
+      "loss": 0.5602,
+      "step": 565
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1175930500030518,
+      "learning_rate": 9.952899994311153e-06,
+      "loss": 0.6814,
+      "step": 566
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2130919694900513,
+      "learning_rate": 9.952615423275584e-06,
+      "loss": 0.6513,
+      "step": 567
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3236795663833618,
+      "learning_rate": 9.952329999252673e-06,
+      "loss": 0.602,
+      "step": 568
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.570998191833496,
+      "learning_rate": 9.952043722291577e-06,
+      "loss": 0.7349,
+      "step": 569
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0845404863357544,
+      "learning_rate": 9.951756592441606e-06,
+      "loss": 0.6471,
+      "step": 570
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2199392318725586,
+      "learning_rate": 9.95146860975221e-06,
+      "loss": 0.6956,
+      "step": 571
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1437824964523315,
+      "learning_rate": 9.951179774272988e-06,
+      "loss": 0.5707,
+      "step": 572
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.277314305305481,
+      "learning_rate": 9.950890086053686e-06,
+      "loss": 0.6485,
+      "step": 573
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1325360536575317,
+      "learning_rate": 9.9505995451442e-06,
+      "loss": 0.574,
+      "step": 574
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1564849615097046,
+      "learning_rate": 9.950308151594563e-06,
+      "loss": 0.595,
+      "step": 575
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2639933824539185,
+      "learning_rate": 9.950015905454969e-06,
+      "loss": 0.712,
+      "step": 576
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.110800266265869,
+      "learning_rate": 9.949722806775746e-06,
+      "loss": 0.6254,
+      "step": 577
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1644377708435059,
+      "learning_rate": 9.949428855607377e-06,
+      "loss": 0.6832,
+      "step": 578
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.162166953086853,
+      "learning_rate": 9.949134052000488e-06,
+      "loss": 0.742,
+      "step": 579
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1934207677841187,
+      "learning_rate": 9.948838396005854e-06,
+      "loss": 0.7117,
+      "step": 580
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3645777702331543,
+      "learning_rate": 9.948541887674396e-06,
+      "loss": 0.6821,
+      "step": 581
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4982502460479736,
+      "learning_rate": 9.94824452705718e-06,
+      "loss": 0.654,
+      "step": 582
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.82792329788208,
+      "learning_rate": 9.947946314205421e-06,
+      "loss": 0.6821,
+      "step": 583
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4962670803070068,
+      "learning_rate": 9.94764724917048e-06,
+      "loss": 0.6436,
+      "step": 584
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.179075837135315,
+      "learning_rate": 9.947347332003865e-06,
+      "loss": 0.6445,
+      "step": 585
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9640148878097534,
+      "learning_rate": 9.94704656275723e-06,
+      "loss": 0.6306,
+      "step": 586
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0042561292648315,
+      "learning_rate": 9.946744941482379e-06,
+      "loss": 0.713,
+      "step": 587
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8437522649765015,
+      "learning_rate": 9.946442468231257e-06,
+      "loss": 0.7496,
+      "step": 588
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3848373889923096,
+      "learning_rate": 9.94613914305596e-06,
+      "loss": 0.6736,
+      "step": 589
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5635427236557007,
+      "learning_rate": 9.94583496600873e-06,
+      "loss": 0.6535,
+      "step": 590
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3483617305755615,
+      "learning_rate": 9.945529937141953e-06,
+      "loss": 0.616,
+      "step": 591
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2669429779052734,
+      "learning_rate": 9.945224056508168e-06,
+      "loss": 0.6904,
+      "step": 592
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7648189067840576,
+      "learning_rate": 9.944917324160055e-06,
+      "loss": 0.6421,
+      "step": 593
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4274234771728516,
+      "learning_rate": 9.944609740150441e-06,
+      "loss": 0.6098,
+      "step": 594
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2724729776382446,
+      "learning_rate": 9.944301304532302e-06,
+      "loss": 0.5873,
+      "step": 595
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.306372880935669,
+      "learning_rate": 9.943992017358761e-06,
+      "loss": 0.6709,
+      "step": 596
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3919581174850464,
+      "learning_rate": 9.943681878683085e-06,
+      "loss": 0.6482,
+      "step": 597
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.333129644393921,
+      "learning_rate": 9.94337088855869e-06,
+      "loss": 0.7064,
+      "step": 598
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.256874680519104,
+      "learning_rate": 9.943059047039137e-06,
+      "loss": 0.6856,
+      "step": 599
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1219730377197266,
+      "learning_rate": 9.942746354178133e-06,
+      "loss": 0.6434,
+      "step": 600
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3746181726455688,
+      "learning_rate": 9.942432810029536e-06,
+      "loss": 0.6124,
+      "step": 601
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2980557680130005,
+      "learning_rate": 9.942118414647346e-06,
+      "loss": 0.6688,
+      "step": 602
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4423713684082031,
+      "learning_rate": 9.941803168085711e-06,
+      "loss": 0.6483,
+      "step": 603
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7107197046279907,
+      "learning_rate": 9.941487070398928e-06,
+      "loss": 0.6004,
+      "step": 604
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.115032434463501,
+      "learning_rate": 9.941170121641434e-06,
+      "loss": 0.5836,
+      "step": 605
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.354966402053833,
+      "learning_rate": 9.940852321867821e-06,
+      "loss": 0.6629,
+      "step": 606
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.070405125617981,
+      "learning_rate": 9.940533671132821e-06,
+      "loss": 0.6276,
+      "step": 607
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.133563756942749,
+      "learning_rate": 9.940214169491318e-06,
+      "loss": 0.715,
+      "step": 608
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1300113201141357,
+      "learning_rate": 9.939893816998337e-06,
+      "loss": 0.6603,
+      "step": 609
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.367214322090149,
+      "learning_rate": 9.939572613709052e-06,
+      "loss": 0.6713,
+      "step": 610
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.45294988155365,
+      "learning_rate": 9.939250559678787e-06,
+      "loss": 0.579,
+      "step": 611
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3254419565200806,
+      "learning_rate": 9.938927654963007e-06,
+      "loss": 0.579,
+      "step": 612
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1809929609298706,
+      "learning_rate": 9.938603899617323e-06,
+      "loss": 0.6424,
+      "step": 613
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.836148977279663,
+      "learning_rate": 9.9382792936975e-06,
+      "loss": 0.598,
+      "step": 614
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1036103963851929,
+      "learning_rate": 9.937953837259444e-06,
+      "loss": 0.5946,
+      "step": 615
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2749247550964355,
+      "learning_rate": 9.937627530359206e-06,
+      "loss": 0.6951,
+      "step": 616
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2078628540039062,
+      "learning_rate": 9.937300373052987e-06,
+      "loss": 0.6501,
+      "step": 617
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.239443302154541,
+      "learning_rate": 9.936972365397133e-06,
+      "loss": 0.6449,
+      "step": 618
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7138959169387817,
+      "learning_rate": 9.936643507448138e-06,
+      "loss": 0.714,
+      "step": 619
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0761655569076538,
+      "learning_rate": 9.93631379926264e-06,
+      "loss": 0.6237,
+      "step": 620
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2124683856964111,
+      "learning_rate": 9.935983240897422e-06,
+      "loss": 0.6217,
+      "step": 621
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4442805051803589,
+      "learning_rate": 9.93565183240942e-06,
+      "loss": 0.6459,
+      "step": 622
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2149344682693481,
+      "learning_rate": 9.935319573855711e-06,
+      "loss": 0.6465,
+      "step": 623
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1919761896133423,
+      "learning_rate": 9.934986465293518e-06,
+      "loss": 0.5959,
+      "step": 624
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3023431301116943,
+      "learning_rate": 9.934652506780214e-06,
+      "loss": 0.6582,
+      "step": 625
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.142003059387207,
+      "learning_rate": 9.934317698373317e-06,
+      "loss": 0.6515,
+      "step": 626
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3569769859313965,
+      "learning_rate": 9.93398204013049e-06,
+      "loss": 0.7018,
+      "step": 627
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1718806028366089,
+      "learning_rate": 9.933645532109544e-06,
+      "loss": 0.5533,
+      "step": 628
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1356302499771118,
+      "learning_rate": 9.933308174368434e-06,
+      "loss": 0.6077,
+      "step": 629
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1580795049667358,
+      "learning_rate": 9.932969966965267e-06,
+      "loss": 0.6484,
+      "step": 630
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.378556728363037,
+      "learning_rate": 9.932630909958287e-06,
+      "loss": 0.6518,
+      "step": 631
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0749226808547974,
+      "learning_rate": 9.932291003405893e-06,
+      "loss": 0.6614,
+      "step": 632
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4288339614868164,
+      "learning_rate": 9.931950247366625e-06,
+      "loss": 0.7169,
+      "step": 633
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.120861291885376,
+      "learning_rate": 9.931608641899172e-06,
+      "loss": 0.6148,
+      "step": 634
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.725447654724121,
+      "learning_rate": 9.931266187062372e-06,
+      "loss": 0.5805,
+      "step": 635
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3710678815841675,
+      "learning_rate": 9.930922882915201e-06,
+      "loss": 0.6566,
+      "step": 636
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3103550672531128,
+      "learning_rate": 9.93057872951679e-06,
+      "loss": 0.6861,
+      "step": 637
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3011488914489746,
+      "learning_rate": 9.93023372692641e-06,
+      "loss": 0.6222,
+      "step": 638
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2335342168807983,
+      "learning_rate": 9.92988787520348e-06,
+      "loss": 0.7169,
+      "step": 639
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.844463586807251,
+      "learning_rate": 9.929541174407568e-06,
+      "loss": 0.6041,
+      "step": 640
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6094704866409302,
+      "learning_rate": 9.929193624598386e-06,
+      "loss": 0.6485,
+      "step": 641
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2318650484085083,
+      "learning_rate": 9.928845225835791e-06,
+      "loss": 0.7301,
+      "step": 642
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3937469720840454,
+      "learning_rate": 9.928495978179788e-06,
+      "loss": 0.658,
+      "step": 643
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3361302614212036,
+      "learning_rate": 9.928145881690527e-06,
+      "loss": 0.6271,
+      "step": 644
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.590607762336731,
+      "learning_rate": 9.927794936428308e-06,
+      "loss": 0.6345,
+      "step": 645
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0950250625610352,
+      "learning_rate": 9.927443142453573e-06,
+      "loss": 0.5692,
+      "step": 646
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.009345531463623,
+      "learning_rate": 9.92709049982691e-06,
+      "loss": 0.6775,
+      "step": 647
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5916540622711182,
+      "learning_rate": 9.926737008609055e-06,
+      "loss": 0.6645,
+      "step": 648
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0630630254745483,
+      "learning_rate": 9.92638266886089e-06,
+      "loss": 0.596,
+      "step": 649
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2478364706039429,
+      "learning_rate": 9.926027480643442e-06,
+      "loss": 0.6118,
+      "step": 650
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2530955076217651,
+      "learning_rate": 9.925671444017887e-06,
+      "loss": 0.658,
+      "step": 651
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3152321577072144,
+      "learning_rate": 9.925314559045543e-06,
+      "loss": 0.6898,
+      "step": 652
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2925292253494263,
+      "learning_rate": 9.924956825787877e-06,
+      "loss": 0.6682,
+      "step": 653
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.472712755203247,
+      "learning_rate": 9.924598244306502e-06,
+      "loss": 0.6382,
+      "step": 654
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6111242771148682,
+      "learning_rate": 9.924238814663174e-06,
+      "loss": 0.5262,
+      "step": 655
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.210485816001892,
+      "learning_rate": 9.9238785369198e-06,
+      "loss": 0.6865,
+      "step": 656
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3140826225280762,
+      "learning_rate": 9.92351741113843e-06,
+      "loss": 0.6448,
+      "step": 657
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.125855803489685,
+      "learning_rate": 9.92315543738126e-06,
+      "loss": 0.6277,
+      "step": 658
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9021656513214111,
+      "learning_rate": 9.922792615710632e-06,
+      "loss": 0.6673,
+      "step": 659
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.304652452468872,
+      "learning_rate": 9.922428946189038e-06,
+      "loss": 0.6477,
+      "step": 660
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2016267776489258,
+      "learning_rate": 9.922064428879108e-06,
+      "loss": 0.7264,
+      "step": 661
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3767685890197754,
+      "learning_rate": 9.921699063843624e-06,
+      "loss": 0.6343,
+      "step": 662
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5952433347702026,
+      "learning_rate": 9.921332851145516e-06,
+      "loss": 0.6332,
+      "step": 663
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1251158714294434,
+      "learning_rate": 9.920965790847852e-06,
+      "loss": 0.6287,
+      "step": 664
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7617466449737549,
+      "learning_rate": 9.920597883013854e-06,
+      "loss": 0.6315,
+      "step": 665
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5821009874343872,
+      "learning_rate": 9.920229127706884e-06,
+      "loss": 0.5471,
+      "step": 666
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1487936973571777,
+      "learning_rate": 9.919859524990456e-06,
+      "loss": 0.7097,
+      "step": 667
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3274495601654053,
+      "learning_rate": 9.919489074928223e-06,
+      "loss": 0.5928,
+      "step": 668
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1028430461883545,
+      "learning_rate": 9.91911777758399e-06,
+      "loss": 0.6415,
+      "step": 669
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2603744268417358,
+      "learning_rate": 9.918745633021706e-06,
+      "loss": 0.7122,
+      "step": 670
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1494251489639282,
+      "learning_rate": 9.918372641305463e-06,
+      "loss": 0.6002,
+      "step": 671
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1268144845962524,
+      "learning_rate": 9.917998802499502e-06,
+      "loss": 0.677,
+      "step": 672
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5660556554794312,
+      "learning_rate": 9.91762411666821e-06,
+      "loss": 0.6431,
+      "step": 673
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1017109155654907,
+      "learning_rate": 9.917248583876116e-06,
+      "loss": 0.7173,
+      "step": 674
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.498296856880188,
+      "learning_rate": 9.916872204187902e-06,
+      "loss": 0.6641,
+      "step": 675
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0133637189865112,
+      "learning_rate": 9.91649497766839e-06,
+      "loss": 0.6279,
+      "step": 676
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.355507731437683,
+      "learning_rate": 9.91611690438255e-06,
+      "loss": 0.5455,
+      "step": 677
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.849907636642456,
+      "learning_rate": 9.915737984395498e-06,
+      "loss": 0.7237,
+      "step": 678
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1791698932647705,
+      "learning_rate": 9.915358217772491e-06,
+      "loss": 0.6523,
+      "step": 679
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3907570838928223,
+      "learning_rate": 9.91497760457894e-06,
+      "loss": 0.6599,
+      "step": 680
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5050311088562012,
+      "learning_rate": 9.914596144880399e-06,
+      "loss": 0.6543,
+      "step": 681
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2648143768310547,
+      "learning_rate": 9.914213838742566e-06,
+      "loss": 0.5308,
+      "step": 682
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1902687549591064,
+      "learning_rate": 9.913830686231281e-06,
+      "loss": 0.6256,
+      "step": 683
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0023192167282104,
+      "learning_rate": 9.913446687412541e-06,
+      "loss": 0.6192,
+      "step": 684
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0021028518676758,
+      "learning_rate": 9.913061842352475e-06,
+      "loss": 0.6225,
+      "step": 685
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0987367630004883,
+      "learning_rate": 9.912676151117371e-06,
+      "loss": 0.6478,
+      "step": 686
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1188229322433472,
+      "learning_rate": 9.912289613773653e-06,
+      "loss": 0.5543,
+      "step": 687
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3304802179336548,
+      "learning_rate": 9.911902230387896e-06,
+      "loss": 0.6396,
+      "step": 688
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0771523714065552,
+      "learning_rate": 9.911514001026816e-06,
+      "loss": 0.6503,
+      "step": 689
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0926355123519897,
+      "learning_rate": 9.911124925757282e-06,
+      "loss": 0.6256,
+      "step": 690
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1526925563812256,
+      "learning_rate": 9.9107350046463e-06,
+      "loss": 0.6386,
+      "step": 691
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0600298643112183,
+      "learning_rate": 9.91034423776103e-06,
+      "loss": 0.6494,
+      "step": 692
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.321643352508545,
+      "learning_rate": 9.90995262516877e-06,
+      "loss": 0.5935,
+      "step": 693
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1695914268493652,
+      "learning_rate": 9.909560166936968e-06,
+      "loss": 0.7052,
+      "step": 694
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3552606105804443,
+      "learning_rate": 9.909166863133218e-06,
+      "loss": 0.6771,
+      "step": 695
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.070600152015686,
+      "learning_rate": 9.908772713825259e-06,
+      "loss": 0.6489,
+      "step": 696
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.541309118270874,
+      "learning_rate": 9.908377719080976e-06,
+      "loss": 0.7063,
+      "step": 697
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2539750337600708,
+      "learning_rate": 9.907981878968395e-06,
+      "loss": 0.7322,
+      "step": 698
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3420991897583008,
+      "learning_rate": 9.907585193555696e-06,
+      "loss": 0.5876,
+      "step": 699
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.159241676330566,
+      "learning_rate": 9.907187662911195e-06,
+      "loss": 0.6435,
+      "step": 700
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.334041953086853,
+      "learning_rate": 9.906789287103364e-06,
+      "loss": 0.67,
+      "step": 701
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2002819776535034,
+      "learning_rate": 9.90639006620081e-06,
+      "loss": 0.6775,
+      "step": 702
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6984732151031494,
+      "learning_rate": 9.905990000272295e-06,
+      "loss": 0.6121,
+      "step": 703
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.382564663887024,
+      "learning_rate": 9.90558908938672e-06,
+      "loss": 0.6936,
+      "step": 704
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2500004768371582,
+      "learning_rate": 9.905187333613134e-06,
+      "loss": 0.6399,
+      "step": 705
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.314711332321167,
+      "learning_rate": 9.904784733020732e-06,
+      "loss": 0.639,
+      "step": 706
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5994131565093994,
+      "learning_rate": 9.904381287678853e-06,
+      "loss": 0.5754,
+      "step": 707
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1765432357788086,
+      "learning_rate": 9.903976997656982e-06,
+      "loss": 0.5971,
+      "step": 708
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9509575366973877,
+      "learning_rate": 9.903571863024752e-06,
+      "loss": 0.6634,
+      "step": 709
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1447938680648804,
+      "learning_rate": 9.903165883851936e-06,
+      "loss": 0.6595,
+      "step": 710
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.166501760482788,
+      "learning_rate": 9.902759060208456e-06,
+      "loss": 0.6014,
+      "step": 711
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1991925239562988,
+      "learning_rate": 9.90235139216438e-06,
+      "loss": 0.6136,
+      "step": 712
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2937411069869995,
+      "learning_rate": 9.901942879789923e-06,
+      "loss": 0.5599,
+      "step": 713
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0895254611968994,
+      "learning_rate": 9.90153352315544e-06,
+      "loss": 0.6231,
+      "step": 714
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2510082721710205,
+      "learning_rate": 9.901123322331434e-06,
+      "loss": 0.6087,
+      "step": 715
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9192419648170471,
+      "learning_rate": 9.900712277388556e-06,
+      "loss": 0.5165,
+      "step": 716
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6960457563400269,
+      "learning_rate": 9.900300388397597e-06,
+      "loss": 0.7067,
+      "step": 717
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0721408128738403,
+      "learning_rate": 9.899887655429501e-06,
+      "loss": 0.5991,
+      "step": 718
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2546442747116089,
+      "learning_rate": 9.89947407855535e-06,
+      "loss": 0.5862,
+      "step": 719
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3291999101638794,
+      "learning_rate": 9.899059657846373e-06,
+      "loss": 0.6478,
+      "step": 720
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.157029628753662,
+      "learning_rate": 9.898644393373946e-06,
+      "loss": 0.6432,
+      "step": 721
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4094780683517456,
+      "learning_rate": 9.898228285209593e-06,
+      "loss": 0.582,
+      "step": 722
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2219111919403076,
+      "learning_rate": 9.897811333424977e-06,
+      "loss": 0.5938,
+      "step": 723
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2780520915985107,
+      "learning_rate": 9.897393538091912e-06,
+      "loss": 0.5572,
+      "step": 724
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0696425437927246,
+      "learning_rate": 9.896974899282352e-06,
+      "loss": 0.6494,
+      "step": 725
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4121901988983154,
+      "learning_rate": 9.896555417068401e-06,
+      "loss": 0.6419,
+      "step": 726
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3450160026550293,
+      "learning_rate": 9.896135091522305e-06,
+      "loss": 0.7238,
+      "step": 727
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0523320436477661,
+      "learning_rate": 9.895713922716457e-06,
+      "loss": 0.5947,
+      "step": 728
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5163980722427368,
+      "learning_rate": 9.895291910723396e-06,
+      "loss": 0.5785,
+      "step": 729
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4099388122558594,
+      "learning_rate": 9.894869055615803e-06,
+      "loss": 0.5904,
+      "step": 730
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9844495058059692,
+      "learning_rate": 9.894445357466507e-06,
+      "loss": 0.6382,
+      "step": 731
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2758938074111938,
+      "learning_rate": 9.894020816348483e-06,
+      "loss": 0.6405,
+      "step": 732
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.246328353881836,
+      "learning_rate": 9.893595432334846e-06,
+      "loss": 0.6006,
+      "step": 733
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2072193622589111,
+      "learning_rate": 9.893169205498864e-06,
+      "loss": 0.6452,
+      "step": 734
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2706694602966309,
+      "learning_rate": 9.892742135913942e-06,
+      "loss": 0.619,
+      "step": 735
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2876821756362915,
+      "learning_rate": 9.892314223653636e-06,
+      "loss": 0.6796,
+      "step": 736
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3346693515777588,
+      "learning_rate": 9.891885468791645e-06,
+      "loss": 0.6494,
+      "step": 737
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2320876121520996,
+      "learning_rate": 9.891455871401814e-06,
+      "loss": 0.6095,
+      "step": 738
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2823511362075806,
+      "learning_rate": 9.89102543155813e-06,
+      "loss": 0.7429,
+      "step": 739
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3542308807373047,
+      "learning_rate": 9.890594149334731e-06,
+      "loss": 0.6251,
+      "step": 740
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.2784230709075928,
+      "learning_rate": 9.890162024805895e-06,
+      "loss": 0.693,
+      "step": 741
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.167242169380188,
+      "learning_rate": 9.889729058046045e-06,
+      "loss": 0.6828,
+      "step": 742
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.067370891571045,
+      "learning_rate": 9.889295249129754e-06,
+      "loss": 0.5998,
+      "step": 743
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2341341972351074,
+      "learning_rate": 9.888860598131733e-06,
+      "loss": 0.5873,
+      "step": 744
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1772515773773193,
+      "learning_rate": 9.888425105126845e-06,
+      "loss": 0.6183,
+      "step": 745
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9825636744499207,
+      "learning_rate": 9.887988770190093e-06,
+      "loss": 0.6205,
+      "step": 746
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7665997743606567,
+      "learning_rate": 9.887551593396628e-06,
+      "loss": 0.6367,
+      "step": 747
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.423568606376648,
+      "learning_rate": 9.887113574821743e-06,
+      "loss": 0.635,
+      "step": 748
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1755050420761108,
+      "learning_rate": 9.88667471454088e-06,
+      "loss": 0.5415,
+      "step": 749
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1046626567840576,
+      "learning_rate": 9.886235012629623e-06,
+      "loss": 0.6031,
+      "step": 750
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0809575319290161,
+      "learning_rate": 9.885794469163702e-06,
+      "loss": 0.6192,
+      "step": 751
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5112090110778809,
+      "learning_rate": 9.88535308421899e-06,
+      "loss": 0.6352,
+      "step": 752
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2142413854599,
+      "learning_rate": 9.884910857871508e-06,
+      "loss": 0.6814,
+      "step": 753
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.40277099609375,
+      "learning_rate": 9.88446779019742e-06,
+      "loss": 0.6682,
+      "step": 754
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2482854127883911,
+      "learning_rate": 9.884023881273038e-06,
+      "loss": 0.7463,
+      "step": 755
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.209838628768921,
+      "learning_rate": 9.883579131174813e-06,
+      "loss": 0.6347,
+      "step": 756
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2466986179351807,
+      "learning_rate": 9.883133539979342e-06,
+      "loss": 0.5901,
+      "step": 757
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2496932744979858,
+      "learning_rate": 9.882687107763374e-06,
+      "loss": 0.6438,
+      "step": 758
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1028106212615967,
+      "learning_rate": 9.882239834603798e-06,
+      "loss": 0.6648,
+      "step": 759
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3057689666748047,
+      "learning_rate": 9.881791720577643e-06,
+      "loss": 0.634,
+      "step": 760
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1680352687835693,
+      "learning_rate": 9.88134276576209e-06,
+      "loss": 0.6429,
+      "step": 761
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.238755702972412,
+      "learning_rate": 9.880892970234464e-06,
+      "loss": 0.7034,
+      "step": 762
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2210910320281982,
+      "learning_rate": 9.880442334072232e-06,
+      "loss": 0.6392,
+      "step": 763
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2049086093902588,
+      "learning_rate": 9.879990857353007e-06,
+      "loss": 0.5435,
+      "step": 764
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2110891342163086,
+      "learning_rate": 9.879538540154545e-06,
+      "loss": 0.6339,
+      "step": 765
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2696014642715454,
+      "learning_rate": 9.87908538255475e-06,
+      "loss": 0.5821,
+      "step": 766
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.348198652267456,
+      "learning_rate": 9.87863138463167e-06,
+      "loss": 0.7056,
+      "step": 767
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1112827062606812,
+      "learning_rate": 9.878176546463495e-06,
+      "loss": 0.5591,
+      "step": 768
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0482698678970337,
+      "learning_rate": 9.877720868128562e-06,
+      "loss": 0.6431,
+      "step": 769
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2081336975097656,
+      "learning_rate": 9.877264349705353e-06,
+      "loss": 0.6349,
+      "step": 770
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1520190238952637,
+      "learning_rate": 9.876806991272497e-06,
+      "loss": 0.6484,
+      "step": 771
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0996439456939697,
+      "learning_rate": 9.876348792908757e-06,
+      "loss": 0.7195,
+      "step": 772
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3486486673355103,
+      "learning_rate": 9.875889754693056e-06,
+      "loss": 0.675,
+      "step": 773
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2174115180969238,
+      "learning_rate": 9.875429876704448e-06,
+      "loss": 0.6746,
+      "step": 774
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3793160915374756,
+      "learning_rate": 9.874969159022143e-06,
+      "loss": 0.6585,
+      "step": 775
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.225388526916504,
+      "learning_rate": 9.874507601725486e-06,
+      "loss": 0.6186,
+      "step": 776
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5237325429916382,
+      "learning_rate": 9.874045204893975e-06,
+      "loss": 0.6844,
+      "step": 777
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6531785726547241,
+      "learning_rate": 9.873581968607243e-06,
+      "loss": 0.617,
+      "step": 778
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1168662309646606,
+      "learning_rate": 9.87311789294508e-06,
+      "loss": 0.6097,
+      "step": 779
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.150830864906311,
+      "learning_rate": 9.872652977987409e-06,
+      "loss": 0.5975,
+      "step": 780
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5129787921905518,
+      "learning_rate": 9.8721872238143e-06,
+      "loss": 0.594,
+      "step": 781
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2734906673431396,
+      "learning_rate": 9.871720630505975e-06,
+      "loss": 0.5761,
+      "step": 782
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1508820056915283,
+      "learning_rate": 9.871253198142792e-06,
+      "loss": 0.4496,
+      "step": 783
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.085600733757019,
+      "learning_rate": 9.870784926805258e-06,
+      "loss": 0.5349,
+      "step": 784
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2703758478164673,
+      "learning_rate": 9.870315816574022e-06,
+      "loss": 0.501,
+      "step": 785
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0857547521591187,
+      "learning_rate": 9.86984586752988e-06,
+      "loss": 0.6313,
+      "step": 786
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0965255498886108,
+      "learning_rate": 9.869375079753772e-06,
+      "loss": 0.6636,
+      "step": 787
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.169797658920288,
+      "learning_rate": 9.868903453326776e-06,
+      "loss": 0.7224,
+      "step": 788
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5188205242156982,
+      "learning_rate": 9.868430988330129e-06,
+      "loss": 0.6751,
+      "step": 789
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0894997119903564,
+      "learning_rate": 9.867957684845196e-06,
+      "loss": 0.6496,
+      "step": 790
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.124399185180664,
+      "learning_rate": 9.867483542953498e-06,
+      "loss": 0.7526,
+      "step": 791
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.100160837173462,
+      "learning_rate": 9.867008562736695e-06,
+      "loss": 0.6073,
+      "step": 792
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3616927862167358,
+      "learning_rate": 9.86653274427659e-06,
+      "loss": 0.6064,
+      "step": 793
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.117180585861206,
+      "learning_rate": 9.86605608765514e-06,
+      "loss": 0.5775,
+      "step": 794
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1300320625305176,
+      "learning_rate": 9.865578592954433e-06,
+      "loss": 0.5791,
+      "step": 795
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2053767442703247,
+      "learning_rate": 9.86510026025671e-06,
+      "loss": 0.6415,
+      "step": 796
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2116667032241821,
+      "learning_rate": 9.864621089644356e-06,
+      "loss": 0.6081,
+      "step": 797
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.282387614250183,
+      "learning_rate": 9.864141081199893e-06,
+      "loss": 0.6175,
+      "step": 798
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0022624731063843,
+      "learning_rate": 9.863660235005999e-06,
+      "loss": 0.6809,
+      "step": 799
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.160461664199829,
+      "learning_rate": 9.863178551145484e-06,
+      "loss": 0.6991,
+      "step": 800
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3416701555252075,
+      "learning_rate": 9.862696029701316e-06,
+      "loss": 0.5662,
+      "step": 801
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1150017976760864,
+      "learning_rate": 9.86221267075659e-06,
+      "loss": 0.5754,
+      "step": 802
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3248844146728516,
+      "learning_rate": 9.861728474394564e-06,
+      "loss": 0.6595,
+      "step": 803
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3990356922149658,
+      "learning_rate": 9.861243440698622e-06,
+      "loss": 0.643,
+      "step": 804
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2016390562057495,
+      "learning_rate": 9.860757569752309e-06,
+      "loss": 0.6551,
+      "step": 805
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7251582145690918,
+      "learning_rate": 9.860270861639303e-06,
+      "loss": 0.6475,
+      "step": 806
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.329265832901001,
+      "learning_rate": 9.859783316443429e-06,
+      "loss": 0.6333,
+      "step": 807
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7398295402526855,
+      "learning_rate": 9.859294934248658e-06,
+      "loss": 0.6681,
+      "step": 808
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1610875129699707,
+      "learning_rate": 9.858805715139102e-06,
+      "loss": 0.643,
+      "step": 809
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1122682094573975,
+      "learning_rate": 9.858315659199023e-06,
+      "loss": 0.599,
+      "step": 810
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0971416234970093,
+      "learning_rate": 9.857824766512818e-06,
+      "loss": 0.5669,
+      "step": 811
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1175240278244019,
+      "learning_rate": 9.857333037165038e-06,
+      "loss": 0.6802,
+      "step": 812
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5853863954544067,
+      "learning_rate": 9.85684047124037e-06,
+      "loss": 0.6693,
+      "step": 813
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.930460810661316,
+      "learning_rate": 9.85634706882365e-06,
+      "loss": 0.5844,
+      "step": 814
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2769625186920166,
+      "learning_rate": 9.855852829999857e-06,
+      "loss": 0.672,
+      "step": 815
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0591260194778442,
+      "learning_rate": 9.855357754854112e-06,
+      "loss": 0.617,
+      "step": 816
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2074815034866333,
+      "learning_rate": 9.854861843471683e-06,
+      "loss": 0.6256,
+      "step": 817
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4503984451293945,
+      "learning_rate": 9.854365095937982e-06,
+      "loss": 0.6393,
+      "step": 818
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3213568925857544,
+      "learning_rate": 9.85386751233856e-06,
+      "loss": 0.74,
+      "step": 819
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.231770396232605,
+      "learning_rate": 9.853369092759118e-06,
+      "loss": 0.6627,
+      "step": 820
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1289067268371582,
+      "learning_rate": 9.852869837285497e-06,
+      "loss": 0.73,
+      "step": 821
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8980963230133057,
+      "learning_rate": 9.852369746003685e-06,
+      "loss": 0.6039,
+      "step": 822
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.139829397201538,
+      "learning_rate": 9.851868818999813e-06,
+      "loss": 0.5732,
+      "step": 823
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1257266998291016,
+      "learning_rate": 9.851367056360154e-06,
+      "loss": 0.6825,
+      "step": 824
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.271867275238037,
+      "learning_rate": 9.85086445817113e-06,
+      "loss": 0.6534,
+      "step": 825
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1219916343688965,
+      "learning_rate": 9.850361024519298e-06,
+      "loss": 0.7195,
+      "step": 826
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4811643362045288,
+      "learning_rate": 9.849856755491367e-06,
+      "loss": 0.6099,
+      "step": 827
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.355452299118042,
+      "learning_rate": 9.849351651174188e-06,
+      "loss": 0.6789,
+      "step": 828
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5336081981658936,
+      "learning_rate": 9.848845711654754e-06,
+      "loss": 0.6725,
+      "step": 829
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1408854722976685,
+      "learning_rate": 9.8483389370202e-06,
+      "loss": 0.6471,
+      "step": 830
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3404558897018433,
+      "learning_rate": 9.847831327357813e-06,
+      "loss": 0.7116,
+      "step": 831
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0383011102676392,
+      "learning_rate": 9.847322882755015e-06,
+      "loss": 0.7415,
+      "step": 832
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4247316122055054,
+      "learning_rate": 9.846813603299377e-06,
+      "loss": 0.6714,
+      "step": 833
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.263272762298584,
+      "learning_rate": 9.846303489078608e-06,
+      "loss": 0.6267,
+      "step": 834
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9594109654426575,
+      "learning_rate": 9.845792540180569e-06,
+      "loss": 0.6712,
+      "step": 835
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1622424125671387,
+      "learning_rate": 9.845280756693262e-06,
+      "loss": 0.6655,
+      "step": 836
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1319442987442017,
+      "learning_rate": 9.844768138704825e-06,
+      "loss": 0.531,
+      "step": 837
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1469734907150269,
+      "learning_rate": 9.844254686303552e-06,
+      "loss": 0.5315,
+      "step": 838
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3223508596420288,
+      "learning_rate": 9.843740399577872e-06,
+      "loss": 0.5838,
+      "step": 839
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5986974239349365,
+      "learning_rate": 9.843225278616363e-06,
+      "loss": 0.7203,
+      "step": 840
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.020804762840271,
+      "learning_rate": 9.842709323507737e-06,
+      "loss": 0.6836,
+      "step": 841
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5095237493515015,
+      "learning_rate": 9.842192534340864e-06,
+      "loss": 0.6815,
+      "step": 842
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7738572359085083,
+      "learning_rate": 9.84167491120475e-06,
+      "loss": 0.6532,
+      "step": 843
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0000008344650269,
+      "learning_rate": 9.84115645418854e-06,
+      "loss": 0.6143,
+      "step": 844
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1826478242874146,
+      "learning_rate": 9.840637163381534e-06,
+      "loss": 0.6351,
+      "step": 845
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4132177829742432,
+      "learning_rate": 9.840117038873165e-06,
+      "loss": 0.744,
+      "step": 846
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2735340595245361,
+      "learning_rate": 9.839596080753015e-06,
+      "loss": 0.7583,
+      "step": 847
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9481948614120483,
+      "learning_rate": 9.83907428911081e-06,
+      "loss": 0.6783,
+      "step": 848
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3666658401489258,
+      "learning_rate": 9.838551664036414e-06,
+      "loss": 0.6517,
+      "step": 849
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2822260856628418,
+      "learning_rate": 9.838028205619843e-06,
+      "loss": 0.6163,
+      "step": 850
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2314839363098145,
+      "learning_rate": 9.837503913951249e-06,
+      "loss": 0.6752,
+      "step": 851
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1731816530227661,
+      "learning_rate": 9.836978789120933e-06,
+      "loss": 0.648,
+      "step": 852
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2162363529205322,
+      "learning_rate": 9.836452831219334e-06,
+      "loss": 0.5972,
+      "step": 853
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1823762655258179,
+      "learning_rate": 9.83592604033704e-06,
+      "loss": 0.5839,
+      "step": 854
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.037787675857544,
+      "learning_rate": 9.835398416564781e-06,
+      "loss": 0.6314,
+      "step": 855
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1973769664764404,
+      "learning_rate": 9.834869959993427e-06,
+      "loss": 0.6461,
+      "step": 856
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4603095054626465,
+      "learning_rate": 9.834340670713994e-06,
+      "loss": 0.548,
+      "step": 857
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1810789108276367,
+      "learning_rate": 9.833810548817644e-06,
+      "loss": 0.5911,
+      "step": 858
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5874276161193848,
+      "learning_rate": 9.833279594395676e-06,
+      "loss": 0.6555,
+      "step": 859
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2780070304870605,
+      "learning_rate": 9.832747807539538e-06,
+      "loss": 0.6349,
+      "step": 860
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2569433450698853,
+      "learning_rate": 9.83221518834082e-06,
+      "loss": 0.7218,
+      "step": 861
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1834901571273804,
+      "learning_rate": 9.831681736891255e-06,
+      "loss": 0.5845,
+      "step": 862
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2185614109039307,
+      "learning_rate": 9.831147453282717e-06,
+      "loss": 0.6264,
+      "step": 863
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.341444969177246,
+      "learning_rate": 9.830612337607227e-06,
+      "loss": 0.7407,
+      "step": 864
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4523122310638428,
+      "learning_rate": 9.830076389956949e-06,
+      "loss": 0.6082,
+      "step": 865
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6930763721466064,
+      "learning_rate": 9.829539610424186e-06,
+      "loss": 0.621,
+      "step": 866
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5043352842330933,
+      "learning_rate": 9.82900199910139e-06,
+      "loss": 0.6928,
+      "step": 867
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6809231042861938,
+      "learning_rate": 9.828463556081153e-06,
+      "loss": 0.6384,
+      "step": 868
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4405546188354492,
+      "learning_rate": 9.827924281456211e-06,
+      "loss": 0.6553,
+      "step": 869
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1939656734466553,
+      "learning_rate": 9.827384175319443e-06,
+      "loss": 0.6822,
+      "step": 870
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.065596103668213,
+      "learning_rate": 9.82684323776387e-06,
+      "loss": 0.6755,
+      "step": 871
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.822675347328186,
+      "learning_rate": 9.82630146888266e-06,
+      "loss": 0.5842,
+      "step": 872
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0467437505722046,
+      "learning_rate": 9.825758868769117e-06,
+      "loss": 0.6246,
+      "step": 873
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2898199558258057,
+      "learning_rate": 9.8252154375167e-06,
+      "loss": 0.6213,
+      "step": 874
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1529115438461304,
+      "learning_rate": 9.824671175218999e-06,
+      "loss": 0.6626,
+      "step": 875
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5466557741165161,
+      "learning_rate": 9.82412608196975e-06,
+      "loss": 0.5785,
+      "step": 876
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.961628794670105,
+      "learning_rate": 9.82358015786284e-06,
+      "loss": 0.5305,
+      "step": 877
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2553181648254395,
+      "learning_rate": 9.823033402992292e-06,
+      "loss": 0.6527,
+      "step": 878
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.308699607849121,
+      "learning_rate": 9.82248581745227e-06,
+      "loss": 0.6968,
+      "step": 879
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1388115882873535,
+      "learning_rate": 9.821937401337086e-06,
+      "loss": 0.71,
+      "step": 880
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1120529174804688,
+      "learning_rate": 9.821388154741198e-06,
+      "loss": 0.7006,
+      "step": 881
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2875672578811646,
+      "learning_rate": 9.820838077759195e-06,
+      "loss": 0.638,
+      "step": 882
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.338017463684082,
+      "learning_rate": 9.820287170485822e-06,
+      "loss": 0.5622,
+      "step": 883
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2129225730895996,
+      "learning_rate": 9.81973543301596e-06,
+      "loss": 0.6085,
+      "step": 884
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.669423222541809,
+      "learning_rate": 9.819182865444633e-06,
+      "loss": 0.6449,
+      "step": 885
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1424647569656372,
+      "learning_rate": 9.81862946786701e-06,
+      "loss": 0.5807,
+      "step": 886
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2412691116333008,
+      "learning_rate": 9.818075240378406e-06,
+      "loss": 0.6645,
+      "step": 887
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2484862804412842,
+      "learning_rate": 9.817520183074273e-06,
+      "loss": 0.6166,
+      "step": 888
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2282848358154297,
+      "learning_rate": 9.816964296050208e-06,
+      "loss": 0.5975,
+      "step": 889
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0314067602157593,
+      "learning_rate": 9.81640757940195e-06,
+      "loss": 0.6313,
+      "step": 890
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.192475438117981,
+      "learning_rate": 9.815850033225384e-06,
+      "loss": 0.6075,
+      "step": 891
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1968879699707031,
+      "learning_rate": 9.815291657616537e-06,
+      "loss": 0.6369,
+      "step": 892
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4490315914154053,
+      "learning_rate": 9.814732452671575e-06,
+      "loss": 0.6103,
+      "step": 893
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.05226731300354,
+      "learning_rate": 9.81417241848681e-06,
+      "loss": 0.6934,
+      "step": 894
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.841873049736023,
+      "learning_rate": 9.813611555158701e-06,
+      "loss": 0.6649,
+      "step": 895
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2420374155044556,
+      "learning_rate": 9.81304986278384e-06,
+      "loss": 0.6011,
+      "step": 896
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2118569612503052,
+      "learning_rate": 9.81248734145897e-06,
+      "loss": 0.6009,
+      "step": 897
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2946357727050781,
+      "learning_rate": 9.811923991280972e-06,
+      "loss": 0.6431,
+      "step": 898
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1636987924575806,
+      "learning_rate": 9.811359812346872e-06,
+      "loss": 0.741,
+      "step": 899
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2798045873641968,
+      "learning_rate": 9.810794804753838e-06,
+      "loss": 0.5853,
+      "step": 900
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2087842226028442,
+      "learning_rate": 9.810228968599185e-06,
+      "loss": 0.675,
+      "step": 901
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2338616847991943,
+      "learning_rate": 9.809662303980361e-06,
+      "loss": 0.62,
+      "step": 902
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2784311771392822,
+      "learning_rate": 9.809094810994966e-06,
+      "loss": 0.6567,
+      "step": 903
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1460515260696411,
+      "learning_rate": 9.808526489740739e-06,
+      "loss": 0.6384,
+      "step": 904
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9330402612686157,
+      "learning_rate": 9.80795734031556e-06,
+      "loss": 0.5802,
+      "step": 905
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.233659029006958,
+      "learning_rate": 9.807387362817455e-06,
+      "loss": 0.7364,
+      "step": 906
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.561555862426758,
+      "learning_rate": 9.806816557344592e-06,
+      "loss": 0.6519,
+      "step": 907
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.063315510749817,
+      "learning_rate": 9.806244923995275e-06,
+      "loss": 0.6765,
+      "step": 908
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2617602348327637,
+      "learning_rate": 9.805672462867966e-06,
+      "loss": 0.6217,
+      "step": 909
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2094106674194336,
+      "learning_rate": 9.80509917406125e-06,
+      "loss": 0.5438,
+      "step": 910
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0630391836166382,
+      "learning_rate": 9.804525057673871e-06,
+      "loss": 0.5533,
+      "step": 911
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.280405879020691,
+      "learning_rate": 9.803950113804705e-06,
+      "loss": 0.6173,
+      "step": 912
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9976212978363037,
+      "learning_rate": 9.803374342552776e-06,
+      "loss": 0.6336,
+      "step": 913
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2395212650299072,
+      "learning_rate": 9.80279774401725e-06,
+      "loss": 0.639,
+      "step": 914
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0815329551696777,
+      "learning_rate": 9.802220318297432e-06,
+      "loss": 0.5816,
+      "step": 915
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3019007444381714,
+      "learning_rate": 9.801642065492773e-06,
+      "loss": 0.5943,
+      "step": 916
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2104711532592773,
+      "learning_rate": 9.801062985702865e-06,
+      "loss": 0.6206,
+      "step": 917
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.823017954826355,
+      "learning_rate": 9.800483079027443e-06,
+      "loss": 0.6293,
+      "step": 918
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2773282527923584,
+      "learning_rate": 9.799902345566384e-06,
+      "loss": 0.5979,
+      "step": 919
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3274632692337036,
+      "learning_rate": 9.799320785419709e-06,
+      "loss": 0.6451,
+      "step": 920
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7359416484832764,
+      "learning_rate": 9.798738398687577e-06,
+      "loss": 0.6662,
+      "step": 921
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.245658278465271,
+      "learning_rate": 9.798155185470296e-06,
+      "loss": 0.726,
+      "step": 922
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.3268086910247803,
+      "learning_rate": 9.797571145868308e-06,
+      "loss": 0.6302,
+      "step": 923
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6008139848709106,
+      "learning_rate": 9.796986279982204e-06,
+      "loss": 0.6036,
+      "step": 924
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4884922504425049,
+      "learning_rate": 9.796400587912717e-06,
+      "loss": 0.5932,
+      "step": 925
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7857106924057007,
+      "learning_rate": 9.795814069760719e-06,
+      "loss": 0.7126,
+      "step": 926
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5720033645629883,
+      "learning_rate": 9.795226725627225e-06,
+      "loss": 0.6674,
+      "step": 927
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4257798194885254,
+      "learning_rate": 9.794638555613396e-06,
+      "loss": 0.6579,
+      "step": 928
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1718521118164062,
+      "learning_rate": 9.794049559820531e-06,
+      "loss": 0.7428,
+      "step": 929
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1296699047088623,
+      "learning_rate": 9.79345973835007e-06,
+      "loss": 0.6206,
+      "step": 930
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.188733458518982,
+      "learning_rate": 9.7928690913036e-06,
+      "loss": 0.6334,
+      "step": 931
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2753381729125977,
+      "learning_rate": 9.792277618782848e-06,
+      "loss": 0.5784,
+      "step": 932
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5454579591751099,
+      "learning_rate": 9.791685320889683e-06,
+      "loss": 0.6151,
+      "step": 933
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.210261344909668,
+      "learning_rate": 9.791092197726118e-06,
+      "loss": 0.6163,
+      "step": 934
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2641156911849976,
+      "learning_rate": 9.790498249394301e-06,
+      "loss": 0.5978,
+      "step": 935
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.322339415550232,
+      "learning_rate": 9.789903475996533e-06,
+      "loss": 0.6684,
+      "step": 936
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0136418342590332,
+      "learning_rate": 9.78930787763525e-06,
+      "loss": 0.6621,
+      "step": 937
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7304836511611938,
+      "learning_rate": 9.78871145441303e-06,
+      "loss": 0.679,
+      "step": 938
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1343514919281006,
+      "learning_rate": 9.788114206432597e-06,
+      "loss": 0.6472,
+      "step": 939
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1854240894317627,
+      "learning_rate": 9.787516133796816e-06,
+      "loss": 0.5903,
+      "step": 940
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0253652334213257,
+      "learning_rate": 9.786917236608687e-06,
+      "loss": 0.6461,
+      "step": 941
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.077979564666748,
+      "learning_rate": 9.786317514971364e-06,
+      "loss": 0.6138,
+      "step": 942
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2683407068252563,
+      "learning_rate": 9.785716968988133e-06,
+      "loss": 0.6181,
+      "step": 943
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0985451936721802,
+      "learning_rate": 9.785115598762429e-06,
+      "loss": 0.6436,
+      "step": 944
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1492769718170166,
+      "learning_rate": 9.784513404397823e-06,
+      "loss": 0.6205,
+      "step": 945
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1750531196594238,
+      "learning_rate": 9.783910385998034e-06,
+      "loss": 0.6014,
+      "step": 946
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1315195560455322,
+      "learning_rate": 9.783306543666916e-06,
+      "loss": 0.6288,
+      "step": 947
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1699100732803345,
+      "learning_rate": 9.78270187750847e-06,
+      "loss": 0.6545,
+      "step": 948
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1685606241226196,
+      "learning_rate": 9.782096387626839e-06,
+      "loss": 0.6522,
+      "step": 949
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6117463111877441,
+      "learning_rate": 9.781490074126305e-06,
+      "loss": 0.593,
+      "step": 950
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1152682304382324,
+      "learning_rate": 9.780882937111292e-06,
+      "loss": 0.6196,
+      "step": 951
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2452787160873413,
+      "learning_rate": 9.780274976686369e-06,
+      "loss": 0.6037,
+      "step": 952
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2231415510177612,
+      "learning_rate": 9.779666192956243e-06,
+      "loss": 0.5567,
+      "step": 953
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1500630378723145,
+      "learning_rate": 9.779056586025766e-06,
+      "loss": 0.6495,
+      "step": 954
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3221664428710938,
+      "learning_rate": 9.77844615599993e-06,
+      "loss": 0.6374,
+      "step": 955
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1846741437911987,
+      "learning_rate": 9.777834902983869e-06,
+      "loss": 0.6598,
+      "step": 956
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.418726921081543,
+      "learning_rate": 9.77722282708286e-06,
+      "loss": 0.7151,
+      "step": 957
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4351948499679565,
+      "learning_rate": 9.776609928402319e-06,
+      "loss": 0.6353,
+      "step": 958
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.146484136581421,
+      "learning_rate": 9.775996207047804e-06,
+      "loss": 0.6261,
+      "step": 959
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1938923597335815,
+      "learning_rate": 9.775381663125021e-06,
+      "loss": 0.6846,
+      "step": 960
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3629497289657593,
+      "learning_rate": 9.774766296739808e-06,
+      "loss": 0.6335,
+      "step": 961
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2427971363067627,
+      "learning_rate": 9.774150107998153e-06,
+      "loss": 0.7141,
+      "step": 962
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.049538016319275,
+      "learning_rate": 9.77353309700618e-06,
+      "loss": 0.6309,
+      "step": 963
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2472624778747559,
+      "learning_rate": 9.772915263870155e-06,
+      "loss": 0.6833,
+      "step": 964
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1295534372329712,
+      "learning_rate": 9.77229660869649e-06,
+      "loss": 0.5353,
+      "step": 965
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.565324306488037,
+      "learning_rate": 9.771677131591734e-06,
+      "loss": 0.6867,
+      "step": 966
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5009647607803345,
+      "learning_rate": 9.771056832662581e-06,
+      "loss": 0.6216,
+      "step": 967
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4722943305969238,
+      "learning_rate": 9.770435712015863e-06,
+      "loss": 0.6958,
+      "step": 968
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2581201791763306,
+      "learning_rate": 9.769813769758556e-06,
+      "loss": 0.455,
+      "step": 969
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1791751384735107,
+      "learning_rate": 9.769191005997781e-06,
+      "loss": 0.6739,
+      "step": 970
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7730474472045898,
+      "learning_rate": 9.76856742084079e-06,
+      "loss": 0.6605,
+      "step": 971
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2831331491470337,
+      "learning_rate": 9.767943014394982e-06,
+      "loss": 0.6259,
+      "step": 972
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7151639461517334,
+      "learning_rate": 9.767317786767907e-06,
+      "loss": 0.6003,
+      "step": 973
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.258449912071228,
+      "learning_rate": 9.76669173806724e-06,
+      "loss": 0.6271,
+      "step": 974
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1486563682556152,
+      "learning_rate": 9.766064868400808e-06,
+      "loss": 0.6293,
+      "step": 975
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.290198564529419,
+      "learning_rate": 9.765437177876576e-06,
+      "loss": 0.6359,
+      "step": 976
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.152216911315918,
+      "learning_rate": 9.764808666602651e-06,
+      "loss": 0.6303,
+      "step": 977
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1518402099609375,
+      "learning_rate": 9.764179334687284e-06,
+      "loss": 0.6305,
+      "step": 978
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1398547887802124,
+      "learning_rate": 9.76354918223886e-06,
+      "loss": 0.5392,
+      "step": 979
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.361898422241211,
+      "learning_rate": 9.762918209365914e-06,
+      "loss": 0.6505,
+      "step": 980
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2657636404037476,
+      "learning_rate": 9.762286416177114e-06,
+      "loss": 0.631,
+      "step": 981
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4761649370193481,
+      "learning_rate": 9.761653802781276e-06,
+      "loss": 0.6788,
+      "step": 982
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5255000591278076,
+      "learning_rate": 9.761020369287358e-06,
+      "loss": 0.572,
+      "step": 983
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4596574306488037,
+      "learning_rate": 9.76038611580445e-06,
+      "loss": 0.6187,
+      "step": 984
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4689172506332397,
+      "learning_rate": 9.759751042441792e-06,
+      "loss": 0.5464,
+      "step": 985
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1188050508499146,
+      "learning_rate": 9.759115149308761e-06,
+      "loss": 0.6616,
+      "step": 986
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3672010898590088,
+      "learning_rate": 9.758478436514882e-06,
+      "loss": 0.7279,
+      "step": 987
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.522152304649353,
+      "learning_rate": 9.75784090416981e-06,
+      "loss": 0.5857,
+      "step": 988
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1300358772277832,
+      "learning_rate": 9.757202552383349e-06,
+      "loss": 0.6825,
+      "step": 989
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.252014398574829,
+      "learning_rate": 9.756563381265443e-06,
+      "loss": 0.6484,
+      "step": 990
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.564777135848999,
+      "learning_rate": 9.755923390926173e-06,
+      "loss": 0.6302,
+      "step": 991
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6608165502548218,
+      "learning_rate": 9.755282581475769e-06,
+      "loss": 0.6619,
+      "step": 992
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.2679226398468018,
+      "learning_rate": 9.754640953024595e-06,
+      "loss": 0.6166,
+      "step": 993
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.119726538658142,
+      "learning_rate": 9.753998505683157e-06,
+      "loss": 0.6459,
+      "step": 994
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1786524057388306,
+      "learning_rate": 9.753355239562106e-06,
+      "loss": 0.6327,
+      "step": 995
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1883645057678223,
+      "learning_rate": 9.75271115477223e-06,
+      "loss": 0.6711,
+      "step": 996
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6034826040267944,
+      "learning_rate": 9.75206625142446e-06,
+      "loss": 0.6559,
+      "step": 997
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1943557262420654,
+      "learning_rate": 9.751420529629869e-06,
+      "loss": 0.5873,
+      "step": 998
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.274271011352539,
+      "learning_rate": 9.750773989499665e-06,
+      "loss": 0.6383,
+      "step": 999
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3029786348342896,
+      "learning_rate": 9.750126631145207e-06,
+      "loss": 0.6523,
+      "step": 1000
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4161441326141357,
+      "learning_rate": 9.749478454677986e-06,
+      "loss": 0.6542,
+      "step": 1001
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0572928190231323,
+      "learning_rate": 9.748829460209638e-06,
+      "loss": 0.6725,
+      "step": 1002
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0757194757461548,
+      "learning_rate": 9.748179647851939e-06,
+      "loss": 0.5555,
+      "step": 1003
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.006316900253296,
+      "learning_rate": 9.747529017716806e-06,
+      "loss": 0.6257,
+      "step": 1004
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1826972961425781,
+      "learning_rate": 9.746877569916297e-06,
+      "loss": 0.6045,
+      "step": 1005
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1543837785720825,
+      "learning_rate": 9.74622530456261e-06,
+      "loss": 0.5696,
+      "step": 1006
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5225355625152588,
+      "learning_rate": 9.745572221768085e-06,
+      "loss": 0.6135,
+      "step": 1007
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3935890197753906,
+      "learning_rate": 9.744918321645203e-06,
+      "loss": 0.6532,
+      "step": 1008
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2201215028762817,
+      "learning_rate": 9.744263604306582e-06,
+      "loss": 0.654,
+      "step": 1009
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4088938236236572,
+      "learning_rate": 9.743608069864988e-06,
+      "loss": 0.6236,
+      "step": 1010
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1685870885849,
+      "learning_rate": 9.742951718433321e-06,
+      "loss": 0.6695,
+      "step": 1011
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9566943645477295,
+      "learning_rate": 9.742294550124625e-06,
+      "loss": 0.6839,
+      "step": 1012
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.222412347793579,
+      "learning_rate": 9.741636565052081e-06,
+      "loss": 0.614,
+      "step": 1013
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.131980299949646,
+      "learning_rate": 9.740977763329018e-06,
+      "loss": 0.598,
+      "step": 1014
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1899813413619995,
+      "learning_rate": 9.740318145068902e-06,
+      "loss": 0.6678,
+      "step": 1015
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1226192712783813,
+      "learning_rate": 9.739657710385332e-06,
+      "loss": 0.6595,
+      "step": 1016
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.9692671298980713,
+      "learning_rate": 9.73899645939206e-06,
+      "loss": 0.6617,
+      "step": 1017
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3050158023834229,
+      "learning_rate": 9.738334392202971e-06,
+      "loss": 0.6292,
+      "step": 1018
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.244711995124817,
+      "learning_rate": 9.737671508932093e-06,
+      "loss": 0.6136,
+      "step": 1019
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4500633478164673,
+      "learning_rate": 9.737007809693595e-06,
+      "loss": 0.6355,
+      "step": 1020
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2068707942962646,
+      "learning_rate": 9.736343294601784e-06,
+      "loss": 0.5697,
+      "step": 1021
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1867763996124268,
+      "learning_rate": 9.735677963771112e-06,
+      "loss": 0.672,
+      "step": 1022
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.561131477355957,
+      "learning_rate": 9.735011817316166e-06,
+      "loss": 0.6256,
+      "step": 1023
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.072286128997803,
+      "learning_rate": 9.734344855351678e-06,
+      "loss": 0.6563,
+      "step": 1024
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.22327721118927,
+      "learning_rate": 9.733677077992515e-06,
+      "loss": 0.6342,
+      "step": 1025
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7818228006362915,
+      "learning_rate": 9.733008485353694e-06,
+      "loss": 0.645,
+      "step": 1026
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4120960235595703,
+      "learning_rate": 9.732339077550364e-06,
+      "loss": 0.6384,
+      "step": 1027
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8205273151397705,
+      "learning_rate": 9.731668854697814e-06,
+      "loss": 0.6097,
+      "step": 1028
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6517037153244019,
+      "learning_rate": 9.730997816911479e-06,
+      "loss": 0.6568,
+      "step": 1029
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1279886960983276,
+      "learning_rate": 9.730325964306931e-06,
+      "loss": 0.5864,
+      "step": 1030
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.175146222114563,
+      "learning_rate": 9.729653296999883e-06,
+      "loss": 0.6494,
+      "step": 1031
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0476584434509277,
+      "learning_rate": 9.72897981510619e-06,
+      "loss": 0.5538,
+      "step": 1032
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0945863723754883,
+      "learning_rate": 9.728305518741844e-06,
+      "loss": 0.589,
+      "step": 1033
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1936923265457153,
+      "learning_rate": 9.727630408022977e-06,
+      "loss": 0.5998,
+      "step": 1034
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1819581985473633,
+      "learning_rate": 9.726954483065864e-06,
+      "loss": 0.6506,
+      "step": 1035
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1652607917785645,
+      "learning_rate": 9.726277743986923e-06,
+      "loss": 0.6516,
+      "step": 1036
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4037582874298096,
+      "learning_rate": 9.725600190902704e-06,
+      "loss": 0.7143,
+      "step": 1037
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1429693698883057,
+      "learning_rate": 9.724921823929905e-06,
+      "loss": 0.6593,
+      "step": 1038
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.021639347076416,
+      "learning_rate": 9.724242643185357e-06,
+      "loss": 0.6161,
+      "step": 1039
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2161128520965576,
+      "learning_rate": 9.72356264878604e-06,
+      "loss": 0.6372,
+      "step": 1040
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1113938093185425,
+      "learning_rate": 9.722881840849066e-06,
+      "loss": 0.63,
+      "step": 1041
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1752476692199707,
+      "learning_rate": 9.722200219491692e-06,
+      "loss": 0.7094,
+      "step": 1042
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.142798662185669,
+      "learning_rate": 9.72151778483131e-06,
+      "loss": 0.6925,
+      "step": 1043
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1876989603042603,
+      "learning_rate": 9.72083453698546e-06,
+      "loss": 0.6793,
+      "step": 1044
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0826714038848877,
+      "learning_rate": 9.720150476071814e-06,
+      "loss": 0.6886,
+      "step": 1045
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.2857930660247803,
+      "learning_rate": 9.71946560220819e-06,
+      "loss": 0.5505,
+      "step": 1046
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1982847452163696,
+      "learning_rate": 9.718779915512542e-06,
+      "loss": 0.651,
+      "step": 1047
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4459384679794312,
+      "learning_rate": 9.718093416102967e-06,
+      "loss": 0.6223,
+      "step": 1048
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1260446310043335,
+      "learning_rate": 9.717406104097702e-06,
+      "loss": 0.6248,
+      "step": 1049
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0635147094726562,
+      "learning_rate": 9.716717979615118e-06,
+      "loss": 0.5919,
+      "step": 1050
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1763317584991455,
+      "learning_rate": 9.716029042773733e-06,
+      "loss": 0.636,
+      "step": 1051
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2086479663848877,
+      "learning_rate": 9.715339293692203e-06,
+      "loss": 0.651,
+      "step": 1052
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3193252086639404,
+      "learning_rate": 9.714648732489322e-06,
+      "loss": 0.633,
+      "step": 1053
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6744786500930786,
+      "learning_rate": 9.713957359284025e-06,
+      "loss": 0.5887,
+      "step": 1054
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2904282808303833,
+      "learning_rate": 9.713265174195388e-06,
+      "loss": 0.68,
+      "step": 1055
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4331204891204834,
+      "learning_rate": 9.712572177342627e-06,
+      "loss": 0.5989,
+      "step": 1056
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9341638684272766,
+      "learning_rate": 9.711878368845094e-06,
+      "loss": 0.647,
+      "step": 1057
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3745675086975098,
+      "learning_rate": 9.711183748822284e-06,
+      "loss": 0.6102,
+      "step": 1058
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2217185497283936,
+      "learning_rate": 9.710488317393832e-06,
+      "loss": 0.6091,
+      "step": 1059
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.1755502223968506,
+      "learning_rate": 9.709792074679511e-06,
+      "loss": 0.6482,
+      "step": 1060
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2285881042480469,
+      "learning_rate": 9.709095020799236e-06,
+      "loss": 0.7853,
+      "step": 1061
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1444462537765503,
+      "learning_rate": 9.70839715587306e-06,
+      "loss": 0.6351,
+      "step": 1062
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0068621635437012,
+      "learning_rate": 9.707698480021174e-06,
+      "loss": 0.7285,
+      "step": 1063
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1805295944213867,
+      "learning_rate": 9.706998993363913e-06,
+      "loss": 0.5867,
+      "step": 1064
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2351415157318115,
+      "learning_rate": 9.70629869602175e-06,
+      "loss": 0.5677,
+      "step": 1065
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.22136652469635,
+      "learning_rate": 9.705597588115295e-06,
+      "loss": 0.647,
+      "step": 1066
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.124104619026184,
+      "learning_rate": 9.7048956697653e-06,
+      "loss": 0.5704,
+      "step": 1067
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2475918531417847,
+      "learning_rate": 9.704192941092657e-06,
+      "loss": 0.6191,
+      "step": 1068
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.219774842262268,
+      "learning_rate": 9.703489402218394e-06,
+      "loss": 0.7046,
+      "step": 1069
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1891069412231445,
+      "learning_rate": 9.702785053263685e-06,
+      "loss": 0.5343,
+      "step": 1070
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.167614459991455,
+      "learning_rate": 9.702079894349838e-06,
+      "loss": 0.6189,
+      "step": 1071
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0407310724258423,
+      "learning_rate": 9.701373925598304e-06,
+      "loss": 0.6745,
+      "step": 1072
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5812993049621582,
+      "learning_rate": 9.70066714713067e-06,
+      "loss": 0.632,
+      "step": 1073
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2237656116485596,
+      "learning_rate": 9.699959559068664e-06,
+      "loss": 0.738,
+      "step": 1074
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2793015241622925,
+      "learning_rate": 9.699251161534153e-06,
+      "loss": 0.7889,
+      "step": 1075
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.167056679725647,
+      "learning_rate": 9.698541954649145e-06,
+      "loss": 0.6292,
+      "step": 1076
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5537493228912354,
+      "learning_rate": 9.69783193853579e-06,
+      "loss": 0.6604,
+      "step": 1077
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2457462549209595,
+      "learning_rate": 9.697121113316367e-06,
+      "loss": 0.618,
+      "step": 1078
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6209555864334106,
+      "learning_rate": 9.696409479113306e-06,
+      "loss": 0.5909,
+      "step": 1079
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1111366748809814,
+      "learning_rate": 9.695697036049173e-06,
+      "loss": 0.6108,
+      "step": 1080
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2718392610549927,
+      "learning_rate": 9.694983784246665e-06,
+      "loss": 0.7539,
+      "step": 1081
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2289928197860718,
+      "learning_rate": 9.69426972382863e-06,
+      "loss": 0.5455,
+      "step": 1082
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0418940782546997,
+      "learning_rate": 9.693554854918052e-06,
+      "loss": 0.5804,
+      "step": 1083
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1238722801208496,
+      "learning_rate": 9.692839177638049e-06,
+      "loss": 0.6448,
+      "step": 1084
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0625381469726562,
+      "learning_rate": 9.692122692111883e-06,
+      "loss": 0.6399,
+      "step": 1085
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4463095664978027,
+      "learning_rate": 9.691405398462956e-06,
+      "loss": 0.6596,
+      "step": 1086
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1603282690048218,
+      "learning_rate": 9.690687296814803e-06,
+      "loss": 0.615,
+      "step": 1087
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1188900470733643,
+      "learning_rate": 9.689968387291107e-06,
+      "loss": 0.5733,
+      "step": 1088
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2128100395202637,
+      "learning_rate": 9.68924867001568e-06,
+      "loss": 0.6144,
+      "step": 1089
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.295868158340454,
+      "learning_rate": 9.688528145112486e-06,
+      "loss": 0.6801,
+      "step": 1090
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2260469198226929,
+      "learning_rate": 9.687806812705616e-06,
+      "loss": 0.5645,
+      "step": 1091
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2258944511413574,
+      "learning_rate": 9.687084672919305e-06,
+      "loss": 0.5889,
+      "step": 1092
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4465066194534302,
+      "learning_rate": 9.686361725877926e-06,
+      "loss": 0.6063,
+      "step": 1093
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3785881996154785,
+      "learning_rate": 9.685637971705998e-06,
+      "loss": 0.5688,
+      "step": 1094
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1346842050552368,
+      "learning_rate": 9.684913410528164e-06,
+      "loss": 0.609,
+      "step": 1095
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1224477291107178,
+      "learning_rate": 9.68418804246922e-06,
+      "loss": 0.6036,
+      "step": 1096
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.019282579421997,
+      "learning_rate": 9.683461867654098e-06,
+      "loss": 0.6989,
+      "step": 1097
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0686557292938232,
+      "learning_rate": 9.682734886207863e-06,
+      "loss": 0.586,
+      "step": 1098
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1529673337936401,
+      "learning_rate": 9.682007098255724e-06,
+      "loss": 0.5873,
+      "step": 1099
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.140615463256836,
+      "learning_rate": 9.681278503923026e-06,
+      "loss": 0.5277,
+      "step": 1100
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3383703231811523,
+      "learning_rate": 9.680549103335259e-06,
+      "loss": 0.6375,
+      "step": 1101
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2095407247543335,
+      "learning_rate": 9.679818896618044e-06,
+      "loss": 0.7485,
+      "step": 1102
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5559945106506348,
+      "learning_rate": 9.679087883897142e-06,
+      "loss": 0.5913,
+      "step": 1103
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2311608791351318,
+      "learning_rate": 9.678356065298463e-06,
+      "loss": 0.6464,
+      "step": 1104
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1227388381958008,
+      "learning_rate": 9.677623440948039e-06,
+      "loss": 0.6501,
+      "step": 1105
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.120994210243225,
+      "learning_rate": 9.676890010972055e-06,
+      "loss": 0.5631,
+      "step": 1106
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5825780630111694,
+      "learning_rate": 9.676155775496828e-06,
+      "loss": 0.6043,
+      "step": 1107
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1823532581329346,
+      "learning_rate": 9.675420734648814e-06,
+      "loss": 0.5822,
+      "step": 1108
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5761380195617676,
+      "learning_rate": 9.674684888554611e-06,
+      "loss": 0.6976,
+      "step": 1109
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2008821964263916,
+      "learning_rate": 9.673948237340951e-06,
+      "loss": 0.5651,
+      "step": 1110
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1293318271636963,
+      "learning_rate": 9.67321078113471e-06,
+      "loss": 0.6455,
+      "step": 1111
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1750528812408447,
+      "learning_rate": 9.6724725200629e-06,
+      "loss": 0.6694,
+      "step": 1112
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3616904020309448,
+      "learning_rate": 9.671733454252668e-06,
+      "loss": 0.6136,
+      "step": 1113
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.360683560371399,
+      "learning_rate": 9.670993583831305e-06,
+      "loss": 0.6378,
+      "step": 1114
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.126110553741455,
+      "learning_rate": 9.67025290892624e-06,
+      "loss": 0.584,
+      "step": 1115
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2391899824142456,
+      "learning_rate": 9.669511429665035e-06,
+      "loss": 0.6807,
+      "step": 1116
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.075810432434082,
+      "learning_rate": 9.668769146175401e-06,
+      "loss": 0.6883,
+      "step": 1117
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9843566417694092,
+      "learning_rate": 9.668026058585175e-06,
+      "loss": 0.6793,
+      "step": 1118
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.44784677028656,
+      "learning_rate": 9.667282167022343e-06,
+      "loss": 0.6444,
+      "step": 1119
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0862517356872559,
+      "learning_rate": 9.666537471615024e-06,
+      "loss": 0.5965,
+      "step": 1120
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.401402235031128,
+      "learning_rate": 9.665791972491476e-06,
+      "loss": 0.6598,
+      "step": 1121
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1191637516021729,
+      "learning_rate": 9.665045669780098e-06,
+      "loss": 0.7505,
+      "step": 1122
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.114385962486267,
+      "learning_rate": 9.664298563609422e-06,
+      "loss": 0.655,
+      "step": 1123
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.621027946472168,
+      "learning_rate": 9.663550654108124e-06,
+      "loss": 0.58,
+      "step": 1124
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0535495281219482,
+      "learning_rate": 9.662801941405016e-06,
+      "loss": 0.5866,
+      "step": 1125
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.036341905593872,
+      "learning_rate": 9.662052425629049e-06,
+      "loss": 0.5704,
+      "step": 1126
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1747620105743408,
+      "learning_rate": 9.661302106909311e-06,
+      "loss": 0.5937,
+      "step": 1127
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3271484375,
+      "learning_rate": 9.66055098537503e-06,
+      "loss": 0.6407,
+      "step": 1128
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1150386333465576,
+      "learning_rate": 9.659799061155573e-06,
+      "loss": 0.6461,
+      "step": 1129
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1988625526428223,
+      "learning_rate": 9.659046334380438e-06,
+      "loss": 0.6438,
+      "step": 1130
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3866018056869507,
+      "learning_rate": 9.658292805179272e-06,
+      "loss": 0.6631,
+      "step": 1131
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2857260704040527,
+      "learning_rate": 9.657538473681855e-06,
+      "loss": 0.6402,
+      "step": 1132
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7626216411590576,
+      "learning_rate": 9.656783340018103e-06,
+      "loss": 0.6901,
+      "step": 1133
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3316880464553833,
+      "learning_rate": 9.656027404318073e-06,
+      "loss": 0.655,
+      "step": 1134
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3176456689834595,
+      "learning_rate": 9.65527066671196e-06,
+      "loss": 0.6157,
+      "step": 1135
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.27821683883667,
+      "learning_rate": 9.654513127330097e-06,
+      "loss": 0.752,
+      "step": 1136
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4629007577896118,
+      "learning_rate": 9.653754786302953e-06,
+      "loss": 0.6666,
+      "step": 1137
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0769317150115967,
+      "learning_rate": 9.652995643761139e-06,
+      "loss": 0.6648,
+      "step": 1138
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0857163667678833,
+      "learning_rate": 9.652235699835402e-06,
+      "loss": 0.5751,
+      "step": 1139
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2306089401245117,
+      "learning_rate": 9.651474954656626e-06,
+      "loss": 0.6447,
+      "step": 1140
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1074614524841309,
+      "learning_rate": 9.650713408355832e-06,
+      "loss": 0.6232,
+      "step": 1141
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.069425106048584,
+      "learning_rate": 9.649951061064184e-06,
+      "loss": 0.6284,
+      "step": 1142
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4408040046691895,
+      "learning_rate": 9.649187912912978e-06,
+      "loss": 0.603,
+      "step": 1143
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.965793490409851,
+      "learning_rate": 9.648423964033653e-06,
+      "loss": 0.609,
+      "step": 1144
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2444179058074951,
+      "learning_rate": 9.647659214557784e-06,
+      "loss": 0.6883,
+      "step": 1145
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3600820302963257,
+      "learning_rate": 9.646893664617082e-06,
+      "loss": 0.6466,
+      "step": 1146
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.187540888786316,
+      "learning_rate": 9.646127314343398e-06,
+      "loss": 0.6567,
+      "step": 1147
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.062788486480713,
+      "learning_rate": 9.64536016386872e-06,
+      "loss": 0.6549,
+      "step": 1148
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4273498058319092,
+      "learning_rate": 9.644592213325174e-06,
+      "loss": 0.573,
+      "step": 1149
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.6031153202056885,
+      "learning_rate": 9.643823462845023e-06,
+      "loss": 0.6893,
+      "step": 1150
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0803120136260986,
+      "learning_rate": 9.643053912560672e-06,
+      "loss": 0.6714,
+      "step": 1151
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4752882719039917,
+      "learning_rate": 9.642283562604658e-06,
+      "loss": 0.6817,
+      "step": 1152
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3621220588684082,
+      "learning_rate": 9.641512413109656e-06,
+      "loss": 0.6725,
+      "step": 1153
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.622391700744629,
+      "learning_rate": 9.640740464208484e-06,
+      "loss": 0.6874,
+      "step": 1154
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2322818040847778,
+      "learning_rate": 9.639967716034095e-06,
+      "loss": 0.6397,
+      "step": 1155
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1383590698242188,
+      "learning_rate": 9.639194168719577e-06,
+      "loss": 0.6025,
+      "step": 1156
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2494959831237793,
+      "learning_rate": 9.638419822398159e-06,
+      "loss": 0.6905,
+      "step": 1157
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1969854831695557,
+      "learning_rate": 9.637644677203205e-06,
+      "loss": 0.5309,
+      "step": 1158
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3315047025680542,
+      "learning_rate": 9.636868733268218e-06,
+      "loss": 0.691,
+      "step": 1159
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0063773393630981,
+      "learning_rate": 9.636091990726843e-06,
+      "loss": 0.5789,
+      "step": 1160
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9836358428001404,
+      "learning_rate": 9.635314449712852e-06,
+      "loss": 0.6658,
+      "step": 1161
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1989636421203613,
+      "learning_rate": 9.634536110360164e-06,
+      "loss": 0.6999,
+      "step": 1162
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0682141780853271,
+      "learning_rate": 9.633756972802832e-06,
+      "loss": 0.6808,
+      "step": 1163
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0833097696304321,
+      "learning_rate": 9.632977037175046e-06,
+      "loss": 0.6291,
+      "step": 1164
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.168710470199585,
+      "learning_rate": 9.632196303611131e-06,
+      "loss": 0.721,
+      "step": 1165
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2726129293441772,
+      "learning_rate": 9.63141477224556e-06,
+      "loss": 0.6047,
+      "step": 1166
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4029903411865234,
+      "learning_rate": 9.63063244321293e-06,
+      "loss": 0.6346,
+      "step": 1167
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2573045492172241,
+      "learning_rate": 9.629849316647983e-06,
+      "loss": 0.5604,
+      "step": 1168
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3267803192138672,
+      "learning_rate": 9.629065392685594e-06,
+      "loss": 0.7928,
+      "step": 1169
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.286224126815796,
+      "learning_rate": 9.628280671460784e-06,
+      "loss": 0.6406,
+      "step": 1170
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2955398559570312,
+      "learning_rate": 9.627495153108698e-06,
+      "loss": 0.6572,
+      "step": 1171
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.24276864528656,
+      "learning_rate": 9.626708837764634e-06,
+      "loss": 0.6414,
+      "step": 1172
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2331914901733398,
+      "learning_rate": 9.62592172556401e-06,
+      "loss": 0.5933,
+      "step": 1173
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0565133094787598,
+      "learning_rate": 9.625133816642395e-06,
+      "loss": 0.7286,
+      "step": 1174
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4214918613433838,
+      "learning_rate": 9.624345111135492e-06,
+      "loss": 0.6858,
+      "step": 1175
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2732834815979004,
+      "learning_rate": 9.623555609179134e-06,
+      "loss": 0.6623,
+      "step": 1176
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.498653769493103,
+      "learning_rate": 9.6227653109093e-06,
+      "loss": 0.6133,
+      "step": 1177
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5062004327774048,
+      "learning_rate": 9.621974216462101e-06,
+      "loss": 0.6736,
+      "step": 1178
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.379712462425232,
+      "learning_rate": 9.621182325973792e-06,
+      "loss": 0.6587,
+      "step": 1179
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.242719292640686,
+      "learning_rate": 9.620389639580753e-06,
+      "loss": 0.7805,
+      "step": 1180
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2989170551300049,
+      "learning_rate": 9.619596157419514e-06,
+      "loss": 0.6942,
+      "step": 1181
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.712216854095459,
+      "learning_rate": 9.618801879626734e-06,
+      "loss": 0.6364,
+      "step": 1182
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.460629940032959,
+      "learning_rate": 9.618006806339209e-06,
+      "loss": 0.6022,
+      "step": 1183
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.011130928993225,
+      "learning_rate": 9.617210937693876e-06,
+      "loss": 0.6697,
+      "step": 1184
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9472834467887878,
+      "learning_rate": 9.616414273827808e-06,
+      "loss": 0.5716,
+      "step": 1185
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0987147092819214,
+      "learning_rate": 9.615616814878213e-06,
+      "loss": 0.6417,
+      "step": 1186
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.320693850517273,
+      "learning_rate": 9.614818560982439e-06,
+      "loss": 0.6858,
+      "step": 1187
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2989826202392578,
+      "learning_rate": 9.614019512277966e-06,
+      "loss": 0.6196,
+      "step": 1188
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.199810266494751,
+      "learning_rate": 9.613219668902417e-06,
+      "loss": 0.7115,
+      "step": 1189
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0295907258987427,
+      "learning_rate": 9.612419030993545e-06,
+      "loss": 0.7165,
+      "step": 1190
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7425963878631592,
+      "learning_rate": 9.611617598689248e-06,
+      "loss": 0.5757,
+      "step": 1191
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9693262577056885,
+      "learning_rate": 9.610815372127553e-06,
+      "loss": 0.534,
+      "step": 1192
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3094513416290283,
+      "learning_rate": 9.610012351446629e-06,
+      "loss": 0.5866,
+      "step": 1193
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1087934970855713,
+      "learning_rate": 9.60920853678478e-06,
+      "loss": 0.6526,
+      "step": 1194
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9788431525230408,
+      "learning_rate": 9.608403928280445e-06,
+      "loss": 0.5877,
+      "step": 1195
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0494896173477173,
+      "learning_rate": 9.607598526072203e-06,
+      "loss": 0.6029,
+      "step": 1196
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3554272651672363,
+      "learning_rate": 9.606792330298767e-06,
+      "loss": 0.6737,
+      "step": 1197
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2967965602874756,
+      "learning_rate": 9.60598534109899e-06,
+      "loss": 0.7037,
+      "step": 1198
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.304180383682251,
+      "learning_rate": 9.605177558611859e-06,
+      "loss": 0.6131,
+      "step": 1199
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.786306619644165,
+      "learning_rate": 9.604368982976495e-06,
+      "loss": 0.6623,
+      "step": 1200
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.139390468597412,
+      "learning_rate": 9.603559614332162e-06,
+      "loss": 0.5915,
+      "step": 1201
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1185011863708496,
+      "learning_rate": 9.602749452818256e-06,
+      "loss": 0.5222,
+      "step": 1202
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.157008409500122,
+      "learning_rate": 9.601938498574312e-06,
+      "loss": 0.6954,
+      "step": 1203
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.68686842918396,
+      "learning_rate": 9.601126751739997e-06,
+      "loss": 0.6317,
+      "step": 1204
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0472536087036133,
+      "learning_rate": 9.600314212455125e-06,
+      "loss": 0.6819,
+      "step": 1205
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.0865917205810547,
+      "learning_rate": 9.599500880859632e-06,
+      "loss": 0.572,
+      "step": 1206
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.095560073852539,
+      "learning_rate": 9.5986867570936e-06,
+      "loss": 0.6667,
+      "step": 1207
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0914700031280518,
+      "learning_rate": 9.597871841297246e-06,
+      "loss": 0.6668,
+      "step": 1208
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1000221967697144,
+      "learning_rate": 9.597056133610923e-06,
+      "loss": 0.6105,
+      "step": 1209
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5570249557495117,
+      "learning_rate": 9.596239634175119e-06,
+      "loss": 0.6424,
+      "step": 1210
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3458069562911987,
+      "learning_rate": 9.595422343130462e-06,
+      "loss": 0.6504,
+      "step": 1211
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.21104896068573,
+      "learning_rate": 9.59460426061771e-06,
+      "loss": 0.6194,
+      "step": 1212
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3381779193878174,
+      "learning_rate": 9.593785386777763e-06,
+      "loss": 0.5563,
+      "step": 1213
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3582426309585571,
+      "learning_rate": 9.592965721751657e-06,
+      "loss": 0.5941,
+      "step": 1214
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5648891925811768,
+      "learning_rate": 9.59214526568056e-06,
+      "loss": 0.5831,
+      "step": 1215
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3309880495071411,
+      "learning_rate": 9.591324018705779e-06,
+      "loss": 0.5284,
+      "step": 1216
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4098135232925415,
+      "learning_rate": 9.590501980968758e-06,
+      "loss": 0.7147,
+      "step": 1217
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2456073760986328,
+      "learning_rate": 9.589679152611078e-06,
+      "loss": 0.5923,
+      "step": 1218
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2749463319778442,
+      "learning_rate": 9.58885553377445e-06,
+      "loss": 0.6064,
+      "step": 1219
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.382809042930603,
+      "learning_rate": 9.58803112460073e-06,
+      "loss": 0.6162,
+      "step": 1220
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5532960891723633,
+      "learning_rate": 9.587205925231904e-06,
+      "loss": 0.6222,
+      "step": 1221
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1475423574447632,
+      "learning_rate": 9.586379935810096e-06,
+      "loss": 0.5918,
+      "step": 1222
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1791549921035767,
+      "learning_rate": 9.585553156477565e-06,
+      "loss": 0.5603,
+      "step": 1223
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2540336847305298,
+      "learning_rate": 9.584725587376708e-06,
+      "loss": 0.6537,
+      "step": 1224
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3627375364303589,
+      "learning_rate": 9.583897228650056e-06,
+      "loss": 0.6247,
+      "step": 1225
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.172321081161499,
+      "learning_rate": 9.583068080440278e-06,
+      "loss": 0.7451,
+      "step": 1226
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1737582683563232,
+      "learning_rate": 9.582238142890179e-06,
+      "loss": 0.6437,
+      "step": 1227
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0629295110702515,
+      "learning_rate": 9.581407416142696e-06,
+      "loss": 0.6622,
+      "step": 1228
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3306161165237427,
+      "learning_rate": 9.580575900340904e-06,
+      "loss": 0.7027,
+      "step": 1229
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.119742512702942,
+      "learning_rate": 9.57974359562802e-06,
+      "loss": 0.6573,
+      "step": 1230
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1853952407836914,
+      "learning_rate": 9.578910502147387e-06,
+      "loss": 0.7601,
+      "step": 1231
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.302206516265869,
+      "learning_rate": 9.578076620042492e-06,
+      "loss": 0.616,
+      "step": 1232
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1335538625717163,
+      "learning_rate": 9.57724194945695e-06,
+      "loss": 0.5944,
+      "step": 1233
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0384563207626343,
+      "learning_rate": 9.576406490534518e-06,
+      "loss": 0.6307,
+      "step": 1234
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1229547262191772,
+      "learning_rate": 9.575570243419087e-06,
+      "loss": 0.585,
+      "step": 1235
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.213402271270752,
+      "learning_rate": 9.574733208254684e-06,
+      "loss": 0.5698,
+      "step": 1236
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1010196208953857,
+      "learning_rate": 9.57389538518547e-06,
+      "loss": 0.593,
+      "step": 1237
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.357301115989685,
+      "learning_rate": 9.573056774355745e-06,
+      "loss": 0.6061,
+      "step": 1238
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.095462679862976,
+      "learning_rate": 9.57221737590994e-06,
+      "loss": 0.6389,
+      "step": 1239
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4295891523361206,
+      "learning_rate": 9.571377189992627e-06,
+      "loss": 0.6627,
+      "step": 1240
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.268510341644287,
+      "learning_rate": 9.570536216748509e-06,
+      "loss": 0.6064,
+      "step": 1241
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1977204084396362,
+      "learning_rate": 9.569694456322428e-06,
+      "loss": 0.6375,
+      "step": 1242
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0961164236068726,
+      "learning_rate": 9.56885190885936e-06,
+      "loss": 0.6977,
+      "step": 1243
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.396316409111023,
+      "learning_rate": 9.568008574504415e-06,
+      "loss": 0.6978,
+      "step": 1244
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.1277964115142822,
+      "learning_rate": 9.56716445340284e-06,
+      "loss": 0.5666,
+      "step": 1245
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1821579933166504,
+      "learning_rate": 9.566319545700021e-06,
+      "loss": 0.6249,
+      "step": 1246
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.187343955039978,
+      "learning_rate": 9.565473851541473e-06,
+      "loss": 0.6518,
+      "step": 1247
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.382359504699707,
+      "learning_rate": 9.564627371072853e-06,
+      "loss": 0.6206,
+      "step": 1248
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0749714374542236,
+      "learning_rate": 9.563780104439945e-06,
+      "loss": 0.6691,
+      "step": 1249
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6738566160202026,
+      "learning_rate": 9.562932051788677e-06,
+      "loss": 0.6744,
+      "step": 1250
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1276246309280396,
+      "learning_rate": 9.56208321326511e-06,
+      "loss": 0.6447,
+      "step": 1251
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1615641117095947,
+      "learning_rate": 9.561233589015435e-06,
+      "loss": 0.4971,
+      "step": 1252
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.220762848854065,
+      "learning_rate": 9.560383179185986e-06,
+      "loss": 0.6402,
+      "step": 1253
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1979705095291138,
+      "learning_rate": 9.559531983923226e-06,
+      "loss": 0.6028,
+      "step": 1254
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3428279161453247,
+      "learning_rate": 9.558680003373762e-06,
+      "loss": 0.65,
+      "step": 1255
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0979018211364746,
+      "learning_rate": 9.557827237684324e-06,
+      "loss": 0.582,
+      "step": 1256
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1736618280410767,
+      "learning_rate": 9.556973687001782e-06,
+      "loss": 0.6034,
+      "step": 1257
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.021039366722107,
+      "learning_rate": 9.55611935147315e-06,
+      "loss": 0.5749,
+      "step": 1258
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1807692050933838,
+      "learning_rate": 9.555264231245566e-06,
+      "loss": 0.652,
+      "step": 1259
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9832674264907837,
+      "learning_rate": 9.554408326466306e-06,
+      "loss": 0.6777,
+      "step": 1260
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0194718837738037,
+      "learning_rate": 9.553551637282785e-06,
+      "loss": 0.6897,
+      "step": 1261
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3178449869155884,
+      "learning_rate": 9.552694163842548e-06,
+      "loss": 0.6381,
+      "step": 1262
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1459720134735107,
+      "learning_rate": 9.55183590629328e-06,
+      "loss": 0.6509,
+      "step": 1263
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.71978759765625,
+      "learning_rate": 9.550976864782792e-06,
+      "loss": 0.6781,
+      "step": 1264
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1249512434005737,
+      "learning_rate": 9.550117039459045e-06,
+      "loss": 0.5807,
+      "step": 1265
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0362927913665771,
+      "learning_rate": 9.549256430470123e-06,
+      "loss": 0.6622,
+      "step": 1266
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.005173683166504,
+      "learning_rate": 9.548395037964247e-06,
+      "loss": 0.6316,
+      "step": 1267
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2679580450057983,
+      "learning_rate": 9.547532862089775e-06,
+      "loss": 0.5937,
+      "step": 1268
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0311354398727417,
+      "learning_rate": 9.5466699029952e-06,
+      "loss": 0.5042,
+      "step": 1269
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9762603044509888,
+      "learning_rate": 9.545806160829147e-06,
+      "loss": 0.6358,
+      "step": 1270
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9036137461662292,
+      "learning_rate": 9.544941635740382e-06,
+      "loss": 0.6065,
+      "step": 1271
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0758832693099976,
+      "learning_rate": 9.544076327877799e-06,
+      "loss": 0.5666,
+      "step": 1272
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2641098499298096,
+      "learning_rate": 9.543210237390429e-06,
+      "loss": 0.6395,
+      "step": 1273
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3716449737548828,
+      "learning_rate": 9.54234336442744e-06,
+      "loss": 0.7927,
+      "step": 1274
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1750115156173706,
+      "learning_rate": 9.541475709138135e-06,
+      "loss": 0.5781,
+      "step": 1275
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.270869493484497,
+      "learning_rate": 9.540607271671947e-06,
+      "loss": 0.7835,
+      "step": 1276
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7524586915969849,
+      "learning_rate": 9.539738052178446e-06,
+      "loss": 0.6404,
+      "step": 1277
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.035844326019287,
+      "learning_rate": 9.538868050807342e-06,
+      "loss": 0.6465,
+      "step": 1278
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.166962742805481,
+      "learning_rate": 9.53799726770847e-06,
+      "loss": 0.5387,
+      "step": 1279
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.160826563835144,
+      "learning_rate": 9.537125703031809e-06,
+      "loss": 0.6005,
+      "step": 1280
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0476465225219727,
+      "learning_rate": 9.536253356927465e-06,
+      "loss": 0.6716,
+      "step": 1281
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0793278217315674,
+      "learning_rate": 9.535380229545684e-06,
+      "loss": 0.7254,
+      "step": 1282
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1872631311416626,
+      "learning_rate": 9.534506321036842e-06,
+      "loss": 0.6274,
+      "step": 1283
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5636075735092163,
+      "learning_rate": 9.533631631551455e-06,
+      "loss": 0.6149,
+      "step": 1284
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2059067487716675,
+      "learning_rate": 9.532756161240166e-06,
+      "loss": 0.6378,
+      "step": 1285
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.154067039489746,
+      "learning_rate": 9.531879910253762e-06,
+      "loss": 0.7198,
+      "step": 1286
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2851213216781616,
+      "learning_rate": 9.531002878743158e-06,
+      "loss": 0.6087,
+      "step": 1287
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1828138828277588,
+      "learning_rate": 9.530125066859404e-06,
+      "loss": 0.7095,
+      "step": 1288
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.218656063079834,
+      "learning_rate": 9.529246474753684e-06,
+      "loss": 0.6411,
+      "step": 1289
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1605652570724487,
+      "learning_rate": 9.528367102577321e-06,
+      "loss": 0.6413,
+      "step": 1290
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1710295677185059,
+      "learning_rate": 9.527486950481765e-06,
+      "loss": 0.6295,
+      "step": 1291
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.972027063369751,
+      "learning_rate": 9.526606018618608e-06,
+      "loss": 0.5839,
+      "step": 1292
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.12654709815979,
+      "learning_rate": 9.52572430713957e-06,
+      "loss": 0.5989,
+      "step": 1293
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2400918006896973,
+      "learning_rate": 9.52484181619651e-06,
+      "loss": 0.5958,
+      "step": 1294
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2699034214019775,
+      "learning_rate": 9.523958545941417e-06,
+      "loss": 0.6808,
+      "step": 1295
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1967825889587402,
+      "learning_rate": 9.523074496526418e-06,
+      "loss": 0.6116,
+      "step": 1296
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2007198333740234,
+      "learning_rate": 9.522189668103771e-06,
+      "loss": 0.564,
+      "step": 1297
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3140076398849487,
+      "learning_rate": 9.521304060825872e-06,
+      "loss": 0.6417,
+      "step": 1298
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6398093700408936,
+      "learning_rate": 9.520417674845246e-06,
+      "loss": 0.6366,
+      "step": 1299
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0856890678405762,
+      "learning_rate": 9.519530510314558e-06,
+      "loss": 0.6931,
+      "step": 1300
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.219240665435791,
+      "learning_rate": 9.518642567386603e-06,
+      "loss": 0.6608,
+      "step": 1301
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2921411991119385,
+      "learning_rate": 9.517753846214309e-06,
+      "loss": 0.6584,
+      "step": 1302
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0977832078933716,
+      "learning_rate": 9.51686434695074e-06,
+      "loss": 0.6219,
+      "step": 1303
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0890233516693115,
+      "learning_rate": 9.515974069749098e-06,
+      "loss": 0.6462,
+      "step": 1304
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3405479192733765,
+      "learning_rate": 9.515083014762714e-06,
+      "loss": 0.5702,
+      "step": 1305
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9640637636184692,
+      "learning_rate": 9.514191182145052e-06,
+      "loss": 0.6735,
+      "step": 1306
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3195781707763672,
+      "learning_rate": 9.513298572049714e-06,
+      "loss": 0.6042,
+      "step": 1307
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.054443120956421,
+      "learning_rate": 9.512405184630432e-06,
+      "loss": 0.6659,
+      "step": 1308
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.662334442138672,
+      "learning_rate": 9.511511020041076e-06,
+      "loss": 0.5981,
+      "step": 1309
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3885622024536133,
+      "learning_rate": 9.510616078435646e-06,
+      "loss": 0.7479,
+      "step": 1310
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1111118793487549,
+      "learning_rate": 9.509720359968279e-06,
+      "loss": 0.6262,
+      "step": 1311
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2182552814483643,
+      "learning_rate": 9.508823864793242e-06,
+      "loss": 0.6059,
+      "step": 1312
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1193416118621826,
+      "learning_rate": 9.507926593064941e-06,
+      "loss": 0.6478,
+      "step": 1313
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3522484302520752,
+      "learning_rate": 9.50702854493791e-06,
+      "loss": 0.6145,
+      "step": 1314
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.071323275566101,
+      "learning_rate": 9.506129720566823e-06,
+      "loss": 0.6033,
+      "step": 1315
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.270236611366272,
+      "learning_rate": 9.50523012010648e-06,
+      "loss": 0.5783,
+      "step": 1316
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0845481157302856,
+      "learning_rate": 9.504329743711822e-06,
+      "loss": 0.6448,
+      "step": 1317
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.229878544807434,
+      "learning_rate": 9.50342859153792e-06,
+      "loss": 0.6432,
+      "step": 1318
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2673510313034058,
+      "learning_rate": 9.50252666373998e-06,
+      "loss": 0.6371,
+      "step": 1319
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5885368585586548,
+      "learning_rate": 9.501623960473336e-06,
+      "loss": 0.669,
+      "step": 1320
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.237269401550293,
+      "learning_rate": 9.500720481893466e-06,
+      "loss": 0.6404,
+      "step": 1321
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3405085802078247,
+      "learning_rate": 9.499816228155972e-06,
+      "loss": 0.625,
+      "step": 1322
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9338662624359131,
+      "learning_rate": 9.498911199416596e-06,
+      "loss": 0.5723,
+      "step": 1323
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9581477642059326,
+      "learning_rate": 9.49800539583121e-06,
+      "loss": 0.5658,
+      "step": 1324
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2048728466033936,
+      "learning_rate": 9.497098817555818e-06,
+      "loss": 0.628,
+      "step": 1325
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.456356406211853,
+      "learning_rate": 9.496191464746564e-06,
+      "loss": 0.6503,
+      "step": 1326
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1231420040130615,
+      "learning_rate": 9.495283337559718e-06,
+      "loss": 0.5696,
+      "step": 1327
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0166943073272705,
+      "learning_rate": 9.494374436151688e-06,
+      "loss": 0.63,
+      "step": 1328
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5634657144546509,
+      "learning_rate": 9.493464760679012e-06,
+      "loss": 0.6734,
+      "step": 1329
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0731712579727173,
+      "learning_rate": 9.492554311298363e-06,
+      "loss": 0.6909,
+      "step": 1330
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0333720445632935,
+      "learning_rate": 9.49164308816655e-06,
+      "loss": 0.6646,
+      "step": 1331
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9412813186645508,
+      "learning_rate": 9.49073109144051e-06,
+      "loss": 0.6295,
+      "step": 1332
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3674019575119019,
+      "learning_rate": 9.489818321277318e-06,
+      "loss": 0.6832,
+      "step": 1333
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2104706764221191,
+      "learning_rate": 9.488904777834178e-06,
+      "loss": 0.6606,
+      "step": 1334
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2919623851776123,
+      "learning_rate": 9.487990461268431e-06,
+      "loss": 0.6844,
+      "step": 1335
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6904457807540894,
+      "learning_rate": 9.48707537173755e-06,
+      "loss": 0.6159,
+      "step": 1336
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9881489872932434,
+      "learning_rate": 9.486159509399138e-06,
+      "loss": 0.5704,
+      "step": 1337
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1692382097244263,
+      "learning_rate": 9.485242874410936e-06,
+      "loss": 0.6502,
+      "step": 1338
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3247543573379517,
+      "learning_rate": 9.484325466930814e-06,
+      "loss": 0.6244,
+      "step": 1339
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2117483615875244,
+      "learning_rate": 9.483407287116777e-06,
+      "loss": 0.5659,
+      "step": 1340
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9740604162216187,
+      "learning_rate": 9.482488335126964e-06,
+      "loss": 0.6486,
+      "step": 1341
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1840041875839233,
+      "learning_rate": 9.481568611119647e-06,
+      "loss": 0.7439,
+      "step": 1342
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2369718551635742,
+      "learning_rate": 9.480648115253225e-06,
+      "loss": 0.5795,
+      "step": 1343
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4914673566818237,
+      "learning_rate": 9.479726847686241e-06,
+      "loss": 0.5856,
+      "step": 1344
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.54400634765625,
+      "learning_rate": 9.478804808577359e-06,
+      "loss": 0.6174,
+      "step": 1345
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0934762954711914,
+      "learning_rate": 9.477881998085386e-06,
+      "loss": 0.6775,
+      "step": 1346
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1581801176071167,
+      "learning_rate": 9.476958416369253e-06,
+      "loss": 0.619,
+      "step": 1347
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2244946956634521,
+      "learning_rate": 9.47603406358803e-06,
+      "loss": 0.5794,
+      "step": 1348
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1591740846633911,
+      "learning_rate": 9.47510893990092e-06,
+      "loss": 0.6477,
+      "step": 1349
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1715471744537354,
+      "learning_rate": 9.474183045467255e-06,
+      "loss": 0.6397,
+      "step": 1350
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2296302318572998,
+      "learning_rate": 9.473256380446501e-06,
+      "loss": 0.7369,
+      "step": 1351
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7023544311523438,
+      "learning_rate": 9.472328944998256e-06,
+      "loss": 0.53,
+      "step": 1352
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0957367420196533,
+      "learning_rate": 9.471400739282258e-06,
+      "loss": 0.5876,
+      "step": 1353
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7862164974212646,
+      "learning_rate": 9.470471763458364e-06,
+      "loss": 0.666,
+      "step": 1354
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7836625576019287,
+      "learning_rate": 9.469542017686574e-06,
+      "loss": 0.6425,
+      "step": 1355
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.832993984222412,
+      "learning_rate": 9.468611502127021e-06,
+      "loss": 0.5768,
+      "step": 1356
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1718406677246094,
+      "learning_rate": 9.467680216939964e-06,
+      "loss": 0.5895,
+      "step": 1357
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.15514075756073,
+      "learning_rate": 9.466748162285797e-06,
+      "loss": 0.5945,
+      "step": 1358
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2014521360397339,
+      "learning_rate": 9.46581533832505e-06,
+      "loss": 0.6036,
+      "step": 1359
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.222567081451416,
+      "learning_rate": 9.464881745218382e-06,
+      "loss": 0.6194,
+      "step": 1360
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3445652723312378,
+      "learning_rate": 9.463947383126586e-06,
+      "loss": 0.6184,
+      "step": 1361
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1077783107757568,
+      "learning_rate": 9.463012252210586e-06,
+      "loss": 0.6802,
+      "step": 1362
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.173924446105957,
+      "learning_rate": 9.46207635263144e-06,
+      "loss": 0.6404,
+      "step": 1363
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.214298129081726,
+      "learning_rate": 9.461139684550335e-06,
+      "loss": 0.6512,
+      "step": 1364
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7848788499832153,
+      "learning_rate": 9.460202248128598e-06,
+      "loss": 0.5918,
+      "step": 1365
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0078115463256836,
+      "learning_rate": 9.45926404352768e-06,
+      "loss": 0.517,
+      "step": 1366
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.179234504699707,
+      "learning_rate": 9.458325070909169e-06,
+      "loss": 0.7112,
+      "step": 1367
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8004214763641357,
+      "learning_rate": 9.457385330434782e-06,
+      "loss": 0.6046,
+      "step": 1368
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1646499633789062,
+      "learning_rate": 9.456444822266373e-06,
+      "loss": 0.5833,
+      "step": 1369
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1174657344818115,
+      "learning_rate": 9.455503546565923e-06,
+      "loss": 0.6349,
+      "step": 1370
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5000091791152954,
+      "learning_rate": 9.45456150349555e-06,
+      "loss": 0.6107,
+      "step": 1371
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1921900510787964,
+      "learning_rate": 9.453618693217498e-06,
+      "loss": 0.6262,
+      "step": 1372
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1785988807678223,
+      "learning_rate": 9.452675115894151e-06,
+      "loss": 0.6661,
+      "step": 1373
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1983801126480103,
+      "learning_rate": 9.45173077168802e-06,
+      "loss": 0.6726,
+      "step": 1374
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2910903692245483,
+      "learning_rate": 9.450785660761747e-06,
+      "loss": 0.6717,
+      "step": 1375
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1190389394760132,
+      "learning_rate": 9.44983978327811e-06,
+      "loss": 0.6953,
+      "step": 1376
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6207910776138306,
+      "learning_rate": 9.448893139400016e-06,
+      "loss": 0.6883,
+      "step": 1377
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0816600322723389,
+      "learning_rate": 9.447945729290507e-06,
+      "loss": 0.6447,
+      "step": 1378
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.4702744483947754,
+      "learning_rate": 9.446997553112753e-06,
+      "loss": 0.5832,
+      "step": 1379
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8984103202819824,
+      "learning_rate": 9.446048611030061e-06,
+      "loss": 0.6695,
+      "step": 1380
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.11966073513031,
+      "learning_rate": 9.445098903205863e-06,
+      "loss": 0.6151,
+      "step": 1381
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.413903832435608,
+      "learning_rate": 9.44414842980373e-06,
+      "loss": 0.5619,
+      "step": 1382
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.174336314201355,
+      "learning_rate": 9.443197190987359e-06,
+      "loss": 0.7763,
+      "step": 1383
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1478570699691772,
+      "learning_rate": 9.442245186920585e-06,
+      "loss": 0.5953,
+      "step": 1384
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0508737564086914,
+      "learning_rate": 9.44129241776737e-06,
+      "loss": 0.6232,
+      "step": 1385
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4047367572784424,
+      "learning_rate": 9.440338883691807e-06,
+      "loss": 0.611,
+      "step": 1386
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1691255569458008,
+      "learning_rate": 9.439384584858125e-06,
+      "loss": 0.6286,
+      "step": 1387
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2848888635635376,
+      "learning_rate": 9.43842952143068e-06,
+      "loss": 0.6801,
+      "step": 1388
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1584075689315796,
+      "learning_rate": 9.437473693573969e-06,
+      "loss": 0.6356,
+      "step": 1389
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8564170598983765,
+      "learning_rate": 9.436517101452607e-06,
+      "loss": 0.6309,
+      "step": 1390
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3005411624908447,
+      "learning_rate": 9.435559745231348e-06,
+      "loss": 0.6847,
+      "step": 1391
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.392625331878662,
+      "learning_rate": 9.434601625075082e-06,
+      "loss": 0.5639,
+      "step": 1392
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3917368650436401,
+      "learning_rate": 9.43364274114882e-06,
+      "loss": 0.7253,
+      "step": 1393
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6492226123809814,
+      "learning_rate": 9.432683093617716e-06,
+      "loss": 0.6228,
+      "step": 1394
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.330742359161377,
+      "learning_rate": 9.431722682647044e-06,
+      "loss": 0.7072,
+      "step": 1395
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6217615604400635,
+      "learning_rate": 9.43076150840222e-06,
+      "loss": 0.5747,
+      "step": 1396
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2197303771972656,
+      "learning_rate": 9.429799571048784e-06,
+      "loss": 0.6557,
+      "step": 1397
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0048884153366089,
+      "learning_rate": 9.428836870752411e-06,
+      "loss": 0.707,
+      "step": 1398
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2903577089309692,
+      "learning_rate": 9.427873407678907e-06,
+      "loss": 0.7498,
+      "step": 1399
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.230726718902588,
+      "learning_rate": 9.426909181994208e-06,
+      "loss": 0.6526,
+      "step": 1400
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1689950227737427,
+      "learning_rate": 9.425944193864382e-06,
+      "loss": 0.6607,
+      "step": 1401
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2050648927688599,
+      "learning_rate": 9.42497844345563e-06,
+      "loss": 0.7587,
+      "step": 1402
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1159378290176392,
+      "learning_rate": 9.424011930934283e-06,
+      "loss": 0.6073,
+      "step": 1403
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.159040927886963,
+      "learning_rate": 9.4230446564668e-06,
+      "loss": 0.6005,
+      "step": 1404
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3513133525848389,
+      "learning_rate": 9.422076620219777e-06,
+      "loss": 0.6888,
+      "step": 1405
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4324787855148315,
+      "learning_rate": 9.42110782235994e-06,
+      "loss": 0.6781,
+      "step": 1406
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0883930921554565,
+      "learning_rate": 9.420138263054143e-06,
+      "loss": 0.6243,
+      "step": 1407
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2407677173614502,
+      "learning_rate": 9.419167942469372e-06,
+      "loss": 0.6807,
+      "step": 1408
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.275864839553833,
+      "learning_rate": 9.418196860772746e-06,
+      "loss": 0.6301,
+      "step": 1409
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.2127833366394043,
+      "learning_rate": 9.417225018131513e-06,
+      "loss": 0.6004,
+      "step": 1410
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.683578610420227,
+      "learning_rate": 9.416252414713056e-06,
+      "loss": 0.5999,
+      "step": 1411
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1180976629257202,
+      "learning_rate": 9.415279050684882e-06,
+      "loss": 0.6629,
+      "step": 1412
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.122758388519287,
+      "learning_rate": 9.414304926214637e-06,
+      "loss": 0.582,
+      "step": 1413
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2862956523895264,
+      "learning_rate": 9.413330041470092e-06,
+      "loss": 0.5903,
+      "step": 1414
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.222602367401123,
+      "learning_rate": 9.412354396619151e-06,
+      "loss": 0.6605,
+      "step": 1415
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1338920593261719,
+      "learning_rate": 9.411377991829851e-06,
+      "loss": 0.5989,
+      "step": 1416
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2186603546142578,
+      "learning_rate": 9.410400827270356e-06,
+      "loss": 0.669,
+      "step": 1417
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.038238763809204,
+      "learning_rate": 9.409422903108963e-06,
+      "loss": 0.6425,
+      "step": 1418
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0504441261291504,
+      "learning_rate": 9.4084442195141e-06,
+      "loss": 0.5588,
+      "step": 1419
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3752822875976562,
+      "learning_rate": 9.407464776654326e-06,
+      "loss": 0.5757,
+      "step": 1420
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1799776554107666,
+      "learning_rate": 9.406484574698328e-06,
+      "loss": 0.7499,
+      "step": 1421
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.192615270614624,
+      "learning_rate": 9.405503613814927e-06,
+      "loss": 0.4951,
+      "step": 1422
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.10984206199646,
+      "learning_rate": 9.404521894173075e-06,
+      "loss": 0.5502,
+      "step": 1423
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2776201963424683,
+      "learning_rate": 9.403539415941852e-06,
+      "loss": 0.7015,
+      "step": 1424
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2026793956756592,
+      "learning_rate": 9.402556179290468e-06,
+      "loss": 0.6019,
+      "step": 1425
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2137603759765625,
+      "learning_rate": 9.401572184388268e-06,
+      "loss": 0.6522,
+      "step": 1426
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.404771327972412,
+      "learning_rate": 9.400587431404726e-06,
+      "loss": 0.6633,
+      "step": 1427
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0881354808807373,
+      "learning_rate": 9.399601920509442e-06,
+      "loss": 0.6346,
+      "step": 1428
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1489696502685547,
+      "learning_rate": 9.398615651872154e-06,
+      "loss": 0.5722,
+      "step": 1429
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.202345609664917,
+      "learning_rate": 9.397628625662724e-06,
+      "loss": 0.6491,
+      "step": 1430
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0614575147628784,
+      "learning_rate": 9.39664084205115e-06,
+      "loss": 0.6658,
+      "step": 1431
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1777621507644653,
+      "learning_rate": 9.395652301207556e-06,
+      "loss": 0.6096,
+      "step": 1432
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9109401106834412,
+      "learning_rate": 9.394663003302197e-06,
+      "loss": 0.5908,
+      "step": 1433
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0313465595245361,
+      "learning_rate": 9.393672948505461e-06,
+      "loss": 0.6819,
+      "step": 1434
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7536683082580566,
+      "learning_rate": 9.392682136987865e-06,
+      "loss": 0.6045,
+      "step": 1435
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0943320989608765,
+      "learning_rate": 9.391690568920055e-06,
+      "loss": 0.5611,
+      "step": 1436
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.105167031288147,
+      "learning_rate": 9.390698244472808e-06,
+      "loss": 0.6414,
+      "step": 1437
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1738401651382446,
+      "learning_rate": 9.389705163817034e-06,
+      "loss": 0.7103,
+      "step": 1438
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1961818933486938,
+      "learning_rate": 9.388711327123769e-06,
+      "loss": 0.6541,
+      "step": 1439
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2521352767944336,
+      "learning_rate": 9.38771673456418e-06,
+      "loss": 0.6377,
+      "step": 1440
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9367631673812866,
+      "learning_rate": 9.386721386309569e-06,
+      "loss": 0.6122,
+      "step": 1441
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1017770767211914,
+      "learning_rate": 9.385725282531364e-06,
+      "loss": 0.5612,
+      "step": 1442
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.280482530593872,
+      "learning_rate": 9.38472842340112e-06,
+      "loss": 0.672,
+      "step": 1443
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5452642440795898,
+      "learning_rate": 9.383730809090528e-06,
+      "loss": 0.6426,
+      "step": 1444
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0411386489868164,
+      "learning_rate": 9.382732439771409e-06,
+      "loss": 0.5851,
+      "step": 1445
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0093711614608765,
+      "learning_rate": 9.381733315615708e-06,
+      "loss": 0.662,
+      "step": 1446
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6546435356140137,
+      "learning_rate": 9.380733436795506e-06,
+      "loss": 0.5422,
+      "step": 1447
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1968332529067993,
+      "learning_rate": 9.379732803483011e-06,
+      "loss": 0.6345,
+      "step": 1448
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.473617672920227,
+      "learning_rate": 9.378731415850561e-06,
+      "loss": 0.5548,
+      "step": 1449
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1579792499542236,
+      "learning_rate": 9.377729274070627e-06,
+      "loss": 0.6089,
+      "step": 1450
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1918591260910034,
+      "learning_rate": 9.376726378315806e-06,
+      "loss": 0.5618,
+      "step": 1451
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.340204119682312,
+      "learning_rate": 9.375722728758826e-06,
+      "loss": 0.5787,
+      "step": 1452
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.30586576461792,
+      "learning_rate": 9.374718325572547e-06,
+      "loss": 0.683,
+      "step": 1453
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.144736409187317,
+      "learning_rate": 9.373713168929954e-06,
+      "loss": 0.6424,
+      "step": 1454
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.587199091911316,
+      "learning_rate": 9.372707259004168e-06,
+      "loss": 0.6204,
+      "step": 1455
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1502057313919067,
+      "learning_rate": 9.371700595968437e-06,
+      "loss": 0.6197,
+      "step": 1456
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1308425664901733,
+      "learning_rate": 9.370693179996133e-06,
+      "loss": 0.5867,
+      "step": 1457
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8421180248260498,
+      "learning_rate": 9.369685011260768e-06,
+      "loss": 0.6537,
+      "step": 1458
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.224798560142517,
+      "learning_rate": 9.368676089935978e-06,
+      "loss": 0.6643,
+      "step": 1459
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2730331420898438,
+      "learning_rate": 9.367666416195526e-06,
+      "loss": 0.6882,
+      "step": 1460
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4292099475860596,
+      "learning_rate": 9.366655990213311e-06,
+      "loss": 0.6895,
+      "step": 1461
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8112746477127075,
+      "learning_rate": 9.365644812163356e-06,
+      "loss": 0.7695,
+      "step": 1462
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.171579360961914,
+      "learning_rate": 9.36463288221982e-06,
+      "loss": 0.648,
+      "step": 1463
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1482553482055664,
+      "learning_rate": 9.363620200556983e-06,
+      "loss": 0.6616,
+      "step": 1464
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1553841829299927,
+      "learning_rate": 9.36260676734926e-06,
+      "loss": 0.5983,
+      "step": 1465
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.32172691822052,
+      "learning_rate": 9.361592582771195e-06,
+      "loss": 0.5921,
+      "step": 1466
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9891936779022217,
+      "learning_rate": 9.36057764699746e-06,
+      "loss": 0.6451,
+      "step": 1467
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3381367921829224,
+      "learning_rate": 9.359561960202857e-06,
+      "loss": 0.6476,
+      "step": 1468
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.140139102935791,
+      "learning_rate": 9.35854552256232e-06,
+      "loss": 0.6169,
+      "step": 1469
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0977221727371216,
+      "learning_rate": 9.357528334250905e-06,
+      "loss": 0.5992,
+      "step": 1470
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3049180507659912,
+      "learning_rate": 9.356510395443804e-06,
+      "loss": 0.5462,
+      "step": 1471
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2993820905685425,
+      "learning_rate": 9.35549170631634e-06,
+      "loss": 0.6289,
+      "step": 1472
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1152647733688354,
+      "learning_rate": 9.354472267043955e-06,
+      "loss": 0.6773,
+      "step": 1473
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.593841552734375,
+      "learning_rate": 9.353452077802233e-06,
+      "loss": 0.6024,
+      "step": 1474
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0295051336288452,
+      "learning_rate": 9.352431138766875e-06,
+      "loss": 0.556,
+      "step": 1475
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2775508165359497,
+      "learning_rate": 9.35140945011372e-06,
+      "loss": 0.6177,
+      "step": 1476
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1952461004257202,
+      "learning_rate": 9.350387012018734e-06,
+      "loss": 0.6117,
+      "step": 1477
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.298227071762085,
+      "learning_rate": 9.34936382465801e-06,
+      "loss": 0.6345,
+      "step": 1478
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2173748016357422,
+      "learning_rate": 9.348339888207771e-06,
+      "loss": 0.6348,
+      "step": 1479
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5317448377609253,
+      "learning_rate": 9.347315202844371e-06,
+      "loss": 0.6326,
+      "step": 1480
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0272226333618164,
+      "learning_rate": 9.346289768744288e-06,
+      "loss": 0.7068,
+      "step": 1481
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.982128381729126,
+      "learning_rate": 9.345263586084135e-06,
+      "loss": 0.5463,
+      "step": 1482
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9588432908058167,
+      "learning_rate": 9.344236655040649e-06,
+      "loss": 0.6088,
+      "step": 1483
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1015230417251587,
+      "learning_rate": 9.343208975790699e-06,
+      "loss": 0.6073,
+      "step": 1484
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1615246534347534,
+      "learning_rate": 9.342180548511283e-06,
+      "loss": 0.5965,
+      "step": 1485
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2907776832580566,
+      "learning_rate": 9.341151373379527e-06,
+      "loss": 0.6704,
+      "step": 1486
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2365469932556152,
+      "learning_rate": 9.340121450572681e-06,
+      "loss": 0.732,
+      "step": 1487
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2405096292495728,
+      "learning_rate": 9.339090780268133e-06,
+      "loss": 0.5716,
+      "step": 1488
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2856647968292236,
+      "learning_rate": 9.338059362643393e-06,
+      "loss": 0.6443,
+      "step": 1489
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0891637802124023,
+      "learning_rate": 9.337027197876103e-06,
+      "loss": 0.6259,
+      "step": 1490
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1378039121627808,
+      "learning_rate": 9.33599428614403e-06,
+      "loss": 0.6413,
+      "step": 1491
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0595463514328003,
+      "learning_rate": 9.334960627625075e-06,
+      "loss": 0.593,
+      "step": 1492
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.307681918144226,
+      "learning_rate": 9.333926222497263e-06,
+      "loss": 0.5555,
+      "step": 1493
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.9554941654205322,
+      "learning_rate": 9.332891070938749e-06,
+      "loss": 0.6792,
+      "step": 1494
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0907893180847168,
+      "learning_rate": 9.331855173127817e-06,
+      "loss": 0.6127,
+      "step": 1495
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8585306406021118,
+      "learning_rate": 9.33081852924288e-06,
+      "loss": 0.5657,
+      "step": 1496
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3523632287979126,
+      "learning_rate": 9.329781139462479e-06,
+      "loss": 0.6613,
+      "step": 1497
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0348542928695679,
+      "learning_rate": 9.328743003965283e-06,
+      "loss": 0.5763,
+      "step": 1498
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.102908968925476,
+      "learning_rate": 9.32770412293009e-06,
+      "loss": 0.6235,
+      "step": 1499
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.521458148956299,
+      "learning_rate": 9.326664496535825e-06,
+      "loss": 0.7002,
+      "step": 1500
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6622076034545898,
+      "learning_rate": 9.325624124961542e-06,
+      "loss": 0.587,
+      "step": 1501
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0094268321990967,
+      "learning_rate": 9.324583008386425e-06,
+      "loss": 0.6609,
+      "step": 1502
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.205629825592041,
+      "learning_rate": 9.323541146989788e-06,
+      "loss": 0.5841,
+      "step": 1503
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3504383563995361,
+      "learning_rate": 9.322498540951067e-06,
+      "loss": 0.7027,
+      "step": 1504
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.421724557876587,
+      "learning_rate": 9.321455190449828e-06,
+      "loss": 0.6843,
+      "step": 1505
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1611723899841309,
+      "learning_rate": 9.32041109566577e-06,
+      "loss": 0.6121,
+      "step": 1506
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.239842176437378,
+      "learning_rate": 9.319366256778717e-06,
+      "loss": 0.7677,
+      "step": 1507
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2777602672576904,
+      "learning_rate": 9.318320673968622e-06,
+      "loss": 0.7695,
+      "step": 1508
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2827951908111572,
+      "learning_rate": 9.317274347415561e-06,
+      "loss": 0.6129,
+      "step": 1509
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.05486261844635,
+      "learning_rate": 9.316227277299748e-06,
+      "loss": 0.6093,
+      "step": 1510
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1314477920532227,
+      "learning_rate": 9.315179463801518e-06,
+      "loss": 0.6235,
+      "step": 1511
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0239319801330566,
+      "learning_rate": 9.314130907101332e-06,
+      "loss": 0.5927,
+      "step": 1512
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2008589506149292,
+      "learning_rate": 9.313081607379786e-06,
+      "loss": 0.6737,
+      "step": 1513
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1876063346862793,
+      "learning_rate": 9.3120315648176e-06,
+      "loss": 0.5764,
+      "step": 1514
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.021418571472168,
+      "learning_rate": 9.310980779595623e-06,
+      "loss": 0.6656,
+      "step": 1515
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.923373818397522,
+      "learning_rate": 9.309929251894828e-06,
+      "loss": 0.6674,
+      "step": 1516
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0926281213760376,
+      "learning_rate": 9.308876981896326e-06,
+      "loss": 0.6407,
+      "step": 1517
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.034195065498352,
+      "learning_rate": 9.307823969781342e-06,
+      "loss": 0.6774,
+      "step": 1518
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0667943954467773,
+      "learning_rate": 9.30677021573124e-06,
+      "loss": 0.7005,
+      "step": 1519
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.644808292388916,
+      "learning_rate": 9.305715719927507e-06,
+      "loss": 0.6978,
+      "step": 1520
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2867639064788818,
+      "learning_rate": 9.30466048255176e-06,
+      "loss": 0.6115,
+      "step": 1521
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1789501905441284,
+      "learning_rate": 9.303604503785737e-06,
+      "loss": 0.7755,
+      "step": 1522
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2231398820877075,
+      "learning_rate": 9.302547783811312e-06,
+      "loss": 0.741,
+      "step": 1523
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3856559991836548,
+      "learning_rate": 9.301490322810487e-06,
+      "loss": 0.632,
+      "step": 1524
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1552003622055054,
+      "learning_rate": 9.300432120965384e-06,
+      "loss": 0.6781,
+      "step": 1525
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2242902517318726,
+      "learning_rate": 9.299373178458255e-06,
+      "loss": 0.6174,
+      "step": 1526
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.118219017982483,
+      "learning_rate": 9.298313495471486e-06,
+      "loss": 0.6034,
+      "step": 1527
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7142438888549805,
+      "learning_rate": 9.297253072187585e-06,
+      "loss": 0.6345,
+      "step": 1528
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1490740776062012,
+      "learning_rate": 9.296191908789186e-06,
+      "loss": 0.6571,
+      "step": 1529
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1473088264465332,
+      "learning_rate": 9.295130005459055e-06,
+      "loss": 0.5983,
+      "step": 1530
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0597072839736938,
+      "learning_rate": 9.294067362380081e-06,
+      "loss": 0.6132,
+      "step": 1531
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.013634443283081,
+      "learning_rate": 9.293003979735284e-06,
+      "loss": 0.5525,
+      "step": 1532
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.376117467880249,
+      "learning_rate": 9.291939857707812e-06,
+      "loss": 0.5656,
+      "step": 1533
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3028066158294678,
+      "learning_rate": 9.290874996480935e-06,
+      "loss": 0.6226,
+      "step": 1534
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4443479776382446,
+      "learning_rate": 9.289809396238054e-06,
+      "loss": 0.5352,
+      "step": 1535
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0546255111694336,
+      "learning_rate": 9.288743057162698e-06,
+      "loss": 0.586,
+      "step": 1536
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3574661016464233,
+      "learning_rate": 9.287675979438526e-06,
+      "loss": 0.6401,
+      "step": 1537
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0744688510894775,
+      "learning_rate": 9.286608163249314e-06,
+      "loss": 0.6846,
+      "step": 1538
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2303810119628906,
+      "learning_rate": 9.285539608778976e-06,
+      "loss": 0.6477,
+      "step": 1539
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0581923723220825,
+      "learning_rate": 9.284470316211545e-06,
+      "loss": 0.6514,
+      "step": 1540
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0219451189041138,
+      "learning_rate": 9.283400285731188e-06,
+      "loss": 0.6836,
+      "step": 1541
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6948091983795166,
+      "learning_rate": 9.282329517522196e-06,
+      "loss": 0.6725,
+      "step": 1542
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1440834999084473,
+      "learning_rate": 9.281258011768985e-06,
+      "loss": 0.5548,
+      "step": 1543
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2690272331237793,
+      "learning_rate": 9.280185768656103e-06,
+      "loss": 0.5611,
+      "step": 1544
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.544965386390686,
+      "learning_rate": 9.279112788368218e-06,
+      "loss": 0.6296,
+      "step": 1545
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2523674964904785,
+      "learning_rate": 9.278039071090135e-06,
+      "loss": 0.671,
+      "step": 1546
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0705721378326416,
+      "learning_rate": 9.276964617006772e-06,
+      "loss": 0.653,
+      "step": 1547
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2598583698272705,
+      "learning_rate": 9.27588942630319e-06,
+      "loss": 0.6127,
+      "step": 1548
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0954736471176147,
+      "learning_rate": 9.274813499164563e-06,
+      "loss": 0.5496,
+      "step": 1549
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3450437784194946,
+      "learning_rate": 9.273736835776199e-06,
+      "loss": 0.6587,
+      "step": 1550
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1436505317687988,
+      "learning_rate": 9.272659436323535e-06,
+      "loss": 0.5947,
+      "step": 1551
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2184737920761108,
+      "learning_rate": 9.271581300992125e-06,
+      "loss": 0.6579,
+      "step": 1552
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0481288433074951,
+      "learning_rate": 9.27050242996766e-06,
+      "loss": 0.5941,
+      "step": 1553
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0766609907150269,
+      "learning_rate": 9.269422823435953e-06,
+      "loss": 0.7407,
+      "step": 1554
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0673633813858032,
+      "learning_rate": 9.268342481582944e-06,
+      "loss": 0.6136,
+      "step": 1555
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2412662506103516,
+      "learning_rate": 9.267261404594698e-06,
+      "loss": 0.6064,
+      "step": 1556
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0127815008163452,
+      "learning_rate": 9.266179592657414e-06,
+      "loss": 0.5653,
+      "step": 1557
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0416415929794312,
+      "learning_rate": 9.265097045957405e-06,
+      "loss": 0.6664,
+      "step": 1558
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0409266948699951,
+      "learning_rate": 9.264013764681123e-06,
+      "loss": 0.5291,
+      "step": 1559
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.155242919921875,
+      "learning_rate": 9.26292974901514e-06,
+      "loss": 0.6171,
+      "step": 1560
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2648565769195557,
+      "learning_rate": 9.261844999146153e-06,
+      "loss": 0.618,
+      "step": 1561
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3313794136047363,
+      "learning_rate": 9.260759515260991e-06,
+      "loss": 0.6499,
+      "step": 1562
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9920714497566223,
+      "learning_rate": 9.259673297546606e-06,
+      "loss": 0.6458,
+      "step": 1563
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.902812123298645,
+      "learning_rate": 9.258586346190077e-06,
+      "loss": 0.5976,
+      "step": 1564
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.122768521308899,
+      "learning_rate": 9.257498661378608e-06,
+      "loss": 0.6489,
+      "step": 1565
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1595557928085327,
+      "learning_rate": 9.256410243299532e-06,
+      "loss": 0.6044,
+      "step": 1566
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0857417583465576,
+      "learning_rate": 9.255321092140305e-06,
+      "loss": 0.6606,
+      "step": 1567
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2797273397445679,
+      "learning_rate": 9.254231208088514e-06,
+      "loss": 0.5624,
+      "step": 1568
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0136442184448242,
+      "learning_rate": 9.253140591331868e-06,
+      "loss": 0.6768,
+      "step": 1569
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0762922763824463,
+      "learning_rate": 9.252049242058202e-06,
+      "loss": 0.693,
+      "step": 1570
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4632151126861572,
+      "learning_rate": 9.250957160455483e-06,
+      "loss": 0.658,
+      "step": 1571
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5439205169677734,
+      "learning_rate": 9.249864346711794e-06,
+      "loss": 0.602,
+      "step": 1572
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.6240251064300537,
+      "learning_rate": 9.248770801015355e-06,
+      "loss": 0.6716,
+      "step": 1573
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2808942794799805,
+      "learning_rate": 9.247676523554503e-06,
+      "loss": 0.6153,
+      "step": 1574
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0833419561386108,
+      "learning_rate": 9.24658151451771e-06,
+      "loss": 0.5805,
+      "step": 1575
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.3130383491516113,
+      "learning_rate": 9.245485774093563e-06,
+      "loss": 0.6547,
+      "step": 1576
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1251847743988037,
+      "learning_rate": 9.244389302470785e-06,
+      "loss": 0.7021,
+      "step": 1577
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2331297397613525,
+      "learning_rate": 9.243292099838222e-06,
+      "loss": 0.7117,
+      "step": 1578
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0849312543869019,
+      "learning_rate": 9.24219416638484e-06,
+      "loss": 0.6178,
+      "step": 1579
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2171435356140137,
+      "learning_rate": 9.24109550229974e-06,
+      "loss": 0.568,
+      "step": 1580
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9565621614456177,
+      "learning_rate": 9.239996107772144e-06,
+      "loss": 0.6697,
+      "step": 1581
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4817333221435547,
+      "learning_rate": 9.238895982991398e-06,
+      "loss": 0.6317,
+      "step": 1582
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3165608644485474,
+      "learning_rate": 9.23779512814698e-06,
+      "loss": 0.6251,
+      "step": 1583
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.244059681892395,
+      "learning_rate": 9.236693543428485e-06,
+      "loss": 0.6704,
+      "step": 1584
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1906737089157104,
+      "learning_rate": 9.235591229025643e-06,
+      "loss": 0.6598,
+      "step": 1585
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.268912434577942,
+      "learning_rate": 9.234488185128304e-06,
+      "loss": 0.5556,
+      "step": 1586
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1951463222503662,
+      "learning_rate": 9.233384411926442e-06,
+      "loss": 0.7621,
+      "step": 1587
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4412353038787842,
+      "learning_rate": 9.232279909610163e-06,
+      "loss": 0.7506,
+      "step": 1588
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2029129266738892,
+      "learning_rate": 9.231174678369695e-06,
+      "loss": 0.6591,
+      "step": 1589
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1525003910064697,
+      "learning_rate": 9.23006871839539e-06,
+      "loss": 0.6395,
+      "step": 1590
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.305783748626709,
+      "learning_rate": 9.228962029877724e-06,
+      "loss": 0.6779,
+      "step": 1591
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1824591159820557,
+      "learning_rate": 9.227854613007308e-06,
+      "loss": 0.7231,
+      "step": 1592
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8407436609268188,
+      "learning_rate": 9.22674646797487e-06,
+      "loss": 0.5819,
+      "step": 1593
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2329612970352173,
+      "learning_rate": 9.225637594971265e-06,
+      "loss": 0.7076,
+      "step": 1594
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.214455246925354,
+      "learning_rate": 9.224527994187471e-06,
+      "loss": 0.6108,
+      "step": 1595
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1704137325286865,
+      "learning_rate": 9.223417665814599e-06,
+      "loss": 0.6927,
+      "step": 1596
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.149852991104126,
+      "learning_rate": 9.222306610043877e-06,
+      "loss": 0.602,
+      "step": 1597
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2092525959014893,
+      "learning_rate": 9.221194827066664e-06,
+      "loss": 0.5947,
+      "step": 1598
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0760937929153442,
+      "learning_rate": 9.22008231707444e-06,
+      "loss": 0.5526,
+      "step": 1599
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1572587490081787,
+      "learning_rate": 9.218969080258816e-06,
+      "loss": 0.6291,
+      "step": 1600
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.240919828414917,
+      "learning_rate": 9.217855116811519e-06,
+      "loss": 0.6204,
+      "step": 1601
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2830955982208252,
+      "learning_rate": 9.21674042692441e-06,
+      "loss": 0.6191,
+      "step": 1602
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.104110598564148,
+      "learning_rate": 9.215625010789469e-06,
+      "loss": 0.6335,
+      "step": 1603
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0325019359588623,
+      "learning_rate": 9.214508868598807e-06,
+      "loss": 0.5526,
+      "step": 1604
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1608110666275024,
+      "learning_rate": 9.213392000544656e-06,
+      "loss": 0.6592,
+      "step": 1605
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1397209167480469,
+      "learning_rate": 9.212274406819373e-06,
+      "loss": 0.6102,
+      "step": 1606
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.098311424255371,
+      "learning_rate": 9.211156087615442e-06,
+      "loss": 0.6441,
+      "step": 1607
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3748606443405151,
+      "learning_rate": 9.210037043125469e-06,
+      "loss": 0.6898,
+      "step": 1608
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2580170631408691,
+      "learning_rate": 9.208917273542188e-06,
+      "loss": 0.6937,
+      "step": 1609
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1992050409317017,
+      "learning_rate": 9.207796779058456e-06,
+      "loss": 0.7689,
+      "step": 1610
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.173262596130371,
+      "learning_rate": 9.206675559867254e-06,
+      "loss": 0.5427,
+      "step": 1611
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1421955823898315,
+      "learning_rate": 9.205553616161692e-06,
+      "loss": 0.648,
+      "step": 1612
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1522105932235718,
+      "learning_rate": 9.204430948135e-06,
+      "loss": 0.6322,
+      "step": 1613
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0081772804260254,
+      "learning_rate": 9.203307555980536e-06,
+      "loss": 0.6584,
+      "step": 1614
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1464192867279053,
+      "learning_rate": 9.20218343989178e-06,
+      "loss": 0.7091,
+      "step": 1615
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0025413036346436,
+      "learning_rate": 9.20105860006234e-06,
+      "loss": 0.6502,
+      "step": 1616
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2751092910766602,
+      "learning_rate": 9.199933036685946e-06,
+      "loss": 0.6899,
+      "step": 1617
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9596514701843262,
+      "learning_rate": 9.198806749956453e-06,
+      "loss": 0.6055,
+      "step": 1618
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1237578392028809,
+      "learning_rate": 9.197679740067842e-06,
+      "loss": 0.7279,
+      "step": 1619
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0528181791305542,
+      "learning_rate": 9.196552007214215e-06,
+      "loss": 0.6048,
+      "step": 1620
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0963466167449951,
+      "learning_rate": 9.195423551589803e-06,
+      "loss": 0.6251,
+      "step": 1621
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2052900791168213,
+      "learning_rate": 9.194294373388962e-06,
+      "loss": 0.7226,
+      "step": 1622
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6359459161758423,
+      "learning_rate": 9.193164472806165e-06,
+      "loss": 0.6516,
+      "step": 1623
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1933037042617798,
+      "learning_rate": 9.192033850036018e-06,
+      "loss": 0.6134,
+      "step": 1624
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.69362211227417,
+      "learning_rate": 9.190902505273247e-06,
+      "loss": 0.6487,
+      "step": 1625
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2256443500518799,
+      "learning_rate": 9.189770438712701e-06,
+      "loss": 0.5863,
+      "step": 1626
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.431551218032837,
+      "learning_rate": 9.188637650549357e-06,
+      "loss": 0.6076,
+      "step": 1627
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1389654874801636,
+      "learning_rate": 9.187504140978316e-06,
+      "loss": 0.6828,
+      "step": 1628
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3692309856414795,
+      "learning_rate": 9.1863699101948e-06,
+      "loss": 0.609,
+      "step": 1629
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4854897260665894,
+      "learning_rate": 9.18523495839416e-06,
+      "loss": 0.5828,
+      "step": 1630
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3503080606460571,
+      "learning_rate": 9.184099285771865e-06,
+      "loss": 0.6239,
+      "step": 1631
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0891162157058716,
+      "learning_rate": 9.182962892523515e-06,
+      "loss": 0.7232,
+      "step": 1632
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.0536813735961914,
+      "learning_rate": 9.181825778844826e-06,
+      "loss": 0.5698,
+      "step": 1633
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0989372730255127,
+      "learning_rate": 9.180687944931646e-06,
+      "loss": 0.7344,
+      "step": 1634
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1523288488388062,
+      "learning_rate": 9.179549390979946e-06,
+      "loss": 0.6036,
+      "step": 1635
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1294163465499878,
+      "learning_rate": 9.178410117185811e-06,
+      "loss": 0.6317,
+      "step": 1636
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0369744300842285,
+      "learning_rate": 9.177270123745466e-06,
+      "loss": 0.6539,
+      "step": 1637
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0784070491790771,
+      "learning_rate": 9.176129410855248e-06,
+      "loss": 0.7082,
+      "step": 1638
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0242846012115479,
+      "learning_rate": 9.17498797871162e-06,
+      "loss": 0.6233,
+      "step": 1639
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3577667474746704,
+      "learning_rate": 9.173845827511176e-06,
+      "loss": 0.7456,
+      "step": 1640
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2746890783309937,
+      "learning_rate": 9.172702957450622e-06,
+      "loss": 0.6227,
+      "step": 1641
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0363303422927856,
+      "learning_rate": 9.171559368726798e-06,
+      "loss": 0.6176,
+      "step": 1642
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.156578540802002,
+      "learning_rate": 9.170415061536661e-06,
+      "loss": 0.5688,
+      "step": 1643
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0655452013015747,
+      "learning_rate": 9.1692700360773e-06,
+      "loss": 0.6034,
+      "step": 1644
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3639652729034424,
+      "learning_rate": 9.168124292545917e-06,
+      "loss": 0.6289,
+      "step": 1645
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4169421195983887,
+      "learning_rate": 9.166977831139845e-06,
+      "loss": 0.6277,
+      "step": 1646
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0737788677215576,
+      "learning_rate": 9.165830652056537e-06,
+      "loss": 0.5081,
+      "step": 1647
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.884833574295044,
+      "learning_rate": 9.164682755493574e-06,
+      "loss": 0.5832,
+      "step": 1648
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2018709182739258,
+      "learning_rate": 9.163534141648658e-06,
+      "loss": 0.602,
+      "step": 1649
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1860284805297852,
+      "learning_rate": 9.162384810719612e-06,
+      "loss": 0.605,
+      "step": 1650
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.697682499885559,
+      "learning_rate": 9.161234762904386e-06,
+      "loss": 0.5906,
+      "step": 1651
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6724482774734497,
+      "learning_rate": 9.160083998401053e-06,
+      "loss": 0.7244,
+      "step": 1652
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3993253707885742,
+      "learning_rate": 9.158932517407806e-06,
+      "loss": 0.703,
+      "step": 1653
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4852396249771118,
+      "learning_rate": 9.15778032012297e-06,
+      "loss": 0.6691,
+      "step": 1654
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.278127670288086,
+      "learning_rate": 9.15662740674498e-06,
+      "loss": 0.6629,
+      "step": 1655
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3181184530258179,
+      "learning_rate": 9.155473777472408e-06,
+      "loss": 0.6775,
+      "step": 1656
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3667001724243164,
+      "learning_rate": 9.15431943250394e-06,
+      "loss": 0.5703,
+      "step": 1657
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3015559911727905,
+      "learning_rate": 9.15316437203839e-06,
+      "loss": 0.65,
+      "step": 1658
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3116071224212646,
+      "learning_rate": 9.152008596274695e-06,
+      "loss": 0.5931,
+      "step": 1659
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.37589430809021,
+      "learning_rate": 9.15085210541191e-06,
+      "loss": 0.5934,
+      "step": 1660
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3749538660049438,
+      "learning_rate": 9.149694899649218e-06,
+      "loss": 0.6775,
+      "step": 1661
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1061633825302124,
+      "learning_rate": 9.148536979185927e-06,
+      "loss": 0.6038,
+      "step": 1662
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.6514101028442383,
+      "learning_rate": 9.147378344221462e-06,
+      "loss": 0.5828,
+      "step": 1663
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.279590129852295,
+      "learning_rate": 9.146218994955378e-06,
+      "loss": 0.548,
+      "step": 1664
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0067777633666992,
+      "learning_rate": 9.145058931587345e-06,
+      "loss": 0.5791,
+      "step": 1665
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2838937044143677,
+      "learning_rate": 9.143898154317164e-06,
+      "loss": 0.5998,
+      "step": 1666
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0655570030212402,
+      "learning_rate": 9.142736663344754e-06,
+      "loss": 0.7026,
+      "step": 1667
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2458590269088745,
+      "learning_rate": 9.141574458870156e-06,
+      "loss": 0.7518,
+      "step": 1668
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3690664768218994,
+      "learning_rate": 9.140411541093539e-06,
+      "loss": 0.5947,
+      "step": 1669
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4112671613693237,
+      "learning_rate": 9.139247910215192e-06,
+      "loss": 0.6794,
+      "step": 1670
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0723782777786255,
+      "learning_rate": 9.138083566435525e-06,
+      "loss": 0.6019,
+      "step": 1671
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5119447708129883,
+      "learning_rate": 9.136918509955074e-06,
+      "loss": 0.6187,
+      "step": 1672
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.17743980884552,
+      "learning_rate": 9.135752740974495e-06,
+      "loss": 0.5778,
+      "step": 1673
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1854586601257324,
+      "learning_rate": 9.13458625969457e-06,
+      "loss": 0.6281,
+      "step": 1674
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9571298360824585,
+      "learning_rate": 9.133419066316198e-06,
+      "loss": 0.5768,
+      "step": 1675
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1036064624786377,
+      "learning_rate": 9.13225116104041e-06,
+      "loss": 0.5698,
+      "step": 1676
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1489185094833374,
+      "learning_rate": 9.131082544068346e-06,
+      "loss": 0.6844,
+      "step": 1677
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2321306467056274,
+      "learning_rate": 9.129913215601286e-06,
+      "loss": 0.6463,
+      "step": 1678
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2575052976608276,
+      "learning_rate": 9.128743175840615e-06,
+      "loss": 0.5624,
+      "step": 1679
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1586254835128784,
+      "learning_rate": 9.127572424987853e-06,
+      "loss": 0.7156,
+      "step": 1680
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7823494672775269,
+      "learning_rate": 9.126400963244636e-06,
+      "loss": 0.5968,
+      "step": 1681
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0591415166854858,
+      "learning_rate": 9.125228790812726e-06,
+      "loss": 0.7415,
+      "step": 1682
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2111726999282837,
+      "learning_rate": 9.124055907894004e-06,
+      "loss": 0.6182,
+      "step": 1683
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.14926278591156,
+      "learning_rate": 9.12288231469048e-06,
+      "loss": 0.6119,
+      "step": 1684
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0378811359405518,
+      "learning_rate": 9.121708011404275e-06,
+      "loss": 0.6497,
+      "step": 1685
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4201372861862183,
+      "learning_rate": 9.120532998237642e-06,
+      "loss": 0.7219,
+      "step": 1686
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1894242763519287,
+      "learning_rate": 9.119357275392954e-06,
+      "loss": 0.7498,
+      "step": 1687
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3526839017868042,
+      "learning_rate": 9.118180843072705e-06,
+      "loss": 0.6255,
+      "step": 1688
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2016130685806274,
+      "learning_rate": 9.117003701479508e-06,
+      "loss": 0.6505,
+      "step": 1689
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1583237648010254,
+      "learning_rate": 9.115825850816106e-06,
+      "loss": 0.5844,
+      "step": 1690
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1343616247177124,
+      "learning_rate": 9.114647291285358e-06,
+      "loss": 0.6809,
+      "step": 1691
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.766956090927124,
+      "learning_rate": 9.113468023090251e-06,
+      "loss": 0.6212,
+      "step": 1692
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2435486316680908,
+      "learning_rate": 9.112288046433883e-06,
+      "loss": 0.6192,
+      "step": 1693
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0251524448394775,
+      "learning_rate": 9.111107361519485e-06,
+      "loss": 0.5751,
+      "step": 1694
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0255200862884521,
+      "learning_rate": 9.109925968550405e-06,
+      "loss": 0.5642,
+      "step": 1695
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2505881786346436,
+      "learning_rate": 9.108743867730115e-06,
+      "loss": 0.6587,
+      "step": 1696
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3478747606277466,
+      "learning_rate": 9.107561059262207e-06,
+      "loss": 0.5907,
+      "step": 1697
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4244675636291504,
+      "learning_rate": 9.106377543350396e-06,
+      "loss": 0.6059,
+      "step": 1698
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9292632937431335,
+      "learning_rate": 9.105193320198518e-06,
+      "loss": 0.5866,
+      "step": 1699
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9981103539466858,
+      "learning_rate": 9.104008390010532e-06,
+      "loss": 0.6674,
+      "step": 1700
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.207874059677124,
+      "learning_rate": 9.102822752990517e-06,
+      "loss": 0.6106,
+      "step": 1701
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0549036264419556,
+      "learning_rate": 9.101636409342676e-06,
+      "loss": 0.583,
+      "step": 1702
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4138764142990112,
+      "learning_rate": 9.100449359271333e-06,
+      "loss": 0.7072,
+      "step": 1703
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2411020994186401,
+      "learning_rate": 9.099261602980933e-06,
+      "loss": 0.6662,
+      "step": 1704
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9189475774765015,
+      "learning_rate": 9.098073140676043e-06,
+      "loss": 0.6922,
+      "step": 1705
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.334502100944519,
+      "learning_rate": 9.096883972561347e-06,
+      "loss": 0.7025,
+      "step": 1706
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3240032196044922,
+      "learning_rate": 9.095694098841662e-06,
+      "loss": 0.6391,
+      "step": 1707
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.224280595779419,
+      "learning_rate": 9.094503519721917e-06,
+      "loss": 0.5683,
+      "step": 1708
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2366915941238403,
+      "learning_rate": 9.093312235407166e-06,
+      "loss": 0.6315,
+      "step": 1709
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2290258407592773,
+      "learning_rate": 9.09212024610258e-06,
+      "loss": 0.6049,
+      "step": 1710
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0810284614562988,
+      "learning_rate": 9.090927552013457e-06,
+      "loss": 0.6721,
+      "step": 1711
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1469855308532715,
+      "learning_rate": 9.089734153345215e-06,
+      "loss": 0.6771,
+      "step": 1712
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5117367506027222,
+      "learning_rate": 9.088540050303392e-06,
+      "loss": 0.6422,
+      "step": 1713
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0608218908309937,
+      "learning_rate": 9.087345243093646e-06,
+      "loss": 0.6211,
+      "step": 1714
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1350992918014526,
+      "learning_rate": 9.086149731921763e-06,
+      "loss": 0.6627,
+      "step": 1715
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1392055749893188,
+      "learning_rate": 9.084953516993642e-06,
+      "loss": 0.574,
+      "step": 1716
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4839973449707031,
+      "learning_rate": 9.083756598515307e-06,
+      "loss": 0.6127,
+      "step": 1717
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2805958986282349,
+      "learning_rate": 9.082558976692904e-06,
+      "loss": 0.6159,
+      "step": 1718
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9901695847511292,
+      "learning_rate": 9.081360651732698e-06,
+      "loss": 0.5861,
+      "step": 1719
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3135817050933838,
+      "learning_rate": 9.080161623841077e-06,
+      "loss": 0.646,
+      "step": 1720
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2557127475738525,
+      "learning_rate": 9.078961893224548e-06,
+      "loss": 0.6455,
+      "step": 1721
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4702677726745605,
+      "learning_rate": 9.07776146008974e-06,
+      "loss": 0.6288,
+      "step": 1722
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4625836610794067,
+      "learning_rate": 9.076560324643405e-06,
+      "loss": 0.7098,
+      "step": 1723
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4593355655670166,
+      "learning_rate": 9.075358487092413e-06,
+      "loss": 0.6319,
+      "step": 1724
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3008965253829956,
+      "learning_rate": 9.074155947643757e-06,
+      "loss": 0.6806,
+      "step": 1725
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2681013345718384,
+      "learning_rate": 9.07295270650455e-06,
+      "loss": 0.575,
+      "step": 1726
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2825214862823486,
+      "learning_rate": 9.071748763882025e-06,
+      "loss": 0.5942,
+      "step": 1727
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.12113356590271,
+      "learning_rate": 9.070544119983536e-06,
+      "loss": 0.6244,
+      "step": 1728
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.394400715827942,
+      "learning_rate": 9.069338775016558e-06,
+      "loss": 0.7336,
+      "step": 1729
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3107945919036865,
+      "learning_rate": 9.06813272918869e-06,
+      "loss": 0.6463,
+      "step": 1730
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9889616966247559,
+      "learning_rate": 9.066925982707647e-06,
+      "loss": 0.5611,
+      "step": 1731
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0906484127044678,
+      "learning_rate": 9.065718535781266e-06,
+      "loss": 0.6513,
+      "step": 1732
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.320393443107605,
+      "learning_rate": 9.064510388617507e-06,
+      "loss": 0.6201,
+      "step": 1733
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2902657985687256,
+      "learning_rate": 9.063301541424447e-06,
+      "loss": 0.662,
+      "step": 1734
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.093662142753601,
+      "learning_rate": 9.062091994410286e-06,
+      "loss": 0.6321,
+      "step": 1735
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2275786399841309,
+      "learning_rate": 9.060881747783347e-06,
+      "loss": 0.5982,
+      "step": 1736
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4073659181594849,
+      "learning_rate": 9.059670801752065e-06,
+      "loss": 0.6775,
+      "step": 1737
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3687350749969482,
+      "learning_rate": 9.058459156525003e-06,
+      "loss": 0.7289,
+      "step": 1738
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.213643193244934,
+      "learning_rate": 9.057246812310844e-06,
+      "loss": 0.6809,
+      "step": 1739
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2109642028808594,
+      "learning_rate": 9.056033769318387e-06,
+      "loss": 0.6457,
+      "step": 1740
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1009825468063354,
+      "learning_rate": 9.054820027756556e-06,
+      "loss": 0.6372,
+      "step": 1741
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8893618583679199,
+      "learning_rate": 9.05360558783439e-06,
+      "loss": 0.6498,
+      "step": 1742
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9988924264907837,
+      "learning_rate": 9.052390449761057e-06,
+      "loss": 0.6498,
+      "step": 1743
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.142358422279358,
+      "learning_rate": 9.051174613745836e-06,
+      "loss": 0.5587,
+      "step": 1744
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.34147047996521,
+      "learning_rate": 9.049958079998132e-06,
+      "loss": 0.6156,
+      "step": 1745
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.456156849861145,
+      "learning_rate": 9.048740848727467e-06,
+      "loss": 0.6106,
+      "step": 1746
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0447920560836792,
+      "learning_rate": 9.047522920143483e-06,
+      "loss": 0.6259,
+      "step": 1747
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.416192054748535,
+      "learning_rate": 9.046304294455945e-06,
+      "loss": 0.5878,
+      "step": 1748
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2161362171173096,
+      "learning_rate": 9.045084971874738e-06,
+      "loss": 0.6203,
+      "step": 1749
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0151557922363281,
+      "learning_rate": 9.043864952609863e-06,
+      "loss": 0.6412,
+      "step": 1750
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4537842273712158,
+      "learning_rate": 9.042644236871445e-06,
+      "loss": 0.6007,
+      "step": 1751
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3863743543624878,
+      "learning_rate": 9.041422824869729e-06,
+      "loss": 0.6443,
+      "step": 1752
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5908379554748535,
+      "learning_rate": 9.040200716815073e-06,
+      "loss": 0.5611,
+      "step": 1753
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6172839403152466,
+      "learning_rate": 9.038977912917963e-06,
+      "loss": 0.6122,
+      "step": 1754
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8064736127853394,
+      "learning_rate": 9.037754413389006e-06,
+      "loss": 0.6604,
+      "step": 1755
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4816584587097168,
+      "learning_rate": 9.03653021843892e-06,
+      "loss": 0.6819,
+      "step": 1756
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1716893911361694,
+      "learning_rate": 9.035305328278549e-06,
+      "loss": 0.6599,
+      "step": 1757
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0522955656051636,
+      "learning_rate": 9.034079743118857e-06,
+      "loss": 0.6139,
+      "step": 1758
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0910663604736328,
+      "learning_rate": 9.032853463170925e-06,
+      "loss": 0.7266,
+      "step": 1759
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1195260286331177,
+      "learning_rate": 9.031626488645955e-06,
+      "loss": 0.5692,
+      "step": 1760
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.161078929901123,
+      "learning_rate": 9.030398819755268e-06,
+      "loss": 0.6371,
+      "step": 1761
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5670857429504395,
+      "learning_rate": 9.029170456710303e-06,
+      "loss": 0.5881,
+      "step": 1762
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.10579252243042,
+      "learning_rate": 9.027941399722626e-06,
+      "loss": 0.6167,
+      "step": 1763
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.177972435951233,
+      "learning_rate": 9.026711649003911e-06,
+      "loss": 0.6471,
+      "step": 1764
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4338864088058472,
+      "learning_rate": 9.025481204765963e-06,
+      "loss": 0.6252,
+      "step": 1765
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.052895188331604,
+      "learning_rate": 9.024250067220697e-06,
+      "loss": 0.6478,
+      "step": 1766
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.422892689704895,
+      "learning_rate": 9.023018236580154e-06,
+      "loss": 0.6252,
+      "step": 1767
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3154759407043457,
+      "learning_rate": 9.02178571305649e-06,
+      "loss": 0.5794,
+      "step": 1768
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0219439268112183,
+      "learning_rate": 9.020552496861982e-06,
+      "loss": 0.6179,
+      "step": 1769
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.240836143493652,
+      "learning_rate": 9.019318588209028e-06,
+      "loss": 0.6145,
+      "step": 1770
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2236559391021729,
+      "learning_rate": 9.018083987310143e-06,
+      "loss": 0.7196,
+      "step": 1771
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.407920479774475,
+      "learning_rate": 9.01684869437796e-06,
+      "loss": 0.5701,
+      "step": 1772
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0528749227523804,
+      "learning_rate": 9.015612709625236e-06,
+      "loss": 0.6588,
+      "step": 1773
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.439102053642273,
+      "learning_rate": 9.014376033264845e-06,
+      "loss": 0.6773,
+      "step": 1774
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9866570830345154,
+      "learning_rate": 9.013138665509776e-06,
+      "loss": 0.6061,
+      "step": 1775
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7695170640945435,
+      "learning_rate": 9.011900606573142e-06,
+      "loss": 0.6408,
+      "step": 1776
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3377400636672974,
+      "learning_rate": 9.010661856668172e-06,
+      "loss": 0.7959,
+      "step": 1777
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6860862970352173,
+      "learning_rate": 9.00942241600822e-06,
+      "loss": 0.55,
+      "step": 1778
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0897938013076782,
+      "learning_rate": 9.00818228480675e-06,
+      "loss": 0.6778,
+      "step": 1779
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2090890407562256,
+      "learning_rate": 9.00694146327735e-06,
+      "loss": 0.7092,
+      "step": 1780
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0635035037994385,
+      "learning_rate": 9.005699951633727e-06,
+      "loss": 0.4973,
+      "step": 1781
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.029817819595337,
+      "learning_rate": 9.004457750089709e-06,
+      "loss": 0.7106,
+      "step": 1782
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2057615518569946,
+      "learning_rate": 9.003214858859234e-06,
+      "loss": 0.7264,
+      "step": 1783
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.096470594406128,
+      "learning_rate": 9.001971278156367e-06,
+      "loss": 0.6101,
+      "step": 1784
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1284945011138916,
+      "learning_rate": 9.000727008195293e-06,
+      "loss": 0.5879,
+      "step": 1785
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3269438743591309,
+      "learning_rate": 8.999482049190308e-06,
+      "loss": 0.6274,
+      "step": 1786
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.038558006286621,
+      "learning_rate": 8.998236401355835e-06,
+      "loss": 0.5697,
+      "step": 1787
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3907923698425293,
+      "learning_rate": 8.996990064906408e-06,
+      "loss": 0.7388,
+      "step": 1788
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2428247928619385,
+      "learning_rate": 8.995743040056683e-06,
+      "loss": 0.6643,
+      "step": 1789
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2280868291854858,
+      "learning_rate": 8.994495327021438e-06,
+      "loss": 0.6722,
+      "step": 1790
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2366783618927002,
+      "learning_rate": 8.993246926015562e-06,
+      "loss": 0.6621,
+      "step": 1791
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1328561305999756,
+      "learning_rate": 8.99199783725407e-06,
+      "loss": 0.5697,
+      "step": 1792
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3674986362457275,
+      "learning_rate": 8.990748060952091e-06,
+      "loss": 0.6256,
+      "step": 1793
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.234496831893921,
+      "learning_rate": 8.989497597324872e-06,
+      "loss": 0.6446,
+      "step": 1794
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1295185089111328,
+      "learning_rate": 8.988246446587781e-06,
+      "loss": 0.6178,
+      "step": 1795
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1279648542404175,
+      "learning_rate": 8.986994608956305e-06,
+      "loss": 0.6405,
+      "step": 1796
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.143036961555481,
+      "learning_rate": 8.985742084646048e-06,
+      "loss": 0.6375,
+      "step": 1797
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3107563257217407,
+      "learning_rate": 8.984488873872728e-06,
+      "loss": 0.6425,
+      "step": 1798
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1655327081680298,
+      "learning_rate": 8.983234976852187e-06,
+      "loss": 0.7571,
+      "step": 1799
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.195489764213562,
+      "learning_rate": 8.981980393800384e-06,
+      "loss": 0.6266,
+      "step": 1800
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3523964881896973,
+      "learning_rate": 8.980725124933396e-06,
+      "loss": 0.6549,
+      "step": 1801
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1617556810379028,
+      "learning_rate": 8.979469170467415e-06,
+      "loss": 0.62,
+      "step": 1802
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.08810293674469,
+      "learning_rate": 8.978212530618756e-06,
+      "loss": 0.6809,
+      "step": 1803
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0384219884872437,
+      "learning_rate": 8.976955205603849e-06,
+      "loss": 0.5748,
+      "step": 1804
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9001789093017578,
+      "learning_rate": 8.975697195639242e-06,
+      "loss": 0.5481,
+      "step": 1805
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.583070993423462,
+      "learning_rate": 8.974438500941603e-06,
+      "loss": 0.6438,
+      "step": 1806
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3476251363754272,
+      "learning_rate": 8.973179121727713e-06,
+      "loss": 0.6684,
+      "step": 1807
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0399832725524902,
+      "learning_rate": 8.97191905821448e-06,
+      "loss": 0.6654,
+      "step": 1808
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1862304210662842,
+      "learning_rate": 8.97065831061892e-06,
+      "loss": 0.6808,
+      "step": 1809
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1737608909606934,
+      "learning_rate": 8.969396879158173e-06,
+      "loss": 0.6566,
+      "step": 1810
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5188488960266113,
+      "learning_rate": 8.968134764049495e-06,
+      "loss": 0.6548,
+      "step": 1811
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1486880779266357,
+      "learning_rate": 8.96687196551026e-06,
+      "loss": 0.6055,
+      "step": 1812
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.047976016998291,
+      "learning_rate": 8.965608483757958e-06,
+      "loss": 0.5543,
+      "step": 1813
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.347322702407837,
+      "learning_rate": 8.964344319010196e-06,
+      "loss": 0.6678,
+      "step": 1814
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.521556854248047,
+      "learning_rate": 8.963079471484707e-06,
+      "loss": 0.6393,
+      "step": 1815
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.818633794784546,
+      "learning_rate": 8.96181394139933e-06,
+      "loss": 0.6287,
+      "step": 1816
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1293917894363403,
+      "learning_rate": 8.960547728972028e-06,
+      "loss": 0.6093,
+      "step": 1817
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.344512462615967,
+      "learning_rate": 8.959280834420882e-06,
+      "loss": 0.5448,
+      "step": 1818
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3196654319763184,
+      "learning_rate": 8.958013257964086e-06,
+      "loss": 0.7376,
+      "step": 1819
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4868218898773193,
+      "learning_rate": 8.956744999819958e-06,
+      "loss": 0.6157,
+      "step": 1820
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2535372972488403,
+      "learning_rate": 8.955476060206928e-06,
+      "loss": 0.5815,
+      "step": 1821
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2110216617584229,
+      "learning_rate": 8.954206439343543e-06,
+      "loss": 0.6472,
+      "step": 1822
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1209574937820435,
+      "learning_rate": 8.95293613744847e-06,
+      "loss": 0.6802,
+      "step": 1823
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.10364830493927,
+      "learning_rate": 8.951665154740495e-06,
+      "loss": 0.6275,
+      "step": 1824
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5042253732681274,
+      "learning_rate": 8.950393491438518e-06,
+      "loss": 0.5856,
+      "step": 1825
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1649775505065918,
+      "learning_rate": 8.949121147761556e-06,
+      "loss": 0.7074,
+      "step": 1826
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2775204181671143,
+      "learning_rate": 8.947848123928747e-06,
+      "loss": 0.6768,
+      "step": 1827
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.109620213508606,
+      "learning_rate": 8.94657442015934e-06,
+      "loss": 0.6156,
+      "step": 1828
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.454298973083496,
+      "learning_rate": 8.945300036672709e-06,
+      "loss": 0.5578,
+      "step": 1829
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3531197309494019,
+      "learning_rate": 8.944024973688334e-06,
+      "loss": 0.637,
+      "step": 1830
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2573683261871338,
+      "learning_rate": 8.942749231425824e-06,
+      "loss": 0.6422,
+      "step": 1831
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0720746517181396,
+      "learning_rate": 8.941472810104898e-06,
+      "loss": 0.6234,
+      "step": 1832
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.301047444343567,
+      "learning_rate": 8.940195709945395e-06,
+      "loss": 0.6037,
+      "step": 1833
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0918940305709839,
+      "learning_rate": 8.938917931167268e-06,
+      "loss": 0.6048,
+      "step": 1834
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1322157382965088,
+      "learning_rate": 8.93763947399059e-06,
+      "loss": 0.6611,
+      "step": 1835
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2955923080444336,
+      "learning_rate": 8.936360338635546e-06,
+      "loss": 0.6222,
+      "step": 1836
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2364767789840698,
+      "learning_rate": 8.935080525322443e-06,
+      "loss": 0.6594,
+      "step": 1837
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4350496530532837,
+      "learning_rate": 8.933800034271706e-06,
+      "loss": 0.6246,
+      "step": 1838
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5731194019317627,
+      "learning_rate": 8.932518865703868e-06,
+      "loss": 0.6258,
+      "step": 1839
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1843388080596924,
+      "learning_rate": 8.931237019839587e-06,
+      "loss": 0.5611,
+      "step": 1840
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2916220426559448,
+      "learning_rate": 8.929954496899636e-06,
+      "loss": 0.6868,
+      "step": 1841
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1749646663665771,
+      "learning_rate": 8.928671297104901e-06,
+      "loss": 0.6049,
+      "step": 1842
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2070494890213013,
+      "learning_rate": 8.927387420676387e-06,
+      "loss": 0.6208,
+      "step": 1843
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1378085613250732,
+      "learning_rate": 8.92610286783522e-06,
+      "loss": 0.7296,
+      "step": 1844
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2360167503356934,
+      "learning_rate": 8.924817638802634e-06,
+      "loss": 0.6041,
+      "step": 1845
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0370523929595947,
+      "learning_rate": 8.923531733799984e-06,
+      "loss": 0.6482,
+      "step": 1846
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1742146015167236,
+      "learning_rate": 8.922245153048742e-06,
+      "loss": 0.5543,
+      "step": 1847
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9824729561805725,
+      "learning_rate": 8.920957896770495e-06,
+      "loss": 0.6877,
+      "step": 1848
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5906468629837036,
+      "learning_rate": 8.919669965186946e-06,
+      "loss": 0.6197,
+      "step": 1849
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0606296062469482,
+      "learning_rate": 8.918381358519916e-06,
+      "loss": 0.6899,
+      "step": 1850
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3511884212493896,
+      "learning_rate": 8.917092076991342e-06,
+      "loss": 0.677,
+      "step": 1851
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.168999195098877,
+      "learning_rate": 8.915802120823274e-06,
+      "loss": 0.5945,
+      "step": 1852
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0430333614349365,
+      "learning_rate": 8.914511490237883e-06,
+      "loss": 0.5796,
+      "step": 1853
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0085581541061401,
+      "learning_rate": 8.913220185457455e-06,
+      "loss": 0.5656,
+      "step": 1854
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1767834424972534,
+      "learning_rate": 8.911928206704388e-06,
+      "loss": 0.6895,
+      "step": 1855
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9236755967140198,
+      "learning_rate": 8.910635554201199e-06,
+      "loss": 0.6282,
+      "step": 1856
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.189577341079712,
+      "learning_rate": 8.909342228170523e-06,
+      "loss": 0.6037,
+      "step": 1857
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.07065749168396,
+      "learning_rate": 8.90804822883511e-06,
+      "loss": 0.6259,
+      "step": 1858
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.224584698677063,
+      "learning_rate": 8.906753556417822e-06,
+      "loss": 0.5854,
+      "step": 1859
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.136846899986267,
+      "learning_rate": 8.905458211141642e-06,
+      "loss": 0.6162,
+      "step": 1860
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1061546802520752,
+      "learning_rate": 8.904162193229667e-06,
+      "loss": 0.6708,
+      "step": 1861
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1807464361190796,
+      "learning_rate": 8.90286550290511e-06,
+      "loss": 0.6058,
+      "step": 1862
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3680741786956787,
+      "learning_rate": 8.901568140391298e-06,
+      "loss": 0.5691,
+      "step": 1863
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.45723557472229,
+      "learning_rate": 8.900270105911676e-06,
+      "loss": 0.6373,
+      "step": 1864
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1721493005752563,
+      "learning_rate": 8.898971399689804e-06,
+      "loss": 0.6277,
+      "step": 1865
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2810320854187012,
+      "learning_rate": 8.89767202194936e-06,
+      "loss": 0.647,
+      "step": 1866
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.231530785560608,
+      "learning_rate": 8.896371972914131e-06,
+      "loss": 0.6348,
+      "step": 1867
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2142878770828247,
+      "learning_rate": 8.895071252808025e-06,
+      "loss": 0.6003,
+      "step": 1868
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8936800956726074,
+      "learning_rate": 8.893769861855068e-06,
+      "loss": 0.6248,
+      "step": 1869
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1831871271133423,
+      "learning_rate": 8.892467800279396e-06,
+      "loss": 0.6167,
+      "step": 1870
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.351455807685852,
+      "learning_rate": 8.891165068305263e-06,
+      "loss": 0.6668,
+      "step": 1871
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.335632562637329,
+      "learning_rate": 8.889861666157038e-06,
+      "loss": 0.6294,
+      "step": 1872
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1500890254974365,
+      "learning_rate": 8.888557594059204e-06,
+      "loss": 0.6425,
+      "step": 1873
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1159440279006958,
+      "learning_rate": 8.887252852236365e-06,
+      "loss": 0.6158,
+      "step": 1874
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4507861137390137,
+      "learning_rate": 8.885947440913232e-06,
+      "loss": 0.6023,
+      "step": 1875
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.169786810874939,
+      "learning_rate": 8.884641360314636e-06,
+      "loss": 0.6775,
+      "step": 1876
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3057630062103271,
+      "learning_rate": 8.883334610665527e-06,
+      "loss": 0.5631,
+      "step": 1877
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4330273866653442,
+      "learning_rate": 8.88202719219096e-06,
+      "loss": 0.6352,
+      "step": 1878
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.038002610206604,
+      "learning_rate": 8.880719105116116e-06,
+      "loss": 0.6006,
+      "step": 1879
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.988121747970581,
+      "learning_rate": 8.879410349666284e-06,
+      "loss": 0.5431,
+      "step": 1880
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.9103357791900635,
+      "learning_rate": 8.87810092606687e-06,
+      "loss": 0.6212,
+      "step": 1881
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3641910552978516,
+      "learning_rate": 8.876790834543398e-06,
+      "loss": 0.7146,
+      "step": 1882
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3473871946334839,
+      "learning_rate": 8.875480075321506e-06,
+      "loss": 0.6231,
+      "step": 1883
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0874309539794922,
+      "learning_rate": 8.87416864862694e-06,
+      "loss": 0.6109,
+      "step": 1884
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3928847312927246,
+      "learning_rate": 8.872856554685569e-06,
+      "loss": 0.6334,
+      "step": 1885
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6365965604782104,
+      "learning_rate": 8.871543793723378e-06,
+      "loss": 0.6463,
+      "step": 1886
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2042670249938965,
+      "learning_rate": 8.870230365966459e-06,
+      "loss": 0.5847,
+      "step": 1887
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.115470290184021,
+      "learning_rate": 8.868916271641025e-06,
+      "loss": 0.6091,
+      "step": 1888
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1585477590560913,
+      "learning_rate": 8.867601510973402e-06,
+      "loss": 0.7887,
+      "step": 1889
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2588825225830078,
+      "learning_rate": 8.86628608419003e-06,
+      "loss": 0.4943,
+      "step": 1890
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.139786720275879,
+      "learning_rate": 8.864969991517465e-06,
+      "loss": 0.5754,
+      "step": 1891
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.179455280303955,
+      "learning_rate": 8.86365323318238e-06,
+      "loss": 0.5672,
+      "step": 1892
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.169512152671814,
+      "learning_rate": 8.862335809411556e-06,
+      "loss": 0.5903,
+      "step": 1893
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.500291347503662,
+      "learning_rate": 8.861017720431893e-06,
+      "loss": 0.7036,
+      "step": 1894
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1209545135498047,
+      "learning_rate": 8.859698966470404e-06,
+      "loss": 0.6231,
+      "step": 1895
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0172909498214722,
+      "learning_rate": 8.858379547754222e-06,
+      "loss": 0.6046,
+      "step": 1896
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.06539785861969,
+      "learning_rate": 8.857059464510586e-06,
+      "loss": 0.5999,
+      "step": 1897
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3168922662734985,
+      "learning_rate": 8.855738716966857e-06,
+      "loss": 0.5925,
+      "step": 1898
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9954319596290588,
+      "learning_rate": 8.854417305350503e-06,
+      "loss": 0.5778,
+      "step": 1899
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1088371276855469,
+      "learning_rate": 8.853095229889112e-06,
+      "loss": 0.5955,
+      "step": 1900
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4451113939285278,
+      "learning_rate": 8.851772490810386e-06,
+      "loss": 0.5305,
+      "step": 1901
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0991573333740234,
+      "learning_rate": 8.850449088342138e-06,
+      "loss": 0.7227,
+      "step": 1902
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.025597333908081,
+      "learning_rate": 8.849125022712297e-06,
+      "loss": 0.5666,
+      "step": 1903
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0423365831375122,
+      "learning_rate": 8.847800294148908e-06,
+      "loss": 0.6944,
+      "step": 1904
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0984046459197998,
+      "learning_rate": 8.846474902880128e-06,
+      "loss": 0.6633,
+      "step": 1905
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.16025710105896,
+      "learning_rate": 8.845148849134228e-06,
+      "loss": 0.7596,
+      "step": 1906
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2320754528045654,
+      "learning_rate": 8.843822133139595e-06,
+      "loss": 0.5805,
+      "step": 1907
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1268361806869507,
+      "learning_rate": 8.842494755124728e-06,
+      "loss": 0.6417,
+      "step": 1908
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4205336570739746,
+      "learning_rate": 8.84116671531824e-06,
+      "loss": 0.684,
+      "step": 1909
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1660581827163696,
+      "learning_rate": 8.839838013948861e-06,
+      "loss": 0.6393,
+      "step": 1910
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.339078426361084,
+      "learning_rate": 8.838508651245432e-06,
+      "loss": 0.6015,
+      "step": 1911
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1819093227386475,
+      "learning_rate": 8.837178627436907e-06,
+      "loss": 0.5398,
+      "step": 1912
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0426304340362549,
+      "learning_rate": 8.835847942752357e-06,
+      "loss": 0.5841,
+      "step": 1913
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4607350826263428,
+      "learning_rate": 8.834516597420968e-06,
+      "loss": 0.5586,
+      "step": 1914
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.166911244392395,
+      "learning_rate": 8.833184591672033e-06,
+      "loss": 0.6171,
+      "step": 1915
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.135672688484192,
+      "learning_rate": 8.831851925734963e-06,
+      "loss": 0.6544,
+      "step": 1916
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9493619799613953,
+      "learning_rate": 8.830518599839286e-06,
+      "loss": 0.6211,
+      "step": 1917
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2883530855178833,
+      "learning_rate": 8.829184614214637e-06,
+      "loss": 0.6184,
+      "step": 1918
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0521215200424194,
+      "learning_rate": 8.82784996909077e-06,
+      "loss": 0.6081,
+      "step": 1919
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2523962259292603,
+      "learning_rate": 8.82651466469755e-06,
+      "loss": 0.6249,
+      "step": 1920
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0267293453216553,
+      "learning_rate": 8.825178701264957e-06,
+      "loss": 0.5879,
+      "step": 1921
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.191742181777954,
+      "learning_rate": 8.82384207902308e-06,
+      "loss": 0.7142,
+      "step": 1922
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.205351710319519,
+      "learning_rate": 8.822504798202128e-06,
+      "loss": 0.6608,
+      "step": 1923
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1526292562484741,
+      "learning_rate": 8.821166859032419e-06,
+      "loss": 0.6627,
+      "step": 1924
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3558284044265747,
+      "learning_rate": 8.819828261744388e-06,
+      "loss": 0.6308,
+      "step": 1925
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1486369371414185,
+      "learning_rate": 8.81848900656858e-06,
+      "loss": 0.6342,
+      "step": 1926
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.2741756439208984,
+      "learning_rate": 8.817149093735654e-06,
+      "loss": 0.5495,
+      "step": 1927
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.269870638847351,
+      "learning_rate": 8.815808523476383e-06,
+      "loss": 0.645,
+      "step": 1928
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1454700231552124,
+      "learning_rate": 8.814467296021652e-06,
+      "loss": 0.6326,
+      "step": 1929
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2435215711593628,
+      "learning_rate": 8.813125411602463e-06,
+      "loss": 0.6898,
+      "step": 1930
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1720244884490967,
+      "learning_rate": 8.811782870449925e-06,
+      "loss": 0.6177,
+      "step": 1931
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.859727382659912,
+      "learning_rate": 8.810439672795266e-06,
+      "loss": 0.6215,
+      "step": 1932
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.085492491722107,
+      "learning_rate": 8.809095818869823e-06,
+      "loss": 0.6814,
+      "step": 1933
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1132259368896484,
+      "learning_rate": 8.807751308905049e-06,
+      "loss": 0.6428,
+      "step": 1934
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0994184017181396,
+      "learning_rate": 8.806406143132507e-06,
+      "loss": 0.6146,
+      "step": 1935
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1863535642623901,
+      "learning_rate": 8.805060321783873e-06,
+      "loss": 0.7188,
+      "step": 1936
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.321311116218567,
+      "learning_rate": 8.803713845090942e-06,
+      "loss": 0.5959,
+      "step": 1937
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1888467073440552,
+      "learning_rate": 8.802366713285612e-06,
+      "loss": 0.6004,
+      "step": 1938
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0325742959976196,
+      "learning_rate": 8.801018926599904e-06,
+      "loss": 0.5642,
+      "step": 1939
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.123363971710205,
+      "learning_rate": 8.799670485265944e-06,
+      "loss": 0.708,
+      "step": 1940
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2410331964492798,
+      "learning_rate": 8.798321389515974e-06,
+      "loss": 0.6157,
+      "step": 1941
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.289198398590088,
+      "learning_rate": 8.796971639582347e-06,
+      "loss": 0.6785,
+      "step": 1942
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5738413333892822,
+      "learning_rate": 8.795621235697531e-06,
+      "loss": 0.6329,
+      "step": 1943
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.011083722114563,
+      "learning_rate": 8.79427017809411e-06,
+      "loss": 0.5909,
+      "step": 1944
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3928797245025635,
+      "learning_rate": 8.792918467004767e-06,
+      "loss": 0.5906,
+      "step": 1945
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9257722496986389,
+      "learning_rate": 8.791566102662315e-06,
+      "loss": 0.5623,
+      "step": 1946
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2923310995101929,
+      "learning_rate": 8.790213085299668e-06,
+      "loss": 0.6385,
+      "step": 1947
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4501917362213135,
+      "learning_rate": 8.788859415149856e-06,
+      "loss": 0.5815,
+      "step": 1948
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2182177305221558,
+      "learning_rate": 8.787505092446022e-06,
+      "loss": 0.6108,
+      "step": 1949
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0286319255828857,
+      "learning_rate": 8.786150117421418e-06,
+      "loss": 0.5499,
+      "step": 1950
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2858717441558838,
+      "learning_rate": 8.784794490309414e-06,
+      "loss": 0.5579,
+      "step": 1951
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9833030104637146,
+      "learning_rate": 8.783438211343487e-06,
+      "loss": 0.6406,
+      "step": 1952
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0329986810684204,
+      "learning_rate": 8.78208128075723e-06,
+      "loss": 0.6064,
+      "step": 1953
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3916441202163696,
+      "learning_rate": 8.780723698784346e-06,
+      "loss": 0.6235,
+      "step": 1954
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7905875444412231,
+      "learning_rate": 8.77936546565865e-06,
+      "loss": 0.6405,
+      "step": 1955
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.13296377658844,
+      "learning_rate": 8.778006581614073e-06,
+      "loss": 0.6188,
+      "step": 1956
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1696197986602783,
+      "learning_rate": 8.776647046884651e-06,
+      "loss": 0.5943,
+      "step": 1957
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2003610134124756,
+      "learning_rate": 8.77528686170454e-06,
+      "loss": 0.6799,
+      "step": 1958
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1605165004730225,
+      "learning_rate": 8.773926026308002e-06,
+      "loss": 0.7245,
+      "step": 1959
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8035575151443481,
+      "learning_rate": 8.772564540929414e-06,
+      "loss": 0.6746,
+      "step": 1960
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1232562065124512,
+      "learning_rate": 8.771202405803263e-06,
+      "loss": 0.6389,
+      "step": 1961
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3930180072784424,
+      "learning_rate": 8.769839621164152e-06,
+      "loss": 0.5986,
+      "step": 1962
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1030189990997314,
+      "learning_rate": 8.768476187246789e-06,
+      "loss": 0.579,
+      "step": 1963
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1885334253311157,
+      "learning_rate": 8.767112104286003e-06,
+      "loss": 0.6221,
+      "step": 1964
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8045053482055664,
+      "learning_rate": 8.765747372516723e-06,
+      "loss": 0.59,
+      "step": 1965
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1481117010116577,
+      "learning_rate": 8.764381992174001e-06,
+      "loss": 0.6523,
+      "step": 1966
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3269027471542358,
+      "learning_rate": 8.763015963492996e-06,
+      "loss": 0.6512,
+      "step": 1967
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9991157054901123,
+      "learning_rate": 8.761649286708975e-06,
+      "loss": 0.5713,
+      "step": 1968
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0284345149993896,
+      "learning_rate": 8.760281962057324e-06,
+      "loss": 0.6516,
+      "step": 1969
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1772648096084595,
+      "learning_rate": 8.758913989773536e-06,
+      "loss": 0.7054,
+      "step": 1970
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2631373405456543,
+      "learning_rate": 8.757545370093216e-06,
+      "loss": 0.727,
+      "step": 1971
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1898554563522339,
+      "learning_rate": 8.756176103252082e-06,
+      "loss": 0.601,
+      "step": 1972
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.192469835281372,
+      "learning_rate": 8.754806189485959e-06,
+      "loss": 0.5665,
+      "step": 1973
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9782890677452087,
+      "learning_rate": 8.75343562903079e-06,
+      "loss": 0.5972,
+      "step": 1974
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5428776741027832,
+      "learning_rate": 8.752064422122625e-06,
+      "loss": 0.6663,
+      "step": 1975
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1398931741714478,
+      "learning_rate": 8.750692568997629e-06,
+      "loss": 0.666,
+      "step": 1976
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2742382287979126,
+      "learning_rate": 8.74932006989207e-06,
+      "loss": 0.6918,
+      "step": 1977
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2991219758987427,
+      "learning_rate": 8.747946925042341e-06,
+      "loss": 0.6471,
+      "step": 1978
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6621350049972534,
+      "learning_rate": 8.746573134684932e-06,
+      "loss": 0.5951,
+      "step": 1979
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5337145328521729,
+      "learning_rate": 8.745198699056452e-06,
+      "loss": 0.5993,
+      "step": 1980
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0378293991088867,
+      "learning_rate": 8.74382361839362e-06,
+      "loss": 0.6211,
+      "step": 1981
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1881955862045288,
+      "learning_rate": 8.742447892933266e-06,
+      "loss": 0.6135,
+      "step": 1982
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.375791311264038,
+      "learning_rate": 8.741071522912331e-06,
+      "loss": 0.7713,
+      "step": 1983
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.263593077659607,
+      "learning_rate": 8.739694508567866e-06,
+      "loss": 0.6795,
+      "step": 1984
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0440095663070679,
+      "learning_rate": 8.738316850137034e-06,
+      "loss": 0.4923,
+      "step": 1985
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.112327218055725,
+      "learning_rate": 8.736938547857109e-06,
+      "loss": 0.6235,
+      "step": 1986
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.110875129699707,
+      "learning_rate": 8.735559601965475e-06,
+      "loss": 0.6164,
+      "step": 1987
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1973077058792114,
+      "learning_rate": 8.734180012699628e-06,
+      "loss": 0.5878,
+      "step": 1988
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1539205312728882,
+      "learning_rate": 8.732799780297174e-06,
+      "loss": 0.7529,
+      "step": 1989
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2167625427246094,
+      "learning_rate": 8.731418904995829e-06,
+      "loss": 0.6641,
+      "step": 1990
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0853583812713623,
+      "learning_rate": 8.730037387033422e-06,
+      "loss": 0.598,
+      "step": 1991
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1032205820083618,
+      "learning_rate": 8.72865522664789e-06,
+      "loss": 0.643,
+      "step": 1992
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2810516357421875,
+      "learning_rate": 8.727272424077284e-06,
+      "loss": 0.5398,
+      "step": 1993
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2989377975463867,
+      "learning_rate": 8.725888979559762e-06,
+      "loss": 0.5852,
+      "step": 1994
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.120161533355713,
+      "learning_rate": 8.724504893333596e-06,
+      "loss": 0.7027,
+      "step": 1995
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3071115016937256,
+      "learning_rate": 8.723120165637165e-06,
+      "loss": 0.5965,
+      "step": 1996
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1043663024902344,
+      "learning_rate": 8.72173479670896e-06,
+      "loss": 0.6167,
+      "step": 1997
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1847256422042847,
+      "learning_rate": 8.720348786787583e-06,
+      "loss": 0.6122,
+      "step": 1998
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1380722522735596,
+      "learning_rate": 8.718962136111749e-06,
+      "loss": 0.6251,
+      "step": 1999
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1924020051956177,
+      "learning_rate": 8.717574844920274e-06,
+      "loss": 0.5994,
+      "step": 2000
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.000066876411438,
+      "learning_rate": 8.716186913452097e-06,
+      "loss": 0.6475,
+      "step": 2001
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0465896129608154,
+      "learning_rate": 8.714798341946258e-06,
+      "loss": 0.6021,
+      "step": 2002
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0609678030014038,
+      "learning_rate": 8.71340913064191e-06,
+      "loss": 0.5697,
+      "step": 2003
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1682281494140625,
+      "learning_rate": 8.712019279778319e-06,
+      "loss": 0.6211,
+      "step": 2004
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4044312238693237,
+      "learning_rate": 8.710628789594855e-06,
+      "loss": 0.6077,
+      "step": 2005
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1312463283538818,
+      "learning_rate": 8.709237660331003e-06,
+      "loss": 0.5921,
+      "step": 2006
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1439876556396484,
+      "learning_rate": 8.70784589222636e-06,
+      "loss": 0.5924,
+      "step": 2007
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3385566473007202,
+      "learning_rate": 8.706453485520622e-06,
+      "loss": 0.5244,
+      "step": 2008
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2704468965530396,
+      "learning_rate": 8.70506044045361e-06,
+      "loss": 0.5937,
+      "step": 2009
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1238104104995728,
+      "learning_rate": 8.703666757265246e-06,
+      "loss": 0.6783,
+      "step": 2010
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1684532165527344,
+      "learning_rate": 8.702272436195562e-06,
+      "loss": 0.6352,
+      "step": 2011
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1618826389312744,
+      "learning_rate": 8.700877477484704e-06,
+      "loss": 0.6676,
+      "step": 2012
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2076810598373413,
+      "learning_rate": 8.699481881372922e-06,
+      "loss": 0.6103,
+      "step": 2013
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1898448467254639,
+      "learning_rate": 8.698085648100581e-06,
+      "loss": 0.5716,
+      "step": 2014
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7246116399765015,
+      "learning_rate": 8.696688777908154e-06,
+      "loss": 0.5495,
+      "step": 2015
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.511298656463623,
+      "learning_rate": 8.695291271036221e-06,
+      "loss": 0.6376,
+      "step": 2016
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2623487710952759,
+      "learning_rate": 8.69389312772548e-06,
+      "loss": 0.6639,
+      "step": 2017
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.220610499382019,
+      "learning_rate": 8.692494348216726e-06,
+      "loss": 0.6363,
+      "step": 2018
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2728426456451416,
+      "learning_rate": 8.691094932750875e-06,
+      "loss": 0.6267,
+      "step": 2019
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6735302209854126,
+      "learning_rate": 8.689694881568945e-06,
+      "loss": 0.6491,
+      "step": 2020
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3414764404296875,
+      "learning_rate": 8.688294194912066e-06,
+      "loss": 0.6314,
+      "step": 2021
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5352545976638794,
+      "learning_rate": 8.686892873021481e-06,
+      "loss": 0.6437,
+      "step": 2022
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3221575021743774,
+      "learning_rate": 8.685490916138536e-06,
+      "loss": 0.6519,
+      "step": 2023
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5757750272750854,
+      "learning_rate": 8.684088324504694e-06,
+      "loss": 0.6486,
+      "step": 2024
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2225582599639893,
+      "learning_rate": 8.682685098361518e-06,
+      "loss": 0.7509,
+      "step": 2025
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0095558166503906,
+      "learning_rate": 8.681281237950688e-06,
+      "loss": 0.6723,
+      "step": 2026
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9800388216972351,
+      "learning_rate": 8.67987674351399e-06,
+      "loss": 0.5663,
+      "step": 2027
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.777536392211914,
+      "learning_rate": 8.678471615293317e-06,
+      "loss": 0.6395,
+      "step": 2028
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0923036336898804,
+      "learning_rate": 8.677065853530679e-06,
+      "loss": 0.5539,
+      "step": 2029
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2774169445037842,
+      "learning_rate": 8.675659458468186e-06,
+      "loss": 0.6695,
+      "step": 2030
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2363957166671753,
+      "learning_rate": 8.67425243034806e-06,
+      "loss": 0.6806,
+      "step": 2031
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3562581539154053,
+      "learning_rate": 8.672844769412637e-06,
+      "loss": 0.6637,
+      "step": 2032
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.290300965309143,
+      "learning_rate": 8.671436475904353e-06,
+      "loss": 0.6027,
+      "step": 2033
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1578395366668701,
+      "learning_rate": 8.670027550065763e-06,
+      "loss": 0.6095,
+      "step": 2034
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.5597174167633057,
+      "learning_rate": 8.668617992139524e-06,
+      "loss": 0.6364,
+      "step": 2035
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9307906031608582,
+      "learning_rate": 8.667207802368403e-06,
+      "loss": 0.671,
+      "step": 2036
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2509186267852783,
+      "learning_rate": 8.665796980995275e-06,
+      "loss": 0.6422,
+      "step": 2037
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.086897373199463,
+      "learning_rate": 8.66438552826313e-06,
+      "loss": 0.6447,
+      "step": 2038
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.172615885734558,
+      "learning_rate": 8.662973444415058e-06,
+      "loss": 0.562,
+      "step": 2039
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.004035234451294,
+      "learning_rate": 8.661560729694262e-06,
+      "loss": 0.515,
+      "step": 2040
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1104687452316284,
+      "learning_rate": 8.660147384344055e-06,
+      "loss": 0.5915,
+      "step": 2041
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2272541522979736,
+      "learning_rate": 8.658733408607856e-06,
+      "loss": 0.5539,
+      "step": 2042
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0483099222183228,
+      "learning_rate": 8.657318802729194e-06,
+      "loss": 0.5457,
+      "step": 2043
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2460724115371704,
+      "learning_rate": 8.655903566951706e-06,
+      "loss": 0.6596,
+      "step": 2044
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8781019449234009,
+      "learning_rate": 8.654487701519139e-06,
+      "loss": 0.5678,
+      "step": 2045
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4468796253204346,
+      "learning_rate": 8.653071206675344e-06,
+      "loss": 0.6264,
+      "step": 2046
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.015211582183838,
+      "learning_rate": 8.651654082664285e-06,
+      "loss": 0.5887,
+      "step": 2047
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4183422327041626,
+      "learning_rate": 8.650236329730034e-06,
+      "loss": 0.6769,
+      "step": 2048
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0147705078125,
+      "learning_rate": 8.648817948116767e-06,
+      "loss": 0.6663,
+      "step": 2049
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3201289176940918,
+      "learning_rate": 8.647398938068775e-06,
+      "loss": 0.6155,
+      "step": 2050
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1508480310440063,
+      "learning_rate": 8.645979299830452e-06,
+      "loss": 0.6329,
+      "step": 2051
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3195370435714722,
+      "learning_rate": 8.644559033646303e-06,
+      "loss": 0.6606,
+      "step": 2052
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2289732694625854,
+      "learning_rate": 8.643138139760935e-06,
+      "loss": 0.5887,
+      "step": 2053
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2333985567092896,
+      "learning_rate": 8.641716618419076e-06,
+      "loss": 0.7597,
+      "step": 2054
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6038957834243774,
+      "learning_rate": 8.640294469865548e-06,
+      "loss": 0.563,
+      "step": 2055
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6648248434066772,
+      "learning_rate": 8.638871694345293e-06,
+      "loss": 0.6064,
+      "step": 2056
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1484403610229492,
+      "learning_rate": 8.637448292103346e-06,
+      "loss": 0.6087,
+      "step": 2057
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.367751955986023,
+      "learning_rate": 8.636024263384868e-06,
+      "loss": 0.6503,
+      "step": 2058
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4821258783340454,
+      "learning_rate": 8.634599608435115e-06,
+      "loss": 0.7098,
+      "step": 2059
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0166525840759277,
+      "learning_rate": 8.633174327499456e-06,
+      "loss": 0.6259,
+      "step": 2060
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2003066539764404,
+      "learning_rate": 8.631748420823365e-06,
+      "loss": 0.7077,
+      "step": 2061
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.448062539100647,
+      "learning_rate": 8.630321888652426e-06,
+      "loss": 0.694,
+      "step": 2062
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.6138148307800293,
+      "learning_rate": 8.628894731232332e-06,
+      "loss": 0.6076,
+      "step": 2063
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1501845121383667,
+      "learning_rate": 8.62746694880888e-06,
+      "loss": 0.5773,
+      "step": 2064
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4958134889602661,
+      "learning_rate": 8.626038541627977e-06,
+      "loss": 0.5991,
+      "step": 2065
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0958569049835205,
+      "learning_rate": 8.624609509935637e-06,
+      "loss": 0.5845,
+      "step": 2066
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1663223505020142,
+      "learning_rate": 8.623179853977984e-06,
+      "loss": 0.5782,
+      "step": 2067
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1979085206985474,
+      "learning_rate": 8.621749574001241e-06,
+      "loss": 0.6255,
+      "step": 2068
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3670035600662231,
+      "learning_rate": 8.620318670251752e-06,
+      "loss": 0.6218,
+      "step": 2069
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0638803243637085,
+      "learning_rate": 8.618887142975956e-06,
+      "loss": 0.696,
+      "step": 2070
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.1462111473083496,
+      "learning_rate": 8.617454992420407e-06,
+      "loss": 0.6679,
+      "step": 2071
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3509697914123535,
+      "learning_rate": 8.616022218831764e-06,
+      "loss": 0.6722,
+      "step": 2072
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0871156454086304,
+      "learning_rate": 8.61458882245679e-06,
+      "loss": 0.5614,
+      "step": 2073
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0885635614395142,
+      "learning_rate": 8.613154803542362e-06,
+      "loss": 0.6469,
+      "step": 2074
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2749704122543335,
+      "learning_rate": 8.611720162335459e-06,
+      "loss": 0.6835,
+      "step": 2075
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2243422269821167,
+      "learning_rate": 8.61028489908317e-06,
+      "loss": 0.5278,
+      "step": 2076
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.17266845703125,
+      "learning_rate": 8.608849014032687e-06,
+      "loss": 0.6967,
+      "step": 2077
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2027003765106201,
+      "learning_rate": 8.607412507431316e-06,
+      "loss": 0.7052,
+      "step": 2078
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0147993564605713,
+      "learning_rate": 8.605975379526463e-06,
+      "loss": 0.6634,
+      "step": 2079
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7742810249328613,
+      "learning_rate": 8.604537630565644e-06,
+      "loss": 0.6074,
+      "step": 2080
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3109650611877441,
+      "learning_rate": 8.603099260796486e-06,
+      "loss": 0.6322,
+      "step": 2081
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1907052993774414,
+      "learning_rate": 8.601660270466714e-06,
+      "loss": 0.6935,
+      "step": 2082
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3169562816619873,
+      "learning_rate": 8.600220659824166e-06,
+      "loss": 0.6313,
+      "step": 2083
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1314212083816528,
+      "learning_rate": 8.598780429116788e-06,
+      "loss": 0.5839,
+      "step": 2084
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2062609195709229,
+      "learning_rate": 8.59733957859263e-06,
+      "loss": 0.6198,
+      "step": 2085
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9913787245750427,
+      "learning_rate": 8.595898108499845e-06,
+      "loss": 0.6855,
+      "step": 2086
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.212138056755066,
+      "learning_rate": 8.594456019086702e-06,
+      "loss": 0.5885,
+      "step": 2087
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0849419832229614,
+      "learning_rate": 8.59301331060157e-06,
+      "loss": 0.5638,
+      "step": 2088
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.176388263702393,
+      "learning_rate": 8.591569983292924e-06,
+      "loss": 0.5899,
+      "step": 2089
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0811898708343506,
+      "learning_rate": 8.590126037409353e-06,
+      "loss": 0.6131,
+      "step": 2090
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1859910488128662,
+      "learning_rate": 8.588681473199543e-06,
+      "loss": 0.5915,
+      "step": 2091
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8125941753387451,
+      "learning_rate": 8.587236290912292e-06,
+      "loss": 0.66,
+      "step": 2092
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1952428817749023,
+      "learning_rate": 8.585790490796502e-06,
+      "loss": 0.5988,
+      "step": 2093
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.172929048538208,
+      "learning_rate": 8.584344073101185e-06,
+      "loss": 0.6717,
+      "step": 2094
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0805124044418335,
+      "learning_rate": 8.582897038075455e-06,
+      "loss": 0.5679,
+      "step": 2095
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1243839263916016,
+      "learning_rate": 8.581449385968536e-06,
+      "loss": 0.6338,
+      "step": 2096
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2426873445510864,
+      "learning_rate": 8.580001117029755e-06,
+      "loss": 0.5016,
+      "step": 2097
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2226959466934204,
+      "learning_rate": 8.57855223150855e-06,
+      "loss": 0.628,
+      "step": 2098
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5704655647277832,
+      "learning_rate": 8.577102729654457e-06,
+      "loss": 0.6069,
+      "step": 2099
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1772339344024658,
+      "learning_rate": 8.575652611717127e-06,
+      "loss": 0.6016,
+      "step": 2100
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0823158025741577,
+      "learning_rate": 8.574201877946314e-06,
+      "loss": 0.6074,
+      "step": 2101
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.330729365348816,
+      "learning_rate": 8.572750528591875e-06,
+      "loss": 0.6382,
+      "step": 2102
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2223377227783203,
+      "learning_rate": 8.571298563903775e-06,
+      "loss": 0.6634,
+      "step": 2103
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.497209906578064,
+      "learning_rate": 8.56984598413209e-06,
+      "loss": 0.6778,
+      "step": 2104
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.1404335498809814,
+      "learning_rate": 8.568392789526992e-06,
+      "loss": 0.6256,
+      "step": 2105
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1916518211364746,
+      "learning_rate": 8.566938980338765e-06,
+      "loss": 0.7344,
+      "step": 2106
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0679166316986084,
+      "learning_rate": 8.565484556817802e-06,
+      "loss": 0.6275,
+      "step": 2107
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1432031393051147,
+      "learning_rate": 8.564029519214594e-06,
+      "loss": 0.6351,
+      "step": 2108
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1834468841552734,
+      "learning_rate": 8.562573867779741e-06,
+      "loss": 0.6035,
+      "step": 2109
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.60970938205719,
+      "learning_rate": 8.561117602763954e-06,
+      "loss": 0.6716,
+      "step": 2110
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1802395582199097,
+      "learning_rate": 8.559660724418041e-06,
+      "loss": 0.5264,
+      "step": 2111
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.7724010944366455,
+      "learning_rate": 8.558203232992923e-06,
+      "loss": 0.6596,
+      "step": 2112
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0022841691970825,
+      "learning_rate": 8.556745128739618e-06,
+      "loss": 0.6672,
+      "step": 2113
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4700846672058105,
+      "learning_rate": 8.55528641190926e-06,
+      "loss": 0.5608,
+      "step": 2114
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1476560831069946,
+      "learning_rate": 8.553827082753084e-06,
+      "loss": 0.5078,
+      "step": 2115
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1997826099395752,
+      "learning_rate": 8.552367141522423e-06,
+      "loss": 0.5914,
+      "step": 2116
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.022222876548767,
+      "learning_rate": 8.550906588468728e-06,
+      "loss": 0.6681,
+      "step": 2117
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3832002878189087,
+      "learning_rate": 8.549445423843548e-06,
+      "loss": 0.6029,
+      "step": 2118
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0408334732055664,
+      "learning_rate": 8.54798364789854e-06,
+      "loss": 0.5953,
+      "step": 2119
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0350518226623535,
+      "learning_rate": 8.546521260885463e-06,
+      "loss": 0.5858,
+      "step": 2120
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.097774624824524,
+      "learning_rate": 8.545058263056186e-06,
+      "loss": 0.6631,
+      "step": 2121
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5874125957489014,
+      "learning_rate": 8.543594654662677e-06,
+      "loss": 0.6732,
+      "step": 2122
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0040942430496216,
+      "learning_rate": 8.542130435957014e-06,
+      "loss": 0.6054,
+      "step": 2123
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.172186255455017,
+      "learning_rate": 8.540665607191383e-06,
+      "loss": 0.6075,
+      "step": 2124
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.285057783126831,
+      "learning_rate": 8.539200168618067e-06,
+      "loss": 0.5955,
+      "step": 2125
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.245982050895691,
+      "learning_rate": 8.537734120489459e-06,
+      "loss": 0.5765,
+      "step": 2126
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9209935665130615,
+      "learning_rate": 8.536267463058055e-06,
+      "loss": 0.65,
+      "step": 2127
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3906763792037964,
+      "learning_rate": 8.534800196576459e-06,
+      "loss": 0.6005,
+      "step": 2128
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8860881328582764,
+      "learning_rate": 8.533332321297374e-06,
+      "loss": 0.6604,
+      "step": 2129
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.790269374847412,
+      "learning_rate": 8.531863837473617e-06,
+      "loss": 0.6665,
+      "step": 2130
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.052520990371704,
+      "learning_rate": 8.530394745358101e-06,
+      "loss": 0.5778,
+      "step": 2131
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5288801193237305,
+      "learning_rate": 8.528925045203847e-06,
+      "loss": 0.6731,
+      "step": 2132
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9974155426025391,
+      "learning_rate": 8.527454737263983e-06,
+      "loss": 0.5731,
+      "step": 2133
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3304848670959473,
+      "learning_rate": 8.52598382179174e-06,
+      "loss": 0.5752,
+      "step": 2134
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.973964512348175,
+      "learning_rate": 8.524512299040451e-06,
+      "loss": 0.6316,
+      "step": 2135
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7174031734466553,
+      "learning_rate": 8.523040169263555e-06,
+      "loss": 0.6612,
+      "step": 2136
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5883331298828125,
+      "learning_rate": 8.5215674327146e-06,
+      "loss": 0.6166,
+      "step": 2137
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9867196083068848,
+      "learning_rate": 8.520094089647233e-06,
+      "loss": 0.557,
+      "step": 2138
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4658268690109253,
+      "learning_rate": 8.518620140315209e-06,
+      "loss": 0.5743,
+      "step": 2139
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9947718381881714,
+      "learning_rate": 8.517145584972383e-06,
+      "loss": 0.558,
+      "step": 2140
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1221495866775513,
+      "learning_rate": 8.515670423872719e-06,
+      "loss": 0.6239,
+      "step": 2141
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.916927695274353,
+      "learning_rate": 8.514194657270283e-06,
+      "loss": 0.6223,
+      "step": 2142
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1117417812347412,
+      "learning_rate": 8.512718285419246e-06,
+      "loss": 0.5865,
+      "step": 2143
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2808547019958496,
+      "learning_rate": 8.511241308573884e-06,
+      "loss": 0.6683,
+      "step": 2144
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0311280488967896,
+      "learning_rate": 8.509763726988573e-06,
+      "loss": 0.5753,
+      "step": 2145
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4085116386413574,
+      "learning_rate": 8.5082855409178e-06,
+      "loss": 0.5724,
+      "step": 2146
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.5789756774902344,
+      "learning_rate": 8.506806750616152e-06,
+      "loss": 0.5779,
+      "step": 2147
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2261545658111572,
+      "learning_rate": 8.505327356338318e-06,
+      "loss": 0.7003,
+      "step": 2148
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2848753929138184,
+      "learning_rate": 8.503847358339094e-06,
+      "loss": 0.534,
+      "step": 2149
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2150418758392334,
+      "learning_rate": 8.502366756873384e-06,
+      "loss": 0.5801,
+      "step": 2150
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.183171033859253,
+      "learning_rate": 8.500885552196187e-06,
+      "loss": 0.6252,
+      "step": 2151
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2070298194885254,
+      "learning_rate": 8.499403744562613e-06,
+      "loss": 0.611,
+      "step": 2152
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6139317750930786,
+      "learning_rate": 8.497921334227872e-06,
+      "loss": 0.6351,
+      "step": 2153
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6720945835113525,
+      "learning_rate": 8.496438321447278e-06,
+      "loss": 0.6369,
+      "step": 2154
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2238283157348633,
+      "learning_rate": 8.49495470647625e-06,
+      "loss": 0.68,
+      "step": 2155
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3400514125823975,
+      "learning_rate": 8.493470489570314e-06,
+      "loss": 0.6404,
+      "step": 2156
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.198529839515686,
+      "learning_rate": 8.491985670985093e-06,
+      "loss": 0.6745,
+      "step": 2157
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.067074775695801,
+      "learning_rate": 8.490500250976314e-06,
+      "loss": 0.6279,
+      "step": 2158
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.331629991531372,
+      "learning_rate": 8.489014229799816e-06,
+      "loss": 0.6106,
+      "step": 2159
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2290736436843872,
+      "learning_rate": 8.487527607711535e-06,
+      "loss": 0.5825,
+      "step": 2160
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1484627723693848,
+      "learning_rate": 8.486040384967509e-06,
+      "loss": 0.5505,
+      "step": 2161
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2274521589279175,
+      "learning_rate": 8.484552561823885e-06,
+      "loss": 0.6092,
+      "step": 2162
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2703810930252075,
+      "learning_rate": 8.483064138536906e-06,
+      "loss": 0.5541,
+      "step": 2163
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1794027090072632,
+      "learning_rate": 8.481575115362926e-06,
+      "loss": 0.6615,
+      "step": 2164
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.208961844444275,
+      "learning_rate": 8.480085492558398e-06,
+      "loss": 0.5989,
+      "step": 2165
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.370457649230957,
+      "learning_rate": 8.47859527037988e-06,
+      "loss": 0.5694,
+      "step": 2166
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.082960605621338,
+      "learning_rate": 8.47710444908403e-06,
+      "loss": 0.5625,
+      "step": 2167
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1672261953353882,
+      "learning_rate": 8.475613028927615e-06,
+      "loss": 0.7696,
+      "step": 2168
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1371452808380127,
+      "learning_rate": 8.4741210101675e-06,
+      "loss": 0.5821,
+      "step": 2169
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1140472888946533,
+      "learning_rate": 8.472628393060654e-06,
+      "loss": 0.5798,
+      "step": 2170
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0004754066467285,
+      "learning_rate": 8.471135177864152e-06,
+      "loss": 0.5552,
+      "step": 2171
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2164877653121948,
+      "learning_rate": 8.469641364835171e-06,
+      "loss": 0.5519,
+      "step": 2172
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.1060073375701904,
+      "learning_rate": 8.468146954230984e-06,
+      "loss": 0.7057,
+      "step": 2173
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1200610399246216,
+      "learning_rate": 8.466651946308979e-06,
+      "loss": 0.6117,
+      "step": 2174
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.214603304862976,
+      "learning_rate": 8.465156341326639e-06,
+      "loss": 0.6462,
+      "step": 2175
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2249237298965454,
+      "learning_rate": 8.46366013954155e-06,
+      "loss": 0.7087,
+      "step": 2176
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9753620028495789,
+      "learning_rate": 8.462163341211404e-06,
+      "loss": 0.6127,
+      "step": 2177
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.143880844116211,
+      "learning_rate": 8.460665946593994e-06,
+      "loss": 0.6184,
+      "step": 2178
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.178911566734314,
+      "learning_rate": 8.459167955947217e-06,
+      "loss": 0.6362,
+      "step": 2179
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4393823146820068,
+      "learning_rate": 8.457669369529067e-06,
+      "loss": 0.7749,
+      "step": 2180
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.1207737922668457,
+      "learning_rate": 8.456170187597647e-06,
+      "loss": 0.6919,
+      "step": 2181
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2718356847763062,
+      "learning_rate": 8.454670410411165e-06,
+      "loss": 0.5967,
+      "step": 2182
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3145051002502441,
+      "learning_rate": 8.453170038227922e-06,
+      "loss": 0.7233,
+      "step": 2183
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1284637451171875,
+      "learning_rate": 8.451669071306326e-06,
+      "loss": 0.629,
+      "step": 2184
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1894221305847168,
+      "learning_rate": 8.450167509904892e-06,
+      "loss": 0.5804,
+      "step": 2185
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2713896036148071,
+      "learning_rate": 8.448665354282233e-06,
+      "loss": 0.6278,
+      "step": 2186
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4796535968780518,
+      "learning_rate": 8.447162604697062e-06,
+      "loss": 0.6252,
+      "step": 2187
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.35587739944458,
+      "learning_rate": 8.445659261408199e-06,
+      "loss": 0.7211,
+      "step": 2188
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.649792194366455,
+      "learning_rate": 8.444155324674564e-06,
+      "loss": 0.6718,
+      "step": 2189
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4509642124176025,
+      "learning_rate": 8.442650794755178e-06,
+      "loss": 0.6138,
+      "step": 2190
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5314606428146362,
+      "learning_rate": 8.44114567190917e-06,
+      "loss": 0.6393,
+      "step": 2191
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.026334524154663,
+      "learning_rate": 8.439639956395763e-06,
+      "loss": 0.5199,
+      "step": 2192
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0074406862258911,
+      "learning_rate": 8.438133648474284e-06,
+      "loss": 0.6609,
+      "step": 2193
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.685986876487732,
+      "learning_rate": 8.43662674840417e-06,
+      "loss": 0.64,
+      "step": 2194
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3212658166885376,
+      "learning_rate": 8.435119256444948e-06,
+      "loss": 0.6855,
+      "step": 2195
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.209119439125061,
+      "learning_rate": 8.433611172856258e-06,
+      "loss": 0.7623,
+      "step": 2196
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0640321969985962,
+      "learning_rate": 8.432102497897832e-06,
+      "loss": 0.5945,
+      "step": 2197
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2218296527862549,
+      "learning_rate": 8.430593231829512e-06,
+      "loss": 0.5931,
+      "step": 2198
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5449182987213135,
+      "learning_rate": 8.429083374911238e-06,
+      "loss": 0.5279,
+      "step": 2199
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1568078994750977,
+      "learning_rate": 8.427572927403049e-06,
+      "loss": 0.5968,
+      "step": 2200
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2066130638122559,
+      "learning_rate": 8.426061889565094e-06,
+      "loss": 0.569,
+      "step": 2201
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4190328121185303,
+      "learning_rate": 8.424550261657614e-06,
+      "loss": 0.6139,
+      "step": 2202
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1991509199142456,
+      "learning_rate": 8.423038043940958e-06,
+      "loss": 0.6717,
+      "step": 2203
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6494675874710083,
+      "learning_rate": 8.421525236675577e-06,
+      "loss": 0.5841,
+      "step": 2204
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3361454010009766,
+      "learning_rate": 8.420011840122016e-06,
+      "loss": 0.6102,
+      "step": 2205
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5138863325119019,
+      "learning_rate": 8.418497854540933e-06,
+      "loss": 0.6498,
+      "step": 2206
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3878949880599976,
+      "learning_rate": 8.416983280193076e-06,
+      "loss": 0.6866,
+      "step": 2207
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4841134548187256,
+      "learning_rate": 8.415468117339302e-06,
+      "loss": 0.6434,
+      "step": 2208
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2668837308883667,
+      "learning_rate": 8.413952366240565e-06,
+      "loss": 0.6114,
+      "step": 2209
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4674394130706787,
+      "learning_rate": 8.412436027157927e-06,
+      "loss": 0.5745,
+      "step": 2210
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2802519798278809,
+      "learning_rate": 8.410919100352543e-06,
+      "loss": 0.564,
+      "step": 2211
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2879337072372437,
+      "learning_rate": 8.409401586085673e-06,
+      "loss": 0.6732,
+      "step": 2212
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2128925323486328,
+      "learning_rate": 8.407883484618679e-06,
+      "loss": 0.6827,
+      "step": 2213
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.12970769405365,
+      "learning_rate": 8.406364796213023e-06,
+      "loss": 0.6273,
+      "step": 2214
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1036797761917114,
+      "learning_rate": 8.404845521130268e-06,
+      "loss": 0.5723,
+      "step": 2215
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1636890172958374,
+      "learning_rate": 8.403325659632076e-06,
+      "loss": 0.6296,
+      "step": 2216
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6161547899246216,
+      "learning_rate": 8.401805211980215e-06,
+      "loss": 0.6374,
+      "step": 2217
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2114921808242798,
+      "learning_rate": 8.400284178436551e-06,
+      "loss": 0.6444,
+      "step": 2218
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.136370062828064,
+      "learning_rate": 8.39876255926305e-06,
+      "loss": 0.6796,
+      "step": 2219
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 10.106849670410156,
+      "learning_rate": 8.397240354721782e-06,
+      "loss": 0.6263,
+      "step": 2220
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2132163047790527,
+      "learning_rate": 8.395717565074913e-06,
+      "loss": 0.6485,
+      "step": 2221
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0989460945129395,
+      "learning_rate": 8.394194190584714e-06,
+      "loss": 0.6106,
+      "step": 2222
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1754080057144165,
+      "learning_rate": 8.392670231513557e-06,
+      "loss": 0.585,
+      "step": 2223
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3587006330490112,
+      "learning_rate": 8.39114568812391e-06,
+      "loss": 0.6329,
+      "step": 2224
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0616475343704224,
+      "learning_rate": 8.389620560678345e-06,
+      "loss": 0.6499,
+      "step": 2225
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4234185218811035,
+      "learning_rate": 8.388094849439536e-06,
+      "loss": 0.5861,
+      "step": 2226
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3070238828659058,
+      "learning_rate": 8.386568554670255e-06,
+      "loss": 0.7642,
+      "step": 2227
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2934914827346802,
+      "learning_rate": 8.385041676633375e-06,
+      "loss": 0.6081,
+      "step": 2228
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.055109739303589,
+      "learning_rate": 8.383514215591868e-06,
+      "loss": 0.5796,
+      "step": 2229
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8594862222671509,
+      "learning_rate": 8.381986171808811e-06,
+      "loss": 0.5904,
+      "step": 2230
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1786482334136963,
+      "learning_rate": 8.380457545547378e-06,
+      "loss": 0.6771,
+      "step": 2231
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8738934993743896,
+      "learning_rate": 8.378928337070844e-06,
+      "loss": 0.5631,
+      "step": 2232
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.595661997795105,
+      "learning_rate": 8.37739854664258e-06,
+      "loss": 0.5903,
+      "step": 2233
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2699490785598755,
+      "learning_rate": 8.375868174526066e-06,
+      "loss": 0.7472,
+      "step": 2234
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3659288883209229,
+      "learning_rate": 8.374337220984879e-06,
+      "loss": 0.5874,
+      "step": 2235
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3856401443481445,
+      "learning_rate": 8.372805686282688e-06,
+      "loss": 0.5405,
+      "step": 2236
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9804805517196655,
+      "learning_rate": 8.371273570683273e-06,
+      "loss": 0.6157,
+      "step": 2237
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1596951484680176,
+      "learning_rate": 8.369740874450511e-06,
+      "loss": 0.6362,
+      "step": 2238
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.279098391532898,
+      "learning_rate": 8.368207597848375e-06,
+      "loss": 0.6012,
+      "step": 2239
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4010846614837646,
+      "learning_rate": 8.36667374114094e-06,
+      "loss": 0.5266,
+      "step": 2240
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9738383293151855,
+      "learning_rate": 8.365139304592384e-06,
+      "loss": 0.6336,
+      "step": 2241
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3508878946304321,
+      "learning_rate": 8.363604288466984e-06,
+      "loss": 0.6368,
+      "step": 2242
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1931465864181519,
+      "learning_rate": 8.362068693029111e-06,
+      "loss": 0.5803,
+      "step": 2243
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0445038080215454,
+      "learning_rate": 8.360532518543241e-06,
+      "loss": 0.5703,
+      "step": 2244
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1057722568511963,
+      "learning_rate": 8.358995765273953e-06,
+      "loss": 0.6061,
+      "step": 2245
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.1264703273773193,
+      "learning_rate": 8.357458433485917e-06,
+      "loss": 0.6588,
+      "step": 2246
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4543817043304443,
+      "learning_rate": 8.355920523443909e-06,
+      "loss": 0.5947,
+      "step": 2247
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9155209064483643,
+      "learning_rate": 8.354382035412803e-06,
+      "loss": 0.5496,
+      "step": 2248
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.251805067062378,
+      "learning_rate": 8.35284296965757e-06,
+      "loss": 0.5388,
+      "step": 2249
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2307274341583252,
+      "learning_rate": 8.351303326443287e-06,
+      "loss": 0.6713,
+      "step": 2250
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3618934154510498,
+      "learning_rate": 8.349763106035123e-06,
+      "loss": 0.6442,
+      "step": 2251
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1276440620422363,
+      "learning_rate": 8.348222308698348e-06,
+      "loss": 0.6474,
+      "step": 2252
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5264719724655151,
+      "learning_rate": 8.34668093469834e-06,
+      "loss": 0.623,
+      "step": 2253
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4504443407058716,
+      "learning_rate": 8.34513898430056e-06,
+      "loss": 0.6397,
+      "step": 2254
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4074475765228271,
+      "learning_rate": 8.343596457770586e-06,
+      "loss": 0.6265,
+      "step": 2255
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2215962409973145,
+      "learning_rate": 8.342053355374082e-06,
+      "loss": 0.708,
+      "step": 2256
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6315170526504517,
+      "learning_rate": 8.340509677376817e-06,
+      "loss": 0.6566,
+      "step": 2257
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2518130540847778,
+      "learning_rate": 8.338965424044658e-06,
+      "loss": 0.7536,
+      "step": 2258
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4042388200759888,
+      "learning_rate": 8.337420595643574e-06,
+      "loss": 0.6747,
+      "step": 2259
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.046313762664795,
+      "learning_rate": 8.335875192439627e-06,
+      "loss": 0.6222,
+      "step": 2260
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1476646661758423,
+      "learning_rate": 8.33432921469898e-06,
+      "loss": 0.6875,
+      "step": 2261
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1270971298217773,
+      "learning_rate": 8.332782662687902e-06,
+      "loss": 0.5926,
+      "step": 2262
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2763153314590454,
+      "learning_rate": 8.331235536672748e-06,
+      "loss": 0.5782,
+      "step": 2263
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3246185779571533,
+      "learning_rate": 8.329687836919986e-06,
+      "loss": 0.6623,
+      "step": 2264
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4971126317977905,
+      "learning_rate": 8.328139563696172e-06,
+      "loss": 0.6468,
+      "step": 2265
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.484387993812561,
+      "learning_rate": 8.326590717267964e-06,
+      "loss": 0.6747,
+      "step": 2266
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.130797266960144,
+      "learning_rate": 8.32504129790212e-06,
+      "loss": 0.6821,
+      "step": 2267
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0058528184890747,
+      "learning_rate": 8.323491305865498e-06,
+      "loss": 0.5863,
+      "step": 2268
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.406995415687561,
+      "learning_rate": 8.321940741425049e-06,
+      "loss": 0.6819,
+      "step": 2269
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8498761653900146,
+      "learning_rate": 8.32038960484783e-06,
+      "loss": 0.7176,
+      "step": 2270
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4135112762451172,
+      "learning_rate": 8.31883789640099e-06,
+      "loss": 0.5951,
+      "step": 2271
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5905379056930542,
+      "learning_rate": 8.317285616351782e-06,
+      "loss": 0.7059,
+      "step": 2272
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0527902841567993,
+      "learning_rate": 8.315732764967552e-06,
+      "loss": 0.6129,
+      "step": 2273
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.336744785308838,
+      "learning_rate": 8.314179342515746e-06,
+      "loss": 0.6918,
+      "step": 2274
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7440773248672485,
+      "learning_rate": 8.312625349263914e-06,
+      "loss": 0.6386,
+      "step": 2275
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.086266279220581,
+      "learning_rate": 8.311070785479699e-06,
+      "loss": 0.6303,
+      "step": 2276
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1209901571273804,
+      "learning_rate": 8.309515651430837e-06,
+      "loss": 0.6507,
+      "step": 2277
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4283497333526611,
+      "learning_rate": 8.307959947385174e-06,
+      "loss": 0.6424,
+      "step": 2278
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2685678005218506,
+      "learning_rate": 8.306403673610646e-06,
+      "loss": 0.5953,
+      "step": 2279
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.394755482673645,
+      "learning_rate": 8.304846830375294e-06,
+      "loss": 0.5581,
+      "step": 2280
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.204455852508545,
+      "learning_rate": 8.303289417947244e-06,
+      "loss": 0.5786,
+      "step": 2281
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3802586793899536,
+      "learning_rate": 8.301731436594737e-06,
+      "loss": 0.5582,
+      "step": 2282
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6759288311004639,
+      "learning_rate": 8.3001728865861e-06,
+      "loss": 0.6505,
+      "step": 2283
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.07804536819458,
+      "learning_rate": 8.298613768189761e-06,
+      "loss": 0.6687,
+      "step": 2284
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4512630701065063,
+      "learning_rate": 8.297054081674247e-06,
+      "loss": 0.6112,
+      "step": 2285
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.449326753616333,
+      "learning_rate": 8.29549382730818e-06,
+      "loss": 0.628,
+      "step": 2286
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.163280725479126,
+      "learning_rate": 8.293933005360288e-06,
+      "loss": 0.6082,
+      "step": 2287
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.181201696395874,
+      "learning_rate": 8.292371616099388e-06,
+      "loss": 0.6441,
+      "step": 2288
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1229546070098877,
+      "learning_rate": 8.290809659794397e-06,
+      "loss": 0.6509,
+      "step": 2289
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2174752950668335,
+      "learning_rate": 8.289247136714328e-06,
+      "loss": 0.6646,
+      "step": 2290
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.106174111366272,
+      "learning_rate": 8.287684047128298e-06,
+      "loss": 0.6211,
+      "step": 2291
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5392602682113647,
+      "learning_rate": 8.286120391305515e-06,
+      "loss": 0.5977,
+      "step": 2292
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2090442180633545,
+      "learning_rate": 8.284556169515287e-06,
+      "loss": 0.6219,
+      "step": 2293
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6874985694885254,
+      "learning_rate": 8.282991382027022e-06,
+      "loss": 0.6698,
+      "step": 2294
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1028480529785156,
+      "learning_rate": 8.281426029110218e-06,
+      "loss": 0.6597,
+      "step": 2295
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1251074075698853,
+      "learning_rate": 8.279860111034478e-06,
+      "loss": 0.6606,
+      "step": 2296
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.540422797203064,
+      "learning_rate": 8.278293628069502e-06,
+      "loss": 0.6238,
+      "step": 2297
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1604392528533936,
+      "learning_rate": 8.276726580485082e-06,
+      "loss": 0.6685,
+      "step": 2298
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4422894716262817,
+      "learning_rate": 8.27515896855111e-06,
+      "loss": 0.6697,
+      "step": 2299
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4221534729003906,
+      "learning_rate": 8.273590792537574e-06,
+      "loss": 0.6167,
+      "step": 2300
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.291616678237915,
+      "learning_rate": 8.272022052714563e-06,
+      "loss": 0.6258,
+      "step": 2301
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9711089134216309,
+      "learning_rate": 8.270452749352261e-06,
+      "loss": 0.5395,
+      "step": 2302
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.145838975906372,
+      "learning_rate": 8.268882882720946e-06,
+      "loss": 0.5787,
+      "step": 2303
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0817339420318604,
+      "learning_rate": 8.267312453090997e-06,
+      "loss": 0.6168,
+      "step": 2304
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2782442569732666,
+      "learning_rate": 8.26574146073289e-06,
+      "loss": 0.673,
+      "step": 2305
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1187020540237427,
+      "learning_rate": 8.264169905917194e-06,
+      "loss": 0.6032,
+      "step": 2306
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3407741785049438,
+      "learning_rate": 8.262597788914579e-06,
+      "loss": 0.6716,
+      "step": 2307
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4748417139053345,
+      "learning_rate": 8.26102510999581e-06,
+      "loss": 0.6684,
+      "step": 2308
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8912734985351562,
+      "learning_rate": 8.259451869431746e-06,
+      "loss": 0.5978,
+      "step": 2309
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9286829233169556,
+      "learning_rate": 8.257878067493351e-06,
+      "loss": 0.6408,
+      "step": 2310
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2534960508346558,
+      "learning_rate": 8.256303704451679e-06,
+      "loss": 0.5862,
+      "step": 2311
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3470648527145386,
+      "learning_rate": 8.254728780577877e-06,
+      "loss": 0.6486,
+      "step": 2312
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1892026662826538,
+      "learning_rate": 8.253153296143199e-06,
+      "loss": 0.6312,
+      "step": 2313
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.13631534576416,
+      "learning_rate": 8.251577251418987e-06,
+      "loss": 0.5599,
+      "step": 2314
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4602019786834717,
+      "learning_rate": 8.250000646676688e-06,
+      "loss": 0.638,
+      "step": 2315
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0925565958023071,
+      "learning_rate": 8.248423482187832e-06,
+      "loss": 0.6265,
+      "step": 2316
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1063237190246582,
+      "learning_rate": 8.246845758224062e-06,
+      "loss": 0.6283,
+      "step": 2317
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2904975414276123,
+      "learning_rate": 8.245267475057102e-06,
+      "loss": 0.6167,
+      "step": 2318
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1062041521072388,
+      "learning_rate": 8.243688632958783e-06,
+      "loss": 0.6804,
+      "step": 2319
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2432068586349487,
+      "learning_rate": 8.242109232201028e-06,
+      "loss": 0.6152,
+      "step": 2320
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.199753761291504,
+      "learning_rate": 8.240529273055852e-06,
+      "loss": 0.6969,
+      "step": 2321
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.155482769012451,
+      "learning_rate": 8.238948755795379e-06,
+      "loss": 0.69,
+      "step": 2322
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2352306842803955,
+      "learning_rate": 8.237367680691817e-06,
+      "loss": 0.6873,
+      "step": 2323
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3099703788757324,
+      "learning_rate": 8.235786048017473e-06,
+      "loss": 0.6747,
+      "step": 2324
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2565714120864868,
+      "learning_rate": 8.234203858044751e-06,
+      "loss": 0.6278,
+      "step": 2325
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7045984268188477,
+      "learning_rate": 8.232621111046154e-06,
+      "loss": 0.6427,
+      "step": 2326
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1133400201797485,
+      "learning_rate": 8.231037807294275e-06,
+      "loss": 0.6679,
+      "step": 2327
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3851282596588135,
+      "learning_rate": 8.229453947061807e-06,
+      "loss": 0.6312,
+      "step": 2328
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1454262733459473,
+      "learning_rate": 8.227869530621538e-06,
+      "loss": 0.6135,
+      "step": 2329
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3275219202041626,
+      "learning_rate": 8.226284558246351e-06,
+      "loss": 0.6552,
+      "step": 2330
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5105386972427368,
+      "learning_rate": 8.224699030209227e-06,
+      "loss": 0.6731,
+      "step": 2331
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1528339385986328,
+      "learning_rate": 8.223112946783237e-06,
+      "loss": 0.5775,
+      "step": 2332
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2737232446670532,
+      "learning_rate": 8.221526308241556e-06,
+      "loss": 0.5996,
+      "step": 2333
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.409363031387329,
+      "learning_rate": 8.219939114857446e-06,
+      "loss": 0.6846,
+      "step": 2334
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1390594244003296,
+      "learning_rate": 8.218351366904273e-06,
+      "loss": 0.6503,
+      "step": 2335
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7003358602523804,
+      "learning_rate": 8.216763064655493e-06,
+      "loss": 0.6394,
+      "step": 2336
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6137683391571045,
+      "learning_rate": 8.215174208384658e-06,
+      "loss": 0.5789,
+      "step": 2337
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.193477988243103,
+      "learning_rate": 8.213584798365416e-06,
+      "loss": 0.5051,
+      "step": 2338
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2043871879577637,
+      "learning_rate": 8.211994834871511e-06,
+      "loss": 0.5932,
+      "step": 2339
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.538806676864624,
+      "learning_rate": 8.21040431817678e-06,
+      "loss": 0.6492,
+      "step": 2340
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2634645700454712,
+      "learning_rate": 8.208813248555163e-06,
+      "loss": 0.6181,
+      "step": 2341
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2859795093536377,
+      "learning_rate": 8.207221626280683e-06,
+      "loss": 0.6666,
+      "step": 2342
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1655632257461548,
+      "learning_rate": 8.205629451627469e-06,
+      "loss": 0.6052,
+      "step": 2343
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 6.354539394378662,
+      "learning_rate": 8.204036724869737e-06,
+      "loss": 0.6652,
+      "step": 2344
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6453803777694702,
+      "learning_rate": 8.202443446281804e-06,
+      "loss": 0.5975,
+      "step": 2345
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1627188920974731,
+      "learning_rate": 8.20084961613808e-06,
+      "loss": 0.6252,
+      "step": 2346
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9727002382278442,
+      "learning_rate": 8.199255234713068e-06,
+      "loss": 0.5488,
+      "step": 2347
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.367080807685852,
+      "learning_rate": 8.197660302281371e-06,
+      "loss": 0.59,
+      "step": 2348
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4162267446517944,
+      "learning_rate": 8.196064819117681e-06,
+      "loss": 0.6033,
+      "step": 2349
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5463422536849976,
+      "learning_rate": 8.194468785496788e-06,
+      "loss": 0.6141,
+      "step": 2350
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.220804214477539,
+      "learning_rate": 8.192872201693575e-06,
+      "loss": 0.6256,
+      "step": 2351
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.263001799583435,
+      "learning_rate": 8.191275067983026e-06,
+      "loss": 0.6838,
+      "step": 2352
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9969754815101624,
+      "learning_rate": 8.189677384640212e-06,
+      "loss": 0.6114,
+      "step": 2353
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2792723178863525,
+      "learning_rate": 8.188079151940299e-06,
+      "loss": 0.6809,
+      "step": 2354
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3698155879974365,
+      "learning_rate": 8.186480370158552e-06,
+      "loss": 0.6168,
+      "step": 2355
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3380119800567627,
+      "learning_rate": 8.18488103957033e-06,
+      "loss": 0.6018,
+      "step": 2356
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2656079530715942,
+      "learning_rate": 8.183281160451083e-06,
+      "loss": 0.5869,
+      "step": 2357
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.461011290550232,
+      "learning_rate": 8.181680733076359e-06,
+      "loss": 0.5784,
+      "step": 2358
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3184751272201538,
+      "learning_rate": 8.180079757721799e-06,
+      "loss": 0.6599,
+      "step": 2359
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.169732689857483,
+      "learning_rate": 8.178478234663139e-06,
+      "loss": 0.6553,
+      "step": 2360
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2968778610229492,
+      "learning_rate": 8.176876164176206e-06,
+      "loss": 0.669,
+      "step": 2361
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.166947603225708,
+      "learning_rate": 8.175273546536929e-06,
+      "loss": 0.6507,
+      "step": 2362
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2829275131225586,
+      "learning_rate": 8.17367038202132e-06,
+      "loss": 0.5774,
+      "step": 2363
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1029894351959229,
+      "learning_rate": 8.172066670905498e-06,
+      "loss": 0.6448,
+      "step": 2364
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3434944152832031,
+      "learning_rate": 8.170462413465666e-06,
+      "loss": 0.6137,
+      "step": 2365
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5390658378601074,
+      "learning_rate": 8.168857609978125e-06,
+      "loss": 0.647,
+      "step": 2366
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3299118280410767,
+      "learning_rate": 8.16725226071927e-06,
+      "loss": 0.6634,
+      "step": 2367
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6107890605926514,
+      "learning_rate": 8.165646365965589e-06,
+      "loss": 0.5927,
+      "step": 2368
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2683249711990356,
+      "learning_rate": 8.164039925993667e-06,
+      "loss": 0.6726,
+      "step": 2369
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6438405513763428,
+      "learning_rate": 8.162432941080178e-06,
+      "loss": 0.619,
+      "step": 2370
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1879276037216187,
+      "learning_rate": 8.160825411501896e-06,
+      "loss": 0.5904,
+      "step": 2371
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1308679580688477,
+      "learning_rate": 8.159217337535682e-06,
+      "loss": 0.6671,
+      "step": 2372
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3544384241104126,
+      "learning_rate": 8.157608719458493e-06,
+      "loss": 0.5496,
+      "step": 2373
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 6.0907440185546875,
+      "learning_rate": 8.155999557547384e-06,
+      "loss": 0.6856,
+      "step": 2374
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3868075609207153,
+      "learning_rate": 8.154389852079501e-06,
+      "loss": 0.6667,
+      "step": 2375
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0029851198196411,
+      "learning_rate": 8.15277960333208e-06,
+      "loss": 0.6324,
+      "step": 2376
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.873964548110962,
+      "learning_rate": 8.151168811582455e-06,
+      "loss": 0.6404,
+      "step": 2377
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0714986324310303,
+      "learning_rate": 8.149557477108051e-06,
+      "loss": 0.6066,
+      "step": 2378
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5956631898880005,
+      "learning_rate": 8.147945600186391e-06,
+      "loss": 0.6414,
+      "step": 2379
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2539238929748535,
+      "learning_rate": 8.146333181095086e-06,
+      "loss": 0.5592,
+      "step": 2380
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2881016731262207,
+      "learning_rate": 8.14472022011184e-06,
+      "loss": 0.5886,
+      "step": 2381
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.226897954940796,
+      "learning_rate": 8.143106717514455e-06,
+      "loss": 0.7155,
+      "step": 2382
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4589011669158936,
+      "learning_rate": 8.141492673580825e-06,
+      "loss": 0.7016,
+      "step": 2383
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2750200033187866,
+      "learning_rate": 8.139878088588934e-06,
+      "loss": 0.6133,
+      "step": 2384
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2676771879196167,
+      "learning_rate": 8.138262962816865e-06,
+      "loss": 0.6297,
+      "step": 2385
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.572501301765442,
+      "learning_rate": 8.136647296542787e-06,
+      "loss": 0.6115,
+      "step": 2386
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.544345736503601,
+      "learning_rate": 8.135031090044966e-06,
+      "loss": 0.6698,
+      "step": 2387
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.247854232788086,
+      "learning_rate": 8.133414343601762e-06,
+      "loss": 0.6356,
+      "step": 2388
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2449547052383423,
+      "learning_rate": 8.131797057491627e-06,
+      "loss": 0.6794,
+      "step": 2389
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5108448266983032,
+      "learning_rate": 8.130179231993105e-06,
+      "loss": 0.6564,
+      "step": 2390
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2895339727401733,
+      "learning_rate": 8.128560867384832e-06,
+      "loss": 0.5596,
+      "step": 2391
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3409855365753174,
+      "learning_rate": 8.126941963945541e-06,
+      "loss": 0.6147,
+      "step": 2392
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2308881282806396,
+      "learning_rate": 8.125322521954055e-06,
+      "loss": 0.5847,
+      "step": 2393
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4125090837478638,
+      "learning_rate": 8.12370254168929e-06,
+      "loss": 0.6779,
+      "step": 2394
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1895586252212524,
+      "learning_rate": 8.122082023430251e-06,
+      "loss": 0.7764,
+      "step": 2395
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.083396077156067,
+      "learning_rate": 8.120460967456043e-06,
+      "loss": 0.6141,
+      "step": 2396
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1598103046417236,
+      "learning_rate": 8.118839374045861e-06,
+      "loss": 0.612,
+      "step": 2397
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1906448602676392,
+      "learning_rate": 8.117217243478988e-06,
+      "loss": 0.6277,
+      "step": 2398
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1712722778320312,
+      "learning_rate": 8.115594576034804e-06,
+      "loss": 0.7004,
+      "step": 2399
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3918662071228027,
+      "learning_rate": 8.113971371992782e-06,
+      "loss": 0.5189,
+      "step": 2400
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.191245436668396,
+      "learning_rate": 8.112347631632484e-06,
+      "loss": 0.6203,
+      "step": 2401
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.500771403312683,
+      "learning_rate": 8.11072335523357e-06,
+      "loss": 0.5052,
+      "step": 2402
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3032631874084473,
+      "learning_rate": 8.109098543075784e-06,
+      "loss": 0.6056,
+      "step": 2403
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1832846403121948,
+      "learning_rate": 8.107473195438969e-06,
+      "loss": 0.6169,
+      "step": 2404
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2408926486968994,
+      "learning_rate": 8.105847312603057e-06,
+      "loss": 0.5983,
+      "step": 2405
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0566129684448242,
+      "learning_rate": 8.104220894848073e-06,
+      "loss": 0.6529,
+      "step": 2406
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.521499752998352,
+      "learning_rate": 8.102593942454138e-06,
+      "loss": 0.6247,
+      "step": 2407
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2609105110168457,
+      "learning_rate": 8.100966455701458e-06,
+      "loss": 0.551,
+      "step": 2408
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3678559064865112,
+      "learning_rate": 8.099338434870336e-06,
+      "loss": 0.6309,
+      "step": 2409
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0221734046936035,
+      "learning_rate": 8.097709880241165e-06,
+      "loss": 0.5807,
+      "step": 2410
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6654701232910156,
+      "learning_rate": 8.09608079209443e-06,
+      "loss": 0.6289,
+      "step": 2411
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2061890363693237,
+      "learning_rate": 8.094451170710708e-06,
+      "loss": 0.6315,
+      "step": 2412
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0013035535812378,
+      "learning_rate": 8.09282101637067e-06,
+      "loss": 0.5325,
+      "step": 2413
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.319288730621338,
+      "learning_rate": 8.091190329355076e-06,
+      "loss": 0.6432,
+      "step": 2414
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1914440393447876,
+      "learning_rate": 8.089559109944777e-06,
+      "loss": 0.5529,
+      "step": 2415
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.162363886833191,
+      "learning_rate": 8.087927358420723e-06,
+      "loss": 0.6628,
+      "step": 2416
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.818101406097412,
+      "learning_rate": 8.086295075063942e-06,
+      "loss": 0.6276,
+      "step": 2417
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2434170246124268,
+      "learning_rate": 8.084662260155567e-06,
+      "loss": 0.6482,
+      "step": 2418
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9897717833518982,
+      "learning_rate": 8.083028913976816e-06,
+      "loss": 0.6063,
+      "step": 2419
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3943613767623901,
+      "learning_rate": 8.081395036808999e-06,
+      "loss": 0.6482,
+      "step": 2420
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2112592458724976,
+      "learning_rate": 8.079760628933518e-06,
+      "loss": 0.6589,
+      "step": 2421
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2801096439361572,
+      "learning_rate": 8.078125690631868e-06,
+      "loss": 0.5905,
+      "step": 2422
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4379853010177612,
+      "learning_rate": 8.076490222185631e-06,
+      "loss": 0.5817,
+      "step": 2423
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9312883615493774,
+      "learning_rate": 8.074854223876487e-06,
+      "loss": 0.6578,
+      "step": 2424
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.0586395263671875,
+      "learning_rate": 8.073217695986203e-06,
+      "loss": 0.5962,
+      "step": 2425
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.1898674964904785,
+      "learning_rate": 8.071580638796634e-06,
+      "loss": 0.6595,
+      "step": 2426
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2980766296386719,
+      "learning_rate": 8.069943052589734e-06,
+      "loss": 0.6066,
+      "step": 2427
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2575178146362305,
+      "learning_rate": 8.068304937647542e-06,
+      "loss": 0.5541,
+      "step": 2428
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1975334882736206,
+      "learning_rate": 8.066666294252189e-06,
+      "loss": 0.6744,
+      "step": 2429
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5488165616989136,
+      "learning_rate": 8.0650271226859e-06,
+      "loss": 0.6803,
+      "step": 2430
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2515982389450073,
+      "learning_rate": 8.063387423230987e-06,
+      "loss": 0.6455,
+      "step": 2431
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3792338371276855,
+      "learning_rate": 8.061747196169855e-06,
+      "loss": 0.5893,
+      "step": 2432
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6035951375961304,
+      "learning_rate": 8.060106441785003e-06,
+      "loss": 0.5758,
+      "step": 2433
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.494800090789795,
+      "learning_rate": 8.058465160359016e-06,
+      "loss": 0.6529,
+      "step": 2434
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2441433668136597,
+      "learning_rate": 8.05682335217457e-06,
+      "loss": 0.5876,
+      "step": 2435
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5442471504211426,
+      "learning_rate": 8.055181017514432e-06,
+      "loss": 0.6346,
+      "step": 2436
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3310129642486572,
+      "learning_rate": 8.053538156661465e-06,
+      "loss": 0.6639,
+      "step": 2437
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.222756028175354,
+      "learning_rate": 8.051894769898615e-06,
+      "loss": 0.6473,
+      "step": 2438
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3063236474990845,
+      "learning_rate": 8.050250857508923e-06,
+      "loss": 0.7353,
+      "step": 2439
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1706254482269287,
+      "learning_rate": 8.04860641977552e-06,
+      "loss": 0.6611,
+      "step": 2440
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5362704992294312,
+      "learning_rate": 8.046961456981625e-06,
+      "loss": 0.6244,
+      "step": 2441
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1532371044158936,
+      "learning_rate": 8.045315969410551e-06,
+      "loss": 0.6642,
+      "step": 2442
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4653925895690918,
+      "learning_rate": 8.043669957345701e-06,
+      "loss": 0.6008,
+      "step": 2443
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.7584710121154785,
+      "learning_rate": 8.042023421070566e-06,
+      "loss": 0.6796,
+      "step": 2444
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.38533616065979,
+      "learning_rate": 8.040376360868727e-06,
+      "loss": 0.7932,
+      "step": 2445
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.1703217029571533,
+      "learning_rate": 8.038728777023858e-06,
+      "loss": 0.5941,
+      "step": 2446
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.146225452423096,
+      "learning_rate": 8.037080669819723e-06,
+      "loss": 0.6693,
+      "step": 2447
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3640509843826294,
+      "learning_rate": 8.035432039540172e-06,
+      "loss": 0.7157,
+      "step": 2448
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1412841081619263,
+      "learning_rate": 8.03378288646915e-06,
+      "loss": 0.4853,
+      "step": 2449
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9124466180801392,
+      "learning_rate": 8.03213321089069e-06,
+      "loss": 0.5493,
+      "step": 2450
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4567307233810425,
+      "learning_rate": 8.030483013088913e-06,
+      "loss": 0.6635,
+      "step": 2451
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.292675495147705,
+      "learning_rate": 8.028832293348036e-06,
+      "loss": 0.7629,
+      "step": 2452
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2411820888519287,
+      "learning_rate": 8.027181051952359e-06,
+      "loss": 0.6375,
+      "step": 2453
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1143420934677124,
+      "learning_rate": 8.025529289186276e-06,
+      "loss": 0.5945,
+      "step": 2454
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4121794700622559,
+      "learning_rate": 8.023877005334268e-06,
+      "loss": 0.5983,
+      "step": 2455
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5333738327026367,
+      "learning_rate": 8.022224200680911e-06,
+      "loss": 0.5918,
+      "step": 2456
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6830614805221558,
+      "learning_rate": 8.020570875510863e-06,
+      "loss": 0.5903,
+      "step": 2457
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3144832849502563,
+      "learning_rate": 8.018917030108874e-06,
+      "loss": 0.665,
+      "step": 2458
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.2713377475738525,
+      "learning_rate": 8.017262664759793e-06,
+      "loss": 0.6369,
+      "step": 2459
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5499030351638794,
+      "learning_rate": 8.015607779748544e-06,
+      "loss": 0.5759,
+      "step": 2460
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8569972515106201,
+      "learning_rate": 8.01395237536015e-06,
+      "loss": 0.6703,
+      "step": 2461
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4563158750534058,
+      "learning_rate": 8.01229645187972e-06,
+      "loss": 0.6818,
+      "step": 2462
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1788994073867798,
+      "learning_rate": 8.010640009592454e-06,
+      "loss": 0.6173,
+      "step": 2463
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.48537015914917,
+      "learning_rate": 8.008983048783639e-06,
+      "loss": 0.5923,
+      "step": 2464
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9768660068511963,
+      "learning_rate": 8.007325569738654e-06,
+      "loss": 0.6231,
+      "step": 2465
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.359007716178894,
+      "learning_rate": 8.005667572742964e-06,
+      "loss": 0.6858,
+      "step": 2466
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6291216611862183,
+      "learning_rate": 8.00400905808213e-06,
+      "loss": 0.6691,
+      "step": 2467
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2007826566696167,
+      "learning_rate": 8.002350026041792e-06,
+      "loss": 0.6349,
+      "step": 2468
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9453998804092407,
+      "learning_rate": 8.000690476907688e-06,
+      "loss": 0.5912,
+      "step": 2469
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2745121717453003,
+      "learning_rate": 7.999030410965642e-06,
+      "loss": 0.6899,
+      "step": 2470
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5417898893356323,
+      "learning_rate": 7.997369828501565e-06,
+      "loss": 0.5747,
+      "step": 2471
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.322919249534607,
+      "learning_rate": 7.995708729801459e-06,
+      "loss": 0.5912,
+      "step": 2472
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.1955883502960205,
+      "learning_rate": 7.994047115151414e-06,
+      "loss": 0.5515,
+      "step": 2473
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2016857862472534,
+      "learning_rate": 7.992384984837608e-06,
+      "loss": 0.5988,
+      "step": 2474
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1391526460647583,
+      "learning_rate": 7.99072233914631e-06,
+      "loss": 0.7175,
+      "step": 2475
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.383262276649475,
+      "learning_rate": 7.98905917836388e-06,
+      "loss": 0.5499,
+      "step": 2476
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3291974067687988,
+      "learning_rate": 7.987395502776762e-06,
+      "loss": 0.5684,
+      "step": 2477
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3303680419921875,
+      "learning_rate": 7.98573131267149e-06,
+      "loss": 0.6121,
+      "step": 2478
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3416320085525513,
+      "learning_rate": 7.984066608334684e-06,
+      "loss": 0.58,
+      "step": 2479
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2517801523208618,
+      "learning_rate": 7.982401390053061e-06,
+      "loss": 0.6141,
+      "step": 2480
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.685667634010315,
+      "learning_rate": 7.980735658113416e-06,
+      "loss": 0.5259,
+      "step": 2481
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3155723810195923,
+      "learning_rate": 7.97906941280264e-06,
+      "loss": 0.5923,
+      "step": 2482
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6275346279144287,
+      "learning_rate": 7.97740265440771e-06,
+      "loss": 0.6981,
+      "step": 2483
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.539910078048706,
+      "learning_rate": 7.975735383215691e-06,
+      "loss": 0.7108,
+      "step": 2484
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0558339357376099,
+      "learning_rate": 7.974067599513737e-06,
+      "loss": 0.6255,
+      "step": 2485
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4272794723510742,
+      "learning_rate": 7.972399303589087e-06,
+      "loss": 0.646,
+      "step": 2486
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5255314111709595,
+      "learning_rate": 7.970730495729075e-06,
+      "loss": 0.6481,
+      "step": 2487
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7075724601745605,
+      "learning_rate": 7.969061176221118e-06,
+      "loss": 0.6343,
+      "step": 2488
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2317755222320557,
+      "learning_rate": 7.96739134535272e-06,
+      "loss": 0.7467,
+      "step": 2489
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4913558959960938,
+      "learning_rate": 7.965721003411477e-06,
+      "loss": 0.6367,
+      "step": 2490
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5637904405593872,
+      "learning_rate": 7.964050150685075e-06,
+      "loss": 0.6279,
+      "step": 2491
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2378253936767578,
+      "learning_rate": 7.962378787461278e-06,
+      "loss": 0.6074,
+      "step": 2492
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2458785772323608,
+      "learning_rate": 7.960706914027947e-06,
+      "loss": 0.6126,
+      "step": 2493
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1778990030288696,
+      "learning_rate": 7.95903453067303e-06,
+      "loss": 0.6499,
+      "step": 2494
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3699263334274292,
+      "learning_rate": 7.95736163768456e-06,
+      "loss": 0.6635,
+      "step": 2495
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1903271675109863,
+      "learning_rate": 7.955688235350659e-06,
+      "loss": 0.6467,
+      "step": 2496
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9940861463546753,
+      "learning_rate": 7.954014323959535e-06,
+      "loss": 0.5876,
+      "step": 2497
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2723509073257446,
+      "learning_rate": 7.952339903799486e-06,
+      "loss": 0.6072,
+      "step": 2498
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.031148076057434,
+      "learning_rate": 7.950664975158898e-06,
+      "loss": 0.6026,
+      "step": 2499
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7933772802352905,
+      "learning_rate": 7.948989538326241e-06,
+      "loss": 0.7027,
+      "step": 2500
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2486510276794434,
+      "learning_rate": 7.947313593590078e-06,
+      "loss": 0.5948,
+      "step": 2501
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5051732063293457,
+      "learning_rate": 7.945637141239054e-06,
+      "loss": 0.6531,
+      "step": 2502
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.544805884361267,
+      "learning_rate": 7.943960181561905e-06,
+      "loss": 0.6543,
+      "step": 2503
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2853175401687622,
+      "learning_rate": 7.942282714847453e-06,
+      "loss": 0.6587,
+      "step": 2504
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3751949071884155,
+      "learning_rate": 7.940604741384607e-06,
+      "loss": 0.5903,
+      "step": 2505
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3573917150497437,
+      "learning_rate": 7.938926261462366e-06,
+      "loss": 0.6358,
+      "step": 2506
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9741413593292236,
+      "learning_rate": 7.937247275369813e-06,
+      "loss": 0.61,
+      "step": 2507
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3532620668411255,
+      "learning_rate": 7.935567783396116e-06,
+      "loss": 0.5756,
+      "step": 2508
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2103369235992432,
+      "learning_rate": 7.933887785830536e-06,
+      "loss": 0.6417,
+      "step": 2509
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3352247476577759,
+      "learning_rate": 7.93220728296242e-06,
+      "loss": 0.6468,
+      "step": 2510
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1334822177886963,
+      "learning_rate": 7.930526275081198e-06,
+      "loss": 0.629,
+      "step": 2511
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.796984314918518,
+      "learning_rate": 7.92884476247639e-06,
+      "loss": 0.637,
+      "step": 2512
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3944237232208252,
+      "learning_rate": 7.927162745437605e-06,
+      "loss": 0.602,
+      "step": 2513
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1634482145309448,
+      "learning_rate": 7.92548022425453e-06,
+      "loss": 0.7354,
+      "step": 2514
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0452791452407837,
+      "learning_rate": 7.923797199216952e-06,
+      "loss": 0.5221,
+      "step": 2515
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.141053557395935,
+      "learning_rate": 7.922113670614733e-06,
+      "loss": 0.6103,
+      "step": 2516
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.371309518814087,
+      "learning_rate": 7.92042963873783e-06,
+      "loss": 0.5625,
+      "step": 2517
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4770622253417969,
+      "learning_rate": 7.91874510387628e-06,
+      "loss": 0.6978,
+      "step": 2518
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2017892599105835,
+      "learning_rate": 7.917060066320213e-06,
+      "loss": 0.5845,
+      "step": 2519
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2020307779312134,
+      "learning_rate": 7.91537452635984e-06,
+      "loss": 0.5964,
+      "step": 2520
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0251448154449463,
+      "learning_rate": 7.913688484285462e-06,
+      "loss": 0.6008,
+      "step": 2521
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8153685331344604,
+      "learning_rate": 7.912001940387466e-06,
+      "loss": 0.736,
+      "step": 2522
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.3994553089141846,
+      "learning_rate": 7.910314894956326e-06,
+      "loss": 0.6798,
+      "step": 2523
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.541852593421936,
+      "learning_rate": 7.908627348282599e-06,
+      "loss": 0.6723,
+      "step": 2524
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0899004936218262,
+      "learning_rate": 7.906939300656929e-06,
+      "loss": 0.6825,
+      "step": 2525
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5158523321151733,
+      "learning_rate": 7.905250752370051e-06,
+      "loss": 0.5933,
+      "step": 2526
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0821444988250732,
+      "learning_rate": 7.903561703712784e-06,
+      "loss": 0.5961,
+      "step": 2527
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3640822172164917,
+      "learning_rate": 7.90187215497603e-06,
+      "loss": 0.6481,
+      "step": 2528
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0577113628387451,
+      "learning_rate": 7.900182106450778e-06,
+      "loss": 0.6615,
+      "step": 2529
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1909829378128052,
+      "learning_rate": 7.898491558428108e-06,
+      "loss": 0.6597,
+      "step": 2530
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1416188478469849,
+      "learning_rate": 7.896800511199182e-06,
+      "loss": 0.7269,
+      "step": 2531
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.210830807685852,
+      "learning_rate": 7.895108965055247e-06,
+      "loss": 0.6678,
+      "step": 2532
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9497708678245544,
+      "learning_rate": 7.893416920287638e-06,
+      "loss": 0.5491,
+      "step": 2533
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2646074295043945,
+      "learning_rate": 7.891724377187774e-06,
+      "loss": 0.5835,
+      "step": 2534
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4362199306488037,
+      "learning_rate": 7.890031336047166e-06,
+      "loss": 0.6361,
+      "step": 2535
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1478391885757446,
+      "learning_rate": 7.8883377971574e-06,
+      "loss": 0.6115,
+      "step": 2536
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1613198518753052,
+      "learning_rate": 7.886643760810156e-06,
+      "loss": 0.6171,
+      "step": 2537
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.247196912765503,
+      "learning_rate": 7.884949227297199e-06,
+      "loss": 0.6298,
+      "step": 2538
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3772872686386108,
+      "learning_rate": 7.883254196910375e-06,
+      "loss": 0.5992,
+      "step": 2539
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6356278657913208,
+      "learning_rate": 7.88155866994162e-06,
+      "loss": 0.6667,
+      "step": 2540
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2067662477493286,
+      "learning_rate": 7.879862646682955e-06,
+      "loss": 0.6492,
+      "step": 2541
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1001278162002563,
+      "learning_rate": 7.878166127426483e-06,
+      "loss": 0.6088,
+      "step": 2542
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9857989549636841,
+      "learning_rate": 7.876469112464395e-06,
+      "loss": 0.6095,
+      "step": 2543
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3550429344177246,
+      "learning_rate": 7.874771602088971e-06,
+      "loss": 0.6157,
+      "step": 2544
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9245561361312866,
+      "learning_rate": 7.873073596592571e-06,
+      "loss": 0.5456,
+      "step": 2545
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.329111099243164,
+      "learning_rate": 7.871375096267641e-06,
+      "loss": 0.6259,
+      "step": 2546
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0523360967636108,
+      "learning_rate": 7.869676101406713e-06,
+      "loss": 0.6267,
+      "step": 2547
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.33540940284729,
+      "learning_rate": 7.867976612302405e-06,
+      "loss": 0.6859,
+      "step": 2548
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6070891618728638,
+      "learning_rate": 7.86627662924742e-06,
+      "loss": 0.6134,
+      "step": 2549
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.088706374168396,
+      "learning_rate": 7.864576152534544e-06,
+      "loss": 0.6603,
+      "step": 2550
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3226513862609863,
+      "learning_rate": 7.862875182456652e-06,
+      "loss": 0.5793,
+      "step": 2551
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0518193244934082,
+      "learning_rate": 7.861173719306697e-06,
+      "loss": 0.5619,
+      "step": 2552
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0934220552444458,
+      "learning_rate": 7.859471763377726e-06,
+      "loss": 0.6339,
+      "step": 2553
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7915741205215454,
+      "learning_rate": 7.857769314962865e-06,
+      "loss": 0.6733,
+      "step": 2554
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3046362400054932,
+      "learning_rate": 7.856066374355326e-06,
+      "loss": 0.6231,
+      "step": 2555
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4739090204238892,
+      "learning_rate": 7.854362941848406e-06,
+      "loss": 0.6874,
+      "step": 2556
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2384624481201172,
+      "learning_rate": 7.852659017735484e-06,
+      "loss": 0.642,
+      "step": 2557
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.99631267786026,
+      "learning_rate": 7.850954602310032e-06,
+      "loss": 0.5691,
+      "step": 2558
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1139394044876099,
+      "learning_rate": 7.849249695865595e-06,
+      "loss": 0.6083,
+      "step": 2559
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2851378917694092,
+      "learning_rate": 7.847544298695812e-06,
+      "loss": 0.7062,
+      "step": 2560
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2008907794952393,
+      "learning_rate": 7.845838411094403e-06,
+      "loss": 0.5856,
+      "step": 2561
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2814170122146606,
+      "learning_rate": 7.844132033355169e-06,
+      "loss": 0.6046,
+      "step": 2562
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5419236421585083,
+      "learning_rate": 7.842425165772003e-06,
+      "loss": 0.6394,
+      "step": 2563
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3178516626358032,
+      "learning_rate": 7.840717808638878e-06,
+      "loss": 0.5834,
+      "step": 2564
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.137864112854004,
+      "learning_rate": 7.839009962249849e-06,
+      "loss": 0.7019,
+      "step": 2565
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5136849880218506,
+      "learning_rate": 7.837301626899059e-06,
+      "loss": 0.6362,
+      "step": 2566
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5012842416763306,
+      "learning_rate": 7.835592802880733e-06,
+      "loss": 0.6163,
+      "step": 2567
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2922011613845825,
+      "learning_rate": 7.833883490489183e-06,
+      "loss": 0.6593,
+      "step": 2568
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0815269947052002,
+      "learning_rate": 7.832173690018803e-06,
+      "loss": 0.641,
+      "step": 2569
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.347912073135376,
+      "learning_rate": 7.83046340176407e-06,
+      "loss": 0.5536,
+      "step": 2570
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4981647729873657,
+      "learning_rate": 7.82875262601955e-06,
+      "loss": 0.7163,
+      "step": 2571
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0309983491897583,
+      "learning_rate": 7.827041363079884e-06,
+      "loss": 0.6477,
+      "step": 2572
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9935951232910156,
+      "learning_rate": 7.825329613239804e-06,
+      "loss": 0.663,
+      "step": 2573
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2904274463653564,
+      "learning_rate": 7.823617376794128e-06,
+      "loss": 0.5916,
+      "step": 2574
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2341314554214478,
+      "learning_rate": 7.82190465403775e-06,
+      "loss": 0.6036,
+      "step": 2575
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3546459674835205,
+      "learning_rate": 7.820191445265653e-06,
+      "loss": 0.6393,
+      "step": 2576
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1507740020751953,
+      "learning_rate": 7.818477750772901e-06,
+      "loss": 0.6983,
+      "step": 2577
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.3480865955352783,
+      "learning_rate": 7.816763570854644e-06,
+      "loss": 0.6082,
+      "step": 2578
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4413642883300781,
+      "learning_rate": 7.815048905806116e-06,
+      "loss": 0.6031,
+      "step": 2579
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2350773811340332,
+      "learning_rate": 7.813333755922631e-06,
+      "loss": 0.6398,
+      "step": 2580
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.1799776554107666,
+      "learning_rate": 7.811618121499591e-06,
+      "loss": 0.6729,
+      "step": 2581
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3736685514450073,
+      "learning_rate": 7.809902002832477e-06,
+      "loss": 0.6516,
+      "step": 2582
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.426729679107666,
+      "learning_rate": 7.808185400216857e-06,
+      "loss": 0.6552,
+      "step": 2583
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6423767805099487,
+      "learning_rate": 7.806468313948379e-06,
+      "loss": 0.6658,
+      "step": 2584
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1850903034210205,
+      "learning_rate": 7.804750744322777e-06,
+      "loss": 0.4809,
+      "step": 2585
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1379947662353516,
+      "learning_rate": 7.80303269163587e-06,
+      "loss": 0.5726,
+      "step": 2586
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.541176438331604,
+      "learning_rate": 7.801314156183554e-06,
+      "loss": 0.5963,
+      "step": 2587
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.38937509059906,
+      "learning_rate": 7.799595138261815e-06,
+      "loss": 0.5938,
+      "step": 2588
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0433942079544067,
+      "learning_rate": 7.797875638166717e-06,
+      "loss": 0.7079,
+      "step": 2589
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3225352764129639,
+      "learning_rate": 7.796155656194409e-06,
+      "loss": 0.6547,
+      "step": 2590
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3787798881530762,
+      "learning_rate": 7.794435192641124e-06,
+      "loss": 0.6904,
+      "step": 2591
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0460058450698853,
+      "learning_rate": 7.792714247803174e-06,
+      "loss": 0.5865,
+      "step": 2592
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1611684560775757,
+      "learning_rate": 7.79099282197696e-06,
+      "loss": 0.7466,
+      "step": 2593
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1869133710861206,
+      "learning_rate": 7.789270915458962e-06,
+      "loss": 0.6646,
+      "step": 2594
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3539026975631714,
+      "learning_rate": 7.787548528545743e-06,
+      "loss": 0.5528,
+      "step": 2595
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1651065349578857,
+      "learning_rate": 7.785825661533947e-06,
+      "loss": 0.6363,
+      "step": 2596
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2617387771606445,
+      "learning_rate": 7.784102314720307e-06,
+      "loss": 0.669,
+      "step": 2597
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5752850770950317,
+      "learning_rate": 7.782378488401632e-06,
+      "loss": 0.5657,
+      "step": 2598
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5625871419906616,
+      "learning_rate": 7.780654182874816e-06,
+      "loss": 0.6113,
+      "step": 2599
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0509943962097168,
+      "learning_rate": 7.778929398436835e-06,
+      "loss": 0.6342,
+      "step": 2600
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2715318202972412,
+      "learning_rate": 7.777204135384749e-06,
+      "loss": 0.5642,
+      "step": 2601
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5984022617340088,
+      "learning_rate": 7.7754783940157e-06,
+      "loss": 0.6316,
+      "step": 2602
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0083099603652954,
+      "learning_rate": 7.773752174626911e-06,
+      "loss": 0.639,
+      "step": 2603
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0958874225616455,
+      "learning_rate": 7.77202547751569e-06,
+      "loss": 0.6832,
+      "step": 2604
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1940189599990845,
+      "learning_rate": 7.770298302979422e-06,
+      "loss": 0.6686,
+      "step": 2605
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1370244026184082,
+      "learning_rate": 7.768570651315582e-06,
+      "loss": 0.5631,
+      "step": 2606
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4601707458496094,
+      "learning_rate": 7.766842522821718e-06,
+      "loss": 0.6766,
+      "step": 2607
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9551748633384705,
+      "learning_rate": 7.76511391779547e-06,
+      "loss": 0.5999,
+      "step": 2608
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3081096410751343,
+      "learning_rate": 7.76338483653455e-06,
+      "loss": 0.5324,
+      "step": 2609
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.218382716178894,
+      "learning_rate": 7.761655279336762e-06,
+      "loss": 0.6574,
+      "step": 2610
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2414915561676025,
+      "learning_rate": 7.759925246499984e-06,
+      "loss": 0.6467,
+      "step": 2611
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1774640083312988,
+      "learning_rate": 7.75819473832218e-06,
+      "loss": 0.6357,
+      "step": 2612
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1158899068832397,
+      "learning_rate": 7.756463755101395e-06,
+      "loss": 0.5574,
+      "step": 2613
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0917646884918213,
+      "learning_rate": 7.754732297135756e-06,
+      "loss": 0.61,
+      "step": 2614
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4236118793487549,
+      "learning_rate": 7.753000364723471e-06,
+      "loss": 0.6406,
+      "step": 2615
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2008647918701172,
+      "learning_rate": 7.75126795816283e-06,
+      "loss": 0.6095,
+      "step": 2616
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.6084930896759033,
+      "learning_rate": 7.749535077752204e-06,
+      "loss": 0.5763,
+      "step": 2617
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2119183540344238,
+      "learning_rate": 7.747801723790046e-06,
+      "loss": 0.5449,
+      "step": 2618
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2325401306152344,
+      "learning_rate": 7.746067896574893e-06,
+      "loss": 0.7414,
+      "step": 2619
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1843703985214233,
+      "learning_rate": 7.744333596405363e-06,
+      "loss": 0.6557,
+      "step": 2620
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0715175867080688,
+      "learning_rate": 7.742598823580149e-06,
+      "loss": 0.63,
+      "step": 2621
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.040466070175171,
+      "learning_rate": 7.740863578398033e-06,
+      "loss": 0.5549,
+      "step": 2622
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.321418285369873,
+      "learning_rate": 7.739127861157878e-06,
+      "loss": 0.6138,
+      "step": 2623
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3999861478805542,
+      "learning_rate": 7.737391672158621e-06,
+      "loss": 0.5579,
+      "step": 2624
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2321044206619263,
+      "learning_rate": 7.73565501169929e-06,
+      "loss": 0.6219,
+      "step": 2625
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1841264963150024,
+      "learning_rate": 7.733917880078988e-06,
+      "loss": 0.5434,
+      "step": 2626
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1282161474227905,
+      "learning_rate": 7.732180277596899e-06,
+      "loss": 0.5728,
+      "step": 2627
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2175527811050415,
+      "learning_rate": 7.730442204552292e-06,
+      "loss": 0.6357,
+      "step": 2628
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1946020126342773,
+      "learning_rate": 7.72870366124451e-06,
+      "loss": 0.6907,
+      "step": 2629
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3675916194915771,
+      "learning_rate": 7.726964647972987e-06,
+      "loss": 0.7276,
+      "step": 2630
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.191794753074646,
+      "learning_rate": 7.725225165037233e-06,
+      "loss": 0.6401,
+      "step": 2631
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3735675811767578,
+      "learning_rate": 7.723485212736835e-06,
+      "loss": 0.7791,
+      "step": 2632
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5447452068328857,
+      "learning_rate": 7.721744791371466e-06,
+      "loss": 0.7008,
+      "step": 2633
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.7708098888397217,
+      "learning_rate": 7.72000390124088e-06,
+      "loss": 0.7301,
+      "step": 2634
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.259595513343811,
+      "learning_rate": 7.718262542644906e-06,
+      "loss": 0.6117,
+      "step": 2635
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.14170503616333,
+      "learning_rate": 7.716520715883463e-06,
+      "loss": 0.6366,
+      "step": 2636
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3256770372390747,
+      "learning_rate": 7.714778421256538e-06,
+      "loss": 0.6714,
+      "step": 2637
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6231776475906372,
+      "learning_rate": 7.71303565906421e-06,
+      "loss": 0.5758,
+      "step": 2638
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.823754906654358,
+      "learning_rate": 7.711292429606635e-06,
+      "loss": 0.6082,
+      "step": 2639
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.233652114868164,
+      "learning_rate": 7.70954873318405e-06,
+      "loss": 0.613,
+      "step": 2640
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0892643928527832,
+      "learning_rate": 7.707804570096769e-06,
+      "loss": 0.5392,
+      "step": 2641
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5985088348388672,
+      "learning_rate": 7.706059940645187e-06,
+      "loss": 0.6543,
+      "step": 2642
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1690950393676758,
+      "learning_rate": 7.704314845129785e-06,
+      "loss": 0.6247,
+      "step": 2643
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.276746392250061,
+      "learning_rate": 7.702569283851117e-06,
+      "loss": 0.6876,
+      "step": 2644
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1838765144348145,
+      "learning_rate": 7.700823257109821e-06,
+      "loss": 0.5706,
+      "step": 2645
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1753582954406738,
+      "learning_rate": 7.699076765206617e-06,
+      "loss": 0.7981,
+      "step": 2646
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.190260410308838,
+      "learning_rate": 7.6973298084423e-06,
+      "loss": 0.6326,
+      "step": 2647
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8324419260025024,
+      "learning_rate": 7.695582387117749e-06,
+      "loss": 0.678,
+      "step": 2648
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1149442195892334,
+      "learning_rate": 7.693834501533921e-06,
+      "loss": 0.6484,
+      "step": 2649
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1948319673538208,
+      "learning_rate": 7.692086151991855e-06,
+      "loss": 0.5729,
+      "step": 2650
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7144792079925537,
+      "learning_rate": 7.690337338792667e-06,
+      "loss": 0.683,
+      "step": 2651
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9441317319869995,
+      "learning_rate": 7.688588062237557e-06,
+      "loss": 0.6081,
+      "step": 2652
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2331782579421997,
+      "learning_rate": 7.6868383226278e-06,
+      "loss": 0.6088,
+      "step": 2653
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6664066314697266,
+      "learning_rate": 7.685088120264754e-06,
+      "loss": 0.6171,
+      "step": 2654
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3156743049621582,
+      "learning_rate": 7.683337455449856e-06,
+      "loss": 0.6272,
+      "step": 2655
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1555067300796509,
+      "learning_rate": 7.681586328484621e-06,
+      "loss": 0.5953,
+      "step": 2656
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3859845399856567,
+      "learning_rate": 7.679834739670649e-06,
+      "loss": 0.634,
+      "step": 2657
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.6940677165985107,
+      "learning_rate": 7.67808268930961e-06,
+      "loss": 0.6804,
+      "step": 2658
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2090493440628052,
+      "learning_rate": 7.676330177703262e-06,
+      "loss": 0.6196,
+      "step": 2659
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1433385610580444,
+      "learning_rate": 7.674577205153441e-06,
+      "loss": 0.6498,
+      "step": 2660
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3377647399902344,
+      "learning_rate": 7.672823771962059e-06,
+      "loss": 0.5986,
+      "step": 2661
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0680303573608398,
+      "learning_rate": 7.671069878431107e-06,
+      "loss": 0.5319,
+      "step": 2662
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3417402505874634,
+      "learning_rate": 7.669315524862662e-06,
+      "loss": 0.6823,
+      "step": 2663
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.365355372428894,
+      "learning_rate": 7.667560711558875e-06,
+      "loss": 0.6342,
+      "step": 2664
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3211275339126587,
+      "learning_rate": 7.665805438821973e-06,
+      "loss": 0.6097,
+      "step": 2665
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.580230474472046,
+      "learning_rate": 7.664049706954271e-06,
+      "loss": 0.6559,
+      "step": 2666
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0732578039169312,
+      "learning_rate": 7.662293516258154e-06,
+      "loss": 0.5693,
+      "step": 2667
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8319809436798096,
+      "learning_rate": 7.660536867036092e-06,
+      "loss": 0.6538,
+      "step": 2668
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.520002007484436,
+      "learning_rate": 7.658779759590634e-06,
+      "loss": 0.5934,
+      "step": 2669
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2555063962936401,
+      "learning_rate": 7.657022194224402e-06,
+      "loss": 0.6547,
+      "step": 2670
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.653493642807007,
+      "learning_rate": 7.655264171240105e-06,
+      "loss": 0.6805,
+      "step": 2671
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.030552625656128,
+      "learning_rate": 7.653505690940522e-06,
+      "loss": 0.5611,
+      "step": 2672
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2763005495071411,
+      "learning_rate": 7.65174675362852e-06,
+      "loss": 0.6387,
+      "step": 2673
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7102898359298706,
+      "learning_rate": 7.649987359607039e-06,
+      "loss": 0.6756,
+      "step": 2674
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2029314041137695,
+      "learning_rate": 7.648227509179095e-06,
+      "loss": 0.7813,
+      "step": 2675
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1546415090560913,
+      "learning_rate": 7.646467202647794e-06,
+      "loss": 0.6681,
+      "step": 2676
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1073461771011353,
+      "learning_rate": 7.644706440316308e-06,
+      "loss": 0.6578,
+      "step": 2677
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2768898010253906,
+      "learning_rate": 7.642945222487892e-06,
+      "loss": 0.7158,
+      "step": 2678
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2239513397216797,
+      "learning_rate": 7.641183549465881e-06,
+      "loss": 0.6005,
+      "step": 2679
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0863401889801025,
+      "learning_rate": 7.639421421553687e-06,
+      "loss": 0.6159,
+      "step": 2680
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.596559762954712,
+      "learning_rate": 7.637658839054805e-06,
+      "loss": 0.6288,
+      "step": 2681
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3041114807128906,
+      "learning_rate": 7.635895802272796e-06,
+      "loss": 0.5878,
+      "step": 2682
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0410126447677612,
+      "learning_rate": 7.634132311511316e-06,
+      "loss": 0.5978,
+      "step": 2683
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9961788654327393,
+      "learning_rate": 7.632368367074083e-06,
+      "loss": 0.5053,
+      "step": 2684
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2974720001220703,
+      "learning_rate": 7.630603969264905e-06,
+      "loss": 0.6118,
+      "step": 2685
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1592649221420288,
+      "learning_rate": 7.628839118387662e-06,
+      "loss": 0.6189,
+      "step": 2686
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.093314528465271,
+      "learning_rate": 7.627073814746315e-06,
+      "loss": 0.6362,
+      "step": 2687
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.052543044090271,
+      "learning_rate": 7.625308058644898e-06,
+      "loss": 0.6143,
+      "step": 2688
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3054975271224976,
+      "learning_rate": 7.623541850387531e-06,
+      "loss": 0.6378,
+      "step": 2689
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0570430755615234,
+      "learning_rate": 7.621775190278407e-06,
+      "loss": 0.5995,
+      "step": 2690
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4440031051635742,
+      "learning_rate": 7.620008078621793e-06,
+      "loss": 0.6158,
+      "step": 2691
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3761212825775146,
+      "learning_rate": 7.618240515722044e-06,
+      "loss": 0.5676,
+      "step": 2692
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1331719160079956,
+      "learning_rate": 7.616472501883583e-06,
+      "loss": 0.6331,
+      "step": 2693
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8078550100326538,
+      "learning_rate": 7.614704037410915e-06,
+      "loss": 0.5285,
+      "step": 2694
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2421823740005493,
+      "learning_rate": 7.6129351226086225e-06,
+      "loss": 0.7476,
+      "step": 2695
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6171449422836304,
+      "learning_rate": 7.6111657577813644e-06,
+      "loss": 0.6853,
+      "step": 2696
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.045614719390869,
+      "learning_rate": 7.609395943233877e-06,
+      "loss": 0.6106,
+      "step": 2697
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2099361419677734,
+      "learning_rate": 7.6076256792709805e-06,
+      "loss": 0.5939,
+      "step": 2698
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3107006549835205,
+      "learning_rate": 7.605854966197559e-06,
+      "loss": 0.6542,
+      "step": 2699
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4774034023284912,
+      "learning_rate": 7.604083804318587e-06,
+      "loss": 0.6179,
+      "step": 2700
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5129690170288086,
+      "learning_rate": 7.60231219393911e-06,
+      "loss": 0.6467,
+      "step": 2701
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1982402801513672,
+      "learning_rate": 7.600540135364252e-06,
+      "loss": 0.6278,
+      "step": 2702
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2928173542022705,
+      "learning_rate": 7.598767628899213e-06,
+      "loss": 0.7124,
+      "step": 2703
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1297682523727417,
+      "learning_rate": 7.596994674849272e-06,
+      "loss": 0.5936,
+      "step": 2704
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0973412990570068,
+      "learning_rate": 7.595221273519784e-06,
+      "loss": 0.7179,
+      "step": 2705
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5251829624176025,
+      "learning_rate": 7.59344742521618e-06,
+      "loss": 0.6319,
+      "step": 2706
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0643885135650635,
+      "learning_rate": 7.591673130243973e-06,
+      "loss": 0.6339,
+      "step": 2707
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0274592638015747,
+      "learning_rate": 7.589898388908745e-06,
+      "loss": 0.6236,
+      "step": 2708
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.272946834564209,
+      "learning_rate": 7.588123201516164e-06,
+      "loss": 0.6329,
+      "step": 2709
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2281945943832397,
+      "learning_rate": 7.5863475683719655e-06,
+      "loss": 0.5959,
+      "step": 2710
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1847566366195679,
+      "learning_rate": 7.584571489781968e-06,
+      "loss": 0.5932,
+      "step": 2711
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0989586114883423,
+      "learning_rate": 7.582794966052064e-06,
+      "loss": 0.6339,
+      "step": 2712
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.114815354347229,
+      "learning_rate": 7.581017997488225e-06,
+      "loss": 0.5454,
+      "step": 2713
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.860734224319458,
+      "learning_rate": 7.579240584396497e-06,
+      "loss": 0.5895,
+      "step": 2714
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2328461408615112,
+      "learning_rate": 7.577462727083002e-06,
+      "loss": 0.6297,
+      "step": 2715
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1195461750030518,
+      "learning_rate": 7.575684425853944e-06,
+      "loss": 0.5487,
+      "step": 2716
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1979864835739136,
+      "learning_rate": 7.573905681015594e-06,
+      "loss": 0.6874,
+      "step": 2717
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1667231321334839,
+      "learning_rate": 7.57212649287431e-06,
+      "loss": 0.5755,
+      "step": 2718
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0767219066619873,
+      "learning_rate": 7.570346861736515e-06,
+      "loss": 0.6358,
+      "step": 2719
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3232417106628418,
+      "learning_rate": 7.568566787908719e-06,
+      "loss": 0.5362,
+      "step": 2720
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3619623184204102,
+      "learning_rate": 7.5667862716975e-06,
+      "loss": 0.6142,
+      "step": 2721
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1791231632232666,
+      "learning_rate": 7.565005313409519e-06,
+      "loss": 0.5849,
+      "step": 2722
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.304181694984436,
+      "learning_rate": 7.563223913351507e-06,
+      "loss": 0.6559,
+      "step": 2723
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1328728199005127,
+      "learning_rate": 7.561442071830279e-06,
+      "loss": 0.5632,
+      "step": 2724
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1923437118530273,
+      "learning_rate": 7.559659789152714e-06,
+      "loss": 0.5817,
+      "step": 2725
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4104726314544678,
+      "learning_rate": 7.557877065625778e-06,
+      "loss": 0.7269,
+      "step": 2726
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1886721849441528,
+      "learning_rate": 7.5560939015565084e-06,
+      "loss": 0.524,
+      "step": 2727
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6470292806625366,
+      "learning_rate": 7.554310297252019e-06,
+      "loss": 0.6261,
+      "step": 2728
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.882947564125061,
+      "learning_rate": 7.552526253019499e-06,
+      "loss": 0.6656,
+      "step": 2729
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4039462804794312,
+      "learning_rate": 7.550741769166215e-06,
+      "loss": 0.6302,
+      "step": 2730
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.5081090927124023,
+      "learning_rate": 7.548956845999504e-06,
+      "loss": 0.652,
+      "step": 2731
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2422338724136353,
+      "learning_rate": 7.547171483826788e-06,
+      "loss": 0.6617,
+      "step": 2732
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.51920485496521,
+      "learning_rate": 7.545385682955558e-06,
+      "loss": 0.5829,
+      "step": 2733
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3961273431777954,
+      "learning_rate": 7.543599443693379e-06,
+      "loss": 0.6472,
+      "step": 2734
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0059709548950195,
+      "learning_rate": 7.541812766347898e-06,
+      "loss": 0.6372,
+      "step": 2735
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0821688175201416,
+      "learning_rate": 7.540025651226832e-06,
+      "loss": 0.618,
+      "step": 2736
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.056602954864502,
+      "learning_rate": 7.538238098637976e-06,
+      "loss": 0.5796,
+      "step": 2737
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2009097337722778,
+      "learning_rate": 7.536450108889197e-06,
+      "loss": 0.5918,
+      "step": 2738
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1832275390625,
+      "learning_rate": 7.534661682288444e-06,
+      "loss": 0.6911,
+      "step": 2739
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2712024450302124,
+      "learning_rate": 7.532872819143735e-06,
+      "loss": 0.57,
+      "step": 2740
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5780545473098755,
+      "learning_rate": 7.531083519763164e-06,
+      "loss": 0.5392,
+      "step": 2741
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6948215961456299,
+      "learning_rate": 7.5292937844549045e-06,
+      "loss": 0.6263,
+      "step": 2742
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.145394206047058,
+      "learning_rate": 7.527503613527198e-06,
+      "loss": 0.5929,
+      "step": 2743
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0543674230575562,
+      "learning_rate": 7.525713007288371e-06,
+      "loss": 0.5308,
+      "step": 2744
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.533777117729187,
+      "learning_rate": 7.523921966046813e-06,
+      "loss": 0.6327,
+      "step": 2745
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6052230596542358,
+      "learning_rate": 7.522130490110999e-06,
+      "loss": 0.6528,
+      "step": 2746
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4751596450805664,
+      "learning_rate": 7.52033857978947e-06,
+      "loss": 0.6004,
+      "step": 2747
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4990054368972778,
+      "learning_rate": 7.51854623539085e-06,
+      "loss": 0.5579,
+      "step": 2748
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8240008354187012,
+      "learning_rate": 7.516753457223831e-06,
+      "loss": 0.6331,
+      "step": 2749
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3702977895736694,
+      "learning_rate": 7.514960245597185e-06,
+      "loss": 0.6204,
+      "step": 2750
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5224536657333374,
+      "learning_rate": 7.5131666008197546e-06,
+      "loss": 0.6065,
+      "step": 2751
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.307070016860962,
+      "learning_rate": 7.51137252320046e-06,
+      "loss": 0.7614,
+      "step": 2752
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2971265316009521,
+      "learning_rate": 7.509578013048293e-06,
+      "loss": 0.5903,
+      "step": 2753
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0652785301208496,
+      "learning_rate": 7.507783070672323e-06,
+      "loss": 0.6088,
+      "step": 2754
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1220684051513672,
+      "learning_rate": 7.505987696381692e-06,
+      "loss": 0.6447,
+      "step": 2755
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2321696281433105,
+      "learning_rate": 7.504191890485615e-06,
+      "loss": 0.6243,
+      "step": 2756
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2906885147094727,
+      "learning_rate": 7.502395653293385e-06,
+      "loss": 0.6823,
+      "step": 2757
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1404194831848145,
+      "learning_rate": 7.500598985114367e-06,
+      "loss": 0.54,
+      "step": 2758
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.316081166267395,
+      "learning_rate": 7.498801886258001e-06,
+      "loss": 0.597,
+      "step": 2759
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2815004587173462,
+      "learning_rate": 7.497004357033799e-06,
+      "loss": 0.6086,
+      "step": 2760
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0035674571990967,
+      "learning_rate": 7.495206397751351e-06,
+      "loss": 0.6706,
+      "step": 2761
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.31266188621521,
+      "learning_rate": 7.493408008720317e-06,
+      "loss": 0.5874,
+      "step": 2762
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.275524377822876,
+      "learning_rate": 7.491609190250435e-06,
+      "loss": 0.5634,
+      "step": 2763
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0364302396774292,
+      "learning_rate": 7.489809942651514e-06,
+      "loss": 0.5669,
+      "step": 2764
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.401902198791504,
+      "learning_rate": 7.488010266233436e-06,
+      "loss": 0.6698,
+      "step": 2765
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1657154560089111,
+      "learning_rate": 7.486210161306161e-06,
+      "loss": 0.585,
+      "step": 2766
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.433221697807312,
+      "learning_rate": 7.48440962817972e-06,
+      "loss": 0.5722,
+      "step": 2767
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1307737827301025,
+      "learning_rate": 7.482608667164218e-06,
+      "loss": 0.6006,
+      "step": 2768
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.26017165184021,
+      "learning_rate": 7.480807278569835e-06,
+      "loss": 0.7052,
+      "step": 2769
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2488316297531128,
+      "learning_rate": 7.479005462706821e-06,
+      "loss": 0.5859,
+      "step": 2770
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1443097591400146,
+      "learning_rate": 7.477203219885505e-06,
+      "loss": 0.632,
+      "step": 2771
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0260684490203857,
+      "learning_rate": 7.4754005504162855e-06,
+      "loss": 0.6807,
+      "step": 2772
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0291800498962402,
+      "learning_rate": 7.4735974546096335e-06,
+      "loss": 0.6448,
+      "step": 2773
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8409909009933472,
+      "learning_rate": 7.4717939327761005e-06,
+      "loss": 0.6562,
+      "step": 2774
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0962021350860596,
+      "learning_rate": 7.4699899852263e-06,
+      "loss": 0.6475,
+      "step": 2775
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6759026050567627,
+      "learning_rate": 7.468185612270932e-06,
+      "loss": 0.6217,
+      "step": 2776
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2551426887512207,
+      "learning_rate": 7.466380814220758e-06,
+      "loss": 0.6965,
+      "step": 2777
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2845110893249512,
+      "learning_rate": 7.464575591386623e-06,
+      "loss": 0.6527,
+      "step": 2778
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.096612572669983,
+      "learning_rate": 7.462769944079433e-06,
+      "loss": 0.6857,
+      "step": 2779
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.328682780265808,
+      "learning_rate": 7.460963872610181e-06,
+      "loss": 0.6129,
+      "step": 2780
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3041599988937378,
+      "learning_rate": 7.45915737728992e-06,
+      "loss": 0.6,
+      "step": 2781
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4073518514633179,
+      "learning_rate": 7.457350458429788e-06,
+      "loss": 0.697,
+      "step": 2782
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2230504751205444,
+      "learning_rate": 7.455543116340985e-06,
+      "loss": 0.665,
+      "step": 2783
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3002861738204956,
+      "learning_rate": 7.453735351334791e-06,
+      "loss": 0.6305,
+      "step": 2784
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2461925745010376,
+      "learning_rate": 7.451927163722557e-06,
+      "loss": 0.6243,
+      "step": 2785
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6128621101379395,
+      "learning_rate": 7.450118553815707e-06,
+      "loss": 0.6508,
+      "step": 2786
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3560428619384766,
+      "learning_rate": 7.448309521925737e-06,
+      "loss": 0.6631,
+      "step": 2787
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.214652180671692,
+      "learning_rate": 7.4465000683642144e-06,
+      "loss": 0.6243,
+      "step": 2788
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.2300350666046143,
+      "learning_rate": 7.444690193442783e-06,
+      "loss": 0.6499,
+      "step": 2789
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3689312934875488,
+      "learning_rate": 7.442879897473156e-06,
+      "loss": 0.6187,
+      "step": 2790
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1015273332595825,
+      "learning_rate": 7.441069180767119e-06,
+      "loss": 0.5566,
+      "step": 2791
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6342823505401611,
+      "learning_rate": 7.439258043636532e-06,
+      "loss": 0.5791,
+      "step": 2792
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.212996482849121,
+      "learning_rate": 7.437446486393327e-06,
+      "loss": 0.535,
+      "step": 2793
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1164084672927856,
+      "learning_rate": 7.435634509349509e-06,
+      "loss": 0.7325,
+      "step": 2794
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.421766996383667,
+      "learning_rate": 7.433822112817151e-06,
+      "loss": 0.5267,
+      "step": 2795
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4409089088439941,
+      "learning_rate": 7.4320092971084044e-06,
+      "loss": 0.5939,
+      "step": 2796
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0344882011413574,
+      "learning_rate": 7.430196062535488e-06,
+      "loss": 0.5837,
+      "step": 2797
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0946214199066162,
+      "learning_rate": 7.428382409410697e-06,
+      "loss": 0.5718,
+      "step": 2798
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3646178245544434,
+      "learning_rate": 7.426568338046394e-06,
+      "loss": 0.5607,
+      "step": 2799
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1723817586898804,
+      "learning_rate": 7.4247538487550154e-06,
+      "loss": 0.6713,
+      "step": 2800
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.326438546180725,
+      "learning_rate": 7.422938941849073e-06,
+      "loss": 0.6182,
+      "step": 2801
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0464811325073242,
+      "learning_rate": 7.421123617641145e-06,
+      "loss": 0.5515,
+      "step": 2802
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1844979524612427,
+      "learning_rate": 7.4193078764438855e-06,
+      "loss": 0.5557,
+      "step": 2803
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2062512636184692,
+      "learning_rate": 7.417491718570018e-06,
+      "loss": 0.6877,
+      "step": 2804
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0828931331634521,
+      "learning_rate": 7.41567514433234e-06,
+      "loss": 0.6256,
+      "step": 2805
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0971540212631226,
+      "learning_rate": 7.413858154043718e-06,
+      "loss": 0.5836,
+      "step": 2806
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0136876106262207,
+      "learning_rate": 7.412040748017094e-06,
+      "loss": 0.6453,
+      "step": 2807
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2080310583114624,
+      "learning_rate": 7.410222926565477e-06,
+      "loss": 0.695,
+      "step": 2808
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1714882850646973,
+      "learning_rate": 7.408404690001949e-06,
+      "loss": 0.6441,
+      "step": 2809
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.214609980583191,
+      "learning_rate": 7.406586038639667e-06,
+      "loss": 0.6754,
+      "step": 2810
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2446070909500122,
+      "learning_rate": 7.404766972791856e-06,
+      "loss": 0.5658,
+      "step": 2811
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5085375308990479,
+      "learning_rate": 7.402947492771811e-06,
+      "loss": 0.6991,
+      "step": 2812
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.154221534729004,
+      "learning_rate": 7.4011275988929046e-06,
+      "loss": 0.6379,
+      "step": 2813
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2423889636993408,
+      "learning_rate": 7.399307291468572e-06,
+      "loss": 0.6439,
+      "step": 2814
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5794448852539062,
+      "learning_rate": 7.3974865708123276e-06,
+      "loss": 0.6704,
+      "step": 2815
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1966044902801514,
+      "learning_rate": 7.395665437237751e-06,
+      "loss": 0.6989,
+      "step": 2816
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.102135419845581,
+      "learning_rate": 7.3938438910584964e-06,
+      "loss": 0.6813,
+      "step": 2817
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2617212533950806,
+      "learning_rate": 7.392021932588289e-06,
+      "loss": 0.6861,
+      "step": 2818
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1992825269699097,
+      "learning_rate": 7.3901995621409224e-06,
+      "loss": 0.5791,
+      "step": 2819
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2392593622207642,
+      "learning_rate": 7.3883767800302665e-06,
+      "loss": 0.7588,
+      "step": 2820
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4153352975845337,
+      "learning_rate": 7.386553586570253e-06,
+      "loss": 0.62,
+      "step": 2821
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2339427471160889,
+      "learning_rate": 7.384729982074894e-06,
+      "loss": 0.6189,
+      "step": 2822
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1156781911849976,
+      "learning_rate": 7.382905966858267e-06,
+      "loss": 0.57,
+      "step": 2823
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5094579458236694,
+      "learning_rate": 7.381081541234522e-06,
+      "loss": 0.6152,
+      "step": 2824
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3721705675125122,
+      "learning_rate": 7.3792567055178766e-06,
+      "loss": 0.5664,
+      "step": 2825
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3138307332992554,
+      "learning_rate": 7.377431460022625e-06,
+      "loss": 0.6344,
+      "step": 2826
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6809388399124146,
+      "learning_rate": 7.375605805063128e-06,
+      "loss": 0.5832,
+      "step": 2827
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2338505983352661,
+      "learning_rate": 7.373779740953817e-06,
+      "loss": 0.5376,
+      "step": 2828
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1951864957809448,
+      "learning_rate": 7.371953268009194e-06,
+      "loss": 0.5897,
+      "step": 2829
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2772202491760254,
+      "learning_rate": 7.370126386543833e-06,
+      "loss": 0.6407,
+      "step": 2830
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6042275428771973,
+      "learning_rate": 7.368299096872377e-06,
+      "loss": 0.6627,
+      "step": 2831
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3527517318725586,
+      "learning_rate": 7.366471399309538e-06,
+      "loss": 0.589,
+      "step": 2832
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1194247007369995,
+      "learning_rate": 7.3646432941701015e-06,
+      "loss": 0.6234,
+      "step": 2833
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0248486995697021,
+      "learning_rate": 7.362814781768919e-06,
+      "loss": 0.551,
+      "step": 2834
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1566435098648071,
+      "learning_rate": 7.360985862420919e-06,
+      "loss": 0.6106,
+      "step": 2835
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.258584976196289,
+      "learning_rate": 7.359156536441091e-06,
+      "loss": 0.6558,
+      "step": 2836
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1506608724594116,
+      "learning_rate": 7.357326804144502e-06,
+      "loss": 0.6423,
+      "step": 2837
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1835181713104248,
+      "learning_rate": 7.355496665846285e-06,
+      "loss": 0.672,
+      "step": 2838
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3529503345489502,
+      "learning_rate": 7.353666121861646e-06,
+      "loss": 0.5715,
+      "step": 2839
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.514220952987671,
+      "learning_rate": 7.3518351725058555e-06,
+      "loss": 0.637,
+      "step": 2840
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.547484278678894,
+      "learning_rate": 7.350003818094261e-06,
+      "loss": 0.6553,
+      "step": 2841
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.531700849533081,
+      "learning_rate": 7.348172058942274e-06,
+      "loss": 0.6472,
+      "step": 2842
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1948038339614868,
+      "learning_rate": 7.346339895365378e-06,
+      "loss": 0.6535,
+      "step": 2843
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2925516366958618,
+      "learning_rate": 7.344507327679127e-06,
+      "loss": 0.5542,
+      "step": 2844
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.377945899963379,
+      "learning_rate": 7.342674356199141e-06,
+      "loss": 0.687,
+      "step": 2845
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2123630046844482,
+      "learning_rate": 7.340840981241115e-06,
+      "loss": 0.5739,
+      "step": 2846
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.087882161140442,
+      "learning_rate": 7.339007203120809e-06,
+      "loss": 0.6949,
+      "step": 2847
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.039701223373413,
+      "learning_rate": 7.3371730221540545e-06,
+      "loss": 0.577,
+      "step": 2848
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.535933494567871,
+      "learning_rate": 7.335338438656752e-06,
+      "loss": 0.6101,
+      "step": 2849
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4305620193481445,
+      "learning_rate": 7.333503452944872e-06,
+      "loss": 0.5939,
+      "step": 2850
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1246757507324219,
+      "learning_rate": 7.331668065334449e-06,
+      "loss": 0.6691,
+      "step": 2851
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2173248529434204,
+      "learning_rate": 7.329832276141597e-06,
+      "loss": 0.7914,
+      "step": 2852
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0413439273834229,
+      "learning_rate": 7.327996085682491e-06,
+      "loss": 0.6105,
+      "step": 2853
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1989972591400146,
+      "learning_rate": 7.326159494273377e-06,
+      "loss": 0.7641,
+      "step": 2854
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2141451835632324,
+      "learning_rate": 7.324322502230571e-06,
+      "loss": 0.6149,
+      "step": 2855
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3219760656356812,
+      "learning_rate": 7.322485109870458e-06,
+      "loss": 0.5544,
+      "step": 2856
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2472971677780151,
+      "learning_rate": 7.320647317509493e-06,
+      "loss": 0.628,
+      "step": 2857
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1860524415969849,
+      "learning_rate": 7.318809125464194e-06,
+      "loss": 0.5704,
+      "step": 2858
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7199898958206177,
+      "learning_rate": 7.316970534051155e-06,
+      "loss": 0.5817,
+      "step": 2859
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1000601053237915,
+      "learning_rate": 7.315131543587035e-06,
+      "loss": 0.5683,
+      "step": 2860
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0615863800048828,
+      "learning_rate": 7.3132921543885646e-06,
+      "loss": 0.5399,
+      "step": 2861
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2328920364379883,
+      "learning_rate": 7.31145236677254e-06,
+      "loss": 0.6103,
+      "step": 2862
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1900357007980347,
+      "learning_rate": 7.309612181055827e-06,
+      "loss": 0.5467,
+      "step": 2863
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1210532188415527,
+      "learning_rate": 7.30777159755536e-06,
+      "loss": 0.6925,
+      "step": 2864
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3341383934020996,
+      "learning_rate": 7.305930616588144e-06,
+      "loss": 0.5621,
+      "step": 2865
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.700534701347351,
+      "learning_rate": 7.304089238471248e-06,
+      "loss": 0.695,
+      "step": 2866
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.206648588180542,
+      "learning_rate": 7.302247463521813e-06,
+      "loss": 0.59,
+      "step": 2867
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0011547803878784,
+      "learning_rate": 7.300405292057048e-06,
+      "loss": 0.5618,
+      "step": 2868
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4958430528640747,
+      "learning_rate": 7.298562724394227e-06,
+      "loss": 0.4853,
+      "step": 2869
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4243090152740479,
+      "learning_rate": 7.296719760850699e-06,
+      "loss": 0.7581,
+      "step": 2870
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1251380443572998,
+      "learning_rate": 7.294876401743873e-06,
+      "loss": 0.6223,
+      "step": 2871
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3010609149932861,
+      "learning_rate": 7.293032647391234e-06,
+      "loss": 0.635,
+      "step": 2872
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1540263891220093,
+      "learning_rate": 7.291188498110328e-06,
+      "loss": 0.611,
+      "step": 2873
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.1613638401031494,
+      "learning_rate": 7.289343954218772e-06,
+      "loss": 0.6606,
+      "step": 2874
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.032372236251831,
+      "learning_rate": 7.287499016034255e-06,
+      "loss": 0.6623,
+      "step": 2875
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2156139612197876,
+      "learning_rate": 7.285653683874527e-06,
+      "loss": 0.6258,
+      "step": 2876
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3962489366531372,
+      "learning_rate": 7.2838079580574076e-06,
+      "loss": 0.5691,
+      "step": 2877
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1535193920135498,
+      "learning_rate": 7.28196183890079e-06,
+      "loss": 0.6384,
+      "step": 2878
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.083708643913269,
+      "learning_rate": 7.2801153267226266e-06,
+      "loss": 0.5573,
+      "step": 2879
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3152596950531006,
+      "learning_rate": 7.278268421840944e-06,
+      "loss": 0.6413,
+      "step": 2880
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2211905717849731,
+      "learning_rate": 7.276421124573834e-06,
+      "loss": 0.608,
+      "step": 2881
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.245803713798523,
+      "learning_rate": 7.274573435239454e-06,
+      "loss": 0.555,
+      "step": 2882
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.663479208946228,
+      "learning_rate": 7.272725354156034e-06,
+      "loss": 0.5815,
+      "step": 2883
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.306326150894165,
+      "learning_rate": 7.270876881641864e-06,
+      "loss": 0.5868,
+      "step": 2884
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.315184235572815,
+      "learning_rate": 7.269028018015311e-06,
+      "loss": 0.6078,
+      "step": 2885
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3038302659988403,
+      "learning_rate": 7.267178763594797e-06,
+      "loss": 0.6068,
+      "step": 2886
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1452289819717407,
+      "learning_rate": 7.265329118698828e-06,
+      "loss": 0.5734,
+      "step": 2887
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0434083938598633,
+      "learning_rate": 7.263479083645961e-06,
+      "loss": 0.6097,
+      "step": 2888
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2132070064544678,
+      "learning_rate": 7.2616286587548305e-06,
+      "loss": 0.5651,
+      "step": 2889
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0834686756134033,
+      "learning_rate": 7.25977784434413e-06,
+      "loss": 0.6347,
+      "step": 2890
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3891212940216064,
+      "learning_rate": 7.25792664073263e-06,
+      "loss": 0.5851,
+      "step": 2891
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2323148250579834,
+      "learning_rate": 7.256075048239158e-06,
+      "loss": 0.7112,
+      "step": 2892
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1010910272598267,
+      "learning_rate": 7.254223067182618e-06,
+      "loss": 0.6333,
+      "step": 2893
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0941414833068848,
+      "learning_rate": 7.252370697881971e-06,
+      "loss": 0.5294,
+      "step": 2894
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3933345079421997,
+      "learning_rate": 7.250517940656253e-06,
+      "loss": 0.6761,
+      "step": 2895
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3738077878952026,
+      "learning_rate": 7.248664795824565e-06,
+      "loss": 0.6789,
+      "step": 2896
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0183112621307373,
+      "learning_rate": 7.246811263706069e-06,
+      "loss": 0.5176,
+      "step": 2897
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2617154121398926,
+      "learning_rate": 7.244957344620003e-06,
+      "loss": 0.7175,
+      "step": 2898
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2727289199829102,
+      "learning_rate": 7.243103038885664e-06,
+      "loss": 0.6891,
+      "step": 2899
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2516908645629883,
+      "learning_rate": 7.241248346822421e-06,
+      "loss": 0.6941,
+      "step": 2900
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.441074013710022,
+      "learning_rate": 7.2393932687497015e-06,
+      "loss": 0.6303,
+      "step": 2901
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.305977702140808,
+      "learning_rate": 7.237537804987013e-06,
+      "loss": 0.7342,
+      "step": 2902
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.399827480316162,
+      "learning_rate": 7.2356819558539124e-06,
+      "loss": 0.6366,
+      "step": 2903
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2844167947769165,
+      "learning_rate": 7.233825721670042e-06,
+      "loss": 0.5527,
+      "step": 2904
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3156801462173462,
+      "learning_rate": 7.231969102755093e-06,
+      "loss": 0.6165,
+      "step": 2905
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2262427806854248,
+      "learning_rate": 7.230112099428832e-06,
+      "loss": 0.578,
+      "step": 2906
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.332987904548645,
+      "learning_rate": 7.228254712011091e-06,
+      "loss": 0.5501,
+      "step": 2907
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1426584720611572,
+      "learning_rate": 7.226396940821767e-06,
+      "loss": 0.5759,
+      "step": 2908
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.636731505393982,
+      "learning_rate": 7.224538786180824e-06,
+      "loss": 0.6399,
+      "step": 2909
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7168676853179932,
+      "learning_rate": 7.222680248408289e-06,
+      "loss": 0.674,
+      "step": 2910
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3796486854553223,
+      "learning_rate": 7.22082132782426e-06,
+      "loss": 0.6683,
+      "step": 2911
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0893101692199707,
+      "learning_rate": 7.218962024748896e-06,
+      "loss": 0.5658,
+      "step": 2912
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2268693447113037,
+      "learning_rate": 7.217102339502426e-06,
+      "loss": 0.6162,
+      "step": 2913
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2177835702896118,
+      "learning_rate": 7.215242272405142e-06,
+      "loss": 0.6081,
+      "step": 2914
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2131762504577637,
+      "learning_rate": 7.213381823777404e-06,
+      "loss": 0.5518,
+      "step": 2915
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3434334993362427,
+      "learning_rate": 7.2115209939396345e-06,
+      "loss": 0.6446,
+      "step": 2916
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2754311561584473,
+      "learning_rate": 7.209659783212326e-06,
+      "loss": 0.6655,
+      "step": 2917
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3804233074188232,
+      "learning_rate": 7.207798191916031e-06,
+      "loss": 0.6211,
+      "step": 2918
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2941632270812988,
+      "learning_rate": 7.205936220371374e-06,
+      "loss": 0.5307,
+      "step": 2919
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4709851741790771,
+      "learning_rate": 7.20407386889904e-06,
+      "loss": 0.5958,
+      "step": 2920
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3792892694473267,
+      "learning_rate": 7.202211137819781e-06,
+      "loss": 0.6584,
+      "step": 2921
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2968083620071411,
+      "learning_rate": 7.2003480274544156e-06,
+      "loss": 0.5921,
+      "step": 2922
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2979252338409424,
+      "learning_rate": 7.198484538123826e-06,
+      "loss": 0.6601,
+      "step": 2923
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3300632238388062,
+      "learning_rate": 7.196620670148961e-06,
+      "loss": 0.7856,
+      "step": 2924
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1163116693496704,
+      "learning_rate": 7.194756423850833e-06,
+      "loss": 0.6127,
+      "step": 2925
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.256772518157959,
+      "learning_rate": 7.192891799550522e-06,
+      "loss": 0.5977,
+      "step": 2926
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.180251955986023,
+      "learning_rate": 7.19102679756917e-06,
+      "loss": 0.6473,
+      "step": 2927
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2117708921432495,
+      "learning_rate": 7.189161418227987e-06,
+      "loss": 0.6191,
+      "step": 2928
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.150450587272644,
+      "learning_rate": 7.187295661848243e-06,
+      "loss": 0.7571,
+      "step": 2929
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3137658834457397,
+      "learning_rate": 7.185429528751285e-06,
+      "loss": 0.5918,
+      "step": 2930
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3359402418136597,
+      "learning_rate": 7.1835630192585085e-06,
+      "loss": 0.5988,
+      "step": 2931
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3789992332458496,
+      "learning_rate": 7.1816961336913835e-06,
+      "loss": 0.6197,
+      "step": 2932
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2217789888381958,
+      "learning_rate": 7.179828872371446e-06,
+      "loss": 0.5909,
+      "step": 2933
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4157116413116455,
+      "learning_rate": 7.17796123562029e-06,
+      "loss": 0.5482,
+      "step": 2934
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2854241132736206,
+      "learning_rate": 7.176093223759581e-06,
+      "loss": 0.5544,
+      "step": 2935
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2791540622711182,
+      "learning_rate": 7.174224837111044e-06,
+      "loss": 0.6145,
+      "step": 2936
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3771109580993652,
+      "learning_rate": 7.172356075996473e-06,
+      "loss": 0.6808,
+      "step": 2937
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5977014303207397,
+      "learning_rate": 7.170486940737722e-06,
+      "loss": 0.7052,
+      "step": 2938
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.30767023563385,
+      "learning_rate": 7.168617431656713e-06,
+      "loss": 0.674,
+      "step": 2939
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5938276052474976,
+      "learning_rate": 7.166747549075428e-06,
+      "loss": 0.6467,
+      "step": 2940
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.34315824508667,
+      "learning_rate": 7.164877293315921e-06,
+      "loss": 0.5739,
+      "step": 2941
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2254630327224731,
+      "learning_rate": 7.163006664700302e-06,
+      "loss": 0.6603,
+      "step": 2942
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1479889154434204,
+      "learning_rate": 7.16113566355075e-06,
+      "loss": 0.6333,
+      "step": 2943
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.111104130744934,
+      "learning_rate": 7.159264290189506e-06,
+      "loss": 0.6006,
+      "step": 2944
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1808959245681763,
+      "learning_rate": 7.157392544938877e-06,
+      "loss": 0.5819,
+      "step": 2945
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5967776775360107,
+      "learning_rate": 7.155520428121233e-06,
+      "loss": 0.649,
+      "step": 2946
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1286388635635376,
+      "learning_rate": 7.153647940059007e-06,
+      "loss": 0.6107,
+      "step": 2947
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.261406421661377,
+      "learning_rate": 7.1517750810746986e-06,
+      "loss": 0.6149,
+      "step": 2948
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5292327404022217,
+      "learning_rate": 7.149901851490869e-06,
+      "loss": 0.6509,
+      "step": 2949
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.308585524559021,
+      "learning_rate": 7.148028251630144e-06,
+      "loss": 0.5977,
+      "step": 2950
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1514670848846436,
+      "learning_rate": 7.146154281815213e-06,
+      "loss": 0.6166,
+      "step": 2951
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0874031782150269,
+      "learning_rate": 7.144279942368829e-06,
+      "loss": 0.6517,
+      "step": 2952
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1332695484161377,
+      "learning_rate": 7.1424052336138094e-06,
+      "loss": 0.5677,
+      "step": 2953
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1511764526367188,
+      "learning_rate": 7.140530155873033e-06,
+      "loss": 0.5318,
+      "step": 2954
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.207506537437439,
+      "learning_rate": 7.138654709469446e-06,
+      "loss": 0.6273,
+      "step": 2955
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4010735750198364,
+      "learning_rate": 7.136778894726055e-06,
+      "loss": 0.6545,
+      "step": 2956
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8122620582580566,
+      "learning_rate": 7.134902711965932e-06,
+      "loss": 0.6774,
+      "step": 2957
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2734031677246094,
+      "learning_rate": 7.133026161512209e-06,
+      "loss": 0.5889,
+      "step": 2958
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1605252027511597,
+      "learning_rate": 7.131149243688086e-06,
+      "loss": 0.7358,
+      "step": 2959
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0211987495422363,
+      "learning_rate": 7.1292719588168225e-06,
+      "loss": 0.5536,
+      "step": 2960
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1764460802078247,
+      "learning_rate": 7.127394307221743e-06,
+      "loss": 0.6021,
+      "step": 2961
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.449872612953186,
+      "learning_rate": 7.125516289226236e-06,
+      "loss": 0.6723,
+      "step": 2962
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2078107595443726,
+      "learning_rate": 7.123637905153749e-06,
+      "loss": 0.5864,
+      "step": 2963
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.082673192024231,
+      "learning_rate": 7.121759155327799e-06,
+      "loss": 0.6392,
+      "step": 2964
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.1781044006347656,
+      "learning_rate": 7.11988004007196e-06,
+      "loss": 0.645,
+      "step": 2965
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0570480823516846,
+      "learning_rate": 7.118000559709872e-06,
+      "loss": 0.56,
+      "step": 2966
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.205815076828003,
+      "learning_rate": 7.1161207145652385e-06,
+      "loss": 0.5751,
+      "step": 2967
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.2318918704986572,
+      "learning_rate": 7.114240504961824e-06,
+      "loss": 0.6159,
+      "step": 2968
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0701419115066528,
+      "learning_rate": 7.112359931223456e-06,
+      "loss": 0.6113,
+      "step": 2969
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3909010887145996,
+      "learning_rate": 7.110478993674023e-06,
+      "loss": 0.6169,
+      "step": 2970
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2270265817642212,
+      "learning_rate": 7.108597692637483e-06,
+      "loss": 0.6437,
+      "step": 2971
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7526935338974,
+      "learning_rate": 7.106716028437848e-06,
+      "loss": 0.5907,
+      "step": 2972
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3577011823654175,
+      "learning_rate": 7.104834001399198e-06,
+      "loss": 0.6427,
+      "step": 2973
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.416495680809021,
+      "learning_rate": 7.102951611845675e-06,
+      "loss": 0.6881,
+      "step": 2974
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.134766936302185,
+      "learning_rate": 7.1010688601014786e-06,
+      "loss": 0.6554,
+      "step": 2975
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2442423105239868,
+      "learning_rate": 7.099185746490878e-06,
+      "loss": 0.5727,
+      "step": 2976
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.257924199104309,
+      "learning_rate": 7.0973022713382e-06,
+      "loss": 0.6091,
+      "step": 2977
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4289751052856445,
+      "learning_rate": 7.0954184349678355e-06,
+      "loss": 0.6965,
+      "step": 2978
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3827462196350098,
+      "learning_rate": 7.0935342377042346e-06,
+      "loss": 0.7763,
+      "step": 2979
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0459548234939575,
+      "learning_rate": 7.091649679871915e-06,
+      "loss": 0.618,
+      "step": 2980
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6075551509857178,
+      "learning_rate": 7.0897647617954536e-06,
+      "loss": 0.5657,
+      "step": 2981
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1650707721710205,
+      "learning_rate": 7.087879483799487e-06,
+      "loss": 0.746,
+      "step": 2982
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0505053997039795,
+      "learning_rate": 7.085993846208718e-06,
+      "loss": 0.6344,
+      "step": 2983
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2678617238998413,
+      "learning_rate": 7.0841078493479066e-06,
+      "loss": 0.6389,
+      "step": 2984
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1751099824905396,
+      "learning_rate": 7.082221493541881e-06,
+      "loss": 0.5815,
+      "step": 2985
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1204549074172974,
+      "learning_rate": 7.080334779115525e-06,
+      "loss": 0.6307,
+      "step": 2986
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.218030571937561,
+      "learning_rate": 7.078447706393788e-06,
+      "loss": 0.6361,
+      "step": 2987
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.160930871963501,
+      "learning_rate": 7.07656027570168e-06,
+      "loss": 0.6299,
+      "step": 2988
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3285592794418335,
+      "learning_rate": 7.07467248736427e-06,
+      "loss": 0.6738,
+      "step": 2989
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2137503623962402,
+      "learning_rate": 7.072784341706696e-06,
+      "loss": 0.6647,
+      "step": 2990
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0857856273651123,
+      "learning_rate": 7.0708958390541485e-06,
+      "loss": 0.5601,
+      "step": 2991
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3946149349212646,
+      "learning_rate": 7.069006979731885e-06,
+      "loss": 0.6107,
+      "step": 2992
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3853344917297363,
+      "learning_rate": 7.067117764065226e-06,
+      "loss": 0.7235,
+      "step": 2993
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.084587574005127,
+      "learning_rate": 7.065228192379545e-06,
+      "loss": 0.5808,
+      "step": 2994
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4154893159866333,
+      "learning_rate": 7.0633382650002854e-06,
+      "loss": 0.7262,
+      "step": 2995
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2447575330734253,
+      "learning_rate": 7.061447982252949e-06,
+      "loss": 0.6407,
+      "step": 2996
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1402736902236938,
+      "learning_rate": 7.0595573444631e-06,
+      "loss": 0.5344,
+      "step": 2997
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4848835468292236,
+      "learning_rate": 7.0576663519563584e-06,
+      "loss": 0.6223,
+      "step": 2998
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1726362705230713,
+      "learning_rate": 7.0557750050584115e-06,
+      "loss": 0.6763,
+      "step": 2999
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2380672693252563,
+      "learning_rate": 7.0538833040950065e-06,
+      "loss": 0.6007,
+      "step": 3000
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.149147629737854,
+      "learning_rate": 7.051991249391947e-06,
+      "loss": 0.5811,
+      "step": 3001
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3859102725982666,
+      "learning_rate": 7.0500988412751044e-06,
+      "loss": 0.5532,
+      "step": 3002
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.054567813873291,
+      "learning_rate": 7.048206080070407e-06,
+      "loss": 0.6674,
+      "step": 3003
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1755194664001465,
+      "learning_rate": 7.046312966103843e-06,
+      "loss": 0.6072,
+      "step": 3004
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3620301485061646,
+      "learning_rate": 7.044419499701462e-06,
+      "loss": 0.6302,
+      "step": 3005
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4070069789886475,
+      "learning_rate": 7.042525681189377e-06,
+      "loss": 0.6966,
+      "step": 3006
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1201298236846924,
+      "learning_rate": 7.0406315108937605e-06,
+      "loss": 0.646,
+      "step": 3007
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.182099461555481,
+      "learning_rate": 7.038736989140843e-06,
+      "loss": 0.6491,
+      "step": 3008
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2445439100265503,
+      "learning_rate": 7.036842116256919e-06,
+      "loss": 0.6441,
+      "step": 3009
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9912326335906982,
+      "learning_rate": 7.034946892568339e-06,
+      "loss": 0.6075,
+      "step": 3010
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.477307915687561,
+      "learning_rate": 7.03305131840152e-06,
+      "loss": 0.5633,
+      "step": 3011
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2548702955245972,
+      "learning_rate": 7.031155394082935e-06,
+      "loss": 0.6102,
+      "step": 3012
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1897332668304443,
+      "learning_rate": 7.029259119939118e-06,
+      "loss": 0.6613,
+      "step": 3013
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2398009300231934,
+      "learning_rate": 7.027362496296662e-06,
+      "loss": 0.6149,
+      "step": 3014
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3539067506790161,
+      "learning_rate": 7.025465523482225e-06,
+      "loss": 0.5495,
+      "step": 3015
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1485373973846436,
+      "learning_rate": 7.023568201822519e-06,
+      "loss": 0.5772,
+      "step": 3016
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3308528661727905,
+      "learning_rate": 7.021670531644324e-06,
+      "loss": 0.5792,
+      "step": 3017
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2606019973754883,
+      "learning_rate": 7.01977251327447e-06,
+      "loss": 0.6744,
+      "step": 3018
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0969258546829224,
+      "learning_rate": 7.017874147039855e-06,
+      "loss": 0.6242,
+      "step": 3019
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1134408712387085,
+      "learning_rate": 7.0159754332674316e-06,
+      "loss": 0.6511,
+      "step": 3020
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1563353538513184,
+      "learning_rate": 7.014076372284217e-06,
+      "loss": 0.712,
+      "step": 3021
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1367099285125732,
+      "learning_rate": 7.012176964417284e-06,
+      "loss": 0.6286,
+      "step": 3022
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.443542718887329,
+      "learning_rate": 7.010277209993769e-06,
+      "loss": 0.6477,
+      "step": 3023
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0937227010726929,
+      "learning_rate": 7.008377109340865e-06,
+      "loss": 0.6286,
+      "step": 3024
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4116562604904175,
+      "learning_rate": 7.006476662785825e-06,
+      "loss": 0.6758,
+      "step": 3025
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9480895400047302,
+      "learning_rate": 7.004575870655963e-06,
+      "loss": 0.5669,
+      "step": 3026
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.328150987625122,
+      "learning_rate": 7.002674733278652e-06,
+      "loss": 0.6269,
+      "step": 3027
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.224358320236206,
+      "learning_rate": 7.000773250981325e-06,
+      "loss": 0.626,
+      "step": 3028
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.514603853225708,
+      "learning_rate": 6.998871424091472e-06,
+      "loss": 0.6341,
+      "step": 3029
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0662037134170532,
+      "learning_rate": 6.996969252936645e-06,
+      "loss": 0.5783,
+      "step": 3030
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2578132152557373,
+      "learning_rate": 6.995066737844454e-06,
+      "loss": 0.6042,
+      "step": 3031
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1706691980361938,
+      "learning_rate": 6.993163879142567e-06,
+      "loss": 0.6136,
+      "step": 3032
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9942347407341003,
+      "learning_rate": 6.991260677158717e-06,
+      "loss": 0.6262,
+      "step": 3033
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1480199098587036,
+      "learning_rate": 6.989357132220686e-06,
+      "loss": 0.5928,
+      "step": 3034
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1092981100082397,
+      "learning_rate": 6.987453244656328e-06,
+      "loss": 0.6324,
+      "step": 3035
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.462157130241394,
+      "learning_rate": 6.985549014793542e-06,
+      "loss": 0.6114,
+      "step": 3036
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9729678630828857,
+      "learning_rate": 6.983644442960299e-06,
+      "loss": 0.644,
+      "step": 3037
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6944612264633179,
+      "learning_rate": 6.9817395294846165e-06,
+      "loss": 0.5726,
+      "step": 3038
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0340354442596436,
+      "learning_rate": 6.979834274694583e-06,
+      "loss": 0.6147,
+      "step": 3039
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3940993547439575,
+      "learning_rate": 6.977928678918335e-06,
+      "loss": 0.6238,
+      "step": 3040
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.217641830444336,
+      "learning_rate": 6.976022742484076e-06,
+      "loss": 0.6713,
+      "step": 3041
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.02081298828125,
+      "learning_rate": 6.974116465720064e-06,
+      "loss": 0.6449,
+      "step": 3042
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6640965938568115,
+      "learning_rate": 6.972209848954617e-06,
+      "loss": 0.6568,
+      "step": 3043
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.213199496269226,
+      "learning_rate": 6.97030289251611e-06,
+      "loss": 0.5438,
+      "step": 3044
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1289682388305664,
+      "learning_rate": 6.968395596732977e-06,
+      "loss": 0.6994,
+      "step": 3045
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1272958517074585,
+      "learning_rate": 6.9664879619337115e-06,
+      "loss": 0.5982,
+      "step": 3046
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7126679420471191,
+      "learning_rate": 6.964579988446866e-06,
+      "loss": 0.5872,
+      "step": 3047
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.227146863937378,
+      "learning_rate": 6.962671676601048e-06,
+      "loss": 0.6131,
+      "step": 3048
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.314324140548706,
+      "learning_rate": 6.960763026724926e-06,
+      "loss": 0.584,
+      "step": 3049
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2320308685302734,
+      "learning_rate": 6.958854039147227e-06,
+      "loss": 0.5748,
+      "step": 3050
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2381142377853394,
+      "learning_rate": 6.956944714196735e-06,
+      "loss": 0.6293,
+      "step": 3051
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3332390785217285,
+      "learning_rate": 6.9550350522022935e-06,
+      "loss": 0.5833,
+      "step": 3052
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1489558219909668,
+      "learning_rate": 6.953125053492801e-06,
+      "loss": 0.5628,
+      "step": 3053
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.03483247756958,
+      "learning_rate": 6.951214718397217e-06,
+      "loss": 0.5724,
+      "step": 3054
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5205141305923462,
+      "learning_rate": 6.9493040472445575e-06,
+      "loss": 0.6287,
+      "step": 3055
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2520228624343872,
+      "learning_rate": 6.947393040363897e-06,
+      "loss": 0.6165,
+      "step": 3056
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1673003435134888,
+      "learning_rate": 6.945481698084366e-06,
+      "loss": 0.6758,
+      "step": 3057
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1318930387496948,
+      "learning_rate": 6.943570020735158e-06,
+      "loss": 0.6002,
+      "step": 3058
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2879729270935059,
+      "learning_rate": 6.941658008645518e-06,
+      "loss": 0.6208,
+      "step": 3059
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3986260890960693,
+      "learning_rate": 6.939745662144751e-06,
+      "loss": 0.5997,
+      "step": 3060
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7398829460144043,
+      "learning_rate": 6.9378329815622215e-06,
+      "loss": 0.6191,
+      "step": 3061
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0883538722991943,
+      "learning_rate": 6.935919967227348e-06,
+      "loss": 0.6555,
+      "step": 3062
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.441008448600769,
+      "learning_rate": 6.93400661946961e-06,
+      "loss": 0.6771,
+      "step": 3063
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.193451166152954,
+      "learning_rate": 6.932092938618541e-06,
+      "loss": 0.6785,
+      "step": 3064
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0834167003631592,
+      "learning_rate": 6.930178925003735e-06,
+      "loss": 0.6142,
+      "step": 3065
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0183688402175903,
+      "learning_rate": 6.928264578954841e-06,
+      "loss": 0.6609,
+      "step": 3066
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2965633869171143,
+      "learning_rate": 6.926349900801568e-06,
+      "loss": 0.6554,
+      "step": 3067
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4456558227539062,
+      "learning_rate": 6.924434890873677e-06,
+      "loss": 0.6321,
+      "step": 3068
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5659126043319702,
+      "learning_rate": 6.922519549500994e-06,
+      "loss": 0.6931,
+      "step": 3069
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1102631092071533,
+      "learning_rate": 6.920603877013393e-06,
+      "loss": 0.6349,
+      "step": 3070
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2046291828155518,
+      "learning_rate": 6.918687873740815e-06,
+      "loss": 0.6647,
+      "step": 3071
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7100063562393188,
+      "learning_rate": 6.916771540013246e-06,
+      "loss": 0.6549,
+      "step": 3072
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1494594812393188,
+      "learning_rate": 6.914854876160741e-06,
+      "loss": 0.619,
+      "step": 3073
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.156104564666748,
+      "learning_rate": 6.912937882513404e-06,
+      "loss": 0.7542,
+      "step": 3074
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.092430591583252,
+      "learning_rate": 6.911020559401399e-06,
+      "loss": 0.5925,
+      "step": 3075
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2527555227279663,
+      "learning_rate": 6.909102907154946e-06,
+      "loss": 0.6515,
+      "step": 3076
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0143706798553467,
+      "learning_rate": 6.90718492610432e-06,
+      "loss": 0.5752,
+      "step": 3077
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.188001036643982,
+      "learning_rate": 6.905266616579857e-06,
+      "loss": 0.5966,
+      "step": 3078
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.197407841682434,
+      "learning_rate": 6.903347978911944e-06,
+      "loss": 0.5806,
+      "step": 3079
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2781100273132324,
+      "learning_rate": 6.9014290134310294e-06,
+      "loss": 0.6362,
+      "step": 3080
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1337027549743652,
+      "learning_rate": 6.899509720467614e-06,
+      "loss": 0.6279,
+      "step": 3081
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2396838665008545,
+      "learning_rate": 6.897590100352261e-06,
+      "loss": 0.5464,
+      "step": 3082
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.947368860244751,
+      "learning_rate": 6.89567015341558e-06,
+      "loss": 0.6205,
+      "step": 3083
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.04353928565979,
+      "learning_rate": 6.893749879988248e-06,
+      "loss": 0.6318,
+      "step": 3084
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1901612281799316,
+      "learning_rate": 6.89182928040099e-06,
+      "loss": 0.5634,
+      "step": 3085
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.215627670288086,
+      "learning_rate": 6.8899083549845914e-06,
+      "loss": 0.6177,
+      "step": 3086
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5702704191207886,
+      "learning_rate": 6.8879871040698935e-06,
+      "loss": 0.6184,
+      "step": 3087
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1875890493392944,
+      "learning_rate": 6.886065527987791e-06,
+      "loss": 0.5802,
+      "step": 3088
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.155051350593567,
+      "learning_rate": 6.884143627069236e-06,
+      "loss": 0.6456,
+      "step": 3089
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4234286546707153,
+      "learning_rate": 6.882221401645239e-06,
+      "loss": 0.6865,
+      "step": 3090
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3715044260025024,
+      "learning_rate": 6.880298852046863e-06,
+      "loss": 0.5297,
+      "step": 3091
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.273527979850769,
+      "learning_rate": 6.878375978605227e-06,
+      "loss": 0.5526,
+      "step": 3092
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.063097596168518,
+      "learning_rate": 6.8764527816515105e-06,
+      "loss": 0.551,
+      "step": 3093
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2897167205810547,
+      "learning_rate": 6.874529261516941e-06,
+      "loss": 0.5874,
+      "step": 3094
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6665688753128052,
+      "learning_rate": 6.872605418532808e-06,
+      "loss": 0.5858,
+      "step": 3095
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0794209241867065,
+      "learning_rate": 6.870681253030453e-06,
+      "loss": 0.6313,
+      "step": 3096
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3062844276428223,
+      "learning_rate": 6.868756765341278e-06,
+      "loss": 0.686,
+      "step": 3097
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1371840238571167,
+      "learning_rate": 6.866831955796731e-06,
+      "loss": 0.525,
+      "step": 3098
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2220420837402344,
+      "learning_rate": 6.864906824728326e-06,
+      "loss": 0.6384,
+      "step": 3099
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.069206237792969,
+      "learning_rate": 6.862981372467626e-06,
+      "loss": 0.6486,
+      "step": 3100
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6746448278427124,
+      "learning_rate": 6.861055599346249e-06,
+      "loss": 0.5713,
+      "step": 3101
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.100169062614441,
+      "learning_rate": 6.859129505695874e-06,
+      "loss": 0.5867,
+      "step": 3102
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2251650094985962,
+      "learning_rate": 6.85720309184823e-06,
+      "loss": 0.61,
+      "step": 3103
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1724867820739746,
+      "learning_rate": 6.855276358135102e-06,
+      "loss": 0.6248,
+      "step": 3104
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0926048755645752,
+      "learning_rate": 6.853349304888331e-06,
+      "loss": 0.5075,
+      "step": 3105
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4889264106750488,
+      "learning_rate": 6.851421932439815e-06,
+      "loss": 0.5743,
+      "step": 3106
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1425611972808838,
+      "learning_rate": 6.8494942411215e-06,
+      "loss": 0.68,
+      "step": 3107
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.138596534729004,
+      "learning_rate": 6.847566231265397e-06,
+      "loss": 0.6524,
+      "step": 3108
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3879531621932983,
+      "learning_rate": 6.845637903203562e-06,
+      "loss": 0.6623,
+      "step": 3109
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7340507507324219,
+      "learning_rate": 6.843709257268112e-06,
+      "loss": 0.6271,
+      "step": 3110
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5906254053115845,
+      "learning_rate": 6.841780293791218e-06,
+      "loss": 0.5912,
+      "step": 3111
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.439652681350708,
+      "learning_rate": 6.839851013105103e-06,
+      "loss": 0.6353,
+      "step": 3112
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2272002696990967,
+      "learning_rate": 6.837921415542048e-06,
+      "loss": 0.6631,
+      "step": 3113
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2197409868240356,
+      "learning_rate": 6.835991501434387e-06,
+      "loss": 0.6065,
+      "step": 3114
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2662650346755981,
+      "learning_rate": 6.834061271114507e-06,
+      "loss": 0.6074,
+      "step": 3115
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.422365665435791,
+      "learning_rate": 6.832130724914852e-06,
+      "loss": 0.676,
+      "step": 3116
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.490148663520813,
+      "learning_rate": 6.830199863167919e-06,
+      "loss": 0.5937,
+      "step": 3117
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.249938726425171,
+      "learning_rate": 6.828268686206259e-06,
+      "loss": 0.6307,
+      "step": 3118
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.534359335899353,
+      "learning_rate": 6.82633719436248e-06,
+      "loss": 0.6081,
+      "step": 3119
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2044472694396973,
+      "learning_rate": 6.82440538796924e-06,
+      "loss": 0.5735,
+      "step": 3120
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0795295238494873,
+      "learning_rate": 6.8224732673592555e-06,
+      "loss": 0.5773,
+      "step": 3121
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2564455270767212,
+      "learning_rate": 6.820540832865293e-06,
+      "loss": 0.592,
+      "step": 3122
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1027977466583252,
+      "learning_rate": 6.818608084820176e-06,
+      "loss": 0.6072,
+      "step": 3123
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.121226191520691,
+      "learning_rate": 6.816675023556781e-06,
+      "loss": 0.6798,
+      "step": 3124
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1786526441574097,
+      "learning_rate": 6.814741649408039e-06,
+      "loss": 0.6565,
+      "step": 3125
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2946691513061523,
+      "learning_rate": 6.812807962706933e-06,
+      "loss": 0.5838,
+      "step": 3126
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.294685959815979,
+      "learning_rate": 6.810873963786501e-06,
+      "loss": 0.5232,
+      "step": 3127
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.5822598934173584,
+      "learning_rate": 6.808939652979839e-06,
+      "loss": 0.6297,
+      "step": 3128
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2533141374588013,
+      "learning_rate": 6.807005030620088e-06,
+      "loss": 0.6384,
+      "step": 3129
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3660823106765747,
+      "learning_rate": 6.805070097040451e-06,
+      "loss": 0.6749,
+      "step": 3130
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3122096061706543,
+      "learning_rate": 6.803134852574177e-06,
+      "loss": 0.6865,
+      "step": 3131
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1023629903793335,
+      "learning_rate": 6.801199297554577e-06,
+      "loss": 0.6321,
+      "step": 3132
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2410534620285034,
+      "learning_rate": 6.799263432315006e-06,
+      "loss": 0.6166,
+      "step": 3133
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.456033706665039,
+      "learning_rate": 6.797327257188882e-06,
+      "loss": 0.6063,
+      "step": 3134
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1916273832321167,
+      "learning_rate": 6.79539077250967e-06,
+      "loss": 0.6006,
+      "step": 3135
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2367602586746216,
+      "learning_rate": 6.793453978610889e-06,
+      "loss": 0.6278,
+      "step": 3136
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3319041728973389,
+      "learning_rate": 6.791516875826115e-06,
+      "loss": 0.6671,
+      "step": 3137
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1352344751358032,
+      "learning_rate": 6.789579464488971e-06,
+      "loss": 0.6093,
+      "step": 3138
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1920417547225952,
+      "learning_rate": 6.787641744933141e-06,
+      "loss": 0.6582,
+      "step": 3139
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1601824760437012,
+      "learning_rate": 6.785703717492355e-06,
+      "loss": 0.5819,
+      "step": 3140
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0924068689346313,
+      "learning_rate": 6.783765382500399e-06,
+      "loss": 0.6438,
+      "step": 3141
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2986295223236084,
+      "learning_rate": 6.781826740291112e-06,
+      "loss": 0.6269,
+      "step": 3142
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9878475666046143,
+      "learning_rate": 6.7798877911983865e-06,
+      "loss": 0.5701,
+      "step": 3143
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0973353385925293,
+      "learning_rate": 6.7779485355561656e-06,
+      "loss": 0.5951,
+      "step": 3144
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1205912828445435,
+      "learning_rate": 6.776008973698449e-06,
+      "loss": 0.6099,
+      "step": 3145
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1063035726547241,
+      "learning_rate": 6.7740691059592844e-06,
+      "loss": 0.5666,
+      "step": 3146
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3901326656341553,
+      "learning_rate": 6.7721289326727765e-06,
+      "loss": 0.6367,
+      "step": 3147
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5512032508850098,
+      "learning_rate": 6.77018845417308e-06,
+      "loss": 0.6069,
+      "step": 3148
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5061674118041992,
+      "learning_rate": 6.768247670794401e-06,
+      "loss": 0.6052,
+      "step": 3149
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2669321298599243,
+      "learning_rate": 6.766306582871004e-06,
+      "loss": 0.593,
+      "step": 3150
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0490036010742188,
+      "learning_rate": 6.764365190737197e-06,
+      "loss": 0.58,
+      "step": 3151
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3043758869171143,
+      "learning_rate": 6.7624234947273495e-06,
+      "loss": 0.6879,
+      "step": 3152
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.293900728225708,
+      "learning_rate": 6.7604814951758764e-06,
+      "loss": 0.6152,
+      "step": 3153
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.1668155193328857,
+      "learning_rate": 6.758539192417251e-06,
+      "loss": 0.5602,
+      "step": 3154
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1483577489852905,
+      "learning_rate": 6.7565965867859914e-06,
+      "loss": 0.6185,
+      "step": 3155
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1590529680252075,
+      "learning_rate": 6.754653678616676e-06,
+      "loss": 0.6432,
+      "step": 3156
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.140228509902954,
+      "learning_rate": 6.752710468243927e-06,
+      "loss": 0.6195,
+      "step": 3157
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4454078674316406,
+      "learning_rate": 6.7507669560024265e-06,
+      "loss": 0.6264,
+      "step": 3158
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2619370222091675,
+      "learning_rate": 6.748823142226902e-06,
+      "loss": 0.7389,
+      "step": 3159
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2213263511657715,
+      "learning_rate": 6.746879027252138e-06,
+      "loss": 0.5855,
+      "step": 3160
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2946794033050537,
+      "learning_rate": 6.74493461141297e-06,
+      "loss": 0.5703,
+      "step": 3161
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.256300926208496,
+      "learning_rate": 6.7429898950442794e-06,
+      "loss": 0.6776,
+      "step": 3162
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.374864101409912,
+      "learning_rate": 6.741044878481009e-06,
+      "loss": 0.6003,
+      "step": 3163
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5750633478164673,
+      "learning_rate": 6.739099562058146e-06,
+      "loss": 0.7092,
+      "step": 3164
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.048159122467041,
+      "learning_rate": 6.737153946110732e-06,
+      "loss": 0.6243,
+      "step": 3165
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.874873399734497,
+      "learning_rate": 6.735208030973858e-06,
+      "loss": 0.5676,
+      "step": 3166
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1411672830581665,
+      "learning_rate": 6.7332618169826725e-06,
+      "loss": 0.5944,
+      "step": 3167
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4558788537979126,
+      "learning_rate": 6.731315304472366e-06,
+      "loss": 0.5991,
+      "step": 3168
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4933147430419922,
+      "learning_rate": 6.7293684937781915e-06,
+      "loss": 0.6637,
+      "step": 3169
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.090269923210144,
+      "learning_rate": 6.727421385235443e-06,
+      "loss": 0.597,
+      "step": 3170
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5020686388015747,
+      "learning_rate": 6.7254739791794735e-06,
+      "loss": 0.6184,
+      "step": 3171
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2008216381072998,
+      "learning_rate": 6.7235262759456824e-06,
+      "loss": 0.5001,
+      "step": 3172
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.278684377670288,
+      "learning_rate": 6.721578275869521e-06,
+      "loss": 0.6604,
+      "step": 3173
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2600221633911133,
+      "learning_rate": 6.719629979286495e-06,
+      "loss": 0.6642,
+      "step": 3174
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3807469606399536,
+      "learning_rate": 6.717681386532158e-06,
+      "loss": 0.5636,
+      "step": 3175
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2727843523025513,
+      "learning_rate": 6.7157324979421145e-06,
+      "loss": 0.6672,
+      "step": 3176
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1497278213500977,
+      "learning_rate": 6.71378331385202e-06,
+      "loss": 0.7587,
+      "step": 3177
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1401848793029785,
+      "learning_rate": 6.711833834597587e-06,
+      "loss": 0.5935,
+      "step": 3178
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0336487293243408,
+      "learning_rate": 6.709884060514568e-06,
+      "loss": 0.5936,
+      "step": 3179
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3813772201538086,
+      "learning_rate": 6.707933991938777e-06,
+      "loss": 0.6915,
+      "step": 3180
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2658019065856934,
+      "learning_rate": 6.705983629206068e-06,
+      "loss": 0.5932,
+      "step": 3181
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.535305142402649,
+      "learning_rate": 6.704032972652357e-06,
+      "loss": 0.6805,
+      "step": 3182
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2994247674942017,
+      "learning_rate": 6.7020820226136e-06,
+      "loss": 0.6971,
+      "step": 3183
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.518620252609253,
+      "learning_rate": 6.700130779425812e-06,
+      "loss": 0.6603,
+      "step": 3184
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4807312488555908,
+      "learning_rate": 6.698179243425053e-06,
+      "loss": 0.5437,
+      "step": 3185
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1867648363113403,
+      "learning_rate": 6.696227414947436e-06,
+      "loss": 0.7545,
+      "step": 3186
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.30292809009552,
+      "learning_rate": 6.694275294329125e-06,
+      "loss": 0.6676,
+      "step": 3187
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5975512266159058,
+      "learning_rate": 6.69232288190633e-06,
+      "loss": 0.6476,
+      "step": 3188
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0806468725204468,
+      "learning_rate": 6.690370178015318e-06,
+      "loss": 0.6023,
+      "step": 3189
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.43095862865448,
+      "learning_rate": 6.688417182992399e-06,
+      "loss": 0.6321,
+      "step": 3190
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.135010838508606,
+      "learning_rate": 6.686463897173942e-06,
+      "loss": 0.6086,
+      "step": 3191
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3971952199935913,
+      "learning_rate": 6.684510320896354e-06,
+      "loss": 0.5269,
+      "step": 3192
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2091296911239624,
+      "learning_rate": 6.6825564544961055e-06,
+      "loss": 0.6593,
+      "step": 3193
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.231992483139038,
+      "learning_rate": 6.680602298309703e-06,
+      "loss": 0.6894,
+      "step": 3194
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6504241228103638,
+      "learning_rate": 6.678647852673717e-06,
+      "loss": 0.6522,
+      "step": 3195
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6003928184509277,
+      "learning_rate": 6.676693117924757e-06,
+      "loss": 0.5055,
+      "step": 3196
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3327730894088745,
+      "learning_rate": 6.674738094399488e-06,
+      "loss": 0.6239,
+      "step": 3197
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1069480180740356,
+      "learning_rate": 6.672782782434622e-06,
+      "loss": 0.5595,
+      "step": 3198
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0996955633163452,
+      "learning_rate": 6.670827182366922e-06,
+      "loss": 0.6455,
+      "step": 3199
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1967326402664185,
+      "learning_rate": 6.668871294533202e-06,
+      "loss": 0.6504,
+      "step": 3200
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4416142702102661,
+      "learning_rate": 6.666915119270322e-06,
+      "loss": 0.7059,
+      "step": 3201
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0445725917816162,
+      "learning_rate": 6.664958656915195e-06,
+      "loss": 0.5707,
+      "step": 3202
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2745569944381714,
+      "learning_rate": 6.663001907804778e-06,
+      "loss": 0.6299,
+      "step": 3203
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.199661374092102,
+      "learning_rate": 6.661044872276086e-06,
+      "loss": 0.6521,
+      "step": 3204
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2170096635818481,
+      "learning_rate": 6.6590875506661764e-06,
+      "loss": 0.5463,
+      "step": 3205
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6790763139724731,
+      "learning_rate": 6.65712994331216e-06,
+      "loss": 0.6315,
+      "step": 3206
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6332316398620605,
+      "learning_rate": 6.655172050551191e-06,
+      "loss": 0.4971,
+      "step": 3207
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.081221580505371,
+      "learning_rate": 6.653213872720481e-06,
+      "loss": 0.7278,
+      "step": 3208
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.130863904953003,
+      "learning_rate": 6.651255410157282e-06,
+      "loss": 0.6655,
+      "step": 3209
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1781548261642456,
+      "learning_rate": 6.649296663198903e-06,
+      "loss": 0.6444,
+      "step": 3210
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1700100898742676,
+      "learning_rate": 6.6473376321826965e-06,
+      "loss": 0.6042,
+      "step": 3211
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1747479438781738,
+      "learning_rate": 6.645378317446066e-06,
+      "loss": 0.6241,
+      "step": 3212
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5794117450714111,
+      "learning_rate": 6.643418719326466e-06,
+      "loss": 0.6243,
+      "step": 3213
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0308579206466675,
+      "learning_rate": 6.6414588381613935e-06,
+      "loss": 0.6249,
+      "step": 3214
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6672332286834717,
+      "learning_rate": 6.6394986742884e-06,
+      "loss": 0.5955,
+      "step": 3215
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2605336904525757,
+      "learning_rate": 6.637538228045084e-06,
+      "loss": 0.6468,
+      "step": 3216
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1389272212982178,
+      "learning_rate": 6.635577499769093e-06,
+      "loss": 0.5899,
+      "step": 3217
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2226412296295166,
+      "learning_rate": 6.633616489798121e-06,
+      "loss": 0.5962,
+      "step": 3218
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1442811489105225,
+      "learning_rate": 6.631655198469915e-06,
+      "loss": 0.7649,
+      "step": 3219
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2384653091430664,
+      "learning_rate": 6.629693626122262e-06,
+      "loss": 0.5698,
+      "step": 3220
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.148032546043396,
+      "learning_rate": 6.62773177309301e-06,
+      "loss": 0.6537,
+      "step": 3221
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6926991939544678,
+      "learning_rate": 6.625769639720045e-06,
+      "loss": 0.5875,
+      "step": 3222
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2067431211471558,
+      "learning_rate": 6.623807226341303e-06,
+      "loss": 0.6435,
+      "step": 3223
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.403552532196045,
+      "learning_rate": 6.621844533294772e-06,
+      "loss": 0.6821,
+      "step": 3224
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3150187730789185,
+      "learning_rate": 6.619881560918485e-06,
+      "loss": 0.5826,
+      "step": 3225
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.287637948989868,
+      "learning_rate": 6.6179183095505265e-06,
+      "loss": 0.6228,
+      "step": 3226
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7602990865707397,
+      "learning_rate": 6.6159547795290214e-06,
+      "loss": 0.5653,
+      "step": 3227
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.151167869567871,
+      "learning_rate": 6.613990971192152e-06,
+      "loss": 0.5097,
+      "step": 3228
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2814674377441406,
+      "learning_rate": 6.6120268848781445e-06,
+      "loss": 0.6207,
+      "step": 3229
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2942862510681152,
+      "learning_rate": 6.610062520925271e-06,
+      "loss": 0.6376,
+      "step": 3230
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.0902457237243652,
+      "learning_rate": 6.608097879671853e-06,
+      "loss": 0.6226,
+      "step": 3231
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3513823747634888,
+      "learning_rate": 6.606132961456264e-06,
+      "loss": 0.6445,
+      "step": 3232
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0098199844360352,
+      "learning_rate": 6.604167766616916e-06,
+      "loss": 0.564,
+      "step": 3233
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3201072216033936,
+      "learning_rate": 6.602202295492277e-06,
+      "loss": 0.5242,
+      "step": 3234
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1454012393951416,
+      "learning_rate": 6.600236548420858e-06,
+      "loss": 0.6332,
+      "step": 3235
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.72659170627594,
+      "learning_rate": 6.59827052574122e-06,
+      "loss": 0.6348,
+      "step": 3236
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.327935814857483,
+      "learning_rate": 6.59630422779197e-06,
+      "loss": 0.5882,
+      "step": 3237
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.219178557395935,
+      "learning_rate": 6.594337654911761e-06,
+      "loss": 0.72,
+      "step": 3238
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1300914287567139,
+      "learning_rate": 6.592370807439299e-06,
+      "loss": 0.6251,
+      "step": 3239
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0695475339889526,
+      "learning_rate": 6.5904036857133315e-06,
+      "loss": 0.5239,
+      "step": 3240
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.289373755455017,
+      "learning_rate": 6.588436290072655e-06,
+      "loss": 0.5697,
+      "step": 3241
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.647684931755066,
+      "learning_rate": 6.586468620856114e-06,
+      "loss": 0.6084,
+      "step": 3242
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1704531908035278,
+      "learning_rate": 6.5845006784025985e-06,
+      "loss": 0.7122,
+      "step": 3243
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6329240798950195,
+      "learning_rate": 6.582532463051048e-06,
+      "loss": 0.6969,
+      "step": 3244
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9630757570266724,
+      "learning_rate": 6.580563975140447e-06,
+      "loss": 0.6295,
+      "step": 3245
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0210407972335815,
+      "learning_rate": 6.578595215009827e-06,
+      "loss": 0.6567,
+      "step": 3246
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.281633734703064,
+      "learning_rate": 6.576626182998267e-06,
+      "loss": 0.5795,
+      "step": 3247
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1632440090179443,
+      "learning_rate": 6.574656879444894e-06,
+      "loss": 0.592,
+      "step": 3248
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1041927337646484,
+      "learning_rate": 6.5726873046888795e-06,
+      "loss": 0.6129,
+      "step": 3249
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.37405526638031,
+      "learning_rate": 6.570717459069442e-06,
+      "loss": 0.7053,
+      "step": 3250
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7148224115371704,
+      "learning_rate": 6.56874734292585e-06,
+      "loss": 0.6797,
+      "step": 3251
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.98465096950531,
+      "learning_rate": 6.5667769565974126e-06,
+      "loss": 0.643,
+      "step": 3252
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3943524360656738,
+      "learning_rate": 6.56480630042349e-06,
+      "loss": 0.6549,
+      "step": 3253
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2270232439041138,
+      "learning_rate": 6.562835374743488e-06,
+      "loss": 0.6189,
+      "step": 3254
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2784587144851685,
+      "learning_rate": 6.5608641798968596e-06,
+      "loss": 0.6567,
+      "step": 3255
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.17548406124115,
+      "learning_rate": 6.558892716223102e-06,
+      "loss": 0.7443,
+      "step": 3256
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4248778820037842,
+      "learning_rate": 6.556920984061759e-06,
+      "loss": 0.6288,
+      "step": 3257
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.57545804977417,
+      "learning_rate": 6.554948983752423e-06,
+      "loss": 0.7004,
+      "step": 3258
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1954271793365479,
+      "learning_rate": 6.552976715634729e-06,
+      "loss": 0.5588,
+      "step": 3259
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.479419231414795,
+      "learning_rate": 6.551004180048361e-06,
+      "loss": 0.6446,
+      "step": 3260
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.04899001121521,
+      "learning_rate": 6.549031377333049e-06,
+      "loss": 0.5631,
+      "step": 3261
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1985399723052979,
+      "learning_rate": 6.5470583078285685e-06,
+      "loss": 0.5939,
+      "step": 3262
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.081529974937439,
+      "learning_rate": 6.545084971874738e-06,
+      "loss": 0.6133,
+      "step": 3263
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3592029809951782,
+      "learning_rate": 6.5431113698114255e-06,
+      "loss": 0.6758,
+      "step": 3264
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.38568115234375,
+      "learning_rate": 6.541137501978547e-06,
+      "loss": 0.6327,
+      "step": 3265
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2125675678253174,
+      "learning_rate": 6.539163368716057e-06,
+      "loss": 0.5552,
+      "step": 3266
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0356985330581665,
+      "learning_rate": 6.537188970363961e-06,
+      "loss": 0.5857,
+      "step": 3267
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4390578269958496,
+      "learning_rate": 6.53521430726231e-06,
+      "loss": 0.6635,
+      "step": 3268
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1902811527252197,
+      "learning_rate": 6.5332393797512e-06,
+      "loss": 0.73,
+      "step": 3269
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1815040111541748,
+      "learning_rate": 6.531264188170769e-06,
+      "loss": 0.678,
+      "step": 3270
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6778348684310913,
+      "learning_rate": 6.529288732861207e-06,
+      "loss": 0.6705,
+      "step": 3271
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5308274030685425,
+      "learning_rate": 6.527313014162745e-06,
+      "loss": 0.6068,
+      "step": 3272
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2709729671478271,
+      "learning_rate": 6.525337032415658e-06,
+      "loss": 0.5982,
+      "step": 3273
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3264434337615967,
+      "learning_rate": 6.523360787960273e-06,
+      "loss": 0.599,
+      "step": 3274
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.82342791557312,
+      "learning_rate": 6.521384281136955e-06,
+      "loss": 0.6239,
+      "step": 3275
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3357826471328735,
+      "learning_rate": 6.519407512286119e-06,
+      "loss": 0.5771,
+      "step": 3276
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4777655601501465,
+      "learning_rate": 6.51743048174822e-06,
+      "loss": 0.6048,
+      "step": 3277
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4173051118850708,
+      "learning_rate": 6.515453189863765e-06,
+      "loss": 0.5204,
+      "step": 3278
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2135194540023804,
+      "learning_rate": 6.513475636973301e-06,
+      "loss": 0.5671,
+      "step": 3279
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3434330224990845,
+      "learning_rate": 6.511497823417418e-06,
+      "loss": 0.576,
+      "step": 3280
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3612425327301025,
+      "learning_rate": 6.50951974953676e-06,
+      "loss": 0.6269,
+      "step": 3281
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.213051438331604,
+      "learning_rate": 6.507541415672007e-06,
+      "loss": 0.6763,
+      "step": 3282
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.151252031326294,
+      "learning_rate": 6.505562822163887e-06,
+      "loss": 0.556,
+      "step": 3283
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.0518524646759033,
+      "learning_rate": 6.503583969353173e-06,
+      "loss": 0.7004,
+      "step": 3284
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3543423414230347,
+      "learning_rate": 6.501604857580681e-06,
+      "loss": 0.6353,
+      "step": 3285
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3768010139465332,
+      "learning_rate": 6.499625487187276e-06,
+      "loss": 0.5926,
+      "step": 3286
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7199901342391968,
+      "learning_rate": 6.497645858513858e-06,
+      "loss": 0.631,
+      "step": 3287
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2099894285202026,
+      "learning_rate": 6.4956659719013835e-06,
+      "loss": 0.5909,
+      "step": 3288
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.031480312347412,
+      "learning_rate": 6.493685827690846e-06,
+      "loss": 0.6396,
+      "step": 3289
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2236201763153076,
+      "learning_rate": 6.491705426223285e-06,
+      "loss": 0.6766,
+      "step": 3290
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6073354482650757,
+      "learning_rate": 6.4897247678397845e-06,
+      "loss": 0.6743,
+      "step": 3291
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1248873472213745,
+      "learning_rate": 6.487743852881472e-06,
+      "loss": 0.5832,
+      "step": 3292
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0797927379608154,
+      "learning_rate": 6.485762681689521e-06,
+      "loss": 0.6391,
+      "step": 3293
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4567921161651611,
+      "learning_rate": 6.483781254605146e-06,
+      "loss": 0.6303,
+      "step": 3294
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9019827842712402,
+      "learning_rate": 6.48179957196961e-06,
+      "loss": 0.5922,
+      "step": 3295
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6069629192352295,
+      "learning_rate": 6.479817634124216e-06,
+      "loss": 0.6879,
+      "step": 3296
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3667546510696411,
+      "learning_rate": 6.477835441410311e-06,
+      "loss": 0.65,
+      "step": 3297
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.415429711341858,
+      "learning_rate": 6.475852994169294e-06,
+      "loss": 0.6479,
+      "step": 3298
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2002050876617432,
+      "learning_rate": 6.473870292742592e-06,
+      "loss": 0.5623,
+      "step": 3299
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5991908311843872,
+      "learning_rate": 6.471887337471693e-06,
+      "loss": 0.6927,
+      "step": 3300
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.442886471748352,
+      "learning_rate": 6.4699041286981155e-06,
+      "loss": 0.6233,
+      "step": 3301
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.619641661643982,
+      "learning_rate": 6.46792066676343e-06,
+      "loss": 0.6882,
+      "step": 3302
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.233254313468933,
+      "learning_rate": 6.465936952009245e-06,
+      "loss": 0.6169,
+      "step": 3303
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1515586376190186,
+      "learning_rate": 6.463952984777218e-06,
+      "loss": 0.5435,
+      "step": 3304
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2402045726776123,
+      "learning_rate": 6.461968765409041e-06,
+      "loss": 0.7376,
+      "step": 3305
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3234292268753052,
+      "learning_rate": 6.4599842942464645e-06,
+      "loss": 0.7769,
+      "step": 3306
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9401755332946777,
+      "learning_rate": 6.457999571631266e-06,
+      "loss": 0.6768,
+      "step": 3307
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.493912696838379,
+      "learning_rate": 6.456014597905278e-06,
+      "loss": 0.5967,
+      "step": 3308
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2056974172592163,
+      "learning_rate": 6.454029373410369e-06,
+      "loss": 0.5962,
+      "step": 3309
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.219796061515808,
+      "learning_rate": 6.452043898488456e-06,
+      "loss": 0.6247,
+      "step": 3310
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3913002014160156,
+      "learning_rate": 6.450058173481493e-06,
+      "loss": 0.5459,
+      "step": 3311
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.591672658920288,
+      "learning_rate": 6.448072198731485e-06,
+      "loss": 0.5614,
+      "step": 3312
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4829857349395752,
+      "learning_rate": 6.4460859745804735e-06,
+      "loss": 0.6168,
+      "step": 3313
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0522713661193848,
+      "learning_rate": 6.444099501370545e-06,
+      "loss": 0.5752,
+      "step": 3314
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.390140175819397,
+      "learning_rate": 6.442112779443832e-06,
+      "loss": 0.6122,
+      "step": 3315
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.25529944896698,
+      "learning_rate": 6.440125809142503e-06,
+      "loss": 0.5556,
+      "step": 3316
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3403065204620361,
+      "learning_rate": 6.438138590808776e-06,
+      "loss": 0.6308,
+      "step": 3317
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.25216543674469,
+      "learning_rate": 6.436151124784906e-06,
+      "loss": 0.6417,
+      "step": 3318
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2277629375457764,
+      "learning_rate": 6.434163411413197e-06,
+      "loss": 0.6425,
+      "step": 3319
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0705773830413818,
+      "learning_rate": 6.432175451035991e-06,
+      "loss": 0.5152,
+      "step": 3320
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.466159701347351,
+      "learning_rate": 6.430187243995674e-06,
+      "loss": 0.6207,
+      "step": 3321
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1634601354599,
+      "learning_rate": 6.428198790634672e-06,
+      "loss": 0.5681,
+      "step": 3322
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6939680576324463,
+      "learning_rate": 6.4262100912954585e-06,
+      "loss": 0.593,
+      "step": 3323
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3650009632110596,
+      "learning_rate": 6.424221146320547e-06,
+      "loss": 0.6422,
+      "step": 3324
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1898249387741089,
+      "learning_rate": 6.422231956052489e-06,
+      "loss": 0.6359,
+      "step": 3325
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2429956197738647,
+      "learning_rate": 6.420242520833886e-06,
+      "loss": 0.6454,
+      "step": 3326
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1976206302642822,
+      "learning_rate": 6.418252841007376e-06,
+      "loss": 0.5683,
+      "step": 3327
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0774952173233032,
+      "learning_rate": 6.416262916915642e-06,
+      "loss": 0.6135,
+      "step": 3328
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.288520336151123,
+      "learning_rate": 6.414272748901405e-06,
+      "loss": 0.6605,
+      "step": 3329
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.0824525356292725,
+      "learning_rate": 6.412282337307436e-06,
+      "loss": 0.5897,
+      "step": 3330
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8679497241973877,
+      "learning_rate": 6.410291682476537e-06,
+      "loss": 0.5329,
+      "step": 3331
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.217800498008728,
+      "learning_rate": 6.4083007847515634e-06,
+      "loss": 0.6684,
+      "step": 3332
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3628228902816772,
+      "learning_rate": 6.406309644475404e-06,
+      "loss": 0.6208,
+      "step": 3333
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4911750555038452,
+      "learning_rate": 6.404318261990992e-06,
+      "loss": 0.6059,
+      "step": 3334
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5707592964172363,
+      "learning_rate": 6.402326637641303e-06,
+      "loss": 0.5948,
+      "step": 3335
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3157973289489746,
+      "learning_rate": 6.400334771769355e-06,
+      "loss": 0.6137,
+      "step": 3336
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2203538417816162,
+      "learning_rate": 6.398342664718204e-06,
+      "loss": 0.607,
+      "step": 3337
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2806761264801025,
+      "learning_rate": 6.396350316830954e-06,
+      "loss": 0.5865,
+      "step": 3338
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.370572805404663,
+      "learning_rate": 6.394357728450741e-06,
+      "loss": 0.6144,
+      "step": 3339
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2665327787399292,
+      "learning_rate": 6.392364899920751e-06,
+      "loss": 0.5818,
+      "step": 3340
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.244334101676941,
+      "learning_rate": 6.39037183158421e-06,
+      "loss": 0.6458,
+      "step": 3341
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2308834791183472,
+      "learning_rate": 6.388378523784379e-06,
+      "loss": 0.7208,
+      "step": 3342
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2577228546142578,
+      "learning_rate": 6.386384976864569e-06,
+      "loss": 0.7069,
+      "step": 3343
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1200151443481445,
+      "learning_rate": 6.384391191168124e-06,
+      "loss": 0.6288,
+      "step": 3344
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3991978168487549,
+      "learning_rate": 6.382397167038438e-06,
+      "loss": 0.616,
+      "step": 3345
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.689049482345581,
+      "learning_rate": 6.3804029048189355e-06,
+      "loss": 0.6469,
+      "step": 3346
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1596959829330444,
+      "learning_rate": 6.378408404853093e-06,
+      "loss": 0.538,
+      "step": 3347
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1679819822311401,
+      "learning_rate": 6.376413667484417e-06,
+      "loss": 0.7596,
+      "step": 3348
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8966916799545288,
+      "learning_rate": 6.374418693056464e-06,
+      "loss": 0.6038,
+      "step": 3349
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.218559741973877,
+      "learning_rate": 6.37242348191283e-06,
+      "loss": 0.6026,
+      "step": 3350
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2759473323822021,
+      "learning_rate": 6.370428034397144e-06,
+      "loss": 0.5726,
+      "step": 3351
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0636835098266602,
+      "learning_rate": 6.368432350853085e-06,
+      "loss": 0.6214,
+      "step": 3352
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.555844783782959,
+      "learning_rate": 6.366436431624368e-06,
+      "loss": 0.622,
+      "step": 3353
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8688308000564575,
+      "learning_rate": 6.3644402770547496e-06,
+      "loss": 0.6356,
+      "step": 3354
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9006859660148621,
+      "learning_rate": 6.3624438874880256e-06,
+      "loss": 0.6228,
+      "step": 3355
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.367267370223999,
+      "learning_rate": 6.360447263268037e-06,
+      "loss": 0.5737,
+      "step": 3356
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1289016008377075,
+      "learning_rate": 6.358450404738656e-06,
+      "loss": 0.6597,
+      "step": 3357
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5956635475158691,
+      "learning_rate": 6.356453312243807e-06,
+      "loss": 0.5691,
+      "step": 3358
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2158634662628174,
+      "learning_rate": 6.354455986127445e-06,
+      "loss": 0.5911,
+      "step": 3359
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.0845446586608887,
+      "learning_rate": 6.352458426733571e-06,
+      "loss": 0.5655,
+      "step": 3360
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3381253480911255,
+      "learning_rate": 6.3504606344062215e-06,
+      "loss": 0.6276,
+      "step": 3361
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.382815957069397,
+      "learning_rate": 6.348462609489477e-06,
+      "loss": 0.6372,
+      "step": 3362
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.607489824295044,
+      "learning_rate": 6.346464352327456e-06,
+      "loss": 0.7139,
+      "step": 3363
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.351166009902954,
+      "learning_rate": 6.3444658632643195e-06,
+      "loss": 0.5938,
+      "step": 3364
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.418544292449951,
+      "learning_rate": 6.342467142644264e-06,
+      "loss": 0.5654,
+      "step": 3365
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2091163396835327,
+      "learning_rate": 6.340468190811531e-06,
+      "loss": 0.6736,
+      "step": 3366
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4336068630218506,
+      "learning_rate": 6.338469008110399e-06,
+      "loss": 0.693,
+      "step": 3367
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4184355735778809,
+      "learning_rate": 6.336469594885183e-06,
+      "loss": 0.6291,
+      "step": 3368
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5481655597686768,
+      "learning_rate": 6.3344699514802465e-06,
+      "loss": 0.6017,
+      "step": 3369
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1255567073822021,
+      "learning_rate": 6.332470078239983e-06,
+      "loss": 0.6172,
+      "step": 3370
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1182490587234497,
+      "learning_rate": 6.330469975508834e-06,
+      "loss": 0.584,
+      "step": 3371
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1752885580062866,
+      "learning_rate": 6.328469643631273e-06,
+      "loss": 0.7249,
+      "step": 3372
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3815187215805054,
+      "learning_rate": 6.326469082951819e-06,
+      "loss": 0.635,
+      "step": 3373
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2596144676208496,
+      "learning_rate": 6.3244682938150255e-06,
+      "loss": 0.5281,
+      "step": 3374
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3565584421157837,
+      "learning_rate": 6.3224672765654905e-06,
+      "loss": 0.6269,
+      "step": 3375
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5856893062591553,
+      "learning_rate": 6.320466031547847e-06,
+      "loss": 0.6211,
+      "step": 3376
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1341127157211304,
+      "learning_rate": 6.31846455910677e-06,
+      "loss": 0.5976,
+      "step": 3377
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2035582065582275,
+      "learning_rate": 6.316462859586971e-06,
+      "loss": 0.5688,
+      "step": 3378
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.234596610069275,
+      "learning_rate": 6.314460933333201e-06,
+      "loss": 0.613,
+      "step": 3379
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.051673412322998,
+      "learning_rate": 6.312458780690254e-06,
+      "loss": 0.4786,
+      "step": 3380
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.180688500404358,
+      "learning_rate": 6.310456402002958e-06,
+      "loss": 0.6154,
+      "step": 3381
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.144311547279358,
+      "learning_rate": 6.308453797616184e-06,
+      "loss": 0.7581,
+      "step": 3382
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3280012607574463,
+      "learning_rate": 6.306450967874836e-06,
+      "loss": 0.6299,
+      "step": 3383
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2186527252197266,
+      "learning_rate": 6.304447913123866e-06,
+      "loss": 0.5792,
+      "step": 3384
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3583048582077026,
+      "learning_rate": 6.3024446337082555e-06,
+      "loss": 0.6775,
+      "step": 3385
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3401097059249878,
+      "learning_rate": 6.300441129973032e-06,
+      "loss": 0.6299,
+      "step": 3386
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4296989440917969,
+      "learning_rate": 6.298437402263254e-06,
+      "loss": 0.6503,
+      "step": 3387
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3577499389648438,
+      "learning_rate": 6.296433450924027e-06,
+      "loss": 0.6072,
+      "step": 3388
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4429153203964233,
+      "learning_rate": 6.2944292763004885e-06,
+      "loss": 0.5407,
+      "step": 3389
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1351635456085205,
+      "learning_rate": 6.292424878737817e-06,
+      "loss": 0.5831,
+      "step": 3390
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.355637550354004,
+      "learning_rate": 6.290420258581229e-06,
+      "loss": 0.6465,
+      "step": 3391
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0525286197662354,
+      "learning_rate": 6.288415416175981e-06,
+      "loss": 0.5857,
+      "step": 3392
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.313124656677246,
+      "learning_rate": 6.286410351867367e-06,
+      "loss": 0.5992,
+      "step": 3393
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.2575294971466064,
+      "learning_rate": 6.284405066000715e-06,
+      "loss": 0.5893,
+      "step": 3394
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1958906650543213,
+      "learning_rate": 6.282399558921398e-06,
+      "loss": 0.5755,
+      "step": 3395
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1793793439865112,
+      "learning_rate": 6.280393830974822e-06,
+      "loss": 0.6098,
+      "step": 3396
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2796311378479004,
+      "learning_rate": 6.278387882506434e-06,
+      "loss": 0.6819,
+      "step": 3397
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8948205709457397,
+      "learning_rate": 6.276381713861717e-06,
+      "loss": 0.6131,
+      "step": 3398
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.889975905418396,
+      "learning_rate": 6.274375325386195e-06,
+      "loss": 0.6074,
+      "step": 3399
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6620714664459229,
+      "learning_rate": 6.272368717425423e-06,
+      "loss": 0.5899,
+      "step": 3400
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.221790075302124,
+      "learning_rate": 6.270361890325003e-06,
+      "loss": 0.5863,
+      "step": 3401
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1769517660140991,
+      "learning_rate": 6.268354844430569e-06,
+      "loss": 0.5553,
+      "step": 3402
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1347368955612183,
+      "learning_rate": 6.266347580087791e-06,
+      "loss": 0.7537,
+      "step": 3403
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2931528091430664,
+      "learning_rate": 6.264340097642385e-06,
+      "loss": 0.5663,
+      "step": 3404
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5123640298843384,
+      "learning_rate": 6.2623323974400944e-06,
+      "loss": 0.571,
+      "step": 3405
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1496232748031616,
+      "learning_rate": 6.260324479826706e-06,
+      "loss": 0.7423,
+      "step": 3406
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2281783819198608,
+      "learning_rate": 6.258316345148042e-06,
+      "loss": 0.6239,
+      "step": 3407
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.229547142982483,
+      "learning_rate": 6.256307993749965e-06,
+      "loss": 0.5493,
+      "step": 3408
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5989558696746826,
+      "learning_rate": 6.254299425978371e-06,
+      "loss": 0.5408,
+      "step": 3409
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.251488208770752,
+      "learning_rate": 6.252290642179197e-06,
+      "loss": 0.6038,
+      "step": 3410
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.296891450881958,
+      "learning_rate": 6.25028164269841e-06,
+      "loss": 0.6108,
+      "step": 3411
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.524884819984436,
+      "learning_rate": 6.248272427882027e-06,
+      "loss": 0.683,
+      "step": 3412
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0750436782836914,
+      "learning_rate": 6.246262998076088e-06,
+      "loss": 0.69,
+      "step": 3413
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3640230894088745,
+      "learning_rate": 6.24425335362668e-06,
+      "loss": 0.6142,
+      "step": 3414
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7753424644470215,
+      "learning_rate": 6.242243494879923e-06,
+      "loss": 0.6733,
+      "step": 3415
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3016899824142456,
+      "learning_rate": 6.240233422181972e-06,
+      "loss": 0.5632,
+      "step": 3416
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.370159387588501,
+      "learning_rate": 6.2382231358790224e-06,
+      "loss": 0.6562,
+      "step": 3417
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9591825008392334,
+      "learning_rate": 6.236212636317305e-06,
+      "loss": 0.6253,
+      "step": 3418
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2164580821990967,
+      "learning_rate": 6.23420192384309e-06,
+      "loss": 0.5545,
+      "step": 3419
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5938736200332642,
+      "learning_rate": 6.232190998802679e-06,
+      "loss": 0.5702,
+      "step": 3420
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2773017883300781,
+      "learning_rate": 6.230179861542413e-06,
+      "loss": 0.6788,
+      "step": 3421
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3805594444274902,
+      "learning_rate": 6.2281685124086714e-06,
+      "loss": 0.5953,
+      "step": 3422
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4324122667312622,
+      "learning_rate": 6.226156951747866e-06,
+      "loss": 0.5984,
+      "step": 3423
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1922259330749512,
+      "learning_rate": 6.224145179906448e-06,
+      "loss": 0.6201,
+      "step": 3424
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.930484652519226,
+      "learning_rate": 6.222133197230904e-06,
+      "loss": 0.6431,
+      "step": 3425
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.209879755973816,
+      "learning_rate": 6.22012100406776e-06,
+      "loss": 0.5888,
+      "step": 3426
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4153494834899902,
+      "learning_rate": 6.21810860076357e-06,
+      "loss": 0.6015,
+      "step": 3427
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4506012201309204,
+      "learning_rate": 6.216095987664935e-06,
+      "loss": 0.6142,
+      "step": 3428
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0460131168365479,
+      "learning_rate": 6.214083165118483e-06,
+      "loss": 0.5344,
+      "step": 3429
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4760301113128662,
+      "learning_rate": 6.212070133470884e-06,
+      "loss": 0.6131,
+      "step": 3430
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3164763450622559,
+      "learning_rate": 6.210056893068839e-06,
+      "loss": 0.607,
+      "step": 3431
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4441686868667603,
+      "learning_rate": 6.208043444259091e-06,
+      "loss": 0.7008,
+      "step": 3432
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0722826719284058,
+      "learning_rate": 6.206029787388412e-06,
+      "loss": 0.5562,
+      "step": 3433
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1804120540618896,
+      "learning_rate": 6.204015922803617e-06,
+      "loss": 0.5505,
+      "step": 3434
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.25029718875885,
+      "learning_rate": 6.202001850851551e-06,
+      "loss": 0.6006,
+      "step": 3435
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1206964254379272,
+      "learning_rate": 6.199987571879101e-06,
+      "loss": 0.7474,
+      "step": 3436
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8449739217758179,
+      "learning_rate": 6.197973086233179e-06,
+      "loss": 0.651,
+      "step": 3437
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.151208758354187,
+      "learning_rate": 6.195958394260744e-06,
+      "loss": 0.5435,
+      "step": 3438
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3795284032821655,
+      "learning_rate": 6.1939434963087845e-06,
+      "loss": 0.6664,
+      "step": 3439
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3607579469680786,
+      "learning_rate": 6.191928392724325e-06,
+      "loss": 0.5944,
+      "step": 3440
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5826529264450073,
+      "learning_rate": 6.1899130838544275e-06,
+      "loss": 0.645,
+      "step": 3441
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0307071208953857,
+      "learning_rate": 6.187897570046184e-06,
+      "loss": 0.5332,
+      "step": 3442
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3897781372070312,
+      "learning_rate": 6.185881851646732e-06,
+      "loss": 0.652,
+      "step": 3443
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9002603888511658,
+      "learning_rate": 6.1838659290032346e-06,
+      "loss": 0.5409,
+      "step": 3444
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2274670600891113,
+      "learning_rate": 6.181849802462895e-06,
+      "loss": 0.5836,
+      "step": 3445
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3531590700149536,
+      "learning_rate": 6.179833472372947e-06,
+      "loss": 0.598,
+      "step": 3446
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2528719902038574,
+      "learning_rate": 6.177816939080666e-06,
+      "loss": 0.5959,
+      "step": 3447
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0928164720535278,
+      "learning_rate": 6.175800202933355e-06,
+      "loss": 0.5685,
+      "step": 3448
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4518722295761108,
+      "learning_rate": 6.17378326427836e-06,
+      "loss": 0.6363,
+      "step": 3449
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.127073884010315,
+      "learning_rate": 6.171766123463054e-06,
+      "loss": 0.79,
+      "step": 3450
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7623597383499146,
+      "learning_rate": 6.1697487808348525e-06,
+      "loss": 0.6219,
+      "step": 3451
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.348456621170044,
+      "learning_rate": 6.167731236741199e-06,
+      "loss": 0.5674,
+      "step": 3452
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6265017986297607,
+      "learning_rate": 6.1657134915295735e-06,
+      "loss": 0.5659,
+      "step": 3453
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.057990550994873,
+      "learning_rate": 6.163695545547494e-06,
+      "loss": 0.6351,
+      "step": 3454
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.3038055896759033,
+      "learning_rate": 6.16167739914251e-06,
+      "loss": 0.6228,
+      "step": 3455
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0282056331634521,
+      "learning_rate": 6.159659052662207e-06,
+      "loss": 0.5298,
+      "step": 3456
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.179836630821228,
+      "learning_rate": 6.157640506454201e-06,
+      "loss": 0.5805,
+      "step": 3457
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.278550148010254,
+      "learning_rate": 6.15562176086615e-06,
+      "loss": 0.6652,
+      "step": 3458
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3121113777160645,
+      "learning_rate": 6.153602816245738e-06,
+      "loss": 0.6242,
+      "step": 3459
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.404284119606018,
+      "learning_rate": 6.1515836729406874e-06,
+      "loss": 0.6148,
+      "step": 3460
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0569359064102173,
+      "learning_rate": 6.149564331298758e-06,
+      "loss": 0.605,
+      "step": 3461
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.422186017036438,
+      "learning_rate": 6.147544791667738e-06,
+      "loss": 0.6296,
+      "step": 3462
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2817769050598145,
+      "learning_rate": 6.145525054395453e-06,
+      "loss": 0.6151,
+      "step": 3463
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.4974191188812256,
+      "learning_rate": 6.143505119829759e-06,
+      "loss": 0.5476,
+      "step": 3464
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.151680827140808,
+      "learning_rate": 6.141484988318554e-06,
+      "loss": 0.588,
+      "step": 3465
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0980405807495117,
+      "learning_rate": 6.139464660209757e-06,
+      "loss": 0.5949,
+      "step": 3466
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.5018017292022705,
+      "learning_rate": 6.137444135851336e-06,
+      "loss": 0.5604,
+      "step": 3467
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1532684564590454,
+      "learning_rate": 6.1354234155912795e-06,
+      "loss": 0.6558,
+      "step": 3468
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3535691499710083,
+      "learning_rate": 6.133402499777621e-06,
+      "loss": 0.6153,
+      "step": 3469
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7174805402755737,
+      "learning_rate": 6.131381388758417e-06,
+      "loss": 0.5909,
+      "step": 3470
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1755709648132324,
+      "learning_rate": 6.1293600828817655e-06,
+      "loss": 0.5663,
+      "step": 3471
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8840426206588745,
+      "learning_rate": 6.127338582495796e-06,
+      "loss": 0.6317,
+      "step": 3472
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3639739751815796,
+      "learning_rate": 6.125316887948668e-06,
+      "loss": 0.6246,
+      "step": 3473
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.087632179260254,
+      "learning_rate": 6.123294999588579e-06,
+      "loss": 0.5259,
+      "step": 3474
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6002978086471558,
+      "learning_rate": 6.1212729177637595e-06,
+      "loss": 0.5469,
+      "step": 3475
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.169373869895935,
+      "learning_rate": 6.119250642822469e-06,
+      "loss": 0.646,
+      "step": 3476
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.6361258029937744,
+      "learning_rate": 6.117228175113006e-06,
+      "loss": 0.6835,
+      "step": 3477
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1000622510910034,
+      "learning_rate": 6.115205514983699e-06,
+      "loss": 0.621,
+      "step": 3478
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3290541172027588,
+      "learning_rate": 6.1131826627829085e-06,
+      "loss": 0.6461,
+      "step": 3479
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3654758930206299,
+      "learning_rate": 6.11115961885903e-06,
+      "loss": 0.6303,
+      "step": 3480
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3180681467056274,
+      "learning_rate": 6.109136383560494e-06,
+      "loss": 0.6891,
+      "step": 3481
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4984737634658813,
+      "learning_rate": 6.107112957235759e-06,
+      "loss": 0.6644,
+      "step": 3482
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4471309185028076,
+      "learning_rate": 6.10508934023332e-06,
+      "loss": 0.5985,
+      "step": 3483
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3030848503112793,
+      "learning_rate": 6.103065532901704e-06,
+      "loss": 0.7308,
+      "step": 3484
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.357759714126587,
+      "learning_rate": 6.101041535589469e-06,
+      "loss": 0.6637,
+      "step": 3485
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5871968269348145,
+      "learning_rate": 6.099017348645212e-06,
+      "loss": 0.6919,
+      "step": 3486
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0771772861480713,
+      "learning_rate": 6.096992972417552e-06,
+      "loss": 0.6352,
+      "step": 3487
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3609187602996826,
+      "learning_rate": 6.094968407255153e-06,
+      "loss": 0.6257,
+      "step": 3488
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1798498630523682,
+      "learning_rate": 6.092943653506701e-06,
+      "loss": 0.59,
+      "step": 3489
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.488000512123108,
+      "learning_rate": 6.090918711520918e-06,
+      "loss": 0.6466,
+      "step": 3490
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2035020589828491,
+      "learning_rate": 6.088893581646562e-06,
+      "loss": 0.5723,
+      "step": 3491
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.004976749420166,
+      "learning_rate": 6.086868264232418e-06,
+      "loss": 0.6055,
+      "step": 3492
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.643667697906494,
+      "learning_rate": 6.084842759627309e-06,
+      "loss": 0.5617,
+      "step": 3493
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.7203829288482666,
+      "learning_rate": 6.082817068180081e-06,
+      "loss": 0.6339,
+      "step": 3494
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1511204242706299,
+      "learning_rate": 6.0807911902396255e-06,
+      "loss": 0.6117,
+      "step": 3495
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4033151865005493,
+      "learning_rate": 6.078765126154854e-06,
+      "loss": 0.5826,
+      "step": 3496
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.21249520778656,
+      "learning_rate": 6.076738876274718e-06,
+      "loss": 0.6407,
+      "step": 3497
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0692471265792847,
+      "learning_rate": 6.074712440948194e-06,
+      "loss": 0.586,
+      "step": 3498
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.459241509437561,
+      "learning_rate": 6.0726858205242974e-06,
+      "loss": 0.581,
+      "step": 3499
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.218548059463501,
+      "learning_rate": 6.070659015352072e-06,
+      "loss": 0.5934,
+      "step": 3500
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2451094388961792,
+      "learning_rate": 6.068632025780592e-06,
+      "loss": 0.6442,
+      "step": 3501
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7080090045928955,
+      "learning_rate": 6.066604852158966e-06,
+      "loss": 0.6385,
+      "step": 3502
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4963022470474243,
+      "learning_rate": 6.064577494836334e-06,
+      "loss": 0.5995,
+      "step": 3503
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1907035112380981,
+      "learning_rate": 6.0625499541618686e-06,
+      "loss": 0.6659,
+      "step": 3504
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9348851442337036,
+      "learning_rate": 6.060522230484769e-06,
+      "loss": 0.5889,
+      "step": 3505
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1017441749572754,
+      "learning_rate": 6.058494324154272e-06,
+      "loss": 0.6732,
+      "step": 3506
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.096999168395996,
+      "learning_rate": 6.056466235519641e-06,
+      "loss": 0.5974,
+      "step": 3507
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2315013408660889,
+      "learning_rate": 6.054437964930175e-06,
+      "loss": 0.6482,
+      "step": 3508
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2487025260925293,
+      "learning_rate": 6.052409512735202e-06,
+      "loss": 0.6511,
+      "step": 3509
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2520768642425537,
+      "learning_rate": 6.0503808792840825e-06,
+      "loss": 0.6488,
+      "step": 3510
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1143486499786377,
+      "learning_rate": 6.048352064926204e-06,
+      "loss": 0.6537,
+      "step": 3511
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.371921181678772,
+      "learning_rate": 6.046323070010993e-06,
+      "loss": 0.5731,
+      "step": 3512
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.219484806060791,
+      "learning_rate": 6.0442938948879006e-06,
+      "loss": 0.5992,
+      "step": 3513
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.838356614112854,
+      "learning_rate": 6.04226453990641e-06,
+      "loss": 0.591,
+      "step": 3514
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2704648971557617,
+      "learning_rate": 6.040235005416037e-06,
+      "loss": 0.6266,
+      "step": 3515
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7678539752960205,
+      "learning_rate": 6.038205291766328e-06,
+      "loss": 0.6643,
+      "step": 3516
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7963787317276,
+      "learning_rate": 6.036175399306861e-06,
+      "loss": 0.628,
+      "step": 3517
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.342311143875122,
+      "learning_rate": 6.034145328387239e-06,
+      "loss": 0.6382,
+      "step": 3518
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.220090627670288,
+      "learning_rate": 6.032115079357108e-06,
+      "loss": 0.5598,
+      "step": 3519
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0853689908981323,
+      "learning_rate": 6.0300846525661305e-06,
+      "loss": 0.5895,
+      "step": 3520
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2728458642959595,
+      "learning_rate": 6.028054048364011e-06,
+      "loss": 0.645,
+      "step": 3521
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1695315837860107,
+      "learning_rate": 6.026023267100474e-06,
+      "loss": 0.5413,
+      "step": 3522
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3852046728134155,
+      "learning_rate": 6.023992309125288e-06,
+      "loss": 0.6413,
+      "step": 3523
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.460791110992432,
+      "learning_rate": 6.021961174788239e-06,
+      "loss": 0.6022,
+      "step": 3524
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5186138153076172,
+      "learning_rate": 6.019929864439149e-06,
+      "loss": 0.6042,
+      "step": 3525
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.498050332069397,
+      "learning_rate": 6.017898378427871e-06,
+      "loss": 0.7024,
+      "step": 3526
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3453694581985474,
+      "learning_rate": 6.015866717104288e-06,
+      "loss": 0.6183,
+      "step": 3527
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3376102447509766,
+      "learning_rate": 6.01383488081831e-06,
+      "loss": 0.6636,
+      "step": 3528
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0858659744262695,
+      "learning_rate": 6.01180286991988e-06,
+      "loss": 0.7165,
+      "step": 3529
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.561975121498108,
+      "learning_rate": 6.009770684758973e-06,
+      "loss": 0.6641,
+      "step": 3530
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.422297716140747,
+      "learning_rate": 6.007738325685588e-06,
+      "loss": 0.6278,
+      "step": 3531
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3532339334487915,
+      "learning_rate": 6.005705793049762e-06,
+      "loss": 0.5749,
+      "step": 3532
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1300181150436401,
+      "learning_rate": 6.0036730872015524e-06,
+      "loss": 0.6098,
+      "step": 3533
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3219990730285645,
+      "learning_rate": 6.001640208491056e-06,
+      "loss": 0.5893,
+      "step": 3534
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.5730981826782227,
+      "learning_rate": 5.9996071572683914e-06,
+      "loss": 0.553,
+      "step": 3535
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.663468837738037,
+      "learning_rate": 5.997573933883714e-06,
+      "loss": 0.5648,
+      "step": 3536
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.451084017753601,
+      "learning_rate": 5.995540538687199e-06,
+      "loss": 0.6901,
+      "step": 3537
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.200803518295288,
+      "learning_rate": 5.993506972029064e-06,
+      "loss": 0.595,
+      "step": 3538
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3902348279953003,
+      "learning_rate": 5.991473234259546e-06,
+      "loss": 0.6697,
+      "step": 3539
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.0828068256378174,
+      "learning_rate": 5.989439325728916e-06,
+      "loss": 0.6027,
+      "step": 3540
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1694917678833008,
+      "learning_rate": 5.987405246787474e-06,
+      "loss": 0.5812,
+      "step": 3541
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7300596237182617,
+      "learning_rate": 5.985370997785547e-06,
+      "loss": 0.5532,
+      "step": 3542
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1379985809326172,
+      "learning_rate": 5.983336579073495e-06,
+      "loss": 0.6126,
+      "step": 3543
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.147905945777893,
+      "learning_rate": 5.981301991001703e-06,
+      "loss": 0.5544,
+      "step": 3544
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.108622431755066,
+      "learning_rate": 5.979267233920589e-06,
+      "loss": 0.5836,
+      "step": 3545
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3943743705749512,
+      "learning_rate": 5.9772323081805985e-06,
+      "loss": 0.6146,
+      "step": 3546
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.359615683555603,
+      "learning_rate": 5.975197214132207e-06,
+      "loss": 0.6164,
+      "step": 3547
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.324974775314331,
+      "learning_rate": 5.9731619521259156e-06,
+      "loss": 0.6021,
+      "step": 3548
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6578129529953003,
+      "learning_rate": 5.97112652251226e-06,
+      "loss": 0.5934,
+      "step": 3549
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3419653177261353,
+      "learning_rate": 5.969090925641799e-06,
+      "loss": 0.7274,
+      "step": 3550
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1800451278686523,
+      "learning_rate": 5.9670551618651225e-06,
+      "loss": 0.7183,
+      "step": 3551
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0364329814910889,
+      "learning_rate": 5.96501923153285e-06,
+      "loss": 0.5909,
+      "step": 3552
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.199339509010315,
+      "learning_rate": 5.9629831349956325e-06,
+      "loss": 0.5893,
+      "step": 3553
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2047299146652222,
+      "learning_rate": 5.960946872604141e-06,
+      "loss": 0.6378,
+      "step": 3554
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2108663320541382,
+      "learning_rate": 5.958910444709083e-06,
+      "loss": 0.5957,
+      "step": 3555
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2545571327209473,
+      "learning_rate": 5.956873851661192e-06,
+      "loss": 0.5564,
+      "step": 3556
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1796643733978271,
+      "learning_rate": 5.954837093811229e-06,
+      "loss": 0.6311,
+      "step": 3557
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4320603609085083,
+      "learning_rate": 5.952800171509985e-06,
+      "loss": 0.6326,
+      "step": 3558
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1937462091445923,
+      "learning_rate": 5.950763085108277e-06,
+      "loss": 0.6688,
+      "step": 3559
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2120587825775146,
+      "learning_rate": 5.948725834956952e-06,
+      "loss": 0.6689,
+      "step": 3560
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.330200433731079,
+      "learning_rate": 5.946688421406886e-06,
+      "loss": 0.5936,
+      "step": 3561
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3225212097167969,
+      "learning_rate": 5.9446508448089825e-06,
+      "loss": 0.6483,
+      "step": 3562
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0410070419311523,
+      "learning_rate": 5.942613105514171e-06,
+      "loss": 0.5963,
+      "step": 3563
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1120356321334839,
+      "learning_rate": 5.940575203873411e-06,
+      "loss": 0.6553,
+      "step": 3564
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1769014596939087,
+      "learning_rate": 5.9385371402376914e-06,
+      "loss": 0.6784,
+      "step": 3565
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.005772590637207,
+      "learning_rate": 5.936498914958025e-06,
+      "loss": 0.6325,
+      "step": 3566
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3367499113082886,
+      "learning_rate": 5.934460528385456e-06,
+      "loss": 0.6275,
+      "step": 3567
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1705892086029053,
+      "learning_rate": 5.932421980871054e-06,
+      "loss": 0.5585,
+      "step": 3568
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4280080795288086,
+      "learning_rate": 5.930383272765918e-06,
+      "loss": 0.5989,
+      "step": 3569
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7109063863754272,
+      "learning_rate": 5.928344404421174e-06,
+      "loss": 0.5889,
+      "step": 3570
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2693599462509155,
+      "learning_rate": 5.926305376187976e-06,
+      "loss": 0.6736,
+      "step": 3571
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3794386386871338,
+      "learning_rate": 5.924266188417503e-06,
+      "loss": 0.5768,
+      "step": 3572
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2329838275909424,
+      "learning_rate": 5.922226841460968e-06,
+      "loss": 0.6015,
+      "step": 3573
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6483304500579834,
+      "learning_rate": 5.920187335669602e-06,
+      "loss": 0.6534,
+      "step": 3574
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2386940717697144,
+      "learning_rate": 5.918147671394674e-06,
+      "loss": 0.5888,
+      "step": 3575
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.224981427192688,
+      "learning_rate": 5.91610784898747e-06,
+      "loss": 0.525,
+      "step": 3576
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.332261085510254,
+      "learning_rate": 5.91406786879931e-06,
+      "loss": 0.5301,
+      "step": 3577
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2706077098846436,
+      "learning_rate": 5.912027731181539e-06,
+      "loss": 0.5354,
+      "step": 3578
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3086779117584229,
+      "learning_rate": 5.909987436485527e-06,
+      "loss": 0.5638,
+      "step": 3579
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.0201406478881836,
+      "learning_rate": 5.907946985062678e-06,
+      "loss": 0.6415,
+      "step": 3580
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.47654390335083,
+      "learning_rate": 5.905906377264413e-06,
+      "loss": 0.6439,
+      "step": 3581
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4309604167938232,
+      "learning_rate": 5.90386561344219e-06,
+      "loss": 0.6235,
+      "step": 3582
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3017194271087646,
+      "learning_rate": 5.901824693947486e-06,
+      "loss": 0.6122,
+      "step": 3583
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1456342935562134,
+      "learning_rate": 5.899783619131809e-06,
+      "loss": 0.5598,
+      "step": 3584
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.045018196105957,
+      "learning_rate": 5.897742389346691e-06,
+      "loss": 0.6183,
+      "step": 3585
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2723512649536133,
+      "learning_rate": 5.8957010049436955e-06,
+      "loss": 0.5801,
+      "step": 3586
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.087829828262329,
+      "learning_rate": 5.893659466274407e-06,
+      "loss": 0.5959,
+      "step": 3587
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7654427289962769,
+      "learning_rate": 5.891617773690439e-06,
+      "loss": 0.6837,
+      "step": 3588
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.581198811531067,
+      "learning_rate": 5.8895759275434326e-06,
+      "loss": 0.5811,
+      "step": 3589
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.239762544631958,
+      "learning_rate": 5.887533928185054e-06,
+      "loss": 0.5911,
+      "step": 3590
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2970668077468872,
+      "learning_rate": 5.8854917759669975e-06,
+      "loss": 0.6369,
+      "step": 3591
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.130383014678955,
+      "learning_rate": 5.88344947124098e-06,
+      "loss": 0.576,
+      "step": 3592
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.180657148361206,
+      "learning_rate": 5.881407014358748e-06,
+      "loss": 0.5828,
+      "step": 3593
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4534460306167603,
+      "learning_rate": 5.879364405672072e-06,
+      "loss": 0.6106,
+      "step": 3594
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3066177368164062,
+      "learning_rate": 5.877321645532752e-06,
+      "loss": 0.6507,
+      "step": 3595
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6938061714172363,
+      "learning_rate": 5.87527873429261e-06,
+      "loss": 0.6446,
+      "step": 3596
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.4884603023529053,
+      "learning_rate": 5.873235672303497e-06,
+      "loss": 0.6266,
+      "step": 3597
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3594398498535156,
+      "learning_rate": 5.871192459917287e-06,
+      "loss": 0.5882,
+      "step": 3598
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6385754346847534,
+      "learning_rate": 5.869149097485887e-06,
+      "loss": 0.7136,
+      "step": 3599
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3254238367080688,
+      "learning_rate": 5.867105585361218e-06,
+      "loss": 0.5929,
+      "step": 3600
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1824506521224976,
+      "learning_rate": 5.865061923895238e-06,
+      "loss": 0.6646,
+      "step": 3601
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4841578006744385,
+      "learning_rate": 5.863018113439925e-06,
+      "loss": 0.6157,
+      "step": 3602
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1699411869049072,
+      "learning_rate": 5.860974154347282e-06,
+      "loss": 0.5475,
+      "step": 3603
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5281282663345337,
+      "learning_rate": 5.858930046969341e-06,
+      "loss": 0.6103,
+      "step": 3604
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1899720430374146,
+      "learning_rate": 5.856885791658158e-06,
+      "loss": 0.6138,
+      "step": 3605
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1821370124816895,
+      "learning_rate": 5.854841388765816e-06,
+      "loss": 0.583,
+      "step": 3606
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.455402135848999,
+      "learning_rate": 5.8527968386444174e-06,
+      "loss": 0.6418,
+      "step": 3607
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7864829301834106,
+      "learning_rate": 5.8507521416461e-06,
+      "loss": 0.5578,
+      "step": 3608
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.501908540725708,
+      "learning_rate": 5.848707298123017e-06,
+      "loss": 0.6361,
+      "step": 3609
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2245491743087769,
+      "learning_rate": 5.846662308427352e-06,
+      "loss": 0.6692,
+      "step": 3610
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.385204792022705,
+      "learning_rate": 5.844617172911313e-06,
+      "loss": 0.639,
+      "step": 3611
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3009589910507202,
+      "learning_rate": 5.8425718919271356e-06,
+      "loss": 0.6582,
+      "step": 3612
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1194837093353271,
+      "learning_rate": 5.8405264658270735e-06,
+      "loss": 0.5595,
+      "step": 3613
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1085587739944458,
+      "learning_rate": 5.838480894963412e-06,
+      "loss": 0.5775,
+      "step": 3614
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2732248306274414,
+      "learning_rate": 5.836435179688461e-06,
+      "loss": 0.5688,
+      "step": 3615
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1878724098205566,
+      "learning_rate": 5.83438932035455e-06,
+      "loss": 0.6376,
+      "step": 3616
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7219138145446777,
+      "learning_rate": 5.8323433173140395e-06,
+      "loss": 0.6421,
+      "step": 3617
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5152086019515991,
+      "learning_rate": 5.830297170919309e-06,
+      "loss": 0.6828,
+      "step": 3618
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.371113657951355,
+      "learning_rate": 5.828250881522769e-06,
+      "loss": 0.6614,
+      "step": 3619
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0953738689422607,
+      "learning_rate": 5.826204449476848e-06,
+      "loss": 0.6094,
+      "step": 3620
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.253603219985962,
+      "learning_rate": 5.824157875134005e-06,
+      "loss": 0.6148,
+      "step": 3621
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2168421745300293,
+      "learning_rate": 5.822111158846718e-06,
+      "loss": 0.6631,
+      "step": 3622
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5407350063323975,
+      "learning_rate": 5.820064300967494e-06,
+      "loss": 0.6399,
+      "step": 3623
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.245928168296814,
+      "learning_rate": 5.8180173018488625e-06,
+      "loss": 0.4972,
+      "step": 3624
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1214627027511597,
+      "learning_rate": 5.815970161843379e-06,
+      "loss": 0.5604,
+      "step": 3625
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.016115427017212,
+      "learning_rate": 5.81392288130362e-06,
+      "loss": 0.6543,
+      "step": 3626
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1197785139083862,
+      "learning_rate": 5.811875460582189e-06,
+      "loss": 0.6655,
+      "step": 3627
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4029167890548706,
+      "learning_rate": 5.80982790003171e-06,
+      "loss": 0.6326,
+      "step": 3628
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0858590602874756,
+      "learning_rate": 5.807780200004838e-06,
+      "loss": 0.6388,
+      "step": 3629
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1979694366455078,
+      "learning_rate": 5.805732360854243e-06,
+      "loss": 0.5826,
+      "step": 3630
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.004387617111206,
+      "learning_rate": 5.803684382932626e-06,
+      "loss": 0.6769,
+      "step": 3631
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0634325742721558,
+      "learning_rate": 5.801636266592712e-06,
+      "loss": 0.629,
+      "step": 3632
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3099150657653809,
+      "learning_rate": 5.799588012187243e-06,
+      "loss": 0.6693,
+      "step": 3633
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3465301990509033,
+      "learning_rate": 5.797539620068992e-06,
+      "loss": 0.6112,
+      "step": 3634
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.356348991394043,
+      "learning_rate": 5.79549109059075e-06,
+      "loss": 0.6714,
+      "step": 3635
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0583562850952148,
+      "learning_rate": 5.793442424105339e-06,
+      "loss": 0.5755,
+      "step": 3636
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.416865348815918,
+      "learning_rate": 5.791393620965597e-06,
+      "loss": 0.6358,
+      "step": 3637
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.24928617477417,
+      "learning_rate": 5.789344681524389e-06,
+      "loss": 0.6334,
+      "step": 3638
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.223817229270935,
+      "learning_rate": 5.7872956061346045e-06,
+      "loss": 0.6467,
+      "step": 3639
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9584287405014038,
+      "learning_rate": 5.785246395149152e-06,
+      "loss": 0.5455,
+      "step": 3640
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9955974221229553,
+      "learning_rate": 5.783197048920972e-06,
+      "loss": 0.6464,
+      "step": 3641
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.016220211982727,
+      "learning_rate": 5.781147567803017e-06,
+      "loss": 0.5736,
+      "step": 3642
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1931318044662476,
+      "learning_rate": 5.779097952148273e-06,
+      "loss": 0.6193,
+      "step": 3643
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3940238952636719,
+      "learning_rate": 5.77704820230974e-06,
+      "loss": 0.691,
+      "step": 3644
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2070850133895874,
+      "learning_rate": 5.77499831864045e-06,
+      "loss": 0.6053,
+      "step": 3645
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3859649896621704,
+      "learning_rate": 5.772948301493451e-06,
+      "loss": 0.6045,
+      "step": 3646
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1878433227539062,
+      "learning_rate": 5.770898151221819e-06,
+      "loss": 0.5722,
+      "step": 3647
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1642296314239502,
+      "learning_rate": 5.768847868178646e-06,
+      "loss": 0.6328,
+      "step": 3648
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 8.618324279785156,
+      "learning_rate": 5.766797452717059e-06,
+      "loss": 0.6084,
+      "step": 3649
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2117552757263184,
+      "learning_rate": 5.764746905190195e-06,
+      "loss": 0.5757,
+      "step": 3650
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7341634035110474,
+      "learning_rate": 5.762696225951222e-06,
+      "loss": 0.6151,
+      "step": 3651
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1232070922851562,
+      "learning_rate": 5.760645415353325e-06,
+      "loss": 0.5258,
+      "step": 3652
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6535828113555908,
+      "learning_rate": 5.7585944737497165e-06,
+      "loss": 0.6518,
+      "step": 3653
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4652224779129028,
+      "learning_rate": 5.756543401493628e-06,
+      "loss": 0.6398,
+      "step": 3654
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3363951444625854,
+      "learning_rate": 5.754492198938318e-06,
+      "loss": 0.6407,
+      "step": 3655
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.152831792831421,
+      "learning_rate": 5.752440866437062e-06,
+      "loss": 0.537,
+      "step": 3656
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4328360557556152,
+      "learning_rate": 5.750389404343159e-06,
+      "loss": 0.6318,
+      "step": 3657
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8562523126602173,
+      "learning_rate": 5.748337813009934e-06,
+      "loss": 0.5985,
+      "step": 3658
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9919341802597046,
+      "learning_rate": 5.74628609279073e-06,
+      "loss": 0.6107,
+      "step": 3659
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3245062828063965,
+      "learning_rate": 5.744234244038918e-06,
+      "loss": 0.526,
+      "step": 3660
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3959466218948364,
+      "learning_rate": 5.742182267107883e-06,
+      "loss": 0.6278,
+      "step": 3661
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1084336042404175,
+      "learning_rate": 5.740130162351039e-06,
+      "loss": 0.5771,
+      "step": 3662
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2701683044433594,
+      "learning_rate": 5.738077930121817e-06,
+      "loss": 0.5676,
+      "step": 3663
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0271109342575073,
+      "learning_rate": 5.736025570773675e-06,
+      "loss": 0.6557,
+      "step": 3664
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5346267223358154,
+      "learning_rate": 5.7339730846600875e-06,
+      "loss": 0.4982,
+      "step": 3665
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2214652299880981,
+      "learning_rate": 5.7319204721345536e-06,
+      "loss": 0.6193,
+      "step": 3666
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1910988092422485,
+      "learning_rate": 5.729867733550597e-06,
+      "loss": 0.5806,
+      "step": 3667
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1290035247802734,
+      "learning_rate": 5.727814869261758e-06,
+      "loss": 0.611,
+      "step": 3668
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1266953945159912,
+      "learning_rate": 5.725761879621601e-06,
+      "loss": 0.628,
+      "step": 3669
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3203330039978027,
+      "learning_rate": 5.723708764983712e-06,
+      "loss": 0.6415,
+      "step": 3670
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3828309774398804,
+      "learning_rate": 5.721655525701699e-06,
+      "loss": 0.5446,
+      "step": 3671
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2213470935821533,
+      "learning_rate": 5.719602162129189e-06,
+      "loss": 0.6274,
+      "step": 3672
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2695316076278687,
+      "learning_rate": 5.7175486746198325e-06,
+      "loss": 0.58,
+      "step": 3673
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0185368061065674,
+      "learning_rate": 5.715495063527301e-06,
+      "loss": 0.6232,
+      "step": 3674
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2112317085266113,
+      "learning_rate": 5.713441329205289e-06,
+      "loss": 0.6422,
+      "step": 3675
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6289145946502686,
+      "learning_rate": 5.711387472007509e-06,
+      "loss": 0.6632,
+      "step": 3676
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5209150314331055,
+      "learning_rate": 5.709333492287698e-06,
+      "loss": 0.6227,
+      "step": 3677
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4312928915023804,
+      "learning_rate": 5.7072793903996085e-06,
+      "loss": 0.66,
+      "step": 3678
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2010070085525513,
+      "learning_rate": 5.705225166697022e-06,
+      "loss": 0.6936,
+      "step": 3679
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0900729894638062,
+      "learning_rate": 5.703170821533733e-06,
+      "loss": 0.6763,
+      "step": 3680
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.209028720855713,
+      "learning_rate": 5.701116355263563e-06,
+      "loss": 0.5245,
+      "step": 3681
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0546902418136597,
+      "learning_rate": 5.69906176824035e-06,
+      "loss": 0.5515,
+      "step": 3682
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3346909284591675,
+      "learning_rate": 5.697007060817958e-06,
+      "loss": 0.57,
+      "step": 3683
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3135395050048828,
+      "learning_rate": 5.694952233350268e-06,
+      "loss": 0.5915,
+      "step": 3684
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2499960660934448,
+      "learning_rate": 5.692897286191179e-06,
+      "loss": 0.5783,
+      "step": 3685
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2707788944244385,
+      "learning_rate": 5.690842219694619e-06,
+      "loss": 0.6796,
+      "step": 3686
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1295020580291748,
+      "learning_rate": 5.688787034214525e-06,
+      "loss": 0.6514,
+      "step": 3687
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.445489764213562,
+      "learning_rate": 5.6867317301048676e-06,
+      "loss": 0.7057,
+      "step": 3688
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3477917909622192,
+      "learning_rate": 5.684676307719626e-06,
+      "loss": 0.5777,
+      "step": 3689
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2112998962402344,
+      "learning_rate": 5.682620767412808e-06,
+      "loss": 0.6103,
+      "step": 3690
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5050208568572998,
+      "learning_rate": 5.6805651095384375e-06,
+      "loss": 0.6307,
+      "step": 3691
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.221442461013794,
+      "learning_rate": 5.678509334450559e-06,
+      "loss": 0.5603,
+      "step": 3692
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4392536878585815,
+      "learning_rate": 5.67645344250324e-06,
+      "loss": 0.6022,
+      "step": 3693
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1971774101257324,
+      "learning_rate": 5.6743974340505645e-06,
+      "loss": 0.5572,
+      "step": 3694
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7652573585510254,
+      "learning_rate": 5.672341309446639e-06,
+      "loss": 0.6265,
+      "step": 3695
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4171807765960693,
+      "learning_rate": 5.670285069045588e-06,
+      "loss": 0.6642,
+      "step": 3696
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8250502347946167,
+      "learning_rate": 5.668228713201559e-06,
+      "loss": 0.6942,
+      "step": 3697
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0191811323165894,
+      "learning_rate": 5.666172242268715e-06,
+      "loss": 0.5493,
+      "step": 3698
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2496845722198486,
+      "learning_rate": 5.664115656601243e-06,
+      "loss": 0.6399,
+      "step": 3699
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2776678800582886,
+      "learning_rate": 5.662058956553348e-06,
+      "loss": 0.6093,
+      "step": 3700
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5822676420211792,
+      "learning_rate": 5.660002142479255e-06,
+      "loss": 0.5506,
+      "step": 3701
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.53562593460083,
+      "learning_rate": 5.657945214733208e-06,
+      "loss": 0.6304,
+      "step": 3702
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2964261770248413,
+      "learning_rate": 5.655888173669472e-06,
+      "loss": 0.6556,
+      "step": 3703
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3869433403015137,
+      "learning_rate": 5.653831019642327e-06,
+      "loss": 0.7004,
+      "step": 3704
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1649458408355713,
+      "learning_rate": 5.6517737530060815e-06,
+      "loss": 0.5943,
+      "step": 3705
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3618543148040771,
+      "learning_rate": 5.649716374115053e-06,
+      "loss": 0.6762,
+      "step": 3706
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0227161645889282,
+      "learning_rate": 5.6476588833235846e-06,
+      "loss": 0.5979,
+      "step": 3707
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4849395751953125,
+      "learning_rate": 5.645601280986038e-06,
+      "loss": 0.5881,
+      "step": 3708
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.639639139175415,
+      "learning_rate": 5.643543567456793e-06,
+      "loss": 0.5817,
+      "step": 3709
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9866191148757935,
+      "learning_rate": 5.64148574309025e-06,
+      "loss": 0.6601,
+      "step": 3710
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.609009027481079,
+      "learning_rate": 5.639427808240825e-06,
+      "loss": 0.5858,
+      "step": 3711
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3619134426116943,
+      "learning_rate": 5.637369763262959e-06,
+      "loss": 0.7077,
+      "step": 3712
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.349634051322937,
+      "learning_rate": 5.635311608511105e-06,
+      "loss": 0.6249,
+      "step": 3713
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1673775911331177,
+      "learning_rate": 5.63325334433974e-06,
+      "loss": 0.6222,
+      "step": 3714
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4045964479446411,
+      "learning_rate": 5.631194971103357e-06,
+      "loss": 0.5758,
+      "step": 3715
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5967957973480225,
+      "learning_rate": 5.6291364891564704e-06,
+      "loss": 0.6434,
+      "step": 3716
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5817066431045532,
+      "learning_rate": 5.62707789885361e-06,
+      "loss": 0.6438,
+      "step": 3717
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5195118188858032,
+      "learning_rate": 5.6250192005493285e-06,
+      "loss": 0.6614,
+      "step": 3718
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3318872451782227,
+      "learning_rate": 5.622960394598194e-06,
+      "loss": 0.598,
+      "step": 3719
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.229845643043518,
+      "learning_rate": 5.620901481354792e-06,
+      "loss": 0.5759,
+      "step": 3720
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2453598976135254,
+      "learning_rate": 5.6188424611737325e-06,
+      "loss": 0.671,
+      "step": 3721
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5366243124008179,
+      "learning_rate": 5.616783334409637e-06,
+      "loss": 0.7277,
+      "step": 3722
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2466989755630493,
+      "learning_rate": 5.614724101417148e-06,
+      "loss": 0.6157,
+      "step": 3723
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4630953073501587,
+      "learning_rate": 5.612664762550927e-06,
+      "loss": 0.6182,
+      "step": 3724
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7095898389816284,
+      "learning_rate": 5.6106053181656535e-06,
+      "loss": 0.6675,
+      "step": 3725
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4824260473251343,
+      "learning_rate": 5.608545768616025e-06,
+      "loss": 0.5571,
+      "step": 3726
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4626764059066772,
+      "learning_rate": 5.6064861142567575e-06,
+      "loss": 0.6716,
+      "step": 3727
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6166949272155762,
+      "learning_rate": 5.604426355442584e-06,
+      "loss": 0.6997,
+      "step": 3728
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.882332444190979,
+      "learning_rate": 5.602366492528256e-06,
+      "loss": 0.6062,
+      "step": 3729
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.456566333770752,
+      "learning_rate": 5.6003065258685444e-06,
+      "loss": 0.5865,
+      "step": 3730
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3899471759796143,
+      "learning_rate": 5.5982464558182335e-06,
+      "loss": 0.5982,
+      "step": 3731
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3733237981796265,
+      "learning_rate": 5.596186282732132e-06,
+      "loss": 0.6052,
+      "step": 3732
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0841834545135498,
+      "learning_rate": 5.59412600696506e-06,
+      "loss": 0.5684,
+      "step": 3733
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2034820318222046,
+      "learning_rate": 5.59206562887186e-06,
+      "loss": 0.5606,
+      "step": 3734
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2253649234771729,
+      "learning_rate": 5.5900051488073896e-06,
+      "loss": 0.5615,
+      "step": 3735
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.54776930809021,
+      "learning_rate": 5.587944567126525e-06,
+      "loss": 0.6376,
+      "step": 3736
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1381703615188599,
+      "learning_rate": 5.585883884184158e-06,
+      "loss": 0.594,
+      "step": 3737
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3217531442642212,
+      "learning_rate": 5.583823100335202e-06,
+      "loss": 0.7467,
+      "step": 3738
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.229705810546875,
+      "learning_rate": 5.581762215934582e-06,
+      "loss": 0.4965,
+      "step": 3739
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.090378761291504,
+      "learning_rate": 5.579701231337247e-06,
+      "loss": 0.6012,
+      "step": 3740
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2407397031784058,
+      "learning_rate": 5.577640146898156e-06,
+      "loss": 0.6052,
+      "step": 3741
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.4652295112609863,
+      "learning_rate": 5.57557896297229e-06,
+      "loss": 0.6058,
+      "step": 3742
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1912661790847778,
+      "learning_rate": 5.5735176799146486e-06,
+      "loss": 0.5969,
+      "step": 3743
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.246437668800354,
+      "learning_rate": 5.571456298080243e-06,
+      "loss": 0.6003,
+      "step": 3744
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1998566389083862,
+      "learning_rate": 5.569394817824106e-06,
+      "loss": 0.612,
+      "step": 3745
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8297500610351562,
+      "learning_rate": 5.567333239501284e-06,
+      "loss": 0.5909,
+      "step": 3746
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1892491579055786,
+      "learning_rate": 5.565271563466845e-06,
+      "loss": 0.5661,
+      "step": 3747
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.076907753944397,
+      "learning_rate": 5.5632097900758676e-06,
+      "loss": 0.5741,
+      "step": 3748
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0066630840301514,
+      "learning_rate": 5.561147919683451e-06,
+      "loss": 0.5962,
+      "step": 3749
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.04477858543396,
+      "learning_rate": 5.559085952644711e-06,
+      "loss": 0.5698,
+      "step": 3750
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.407403588294983,
+      "learning_rate": 5.5570238893147795e-06,
+      "loss": 0.6276,
+      "step": 3751
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.328902244567871,
+      "learning_rate": 5.554961730048806e-06,
+      "loss": 0.579,
+      "step": 3752
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0437157154083252,
+      "learning_rate": 5.5528994752019535e-06,
+      "loss": 0.6284,
+      "step": 3753
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2737390995025635,
+      "learning_rate": 5.550837125129406e-06,
+      "loss": 0.6508,
+      "step": 3754
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1763293743133545,
+      "learning_rate": 5.548774680186359e-06,
+      "loss": 0.5689,
+      "step": 3755
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2019954919815063,
+      "learning_rate": 5.5467121407280275e-06,
+      "loss": 0.7561,
+      "step": 3756
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4074658155441284,
+      "learning_rate": 5.544649507109642e-06,
+      "loss": 0.6113,
+      "step": 3757
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.317198395729065,
+      "learning_rate": 5.5425867796864496e-06,
+      "loss": 0.6282,
+      "step": 3758
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7032959461212158,
+      "learning_rate": 5.54052395881371e-06,
+      "loss": 0.6331,
+      "step": 3759
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3925243616104126,
+      "learning_rate": 5.5384610448467095e-06,
+      "loss": 0.563,
+      "step": 3760
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.3951807022094727,
+      "learning_rate": 5.536398038140735e-06,
+      "loss": 0.596,
+      "step": 3761
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2922788858413696,
+      "learning_rate": 5.534334939051104e-06,
+      "loss": 0.5835,
+      "step": 3762
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4560922384262085,
+      "learning_rate": 5.53227174793314e-06,
+      "loss": 0.5856,
+      "step": 3763
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7121245861053467,
+      "learning_rate": 5.530208465142186e-06,
+      "loss": 0.5596,
+      "step": 3764
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4115746021270752,
+      "learning_rate": 5.5281450910336e-06,
+      "loss": 0.6768,
+      "step": 3765
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.208176612854004,
+      "learning_rate": 5.526081625962758e-06,
+      "loss": 0.5841,
+      "step": 3766
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.182554841041565,
+      "learning_rate": 5.524018070285047e-06,
+      "loss": 0.6007,
+      "step": 3767
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3877570629119873,
+      "learning_rate": 5.521954424355876e-06,
+      "loss": 0.6617,
+      "step": 3768
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7248724699020386,
+      "learning_rate": 5.519890688530666e-06,
+      "loss": 0.6391,
+      "step": 3769
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0161206722259521,
+      "learning_rate": 5.517826863164849e-06,
+      "loss": 0.4795,
+      "step": 3770
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1962370872497559,
+      "learning_rate": 5.515762948613882e-06,
+      "loss": 0.6225,
+      "step": 3771
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3718147277832031,
+      "learning_rate": 5.51369894523323e-06,
+      "loss": 0.6185,
+      "step": 3772
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4110352993011475,
+      "learning_rate": 5.5116348533783755e-06,
+      "loss": 0.6968,
+      "step": 3773
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2778741121292114,
+      "learning_rate": 5.509570673404819e-06,
+      "loss": 0.5913,
+      "step": 3774
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9584821462631226,
+      "learning_rate": 5.50750640566807e-06,
+      "loss": 0.5655,
+      "step": 3775
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1385620832443237,
+      "learning_rate": 5.505442050523655e-06,
+      "loss": 0.703,
+      "step": 3776
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.690279483795166,
+      "learning_rate": 5.503377608327124e-06,
+      "loss": 0.5459,
+      "step": 3777
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3032708168029785,
+      "learning_rate": 5.50131307943403e-06,
+      "loss": 0.6358,
+      "step": 3778
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.129211187362671,
+      "learning_rate": 5.499248464199949e-06,
+      "loss": 0.5988,
+      "step": 3779
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3283401727676392,
+      "learning_rate": 5.497183762980467e-06,
+      "loss": 0.6212,
+      "step": 3780
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4811829328536987,
+      "learning_rate": 5.495118976131187e-06,
+      "loss": 0.5867,
+      "step": 3781
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.861901044845581,
+      "learning_rate": 5.493054104007728e-06,
+      "loss": 0.6727,
+      "step": 3782
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4480305910110474,
+      "learning_rate": 5.49098914696572e-06,
+      "loss": 0.6174,
+      "step": 3783
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.423263669013977,
+      "learning_rate": 5.488924105360812e-06,
+      "loss": 0.6232,
+      "step": 3784
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5136548280715942,
+      "learning_rate": 5.486858979548663e-06,
+      "loss": 0.5616,
+      "step": 3785
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1748825311660767,
+      "learning_rate": 5.484793769884953e-06,
+      "loss": 0.5792,
+      "step": 3786
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2067316770553589,
+      "learning_rate": 5.482728476725369e-06,
+      "loss": 0.5731,
+      "step": 3787
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8080260753631592,
+      "learning_rate": 5.480663100425616e-06,
+      "loss": 0.6952,
+      "step": 3788
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5747623443603516,
+      "learning_rate": 5.478597641341414e-06,
+      "loss": 0.5669,
+      "step": 3789
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.224440574645996,
+      "learning_rate": 5.476532099828498e-06,
+      "loss": 0.5727,
+      "step": 3790
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2634854316711426,
+      "learning_rate": 5.474466476242611e-06,
+      "loss": 0.6208,
+      "step": 3791
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3283199071884155,
+      "learning_rate": 5.472400770939519e-06,
+      "loss": 0.6815,
+      "step": 3792
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3618751764297485,
+      "learning_rate": 5.470334984274995e-06,
+      "loss": 0.58,
+      "step": 3793
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.400259256362915,
+      "learning_rate": 5.468269116604831e-06,
+      "loss": 0.6144,
+      "step": 3794
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.376218557357788,
+      "learning_rate": 5.466203168284829e-06,
+      "loss": 0.6464,
+      "step": 3795
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2561371326446533,
+      "learning_rate": 5.464137139670806e-06,
+      "loss": 0.5849,
+      "step": 3796
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5874736309051514,
+      "learning_rate": 5.4620710311185976e-06,
+      "loss": 0.6173,
+      "step": 3797
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9282337427139282,
+      "learning_rate": 5.460004842984044e-06,
+      "loss": 0.5781,
+      "step": 3798
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2490649223327637,
+      "learning_rate": 5.457938575623008e-06,
+      "loss": 0.6326,
+      "step": 3799
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4292243719100952,
+      "learning_rate": 5.455872229391357e-06,
+      "loss": 0.5963,
+      "step": 3800
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1780860424041748,
+      "learning_rate": 5.453805804644984e-06,
+      "loss": 0.5695,
+      "step": 3801
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3638889789581299,
+      "learning_rate": 5.451739301739782e-06,
+      "loss": 0.5805,
+      "step": 3802
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8172688484191895,
+      "learning_rate": 5.449672721031668e-06,
+      "loss": 0.5556,
+      "step": 3803
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3854845762252808,
+      "learning_rate": 5.447606062876569e-06,
+      "loss": 0.6247,
+      "step": 3804
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3911125659942627,
+      "learning_rate": 5.445539327630423e-06,
+      "loss": 0.5315,
+      "step": 3805
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.301125407218933,
+      "learning_rate": 5.443472515649184e-06,
+      "loss": 0.6608,
+      "step": 3806
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.013878583908081,
+      "learning_rate": 5.441405627288817e-06,
+      "loss": 0.5742,
+      "step": 3807
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.087625026702881,
+      "learning_rate": 5.439338662905305e-06,
+      "loss": 0.6331,
+      "step": 3808
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1464579105377197,
+      "learning_rate": 5.437271622854635e-06,
+      "loss": 0.6458,
+      "step": 3809
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.453547477722168,
+      "learning_rate": 5.435204507492819e-06,
+      "loss": 0.6398,
+      "step": 3810
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6091513633728027,
+      "learning_rate": 5.433137317175868e-06,
+      "loss": 0.5774,
+      "step": 3811
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2415236234664917,
+      "learning_rate": 5.431070052259821e-06,
+      "loss": 0.7783,
+      "step": 3812
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.310966968536377,
+      "learning_rate": 5.42900271310072e-06,
+      "loss": 0.7047,
+      "step": 3813
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3866527080535889,
+      "learning_rate": 5.426935300054621e-06,
+      "loss": 0.6902,
+      "step": 3814
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3021939992904663,
+      "learning_rate": 5.424867813477593e-06,
+      "loss": 0.5693,
+      "step": 3815
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2108157873153687,
+      "learning_rate": 5.422800253725722e-06,
+      "loss": 0.5726,
+      "step": 3816
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.936347246170044,
+      "learning_rate": 5.4207326211550995e-06,
+      "loss": 0.6049,
+      "step": 3817
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1145234107971191,
+      "learning_rate": 5.418664916121835e-06,
+      "loss": 0.5592,
+      "step": 3818
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.271641254425049,
+      "learning_rate": 5.416597138982048e-06,
+      "loss": 0.5187,
+      "step": 3819
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0878288745880127,
+      "learning_rate": 5.414529290091872e-06,
+      "loss": 0.6153,
+      "step": 3820
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5007294416427612,
+      "learning_rate": 5.412461369807451e-06,
+      "loss": 0.621,
+      "step": 3821
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5491704940795898,
+      "learning_rate": 5.4103933784849435e-06,
+      "loss": 0.723,
+      "step": 3822
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1629546880722046,
+      "learning_rate": 5.408325316480518e-06,
+      "loss": 0.5476,
+      "step": 3823
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2649697065353394,
+      "learning_rate": 5.406257184150355e-06,
+      "loss": 0.6794,
+      "step": 3824
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6518555879592896,
+      "learning_rate": 5.404188981850651e-06,
+      "loss": 0.6568,
+      "step": 3825
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.191433310508728,
+      "learning_rate": 5.4021207099376095e-06,
+      "loss": 0.6148,
+      "step": 3826
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2688812017440796,
+      "learning_rate": 5.40005236876745e-06,
+      "loss": 0.6096,
+      "step": 3827
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7701101303100586,
+      "learning_rate": 5.3979839586964e-06,
+      "loss": 0.5587,
+      "step": 3828
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3165203332901,
+      "learning_rate": 5.3959154800807025e-06,
+      "loss": 0.5454,
+      "step": 3829
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3157920837402344,
+      "learning_rate": 5.393846933276612e-06,
+      "loss": 0.6478,
+      "step": 3830
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2996734380722046,
+      "learning_rate": 5.391778318640392e-06,
+      "loss": 0.5351,
+      "step": 3831
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2892003059387207,
+      "learning_rate": 5.38970963652832e-06,
+      "loss": 0.5722,
+      "step": 3832
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.160461187362671,
+      "learning_rate": 5.387640887296683e-06,
+      "loss": 0.6198,
+      "step": 3833
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.233054757118225,
+      "learning_rate": 5.385572071301785e-06,
+      "loss": 0.5894,
+      "step": 3834
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3636419773101807,
+      "learning_rate": 5.383503188899932e-06,
+      "loss": 0.6296,
+      "step": 3835
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6694684028625488,
+      "learning_rate": 5.381434240447451e-06,
+      "loss": 0.6466,
+      "step": 3836
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.753450632095337,
+      "learning_rate": 5.379365226300673e-06,
+      "loss": 0.6464,
+      "step": 3837
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.236421823501587,
+      "learning_rate": 5.377296146815949e-06,
+      "loss": 0.5271,
+      "step": 3838
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3160569667816162,
+      "learning_rate": 5.37522700234963e-06,
+      "loss": 0.7266,
+      "step": 3839
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5594069957733154,
+      "learning_rate": 5.373157793258088e-06,
+      "loss": 0.6761,
+      "step": 3840
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5676730871200562,
+      "learning_rate": 5.3710885198977004e-06,
+      "loss": 0.5312,
+      "step": 3841
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.762484073638916,
+      "learning_rate": 5.369019182624858e-06,
+      "loss": 0.6218,
+      "step": 3842
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2499971389770508,
+      "learning_rate": 5.366949781795961e-06,
+      "loss": 0.6244,
+      "step": 3843
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2467316389083862,
+      "learning_rate": 5.364880317767424e-06,
+      "loss": 0.6309,
+      "step": 3844
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1873960494995117,
+      "learning_rate": 5.362810790895668e-06,
+      "loss": 0.5959,
+      "step": 3845
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2639389038085938,
+      "learning_rate": 5.360741201537127e-06,
+      "loss": 0.6166,
+      "step": 3846
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4507334232330322,
+      "learning_rate": 5.358671550048249e-06,
+      "loss": 0.5928,
+      "step": 3847
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1256029605865479,
+      "learning_rate": 5.356601836785484e-06,
+      "loss": 0.5958,
+      "step": 3848
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2243516445159912,
+      "learning_rate": 5.354532062105303e-06,
+      "loss": 0.5836,
+      "step": 3849
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5672364234924316,
+      "learning_rate": 5.352462226364179e-06,
+      "loss": 0.6359,
+      "step": 3850
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.424888014793396,
+      "learning_rate": 5.350392329918601e-06,
+      "loss": 0.6901,
+      "step": 3851
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.7046003341674805,
+      "learning_rate": 5.348322373125065e-06,
+      "loss": 0.6442,
+      "step": 3852
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4008851051330566,
+      "learning_rate": 5.346252356340082e-06,
+      "loss": 0.5696,
+      "step": 3853
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6207784414291382,
+      "learning_rate": 5.344182279920167e-06,
+      "loss": 0.5765,
+      "step": 3854
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.579346776008606,
+      "learning_rate": 5.342112144221851e-06,
+      "loss": 0.6231,
+      "step": 3855
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2518047094345093,
+      "learning_rate": 5.340041949601672e-06,
+      "loss": 0.5404,
+      "step": 3856
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0116968154907227,
+      "learning_rate": 5.337971696416178e-06,
+      "loss": 0.7541,
+      "step": 3857
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3027863502502441,
+      "learning_rate": 5.33590138502193e-06,
+      "loss": 0.626,
+      "step": 3858
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.156329870223999,
+      "learning_rate": 5.3338310157754934e-06,
+      "loss": 0.6029,
+      "step": 3859
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7596616744995117,
+      "learning_rate": 5.331760589033452e-06,
+      "loss": 0.6789,
+      "step": 3860
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0322740077972412,
+      "learning_rate": 5.329690105152392e-06,
+      "loss": 0.637,
+      "step": 3861
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2053539752960205,
+      "learning_rate": 5.3276195644889115e-06,
+      "loss": 0.5274,
+      "step": 3862
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.20881986618042,
+      "learning_rate": 5.325548967399621e-06,
+      "loss": 0.636,
+      "step": 3863
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.379366159439087,
+      "learning_rate": 5.323478314241138e-06,
+      "loss": 0.598,
+      "step": 3864
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3362188339233398,
+      "learning_rate": 5.321407605370091e-06,
+      "loss": 0.7068,
+      "step": 3865
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4871771335601807,
+      "learning_rate": 5.319336841143116e-06,
+      "loss": 0.6618,
+      "step": 3866
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.288758397102356,
+      "learning_rate": 5.31726602191686e-06,
+      "loss": 0.7501,
+      "step": 3867
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3813121318817139,
+      "learning_rate": 5.315195148047981e-06,
+      "loss": 0.6301,
+      "step": 3868
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2369877099990845,
+      "learning_rate": 5.313124219893145e-06,
+      "loss": 0.5641,
+      "step": 3869
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0300804376602173,
+      "learning_rate": 5.311053237809026e-06,
+      "loss": 0.6737,
+      "step": 3870
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3140389919281006,
+      "learning_rate": 5.30898220215231e-06,
+      "loss": 0.5421,
+      "step": 3871
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4346362352371216,
+      "learning_rate": 5.306911113279689e-06,
+      "loss": 0.6743,
+      "step": 3872
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0767395496368408,
+      "learning_rate": 5.3048399715478675e-06,
+      "loss": 0.6462,
+      "step": 3873
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3019919395446777,
+      "learning_rate": 5.302768777313556e-06,
+      "loss": 0.7012,
+      "step": 3874
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2964775562286377,
+      "learning_rate": 5.300697530933479e-06,
+      "loss": 0.4923,
+      "step": 3875
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0694615840911865,
+      "learning_rate": 5.2986262327643615e-06,
+      "loss": 0.5454,
+      "step": 3876
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3932045698165894,
+      "learning_rate": 5.296554883162947e-06,
+      "loss": 0.5652,
+      "step": 3877
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.147936463356018,
+      "learning_rate": 5.29448348248598e-06,
+      "loss": 0.6513,
+      "step": 3878
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2141468524932861,
+      "learning_rate": 5.29241203109022e-06,
+      "loss": 0.5956,
+      "step": 3879
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1196589469909668,
+      "learning_rate": 5.2903405293324316e-06,
+      "loss": 0.5876,
+      "step": 3880
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.196601152420044,
+      "learning_rate": 5.288268977569386e-06,
+      "loss": 0.6035,
+      "step": 3881
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4775373935699463,
+      "learning_rate": 5.286197376157871e-06,
+      "loss": 0.6348,
+      "step": 3882
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.283056378364563,
+      "learning_rate": 5.284125725454674e-06,
+      "loss": 0.64,
+      "step": 3883
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.263865351676941,
+      "learning_rate": 5.282054025816597e-06,
+      "loss": 0.6002,
+      "step": 3884
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2261013984680176,
+      "learning_rate": 5.279982277600445e-06,
+      "loss": 0.5812,
+      "step": 3885
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.103773355484009,
+      "learning_rate": 5.277910481163038e-06,
+      "loss": 0.6105,
+      "step": 3886
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.412146806716919,
+      "learning_rate": 5.275838636861198e-06,
+      "loss": 0.6886,
+      "step": 3887
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.211158037185669,
+      "learning_rate": 5.273766745051761e-06,
+      "loss": 0.5951,
+      "step": 3888
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4581794738769531,
+      "learning_rate": 5.271694806091564e-06,
+      "loss": 0.64,
+      "step": 3889
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1060121059417725,
+      "learning_rate": 5.269622820337462e-06,
+      "loss": 0.5375,
+      "step": 3890
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.2724769115448,
+      "learning_rate": 5.267550788146308e-06,
+      "loss": 0.5919,
+      "step": 3891
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2122722864151,
+      "learning_rate": 5.26547870987497e-06,
+      "loss": 0.6125,
+      "step": 3892
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1100267171859741,
+      "learning_rate": 5.26340658588032e-06,
+      "loss": 0.5679,
+      "step": 3893
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2956206798553467,
+      "learning_rate": 5.261334416519239e-06,
+      "loss": 0.6334,
+      "step": 3894
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3694170713424683,
+      "learning_rate": 5.259262202148616e-06,
+      "loss": 0.6584,
+      "step": 3895
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3362345695495605,
+      "learning_rate": 5.257189943125349e-06,
+      "loss": 0.5786,
+      "step": 3896
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2397748231887817,
+      "learning_rate": 5.255117639806343e-06,
+      "loss": 0.5881,
+      "step": 3897
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4001562595367432,
+      "learning_rate": 5.253045292548508e-06,
+      "loss": 0.646,
+      "step": 3898
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3735246658325195,
+      "learning_rate": 5.250972901708765e-06,
+      "loss": 0.8063,
+      "step": 3899
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1073076725006104,
+      "learning_rate": 5.248900467644041e-06,
+      "loss": 0.6368,
+      "step": 3900
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3196773529052734,
+      "learning_rate": 5.246827990711269e-06,
+      "loss": 0.5578,
+      "step": 3901
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3190702199935913,
+      "learning_rate": 5.244755471267394e-06,
+      "loss": 0.5742,
+      "step": 3902
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3288342952728271,
+      "learning_rate": 5.242682909669364e-06,
+      "loss": 0.5448,
+      "step": 3903
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3225739002227783,
+      "learning_rate": 5.240610306274134e-06,
+      "loss": 0.6489,
+      "step": 3904
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.309205174446106,
+      "learning_rate": 5.238537661438671e-06,
+      "loss": 0.6241,
+      "step": 3905
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2318326234817505,
+      "learning_rate": 5.236464975519944e-06,
+      "loss": 0.5419,
+      "step": 3906
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4670156240463257,
+      "learning_rate": 5.234392248874931e-06,
+      "loss": 0.6127,
+      "step": 3907
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.364734411239624,
+      "learning_rate": 5.232319481860618e-06,
+      "loss": 0.6175,
+      "step": 3908
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1357091665267944,
+      "learning_rate": 5.2302466748339954e-06,
+      "loss": 0.6176,
+      "step": 3909
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4520831108093262,
+      "learning_rate": 5.228173828152063e-06,
+      "loss": 0.5557,
+      "step": 3910
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7567130327224731,
+      "learning_rate": 5.226100942171827e-06,
+      "loss": 0.5868,
+      "step": 3911
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0778729915618896,
+      "learning_rate": 5.2240280172503e-06,
+      "loss": 0.6589,
+      "step": 3912
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.256453037261963,
+      "learning_rate": 5.221955053744498e-06,
+      "loss": 0.6077,
+      "step": 3913
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.119685649871826,
+      "learning_rate": 5.219882052011451e-06,
+      "loss": 0.6429,
+      "step": 3914
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4709926843643188,
+      "learning_rate": 5.217809012408191e-06,
+      "loss": 0.621,
+      "step": 3915
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3248754739761353,
+      "learning_rate": 5.215735935291755e-06,
+      "loss": 0.656,
+      "step": 3916
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7260463237762451,
+      "learning_rate": 5.213662821019187e-06,
+      "loss": 0.6932,
+      "step": 3917
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4484518766403198,
+      "learning_rate": 5.211589669947544e-06,
+      "loss": 0.5989,
+      "step": 3918
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3106611967086792,
+      "learning_rate": 5.209516482433879e-06,
+      "loss": 0.682,
+      "step": 3919
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3350433111190796,
+      "learning_rate": 5.207443258835261e-06,
+      "loss": 0.6008,
+      "step": 3920
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8616951704025269,
+      "learning_rate": 5.205369999508756e-06,
+      "loss": 0.5976,
+      "step": 3921
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2377861738204956,
+      "learning_rate": 5.203296704811443e-06,
+      "loss": 0.5425,
+      "step": 3922
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.428403615951538,
+      "learning_rate": 5.201223375100406e-06,
+      "loss": 0.6628,
+      "step": 3923
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4202046394348145,
+      "learning_rate": 5.199150010732731e-06,
+      "loss": 0.567,
+      "step": 3924
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4062016010284424,
+      "learning_rate": 5.1970766120655155e-06,
+      "loss": 0.6508,
+      "step": 3925
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2854269742965698,
+      "learning_rate": 5.195003179455859e-06,
+      "loss": 0.6857,
+      "step": 3926
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.979161024093628,
+      "learning_rate": 5.192929713260869e-06,
+      "loss": 0.6199,
+      "step": 3927
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3159161806106567,
+      "learning_rate": 5.190856213837656e-06,
+      "loss": 0.6078,
+      "step": 3928
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0987286567687988,
+      "learning_rate": 5.1887826815433415e-06,
+      "loss": 0.5123,
+      "step": 3929
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5091686248779297,
+      "learning_rate": 5.186709116735046e-06,
+      "loss": 0.596,
+      "step": 3930
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3786001205444336,
+      "learning_rate": 5.1846355197699e-06,
+      "loss": 0.6173,
+      "step": 3931
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4815407991409302,
+      "learning_rate": 5.182561891005039e-06,
+      "loss": 0.5635,
+      "step": 3932
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4058189392089844,
+      "learning_rate": 5.180488230797602e-06,
+      "loss": 0.5896,
+      "step": 3933
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2490657567977905,
+      "learning_rate": 5.178414539504737e-06,
+      "loss": 0.596,
+      "step": 3934
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.224664330482483,
+      "learning_rate": 5.176340817483592e-06,
+      "loss": 0.6107,
+      "step": 3935
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3340404033660889,
+      "learning_rate": 5.174267065091329e-06,
+      "loss": 0.5858,
+      "step": 3936
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.558334469795227,
+      "learning_rate": 5.172193282685104e-06,
+      "loss": 0.5914,
+      "step": 3937
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1823575496673584,
+      "learning_rate": 5.170119470622086e-06,
+      "loss": 0.5338,
+      "step": 3938
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.632826328277588,
+      "learning_rate": 5.168045629259446e-06,
+      "loss": 0.7068,
+      "step": 3939
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.282450556755066,
+      "learning_rate": 5.1659717589543635e-06,
+      "loss": 0.6205,
+      "step": 3940
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.351825714111328,
+      "learning_rate": 5.163897860064019e-06,
+      "loss": 0.6188,
+      "step": 3941
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1424978971481323,
+      "learning_rate": 5.1618239329456e-06,
+      "loss": 0.5913,
+      "step": 3942
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2732305526733398,
+      "learning_rate": 5.159749977956295e-06,
+      "loss": 0.572,
+      "step": 3943
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.20172917842865,
+      "learning_rate": 5.157675995453306e-06,
+      "loss": 0.7332,
+      "step": 3944
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1711151599884033,
+      "learning_rate": 5.155601985793831e-06,
+      "loss": 0.5454,
+      "step": 3945
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4528971910476685,
+      "learning_rate": 5.153527949335077e-06,
+      "loss": 0.5712,
+      "step": 3946
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6643458604812622,
+      "learning_rate": 5.151453886434255e-06,
+      "loss": 0.6729,
+      "step": 3947
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3720951080322266,
+      "learning_rate": 5.149379797448577e-06,
+      "loss": 0.6249,
+      "step": 3948
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.305267572402954,
+      "learning_rate": 5.147305682735266e-06,
+      "loss": 0.5595,
+      "step": 3949
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.398755669593811,
+      "learning_rate": 5.145231542651547e-06,
+      "loss": 0.5929,
+      "step": 3950
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1850718259811401,
+      "learning_rate": 5.143157377554645e-06,
+      "loss": 0.6145,
+      "step": 3951
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3608309030532837,
+      "learning_rate": 5.1410831878017945e-06,
+      "loss": 0.6691,
+      "step": 3952
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1051334142684937,
+      "learning_rate": 5.139008973750234e-06,
+      "loss": 0.6059,
+      "step": 3953
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0188237428665161,
+      "learning_rate": 5.136934735757202e-06,
+      "loss": 0.6137,
+      "step": 3954
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2129838466644287,
+      "learning_rate": 5.1348604741799455e-06,
+      "loss": 0.5999,
+      "step": 3955
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.062683343887329,
+      "learning_rate": 5.1327861893757125e-06,
+      "loss": 0.5963,
+      "step": 3956
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2500804662704468,
+      "learning_rate": 5.1307118817017575e-06,
+      "loss": 0.6215,
+      "step": 3957
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3206661939620972,
+      "learning_rate": 5.12863755151534e-06,
+      "loss": 0.7026,
+      "step": 3958
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.201270580291748,
+      "learning_rate": 5.1265631991737165e-06,
+      "loss": 0.6585,
+      "step": 3959
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1174263954162598,
+      "learning_rate": 5.124488825034155e-06,
+      "loss": 0.5546,
+      "step": 3960
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.306179165840149,
+      "learning_rate": 5.122414429453923e-06,
+      "loss": 0.6222,
+      "step": 3961
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.289539098739624,
+      "learning_rate": 5.120340012790296e-06,
+      "loss": 0.6138,
+      "step": 3962
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3443716764450073,
+      "learning_rate": 5.118265575400546e-06,
+      "loss": 0.5832,
+      "step": 3963
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2395848035812378,
+      "learning_rate": 5.116191117641955e-06,
+      "loss": 0.6501,
+      "step": 3964
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1485309600830078,
+      "learning_rate": 5.114116639871804e-06,
+      "loss": 0.579,
+      "step": 3965
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2162903547286987,
+      "learning_rate": 5.112042142447384e-06,
+      "loss": 0.5577,
+      "step": 3966
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1788960695266724,
+      "learning_rate": 5.10996762572598e-06,
+      "loss": 0.5424,
+      "step": 3967
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1972217559814453,
+      "learning_rate": 5.10789309006489e-06,
+      "loss": 0.5713,
+      "step": 3968
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1885600090026855,
+      "learning_rate": 5.105818535821406e-06,
+      "loss": 0.5696,
+      "step": 3969
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4530361890792847,
+      "learning_rate": 5.103743963352832e-06,
+      "loss": 0.6195,
+      "step": 3970
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3970437049865723,
+      "learning_rate": 5.101669373016469e-06,
+      "loss": 0.7098,
+      "step": 3971
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6319563388824463,
+      "learning_rate": 5.099594765169621e-06,
+      "loss": 0.5933,
+      "step": 3972
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.417358636856079,
+      "learning_rate": 5.097520140169599e-06,
+      "loss": 0.6566,
+      "step": 3973
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.314659833908081,
+      "learning_rate": 5.095445498373717e-06,
+      "loss": 0.6062,
+      "step": 3974
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4447764158248901,
+      "learning_rate": 5.0933708401392864e-06,
+      "loss": 0.5829,
+      "step": 3975
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1412814855575562,
+      "learning_rate": 5.091296165823627e-06,
+      "loss": 0.5649,
+      "step": 3976
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.5301032066345215,
+      "learning_rate": 5.08922147578406e-06,
+      "loss": 0.5886,
+      "step": 3977
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2669841051101685,
+      "learning_rate": 5.0871467703779054e-06,
+      "loss": 0.5541,
+      "step": 3978
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5283225774765015,
+      "learning_rate": 5.0850720499624915e-06,
+      "loss": 0.5673,
+      "step": 3979
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3331176042556763,
+      "learning_rate": 5.082997314895146e-06,
+      "loss": 0.6687,
+      "step": 3980
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5164918899536133,
+      "learning_rate": 5.080922565533201e-06,
+      "loss": 0.6221,
+      "step": 3981
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4119714498519897,
+      "learning_rate": 5.078847802233987e-06,
+      "loss": 0.5823,
+      "step": 3982
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.263611912727356,
+      "learning_rate": 5.076773025354843e-06,
+      "loss": 0.6335,
+      "step": 3983
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6149969100952148,
+      "learning_rate": 5.074698235253106e-06,
+      "loss": 0.6478,
+      "step": 3984
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2421183586120605,
+      "learning_rate": 5.072623432286116e-06,
+      "loss": 0.7456,
+      "step": 3985
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2369145154953003,
+      "learning_rate": 5.070548616811216e-06,
+      "loss": 0.5444,
+      "step": 3986
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4720882177352905,
+      "learning_rate": 5.0684737891857505e-06,
+      "loss": 0.5883,
+      "step": 3987
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7480475902557373,
+      "learning_rate": 5.066398949767068e-06,
+      "loss": 0.5908,
+      "step": 3988
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.277265191078186,
+      "learning_rate": 5.064324098912513e-06,
+      "loss": 0.5804,
+      "step": 3989
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.826635479927063,
+      "learning_rate": 5.062249236979442e-06,
+      "loss": 0.6328,
+      "step": 3990
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3039029836654663,
+      "learning_rate": 5.060174364325202e-06,
+      "loss": 0.6261,
+      "step": 3991
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1858707666397095,
+      "learning_rate": 5.058099481307154e-06,
+      "loss": 0.5912,
+      "step": 3992
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1001001596450806,
+      "learning_rate": 5.05602458828265e-06,
+      "loss": 0.5876,
+      "step": 3993
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4584338665008545,
+      "learning_rate": 5.053949685609051e-06,
+      "loss": 0.5939,
+      "step": 3994
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3201522827148438,
+      "learning_rate": 5.051874773643713e-06,
+      "loss": 0.654,
+      "step": 3995
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.044450044631958,
+      "learning_rate": 5.049799852744001e-06,
+      "loss": 0.58,
+      "step": 3996
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.269258975982666,
+      "learning_rate": 5.047724923267277e-06,
+      "loss": 0.5994,
+      "step": 3997
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2389514446258545,
+      "learning_rate": 5.045649985570904e-06,
+      "loss": 0.5532,
+      "step": 3998
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3213895559310913,
+      "learning_rate": 5.0435750400122485e-06,
+      "loss": 0.6093,
+      "step": 3999
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4534727334976196,
+      "learning_rate": 5.041500086948677e-06,
+      "loss": 0.6433,
+      "step": 4000
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4285792112350464,
+      "learning_rate": 5.039425126737563e-06,
+      "loss": 0.6478,
+      "step": 4001
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.21049165725708,
+      "learning_rate": 5.0373501597362685e-06,
+      "loss": 0.6199,
+      "step": 4002
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1804089546203613,
+      "learning_rate": 5.03527518630217e-06,
+      "loss": 0.5829,
+      "step": 4003
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5666120052337646,
+      "learning_rate": 5.033200206792637e-06,
+      "loss": 0.616,
+      "step": 4004
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9499365091323853,
+      "learning_rate": 5.031125221565044e-06,
+      "loss": 0.6225,
+      "step": 4005
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.4018373489379883,
+      "learning_rate": 5.029050230976763e-06,
+      "loss": 0.6091,
+      "step": 4006
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8645708560943604,
+      "learning_rate": 5.026975235385172e-06,
+      "loss": 0.6104,
+      "step": 4007
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2363499402999878,
+      "learning_rate": 5.024900235147643e-06,
+      "loss": 0.6544,
+      "step": 4008
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.315367341041565,
+      "learning_rate": 5.022825230621555e-06,
+      "loss": 0.6242,
+      "step": 4009
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2281227111816406,
+      "learning_rate": 5.020750222164286e-06,
+      "loss": 0.5227,
+      "step": 4010
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3124263286590576,
+      "learning_rate": 5.0186752101332124e-06,
+      "loss": 0.6848,
+      "step": 4011
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.246238350868225,
+      "learning_rate": 5.016600194885714e-06,
+      "loss": 0.5938,
+      "step": 4012
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2616819143295288,
+      "learning_rate": 5.014525176779168e-06,
+      "loss": 0.6353,
+      "step": 4013
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8460921049118042,
+      "learning_rate": 5.012450156170957e-06,
+      "loss": 0.5932,
+      "step": 4014
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.32743501663208,
+      "learning_rate": 5.0103751334184595e-06,
+      "loss": 0.6467,
+      "step": 4015
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.421920657157898,
+      "learning_rate": 5.008300108879055e-06,
+      "loss": 0.521,
+      "step": 4016
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3539751768112183,
+      "learning_rate": 5.006225082910126e-06,
+      "loss": 0.5817,
+      "step": 4017
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1650025844573975,
+      "learning_rate": 5.004150055869053e-06,
+      "loss": 0.6017,
+      "step": 4018
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.121670126914978,
+      "learning_rate": 5.0020750281132165e-06,
+      "loss": 0.546,
+      "step": 4019
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4239189624786377,
+      "learning_rate": 5e-06,
+      "loss": 0.5849,
+      "step": 4020
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4270631074905396,
+      "learning_rate": 4.997924971886784e-06,
+      "loss": 0.5985,
+      "step": 4021
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1389853954315186,
+      "learning_rate": 4.995849944130948e-06,
+      "loss": 0.5686,
+      "step": 4022
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3634734153747559,
+      "learning_rate": 4.993774917089876e-06,
+      "loss": 0.6108,
+      "step": 4023
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4653452634811401,
+      "learning_rate": 4.991699891120947e-06,
+      "loss": 0.5871,
+      "step": 4024
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2894748449325562,
+      "learning_rate": 4.989624866581544e-06,
+      "loss": 0.575,
+      "step": 4025
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3853498697280884,
+      "learning_rate": 4.987549843829045e-06,
+      "loss": 0.6293,
+      "step": 4026
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2562291622161865,
+      "learning_rate": 4.985474823220835e-06,
+      "loss": 0.6276,
+      "step": 4027
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1109954118728638,
+      "learning_rate": 4.983399805114289e-06,
+      "loss": 0.5868,
+      "step": 4028
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2795517444610596,
+      "learning_rate": 4.981324789866788e-06,
+      "loss": 0.5745,
+      "step": 4029
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9505984783172607,
+      "learning_rate": 4.979249777835715e-06,
+      "loss": 0.5709,
+      "step": 4030
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2079765796661377,
+      "learning_rate": 4.977174769378445e-06,
+      "loss": 0.6797,
+      "step": 4031
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.466506004333496,
+      "learning_rate": 4.975099764852359e-06,
+      "loss": 0.6298,
+      "step": 4032
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.308029294013977,
+      "learning_rate": 4.973024764614829e-06,
+      "loss": 0.5299,
+      "step": 4033
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.230959892272949,
+      "learning_rate": 4.970949769023238e-06,
+      "loss": 0.6072,
+      "step": 4034
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2437947988510132,
+      "learning_rate": 4.968874778434957e-06,
+      "loss": 0.6981,
+      "step": 4035
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1782584190368652,
+      "learning_rate": 4.966799793207364e-06,
+      "loss": 0.5982,
+      "step": 4036
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3160803318023682,
+      "learning_rate": 4.964724813697831e-06,
+      "loss": 0.6438,
+      "step": 4037
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.471976637840271,
+      "learning_rate": 4.962649840263733e-06,
+      "loss": 0.5819,
+      "step": 4038
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1718621253967285,
+      "learning_rate": 4.960574873262439e-06,
+      "loss": 0.545,
+      "step": 4039
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1031982898712158,
+      "learning_rate": 4.9584999130513235e-06,
+      "loss": 0.551,
+      "step": 4040
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2355008125305176,
+      "learning_rate": 4.956424959987753e-06,
+      "loss": 0.538,
+      "step": 4041
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5801515579223633,
+      "learning_rate": 4.954350014429099e-06,
+      "loss": 0.622,
+      "step": 4042
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3405274152755737,
+      "learning_rate": 4.952275076732726e-06,
+      "loss": 0.6235,
+      "step": 4043
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0359275341033936,
+      "learning_rate": 4.950200147256002e-06,
+      "loss": 0.743,
+      "step": 4044
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0925586223602295,
+      "learning_rate": 4.948125226356288e-06,
+      "loss": 0.662,
+      "step": 4045
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.484156847000122,
+      "learning_rate": 4.94605031439095e-06,
+      "loss": 0.6197,
+      "step": 4046
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1978555917739868,
+      "learning_rate": 4.943975411717351e-06,
+      "loss": 0.6148,
+      "step": 4047
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3440272808074951,
+      "learning_rate": 4.941900518692846e-06,
+      "loss": 0.5443,
+      "step": 4048
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8050079345703125,
+      "learning_rate": 4.939825635674798e-06,
+      "loss": 0.6654,
+      "step": 4049
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0395833253860474,
+      "learning_rate": 4.93775076302056e-06,
+      "loss": 0.6051,
+      "step": 4050
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.008521556854248,
+      "learning_rate": 4.935675901087488e-06,
+      "loss": 0.6138,
+      "step": 4051
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.4628641605377197,
+      "learning_rate": 4.933601050232935e-06,
+      "loss": 0.5635,
+      "step": 4052
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2769087553024292,
+      "learning_rate": 4.931526210814251e-06,
+      "loss": 0.6141,
+      "step": 4053
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3242411613464355,
+      "learning_rate": 4.929451383188785e-06,
+      "loss": 0.6098,
+      "step": 4054
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0932570695877075,
+      "learning_rate": 4.927376567713886e-06,
+      "loss": 0.6458,
+      "step": 4055
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3448100090026855,
+      "learning_rate": 4.925301764746895e-06,
+      "loss": 0.5961,
+      "step": 4056
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2128664255142212,
+      "learning_rate": 4.923226974645158e-06,
+      "loss": 0.5711,
+      "step": 4057
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6496801376342773,
+      "learning_rate": 4.921152197766014e-06,
+      "loss": 0.5831,
+      "step": 4058
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2741731405258179,
+      "learning_rate": 4.919077434466802e-06,
+      "loss": 0.6547,
+      "step": 4059
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3321490287780762,
+      "learning_rate": 4.917002685104855e-06,
+      "loss": 0.5447,
+      "step": 4060
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.141733169555664,
+      "learning_rate": 4.914927950037511e-06,
+      "loss": 0.5685,
+      "step": 4061
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9314861297607422,
+      "learning_rate": 4.912853229622096e-06,
+      "loss": 0.5964,
+      "step": 4062
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4189141988754272,
+      "learning_rate": 4.910778524215941e-06,
+      "loss": 0.586,
+      "step": 4063
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2865058183670044,
+      "learning_rate": 4.908703834176373e-06,
+      "loss": 0.5715,
+      "step": 4064
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4331640005111694,
+      "learning_rate": 4.906629159860713e-06,
+      "loss": 0.6381,
+      "step": 4065
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0844879150390625,
+      "learning_rate": 4.904554501626284e-06,
+      "loss": 0.5739,
+      "step": 4066
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2306838035583496,
+      "learning_rate": 4.9024798598304006e-06,
+      "loss": 0.7441,
+      "step": 4067
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.091158390045166,
+      "learning_rate": 4.90040523483038e-06,
+      "loss": 0.6085,
+      "step": 4068
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3969615697860718,
+      "learning_rate": 4.898330626983533e-06,
+      "loss": 0.6604,
+      "step": 4069
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0094318389892578,
+      "learning_rate": 4.89625603664717e-06,
+      "loss": 0.5522,
+      "step": 4070
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2802209854125977,
+      "learning_rate": 4.894181464178595e-06,
+      "loss": 0.6044,
+      "step": 4071
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6502206325531006,
+      "learning_rate": 4.892106909935111e-06,
+      "loss": 0.6496,
+      "step": 4072
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3228180408477783,
+      "learning_rate": 4.890032374274021e-06,
+      "loss": 0.6288,
+      "step": 4073
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3824211359024048,
+      "learning_rate": 4.887957857552617e-06,
+      "loss": 0.6468,
+      "step": 4074
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9041395783424377,
+      "learning_rate": 4.885883360128197e-06,
+      "loss": 0.5782,
+      "step": 4075
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6577094793319702,
+      "learning_rate": 4.883808882358047e-06,
+      "loss": 0.6067,
+      "step": 4076
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2962143421173096,
+      "learning_rate": 4.881734424599456e-06,
+      "loss": 0.6659,
+      "step": 4077
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3377918004989624,
+      "learning_rate": 4.879659987209707e-06,
+      "loss": 0.6366,
+      "step": 4078
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.1966018676757812,
+      "learning_rate": 4.877585570546078e-06,
+      "loss": 0.6419,
+      "step": 4079
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3197449445724487,
+      "learning_rate": 4.875511174965846e-06,
+      "loss": 0.6256,
+      "step": 4080
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4747912883758545,
+      "learning_rate": 4.8734368008262835e-06,
+      "loss": 0.6736,
+      "step": 4081
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2974356412887573,
+      "learning_rate": 4.871362448484662e-06,
+      "loss": 0.5823,
+      "step": 4082
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4338997602462769,
+      "learning_rate": 4.869288118298242e-06,
+      "loss": 0.5833,
+      "step": 4083
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3909071683883667,
+      "learning_rate": 4.867213810624288e-06,
+      "loss": 0.5486,
+      "step": 4084
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.427019476890564,
+      "learning_rate": 4.865139525820055e-06,
+      "loss": 0.555,
+      "step": 4085
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.493001937866211,
+      "learning_rate": 4.8630652642428e-06,
+      "loss": 0.6544,
+      "step": 4086
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.26246178150177,
+      "learning_rate": 4.860991026249768e-06,
+      "loss": 0.6591,
+      "step": 4087
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4320074319839478,
+      "learning_rate": 4.858916812198206e-06,
+      "loss": 0.6442,
+      "step": 4088
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3701486587524414,
+      "learning_rate": 4.856842622445356e-06,
+      "loss": 0.5566,
+      "step": 4089
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2770544290542603,
+      "learning_rate": 4.854768457348456e-06,
+      "loss": 0.6557,
+      "step": 4090
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5645573139190674,
+      "learning_rate": 4.852694317264735e-06,
+      "loss": 0.6767,
+      "step": 4091
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.349003791809082,
+      "learning_rate": 4.850620202551425e-06,
+      "loss": 0.6173,
+      "step": 4092
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3395682573318481,
+      "learning_rate": 4.848546113565748e-06,
+      "loss": 0.5781,
+      "step": 4093
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.317745566368103,
+      "learning_rate": 4.846472050664925e-06,
+      "loss": 0.6592,
+      "step": 4094
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.009605884552002,
+      "learning_rate": 4.84439801420617e-06,
+      "loss": 0.5211,
+      "step": 4095
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3194994926452637,
+      "learning_rate": 4.842324004546696e-06,
+      "loss": 0.6196,
+      "step": 4096
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.368104338645935,
+      "learning_rate": 4.8402500220437054e-06,
+      "loss": 0.6548,
+      "step": 4097
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1761075258255005,
+      "learning_rate": 4.838176067054401e-06,
+      "loss": 0.6217,
+      "step": 4098
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5087071657180786,
+      "learning_rate": 4.836102139935982e-06,
+      "loss": 0.5141,
+      "step": 4099
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0981336832046509,
+      "learning_rate": 4.8340282410456365e-06,
+      "loss": 0.5627,
+      "step": 4100
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1873528957366943,
+      "learning_rate": 4.831954370740554e-06,
+      "loss": 0.6115,
+      "step": 4101
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1071884632110596,
+      "learning_rate": 4.829880529377915e-06,
+      "loss": 0.5372,
+      "step": 4102
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2068363428115845,
+      "learning_rate": 4.8278067173148975e-06,
+      "loss": 0.6624,
+      "step": 4103
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3392771482467651,
+      "learning_rate": 4.825732934908672e-06,
+      "loss": 0.6454,
+      "step": 4104
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.0413925647735596,
+      "learning_rate": 4.8236591825164085e-06,
+      "loss": 0.6027,
+      "step": 4105
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4464915990829468,
+      "learning_rate": 4.821585460495264e-06,
+      "loss": 0.6546,
+      "step": 4106
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.405840277671814,
+      "learning_rate": 4.819511769202399e-06,
+      "loss": 0.5513,
+      "step": 4107
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2467281818389893,
+      "learning_rate": 4.817438108994963e-06,
+      "loss": 0.5584,
+      "step": 4108
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.293278455734253,
+      "learning_rate": 4.815364480230103e-06,
+      "loss": 0.6054,
+      "step": 4109
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4756916761398315,
+      "learning_rate": 4.813290883264956e-06,
+      "loss": 0.6539,
+      "step": 4110
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.106099009513855,
+      "learning_rate": 4.811217318456661e-06,
+      "loss": 0.7242,
+      "step": 4111
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1641430854797363,
+      "learning_rate": 4.809143786162345e-06,
+      "loss": 0.605,
+      "step": 4112
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.113264799118042,
+      "learning_rate": 4.807070286739134e-06,
+      "loss": 0.753,
+      "step": 4113
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1319680213928223,
+      "learning_rate": 4.804996820544144e-06,
+      "loss": 0.7126,
+      "step": 4114
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2857635021209717,
+      "learning_rate": 4.8029233879344845e-06,
+      "loss": 0.6106,
+      "step": 4115
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3838752508163452,
+      "learning_rate": 4.800849989267269e-06,
+      "loss": 0.5595,
+      "step": 4116
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 6.1764984130859375,
+      "learning_rate": 4.798776624899595e-06,
+      "loss": 0.5441,
+      "step": 4117
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6189913749694824,
+      "learning_rate": 4.796703295188557e-06,
+      "loss": 0.6184,
+      "step": 4118
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.234065532684326,
+      "learning_rate": 4.7946300004912454e-06,
+      "loss": 0.671,
+      "step": 4119
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2225993871688843,
+      "learning_rate": 4.7925567411647405e-06,
+      "loss": 0.6519,
+      "step": 4120
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2532271146774292,
+      "learning_rate": 4.790483517566122e-06,
+      "loss": 0.5927,
+      "step": 4121
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2144299745559692,
+      "learning_rate": 4.788410330052457e-06,
+      "loss": 0.724,
+      "step": 4122
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2604823112487793,
+      "learning_rate": 4.7863371789808135e-06,
+      "loss": 0.5651,
+      "step": 4123
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.168423056602478,
+      "learning_rate": 4.784264064708247e-06,
+      "loss": 0.5987,
+      "step": 4124
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.171149730682373,
+      "learning_rate": 4.782190987591811e-06,
+      "loss": 0.6428,
+      "step": 4125
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.292298674583435,
+      "learning_rate": 4.7801179479885495e-06,
+      "loss": 0.6348,
+      "step": 4126
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.250784993171692,
+      "learning_rate": 4.778044946255503e-06,
+      "loss": 0.5781,
+      "step": 4127
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4512914419174194,
+      "learning_rate": 4.775971982749703e-06,
+      "loss": 0.5359,
+      "step": 4128
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.359748363494873,
+      "learning_rate": 4.773899057828176e-06,
+      "loss": 0.5784,
+      "step": 4129
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3749535083770752,
+      "learning_rate": 4.771826171847939e-06,
+      "loss": 0.6338,
+      "step": 4130
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.980396032333374,
+      "learning_rate": 4.769753325166008e-06,
+      "loss": 0.6662,
+      "step": 4131
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5456019639968872,
+      "learning_rate": 4.7676805181393835e-06,
+      "loss": 0.6045,
+      "step": 4132
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.6582868099212646,
+      "learning_rate": 4.76560775112507e-06,
+      "loss": 0.5481,
+      "step": 4133
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1082854270935059,
+      "learning_rate": 4.763535024480057e-06,
+      "loss": 0.6318,
+      "step": 4134
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.2785942554473877,
+      "learning_rate": 4.761462338561329e-06,
+      "loss": 0.611,
+      "step": 4135
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0957773923873901,
+      "learning_rate": 4.759389693725867e-06,
+      "loss": 0.706,
+      "step": 4136
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9537548422813416,
+      "learning_rate": 4.757317090330638e-06,
+      "loss": 0.5881,
+      "step": 4137
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0327272415161133,
+      "learning_rate": 4.755244528732608e-06,
+      "loss": 0.5027,
+      "step": 4138
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2874075174331665,
+      "learning_rate": 4.753172009288732e-06,
+      "loss": 0.6377,
+      "step": 4139
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2741062641143799,
+      "learning_rate": 4.751099532355962e-06,
+      "loss": 0.6127,
+      "step": 4140
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2608203887939453,
+      "learning_rate": 4.749027098291237e-06,
+      "loss": 0.583,
+      "step": 4141
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.517383337020874,
+      "learning_rate": 4.7469547074514946e-06,
+      "loss": 0.5961,
+      "step": 4142
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.231284499168396,
+      "learning_rate": 4.7448823601936585e-06,
+      "loss": 0.583,
+      "step": 4143
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1564208269119263,
+      "learning_rate": 4.742810056874652e-06,
+      "loss": 0.6192,
+      "step": 4144
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2633081674575806,
+      "learning_rate": 4.740737797851385e-06,
+      "loss": 0.5658,
+      "step": 4145
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1155922412872314,
+      "learning_rate": 4.7386655834807634e-06,
+      "loss": 0.5894,
+      "step": 4146
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3662333488464355,
+      "learning_rate": 4.736593414119682e-06,
+      "loss": 0.6027,
+      "step": 4147
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2502559423446655,
+      "learning_rate": 4.734521290125032e-06,
+      "loss": 0.5672,
+      "step": 4148
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3026888370513916,
+      "learning_rate": 4.732449211853693e-06,
+      "loss": 0.5624,
+      "step": 4149
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1376659870147705,
+      "learning_rate": 4.730377179662538e-06,
+      "loss": 0.5888,
+      "step": 4150
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2101504802703857,
+      "learning_rate": 4.728305193908436e-06,
+      "loss": 0.6549,
+      "step": 4151
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.048017144203186,
+      "learning_rate": 4.72623325494824e-06,
+      "loss": 0.6031,
+      "step": 4152
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2536982297897339,
+      "learning_rate": 4.7241613631388034e-06,
+      "loss": 0.5735,
+      "step": 4153
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2073613405227661,
+      "learning_rate": 4.722089518836964e-06,
+      "loss": 0.6656,
+      "step": 4154
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3333324193954468,
+      "learning_rate": 4.720017722399557e-06,
+      "loss": 0.6033,
+      "step": 4155
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8105237483978271,
+      "learning_rate": 4.717945974183405e-06,
+      "loss": 0.4994,
+      "step": 4156
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.501945972442627,
+      "learning_rate": 4.715874274545328e-06,
+      "loss": 0.5396,
+      "step": 4157
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8984172344207764,
+      "learning_rate": 4.71380262384213e-06,
+      "loss": 0.549,
+      "step": 4158
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.550038456916809,
+      "learning_rate": 4.711731022430615e-06,
+      "loss": 0.5852,
+      "step": 4159
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7350627183914185,
+      "learning_rate": 4.70965947066757e-06,
+      "loss": 0.6284,
+      "step": 4160
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5158087015151978,
+      "learning_rate": 4.707587968909782e-06,
+      "loss": 0.5684,
+      "step": 4161
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1972614526748657,
+      "learning_rate": 4.705516517514021e-06,
+      "loss": 0.5638,
+      "step": 4162
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1058855056762695,
+      "learning_rate": 4.703445116837055e-06,
+      "loss": 0.5002,
+      "step": 4163
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3835372924804688,
+      "learning_rate": 4.701373767235641e-06,
+      "loss": 0.6697,
+      "step": 4164
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.174512267112732,
+      "learning_rate": 4.699302469066524e-06,
+      "loss": 0.6428,
+      "step": 4165
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.310792326927185,
+      "learning_rate": 4.6972312226864445e-06,
+      "loss": 0.652,
+      "step": 4166
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3059464693069458,
+      "learning_rate": 4.6951600284521324e-06,
+      "loss": 0.5807,
+      "step": 4167
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.334140658378601,
+      "learning_rate": 4.6930888867203115e-06,
+      "loss": 0.6129,
+      "step": 4168
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4384865760803223,
+      "learning_rate": 4.691017797847692e-06,
+      "loss": 0.6122,
+      "step": 4169
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3629262447357178,
+      "learning_rate": 4.688946762190975e-06,
+      "loss": 0.6201,
+      "step": 4170
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.473082423210144,
+      "learning_rate": 4.686875780106856e-06,
+      "loss": 0.5797,
+      "step": 4171
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.335864782333374,
+      "learning_rate": 4.68480485195202e-06,
+      "loss": 0.6164,
+      "step": 4172
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.127086639404297,
+      "learning_rate": 4.682733978083142e-06,
+      "loss": 0.6167,
+      "step": 4173
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2139432430267334,
+      "learning_rate": 4.680663158856886e-06,
+      "loss": 0.6562,
+      "step": 4174
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4341139793395996,
+      "learning_rate": 4.678592394629912e-06,
+      "loss": 0.6285,
+      "step": 4175
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3252298831939697,
+      "learning_rate": 4.676521685758863e-06,
+      "loss": 0.6012,
+      "step": 4176
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3366296291351318,
+      "learning_rate": 4.6744510326003805e-06,
+      "loss": 0.6053,
+      "step": 4177
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.302324652671814,
+      "learning_rate": 4.672380435511089e-06,
+      "loss": 0.6217,
+      "step": 4178
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.673362374305725,
+      "learning_rate": 4.67030989484761e-06,
+      "loss": 0.6395,
+      "step": 4179
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6342353820800781,
+      "learning_rate": 4.668239410966549e-06,
+      "loss": 0.5935,
+      "step": 4180
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6448659896850586,
+      "learning_rate": 4.666168984224508e-06,
+      "loss": 0.555,
+      "step": 4181
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3737086057662964,
+      "learning_rate": 4.664098614978073e-06,
+      "loss": 0.5609,
+      "step": 4182
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.202628254890442,
+      "learning_rate": 4.662028303583823e-06,
+      "loss": 0.6397,
+      "step": 4183
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1492021083831787,
+      "learning_rate": 4.6599580503983295e-06,
+      "loss": 0.6714,
+      "step": 4184
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1646534204483032,
+      "learning_rate": 4.657887855778149e-06,
+      "loss": 0.6116,
+      "step": 4185
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2870683670043945,
+      "learning_rate": 4.655817720079834e-06,
+      "loss": 0.5639,
+      "step": 4186
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.9178974628448486,
+      "learning_rate": 4.6537476436599184e-06,
+      "loss": 0.6177,
+      "step": 4187
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0415689945220947,
+      "learning_rate": 4.651677626874936e-06,
+      "loss": 0.5541,
+      "step": 4188
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0987578630447388,
+      "learning_rate": 4.6496076700814e-06,
+      "loss": 0.6257,
+      "step": 4189
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4872864484786987,
+      "learning_rate": 4.647537773635823e-06,
+      "loss": 0.6029,
+      "step": 4190
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2052521705627441,
+      "learning_rate": 4.645467937894699e-06,
+      "loss": 0.6159,
+      "step": 4191
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0160893201828003,
+      "learning_rate": 4.643398163214517e-06,
+      "loss": 0.6487,
+      "step": 4192
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.498758316040039,
+      "learning_rate": 4.641328449951753e-06,
+      "loss": 0.5912,
+      "step": 4193
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1858739852905273,
+      "learning_rate": 4.6392587984628735e-06,
+      "loss": 0.7092,
+      "step": 4194
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1768958568572998,
+      "learning_rate": 4.637189209104333e-06,
+      "loss": 0.5921,
+      "step": 4195
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3898215293884277,
+      "learning_rate": 4.635119682232577e-06,
+      "loss": 0.6398,
+      "step": 4196
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3137660026550293,
+      "learning_rate": 4.63305021820404e-06,
+      "loss": 0.6571,
+      "step": 4197
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1577122211456299,
+      "learning_rate": 4.6309808173751445e-06,
+      "loss": 0.7469,
+      "step": 4198
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4007368087768555,
+      "learning_rate": 4.628911480102301e-06,
+      "loss": 0.6021,
+      "step": 4199
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1606431007385254,
+      "learning_rate": 4.626842206741912e-06,
+      "loss": 0.6898,
+      "step": 4200
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4867026805877686,
+      "learning_rate": 4.62477299765037e-06,
+      "loss": 0.6388,
+      "step": 4201
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.663015365600586,
+      "learning_rate": 4.622703853184052e-06,
+      "loss": 0.5711,
+      "step": 4202
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1145719289779663,
+      "learning_rate": 4.620634773699327e-06,
+      "loss": 0.603,
+      "step": 4203
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4312100410461426,
+      "learning_rate": 4.61856575955255e-06,
+      "loss": 0.6286,
+      "step": 4204
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1849942207336426,
+      "learning_rate": 4.6164968111000695e-06,
+      "loss": 0.6667,
+      "step": 4205
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0929937362670898,
+      "learning_rate": 4.614427928698217e-06,
+      "loss": 0.6663,
+      "step": 4206
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.972935140132904,
+      "learning_rate": 4.612359112703318e-06,
+      "loss": 0.5812,
+      "step": 4207
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4837443828582764,
+      "learning_rate": 4.610290363471681e-06,
+      "loss": 0.6155,
+      "step": 4208
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6648123264312744,
+      "learning_rate": 4.608221681359609e-06,
+      "loss": 0.5247,
+      "step": 4209
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3424453735351562,
+      "learning_rate": 4.606153066723389e-06,
+      "loss": 0.6489,
+      "step": 4210
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4238646030426025,
+      "learning_rate": 4.604084519919298e-06,
+      "loss": 0.5952,
+      "step": 4211
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4087281227111816,
+      "learning_rate": 4.602016041303601e-06,
+      "loss": 0.6647,
+      "step": 4212
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5249015092849731,
+      "learning_rate": 4.599947631232552e-06,
+      "loss": 0.6362,
+      "step": 4213
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3257591724395752,
+      "learning_rate": 4.597879290062393e-06,
+      "loss": 0.5909,
+      "step": 4214
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1432000398635864,
+      "learning_rate": 4.595811018149351e-06,
+      "loss": 0.5854,
+      "step": 4215
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2500845193862915,
+      "learning_rate": 4.5937428158496475e-06,
+      "loss": 0.5857,
+      "step": 4216
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0443065166473389,
+      "learning_rate": 4.591674683519483e-06,
+      "loss": 0.6195,
+      "step": 4217
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2430527210235596,
+      "learning_rate": 4.589606621515057e-06,
+      "loss": 0.5721,
+      "step": 4218
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3105758428573608,
+      "learning_rate": 4.5875386301925495e-06,
+      "loss": 0.4901,
+      "step": 4219
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6490211486816406,
+      "learning_rate": 4.5854707099081285e-06,
+      "loss": 0.6874,
+      "step": 4220
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3691102266311646,
+      "learning_rate": 4.583402861017953e-06,
+      "loss": 0.594,
+      "step": 4221
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2332993745803833,
+      "learning_rate": 4.5813350838781665e-06,
+      "loss": 0.6166,
+      "step": 4222
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3578057289123535,
+      "learning_rate": 4.579267378844902e-06,
+      "loss": 0.6297,
+      "step": 4223
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7097007036209106,
+      "learning_rate": 4.577199746274279e-06,
+      "loss": 0.6119,
+      "step": 4224
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.160226583480835,
+      "learning_rate": 4.575132186522408e-06,
+      "loss": 0.5266,
+      "step": 4225
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.177698016166687,
+      "learning_rate": 4.5730646999453805e-06,
+      "loss": 0.6148,
+      "step": 4226
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4060474634170532,
+      "learning_rate": 4.570997286899282e-06,
+      "loss": 0.5874,
+      "step": 4227
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2073032855987549,
+      "learning_rate": 4.56892994774018e-06,
+      "loss": 0.5137,
+      "step": 4228
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2490034103393555,
+      "learning_rate": 4.566862682824133e-06,
+      "loss": 0.58,
+      "step": 4229
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3035485744476318,
+      "learning_rate": 4.564795492507184e-06,
+      "loss": 0.6719,
+      "step": 4230
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.219054102897644,
+      "learning_rate": 4.562728377145367e-06,
+      "loss": 0.6331,
+      "step": 4231
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.202539324760437,
+      "learning_rate": 4.560661337094698e-06,
+      "loss": 0.5764,
+      "step": 4232
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1667770147323608,
+      "learning_rate": 4.558594372711185e-06,
+      "loss": 0.5885,
+      "step": 4233
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0304358005523682,
+      "learning_rate": 4.556527484350819e-06,
+      "loss": 0.5146,
+      "step": 4234
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0976370573043823,
+      "learning_rate": 4.554460672369578e-06,
+      "loss": 0.574,
+      "step": 4235
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.336747646331787,
+      "learning_rate": 4.552393937123432e-06,
+      "loss": 0.5512,
+      "step": 4236
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2262814044952393,
+      "learning_rate": 4.550327278968333e-06,
+      "loss": 0.5616,
+      "step": 4237
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2212550640106201,
+      "learning_rate": 4.548260698260219e-06,
+      "loss": 0.6184,
+      "step": 4238
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4734658002853394,
+      "learning_rate": 4.546194195355018e-06,
+      "loss": 0.5742,
+      "step": 4239
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4658452272415161,
+      "learning_rate": 4.544127770608644e-06,
+      "loss": 0.6211,
+      "step": 4240
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2137492895126343,
+      "learning_rate": 4.542061424376995e-06,
+      "loss": 0.5261,
+      "step": 4241
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2533836364746094,
+      "learning_rate": 4.539995157015957e-06,
+      "loss": 0.6744,
+      "step": 4242
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.436880111694336,
+      "learning_rate": 4.537928968881404e-06,
+      "loss": 0.6731,
+      "step": 4243
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1234179735183716,
+      "learning_rate": 4.535862860329195e-06,
+      "loss": 0.5358,
+      "step": 4244
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4436432123184204,
+      "learning_rate": 4.533796831715172e-06,
+      "loss": 0.6199,
+      "step": 4245
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6011847257614136,
+      "learning_rate": 4.531730883395171e-06,
+      "loss": 0.5899,
+      "step": 4246
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1307777166366577,
+      "learning_rate": 4.529665015725006e-06,
+      "loss": 0.6098,
+      "step": 4247
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4225749969482422,
+      "learning_rate": 4.527599229060483e-06,
+      "loss": 0.6267,
+      "step": 4248
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1333550214767456,
+      "learning_rate": 4.5255335237573905e-06,
+      "loss": 0.6978,
+      "step": 4249
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0923420190811157,
+      "learning_rate": 4.5234679001715055e-06,
+      "loss": 0.6762,
+      "step": 4250
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4398398399353027,
+      "learning_rate": 4.521402358658587e-06,
+      "loss": 0.684,
+      "step": 4251
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3944870233535767,
+      "learning_rate": 4.519336899574384e-06,
+      "loss": 0.6649,
+      "step": 4252
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.357353925704956,
+      "learning_rate": 4.517271523274632e-06,
+      "loss": 0.5606,
+      "step": 4253
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7038229703903198,
+      "learning_rate": 4.515206230115047e-06,
+      "loss": 0.6472,
+      "step": 4254
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4553661346435547,
+      "learning_rate": 4.5131410204513375e-06,
+      "loss": 0.5576,
+      "step": 4255
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0828001499176025,
+      "learning_rate": 4.511075894639189e-06,
+      "loss": 0.6354,
+      "step": 4256
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3921860456466675,
+      "learning_rate": 4.509010853034281e-06,
+      "loss": 0.5799,
+      "step": 4257
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2910650968551636,
+      "learning_rate": 4.506945895992274e-06,
+      "loss": 0.6111,
+      "step": 4258
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2620890140533447,
+      "learning_rate": 4.5048810238688145e-06,
+      "loss": 0.7103,
+      "step": 4259
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2186883687973022,
+      "learning_rate": 4.502816237019534e-06,
+      "loss": 0.6864,
+      "step": 4260
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.271498680114746,
+      "learning_rate": 4.5007515358000525e-06,
+      "loss": 0.5766,
+      "step": 4261
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3539847135543823,
+      "learning_rate": 4.498686920565972e-06,
+      "loss": 0.6176,
+      "step": 4262
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1289831399917603,
+      "learning_rate": 4.496622391672878e-06,
+      "loss": 0.5878,
+      "step": 4263
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1196657419204712,
+      "learning_rate": 4.494557949476347e-06,
+      "loss": 0.5116,
+      "step": 4264
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3079795837402344,
+      "learning_rate": 4.492493594331934e-06,
+      "loss": 0.524,
+      "step": 4265
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2875686883926392,
+      "learning_rate": 4.490429326595185e-06,
+      "loss": 0.5328,
+      "step": 4266
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1164149045944214,
+      "learning_rate": 4.488365146621626e-06,
+      "loss": 0.6377,
+      "step": 4267
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.228834867477417,
+      "learning_rate": 4.486301054766773e-06,
+      "loss": 0.5413,
+      "step": 4268
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0576958656311035,
+      "learning_rate": 4.484237051386119e-06,
+      "loss": 0.6451,
+      "step": 4269
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1337889432907104,
+      "learning_rate": 4.482173136835152e-06,
+      "loss": 0.68,
+      "step": 4270
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1818243265151978,
+      "learning_rate": 4.480109311469336e-06,
+      "loss": 0.5929,
+      "step": 4271
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.773174524307251,
+      "learning_rate": 4.4780455756441245e-06,
+      "loss": 0.5298,
+      "step": 4272
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5445469617843628,
+      "learning_rate": 4.475981929714953e-06,
+      "loss": 0.5569,
+      "step": 4273
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3637791872024536,
+      "learning_rate": 4.473918374037244e-06,
+      "loss": 0.6169,
+      "step": 4274
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1337825059890747,
+      "learning_rate": 4.471854908966402e-06,
+      "loss": 0.5918,
+      "step": 4275
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2219849824905396,
+      "learning_rate": 4.469791534857816e-06,
+      "loss": 0.5439,
+      "step": 4276
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.176208734512329,
+      "learning_rate": 4.467728252066862e-06,
+      "loss": 0.6635,
+      "step": 4277
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3340530395507812,
+      "learning_rate": 4.465665060948897e-06,
+      "loss": 0.5424,
+      "step": 4278
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.166491985321045,
+      "learning_rate": 4.4636019618592655e-06,
+      "loss": 0.665,
+      "step": 4279
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9526617527008057,
+      "learning_rate": 4.461538955153292e-06,
+      "loss": 0.6259,
+      "step": 4280
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2087116241455078,
+      "learning_rate": 4.4594760411862905e-06,
+      "loss": 0.7804,
+      "step": 4281
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6400904655456543,
+      "learning_rate": 4.457413220313553e-06,
+      "loss": 0.6266,
+      "step": 4282
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6623104810714722,
+      "learning_rate": 4.455350492890361e-06,
+      "loss": 0.6399,
+      "step": 4283
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3035902976989746,
+      "learning_rate": 4.453287859271975e-06,
+      "loss": 0.5671,
+      "step": 4284
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1311049461364746,
+      "learning_rate": 4.451225319813644e-06,
+      "loss": 0.5505,
+      "step": 4285
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.258715033531189,
+      "learning_rate": 4.449162874870595e-06,
+      "loss": 0.62,
+      "step": 4286
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3356287479400635,
+      "learning_rate": 4.4471005247980464e-06,
+      "loss": 0.6583,
+      "step": 4287
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2244027853012085,
+      "learning_rate": 4.445038269951195e-06,
+      "loss": 0.5537,
+      "step": 4288
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3109939098358154,
+      "learning_rate": 4.4429761106852204e-06,
+      "loss": 0.5879,
+      "step": 4289
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1747775077819824,
+      "learning_rate": 4.44091404735529e-06,
+      "loss": 0.5629,
+      "step": 4290
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4858828783035278,
+      "learning_rate": 4.4388520803165495e-06,
+      "loss": 0.6296,
+      "step": 4291
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.4959919452667236,
+      "learning_rate": 4.436790209924134e-06,
+      "loss": 0.6197,
+      "step": 4292
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1516481637954712,
+      "learning_rate": 4.434728436533156e-06,
+      "loss": 0.7115,
+      "step": 4293
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2591958045959473,
+      "learning_rate": 4.4326667604987165e-06,
+      "loss": 0.571,
+      "step": 4294
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4871764183044434,
+      "learning_rate": 4.430605182175895e-06,
+      "loss": 0.5495,
+      "step": 4295
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2093896865844727,
+      "learning_rate": 4.428543701919758e-06,
+      "loss": 0.5496,
+      "step": 4296
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2225619554519653,
+      "learning_rate": 4.426482320085352e-06,
+      "loss": 0.5976,
+      "step": 4297
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4069699048995972,
+      "learning_rate": 4.424421037027711e-06,
+      "loss": 0.5725,
+      "step": 4298
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7598546743392944,
+      "learning_rate": 4.422359853101846e-06,
+      "loss": 0.5402,
+      "step": 4299
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1938350200653076,
+      "learning_rate": 4.420298768662756e-06,
+      "loss": 0.5808,
+      "step": 4300
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3357658386230469,
+      "learning_rate": 4.418237784065419e-06,
+      "loss": 0.6606,
+      "step": 4301
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4759631156921387,
+      "learning_rate": 4.416176899664801e-06,
+      "loss": 0.6102,
+      "step": 4302
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2692142724990845,
+      "learning_rate": 4.4141161158158426e-06,
+      "loss": 0.6524,
+      "step": 4303
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1206196546554565,
+      "learning_rate": 4.412055432873475e-06,
+      "loss": 0.6809,
+      "step": 4304
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9297462701797485,
+      "learning_rate": 4.409994851192611e-06,
+      "loss": 0.5622,
+      "step": 4305
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0943596363067627,
+      "learning_rate": 4.40793437112814e-06,
+      "loss": 0.5449,
+      "step": 4306
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.434758186340332,
+      "learning_rate": 4.4058739930349406e-06,
+      "loss": 0.6168,
+      "step": 4307
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8183618783950806,
+      "learning_rate": 4.403813717267869e-06,
+      "loss": 0.5821,
+      "step": 4308
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0242024660110474,
+      "learning_rate": 4.401753544181767e-06,
+      "loss": 0.6293,
+      "step": 4309
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1580708026885986,
+      "learning_rate": 4.399693474131456e-06,
+      "loss": 0.6038,
+      "step": 4310
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2843397855758667,
+      "learning_rate": 4.3976335074717446e-06,
+      "loss": 0.6346,
+      "step": 4311
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1573116779327393,
+      "learning_rate": 4.3955736445574176e-06,
+      "loss": 0.6021,
+      "step": 4312
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.339087963104248,
+      "learning_rate": 4.393513885743243e-06,
+      "loss": 0.5863,
+      "step": 4313
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3947805166244507,
+      "learning_rate": 4.391454231383976e-06,
+      "loss": 0.649,
+      "step": 4314
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3123061656951904,
+      "learning_rate": 4.389394681834348e-06,
+      "loss": 0.6246,
+      "step": 4315
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.151753306388855,
+      "learning_rate": 4.387335237449076e-06,
+      "loss": 0.709,
+      "step": 4316
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.03976571559906,
+      "learning_rate": 4.385275898582855e-06,
+      "loss": 0.549,
+      "step": 4317
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5990352630615234,
+      "learning_rate": 4.383216665590366e-06,
+      "loss": 0.7688,
+      "step": 4318
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2555058002471924,
+      "learning_rate": 4.381157538826269e-06,
+      "loss": 0.6593,
+      "step": 4319
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5851774215698242,
+      "learning_rate": 4.379098518645207e-06,
+      "loss": 0.6126,
+      "step": 4320
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2325047254562378,
+      "learning_rate": 4.377039605401807e-06,
+      "loss": 0.6097,
+      "step": 4321
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3471224308013916,
+      "learning_rate": 4.374980799450672e-06,
+      "loss": 0.5994,
+      "step": 4322
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.4533209800720215,
+      "learning_rate": 4.372922101146391e-06,
+      "loss": 0.5687,
+      "step": 4323
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.748429536819458,
+      "learning_rate": 4.370863510843531e-06,
+      "loss": 0.6614,
+      "step": 4324
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2647942304611206,
+      "learning_rate": 4.368805028896645e-06,
+      "loss": 0.6273,
+      "step": 4325
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4914000034332275,
+      "learning_rate": 4.366746655660262e-06,
+      "loss": 0.7031,
+      "step": 4326
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.345897912979126,
+      "learning_rate": 4.364688391488897e-06,
+      "loss": 0.5652,
+      "step": 4327
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1223891973495483,
+      "learning_rate": 4.362630236737043e-06,
+      "loss": 0.6655,
+      "step": 4328
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1209384202957153,
+      "learning_rate": 4.360572191759176e-06,
+      "loss": 0.6175,
+      "step": 4329
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4959664344787598,
+      "learning_rate": 4.358514256909751e-06,
+      "loss": 0.6079,
+      "step": 4330
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6316180229187012,
+      "learning_rate": 4.356456432543208e-06,
+      "loss": 0.6025,
+      "step": 4331
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7658164501190186,
+      "learning_rate": 4.354398719013964e-06,
+      "loss": 0.5352,
+      "step": 4332
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.239953875541687,
+      "learning_rate": 4.352341116676418e-06,
+      "loss": 0.6356,
+      "step": 4333
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2205899953842163,
+      "learning_rate": 4.350283625884949e-06,
+      "loss": 0.7296,
+      "step": 4334
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6176862716674805,
+      "learning_rate": 4.348226246993922e-06,
+      "loss": 0.6232,
+      "step": 4335
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2962661981582642,
+      "learning_rate": 4.346168980357674e-06,
+      "loss": 0.6093,
+      "step": 4336
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1365058422088623,
+      "learning_rate": 4.344111826330529e-06,
+      "loss": 0.6139,
+      "step": 4337
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4743257761001587,
+      "learning_rate": 4.342054785266792e-06,
+      "loss": 0.6462,
+      "step": 4338
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2239124774932861,
+      "learning_rate": 4.339997857520745e-06,
+      "loss": 0.58,
+      "step": 4339
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2178657054901123,
+      "learning_rate": 4.337941043446653e-06,
+      "loss": 0.6802,
+      "step": 4340
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9617922902107239,
+      "learning_rate": 4.335884343398757e-06,
+      "loss": 0.6315,
+      "step": 4341
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1992239952087402,
+      "learning_rate": 4.333827757731286e-06,
+      "loss": 0.6106,
+      "step": 4342
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2823426723480225,
+      "learning_rate": 4.331771286798442e-06,
+      "loss": 0.7401,
+      "step": 4343
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0556762218475342,
+      "learning_rate": 4.329714930954414e-06,
+      "loss": 0.5582,
+      "step": 4344
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2360563278198242,
+      "learning_rate": 4.327658690553362e-06,
+      "loss": 0.5358,
+      "step": 4345
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4432363510131836,
+      "learning_rate": 4.325602565949437e-06,
+      "loss": 0.6498,
+      "step": 4346
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0684760808944702,
+      "learning_rate": 4.3235465574967615e-06,
+      "loss": 0.5381,
+      "step": 4347
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2283833026885986,
+      "learning_rate": 4.321490665549442e-06,
+      "loss": 0.525,
+      "step": 4348
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2738137245178223,
+      "learning_rate": 4.319434890461563e-06,
+      "loss": 0.6444,
+      "step": 4349
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3298536539077759,
+      "learning_rate": 4.317379232587194e-06,
+      "loss": 0.6234,
+      "step": 4350
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.158233642578125,
+      "learning_rate": 4.315323692280375e-06,
+      "loss": 0.5949,
+      "step": 4351
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.0666298866271973,
+      "learning_rate": 4.313268269895134e-06,
+      "loss": 0.5992,
+      "step": 4352
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5501093864440918,
+      "learning_rate": 4.3112129657854755e-06,
+      "loss": 0.5746,
+      "step": 4353
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7205291986465454,
+      "learning_rate": 4.3091577803053816e-06,
+      "loss": 0.6069,
+      "step": 4354
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.564469337463379,
+      "learning_rate": 4.3071027138088206e-06,
+      "loss": 0.5885,
+      "step": 4355
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.302836298942566,
+      "learning_rate": 4.305047766649733e-06,
+      "loss": 0.6223,
+      "step": 4356
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3741014003753662,
+      "learning_rate": 4.302992939182042e-06,
+      "loss": 0.5507,
+      "step": 4357
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.358909249305725,
+      "learning_rate": 4.30093823175965e-06,
+      "loss": 0.6304,
+      "step": 4358
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1862109899520874,
+      "learning_rate": 4.298883644736438e-06,
+      "loss": 0.7494,
+      "step": 4359
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1762068271636963,
+      "learning_rate": 4.296829178466268e-06,
+      "loss": 0.6076,
+      "step": 4360
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2821547985076904,
+      "learning_rate": 4.294774833302981e-06,
+      "loss": 0.627,
+      "step": 4361
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2723315954208374,
+      "learning_rate": 4.292720609600393e-06,
+      "loss": 0.6285,
+      "step": 4362
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2723183631896973,
+      "learning_rate": 4.290666507712304e-06,
+      "loss": 0.6313,
+      "step": 4363
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1899631023406982,
+      "learning_rate": 4.288612527992492e-06,
+      "loss": 0.6145,
+      "step": 4364
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3209925889968872,
+      "learning_rate": 4.286558670794712e-06,
+      "loss": 0.5391,
+      "step": 4365
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0585249662399292,
+      "learning_rate": 4.284504936472701e-06,
+      "loss": 0.5716,
+      "step": 4366
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8191659450531006,
+      "learning_rate": 4.28245132538017e-06,
+      "loss": 0.5629,
+      "step": 4367
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2652943134307861,
+      "learning_rate": 4.2803978378708145e-06,
+      "loss": 0.6255,
+      "step": 4368
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.146340250968933,
+      "learning_rate": 4.278344474298304e-06,
+      "loss": 0.726,
+      "step": 4369
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.114557147026062,
+      "learning_rate": 4.276291235016291e-06,
+      "loss": 0.6706,
+      "step": 4370
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4655970335006714,
+      "learning_rate": 4.274238120378401e-06,
+      "loss": 0.5154,
+      "step": 4371
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.245018720626831,
+      "learning_rate": 4.272185130738243e-06,
+      "loss": 0.5518,
+      "step": 4372
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4143431186676025,
+      "learning_rate": 4.270132266449404e-06,
+      "loss": 0.589,
+      "step": 4373
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2347571849822998,
+      "learning_rate": 4.268079527865447e-06,
+      "loss": 0.712,
+      "step": 4374
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2976598739624023,
+      "learning_rate": 4.266026915339915e-06,
+      "loss": 0.5632,
+      "step": 4375
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4560409784317017,
+      "learning_rate": 4.263974429226327e-06,
+      "loss": 0.6189,
+      "step": 4376
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3681960105895996,
+      "learning_rate": 4.261922069878185e-06,
+      "loss": 0.4996,
+      "step": 4377
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.385707139968872,
+      "learning_rate": 4.259869837648963e-06,
+      "loss": 0.615,
+      "step": 4378
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.20070743560791,
+      "learning_rate": 4.2578177328921185e-06,
+      "loss": 0.5422,
+      "step": 4379
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4489797353744507,
+      "learning_rate": 4.255765755961083e-06,
+      "loss": 0.6436,
+      "step": 4380
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.207725167274475,
+      "learning_rate": 4.253713907209271e-06,
+      "loss": 0.5272,
+      "step": 4381
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0956413745880127,
+      "learning_rate": 4.251662186990067e-06,
+      "loss": 0.5381,
+      "step": 4382
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.241393804550171,
+      "learning_rate": 4.249610595656843e-06,
+      "loss": 0.6241,
+      "step": 4383
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1352612972259521,
+      "learning_rate": 4.24755913356294e-06,
+      "loss": 0.6223,
+      "step": 4384
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1019214391708374,
+      "learning_rate": 4.245507801061684e-06,
+      "loss": 0.7584,
+      "step": 4385
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3050336837768555,
+      "learning_rate": 4.243456598506373e-06,
+      "loss": 0.6316,
+      "step": 4386
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0004106760025024,
+      "learning_rate": 4.241405526250285e-06,
+      "loss": 0.5286,
+      "step": 4387
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.244195818901062,
+      "learning_rate": 4.239354584646677e-06,
+      "loss": 0.5851,
+      "step": 4388
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2075914144515991,
+      "learning_rate": 4.2373037740487785e-06,
+      "loss": 0.5609,
+      "step": 4389
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2431285381317139,
+      "learning_rate": 4.235253094809804e-06,
+      "loss": 0.5163,
+      "step": 4390
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.424965739250183,
+      "learning_rate": 4.233202547282941e-06,
+      "loss": 0.5798,
+      "step": 4391
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2569266557693481,
+      "learning_rate": 4.231152131821353e-06,
+      "loss": 0.6726,
+      "step": 4392
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3034515380859375,
+      "learning_rate": 4.2291018487781825e-06,
+      "loss": 0.6889,
+      "step": 4393
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1181299686431885,
+      "learning_rate": 4.227051698506551e-06,
+      "loss": 0.622,
+      "step": 4394
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.356863260269165,
+      "learning_rate": 4.225001681359552e-06,
+      "loss": 0.6016,
+      "step": 4395
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.473215937614441,
+      "learning_rate": 4.222951797690262e-06,
+      "loss": 0.603,
+      "step": 4396
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1657586097717285,
+      "learning_rate": 4.220902047851729e-06,
+      "loss": 0.6301,
+      "step": 4397
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2193200588226318,
+      "learning_rate": 4.218852432196984e-06,
+      "loss": 0.5981,
+      "step": 4398
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.201629400253296,
+      "learning_rate": 4.21680295107903e-06,
+      "loss": 0.5681,
+      "step": 4399
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4651731252670288,
+      "learning_rate": 4.2147536048508485e-06,
+      "loss": 0.5787,
+      "step": 4400
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9583895206451416,
+      "learning_rate": 4.212704393865398e-06,
+      "loss": 0.521,
+      "step": 4401
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2025455236434937,
+      "learning_rate": 4.210655318475613e-06,
+      "loss": 0.6295,
+      "step": 4402
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3904603719711304,
+      "learning_rate": 4.208606379034405e-06,
+      "loss": 0.584,
+      "step": 4403
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3118081092834473,
+      "learning_rate": 4.206557575894664e-06,
+      "loss": 0.6188,
+      "step": 4404
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2359284162521362,
+      "learning_rate": 4.204508909409253e-06,
+      "loss": 0.5612,
+      "step": 4405
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6139267683029175,
+      "learning_rate": 4.202460379931009e-06,
+      "loss": 0.5957,
+      "step": 4406
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1521854400634766,
+      "learning_rate": 4.200411987812758e-06,
+      "loss": 0.613,
+      "step": 4407
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0886516571044922,
+      "learning_rate": 4.198363733407289e-06,
+      "loss": 0.6028,
+      "step": 4408
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.284321904182434,
+      "learning_rate": 4.196315617067374e-06,
+      "loss": 0.5777,
+      "step": 4409
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2444514036178589,
+      "learning_rate": 4.194267639145758e-06,
+      "loss": 0.5497,
+      "step": 4410
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4131932258605957,
+      "learning_rate": 4.192219799995164e-06,
+      "loss": 0.6017,
+      "step": 4411
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.238851308822632,
+      "learning_rate": 4.190172099968291e-06,
+      "loss": 0.5908,
+      "step": 4412
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.365478754043579,
+      "learning_rate": 4.1881245394178125e-06,
+      "loss": 0.5623,
+      "step": 4413
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6913090944290161,
+      "learning_rate": 4.186077118696381e-06,
+      "loss": 0.6189,
+      "step": 4414
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0305039882659912,
+      "learning_rate": 4.184029838156622e-06,
+      "loss": 0.6424,
+      "step": 4415
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9532804489135742,
+      "learning_rate": 4.181982698151138e-06,
+      "loss": 0.6773,
+      "step": 4416
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6363182067871094,
+      "learning_rate": 4.179935699032507e-06,
+      "loss": 0.5869,
+      "step": 4417
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2440401315689087,
+      "learning_rate": 4.177888841153285e-06,
+      "loss": 0.6201,
+      "step": 4418
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4306551218032837,
+      "learning_rate": 4.175842124865998e-06,
+      "loss": 0.6048,
+      "step": 4419
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1270860433578491,
+      "learning_rate": 4.1737955505231546e-06,
+      "loss": 0.6119,
+      "step": 4420
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3283473253250122,
+      "learning_rate": 4.171749118477234e-06,
+      "loss": 0.549,
+      "step": 4421
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4524391889572144,
+      "learning_rate": 4.1697028290806935e-06,
+      "loss": 0.5849,
+      "step": 4422
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.351607084274292,
+      "learning_rate": 4.167656682685962e-06,
+      "loss": 0.6161,
+      "step": 4423
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2113388776779175,
+      "learning_rate": 4.165610679645451e-06,
+      "loss": 0.5219,
+      "step": 4424
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.863964557647705,
+      "learning_rate": 4.1635648203115405e-06,
+      "loss": 0.5498,
+      "step": 4425
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9458630084991455,
+      "learning_rate": 4.161519105036588e-06,
+      "loss": 0.4855,
+      "step": 4426
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1281429529190063,
+      "learning_rate": 4.159473534172927e-06,
+      "loss": 0.5676,
+      "step": 4427
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1383236646652222,
+      "learning_rate": 4.157428108072866e-06,
+      "loss": 0.6055,
+      "step": 4428
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.510728120803833,
+      "learning_rate": 4.155382827088688e-06,
+      "loss": 0.5496,
+      "step": 4429
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6128290891647339,
+      "learning_rate": 4.1533376915726495e-06,
+      "loss": 0.5808,
+      "step": 4430
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0385468006134033,
+      "learning_rate": 4.151292701876986e-06,
+      "loss": 0.6876,
+      "step": 4431
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1657432317733765,
+      "learning_rate": 4.149247858353902e-06,
+      "loss": 0.5627,
+      "step": 4432
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0777486562728882,
+      "learning_rate": 4.147203161355583e-06,
+      "loss": 0.5524,
+      "step": 4433
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4103708267211914,
+      "learning_rate": 4.145158611234186e-06,
+      "loss": 0.5984,
+      "step": 4434
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.323095679283142,
+      "learning_rate": 4.143114208341843e-06,
+      "loss": 0.7533,
+      "step": 4435
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0617965459823608,
+      "learning_rate": 4.14106995303066e-06,
+      "loss": 0.5806,
+      "step": 4436
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3179047107696533,
+      "learning_rate": 4.1390258456527195e-06,
+      "loss": 0.5494,
+      "step": 4437
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.672276020050049,
+      "learning_rate": 4.136981886560078e-06,
+      "loss": 0.5966,
+      "step": 4438
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2726572751998901,
+      "learning_rate": 4.134938076104764e-06,
+      "loss": 0.6048,
+      "step": 4439
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0546799898147583,
+      "learning_rate": 4.132894414638782e-06,
+      "loss": 0.6254,
+      "step": 4440
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2751909494400024,
+      "learning_rate": 4.130850902514114e-06,
+      "loss": 0.5718,
+      "step": 4441
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2318024635314941,
+      "learning_rate": 4.128807540082714e-06,
+      "loss": 0.5496,
+      "step": 4442
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.266778826713562,
+      "learning_rate": 4.126764327696504e-06,
+      "loss": 0.5568,
+      "step": 4443
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3783042430877686,
+      "learning_rate": 4.124721265707392e-06,
+      "loss": 0.5641,
+      "step": 4444
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1089837551116943,
+      "learning_rate": 4.12267835446725e-06,
+      "loss": 0.6204,
+      "step": 4445
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.281911611557007,
+      "learning_rate": 4.12063559432793e-06,
+      "loss": 0.7028,
+      "step": 4446
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1332265138626099,
+      "learning_rate": 4.118592985641254e-06,
+      "loss": 0.6078,
+      "step": 4447
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2196893692016602,
+      "learning_rate": 4.116550528759023e-06,
+      "loss": 0.5784,
+      "step": 4448
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2832058668136597,
+      "learning_rate": 4.114508224033004e-06,
+      "loss": 0.6051,
+      "step": 4449
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4253997802734375,
+      "learning_rate": 4.112466071814947e-06,
+      "loss": 0.6712,
+      "step": 4450
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.711785912513733,
+      "learning_rate": 4.110424072456568e-06,
+      "loss": 0.6563,
+      "step": 4451
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2868982553482056,
+      "learning_rate": 4.108382226309563e-06,
+      "loss": 0.5904,
+      "step": 4452
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2525250911712646,
+      "learning_rate": 4.106340533725595e-06,
+      "loss": 0.6214,
+      "step": 4453
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.175255537033081,
+      "learning_rate": 4.104298995056307e-06,
+      "loss": 0.5653,
+      "step": 4454
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.409961223602295,
+      "learning_rate": 4.102257610653311e-06,
+      "loss": 0.605,
+      "step": 4455
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.28078293800354,
+      "learning_rate": 4.100216380868194e-06,
+      "loss": 0.5668,
+      "step": 4456
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1241258382797241,
+      "learning_rate": 4.098175306052515e-06,
+      "loss": 0.6365,
+      "step": 4457
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2419236898422241,
+      "learning_rate": 4.09613438655781e-06,
+      "loss": 0.6215,
+      "step": 4458
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4537636041641235,
+      "learning_rate": 4.0940936227355866e-06,
+      "loss": 0.5763,
+      "step": 4459
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1015124320983887,
+      "learning_rate": 4.0920530149373235e-06,
+      "loss": 0.5656,
+      "step": 4460
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.080503225326538,
+      "learning_rate": 4.090012563514473e-06,
+      "loss": 0.606,
+      "step": 4461
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4011242389678955,
+      "learning_rate": 4.087972268818463e-06,
+      "loss": 0.6446,
+      "step": 4462
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2129254341125488,
+      "learning_rate": 4.085932131200691e-06,
+      "loss": 0.5713,
+      "step": 4463
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1942564249038696,
+      "learning_rate": 4.083892151012531e-06,
+      "loss": 0.6235,
+      "step": 4464
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2527570724487305,
+      "learning_rate": 4.081852328605327e-06,
+      "loss": 0.5275,
+      "step": 4465
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1266459226608276,
+      "learning_rate": 4.079812664330398e-06,
+      "loss": 0.6578,
+      "step": 4466
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1888070106506348,
+      "learning_rate": 4.0777731585390335e-06,
+      "loss": 0.5591,
+      "step": 4467
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3004099130630493,
+      "learning_rate": 4.0757338115824975e-06,
+      "loss": 0.5872,
+      "step": 4468
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.164547920227051,
+      "learning_rate": 4.073694623812026e-06,
+      "loss": 0.5892,
+      "step": 4469
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.132218599319458,
+      "learning_rate": 4.071655595578829e-06,
+      "loss": 0.5648,
+      "step": 4470
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.187025547027588,
+      "learning_rate": 4.0696167272340845e-06,
+      "loss": 0.5044,
+      "step": 4471
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.381439208984375,
+      "learning_rate": 4.0675780191289496e-06,
+      "loss": 0.6377,
+      "step": 4472
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3431545495986938,
+      "learning_rate": 4.065539471614547e-06,
+      "loss": 0.5823,
+      "step": 4473
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.775613784790039,
+      "learning_rate": 4.063501085041976e-06,
+      "loss": 0.5882,
+      "step": 4474
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6177005767822266,
+      "learning_rate": 4.06146285976231e-06,
+      "loss": 0.6739,
+      "step": 4475
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1213589906692505,
+      "learning_rate": 4.059424796126589e-06,
+      "loss": 0.6788,
+      "step": 4476
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4037939310073853,
+      "learning_rate": 4.0573868944858306e-06,
+      "loss": 0.5024,
+      "step": 4477
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1332634687423706,
+      "learning_rate": 4.055349155191018e-06,
+      "loss": 0.7279,
+      "step": 4478
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2475149631500244,
+      "learning_rate": 4.0533115785931146e-06,
+      "loss": 0.6189,
+      "step": 4479
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.490715742111206,
+      "learning_rate": 4.051274165043049e-06,
+      "loss": 0.6506,
+      "step": 4480
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3428494930267334,
+      "learning_rate": 4.049236914891726e-06,
+      "loss": 0.56,
+      "step": 4481
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3903584480285645,
+      "learning_rate": 4.047199828490017e-06,
+      "loss": 0.5654,
+      "step": 4482
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2388803958892822,
+      "learning_rate": 4.045162906188773e-06,
+      "loss": 0.5477,
+      "step": 4483
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5961958169937134,
+      "learning_rate": 4.04312614833881e-06,
+      "loss": 0.574,
+      "step": 4484
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4322210550308228,
+      "learning_rate": 4.041089555290919e-06,
+      "loss": 0.5648,
+      "step": 4485
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3006025552749634,
+      "learning_rate": 4.039053127395861e-06,
+      "loss": 0.6208,
+      "step": 4486
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0778475999832153,
+      "learning_rate": 4.037016865004371e-06,
+      "loss": 0.5335,
+      "step": 4487
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3122655153274536,
+      "learning_rate": 4.0349807684671506e-06,
+      "loss": 0.606,
+      "step": 4488
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.462798833847046,
+      "learning_rate": 4.03294483813488e-06,
+      "loss": 0.6424,
+      "step": 4489
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.512628436088562,
+      "learning_rate": 4.030909074358204e-06,
+      "loss": 0.5838,
+      "step": 4490
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.626233696937561,
+      "learning_rate": 4.028873477487741e-06,
+      "loss": 0.6137,
+      "step": 4491
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5814064741134644,
+      "learning_rate": 4.026838047874084e-06,
+      "loss": 0.6584,
+      "step": 4492
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6172866821289062,
+      "learning_rate": 4.024802785867793e-06,
+      "loss": 0.6242,
+      "step": 4493
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1439348459243774,
+      "learning_rate": 4.0227676918194015e-06,
+      "loss": 0.7038,
+      "step": 4494
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3968899250030518,
+      "learning_rate": 4.020732766079411e-06,
+      "loss": 0.6014,
+      "step": 4495
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4990570545196533,
+      "learning_rate": 4.018698008998298e-06,
+      "loss": 0.623,
+      "step": 4496
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8007217645645142,
+      "learning_rate": 4.0166634209265065e-06,
+      "loss": 0.6442,
+      "step": 4497
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3919003009796143,
+      "learning_rate": 4.014629002214454e-06,
+      "loss": 0.46,
+      "step": 4498
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3041036128997803,
+      "learning_rate": 4.012594753212528e-06,
+      "loss": 0.6084,
+      "step": 4499
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.221053123474121,
+      "learning_rate": 4.010560674271085e-06,
+      "loss": 0.5738,
+      "step": 4500
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5155267715454102,
+      "learning_rate": 4.0085267657404544e-06,
+      "loss": 0.7307,
+      "step": 4501
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1297587156295776,
+      "learning_rate": 4.006493027970938e-06,
+      "loss": 0.6133,
+      "step": 4502
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.395195484161377,
+      "learning_rate": 4.004459461312802e-06,
+      "loss": 0.7651,
+      "step": 4503
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0992610454559326,
+      "learning_rate": 4.0024260661162895e-06,
+      "loss": 0.4897,
+      "step": 4504
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1246989965438843,
+      "learning_rate": 4.000392842731611e-06,
+      "loss": 0.5924,
+      "step": 4505
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2189942598342896,
+      "learning_rate": 3.998359791508946e-06,
+      "loss": 0.6192,
+      "step": 4506
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5760741233825684,
+      "learning_rate": 3.996326912798449e-06,
+      "loss": 0.6763,
+      "step": 4507
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6541393995285034,
+      "learning_rate": 3.994294206950241e-06,
+      "loss": 0.6857,
+      "step": 4508
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0716215372085571,
+      "learning_rate": 3.992261674314411e-06,
+      "loss": 0.6769,
+      "step": 4509
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1231220960617065,
+      "learning_rate": 3.990229315241028e-06,
+      "loss": 0.5404,
+      "step": 4510
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6681180000305176,
+      "learning_rate": 3.988197130080121e-06,
+      "loss": 0.5339,
+      "step": 4511
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.169209361076355,
+      "learning_rate": 3.986165119181692e-06,
+      "loss": 0.5946,
+      "step": 4512
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.410796046257019,
+      "learning_rate": 3.9841332828957135e-06,
+      "loss": 0.6796,
+      "step": 4513
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1149036884307861,
+      "learning_rate": 3.98210162157213e-06,
+      "loss": 0.5849,
+      "step": 4514
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4105883836746216,
+      "learning_rate": 3.980070135560852e-06,
+      "loss": 0.5759,
+      "step": 4515
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4017486572265625,
+      "learning_rate": 3.978038825211763e-06,
+      "loss": 0.5776,
+      "step": 4516
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.292477011680603,
+      "learning_rate": 3.9760076908747134e-06,
+      "loss": 0.7173,
+      "step": 4517
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4200750589370728,
+      "learning_rate": 3.973976732899526e-06,
+      "loss": 0.5984,
+      "step": 4518
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.526354432106018,
+      "learning_rate": 3.971945951635992e-06,
+      "loss": 0.5828,
+      "step": 4519
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.998432457447052,
+      "learning_rate": 3.969915347433871e-06,
+      "loss": 0.5829,
+      "step": 4520
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.682458400726318,
+      "learning_rate": 3.967884920642895e-06,
+      "loss": 0.6694,
+      "step": 4521
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5824471712112427,
+      "learning_rate": 3.965854671612762e-06,
+      "loss": 0.5997,
+      "step": 4522
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 6.456887245178223,
+      "learning_rate": 3.963824600693143e-06,
+      "loss": 0.6215,
+      "step": 4523
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.255520224571228,
+      "learning_rate": 3.961794708233675e-06,
+      "loss": 0.7175,
+      "step": 4524
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1447798013687134,
+      "learning_rate": 3.959764994583965e-06,
+      "loss": 0.5813,
+      "step": 4525
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.238195538520813,
+      "learning_rate": 3.957735460093591e-06,
+      "loss": 0.5689,
+      "step": 4526
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5053225755691528,
+      "learning_rate": 3.955706105112101e-06,
+      "loss": 0.591,
+      "step": 4527
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2886848449707031,
+      "learning_rate": 3.953676929989008e-06,
+      "loss": 0.5764,
+      "step": 4528
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.453669786453247,
+      "learning_rate": 3.951647935073796e-06,
+      "loss": 0.6344,
+      "step": 4529
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3581910133361816,
+      "learning_rate": 3.949619120715918e-06,
+      "loss": 0.5262,
+      "step": 4530
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4336278438568115,
+      "learning_rate": 3.947590487264799e-06,
+      "loss": 0.5272,
+      "step": 4531
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4319508075714111,
+      "learning_rate": 3.945562035069826e-06,
+      "loss": 0.6207,
+      "step": 4532
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5566030740737915,
+      "learning_rate": 3.9435337644803605e-06,
+      "loss": 0.6633,
+      "step": 4533
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4970437288284302,
+      "learning_rate": 3.94150567584573e-06,
+      "loss": 0.6113,
+      "step": 4534
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2610963582992554,
+      "learning_rate": 3.9394777695152335e-06,
+      "loss": 0.6422,
+      "step": 4535
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2423081398010254,
+      "learning_rate": 3.937450045838133e-06,
+      "loss": 0.6177,
+      "step": 4536
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.597859501838684,
+      "learning_rate": 3.935422505163667e-06,
+      "loss": 0.6175,
+      "step": 4537
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2870022058486938,
+      "learning_rate": 3.933395147841035e-06,
+      "loss": 0.5798,
+      "step": 4538
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2589836120605469,
+      "learning_rate": 3.931367974219411e-06,
+      "loss": 0.5974,
+      "step": 4539
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2614821195602417,
+      "learning_rate": 3.9293409846479305e-06,
+      "loss": 0.6749,
+      "step": 4540
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2686032056808472,
+      "learning_rate": 3.927314179475705e-06,
+      "loss": 0.5984,
+      "step": 4541
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2367477416992188,
+      "learning_rate": 3.925287559051808e-06,
+      "loss": 0.6273,
+      "step": 4542
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1769745349884033,
+      "learning_rate": 3.923261123725283e-06,
+      "loss": 0.5428,
+      "step": 4543
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3887510299682617,
+      "learning_rate": 3.921234873845146e-06,
+      "loss": 0.5769,
+      "step": 4544
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.007226586341858,
+      "learning_rate": 3.9192088097603745e-06,
+      "loss": 0.6352,
+      "step": 4545
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2268894910812378,
+      "learning_rate": 3.917182931819918e-06,
+      "loss": 0.6749,
+      "step": 4546
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3286808729171753,
+      "learning_rate": 3.915157240372693e-06,
+      "loss": 0.5088,
+      "step": 4547
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3561018705368042,
+      "learning_rate": 3.913131735767583e-06,
+      "loss": 0.5868,
+      "step": 4548
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3705759048461914,
+      "learning_rate": 3.911106418353439e-06,
+      "loss": 0.6115,
+      "step": 4549
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.358817458152771,
+      "learning_rate": 3.909081288479083e-06,
+      "loss": 0.5725,
+      "step": 4550
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.503664493560791,
+      "learning_rate": 3.907056346493301e-06,
+      "loss": 0.6032,
+      "step": 4551
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1593071222305298,
+      "learning_rate": 3.905031592744849e-06,
+      "loss": 0.629,
+      "step": 4552
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9405057430267334,
+      "learning_rate": 3.9030070275824486e-06,
+      "loss": 0.5977,
+      "step": 4553
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4616215229034424,
+      "learning_rate": 3.90098265135479e-06,
+      "loss": 0.6383,
+      "step": 4554
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.344404697418213,
+      "learning_rate": 3.898958464410532e-06,
+      "loss": 0.6543,
+      "step": 4555
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2103294134140015,
+      "learning_rate": 3.896934467098298e-06,
+      "loss": 0.598,
+      "step": 4556
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1870449781417847,
+      "learning_rate": 3.894910659766682e-06,
+      "loss": 0.6187,
+      "step": 4557
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.1117770671844482,
+      "learning_rate": 3.892887042764243e-06,
+      "loss": 0.6356,
+      "step": 4558
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8922291994094849,
+      "learning_rate": 3.890863616439509e-06,
+      "loss": 0.5096,
+      "step": 4559
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3100972175598145,
+      "learning_rate": 3.888840381140971e-06,
+      "loss": 0.6282,
+      "step": 4560
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8961695432662964,
+      "learning_rate": 3.886817337217092e-06,
+      "loss": 0.5745,
+      "step": 4561
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7968751192092896,
+      "learning_rate": 3.884794485016302e-06,
+      "loss": 0.4665,
+      "step": 4562
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0768951177597046,
+      "learning_rate": 3.882771824886994e-06,
+      "loss": 0.583,
+      "step": 4563
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.303018569946289,
+      "learning_rate": 3.8807493571775315e-06,
+      "loss": 0.5878,
+      "step": 4564
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1549627780914307,
+      "learning_rate": 3.878727082236241e-06,
+      "loss": 0.6445,
+      "step": 4565
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1280014514923096,
+      "learning_rate": 3.876705000411422e-06,
+      "loss": 0.627,
+      "step": 4566
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0545580387115479,
+      "learning_rate": 3.874683112051333e-06,
+      "loss": 0.6427,
+      "step": 4567
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2090257406234741,
+      "learning_rate": 3.872661417504207e-06,
+      "loss": 0.7049,
+      "step": 4568
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2235411405563354,
+      "learning_rate": 3.870639917118235e-06,
+      "loss": 0.7095,
+      "step": 4569
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0587208271026611,
+      "learning_rate": 3.868618611241584e-06,
+      "loss": 0.5291,
+      "step": 4570
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2914259433746338,
+      "learning_rate": 3.866597500222381e-06,
+      "loss": 0.5862,
+      "step": 4571
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3580983877182007,
+      "learning_rate": 3.864576584408722e-06,
+      "loss": 0.515,
+      "step": 4572
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3612672090530396,
+      "learning_rate": 3.862555864148666e-06,
+      "loss": 0.6386,
+      "step": 4573
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3382236957550049,
+      "learning_rate": 3.860535339790245e-06,
+      "loss": 0.5627,
+      "step": 4574
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3669838905334473,
+      "learning_rate": 3.85851501168145e-06,
+      "loss": 0.6867,
+      "step": 4575
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.110496997833252,
+      "learning_rate": 3.856494880170243e-06,
+      "loss": 0.5651,
+      "step": 4576
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1593047380447388,
+      "learning_rate": 3.854474945604549e-06,
+      "loss": 0.4934,
+      "step": 4577
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3233683109283447,
+      "learning_rate": 3.852455208332262e-06,
+      "loss": 0.6082,
+      "step": 4578
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.193565011024475,
+      "learning_rate": 3.850435668701243e-06,
+      "loss": 0.531,
+      "step": 4579
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1400476694107056,
+      "learning_rate": 3.8484163270593125e-06,
+      "loss": 0.5821,
+      "step": 4580
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.239013910293579,
+      "learning_rate": 3.846397183754265e-06,
+      "loss": 0.5321,
+      "step": 4581
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.63629150390625,
+      "learning_rate": 3.844378239133852e-06,
+      "loss": 0.5889,
+      "step": 4582
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2488521337509155,
+      "learning_rate": 3.8423594935458e-06,
+      "loss": 0.6494,
+      "step": 4583
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3007025718688965,
+      "learning_rate": 3.840340947337795e-06,
+      "loss": 0.6275,
+      "step": 4584
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5868581533432007,
+      "learning_rate": 3.838322600857491e-06,
+      "loss": 0.6701,
+      "step": 4585
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9399994015693665,
+      "learning_rate": 3.8363044544525065e-06,
+      "loss": 0.5922,
+      "step": 4586
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3345903158187866,
+      "learning_rate": 3.834286508470428e-06,
+      "loss": 0.6308,
+      "step": 4587
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9581162929534912,
+      "learning_rate": 3.832268763258803e-06,
+      "loss": 0.6132,
+      "step": 4588
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2127094268798828,
+      "learning_rate": 3.83025121916515e-06,
+      "loss": 0.6929,
+      "step": 4589
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5069866180419922,
+      "learning_rate": 3.8282338765369466e-06,
+      "loss": 0.5396,
+      "step": 4590
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1287541389465332,
+      "learning_rate": 3.8262167357216426e-06,
+      "loss": 0.6039,
+      "step": 4591
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0839670896530151,
+      "learning_rate": 3.824199797066646e-06,
+      "loss": 0.6734,
+      "step": 4592
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1110575199127197,
+      "learning_rate": 3.822183060919337e-06,
+      "loss": 0.6552,
+      "step": 4593
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0983967781066895,
+      "learning_rate": 3.820166527627054e-06,
+      "loss": 0.5722,
+      "step": 4594
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3963110446929932,
+      "learning_rate": 3.818150197537106e-06,
+      "loss": 0.6369,
+      "step": 4595
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2948497533798218,
+      "learning_rate": 3.816134070996766e-06,
+      "loss": 0.5612,
+      "step": 4596
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1768401861190796,
+      "learning_rate": 3.8141181483532676e-06,
+      "loss": 0.586,
+      "step": 4597
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5991896390914917,
+      "learning_rate": 3.8121024299538156e-06,
+      "loss": 0.6508,
+      "step": 4598
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1931565999984741,
+      "learning_rate": 3.8100869161455746e-06,
+      "loss": 0.5943,
+      "step": 4599
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2148854732513428,
+      "learning_rate": 3.8080716072756767e-06,
+      "loss": 0.648,
+      "step": 4600
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4304548501968384,
+      "learning_rate": 3.8060565036912167e-06,
+      "loss": 0.674,
+      "step": 4601
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.529685616493225,
+      "learning_rate": 3.8040416057392577e-06,
+      "loss": 0.5341,
+      "step": 4602
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5385857820510864,
+      "learning_rate": 3.802026913766823e-06,
+      "loss": 0.678,
+      "step": 4603
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.180269479751587,
+      "learning_rate": 3.8000124281209015e-06,
+      "loss": 0.5744,
+      "step": 4604
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1694934368133545,
+      "learning_rate": 3.7979981491484496e-06,
+      "loss": 0.5707,
+      "step": 4605
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2115732431411743,
+      "learning_rate": 3.795984077196384e-06,
+      "loss": 0.5016,
+      "step": 4606
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5277433395385742,
+      "learning_rate": 3.7939702126115895e-06,
+      "loss": 0.6791,
+      "step": 4607
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4278790950775146,
+      "learning_rate": 3.7919565557409115e-06,
+      "loss": 0.6131,
+      "step": 4608
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4480127096176147,
+      "learning_rate": 3.789943106931164e-06,
+      "loss": 0.532,
+      "step": 4609
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3539552688598633,
+      "learning_rate": 3.7879298665291194e-06,
+      "loss": 0.5845,
+      "step": 4610
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.717895030975342,
+      "learning_rate": 3.7859168348815177e-06,
+      "loss": 0.5879,
+      "step": 4611
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.146493911743164,
+      "learning_rate": 3.7839040123350664e-06,
+      "loss": 0.7074,
+      "step": 4612
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0468714237213135,
+      "learning_rate": 3.7818913992364298e-06,
+      "loss": 0.4987,
+      "step": 4613
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3573884963989258,
+      "learning_rate": 3.7798789959322417e-06,
+      "loss": 0.7328,
+      "step": 4614
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.123680591583252,
+      "learning_rate": 3.7778668027690957e-06,
+      "loss": 0.603,
+      "step": 4615
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4465726613998413,
+      "learning_rate": 3.7758548200935537e-06,
+      "loss": 0.5605,
+      "step": 4616
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3835780620574951,
+      "learning_rate": 3.7738430482521355e-06,
+      "loss": 0.6137,
+      "step": 4617
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2412621974945068,
+      "learning_rate": 3.771831487591331e-06,
+      "loss": 0.5755,
+      "step": 4618
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3044785261154175,
+      "learning_rate": 3.7698201384575883e-06,
+      "loss": 0.6081,
+      "step": 4619
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.197967290878296,
+      "learning_rate": 3.767809001197323e-06,
+      "loss": 0.6496,
+      "step": 4620
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3317675590515137,
+      "learning_rate": 3.7657980761569114e-06,
+      "loss": 0.6448,
+      "step": 4621
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4491106271743774,
+      "learning_rate": 3.763787363682696e-06,
+      "loss": 0.7453,
+      "step": 4622
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.172399640083313,
+      "learning_rate": 3.7617768641209797e-06,
+      "loss": 0.5975,
+      "step": 4623
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3938887119293213,
+      "learning_rate": 3.7597665778180307e-06,
+      "loss": 0.6943,
+      "step": 4624
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1383371353149414,
+      "learning_rate": 3.75775650512008e-06,
+      "loss": 0.5426,
+      "step": 4625
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.595054268836975,
+      "learning_rate": 3.755746646373322e-06,
+      "loss": 0.6339,
+      "step": 4626
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3873826265335083,
+      "learning_rate": 3.7537370019239135e-06,
+      "loss": 0.597,
+      "step": 4627
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1629124879837036,
+      "learning_rate": 3.7517275721179736e-06,
+      "loss": 0.6172,
+      "step": 4628
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3999043703079224,
+      "learning_rate": 3.7497183573015893e-06,
+      "loss": 0.5355,
+      "step": 4629
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1866925954818726,
+      "learning_rate": 3.7477093578208047e-06,
+      "loss": 0.5889,
+      "step": 4630
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3769731521606445,
+      "learning_rate": 3.74570057402163e-06,
+      "loss": 0.5985,
+      "step": 4631
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4890729188919067,
+      "learning_rate": 3.743692006250036e-06,
+      "loss": 0.5906,
+      "step": 4632
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5705612897872925,
+      "learning_rate": 3.741683654851959e-06,
+      "loss": 0.6187,
+      "step": 4633
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6368554830551147,
+      "learning_rate": 3.739675520173296e-06,
+      "loss": 0.6418,
+      "step": 4634
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1601102352142334,
+      "learning_rate": 3.737667602559908e-06,
+      "loss": 0.5984,
+      "step": 4635
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.6461944580078125,
+      "learning_rate": 3.7356599023576166e-06,
+      "loss": 0.5892,
+      "step": 4636
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2127076387405396,
+      "learning_rate": 3.7336524199122094e-06,
+      "loss": 0.5839,
+      "step": 4637
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.202880859375,
+      "learning_rate": 3.7316451555694327e-06,
+      "loss": 0.6829,
+      "step": 4638
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4662011861801147,
+      "learning_rate": 3.7296381096749983e-06,
+      "loss": 0.5418,
+      "step": 4639
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1767808198928833,
+      "learning_rate": 3.7276312825745775e-06,
+      "loss": 0.5559,
+      "step": 4640
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2574052810668945,
+      "learning_rate": 3.7256246746138082e-06,
+      "loss": 0.6815,
+      "step": 4641
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4866012334823608,
+      "learning_rate": 3.7236182861382843e-06,
+      "loss": 0.6487,
+      "step": 4642
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2232147455215454,
+      "learning_rate": 3.721612117493568e-06,
+      "loss": 0.6718,
+      "step": 4643
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3642147779464722,
+      "learning_rate": 3.7196061690251795e-06,
+      "loss": 0.5534,
+      "step": 4644
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.121066927909851,
+      "learning_rate": 3.7176004410786047e-06,
+      "loss": 0.5487,
+      "step": 4645
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6986517906188965,
+      "learning_rate": 3.7155949339992856e-06,
+      "loss": 0.7287,
+      "step": 4646
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4121288061141968,
+      "learning_rate": 3.713589648132634e-06,
+      "loss": 0.5755,
+      "step": 4647
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3280876874923706,
+      "learning_rate": 3.7115845838240193e-06,
+      "loss": 0.5657,
+      "step": 4648
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3368446826934814,
+      "learning_rate": 3.7095797414187707e-06,
+      "loss": 0.6326,
+      "step": 4649
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0597262382507324,
+      "learning_rate": 3.707575121262185e-06,
+      "loss": 0.5701,
+      "step": 4650
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1828985214233398,
+      "learning_rate": 3.7055707236995123e-06,
+      "loss": 0.5832,
+      "step": 4651
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4886174201965332,
+      "learning_rate": 3.7035665490759743e-06,
+      "loss": 0.6071,
+      "step": 4652
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7236698865890503,
+      "learning_rate": 3.7015625977367476e-06,
+      "loss": 0.6029,
+      "step": 4653
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.32156240940094,
+      "learning_rate": 3.6995588700269697e-06,
+      "loss": 0.5823,
+      "step": 4654
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.207321047782898,
+      "learning_rate": 3.6975553662917453e-06,
+      "loss": 0.5992,
+      "step": 4655
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2201660871505737,
+      "learning_rate": 3.695552086876135e-06,
+      "loss": 0.4492,
+      "step": 4656
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4512239694595337,
+      "learning_rate": 3.6935490321251655e-06,
+      "loss": 0.6295,
+      "step": 4657
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6261647939682007,
+      "learning_rate": 3.691546202383819e-06,
+      "loss": 0.5875,
+      "step": 4658
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0876991748809814,
+      "learning_rate": 3.689543597997044e-06,
+      "loss": 0.5359,
+      "step": 4659
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2004252672195435,
+      "learning_rate": 3.687541219309748e-06,
+      "loss": 0.5935,
+      "step": 4660
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3604297637939453,
+      "learning_rate": 3.685539066666802e-06,
+      "loss": 0.5974,
+      "step": 4661
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2416634559631348,
+      "learning_rate": 3.683537140413032e-06,
+      "loss": 0.6314,
+      "step": 4662
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1395695209503174,
+      "learning_rate": 3.6815354408932314e-06,
+      "loss": 0.6032,
+      "step": 4663
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3960171937942505,
+      "learning_rate": 3.6795339684521535e-06,
+      "loss": 0.6214,
+      "step": 4664
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3174827098846436,
+      "learning_rate": 3.67753272343451e-06,
+      "loss": 0.6025,
+      "step": 4665
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3568823337554932,
+      "learning_rate": 3.675531706184975e-06,
+      "loss": 0.6243,
+      "step": 4666
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2531613111495972,
+      "learning_rate": 3.6735309170481825e-06,
+      "loss": 0.588,
+      "step": 4667
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1361048221588135,
+      "learning_rate": 3.6715303563687286e-06,
+      "loss": 0.6366,
+      "step": 4668
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0308409929275513,
+      "learning_rate": 3.6695300244911676e-06,
+      "loss": 0.485,
+      "step": 4669
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2207977771759033,
+      "learning_rate": 3.667529921760018e-06,
+      "loss": 0.6058,
+      "step": 4670
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.139939546585083,
+      "learning_rate": 3.6655300485197556e-06,
+      "loss": 0.5526,
+      "step": 4671
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3506957292556763,
+      "learning_rate": 3.663530405114818e-06,
+      "loss": 0.6159,
+      "step": 4672
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.240322232246399,
+      "learning_rate": 3.6615309918896034e-06,
+      "loss": 0.5767,
+      "step": 4673
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.595827579498291,
+      "learning_rate": 3.6595318091884707e-06,
+      "loss": 0.6388,
+      "step": 4674
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5169620513916016,
+      "learning_rate": 3.6575328573557367e-06,
+      "loss": 0.6659,
+      "step": 4675
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1293915510177612,
+      "learning_rate": 3.655534136735682e-06,
+      "loss": 0.5856,
+      "step": 4676
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3506858348846436,
+      "learning_rate": 3.6535356476725447e-06,
+      "loss": 0.6416,
+      "step": 4677
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2166670560836792,
+      "learning_rate": 3.6515373905105254e-06,
+      "loss": 0.6027,
+      "step": 4678
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.4438283443450928,
+      "learning_rate": 3.6495393655937806e-06,
+      "loss": 0.6248,
+      "step": 4679
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3515195846557617,
+      "learning_rate": 3.6475415732664297e-06,
+      "loss": 0.5362,
+      "step": 4680
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5036596059799194,
+      "learning_rate": 3.6455440138725553e-06,
+      "loss": 0.6032,
+      "step": 4681
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2506636381149292,
+      "learning_rate": 3.6435466877561933e-06,
+      "loss": 0.6498,
+      "step": 4682
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2424179315567017,
+      "learning_rate": 3.6415495952613446e-06,
+      "loss": 0.6529,
+      "step": 4683
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3675919771194458,
+      "learning_rate": 3.639552736731965e-06,
+      "loss": 0.5896,
+      "step": 4684
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2366334199905396,
+      "learning_rate": 3.6375561125119752e-06,
+      "loss": 0.7092,
+      "step": 4685
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2184380292892456,
+      "learning_rate": 3.635559722945252e-06,
+      "loss": 0.6131,
+      "step": 4686
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.219530701637268,
+      "learning_rate": 3.6335635683756343e-06,
+      "loss": 0.5639,
+      "step": 4687
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.363353967666626,
+      "learning_rate": 3.6315676491469165e-06,
+      "loss": 0.5838,
+      "step": 4688
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2268239259719849,
+      "learning_rate": 3.629571965602858e-06,
+      "loss": 0.5609,
+      "step": 4689
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1857258081436157,
+      "learning_rate": 3.6275765180871723e-06,
+      "loss": 0.6544,
+      "step": 4690
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2621301412582397,
+      "learning_rate": 3.625581306943537e-06,
+      "loss": 0.5046,
+      "step": 4691
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.828747034072876,
+      "learning_rate": 3.623586332515584e-06,
+      "loss": 0.6152,
+      "step": 4692
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2840453386306763,
+      "learning_rate": 3.6215915951469105e-06,
+      "loss": 0.7242,
+      "step": 4693
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.8199918270111084,
+      "learning_rate": 3.6195970951810653e-06,
+      "loss": 0.5506,
+      "step": 4694
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6158698797225952,
+      "learning_rate": 3.6176028329615654e-06,
+      "loss": 0.6238,
+      "step": 4695
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.960234522819519,
+      "learning_rate": 3.615608808831877e-06,
+      "loss": 0.6093,
+      "step": 4696
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1626479625701904,
+      "learning_rate": 3.6136150231354317e-06,
+      "loss": 0.6119,
+      "step": 4697
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4083420038223267,
+      "learning_rate": 3.611621476215621e-06,
+      "loss": 0.5713,
+      "step": 4698
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3749487400054932,
+      "learning_rate": 3.609628168415791e-06,
+      "loss": 0.6795,
+      "step": 4699
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4019120931625366,
+      "learning_rate": 3.6076351000792487e-06,
+      "loss": 0.643,
+      "step": 4700
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2229111194610596,
+      "learning_rate": 3.60564227154926e-06,
+      "loss": 0.6225,
+      "step": 4701
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.121282696723938,
+      "learning_rate": 3.6036496831690483e-06,
+      "loss": 0.5681,
+      "step": 4702
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2437260150909424,
+      "learning_rate": 3.601657335281797e-06,
+      "loss": 0.6019,
+      "step": 4703
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2536563873291016,
+      "learning_rate": 3.5996652282306467e-06,
+      "loss": 0.6497,
+      "step": 4704
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2499635219573975,
+      "learning_rate": 3.5976733623586986e-06,
+      "loss": 0.574,
+      "step": 4705
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0934405326843262,
+      "learning_rate": 3.5956817380090092e-06,
+      "loss": 0.617,
+      "step": 4706
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4827922582626343,
+      "learning_rate": 3.5936903555245983e-06,
+      "loss": 0.6322,
+      "step": 4707
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.30734121799469,
+      "learning_rate": 3.5916992152484382e-06,
+      "loss": 0.6527,
+      "step": 4708
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2423964738845825,
+      "learning_rate": 3.5897083175234644e-06,
+      "loss": 0.632,
+      "step": 4709
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3838889598846436,
+      "learning_rate": 3.587717662692567e-06,
+      "loss": 0.6033,
+      "step": 4710
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2591181993484497,
+      "learning_rate": 3.5857272510985964e-06,
+      "loss": 0.6277,
+      "step": 4711
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1490346193313599,
+      "learning_rate": 3.583737083084361e-06,
+      "loss": 0.5655,
+      "step": 4712
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.445289134979248,
+      "learning_rate": 3.5817471589926266e-06,
+      "loss": 0.6682,
+      "step": 4713
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6664557456970215,
+      "learning_rate": 3.579757479166115e-06,
+      "loss": 0.5634,
+      "step": 4714
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.124857783317566,
+      "learning_rate": 3.5777680439475115e-06,
+      "loss": 0.7383,
+      "step": 4715
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.228276014328003,
+      "learning_rate": 3.5757788536794547e-06,
+      "loss": 0.5617,
+      "step": 4716
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1751571893692017,
+      "learning_rate": 3.573789908704542e-06,
+      "loss": 0.5534,
+      "step": 4717
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3536722660064697,
+      "learning_rate": 3.5718012093653294e-06,
+      "loss": 0.568,
+      "step": 4718
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.338544249534607,
+      "learning_rate": 3.5698127560043277e-06,
+      "loss": 0.554,
+      "step": 4719
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0579848289489746,
+      "learning_rate": 3.567824548964011e-06,
+      "loss": 0.4984,
+      "step": 4720
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7330485582351685,
+      "learning_rate": 3.565836588586804e-06,
+      "loss": 0.6076,
+      "step": 4721
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2623368501663208,
+      "learning_rate": 3.5638488752150956e-06,
+      "loss": 0.6481,
+      "step": 4722
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1102206707000732,
+      "learning_rate": 3.5618614091912264e-06,
+      "loss": 0.5571,
+      "step": 4723
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2276370525360107,
+      "learning_rate": 3.559874190857499e-06,
+      "loss": 0.7092,
+      "step": 4724
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.190426230430603,
+      "learning_rate": 3.55788722055617e-06,
+      "loss": 0.6076,
+      "step": 4725
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1113755702972412,
+      "learning_rate": 3.555900498629456e-06,
+      "loss": 0.6253,
+      "step": 4726
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.146645426750183,
+      "learning_rate": 3.5539140254195277e-06,
+      "loss": 0.5932,
+      "step": 4727
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3581435680389404,
+      "learning_rate": 3.5519278012685164e-06,
+      "loss": 0.6639,
+      "step": 4728
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.19074285030365,
+      "learning_rate": 3.5499418265185082e-06,
+      "loss": 0.6644,
+      "step": 4729
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2494276762008667,
+      "learning_rate": 3.547956101511547e-06,
+      "loss": 0.5994,
+      "step": 4730
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3655177354812622,
+      "learning_rate": 3.545970626589631e-06,
+      "loss": 0.6564,
+      "step": 4731
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3489915132522583,
+      "learning_rate": 3.5439854020947216e-06,
+      "loss": 0.7015,
+      "step": 4732
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1399732828140259,
+      "learning_rate": 3.5420004283687338e-06,
+      "loss": 0.594,
+      "step": 4733
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2338383197784424,
+      "learning_rate": 3.5400157057535354e-06,
+      "loss": 0.6912,
+      "step": 4734
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1922136545181274,
+      "learning_rate": 3.5380312345909582e-06,
+      "loss": 0.5775,
+      "step": 4735
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2153799533843994,
+      "learning_rate": 3.5360470152227846e-06,
+      "loss": 0.5943,
+      "step": 4736
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3712053298950195,
+      "learning_rate": 3.5340630479907567e-06,
+      "loss": 0.6236,
+      "step": 4737
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0961687564849854,
+      "learning_rate": 3.532079333236571e-06,
+      "loss": 0.5619,
+      "step": 4738
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1054195165634155,
+      "learning_rate": 3.5300958713018858e-06,
+      "loss": 0.5613,
+      "step": 4739
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.196414589881897,
+      "learning_rate": 3.528112662528308e-06,
+      "loss": 0.6143,
+      "step": 4740
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1598870754241943,
+      "learning_rate": 3.5261297072574085e-06,
+      "loss": 0.6768,
+      "step": 4741
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3299384117126465,
+      "learning_rate": 3.524147005830708e-06,
+      "loss": 0.6039,
+      "step": 4742
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8477388620376587,
+      "learning_rate": 3.522164558589689e-06,
+      "loss": 0.5782,
+      "step": 4743
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4437540769577026,
+      "learning_rate": 3.5201823658757855e-06,
+      "loss": 0.6431,
+      "step": 4744
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.136198878288269,
+      "learning_rate": 3.5182004280303927e-06,
+      "loss": 0.5964,
+      "step": 4745
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2844723463058472,
+      "learning_rate": 3.516218745394857e-06,
+      "loss": 0.5537,
+      "step": 4746
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1594425439834595,
+      "learning_rate": 3.5142373183104823e-06,
+      "loss": 0.7255,
+      "step": 4747
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6344101428985596,
+      "learning_rate": 3.5122561471185292e-06,
+      "loss": 0.5981,
+      "step": 4748
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.394809603691101,
+      "learning_rate": 3.5102752321602163e-06,
+      "loss": 0.5609,
+      "step": 4749
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0609498023986816,
+      "learning_rate": 3.508294573776716e-06,
+      "loss": 0.5798,
+      "step": 4750
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3028736114501953,
+      "learning_rate": 3.5063141723091555e-06,
+      "loss": 0.5482,
+      "step": 4751
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1946570873260498,
+      "learning_rate": 3.504334028098617e-06,
+      "loss": 0.5992,
+      "step": 4752
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6412652730941772,
+      "learning_rate": 3.502354141486143e-06,
+      "loss": 0.6044,
+      "step": 4753
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.494856119155884,
+      "learning_rate": 3.5003745128127263e-06,
+      "loss": 0.5425,
+      "step": 4754
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1414670944213867,
+      "learning_rate": 3.4983951424193196e-06,
+      "loss": 0.756,
+      "step": 4755
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2341411113739014,
+      "learning_rate": 3.4964160306468274e-06,
+      "loss": 0.5684,
+      "step": 4756
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4616705179214478,
+      "learning_rate": 3.4944371778361143e-06,
+      "loss": 0.5613,
+      "step": 4757
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.233243465423584,
+      "learning_rate": 3.4924585843279933e-06,
+      "loss": 0.6797,
+      "step": 4758
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.328782081604004,
+      "learning_rate": 3.4904802504632406e-06,
+      "loss": 0.6308,
+      "step": 4759
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.412166714668274,
+      "learning_rate": 3.488502176582582e-06,
+      "loss": 0.5519,
+      "step": 4760
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.19275963306427,
+      "learning_rate": 3.4865243630267022e-06,
+      "loss": 0.5534,
+      "step": 4761
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3603435754776,
+      "learning_rate": 3.484546810136237e-06,
+      "loss": 0.71,
+      "step": 4762
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1178488731384277,
+      "learning_rate": 3.4825695182517823e-06,
+      "loss": 0.6962,
+      "step": 4763
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.11331307888031,
+      "learning_rate": 3.4805924877138837e-06,
+      "loss": 0.6168,
+      "step": 4764
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4276987314224243,
+      "learning_rate": 3.478615718863047e-06,
+      "loss": 0.6107,
+      "step": 4765
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1094125509262085,
+      "learning_rate": 3.4766392120397274e-06,
+      "loss": 0.5935,
+      "step": 4766
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.393486738204956,
+      "learning_rate": 3.4746629675843413e-06,
+      "loss": 0.6608,
+      "step": 4767
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3646267652511597,
+      "learning_rate": 3.472686985837257e-06,
+      "loss": 0.7118,
+      "step": 4768
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2422285079956055,
+      "learning_rate": 3.470711267138794e-06,
+      "loss": 0.5005,
+      "step": 4769
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4949270486831665,
+      "learning_rate": 3.4687358118292325e-06,
+      "loss": 0.5937,
+      "step": 4770
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2646280527114868,
+      "learning_rate": 3.4667606202488014e-06,
+      "loss": 0.5595,
+      "step": 4771
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8542555570602417,
+      "learning_rate": 3.464785692737691e-06,
+      "loss": 0.6481,
+      "step": 4772
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4289233684539795,
+      "learning_rate": 3.4628110296360394e-06,
+      "loss": 0.6218,
+      "step": 4773
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.189062237739563,
+      "learning_rate": 3.460836631283945e-06,
+      "loss": 0.5669,
+      "step": 4774
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6110448837280273,
+      "learning_rate": 3.4588624980214547e-06,
+      "loss": 0.6313,
+      "step": 4775
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2245473861694336,
+      "learning_rate": 3.4568886301885753e-06,
+      "loss": 0.651,
+      "step": 4776
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3276503086090088,
+      "learning_rate": 3.4549150281252635e-06,
+      "loss": 0.6339,
+      "step": 4777
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3305314779281616,
+      "learning_rate": 3.4529416921714344e-06,
+      "loss": 0.6083,
+      "step": 4778
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6480473279953003,
+      "learning_rate": 3.450968622666952e-06,
+      "loss": 0.6153,
+      "step": 4779
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2591640949249268,
+      "learning_rate": 3.4489958199516404e-06,
+      "loss": 0.7308,
+      "step": 4780
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.382493019104004,
+      "learning_rate": 3.4470232843652728e-06,
+      "loss": 0.6487,
+      "step": 4781
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3461976051330566,
+      "learning_rate": 3.4450510162475797e-06,
+      "loss": 0.664,
+      "step": 4782
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8919711112976074,
+      "learning_rate": 3.4430790159382414e-06,
+      "loss": 0.6032,
+      "step": 4783
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4023423194885254,
+      "learning_rate": 3.441107283776899e-06,
+      "loss": 0.6103,
+      "step": 4784
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.211699366569519,
+      "learning_rate": 3.4391358201031412e-06,
+      "loss": 0.5951,
+      "step": 4785
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2939444780349731,
+      "learning_rate": 3.4371646252565114e-06,
+      "loss": 0.6142,
+      "step": 4786
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2708574533462524,
+      "learning_rate": 3.4351936995765112e-06,
+      "loss": 0.5485,
+      "step": 4787
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2561215162277222,
+      "learning_rate": 3.4332230434025887e-06,
+      "loss": 0.5625,
+      "step": 4788
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0576465129852295,
+      "learning_rate": 3.4312526570741524e-06,
+      "loss": 0.693,
+      "step": 4789
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.220123052597046,
+      "learning_rate": 3.4292825409305586e-06,
+      "loss": 0.641,
+      "step": 4790
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1228139400482178,
+      "learning_rate": 3.4273126953111226e-06,
+      "loss": 0.7457,
+      "step": 4791
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.302809715270996,
+      "learning_rate": 3.425343120555107e-06,
+      "loss": 0.6077,
+      "step": 4792
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.169256567955017,
+      "learning_rate": 3.4233738170017338e-06,
+      "loss": 0.6243,
+      "step": 4793
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4685825109481812,
+      "learning_rate": 3.4214047849901743e-06,
+      "loss": 0.566,
+      "step": 4794
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.632744550704956,
+      "learning_rate": 3.4194360248595547e-06,
+      "loss": 0.5817,
+      "step": 4795
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2418441772460938,
+      "learning_rate": 3.417467536948954e-06,
+      "loss": 0.5698,
+      "step": 4796
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.720788598060608,
+      "learning_rate": 3.415499321597403e-06,
+      "loss": 0.6484,
+      "step": 4797
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3634127378463745,
+      "learning_rate": 3.4135313791438885e-06,
+      "loss": 0.7072,
+      "step": 4798
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0684276819229126,
+      "learning_rate": 3.411563709927347e-06,
+      "loss": 0.6948,
+      "step": 4799
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3760558366775513,
+      "learning_rate": 3.409596314286669e-06,
+      "loss": 0.5563,
+      "step": 4800
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.910668969154358,
+      "learning_rate": 3.4076291925607017e-06,
+      "loss": 0.692,
+      "step": 4801
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1838176250457764,
+      "learning_rate": 3.4056623450882388e-06,
+      "loss": 0.6427,
+      "step": 4802
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3909626007080078,
+      "learning_rate": 3.403695772208032e-06,
+      "loss": 0.5647,
+      "step": 4803
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0790553092956543,
+      "learning_rate": 3.4017294742587812e-06,
+      "loss": 0.6149,
+      "step": 4804
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1320269107818604,
+      "learning_rate": 3.399763451579144e-06,
+      "loss": 0.7715,
+      "step": 4805
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1932052373886108,
+      "learning_rate": 3.3977977045077247e-06,
+      "loss": 0.6292,
+      "step": 4806
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2008845806121826,
+      "learning_rate": 3.3958322333830864e-06,
+      "loss": 0.6292,
+      "step": 4807
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3319114446640015,
+      "learning_rate": 3.393867038543738e-06,
+      "loss": 0.5772,
+      "step": 4808
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5973466634750366,
+      "learning_rate": 3.3919021203281475e-06,
+      "loss": 0.6426,
+      "step": 4809
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.542841911315918,
+      "learning_rate": 3.389937479074731e-06,
+      "loss": 0.5787,
+      "step": 4810
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9632922410964966,
+      "learning_rate": 3.3879731151218575e-06,
+      "loss": 0.5883,
+      "step": 4811
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3410135507583618,
+      "learning_rate": 3.3860090288078496e-06,
+      "loss": 0.6418,
+      "step": 4812
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.061713695526123,
+      "learning_rate": 3.3840452204709806e-06,
+      "loss": 0.6125,
+      "step": 4813
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0856437683105469,
+      "learning_rate": 3.382081690449477e-06,
+      "loss": 0.5871,
+      "step": 4814
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.18121337890625,
+      "learning_rate": 3.3801184390815173e-06,
+      "loss": 0.5499,
+      "step": 4815
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4576302766799927,
+      "learning_rate": 3.37815546670523e-06,
+      "loss": 0.5686,
+      "step": 4816
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9863116145133972,
+      "learning_rate": 3.3761927736586976e-06,
+      "loss": 0.58,
+      "step": 4817
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1071988344192505,
+      "learning_rate": 3.3742303602799565e-06,
+      "loss": 0.6325,
+      "step": 4818
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.518639087677002,
+      "learning_rate": 3.3722682269069906e-06,
+      "loss": 0.5877,
+      "step": 4819
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3329269886016846,
+      "learning_rate": 3.370306373877738e-06,
+      "loss": 0.5854,
+      "step": 4820
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4260133504867554,
+      "learning_rate": 3.368344801530087e-06,
+      "loss": 0.6094,
+      "step": 4821
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.498528242111206,
+      "learning_rate": 3.3663835102018803e-06,
+      "loss": 0.5885,
+      "step": 4822
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4872510433197021,
+      "learning_rate": 3.364422500230908e-06,
+      "loss": 0.581,
+      "step": 4823
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.251771092414856,
+      "learning_rate": 3.3624617719549178e-06,
+      "loss": 0.6102,
+      "step": 4824
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.566983938217163,
+      "learning_rate": 3.3605013257116016e-06,
+      "loss": 0.6619,
+      "step": 4825
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4458528757095337,
+      "learning_rate": 3.3585411618386086e-06,
+      "loss": 0.5325,
+      "step": 4826
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2375273704528809,
+      "learning_rate": 3.356581280673536e-06,
+      "loss": 0.5656,
+      "step": 4827
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2488701343536377,
+      "learning_rate": 3.3546216825539347e-06,
+      "loss": 0.5597,
+      "step": 4828
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3934226036071777,
+      "learning_rate": 3.3526623678173043e-06,
+      "loss": 0.5244,
+      "step": 4829
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.497615098953247,
+      "learning_rate": 3.350703336801099e-06,
+      "loss": 0.5815,
+      "step": 4830
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1210520267486572,
+      "learning_rate": 3.3487445898427195e-06,
+      "loss": 0.5517,
+      "step": 4831
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2002229690551758,
+      "learning_rate": 3.346786127279522e-06,
+      "loss": 0.6144,
+      "step": 4832
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2714262008666992,
+      "learning_rate": 3.34482794944881e-06,
+      "loss": 0.5459,
+      "step": 4833
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2544283866882324,
+      "learning_rate": 3.3428700566878407e-06,
+      "loss": 0.5844,
+      "step": 4834
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3227342367172241,
+      "learning_rate": 3.340912449333824e-06,
+      "loss": 0.5983,
+      "step": 4835
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1427699327468872,
+      "learning_rate": 3.3389551277239143e-06,
+      "loss": 0.5442,
+      "step": 4836
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4273881912231445,
+      "learning_rate": 3.3369980921952227e-06,
+      "loss": 0.5977,
+      "step": 4837
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.286092758178711,
+      "learning_rate": 3.335041343084807e-06,
+      "loss": 0.7324,
+      "step": 4838
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.303217887878418,
+      "learning_rate": 3.3330848807296796e-06,
+      "loss": 0.6638,
+      "step": 4839
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4378916025161743,
+      "learning_rate": 3.331128705466799e-06,
+      "loss": 0.63,
+      "step": 4840
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.227399468421936,
+      "learning_rate": 3.3291728176330786e-06,
+      "loss": 0.625,
+      "step": 4841
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3684594631195068,
+      "learning_rate": 3.327217217565379e-06,
+      "loss": 0.6153,
+      "step": 4842
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5865708589553833,
+      "learning_rate": 3.325261905600514e-06,
+      "loss": 0.6505,
+      "step": 4843
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1837784051895142,
+      "learning_rate": 3.3233068820752447e-06,
+      "loss": 0.5805,
+      "step": 4844
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0995562076568604,
+      "learning_rate": 3.321352147326285e-06,
+      "loss": 0.7422,
+      "step": 4845
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.242751121520996,
+      "learning_rate": 3.3193977016902988e-06,
+      "loss": 0.6068,
+      "step": 4846
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5467709302902222,
+      "learning_rate": 3.317443545503898e-06,
+      "loss": 0.5951,
+      "step": 4847
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.178579568862915,
+      "learning_rate": 3.315489679103648e-06,
+      "loss": 0.515,
+      "step": 4848
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3723570108413696,
+      "learning_rate": 3.3135361028260604e-06,
+      "loss": 0.5833,
+      "step": 4849
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.414771795272827,
+      "learning_rate": 3.3115828170076026e-06,
+      "loss": 0.6358,
+      "step": 4850
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.168471336364746,
+      "learning_rate": 3.3096298219846835e-06,
+      "loss": 0.5542,
+      "step": 4851
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5970467329025269,
+      "learning_rate": 3.3076771180936707e-06,
+      "loss": 0.7044,
+      "step": 4852
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1945412158966064,
+      "learning_rate": 3.305724705670877e-06,
+      "loss": 0.5253,
+      "step": 4853
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2077255249023438,
+      "learning_rate": 3.3037725850525648e-06,
+      "loss": 0.6126,
+      "step": 4854
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2469735145568848,
+      "learning_rate": 3.3018207565749484e-06,
+      "loss": 0.5933,
+      "step": 4855
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3545241355895996,
+      "learning_rate": 3.2998692205741893e-06,
+      "loss": 0.5633,
+      "step": 4856
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2162749767303467,
+      "learning_rate": 3.2979179773864013e-06,
+      "loss": 0.5882,
+      "step": 4857
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2614243030548096,
+      "learning_rate": 3.295967027347645e-06,
+      "loss": 0.5937,
+      "step": 4858
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.402371883392334,
+      "learning_rate": 3.2940163707939333e-06,
+      "loss": 0.5585,
+      "step": 4859
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.107991099357605,
+      "learning_rate": 3.2920660080612245e-06,
+      "loss": 0.6597,
+      "step": 4860
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2759180068969727,
+      "learning_rate": 3.2901159394854324e-06,
+      "loss": 0.5066,
+      "step": 4861
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3145313262939453,
+      "learning_rate": 3.2881661654024144e-06,
+      "loss": 0.5645,
+      "step": 4862
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1362555027008057,
+      "learning_rate": 3.2862166861479806e-06,
+      "loss": 0.5143,
+      "step": 4863
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.101493000984192,
+      "learning_rate": 3.284267502057888e-06,
+      "loss": 0.6588,
+      "step": 4864
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3010179996490479,
+      "learning_rate": 3.2823186134678455e-06,
+      "loss": 0.5626,
+      "step": 4865
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1528490781784058,
+      "learning_rate": 3.280370020713507e-06,
+      "loss": 0.6245,
+      "step": 4866
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2665293216705322,
+      "learning_rate": 3.2784217241304815e-06,
+      "loss": 0.5826,
+      "step": 4867
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4382033348083496,
+      "learning_rate": 3.2764737240543192e-06,
+      "loss": 0.565,
+      "step": 4868
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.322492003440857,
+      "learning_rate": 3.2745260208205273e-06,
+      "loss": 0.5892,
+      "step": 4869
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3495713472366333,
+      "learning_rate": 3.2725786147645577e-06,
+      "loss": 0.6144,
+      "step": 4870
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2086135149002075,
+      "learning_rate": 3.2706315062218085e-06,
+      "loss": 0.5433,
+      "step": 4871
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6504979133605957,
+      "learning_rate": 3.268684695527634e-06,
+      "loss": 0.5627,
+      "step": 4872
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2472084760665894,
+      "learning_rate": 3.2667381830173287e-06,
+      "loss": 0.5854,
+      "step": 4873
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8455642461776733,
+      "learning_rate": 3.2647919690261433e-06,
+      "loss": 0.5472,
+      "step": 4874
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2269060611724854,
+      "learning_rate": 3.26284605388927e-06,
+      "loss": 0.5151,
+      "step": 4875
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9887154698371887,
+      "learning_rate": 3.2609004379418564e-06,
+      "loss": 0.6177,
+      "step": 4876
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5012866258621216,
+      "learning_rate": 3.2589551215189925e-06,
+      "loss": 0.694,
+      "step": 4877
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2509353160858154,
+      "learning_rate": 3.257010104955722e-06,
+      "loss": 0.6014,
+      "step": 4878
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3519951105117798,
+      "learning_rate": 3.255065388587032e-06,
+      "loss": 0.6751,
+      "step": 4879
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2042193412780762,
+      "learning_rate": 3.253120972747863e-06,
+      "loss": 0.5931,
+      "step": 4880
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2597451210021973,
+      "learning_rate": 3.251176857773099e-06,
+      "loss": 0.6325,
+      "step": 4881
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.296736478805542,
+      "learning_rate": 3.249233043997576e-06,
+      "loss": 0.6089,
+      "step": 4882
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2353603839874268,
+      "learning_rate": 3.2472895317560744e-06,
+      "loss": 0.6336,
+      "step": 4883
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2520846128463745,
+      "learning_rate": 3.2453463213833267e-06,
+      "loss": 0.5977,
+      "step": 4884
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.338379979133606,
+      "learning_rate": 3.2434034132140085e-06,
+      "loss": 0.6035,
+      "step": 4885
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2667759656906128,
+      "learning_rate": 3.241460807582749e-06,
+      "loss": 0.7181,
+      "step": 4886
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2850943803787231,
+      "learning_rate": 3.2395185048241235e-06,
+      "loss": 0.4562,
+      "step": 4887
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7538964748382568,
+      "learning_rate": 3.2375765052726505e-06,
+      "loss": 0.5707,
+      "step": 4888
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2799808979034424,
+      "learning_rate": 3.2356348092628038e-06,
+      "loss": 0.692,
+      "step": 4889
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.681443452835083,
+      "learning_rate": 3.2336934171289974e-06,
+      "loss": 0.5608,
+      "step": 4890
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5989010334014893,
+      "learning_rate": 3.2317523292055998e-06,
+      "loss": 0.5823,
+      "step": 4891
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2030177116394043,
+      "learning_rate": 3.2298115458269212e-06,
+      "loss": 0.6447,
+      "step": 4892
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3958220481872559,
+      "learning_rate": 3.227871067327225e-06,
+      "loss": 0.6051,
+      "step": 4893
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4750139713287354,
+      "learning_rate": 3.225930894040717e-06,
+      "loss": 0.6222,
+      "step": 4894
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5289214849472046,
+      "learning_rate": 3.2239910263015524e-06,
+      "loss": 0.626,
+      "step": 4895
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1837035417556763,
+      "learning_rate": 3.222051464443836e-06,
+      "loss": 0.5143,
+      "step": 4896
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1531528234481812,
+      "learning_rate": 3.220112208801615e-06,
+      "loss": 0.6023,
+      "step": 4897
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0396391153335571,
+      "learning_rate": 3.21817325970889e-06,
+      "loss": 0.5225,
+      "step": 4898
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2080007791519165,
+      "learning_rate": 3.216234617499603e-06,
+      "loss": 0.4965,
+      "step": 4899
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1739264726638794,
+      "learning_rate": 3.2142962825076477e-06,
+      "loss": 0.5772,
+      "step": 4900
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0638878345489502,
+      "learning_rate": 3.2123582550668608e-06,
+      "loss": 0.6668,
+      "step": 4901
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3319416046142578,
+      "learning_rate": 3.210420535511031e-06,
+      "loss": 0.5559,
+      "step": 4902
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.216744065284729,
+      "learning_rate": 3.2084831241738866e-06,
+      "loss": 0.4514,
+      "step": 4903
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1501775979995728,
+      "learning_rate": 3.206546021389111e-06,
+      "loss": 0.5571,
+      "step": 4904
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2810231447219849,
+      "learning_rate": 3.2046092274903316e-06,
+      "loss": 0.7026,
+      "step": 4905
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1379876136779785,
+      "learning_rate": 3.2026727428111186e-06,
+      "loss": 0.5862,
+      "step": 4906
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.143991231918335,
+      "learning_rate": 3.200736567684995e-06,
+      "loss": 0.5613,
+      "step": 4907
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5698902606964111,
+      "learning_rate": 3.198800702445425e-06,
+      "loss": 0.5612,
+      "step": 4908
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3888661861419678,
+      "learning_rate": 3.196865147425824e-06,
+      "loss": 0.6665,
+      "step": 4909
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4403868913650513,
+      "learning_rate": 3.19492990295955e-06,
+      "loss": 0.5879,
+      "step": 4910
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.532800316810608,
+      "learning_rate": 3.1929949693799134e-06,
+      "loss": 0.7634,
+      "step": 4911
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1193608045578003,
+      "learning_rate": 3.1910603470201616e-06,
+      "loss": 0.6415,
+      "step": 4912
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3783601522445679,
+      "learning_rate": 3.189126036213499e-06,
+      "loss": 0.6293,
+      "step": 4913
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1927382946014404,
+      "learning_rate": 3.1871920372930687e-06,
+      "loss": 0.5678,
+      "step": 4914
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2124135494232178,
+      "learning_rate": 3.185258350591963e-06,
+      "loss": 0.6184,
+      "step": 4915
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0941423177719116,
+      "learning_rate": 3.1833249764432206e-06,
+      "loss": 0.5018,
+      "step": 4916
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1445602178573608,
+      "learning_rate": 3.1813919151798265e-06,
+      "loss": 0.654,
+      "step": 4917
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6601332426071167,
+      "learning_rate": 3.1794591671347087e-06,
+      "loss": 0.5667,
+      "step": 4918
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3656716346740723,
+      "learning_rate": 3.177526732640747e-06,
+      "loss": 0.5714,
+      "step": 4919
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.346827507019043,
+      "learning_rate": 3.1755946120307605e-06,
+      "loss": 0.583,
+      "step": 4920
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2119982242584229,
+      "learning_rate": 3.173662805637521e-06,
+      "loss": 0.5986,
+      "step": 4921
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3897769451141357,
+      "learning_rate": 3.1717313137937415e-06,
+      "loss": 0.5864,
+      "step": 4922
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1082748174667358,
+      "learning_rate": 3.1698001368320817e-06,
+      "loss": 0.5421,
+      "step": 4923
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4113824367523193,
+      "learning_rate": 3.16786927508515e-06,
+      "loss": 0.588,
+      "step": 4924
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2338011264801025,
+      "learning_rate": 3.1659387288854937e-06,
+      "loss": 0.5927,
+      "step": 4925
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.165490746498108,
+      "learning_rate": 3.164008498565615e-06,
+      "loss": 0.5956,
+      "step": 4926
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.206478476524353,
+      "learning_rate": 3.1620785844579526e-06,
+      "loss": 0.589,
+      "step": 4927
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.202146291732788,
+      "learning_rate": 3.160148986894899e-06,
+      "loss": 0.6487,
+      "step": 4928
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1670382022857666,
+      "learning_rate": 3.1582197062087837e-06,
+      "loss": 0.495,
+      "step": 4929
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.248085618019104,
+      "learning_rate": 3.15629074273189e-06,
+      "loss": 0.6223,
+      "step": 4930
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2666771411895752,
+      "learning_rate": 3.15436209679644e-06,
+      "loss": 0.6845,
+      "step": 4931
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1898845434188843,
+      "learning_rate": 3.1524337687346065e-06,
+      "loss": 0.5541,
+      "step": 4932
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.34114670753479,
+      "learning_rate": 3.150505758878501e-06,
+      "loss": 0.5768,
+      "step": 4933
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2350118160247803,
+      "learning_rate": 3.1485780675601878e-06,
+      "loss": 0.5804,
+      "step": 4934
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3063546419143677,
+      "learning_rate": 3.1466506951116697e-06,
+      "loss": 0.6739,
+      "step": 4935
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5891927480697632,
+      "learning_rate": 3.1447236418648997e-06,
+      "loss": 0.5966,
+      "step": 4936
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3554737567901611,
+      "learning_rate": 3.1427969081517705e-06,
+      "loss": 0.5389,
+      "step": 4937
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4271050691604614,
+      "learning_rate": 3.1408704943041257e-06,
+      "loss": 0.6583,
+      "step": 4938
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2718795537948608,
+      "learning_rate": 3.1389444006537517e-06,
+      "loss": 0.5947,
+      "step": 4939
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1423945426940918,
+      "learning_rate": 3.1370186275323756e-06,
+      "loss": 0.5624,
+      "step": 4940
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.107884168624878,
+      "learning_rate": 3.135093175271676e-06,
+      "loss": 0.6333,
+      "step": 4941
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4218220710754395,
+      "learning_rate": 3.1331680442032697e-06,
+      "loss": 0.6028,
+      "step": 4942
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4798823595046997,
+      "learning_rate": 3.131243234658724e-06,
+      "loss": 0.5927,
+      "step": 4943
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1230974197387695,
+      "learning_rate": 3.1293187469695472e-06,
+      "loss": 0.5974,
+      "step": 4944
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2540388107299805,
+      "learning_rate": 3.127394581467193e-06,
+      "loss": 0.5366,
+      "step": 4945
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.489526629447937,
+      "learning_rate": 3.1254707384830607e-06,
+      "loss": 0.6562,
+      "step": 4946
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.336727261543274,
+      "learning_rate": 3.123547218348491e-06,
+      "loss": 0.6517,
+      "step": 4947
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2091395854949951,
+      "learning_rate": 3.121624021394774e-06,
+      "loss": 0.6301,
+      "step": 4948
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2998344898223877,
+      "learning_rate": 3.1197011479531386e-06,
+      "loss": 0.7585,
+      "step": 4949
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2029203176498413,
+      "learning_rate": 3.1177785983547633e-06,
+      "loss": 0.5922,
+      "step": 4950
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2441179752349854,
+      "learning_rate": 3.1158563729307658e-06,
+      "loss": 0.6311,
+      "step": 4951
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2800577878952026,
+      "learning_rate": 3.113934472012212e-06,
+      "loss": 0.5742,
+      "step": 4952
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2300503253936768,
+      "learning_rate": 3.112012895930109e-06,
+      "loss": 0.5904,
+      "step": 4953
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1221928596496582,
+      "learning_rate": 3.110091645015409e-06,
+      "loss": 0.5441,
+      "step": 4954
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1386147737503052,
+      "learning_rate": 3.1081707195990115e-06,
+      "loss": 0.6479,
+      "step": 4955
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9600458145141602,
+      "learning_rate": 3.1062501200117536e-06,
+      "loss": 0.4795,
+      "step": 4956
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1221802234649658,
+      "learning_rate": 3.1043298465844207e-06,
+      "loss": 0.5593,
+      "step": 4957
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3925119638442993,
+      "learning_rate": 3.1024098996477407e-06,
+      "loss": 0.5899,
+      "step": 4958
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0606712102890015,
+      "learning_rate": 3.1004902795323867e-06,
+      "loss": 0.4922,
+      "step": 4959
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3476324081420898,
+      "learning_rate": 3.098570986568972e-06,
+      "loss": 0.6115,
+      "step": 4960
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3782424926757812,
+      "learning_rate": 3.096652021088057e-06,
+      "loss": 0.6623,
+      "step": 4961
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2778935432434082,
+      "learning_rate": 3.0947333834201443e-06,
+      "loss": 0.6427,
+      "step": 4962
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.0849034786224365,
+      "learning_rate": 3.092815073895681e-06,
+      "loss": 0.573,
+      "step": 4963
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0979350805282593,
+      "learning_rate": 3.0908970928450555e-06,
+      "loss": 0.5269,
+      "step": 4964
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2356542348861694,
+      "learning_rate": 3.0889794405986024e-06,
+      "loss": 0.5587,
+      "step": 4965
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2701390981674194,
+      "learning_rate": 3.087062117486597e-06,
+      "loss": 0.6439,
+      "step": 4966
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3336454629898071,
+      "learning_rate": 3.0851451238392604e-06,
+      "loss": 0.5936,
+      "step": 4967
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7004276514053345,
+      "learning_rate": 3.0832284599867544e-06,
+      "loss": 0.5704,
+      "step": 4968
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0348007678985596,
+      "learning_rate": 3.0813121262591885e-06,
+      "loss": 0.5852,
+      "step": 4969
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3528045415878296,
+      "learning_rate": 3.0793961229866077e-06,
+      "loss": 0.6092,
+      "step": 4970
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1565823554992676,
+      "learning_rate": 3.0774804504990064e-06,
+      "loss": 0.6212,
+      "step": 4971
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.185302734375,
+      "learning_rate": 3.0755651091263233e-06,
+      "loss": 0.5614,
+      "step": 4972
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3281807899475098,
+      "learning_rate": 3.073650099198433e-06,
+      "loss": 0.6672,
+      "step": 4973
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.251400351524353,
+      "learning_rate": 3.07173542104516e-06,
+      "loss": 0.7054,
+      "step": 4974
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.98326575756073,
+      "learning_rate": 3.069821074996266e-06,
+      "loss": 0.6104,
+      "step": 4975
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5410773754119873,
+      "learning_rate": 3.067907061381461e-06,
+      "loss": 0.6519,
+      "step": 4976
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4055252075195312,
+      "learning_rate": 3.0659933805303914e-06,
+      "loss": 0.607,
+      "step": 4977
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.695743203163147,
+      "learning_rate": 3.0640800327726537e-06,
+      "loss": 0.6483,
+      "step": 4978
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2075164318084717,
+      "learning_rate": 3.06216701843778e-06,
+      "loss": 0.5343,
+      "step": 4979
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.335493564605713,
+      "learning_rate": 3.060254337855251e-06,
+      "loss": 0.6638,
+      "step": 4980
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1143360137939453,
+      "learning_rate": 3.0583419913544833e-06,
+      "loss": 0.5355,
+      "step": 4981
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7671163082122803,
+      "learning_rate": 3.056429979264844e-06,
+      "loss": 0.5906,
+      "step": 4982
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.333821415901184,
+      "learning_rate": 3.0545183019156345e-06,
+      "loss": 0.6186,
+      "step": 4983
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1221269369125366,
+      "learning_rate": 3.052606959636106e-06,
+      "loss": 0.634,
+      "step": 4984
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.378235936164856,
+      "learning_rate": 3.0506959527554445e-06,
+      "loss": 0.5971,
+      "step": 4985
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.267785668373108,
+      "learning_rate": 3.0487852816027853e-06,
+      "loss": 0.5958,
+      "step": 4986
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2087866067886353,
+      "learning_rate": 3.046874946507201e-06,
+      "loss": 0.6207,
+      "step": 4987
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.871625542640686,
+      "learning_rate": 3.0449649477977073e-06,
+      "loss": 0.5814,
+      "step": 4988
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2489285469055176,
+      "learning_rate": 3.0430552858032647e-06,
+      "loss": 0.6081,
+      "step": 4989
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1966968774795532,
+      "learning_rate": 3.0411459608527727e-06,
+      "loss": 0.5906,
+      "step": 4990
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.507742166519165,
+      "learning_rate": 3.039236973275075e-06,
+      "loss": 0.628,
+      "step": 4991
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4480714797973633,
+      "learning_rate": 3.037328323398953e-06,
+      "loss": 0.5951,
+      "step": 4992
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2698642015457153,
+      "learning_rate": 3.035420011553136e-06,
+      "loss": 0.5878,
+      "step": 4993
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0168004035949707,
+      "learning_rate": 3.0335120380662897e-06,
+      "loss": 0.6092,
+      "step": 4994
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1975743770599365,
+      "learning_rate": 3.0316044032670245e-06,
+      "loss": 0.607,
+      "step": 4995
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.311187982559204,
+      "learning_rate": 3.0296971074838923e-06,
+      "loss": 0.5956,
+      "step": 4996
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.199485182762146,
+      "learning_rate": 3.027790151045384e-06,
+      "loss": 0.7287,
+      "step": 4997
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2564228773117065,
+      "learning_rate": 3.0258835342799362e-06,
+      "loss": 0.6196,
+      "step": 4998
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.117231011390686,
+      "learning_rate": 3.023977257515924e-06,
+      "loss": 0.5185,
+      "step": 4999
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2204550504684448,
+      "learning_rate": 3.022071321081666e-06,
+      "loss": 0.5253,
+      "step": 5000
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2541391849517822,
+      "learning_rate": 3.020165725305419e-06,
+      "loss": 0.6425,
+      "step": 5001
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.467839241027832,
+      "learning_rate": 3.018260470515385e-06,
+      "loss": 0.6367,
+      "step": 5002
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1253759860992432,
+      "learning_rate": 3.016355557039704e-06,
+      "loss": 0.5926,
+      "step": 5003
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2686272859573364,
+      "learning_rate": 3.0144509852064597e-06,
+      "loss": 0.534,
+      "step": 5004
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.152122139930725,
+      "learning_rate": 3.0125467553436737e-06,
+      "loss": 0.5331,
+      "step": 5005
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1917579174041748,
+      "learning_rate": 3.0106428677793133e-06,
+      "loss": 0.596,
+      "step": 5006
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1018012762069702,
+      "learning_rate": 3.008739322841285e-06,
+      "loss": 0.6628,
+      "step": 5007
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.526096224784851,
+      "learning_rate": 3.0068361208574336e-06,
+      "loss": 0.604,
+      "step": 5008
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2660118341445923,
+      "learning_rate": 3.0049332621555483e-06,
+      "loss": 0.6269,
+      "step": 5009
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.051699161529541,
+      "learning_rate": 3.003030747063357e-06,
+      "loss": 0.5862,
+      "step": 5010
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3319251537322998,
+      "learning_rate": 3.0011285759085296e-06,
+      "loss": 0.5735,
+      "step": 5011
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.586374282836914,
+      "learning_rate": 2.9992267490186766e-06,
+      "loss": 0.6117,
+      "step": 5012
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3304932117462158,
+      "learning_rate": 2.9973252667213494e-06,
+      "loss": 0.6032,
+      "step": 5013
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2678347826004028,
+      "learning_rate": 2.995424129344038e-06,
+      "loss": 0.4959,
+      "step": 5014
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3165990114212036,
+      "learning_rate": 2.993523337214177e-06,
+      "loss": 0.6297,
+      "step": 5015
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1281373500823975,
+      "learning_rate": 2.9916228906591366e-06,
+      "loss": 0.5204,
+      "step": 5016
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.138863205909729,
+      "learning_rate": 2.9897227900062327e-06,
+      "loss": 0.6399,
+      "step": 5017
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1743173599243164,
+      "learning_rate": 2.9878230355827166e-06,
+      "loss": 0.5562,
+      "step": 5018
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.163933277130127,
+      "learning_rate": 2.985923627715785e-06,
+      "loss": 0.614,
+      "step": 5019
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2621997594833374,
+      "learning_rate": 2.9840245667325697e-06,
+      "loss": 0.6404,
+      "step": 5020
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3806610107421875,
+      "learning_rate": 2.982125852960148e-06,
+      "loss": 0.6053,
+      "step": 5021
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8941912651062012,
+      "learning_rate": 2.9802274867255306e-06,
+      "loss": 0.6629,
+      "step": 5022
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2895264625549316,
+      "learning_rate": 2.9783294683556764e-06,
+      "loss": 0.6593,
+      "step": 5023
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5828664302825928,
+      "learning_rate": 2.9764317981774804e-06,
+      "loss": 0.6568,
+      "step": 5024
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4575058221817017,
+      "learning_rate": 2.9745344765177753e-06,
+      "loss": 0.6443,
+      "step": 5025
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.212241530418396,
+      "learning_rate": 2.9726375037033396e-06,
+      "loss": 0.5902,
+      "step": 5026
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.538864016532898,
+      "learning_rate": 2.9707408800608837e-06,
+      "loss": 0.5993,
+      "step": 5027
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0669389963150024,
+      "learning_rate": 2.968844605917067e-06,
+      "loss": 0.5779,
+      "step": 5028
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2678580284118652,
+      "learning_rate": 2.9669486815984807e-06,
+      "loss": 0.6404,
+      "step": 5029
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2878882884979248,
+      "learning_rate": 2.965053107431662e-06,
+      "loss": 0.5885,
+      "step": 5030
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1472769975662231,
+      "learning_rate": 2.9631578837430826e-06,
+      "loss": 0.6865,
+      "step": 5031
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0584124326705933,
+      "learning_rate": 2.9612630108591576e-06,
+      "loss": 0.5281,
+      "step": 5032
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6062891483306885,
+      "learning_rate": 2.9593684891062403e-06,
+      "loss": 0.5883,
+      "step": 5033
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0802760124206543,
+      "learning_rate": 2.9574743188106235e-06,
+      "loss": 0.5788,
+      "step": 5034
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1777443885803223,
+      "learning_rate": 2.9555805002985394e-06,
+      "loss": 0.5501,
+      "step": 5035
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3648144006729126,
+      "learning_rate": 2.9536870338961597e-06,
+      "loss": 0.6221,
+      "step": 5036
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4176534414291382,
+      "learning_rate": 2.9517939199295965e-06,
+      "loss": 0.6233,
+      "step": 5037
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1230452060699463,
+      "learning_rate": 2.9499011587248972e-06,
+      "loss": 0.5559,
+      "step": 5038
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0202972888946533,
+      "learning_rate": 2.948008750608055e-06,
+      "loss": 0.6254,
+      "step": 5039
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4414222240447998,
+      "learning_rate": 2.9461166959049943e-06,
+      "loss": 0.5967,
+      "step": 5040
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1397916078567505,
+      "learning_rate": 2.9442249949415893e-06,
+      "loss": 0.4967,
+      "step": 5041
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.192447543144226,
+      "learning_rate": 2.942333648043643e-06,
+      "loss": 0.5351,
+      "step": 5042
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.623815894126892,
+      "learning_rate": 2.9404426555369012e-06,
+      "loss": 0.5818,
+      "step": 5043
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3485163450241089,
+      "learning_rate": 2.9385520177470517e-06,
+      "loss": 0.5429,
+      "step": 5044
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3247160911560059,
+      "learning_rate": 2.936661734999715e-06,
+      "loss": 0.5124,
+      "step": 5045
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.489458441734314,
+      "learning_rate": 2.934771807620457e-06,
+      "loss": 0.6052,
+      "step": 5046
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4193159341812134,
+      "learning_rate": 2.932882235934776e-06,
+      "loss": 0.5938,
+      "step": 5047
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.929944396018982,
+      "learning_rate": 2.9309930202681156e-06,
+      "loss": 0.5478,
+      "step": 5048
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3525385856628418,
+      "learning_rate": 2.9291041609458528e-06,
+      "loss": 0.5828,
+      "step": 5049
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.367734670639038,
+      "learning_rate": 2.9272156582933063e-06,
+      "loss": 0.5198,
+      "step": 5050
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2847005128860474,
+      "learning_rate": 2.9253275126357313e-06,
+      "loss": 0.5772,
+      "step": 5051
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4280390739440918,
+      "learning_rate": 2.9234397242983232e-06,
+      "loss": 0.5525,
+      "step": 5052
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.293440580368042,
+      "learning_rate": 2.921552293606214e-06,
+      "loss": 0.5723,
+      "step": 5053
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1732693910598755,
+      "learning_rate": 2.919665220884478e-06,
+      "loss": 0.6308,
+      "step": 5054
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4701364040374756,
+      "learning_rate": 2.917778506458121e-06,
+      "loss": 0.5309,
+      "step": 5055
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3998725414276123,
+      "learning_rate": 2.9158921506520943e-06,
+      "loss": 0.5675,
+      "step": 5056
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3049033880233765,
+      "learning_rate": 2.9140061537912835e-06,
+      "loss": 0.6115,
+      "step": 5057
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.081899881362915,
+      "learning_rate": 2.9121205162005134e-06,
+      "loss": 0.5292,
+      "step": 5058
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.247861623764038,
+      "learning_rate": 2.9102352382045464e-06,
+      "loss": 0.5237,
+      "step": 5059
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3318337202072144,
+      "learning_rate": 2.908350320128086e-06,
+      "loss": 0.6684,
+      "step": 5060
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.313628077507019,
+      "learning_rate": 2.906465762295766e-06,
+      "loss": 0.6259,
+      "step": 5061
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4944661855697632,
+      "learning_rate": 2.904581565032166e-06,
+      "loss": 0.5594,
+      "step": 5062
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2283939123153687,
+      "learning_rate": 2.902697728661801e-06,
+      "loss": 0.5856,
+      "step": 5063
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3918577432632446,
+      "learning_rate": 2.9008142535091246e-06,
+      "loss": 0.5818,
+      "step": 5064
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.163896918296814,
+      "learning_rate": 2.898931139898523e-06,
+      "loss": 0.6553,
+      "step": 5065
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1158233880996704,
+      "learning_rate": 2.897048388154328e-06,
+      "loss": 0.5073,
+      "step": 5066
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0618435144424438,
+      "learning_rate": 2.895165998600803e-06,
+      "loss": 0.5781,
+      "step": 5067
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5787274837493896,
+      "learning_rate": 2.893283971562154e-06,
+      "loss": 0.597,
+      "step": 5068
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4554435014724731,
+      "learning_rate": 2.891402307362519e-06,
+      "loss": 0.5957,
+      "step": 5069
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2838584184646606,
+      "learning_rate": 2.8895210063259775e-06,
+      "loss": 0.565,
+      "step": 5070
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2869141101837158,
+      "learning_rate": 2.887640068776546e-06,
+      "loss": 0.5993,
+      "step": 5071
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1139177083969116,
+      "learning_rate": 2.885759495038179e-06,
+      "loss": 0.5741,
+      "step": 5072
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.5964291095733643,
+      "learning_rate": 2.883879285434763e-06,
+      "loss": 0.6509,
+      "step": 5073
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9919441938400269,
+      "learning_rate": 2.8819994402901276e-06,
+      "loss": 0.4702,
+      "step": 5074
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2516149282455444,
+      "learning_rate": 2.8801199599280417e-06,
+      "loss": 0.6224,
+      "step": 5075
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0911529064178467,
+      "learning_rate": 2.878240844672202e-06,
+      "loss": 0.5022,
+      "step": 5076
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3920502662658691,
+      "learning_rate": 2.876362094846251e-06,
+      "loss": 0.6245,
+      "step": 5077
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.443747878074646,
+      "learning_rate": 2.874483710773765e-06,
+      "loss": 0.5352,
+      "step": 5078
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.816619634628296,
+      "learning_rate": 2.8726056927782587e-06,
+      "loss": 0.6492,
+      "step": 5079
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.262075662612915,
+      "learning_rate": 2.8707280411831796e-06,
+      "loss": 0.5654,
+      "step": 5080
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3001986742019653,
+      "learning_rate": 2.868850756311915e-06,
+      "loss": 0.5528,
+      "step": 5081
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3010063171386719,
+      "learning_rate": 2.8669738384877915e-06,
+      "loss": 0.5112,
+      "step": 5082
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.350829839706421,
+      "learning_rate": 2.8650972880340704e-06,
+      "loss": 0.6077,
+      "step": 5083
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1886601448059082,
+      "learning_rate": 2.8632211052739463e-06,
+      "loss": 0.6004,
+      "step": 5084
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.6389498710632324,
+      "learning_rate": 2.861345290530555e-06,
+      "loss": 0.6008,
+      "step": 5085
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.1080713272094727,
+      "learning_rate": 2.8594698441269696e-06,
+      "loss": 0.5678,
+      "step": 5086
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1639949083328247,
+      "learning_rate": 2.8575947663861935e-06,
+      "loss": 0.6014,
+      "step": 5087
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.455990195274353,
+      "learning_rate": 2.855720057631173e-06,
+      "loss": 0.5077,
+      "step": 5088
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3436378240585327,
+      "learning_rate": 2.853845718184789e-06,
+      "loss": 0.5635,
+      "step": 5089
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1918082237243652,
+      "learning_rate": 2.851971748369859e-06,
+      "loss": 0.5182,
+      "step": 5090
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5144091844558716,
+      "learning_rate": 2.8500981485091313e-06,
+      "loss": 0.69,
+      "step": 5091
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.447807788848877,
+      "learning_rate": 2.848224918925301e-06,
+      "loss": 0.6154,
+      "step": 5092
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.535978078842163,
+      "learning_rate": 2.8463520599409945e-06,
+      "loss": 0.5446,
+      "step": 5093
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2916719913482666,
+      "learning_rate": 2.844479571878769e-06,
+      "loss": 0.5821,
+      "step": 5094
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3413225412368774,
+      "learning_rate": 2.8426074550611237e-06,
+      "loss": 0.5991,
+      "step": 5095
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7014421224594116,
+      "learning_rate": 2.840735709810495e-06,
+      "loss": 0.6615,
+      "step": 5096
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.217027187347412,
+      "learning_rate": 2.838864336449253e-06,
+      "loss": 0.7111,
+      "step": 5097
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2105580568313599,
+      "learning_rate": 2.8369933352997003e-06,
+      "loss": 0.552,
+      "step": 5098
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.383817434310913,
+      "learning_rate": 2.8351227066840805e-06,
+      "loss": 0.6348,
+      "step": 5099
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2009333372116089,
+      "learning_rate": 2.8332524509245718e-06,
+      "loss": 0.539,
+      "step": 5100
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.331572413444519,
+      "learning_rate": 2.8313825683432906e-06,
+      "loss": 0.5791,
+      "step": 5101
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6446709632873535,
+      "learning_rate": 2.8295130592622797e-06,
+      "loss": 0.6404,
+      "step": 5102
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5512125492095947,
+      "learning_rate": 2.8276439240035287e-06,
+      "loss": 0.542,
+      "step": 5103
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3419227600097656,
+      "learning_rate": 2.8257751628889564e-06,
+      "loss": 0.5675,
+      "step": 5104
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3091492652893066,
+      "learning_rate": 2.8239067762404216e-06,
+      "loss": 0.6319,
+      "step": 5105
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.235048770904541,
+      "learning_rate": 2.822038764379712e-06,
+      "loss": 0.6524,
+      "step": 5106
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.292203664779663,
+      "learning_rate": 2.820171127628557e-06,
+      "loss": 0.5984,
+      "step": 5107
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.168729543685913,
+      "learning_rate": 2.818303866308618e-06,
+      "loss": 0.6435,
+      "step": 5108
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1043180227279663,
+      "learning_rate": 2.8164369807414936e-06,
+      "loss": 0.6302,
+      "step": 5109
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3122977018356323,
+      "learning_rate": 2.8145704712487167e-06,
+      "loss": 0.6408,
+      "step": 5110
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1558114290237427,
+      "learning_rate": 2.8127043381517553e-06,
+      "loss": 0.5507,
+      "step": 5111
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5054841041564941,
+      "learning_rate": 2.810838581772015e-06,
+      "loss": 0.6062,
+      "step": 5112
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2152769565582275,
+      "learning_rate": 2.8089732024308316e-06,
+      "loss": 0.623,
+      "step": 5113
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.117659091949463,
+      "learning_rate": 2.807108200449479e-06,
+      "loss": 0.5331,
+      "step": 5114
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4485636949539185,
+      "learning_rate": 2.805243576149167e-06,
+      "loss": 0.6216,
+      "step": 5115
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1245951652526855,
+      "learning_rate": 2.8033793298510415e-06,
+      "loss": 0.6768,
+      "step": 5116
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2359367609024048,
+      "learning_rate": 2.8015154618761754e-06,
+      "loss": 0.611,
+      "step": 5117
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2246320247650146,
+      "learning_rate": 2.7996519725455857e-06,
+      "loss": 0.5716,
+      "step": 5118
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1994571685791016,
+      "learning_rate": 2.7977888621802196e-06,
+      "loss": 0.5599,
+      "step": 5119
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1807100772857666,
+      "learning_rate": 2.7959261311009623e-06,
+      "loss": 0.583,
+      "step": 5120
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2274185419082642,
+      "learning_rate": 2.794063779628628e-06,
+      "loss": 0.5591,
+      "step": 5121
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1783159971237183,
+      "learning_rate": 2.79220180808397e-06,
+      "loss": 0.5576,
+      "step": 5122
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.355054497718811,
+      "learning_rate": 2.790340216787676e-06,
+      "loss": 0.6449,
+      "step": 5123
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0858644247055054,
+      "learning_rate": 2.788479006060368e-06,
+      "loss": 0.7109,
+      "step": 5124
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1023110151290894,
+      "learning_rate": 2.7866181762225964e-06,
+      "loss": 0.5975,
+      "step": 5125
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.26510751247406,
+      "learning_rate": 2.7847577275948573e-06,
+      "loss": 0.6797,
+      "step": 5126
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3188729286193848,
+      "learning_rate": 2.7828976604975756e-06,
+      "loss": 0.6283,
+      "step": 5127
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3665924072265625,
+      "learning_rate": 2.7810379752511045e-06,
+      "loss": 0.539,
+      "step": 5128
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.152752161026001,
+      "learning_rate": 2.779178672175741e-06,
+      "loss": 0.5775,
+      "step": 5129
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2595551013946533,
+      "learning_rate": 2.777319751591711e-06,
+      "loss": 0.6191,
+      "step": 5130
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1859368085861206,
+      "learning_rate": 2.7754612138191784e-06,
+      "loss": 0.5934,
+      "step": 5131
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0668327808380127,
+      "learning_rate": 2.7736030591782337e-06,
+      "loss": 0.7114,
+      "step": 5132
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3377395868301392,
+      "learning_rate": 2.7717452879889094e-06,
+      "loss": 0.6773,
+      "step": 5133
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3250497579574585,
+      "learning_rate": 2.7698879005711684e-06,
+      "loss": 0.5363,
+      "step": 5134
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2125672101974487,
+      "learning_rate": 2.768030897244909e-06,
+      "loss": 0.5669,
+      "step": 5135
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2120472192764282,
+      "learning_rate": 2.76617427832996e-06,
+      "loss": 0.5805,
+      "step": 5136
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2930200099945068,
+      "learning_rate": 2.764318044146087e-06,
+      "loss": 0.6083,
+      "step": 5137
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3229888677597046,
+      "learning_rate": 2.762462195012991e-06,
+      "loss": 0.5454,
+      "step": 5138
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3981189727783203,
+      "learning_rate": 2.7606067312503006e-06,
+      "loss": 0.6181,
+      "step": 5139
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3310023546218872,
+      "learning_rate": 2.7587516531775826e-06,
+      "loss": 0.5589,
+      "step": 5140
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1334054470062256,
+      "learning_rate": 2.7568969611143377e-06,
+      "loss": 0.5911,
+      "step": 5141
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3404936790466309,
+      "learning_rate": 2.755042655379998e-06,
+      "loss": 0.624,
+      "step": 5142
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3292677402496338,
+      "learning_rate": 2.7531887362939314e-06,
+      "loss": 0.598,
+      "step": 5143
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5356954336166382,
+      "learning_rate": 2.751335204175436e-06,
+      "loss": 0.6055,
+      "step": 5144
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.472981333732605,
+      "learning_rate": 2.7494820593437483e-06,
+      "loss": 0.6197,
+      "step": 5145
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6259320974349976,
+      "learning_rate": 2.74762930211803e-06,
+      "loss": 0.5957,
+      "step": 5146
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2277649641036987,
+      "learning_rate": 2.745776932817384e-06,
+      "loss": 0.5525,
+      "step": 5147
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.352315068244934,
+      "learning_rate": 2.743924951760842e-06,
+      "loss": 0.5822,
+      "step": 5148
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4033784866333008,
+      "learning_rate": 2.7420733592673727e-06,
+      "loss": 0.5787,
+      "step": 5149
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3330798149108887,
+      "learning_rate": 2.740222155655871e-06,
+      "loss": 0.5817,
+      "step": 5150
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3742144107818604,
+      "learning_rate": 2.7383713412451716e-06,
+      "loss": 0.5179,
+      "step": 5151
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.388723373413086,
+      "learning_rate": 2.736520916354039e-06,
+      "loss": 0.4809,
+      "step": 5152
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3881044387817383,
+      "learning_rate": 2.734670881301174e-06,
+      "loss": 0.5648,
+      "step": 5153
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7049732208251953,
+      "learning_rate": 2.732821236405203e-06,
+      "loss": 0.5736,
+      "step": 5154
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0168077945709229,
+      "learning_rate": 2.730971981984692e-06,
+      "loss": 0.6677,
+      "step": 5155
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2047303915023804,
+      "learning_rate": 2.729123118358137e-06,
+      "loss": 0.6059,
+      "step": 5156
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1460449695587158,
+      "learning_rate": 2.7272746458439705e-06,
+      "loss": 0.6915,
+      "step": 5157
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4450008869171143,
+      "learning_rate": 2.7254265647605483e-06,
+      "loss": 0.6629,
+      "step": 5158
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5762354135513306,
+      "learning_rate": 2.723578875426166e-06,
+      "loss": 0.6685,
+      "step": 5159
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3964108228683472,
+      "learning_rate": 2.721731578159057e-06,
+      "loss": 0.6259,
+      "step": 5160
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.066676139831543,
+      "learning_rate": 2.7198846732773743e-06,
+      "loss": 0.7189,
+      "step": 5161
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1809970140457153,
+      "learning_rate": 2.718038161099211e-06,
+      "loss": 0.6223,
+      "step": 5162
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1365050077438354,
+      "learning_rate": 2.716192041942592e-06,
+      "loss": 0.5089,
+      "step": 5163
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1270561218261719,
+      "learning_rate": 2.7143463161254755e-06,
+      "loss": 0.555,
+      "step": 5164
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.720628261566162,
+      "learning_rate": 2.712500983965747e-06,
+      "loss": 0.6347,
+      "step": 5165
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4290040731430054,
+      "learning_rate": 2.710656045781228e-06,
+      "loss": 0.5405,
+      "step": 5166
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.559084415435791,
+      "learning_rate": 2.7088115018896725e-06,
+      "loss": 0.6434,
+      "step": 5167
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1094157695770264,
+      "learning_rate": 2.706967352608768e-06,
+      "loss": 0.5956,
+      "step": 5168
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2786990404129028,
+      "learning_rate": 2.7051235982561275e-06,
+      "loss": 0.515,
+      "step": 5169
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1976306438446045,
+      "learning_rate": 2.703280239149302e-06,
+      "loss": 0.5842,
+      "step": 5170
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3613533973693848,
+      "learning_rate": 2.701437275605773e-06,
+      "loss": 0.6508,
+      "step": 5171
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.403113842010498,
+      "learning_rate": 2.699594707942955e-06,
+      "loss": 0.6158,
+      "step": 5172
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.390093445777893,
+      "learning_rate": 2.6977525364781887e-06,
+      "loss": 0.6485,
+      "step": 5173
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1422367095947266,
+      "learning_rate": 2.695910761528754e-06,
+      "loss": 0.6894,
+      "step": 5174
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3628041744232178,
+      "learning_rate": 2.694069383411857e-06,
+      "loss": 0.5549,
+      "step": 5175
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.215430498123169,
+      "learning_rate": 2.692228402444642e-06,
+      "loss": 0.5934,
+      "step": 5176
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1754097938537598,
+      "learning_rate": 2.6903878189441734e-06,
+      "loss": 0.5572,
+      "step": 5177
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1725589036941528,
+      "learning_rate": 2.6885476332274598e-06,
+      "loss": 0.608,
+      "step": 5178
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1601642370224,
+      "learning_rate": 2.6867078456114367e-06,
+      "loss": 0.6104,
+      "step": 5179
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.180969476699829,
+      "learning_rate": 2.6848684564129657e-06,
+      "loss": 0.5651,
+      "step": 5180
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.457452654838562,
+      "learning_rate": 2.683029465948846e-06,
+      "loss": 0.6624,
+      "step": 5181
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2275680303573608,
+      "learning_rate": 2.6811908745358068e-06,
+      "loss": 0.7376,
+      "step": 5182
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2394523620605469,
+      "learning_rate": 2.6793526824905102e-06,
+      "loss": 0.6944,
+      "step": 5183
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4745672941207886,
+      "learning_rate": 2.677514890129543e-06,
+      "loss": 0.5372,
+      "step": 5184
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3252058029174805,
+      "learning_rate": 2.6756774977694295e-06,
+      "loss": 0.6018,
+      "step": 5185
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.290786623954773,
+      "learning_rate": 2.6738405057266255e-06,
+      "loss": 0.613,
+      "step": 5186
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.545142650604248,
+      "learning_rate": 2.6720039143175116e-06,
+      "loss": 0.6156,
+      "step": 5187
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6830646991729736,
+      "learning_rate": 2.6701677238584046e-06,
+      "loss": 0.5804,
+      "step": 5188
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2544286251068115,
+      "learning_rate": 2.6683319346655523e-06,
+      "loss": 0.5945,
+      "step": 5189
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4033396244049072,
+      "learning_rate": 2.666496547055133e-06,
+      "loss": 0.5992,
+      "step": 5190
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.134125828742981,
+      "learning_rate": 2.6646615613432507e-06,
+      "loss": 0.6052,
+      "step": 5191
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1849713325500488,
+      "learning_rate": 2.6628269778459475e-06,
+      "loss": 0.6269,
+      "step": 5192
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2358061075210571,
+      "learning_rate": 2.660992796879193e-06,
+      "loss": 0.6058,
+      "step": 5193
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.204371452331543,
+      "learning_rate": 2.659159018758886e-06,
+      "loss": 0.5892,
+      "step": 5194
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3742873668670654,
+      "learning_rate": 2.6573256438008594e-06,
+      "loss": 0.5877,
+      "step": 5195
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4056979417800903,
+      "learning_rate": 2.655492672320874e-06,
+      "loss": 0.6435,
+      "step": 5196
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.357175350189209,
+      "learning_rate": 2.653660104634624e-06,
+      "loss": 0.6442,
+      "step": 5197
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4769208431243896,
+      "learning_rate": 2.6518279410577276e-06,
+      "loss": 0.5422,
+      "step": 5198
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2639411687850952,
+      "learning_rate": 2.6499961819057396e-06,
+      "loss": 0.6191,
+      "step": 5199
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4961445331573486,
+      "learning_rate": 2.648164827494144e-06,
+      "loss": 0.6537,
+      "step": 5200
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1992703676223755,
+      "learning_rate": 2.6463338781383563e-06,
+      "loss": 0.5623,
+      "step": 5201
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.429692029953003,
+      "learning_rate": 2.6445033341537164e-06,
+      "loss": 0.5856,
+      "step": 5202
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4961892366409302,
+      "learning_rate": 2.642673195855499e-06,
+      "loss": 0.6466,
+      "step": 5203
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2756520509719849,
+      "learning_rate": 2.6408434635589096e-06,
+      "loss": 0.5343,
+      "step": 5204
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2954007387161255,
+      "learning_rate": 2.6390141375790834e-06,
+      "loss": 0.6814,
+      "step": 5205
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1817203760147095,
+      "learning_rate": 2.637185218231082e-06,
+      "loss": 0.5147,
+      "step": 5206
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.356261134147644,
+      "learning_rate": 2.6353567058299006e-06,
+      "loss": 0.6249,
+      "step": 5207
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4610151052474976,
+      "learning_rate": 2.633528600690463e-06,
+      "loss": 0.6509,
+      "step": 5208
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2745957374572754,
+      "learning_rate": 2.6317009031276264e-06,
+      "loss": 0.7069,
+      "step": 5209
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1911793947219849,
+      "learning_rate": 2.6298736134561686e-06,
+      "loss": 0.5957,
+      "step": 5210
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2115494012832642,
+      "learning_rate": 2.6280467319908052e-06,
+      "loss": 0.4903,
+      "step": 5211
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1515758037567139,
+      "learning_rate": 2.6262202590461843e-06,
+      "loss": 0.6614,
+      "step": 5212
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.148162841796875,
+      "learning_rate": 2.6243941949368728e-06,
+      "loss": 0.6236,
+      "step": 5213
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7739685773849487,
+      "learning_rate": 2.622568539977375e-06,
+      "loss": 0.6557,
+      "step": 5214
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.455483078956604,
+      "learning_rate": 2.620743294482123e-06,
+      "loss": 0.6547,
+      "step": 5215
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.486600399017334,
+      "learning_rate": 2.6189184587654813e-06,
+      "loss": 0.6379,
+      "step": 5216
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2012253999710083,
+      "learning_rate": 2.617094033141735e-06,
+      "loss": 0.546,
+      "step": 5217
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.324657917022705,
+      "learning_rate": 2.615270017925107e-06,
+      "loss": 0.6812,
+      "step": 5218
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.661184310913086,
+      "learning_rate": 2.6134464134297476e-06,
+      "loss": 0.6015,
+      "step": 5219
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1693406105041504,
+      "learning_rate": 2.611623219969737e-06,
+      "loss": 0.5996,
+      "step": 5220
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1396194696426392,
+      "learning_rate": 2.609800437859078e-06,
+      "loss": 0.6311,
+      "step": 5221
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.285706877708435,
+      "learning_rate": 2.607978067411712e-06,
+      "loss": 0.6305,
+      "step": 5222
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.161902666091919,
+      "learning_rate": 2.606156108941504e-06,
+      "loss": 0.6345,
+      "step": 5223
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4399522542953491,
+      "learning_rate": 2.6043345627622513e-06,
+      "loss": 0.5897,
+      "step": 5224
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.349932312965393,
+      "learning_rate": 2.6025134291876745e-06,
+      "loss": 0.6166,
+      "step": 5225
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8268938064575195,
+      "learning_rate": 2.6006927085314295e-06,
+      "loss": 0.6521,
+      "step": 5226
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7306489944458008,
+      "learning_rate": 2.5988724011070967e-06,
+      "loss": 0.5793,
+      "step": 5227
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.346104621887207,
+      "learning_rate": 2.597052507228189e-06,
+      "loss": 0.596,
+      "step": 5228
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2005935907363892,
+      "learning_rate": 2.5952330272081446e-06,
+      "loss": 0.6136,
+      "step": 5229
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2340837717056274,
+      "learning_rate": 2.5934139613603326e-06,
+      "loss": 0.5841,
+      "step": 5230
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3165960311889648,
+      "learning_rate": 2.591595309998052e-06,
+      "loss": 0.59,
+      "step": 5231
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0761138200759888,
+      "learning_rate": 2.5897770734345253e-06,
+      "loss": 0.5112,
+      "step": 5232
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1540740728378296,
+      "learning_rate": 2.5879592519829065e-06,
+      "loss": 0.5531,
+      "step": 5233
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2632431983947754,
+      "learning_rate": 2.586141845956284e-06,
+      "loss": 0.5473,
+      "step": 5234
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3133299350738525,
+      "learning_rate": 2.5843248556676615e-06,
+      "loss": 0.5852,
+      "step": 5235
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.290602207183838,
+      "learning_rate": 2.582508281429983e-06,
+      "loss": 0.5788,
+      "step": 5236
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1755783557891846,
+      "learning_rate": 2.5806921235561157e-06,
+      "loss": 0.528,
+      "step": 5237
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6222681999206543,
+      "learning_rate": 2.578876382358857e-06,
+      "loss": 0.6104,
+      "step": 5238
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2842351198196411,
+      "learning_rate": 2.5770610581509292e-06,
+      "loss": 0.528,
+      "step": 5239
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1351863145828247,
+      "learning_rate": 2.5752461512449854e-06,
+      "loss": 0.6943,
+      "step": 5240
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.076362133026123,
+      "learning_rate": 2.5734316619536072e-06,
+      "loss": 0.611,
+      "step": 5241
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2188138961791992,
+      "learning_rate": 2.571617590589306e-06,
+      "loss": 0.6463,
+      "step": 5242
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3224034309387207,
+      "learning_rate": 2.5698039374645134e-06,
+      "loss": 0.6016,
+      "step": 5243
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2040905952453613,
+      "learning_rate": 2.5679907028915972e-06,
+      "loss": 0.5456,
+      "step": 5244
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2921534776687622,
+      "learning_rate": 2.5661778871828503e-06,
+      "loss": 0.6099,
+      "step": 5245
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2232110500335693,
+      "learning_rate": 2.564365490650493e-06,
+      "loss": 0.6297,
+      "step": 5246
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3627408742904663,
+      "learning_rate": 2.562553513606674e-06,
+      "loss": 0.5766,
+      "step": 5247
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.229413390159607,
+      "learning_rate": 2.5607419563634682e-06,
+      "loss": 0.5605,
+      "step": 5248
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2325100898742676,
+      "learning_rate": 2.558930819232883e-06,
+      "loss": 0.6456,
+      "step": 5249
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1989376544952393,
+      "learning_rate": 2.5571201025268468e-06,
+      "loss": 0.641,
+      "step": 5250
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2480047941207886,
+      "learning_rate": 2.5553098065572186e-06,
+      "loss": 0.5573,
+      "step": 5251
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2484368085861206,
+      "learning_rate": 2.553499931635786e-06,
+      "loss": 0.6038,
+      "step": 5252
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4819079637527466,
+      "learning_rate": 2.5516904780742658e-06,
+      "loss": 0.6092,
+      "step": 5253
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2783153057098389,
+      "learning_rate": 2.5498814461842942e-06,
+      "loss": 0.6426,
+      "step": 5254
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3205689191818237,
+      "learning_rate": 2.548072836277443e-06,
+      "loss": 0.6066,
+      "step": 5255
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2751951217651367,
+      "learning_rate": 2.5462646486652094e-06,
+      "loss": 0.5752,
+      "step": 5256
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.27857506275177,
+      "learning_rate": 2.5444568836590174e-06,
+      "loss": 0.667,
+      "step": 5257
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5317552089691162,
+      "learning_rate": 2.5426495415702146e-06,
+      "loss": 0.6109,
+      "step": 5258
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.290687918663025,
+      "learning_rate": 2.54084262271008e-06,
+      "loss": 0.6179,
+      "step": 5259
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1564019918441772,
+      "learning_rate": 2.5390361273898207e-06,
+      "loss": 0.5787,
+      "step": 5260
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4289751052856445,
+      "learning_rate": 2.537230055920569e-06,
+      "loss": 0.5601,
+      "step": 5261
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3122379779815674,
+      "learning_rate": 2.5354244086133784e-06,
+      "loss": 0.6218,
+      "step": 5262
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3078715801239014,
+      "learning_rate": 2.533619185779241e-06,
+      "loss": 0.5187,
+      "step": 5263
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.357385277748108,
+      "learning_rate": 2.531814387729069e-06,
+      "loss": 0.648,
+      "step": 5264
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2903344631195068,
+      "learning_rate": 2.5300100147737007e-06,
+      "loss": 0.5953,
+      "step": 5265
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2929610013961792,
+      "learning_rate": 2.5282060672239016e-06,
+      "loss": 0.519,
+      "step": 5266
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.800912618637085,
+      "learning_rate": 2.526402545390367e-06,
+      "loss": 0.6361,
+      "step": 5267
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.328589916229248,
+      "learning_rate": 2.524599449583718e-06,
+      "loss": 0.6149,
+      "step": 5268
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3592402935028076,
+      "learning_rate": 2.5227967801144972e-06,
+      "loss": 0.5711,
+      "step": 5269
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2721518278121948,
+      "learning_rate": 2.5209945372931798e-06,
+      "loss": 0.5685,
+      "step": 5270
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2351329326629639,
+      "learning_rate": 2.519192721430166e-06,
+      "loss": 0.5586,
+      "step": 5271
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.6479928493499756,
+      "learning_rate": 2.5173913328357836e-06,
+      "loss": 0.5282,
+      "step": 5272
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0487334728240967,
+      "learning_rate": 2.515590371820281e-06,
+      "loss": 0.6852,
+      "step": 5273
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3234119415283203,
+      "learning_rate": 2.5137898386938396e-06,
+      "loss": 0.5542,
+      "step": 5274
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2834035158157349,
+      "learning_rate": 2.511989733766565e-06,
+      "loss": 0.5604,
+      "step": 5275
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3077657222747803,
+      "learning_rate": 2.510190057348489e-06,
+      "loss": 0.656,
+      "step": 5276
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3545429706573486,
+      "learning_rate": 2.508390809749567e-06,
+      "loss": 0.5638,
+      "step": 5277
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1976977586746216,
+      "learning_rate": 2.5065919912796845e-06,
+      "loss": 0.5891,
+      "step": 5278
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1058452129364014,
+      "learning_rate": 2.5047936022486503e-06,
+      "loss": 0.5514,
+      "step": 5279
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2233741283416748,
+      "learning_rate": 2.5029956429662017e-06,
+      "loss": 0.5231,
+      "step": 5280
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.187469244003296,
+      "learning_rate": 2.5011981137419993e-06,
+      "loss": 0.6487,
+      "step": 5281
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.358062744140625,
+      "learning_rate": 2.4994010148856325e-06,
+      "loss": 0.6192,
+      "step": 5282
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3537427186965942,
+      "learning_rate": 2.4976043467066164e-06,
+      "loss": 0.5947,
+      "step": 5283
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1666181087493896,
+      "learning_rate": 2.495808109514386e-06,
+      "loss": 0.5528,
+      "step": 5284
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.244334101676941,
+      "learning_rate": 2.4940123036183095e-06,
+      "loss": 0.5853,
+      "step": 5285
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.467995524406433,
+      "learning_rate": 2.4922169293276793e-06,
+      "loss": 0.5277,
+      "step": 5286
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.112685203552246,
+      "learning_rate": 2.4904219869517083e-06,
+      "loss": 0.6663,
+      "step": 5287
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6050626039505005,
+      "learning_rate": 2.488627476799542e-06,
+      "loss": 0.583,
+      "step": 5288
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.614817500114441,
+      "learning_rate": 2.486833399180246e-06,
+      "loss": 0.5568,
+      "step": 5289
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.259895920753479,
+      "learning_rate": 2.485039754402817e-06,
+      "loss": 0.6091,
+      "step": 5290
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1175788640975952,
+      "learning_rate": 2.4832465427761704e-06,
+      "loss": 0.6088,
+      "step": 5291
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3290629386901855,
+      "learning_rate": 2.481453764609152e-06,
+      "loss": 0.6004,
+      "step": 5292
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1337230205535889,
+      "learning_rate": 2.4796614202105313e-06,
+      "loss": 0.4849,
+      "step": 5293
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3604735136032104,
+      "learning_rate": 2.477869509889005e-06,
+      "loss": 0.5763,
+      "step": 5294
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.551064372062683,
+      "learning_rate": 2.476078033953189e-06,
+      "loss": 0.5731,
+      "step": 5295
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4335790872573853,
+      "learning_rate": 2.4742869927116297e-06,
+      "loss": 0.5623,
+      "step": 5296
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4575905799865723,
+      "learning_rate": 2.4724963864728026e-06,
+      "loss": 0.6491,
+      "step": 5297
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3047614097595215,
+      "learning_rate": 2.470706215545097e-06,
+      "loss": 0.5279,
+      "step": 5298
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3483816385269165,
+      "learning_rate": 2.4689164802368366e-06,
+      "loss": 0.636,
+      "step": 5299
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.285593032836914,
+      "learning_rate": 2.4671271808562664e-06,
+      "loss": 0.5903,
+      "step": 5300
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0162348747253418,
+      "learning_rate": 2.465338317711558e-06,
+      "loss": 0.5889,
+      "step": 5301
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4397096633911133,
+      "learning_rate": 2.4635498911108042e-06,
+      "loss": 0.5798,
+      "step": 5302
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4754639863967896,
+      "learning_rate": 2.4617619013620257e-06,
+      "loss": 0.6134,
+      "step": 5303
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1740292310714722,
+      "learning_rate": 2.4599743487731686e-06,
+      "loss": 0.5692,
+      "step": 5304
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4390628337860107,
+      "learning_rate": 2.458187233652104e-06,
+      "loss": 0.6033,
+      "step": 5305
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.326101541519165,
+      "learning_rate": 2.456400556306622e-06,
+      "loss": 0.5816,
+      "step": 5306
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6704646348953247,
+      "learning_rate": 2.454614317044443e-06,
+      "loss": 0.5938,
+      "step": 5307
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2074356079101562,
+      "learning_rate": 2.452828516173212e-06,
+      "loss": 0.5318,
+      "step": 5308
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.418071985244751,
+      "learning_rate": 2.451043154000497e-06,
+      "loss": 0.6688,
+      "step": 5309
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3660101890563965,
+      "learning_rate": 2.449258230833788e-06,
+      "loss": 0.5901,
+      "step": 5310
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3401819467544556,
+      "learning_rate": 2.4474737469805026e-06,
+      "loss": 0.6244,
+      "step": 5311
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1151803731918335,
+      "learning_rate": 2.445689702747982e-06,
+      "loss": 0.6078,
+      "step": 5312
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3116505146026611,
+      "learning_rate": 2.443906098443494e-06,
+      "loss": 0.5469,
+      "step": 5313
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2464821338653564,
+      "learning_rate": 2.4421229343742224e-06,
+      "loss": 0.573,
+      "step": 5314
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6851762533187866,
+      "learning_rate": 2.4403402108472863e-06,
+      "loss": 0.6493,
+      "step": 5315
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3235703706741333,
+      "learning_rate": 2.4385579281697235e-06,
+      "loss": 0.5459,
+      "step": 5316
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.267038106918335,
+      "learning_rate": 2.436776086648493e-06,
+      "loss": 0.6437,
+      "step": 5317
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.639235258102417,
+      "learning_rate": 2.4349946865904815e-06,
+      "loss": 0.5845,
+      "step": 5318
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2454874515533447,
+      "learning_rate": 2.4332137283024997e-06,
+      "loss": 0.537,
+      "step": 5319
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.361704707145691,
+      "learning_rate": 2.431433212091284e-06,
+      "loss": 0.5855,
+      "step": 5320
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4523319005966187,
+      "learning_rate": 2.4296531382634864e-06,
+      "loss": 0.626,
+      "step": 5321
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2574588060379028,
+      "learning_rate": 2.427873507125692e-06,
+      "loss": 0.6157,
+      "step": 5322
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.1008527278900146,
+      "learning_rate": 2.4260943189844055e-06,
+      "loss": 0.593,
+      "step": 5323
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3145480155944824,
+      "learning_rate": 2.4243155741460583e-06,
+      "loss": 0.5892,
+      "step": 5324
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1647745370864868,
+      "learning_rate": 2.422537272916998e-06,
+      "loss": 0.5774,
+      "step": 5325
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1349399089813232,
+      "learning_rate": 2.4207594156035042e-06,
+      "loss": 0.5195,
+      "step": 5326
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4423859119415283,
+      "learning_rate": 2.4189820025117762e-06,
+      "loss": 0.6294,
+      "step": 5327
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2302815914154053,
+      "learning_rate": 2.417205033947938e-06,
+      "loss": 0.5202,
+      "step": 5328
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4219542741775513,
+      "learning_rate": 2.415428510218035e-06,
+      "loss": 0.6441,
+      "step": 5329
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2709050178527832,
+      "learning_rate": 2.413652431628036e-06,
+      "loss": 0.6086,
+      "step": 5330
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3104877471923828,
+      "learning_rate": 2.4118767984838376e-06,
+      "loss": 0.5737,
+      "step": 5331
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.25847327709198,
+      "learning_rate": 2.4101016110912547e-06,
+      "loss": 0.5736,
+      "step": 5332
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.170403003692627,
+      "learning_rate": 2.4083268697560276e-06,
+      "loss": 0.5607,
+      "step": 5333
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3111367225646973,
+      "learning_rate": 2.406552574783821e-06,
+      "loss": 0.5241,
+      "step": 5334
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4980742931365967,
+      "learning_rate": 2.404778726480218e-06,
+      "loss": 0.6526,
+      "step": 5335
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3784537315368652,
+      "learning_rate": 2.4030053251507298e-06,
+      "loss": 0.5576,
+      "step": 5336
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2792404890060425,
+      "learning_rate": 2.401232371100788e-06,
+      "loss": 0.6557,
+      "step": 5337
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.282593846321106,
+      "learning_rate": 2.3994598646357505e-06,
+      "loss": 0.5845,
+      "step": 5338
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3621476888656616,
+      "learning_rate": 2.397687806060891e-06,
+      "loss": 0.5418,
+      "step": 5339
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4283099174499512,
+      "learning_rate": 2.3959161956814136e-06,
+      "loss": 0.588,
+      "step": 5340
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1404139995574951,
+      "learning_rate": 2.394145033802441e-06,
+      "loss": 0.577,
+      "step": 5341
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.461607813835144,
+      "learning_rate": 2.392374320729023e-06,
+      "loss": 0.6742,
+      "step": 5342
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1831786632537842,
+      "learning_rate": 2.390604056766124e-06,
+      "loss": 0.6059,
+      "step": 5343
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6188664436340332,
+      "learning_rate": 2.3888342422186376e-06,
+      "loss": 0.5266,
+      "step": 5344
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6657450199127197,
+      "learning_rate": 2.3870648773913796e-06,
+      "loss": 0.6137,
+      "step": 5345
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1973358392715454,
+      "learning_rate": 2.3852959625890888e-06,
+      "loss": 0.6311,
+      "step": 5346
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5780502557754517,
+      "learning_rate": 2.38352749811642e-06,
+      "loss": 0.6452,
+      "step": 5347
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6269686222076416,
+      "learning_rate": 2.3817594842779566e-06,
+      "loss": 0.5092,
+      "step": 5348
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3493552207946777,
+      "learning_rate": 2.379991921378208e-06,
+      "loss": 0.6042,
+      "step": 5349
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.241943359375,
+      "learning_rate": 2.378224809721595e-06,
+      "loss": 0.5149,
+      "step": 5350
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2631124258041382,
+      "learning_rate": 2.3764581496124693e-06,
+      "loss": 0.5638,
+      "step": 5351
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1910855770111084,
+      "learning_rate": 2.374691941355102e-06,
+      "loss": 0.5865,
+      "step": 5352
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.200222015380859,
+      "learning_rate": 2.372926185253688e-06,
+      "loss": 0.5512,
+      "step": 5353
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0969185829162598,
+      "learning_rate": 2.3711608816123393e-06,
+      "loss": 0.5978,
+      "step": 5354
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5703184604644775,
+      "learning_rate": 2.3693960307350953e-06,
+      "loss": 0.5803,
+      "step": 5355
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.877293586730957,
+      "learning_rate": 2.367631632925917e-06,
+      "loss": 0.5711,
+      "step": 5356
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0474374294281006,
+      "learning_rate": 2.365867688488686e-06,
+      "loss": 0.5735,
+      "step": 5357
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2668507099151611,
+      "learning_rate": 2.364104197727204e-06,
+      "loss": 0.5971,
+      "step": 5358
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.335062861442566,
+      "learning_rate": 2.362341160945197e-06,
+      "loss": 0.5533,
+      "step": 5359
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.19700026512146,
+      "learning_rate": 2.360578578446312e-06,
+      "loss": 0.5404,
+      "step": 5360
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3640122413635254,
+      "learning_rate": 2.358816450534121e-06,
+      "loss": 0.636,
+      "step": 5361
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0688934326171875,
+      "learning_rate": 2.3570547775121106e-06,
+      "loss": 0.6033,
+      "step": 5362
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1979340314865112,
+      "learning_rate": 2.3552935596836947e-06,
+      "loss": 0.554,
+      "step": 5363
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4938658475875854,
+      "learning_rate": 2.353532797352207e-06,
+      "loss": 0.5977,
+      "step": 5364
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.255603551864624,
+      "learning_rate": 2.3517724908209037e-06,
+      "loss": 0.5904,
+      "step": 5365
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.557724118232727,
+      "learning_rate": 2.3500126403929624e-06,
+      "loss": 0.6582,
+      "step": 5366
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2635000944137573,
+      "learning_rate": 2.3482532463714803e-06,
+      "loss": 0.5664,
+      "step": 5367
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5641952753067017,
+      "learning_rate": 2.3464943090594794e-06,
+      "loss": 0.6332,
+      "step": 5368
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1680552959442139,
+      "learning_rate": 2.3447358287598975e-06,
+      "loss": 0.625,
+      "step": 5369
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9742357730865479,
+      "learning_rate": 2.3429778057755994e-06,
+      "loss": 0.5614,
+      "step": 5370
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2470797300338745,
+      "learning_rate": 2.3412202404093675e-06,
+      "loss": 0.5732,
+      "step": 5371
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2727478742599487,
+      "learning_rate": 2.3394631329639094e-06,
+      "loss": 0.5777,
+      "step": 5372
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1417211294174194,
+      "learning_rate": 2.337706483741848e-06,
+      "loss": 0.6026,
+      "step": 5373
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1219482421875,
+      "learning_rate": 2.3359502930457306e-06,
+      "loss": 0.5525,
+      "step": 5374
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3681706190109253,
+      "learning_rate": 2.334194561178027e-06,
+      "loss": 0.5578,
+      "step": 5375
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2046383619308472,
+      "learning_rate": 2.3324392884411277e-06,
+      "loss": 0.5576,
+      "step": 5376
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.306382894515991,
+      "learning_rate": 2.3306844751373384e-06,
+      "loss": 0.5266,
+      "step": 5377
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.108659267425537,
+      "learning_rate": 2.328930121568893e-06,
+      "loss": 0.6962,
+      "step": 5378
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2322014570236206,
+      "learning_rate": 2.3271762280379446e-06,
+      "loss": 0.5757,
+      "step": 5379
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.400467038154602,
+      "learning_rate": 2.3254227948465613e-06,
+      "loss": 0.5913,
+      "step": 5380
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3882756233215332,
+      "learning_rate": 2.3236698222967392e-06,
+      "loss": 0.5503,
+      "step": 5381
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5921300649642944,
+      "learning_rate": 2.321917310690392e-06,
+      "loss": 0.5449,
+      "step": 5382
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7357591390609741,
+      "learning_rate": 2.3201652603293535e-06,
+      "loss": 0.5941,
+      "step": 5383
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2445812225341797,
+      "learning_rate": 2.318413671515379e-06,
+      "loss": 0.5878,
+      "step": 5384
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2913782596588135,
+      "learning_rate": 2.3166625445501444e-06,
+      "loss": 0.6032,
+      "step": 5385
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1985303163528442,
+      "learning_rate": 2.3149118797352484e-06,
+      "loss": 0.5549,
+      "step": 5386
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0271339416503906,
+      "learning_rate": 2.3131616773722015e-06,
+      "loss": 0.5392,
+      "step": 5387
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3953230381011963,
+      "learning_rate": 2.3114119377624444e-06,
+      "loss": 0.6916,
+      "step": 5388
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.073756456375122,
+      "learning_rate": 2.309662661207334e-06,
+      "loss": 0.6164,
+      "step": 5389
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0587557554244995,
+      "learning_rate": 2.3079138480081474e-06,
+      "loss": 0.5769,
+      "step": 5390
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3357759714126587,
+      "learning_rate": 2.3061654984660808e-06,
+      "loss": 0.6175,
+      "step": 5391
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4631446599960327,
+      "learning_rate": 2.3044176128822533e-06,
+      "loss": 0.5939,
+      "step": 5392
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3898251056671143,
+      "learning_rate": 2.3026701915577017e-06,
+      "loss": 0.5235,
+      "step": 5393
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3785183429718018,
+      "learning_rate": 2.3009232347933858e-06,
+      "loss": 0.647,
+      "step": 5394
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1423410177230835,
+      "learning_rate": 2.299176742890181e-06,
+      "loss": 0.5851,
+      "step": 5395
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.523036003112793,
+      "learning_rate": 2.297430716148885e-06,
+      "loss": 0.5871,
+      "step": 5396
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1296528577804565,
+      "learning_rate": 2.295685154870217e-06,
+      "loss": 0.5476,
+      "step": 5397
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4960590600967407,
+      "learning_rate": 2.2939400593548157e-06,
+      "loss": 0.6542,
+      "step": 5398
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2516299486160278,
+      "learning_rate": 2.2921954299032324e-06,
+      "loss": 0.5171,
+      "step": 5399
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3827989101409912,
+      "learning_rate": 2.29045126681595e-06,
+      "loss": 0.6073,
+      "step": 5400
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7401831150054932,
+      "learning_rate": 2.2887075703933654e-06,
+      "loss": 0.5981,
+      "step": 5401
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0633769035339355,
+      "learning_rate": 2.2869643409357907e-06,
+      "loss": 0.7061,
+      "step": 5402
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2310799360275269,
+      "learning_rate": 2.2852215787434636e-06,
+      "loss": 0.5535,
+      "step": 5403
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2269682884216309,
+      "learning_rate": 2.28347928411654e-06,
+      "loss": 0.606,
+      "step": 5404
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3008508682250977,
+      "learning_rate": 2.2817374573550956e-06,
+      "loss": 0.5632,
+      "step": 5405
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3585628271102905,
+      "learning_rate": 2.2799960987591217e-06,
+      "loss": 0.604,
+      "step": 5406
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6831282377243042,
+      "learning_rate": 2.278255208628534e-06,
+      "loss": 0.6772,
+      "step": 5407
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3406838178634644,
+      "learning_rate": 2.2765147872631656e-06,
+      "loss": 0.6374,
+      "step": 5408
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5567865371704102,
+      "learning_rate": 2.274774834962769e-06,
+      "loss": 0.6538,
+      "step": 5409
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.564103364944458,
+      "learning_rate": 2.273035352027013e-06,
+      "loss": 0.6838,
+      "step": 5410
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5919641256332397,
+      "learning_rate": 2.271296338755491e-06,
+      "loss": 0.6681,
+      "step": 5411
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4023327827453613,
+      "learning_rate": 2.269557795447711e-06,
+      "loss": 0.5633,
+      "step": 5412
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1743390560150146,
+      "learning_rate": 2.267819722403104e-06,
+      "loss": 0.519,
+      "step": 5413
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3634856939315796,
+      "learning_rate": 2.2660821199210147e-06,
+      "loss": 0.6134,
+      "step": 5414
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3316655158996582,
+      "learning_rate": 2.2643449883007113e-06,
+      "loss": 0.6312,
+      "step": 5415
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.128868818283081,
+      "learning_rate": 2.26260832784138e-06,
+      "loss": 0.45,
+      "step": 5416
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.427899718284607,
+      "learning_rate": 2.2608721388421233e-06,
+      "loss": 0.6127,
+      "step": 5417
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 16.15768051147461,
+      "learning_rate": 2.259136421601967e-06,
+      "loss": 0.6433,
+      "step": 5418
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.452759027481079,
+      "learning_rate": 2.257401176419851e-06,
+      "loss": 0.5636,
+      "step": 5419
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5467267036437988,
+      "learning_rate": 2.2556664035946396e-06,
+      "loss": 0.6037,
+      "step": 5420
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.701102375984192,
+      "learning_rate": 2.253932103425107e-06,
+      "loss": 0.5748,
+      "step": 5421
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.346367359161377,
+      "learning_rate": 2.2521982762099544e-06,
+      "loss": 0.6572,
+      "step": 5422
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1707795858383179,
+      "learning_rate": 2.2504649222477977e-06,
+      "loss": 0.6369,
+      "step": 5423
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1806703805923462,
+      "learning_rate": 2.248732041837173e-06,
+      "loss": 0.616,
+      "step": 5424
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6231908798217773,
+      "learning_rate": 2.2469996352765307e-06,
+      "loss": 0.5244,
+      "step": 5425
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.244320034980774,
+      "learning_rate": 2.2452677028642445e-06,
+      "loss": 0.6273,
+      "step": 5426
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.8792548179626465,
+      "learning_rate": 2.2435362448986058e-06,
+      "loss": 0.6268,
+      "step": 5427
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.297541618347168,
+      "learning_rate": 2.241805261677821e-06,
+      "loss": 0.5361,
+      "step": 5428
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4641773700714111,
+      "learning_rate": 2.240074753500017e-06,
+      "loss": 0.5059,
+      "step": 5429
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5284756422042847,
+      "learning_rate": 2.2383447206632385e-06,
+      "loss": 0.5743,
+      "step": 5430
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.548749327659607,
+      "learning_rate": 2.2366151634654513e-06,
+      "loss": 0.6469,
+      "step": 5431
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4602359533309937,
+      "learning_rate": 2.2348860822045327e-06,
+      "loss": 0.5945,
+      "step": 5432
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2962963581085205,
+      "learning_rate": 2.233157477178281e-06,
+      "loss": 0.63,
+      "step": 5433
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1280606985092163,
+      "learning_rate": 2.23142934868442e-06,
+      "loss": 0.5783,
+      "step": 5434
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4649392366409302,
+      "learning_rate": 2.229701697020578e-06,
+      "loss": 0.6784,
+      "step": 5435
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.415880799293518,
+      "learning_rate": 2.2279745224843108e-06,
+      "loss": 0.6329,
+      "step": 5436
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3583704233169556,
+      "learning_rate": 2.2262478253730885e-06,
+      "loss": 0.5051,
+      "step": 5437
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3856008052825928,
+      "learning_rate": 2.224521605984302e-06,
+      "loss": 0.6311,
+      "step": 5438
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.266484022140503,
+      "learning_rate": 2.2227958646152525e-06,
+      "loss": 0.6008,
+      "step": 5439
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2216291427612305,
+      "learning_rate": 2.2210706015631666e-06,
+      "loss": 0.5565,
+      "step": 5440
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1825257539749146,
+      "learning_rate": 2.2193458171251857e-06,
+      "loss": 0.6352,
+      "step": 5441
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9359205365180969,
+      "learning_rate": 2.217621511598371e-06,
+      "loss": 0.5556,
+      "step": 5442
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3200385570526123,
+      "learning_rate": 2.2158976852796956e-06,
+      "loss": 0.5984,
+      "step": 5443
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6388661861419678,
+      "learning_rate": 2.2141743384660542e-06,
+      "loss": 0.6564,
+      "step": 5444
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.454256296157837,
+      "learning_rate": 2.212451471454259e-06,
+      "loss": 0.5588,
+      "step": 5445
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3445405960083008,
+      "learning_rate": 2.2107290845410405e-06,
+      "loss": 0.5404,
+      "step": 5446
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1842902898788452,
+      "learning_rate": 2.209007178023042e-06,
+      "loss": 0.6045,
+      "step": 5447
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4644899368286133,
+      "learning_rate": 2.2072857521968276e-06,
+      "loss": 0.5565,
+      "step": 5448
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2165268659591675,
+      "learning_rate": 2.205564807358878e-06,
+      "loss": 0.6259,
+      "step": 5449
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3028157949447632,
+      "learning_rate": 2.203844343805594e-06,
+      "loss": 0.5122,
+      "step": 5450
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.135292410850525,
+      "learning_rate": 2.2021243618332833e-06,
+      "loss": 0.54,
+      "step": 5451
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.196251630783081,
+      "learning_rate": 2.2004048617381844e-06,
+      "loss": 0.4666,
+      "step": 5452
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4005775451660156,
+      "learning_rate": 2.1986858438164464e-06,
+      "loss": 0.5846,
+      "step": 5453
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7641587257385254,
+      "learning_rate": 2.196967308364131e-06,
+      "loss": 0.578,
+      "step": 5454
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2260916233062744,
+      "learning_rate": 2.1952492556772226e-06,
+      "loss": 0.5843,
+      "step": 5455
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4907604455947876,
+      "learning_rate": 2.1935316860516214e-06,
+      "loss": 0.6536,
+      "step": 5456
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4494532346725464,
+      "learning_rate": 2.191814599783146e-06,
+      "loss": 0.5521,
+      "step": 5457
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3400883674621582,
+      "learning_rate": 2.190097997167525e-06,
+      "loss": 0.576,
+      "step": 5458
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1813629865646362,
+      "learning_rate": 2.1883818785004108e-06,
+      "loss": 0.7461,
+      "step": 5459
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.175892949104309,
+      "learning_rate": 2.1866662440773694e-06,
+      "loss": 0.6967,
+      "step": 5460
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5057560205459595,
+      "learning_rate": 2.1849510941938864e-06,
+      "loss": 0.637,
+      "step": 5461
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2532135248184204,
+      "learning_rate": 2.1832364291453572e-06,
+      "loss": 0.6955,
+      "step": 5462
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2681691646575928,
+      "learning_rate": 2.181522249227101e-06,
+      "loss": 0.6327,
+      "step": 5463
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4270204305648804,
+      "learning_rate": 2.179808554734349e-06,
+      "loss": 0.5954,
+      "step": 5464
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4886845350265503,
+      "learning_rate": 2.1780953459622526e-06,
+      "loss": 0.5569,
+      "step": 5465
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3294905424118042,
+      "learning_rate": 2.1763826232058745e-06,
+      "loss": 0.6214,
+      "step": 5466
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3741010427474976,
+      "learning_rate": 2.174670386760197e-06,
+      "loss": 0.6151,
+      "step": 5467
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0721983909606934,
+      "learning_rate": 2.1729586369201184e-06,
+      "loss": 0.7426,
+      "step": 5468
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.39492666721344,
+      "learning_rate": 2.1712473739804524e-06,
+      "loss": 0.6307,
+      "step": 5469
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.437631368637085,
+      "learning_rate": 2.16953659823593e-06,
+      "loss": 0.5988,
+      "step": 5470
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.322919487953186,
+      "learning_rate": 2.1678263099811973e-06,
+      "loss": 0.6527,
+      "step": 5471
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1208932399749756,
+      "learning_rate": 2.166116509510819e-06,
+      "loss": 0.5896,
+      "step": 5472
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3424168825149536,
+      "learning_rate": 2.1644071971192687e-06,
+      "loss": 0.5752,
+      "step": 5473
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3916758298873901,
+      "learning_rate": 2.162698373100943e-06,
+      "loss": 0.6444,
+      "step": 5474
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0904909372329712,
+      "learning_rate": 2.1609900377501526e-06,
+      "loss": 0.6417,
+      "step": 5475
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8144199848175049,
+      "learning_rate": 2.1592821913611246e-06,
+      "loss": 0.5684,
+      "step": 5476
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6254833936691284,
+      "learning_rate": 2.157574834227998e-06,
+      "loss": 0.6259,
+      "step": 5477
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1566519737243652,
+      "learning_rate": 2.1558679666448314e-06,
+      "loss": 0.5346,
+      "step": 5478
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5531538724899292,
+      "learning_rate": 2.1541615889056e-06,
+      "loss": 0.6,
+      "step": 5479
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1958260536193848,
+      "learning_rate": 2.1524557013041897e-06,
+      "loss": 0.6024,
+      "step": 5480
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4889098405838013,
+      "learning_rate": 2.1507503041344063e-06,
+      "loss": 0.6399,
+      "step": 5481
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2659668922424316,
+      "learning_rate": 2.14904539768997e-06,
+      "loss": 0.5371,
+      "step": 5482
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2588521242141724,
+      "learning_rate": 2.1473409822645174e-06,
+      "loss": 0.6951,
+      "step": 5483
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2568881511688232,
+      "learning_rate": 2.1456370581515967e-06,
+      "loss": 0.5739,
+      "step": 5484
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5543744564056396,
+      "learning_rate": 2.143933625644674e-06,
+      "loss": 0.5692,
+      "step": 5485
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.441428542137146,
+      "learning_rate": 2.142230685037136e-06,
+      "loss": 0.6467,
+      "step": 5486
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2142436504364014,
+      "learning_rate": 2.140528236622274e-06,
+      "loss": 0.4952,
+      "step": 5487
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1973832845687866,
+      "learning_rate": 2.138826280693303e-06,
+      "loss": 0.5627,
+      "step": 5488
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.733954668045044,
+      "learning_rate": 2.137124817543349e-06,
+      "loss": 0.604,
+      "step": 5489
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3623613119125366,
+      "learning_rate": 2.1354238474654574e-06,
+      "loss": 0.5886,
+      "step": 5490
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.325785517692566,
+      "learning_rate": 2.133723370752581e-06,
+      "loss": 0.578,
+      "step": 5491
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2375460863113403,
+      "learning_rate": 2.1320233876975955e-06,
+      "loss": 0.5326,
+      "step": 5492
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4318372011184692,
+      "learning_rate": 2.130323898593287e-06,
+      "loss": 0.5755,
+      "step": 5493
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3102978467941284,
+      "learning_rate": 2.128624903732361e-06,
+      "loss": 0.6431,
+      "step": 5494
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4908846616744995,
+      "learning_rate": 2.12692640340743e-06,
+      "loss": 0.5952,
+      "step": 5495
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2359824180603027,
+      "learning_rate": 2.1252283979110292e-06,
+      "loss": 0.6169,
+      "step": 5496
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3551256656646729,
+      "learning_rate": 2.1235308875356048e-06,
+      "loss": 0.6417,
+      "step": 5497
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.139798641204834,
+      "learning_rate": 2.1218338725735203e-06,
+      "loss": 0.5594,
+      "step": 5498
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2601267099380493,
+      "learning_rate": 2.1201373533170484e-06,
+      "loss": 0.6317,
+      "step": 5499
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3188377618789673,
+      "learning_rate": 2.1184413300583823e-06,
+      "loss": 0.6327,
+      "step": 5500
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4169940948486328,
+      "learning_rate": 2.116745803089627e-06,
+      "loss": 0.6216,
+      "step": 5501
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5327837467193604,
+      "learning_rate": 2.115050772702803e-06,
+      "loss": 0.6618,
+      "step": 5502
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3896936178207397,
+      "learning_rate": 2.1133562391898444e-06,
+      "loss": 0.5753,
+      "step": 5503
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2866517305374146,
+      "learning_rate": 2.1116622028426004e-06,
+      "loss": 0.6075,
+      "step": 5504
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4563747644424438,
+      "learning_rate": 2.109968663952836e-06,
+      "loss": 0.6092,
+      "step": 5505
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2761017084121704,
+      "learning_rate": 2.108275622812226e-06,
+      "loss": 0.6336,
+      "step": 5506
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3175703287124634,
+      "learning_rate": 2.1065830797123628e-06,
+      "loss": 0.5825,
+      "step": 5507
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1784989833831787,
+      "learning_rate": 2.1048910349447533e-06,
+      "loss": 0.5514,
+      "step": 5508
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3874294757843018,
+      "learning_rate": 2.1031994888008196e-06,
+      "loss": 0.5477,
+      "step": 5509
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0878545045852661,
+      "learning_rate": 2.1015084415718926e-06,
+      "loss": 0.5669,
+      "step": 5510
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1752663850784302,
+      "learning_rate": 2.0998178935492223e-06,
+      "loss": 0.5278,
+      "step": 5511
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5478566884994507,
+      "learning_rate": 2.098127845023972e-06,
+      "loss": 0.5717,
+      "step": 5512
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4689644575119019,
+      "learning_rate": 2.0964382962872187e-06,
+      "loss": 0.6032,
+      "step": 5513
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3008317947387695,
+      "learning_rate": 2.0947492476299503e-06,
+      "loss": 0.6083,
+      "step": 5514
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0789536237716675,
+      "learning_rate": 2.0930606993430722e-06,
+      "loss": 0.5137,
+      "step": 5515
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4409828186035156,
+      "learning_rate": 2.091372651717404e-06,
+      "loss": 0.6374,
+      "step": 5516
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2810938358306885,
+      "learning_rate": 2.0896851050436774e-06,
+      "loss": 0.6584,
+      "step": 5517
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3894275426864624,
+      "learning_rate": 2.0879980596125353e-06,
+      "loss": 0.5872,
+      "step": 5518
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4865697622299194,
+      "learning_rate": 2.086311515714537e-06,
+      "loss": 0.4925,
+      "step": 5519
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.483933448791504,
+      "learning_rate": 2.084625473640161e-06,
+      "loss": 0.6136,
+      "step": 5520
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3818782567977905,
+      "learning_rate": 2.0829399336797883e-06,
+      "loss": 0.6241,
+      "step": 5521
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9844701290130615,
+      "learning_rate": 2.08125489612372e-06,
+      "loss": 0.641,
+      "step": 5522
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4045764207839966,
+      "learning_rate": 2.079570361262171e-06,
+      "loss": 0.5807,
+      "step": 5523
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5632199048995972,
+      "learning_rate": 2.0778863293852686e-06,
+      "loss": 0.607,
+      "step": 5524
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.273563265800476,
+      "learning_rate": 2.07620280078305e-06,
+      "loss": 0.5969,
+      "step": 5525
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6909412145614624,
+      "learning_rate": 2.0745197757454704e-06,
+      "loss": 0.6227,
+      "step": 5526
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1659879684448242,
+      "learning_rate": 2.0728372545623987e-06,
+      "loss": 0.559,
+      "step": 5527
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.438142776489258,
+      "learning_rate": 2.0711552375236115e-06,
+      "loss": 0.5949,
+      "step": 5528
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3141118288040161,
+      "learning_rate": 2.0694737249188036e-06,
+      "loss": 0.4984,
+      "step": 5529
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1047041416168213,
+      "learning_rate": 2.0677927170375812e-06,
+      "loss": 0.6739,
+      "step": 5530
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4703797101974487,
+      "learning_rate": 2.0661122141694655e-06,
+      "loss": 0.5354,
+      "step": 5531
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1246980428695679,
+      "learning_rate": 2.0644322166038867e-06,
+      "loss": 0.5546,
+      "step": 5532
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.181230902671814,
+      "learning_rate": 2.06275272463019e-06,
+      "loss": 0.6074,
+      "step": 5533
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4038329124450684,
+      "learning_rate": 2.061073738537635e-06,
+      "loss": 0.5793,
+      "step": 5534
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2831275463104248,
+      "learning_rate": 2.059395258615394e-06,
+      "loss": 0.6463,
+      "step": 5535
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1998370885849,
+      "learning_rate": 2.057717285152547e-06,
+      "loss": 0.743,
+      "step": 5536
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8671464920043945,
+      "learning_rate": 2.056039818438095e-06,
+      "loss": 0.5823,
+      "step": 5537
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3943527936935425,
+      "learning_rate": 2.0543628587609472e-06,
+      "loss": 0.5931,
+      "step": 5538
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3470430374145508,
+      "learning_rate": 2.052686406409923e-06,
+      "loss": 0.5477,
+      "step": 5539
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0583560466766357,
+      "learning_rate": 2.0510104616737597e-06,
+      "loss": 0.6012,
+      "step": 5540
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5672448873519897,
+      "learning_rate": 2.0493350248411033e-06,
+      "loss": 0.6135,
+      "step": 5541
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6713660955429077,
+      "learning_rate": 2.047660096200516e-06,
+      "loss": 0.653,
+      "step": 5542
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2293089628219604,
+      "learning_rate": 2.045985676040467e-06,
+      "loss": 0.5229,
+      "step": 5543
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.250230312347412,
+      "learning_rate": 2.044311764649343e-06,
+      "loss": 0.6772,
+      "step": 5544
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6269879341125488,
+      "learning_rate": 2.042638362315441e-06,
+      "loss": 0.6468,
+      "step": 5545
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4112759828567505,
+      "learning_rate": 2.040965469326972e-06,
+      "loss": 0.6261,
+      "step": 5546
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4061042070388794,
+      "learning_rate": 2.0392930859720545e-06,
+      "loss": 0.5911,
+      "step": 5547
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2331053018569946,
+      "learning_rate": 2.037621212538724e-06,
+      "loss": 0.5203,
+      "step": 5548
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3158512115478516,
+      "learning_rate": 2.035949849314928e-06,
+      "loss": 0.4839,
+      "step": 5549
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7110259532928467,
+      "learning_rate": 2.0342789965885246e-06,
+      "loss": 0.601,
+      "step": 5550
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6905916929244995,
+      "learning_rate": 2.0326086546472823e-06,
+      "loss": 0.6474,
+      "step": 5551
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.428787112236023,
+      "learning_rate": 2.0309388237788847e-06,
+      "loss": 0.5905,
+      "step": 5552
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4208601713180542,
+      "learning_rate": 2.029269504270926e-06,
+      "loss": 0.6783,
+      "step": 5553
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5845845937728882,
+      "learning_rate": 2.027600696410913e-06,
+      "loss": 0.6614,
+      "step": 5554
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4801743030548096,
+      "learning_rate": 2.025932400486264e-06,
+      "loss": 0.6549,
+      "step": 5555
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0526586771011353,
+      "learning_rate": 2.0242646167843083e-06,
+      "loss": 0.5649,
+      "step": 5556
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4187215566635132,
+      "learning_rate": 2.022597345592291e-06,
+      "loss": 0.6142,
+      "step": 5557
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.489784836769104,
+      "learning_rate": 2.02093058719736e-06,
+      "loss": 0.5635,
+      "step": 5558
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1324917078018188,
+      "learning_rate": 2.0192643418865843e-06,
+      "loss": 0.512,
+      "step": 5559
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.250471591949463,
+      "learning_rate": 2.0175986099469395e-06,
+      "loss": 0.6027,
+      "step": 5560
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.311180830001831,
+      "learning_rate": 2.0159333916653166e-06,
+      "loss": 0.5209,
+      "step": 5561
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4973576068878174,
+      "learning_rate": 2.0142686873285124e-06,
+      "loss": 0.5737,
+      "step": 5562
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.524202823638916,
+      "learning_rate": 2.012604497223239e-06,
+      "loss": 0.5775,
+      "step": 5563
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5545145273208618,
+      "learning_rate": 2.01094082163612e-06,
+      "loss": 0.6578,
+      "step": 5564
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5071074962615967,
+      "learning_rate": 2.009277660853691e-06,
+      "loss": 0.7238,
+      "step": 5565
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6860618591308594,
+      "learning_rate": 2.007615015162395e-06,
+      "loss": 0.6873,
+      "step": 5566
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1676408052444458,
+      "learning_rate": 2.0059528848485895e-06,
+      "loss": 0.6382,
+      "step": 5567
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1281388998031616,
+      "learning_rate": 2.0042912701985436e-06,
+      "loss": 0.6241,
+      "step": 5568
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5055862665176392,
+      "learning_rate": 2.002630171498438e-06,
+      "loss": 0.6282,
+      "step": 5569
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5386683940887451,
+      "learning_rate": 2.0009695890343583e-06,
+      "loss": 0.6475,
+      "step": 5570
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3176437616348267,
+      "learning_rate": 1.999309523092311e-06,
+      "loss": 0.5033,
+      "step": 5571
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1751028299331665,
+      "learning_rate": 1.997649973958208e-06,
+      "loss": 0.5784,
+      "step": 5572
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2589142322540283,
+      "learning_rate": 1.9959909419178713e-06,
+      "loss": 0.6164,
+      "step": 5573
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3741481304168701,
+      "learning_rate": 1.9943324272570356e-06,
+      "loss": 0.5355,
+      "step": 5574
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4822330474853516,
+      "learning_rate": 1.992674430261349e-06,
+      "loss": 0.5507,
+      "step": 5575
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3982889652252197,
+      "learning_rate": 1.9910169512163634e-06,
+      "loss": 0.6258,
+      "step": 5576
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3766522407531738,
+      "learning_rate": 1.9893599904075485e-06,
+      "loss": 0.594,
+      "step": 5577
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1951097249984741,
+      "learning_rate": 1.987703548120281e-06,
+      "loss": 0.5951,
+      "step": 5578
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.265201449394226,
+      "learning_rate": 1.9860476246398526e-06,
+      "loss": 0.6276,
+      "step": 5579
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.538569927215576,
+      "learning_rate": 1.984392220251458e-06,
+      "loss": 0.6001,
+      "step": 5580
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.507003903388977,
+      "learning_rate": 1.982737335240209e-06,
+      "loss": 0.6099,
+      "step": 5581
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2398253679275513,
+      "learning_rate": 1.9810829698911256e-06,
+      "loss": 0.5759,
+      "step": 5582
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2879706621170044,
+      "learning_rate": 1.9794291244891406e-06,
+      "loss": 0.5272,
+      "step": 5583
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5107345581054688,
+      "learning_rate": 1.977775799319092e-06,
+      "loss": 0.6362,
+      "step": 5584
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.378507137298584,
+      "learning_rate": 1.976122994665733e-06,
+      "loss": 0.6266,
+      "step": 5585
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3231158256530762,
+      "learning_rate": 1.9744707108137257e-06,
+      "loss": 0.6275,
+      "step": 5586
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7076232433319092,
+      "learning_rate": 1.9728189480476436e-06,
+      "loss": 0.6385,
+      "step": 5587
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4472925662994385,
+      "learning_rate": 1.9711677066519645e-06,
+      "loss": 0.6796,
+      "step": 5588
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3523715734481812,
+      "learning_rate": 1.9695169869110864e-06,
+      "loss": 0.6968,
+      "step": 5589
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.466336727142334,
+      "learning_rate": 1.967866789109313e-06,
+      "loss": 0.5789,
+      "step": 5590
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.324497103691101,
+      "learning_rate": 1.9662171135308517e-06,
+      "loss": 0.5396,
+      "step": 5591
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0975810289382935,
+      "learning_rate": 1.9645679604598297e-06,
+      "loss": 0.5384,
+      "step": 5592
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4148670434951782,
+      "learning_rate": 1.9629193301802786e-06,
+      "loss": 0.5924,
+      "step": 5593
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2906337976455688,
+      "learning_rate": 1.9612712229761434e-06,
+      "loss": 0.6046,
+      "step": 5594
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1885030269622803,
+      "learning_rate": 1.959623639131274e-06,
+      "loss": 0.5818,
+      "step": 5595
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5290043354034424,
+      "learning_rate": 1.9579765789294354e-06,
+      "loss": 0.6608,
+      "step": 5596
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3768508434295654,
+      "learning_rate": 1.9563300426542992e-06,
+      "loss": 0.5653,
+      "step": 5597
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3716613054275513,
+      "learning_rate": 1.95468403058945e-06,
+      "loss": 0.666,
+      "step": 5598
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6025148630142212,
+      "learning_rate": 1.9530385430183763e-06,
+      "loss": 0.6309,
+      "step": 5599
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2713145017623901,
+      "learning_rate": 1.9513935802244816e-06,
+      "loss": 0.6173,
+      "step": 5600
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2051814794540405,
+      "learning_rate": 1.9497491424910774e-06,
+      "loss": 0.5658,
+      "step": 5601
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1207643747329712,
+      "learning_rate": 1.948105230101387e-06,
+      "loss": 0.5222,
+      "step": 5602
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8980365991592407,
+      "learning_rate": 1.9464618433385367e-06,
+      "loss": 0.5671,
+      "step": 5603
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4103658199310303,
+      "learning_rate": 1.9448189824855684e-06,
+      "loss": 0.6214,
+      "step": 5604
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3645200729370117,
+      "learning_rate": 1.9431766478254315e-06,
+      "loss": 0.5669,
+      "step": 5605
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2241871356964111,
+      "learning_rate": 1.941534839640985e-06,
+      "loss": 0.5364,
+      "step": 5606
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.241102933883667,
+      "learning_rate": 1.9398935582149964e-06,
+      "loss": 0.6437,
+      "step": 5607
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.057539939880371,
+      "learning_rate": 1.9382528038301437e-06,
+      "loss": 0.622,
+      "step": 5608
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2099688053131104,
+      "learning_rate": 1.9366125767690145e-06,
+      "loss": 0.5541,
+      "step": 5609
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6069209575653076,
+      "learning_rate": 1.9349728773141017e-06,
+      "loss": 0.5802,
+      "step": 5610
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1698077917099,
+      "learning_rate": 1.9333337057478123e-06,
+      "loss": 0.5953,
+      "step": 5611
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2990256547927856,
+      "learning_rate": 1.9316950623524595e-06,
+      "loss": 0.6457,
+      "step": 5612
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0718307495117188,
+      "learning_rate": 1.930056947410268e-06,
+      "loss": 0.5546,
+      "step": 5613
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0749974250793457,
+      "learning_rate": 1.928419361203367e-06,
+      "loss": 0.6744,
+      "step": 5614
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2324750423431396,
+      "learning_rate": 1.9267823040137983e-06,
+      "loss": 0.6432,
+      "step": 5615
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1767945289611816,
+      "learning_rate": 1.9251457761235127e-06,
+      "loss": 0.6066,
+      "step": 5616
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.249362826347351,
+      "learning_rate": 1.92350977781437e-06,
+      "loss": 0.5928,
+      "step": 5617
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3877533674240112,
+      "learning_rate": 1.9218743093681346e-06,
+      "loss": 0.5676,
+      "step": 5618
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1825809478759766,
+      "learning_rate": 1.920239371066484e-06,
+      "loss": 0.6075,
+      "step": 5619
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3485913276672363,
+      "learning_rate": 1.9186049631910047e-06,
+      "loss": 0.616,
+      "step": 5620
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5328774452209473,
+      "learning_rate": 1.916971086023188e-06,
+      "loss": 0.6823,
+      "step": 5621
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6213551759719849,
+      "learning_rate": 1.915337739844434e-06,
+      "loss": 0.5982,
+      "step": 5622
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1525039672851562,
+      "learning_rate": 1.9137049249360583e-06,
+      "loss": 0.568,
+      "step": 5623
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1615984439849854,
+      "learning_rate": 1.9120726415792795e-06,
+      "loss": 0.6214,
+      "step": 5624
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1085628271102905,
+      "learning_rate": 1.9104408900552226e-06,
+      "loss": 0.6147,
+      "step": 5625
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7933579683303833,
+      "learning_rate": 1.9088096706449245e-06,
+      "loss": 0.6051,
+      "step": 5626
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3298068046569824,
+      "learning_rate": 1.9071789836293313e-06,
+      "loss": 0.5936,
+      "step": 5627
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3054797649383545,
+      "learning_rate": 1.9055488292892927e-06,
+      "loss": 0.6075,
+      "step": 5628
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7392698526382446,
+      "learning_rate": 1.9039192079055712e-06,
+      "loss": 0.6097,
+      "step": 5629
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3165106773376465,
+      "learning_rate": 1.902290119758836e-06,
+      "loss": 0.526,
+      "step": 5630
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.294399380683899,
+      "learning_rate": 1.9006615651296662e-06,
+      "loss": 0.4764,
+      "step": 5631
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2079136371612549,
+      "learning_rate": 1.8990335442985436e-06,
+      "loss": 0.6266,
+      "step": 5632
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8871737718582153,
+      "learning_rate": 1.897406057545863e-06,
+      "loss": 0.6647,
+      "step": 5633
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2210822105407715,
+      "learning_rate": 1.895779105151927e-06,
+      "loss": 0.5954,
+      "step": 5634
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3488540649414062,
+      "learning_rate": 1.8941526873969463e-06,
+      "loss": 0.5809,
+      "step": 5635
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.655834674835205,
+      "learning_rate": 1.8925268045610345e-06,
+      "loss": 0.5514,
+      "step": 5636
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7508749961853027,
+      "learning_rate": 1.890901456924219e-06,
+      "loss": 0.5246,
+      "step": 5637
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.218916416168213,
+      "learning_rate": 1.8892766447664323e-06,
+      "loss": 0.5255,
+      "step": 5638
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2823469638824463,
+      "learning_rate": 1.8876523683675163e-06,
+      "loss": 0.6096,
+      "step": 5639
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.282475233078003,
+      "learning_rate": 1.8860286280072187e-06,
+      "loss": 0.5822,
+      "step": 5640
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4252816438674927,
+      "learning_rate": 1.884405423965196e-06,
+      "loss": 0.5581,
+      "step": 5641
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3148558139801025,
+      "learning_rate": 1.8827827565210143e-06,
+      "loss": 0.5631,
+      "step": 5642
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.354689359664917,
+      "learning_rate": 1.881160625954141e-06,
+      "loss": 0.5763,
+      "step": 5643
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8449751138687134,
+      "learning_rate": 1.8795390325439572e-06,
+      "loss": 0.6018,
+      "step": 5644
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3221594095230103,
+      "learning_rate": 1.8779179765697491e-06,
+      "loss": 0.6552,
+      "step": 5645
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4569640159606934,
+      "learning_rate": 1.8762974583107129e-06,
+      "loss": 0.5821,
+      "step": 5646
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3788806200027466,
+      "learning_rate": 1.8746774780459465e-06,
+      "loss": 0.5294,
+      "step": 5647
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5517175197601318,
+      "learning_rate": 1.8730580360544593e-06,
+      "loss": 0.6109,
+      "step": 5648
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6753023862838745,
+      "learning_rate": 1.8714391326151681e-06,
+      "loss": 0.5427,
+      "step": 5649
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5441654920578003,
+      "learning_rate": 1.8698207680068974e-06,
+      "loss": 0.6389,
+      "step": 5650
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1922584772109985,
+      "learning_rate": 1.8682029425083748e-06,
+      "loss": 0.5781,
+      "step": 5651
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3633989095687866,
+      "learning_rate": 1.8665856563982392e-06,
+      "loss": 0.5629,
+      "step": 5652
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3253165483474731,
+      "learning_rate": 1.864968909955035e-06,
+      "loss": 0.554,
+      "step": 5653
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3885520696640015,
+      "learning_rate": 1.8633527034572164e-06,
+      "loss": 0.6109,
+      "step": 5654
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.899064302444458,
+      "learning_rate": 1.8617370371831373e-06,
+      "loss": 0.6412,
+      "step": 5655
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9599201679229736,
+      "learning_rate": 1.8601219114110646e-06,
+      "loss": 0.5653,
+      "step": 5656
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2851388454437256,
+      "learning_rate": 1.858507326419176e-06,
+      "loss": 0.5802,
+      "step": 5657
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6403940916061401,
+      "learning_rate": 1.8568932824855457e-06,
+      "loss": 0.5969,
+      "step": 5658
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4266154766082764,
+      "learning_rate": 1.8552797798881611e-06,
+      "loss": 0.554,
+      "step": 5659
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2188271284103394,
+      "learning_rate": 1.8536668189049156e-06,
+      "loss": 0.5139,
+      "step": 5660
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4466283321380615,
+      "learning_rate": 1.8520543998136104e-06,
+      "loss": 0.6892,
+      "step": 5661
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.507656455039978,
+      "learning_rate": 1.8504425228919492e-06,
+      "loss": 0.6445,
+      "step": 5662
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2851141691207886,
+      "learning_rate": 1.8488311884175458e-06,
+      "loss": 0.69,
+      "step": 5663
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.239793062210083,
+      "learning_rate": 1.8472203966679208e-06,
+      "loss": 0.65,
+      "step": 5664
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2325901985168457,
+      "learning_rate": 1.8456101479205012e-06,
+      "loss": 0.5284,
+      "step": 5665
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.43928861618042,
+      "learning_rate": 1.8440004424526165e-06,
+      "loss": 0.6087,
+      "step": 5666
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4355437755584717,
+      "learning_rate": 1.8423912805415078e-06,
+      "loss": 0.6277,
+      "step": 5667
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8221622705459595,
+      "learning_rate": 1.84078266246432e-06,
+      "loss": 0.5687,
+      "step": 5668
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6198118925094604,
+      "learning_rate": 1.839174588498107e-06,
+      "loss": 0.5901,
+      "step": 5669
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7814929485321045,
+      "learning_rate": 1.837567058919823e-06,
+      "loss": 0.5777,
+      "step": 5670
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.416741967201233,
+      "learning_rate": 1.8359600740063344e-06,
+      "loss": 0.5138,
+      "step": 5671
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3309919834136963,
+      "learning_rate": 1.8343536340344136e-06,
+      "loss": 0.5628,
+      "step": 5672
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2797962427139282,
+      "learning_rate": 1.8327477392807314e-06,
+      "loss": 0.5647,
+      "step": 5673
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2392196655273438,
+      "learning_rate": 1.831142390021876e-06,
+      "loss": 0.5267,
+      "step": 5674
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2258974313735962,
+      "learning_rate": 1.8295375865343363e-06,
+      "loss": 0.631,
+      "step": 5675
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6016523838043213,
+      "learning_rate": 1.8279333290945035e-06,
+      "loss": 0.6077,
+      "step": 5676
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2385517358779907,
+      "learning_rate": 1.8263296179786798e-06,
+      "loss": 0.6142,
+      "step": 5677
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3300591707229614,
+      "learning_rate": 1.8247264534630727e-06,
+      "loss": 0.5558,
+      "step": 5678
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3536458015441895,
+      "learning_rate": 1.8231238358237952e-06,
+      "loss": 0.5652,
+      "step": 5679
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.294861078262329,
+      "learning_rate": 1.8215217653368632e-06,
+      "loss": 0.6216,
+      "step": 5680
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1894181966781616,
+      "learning_rate": 1.8199202422782026e-06,
+      "loss": 0.5569,
+      "step": 5681
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3334403038024902,
+      "learning_rate": 1.8183192669236422e-06,
+      "loss": 0.6186,
+      "step": 5682
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4797943830490112,
+      "learning_rate": 1.8167188395489194e-06,
+      "loss": 0.5627,
+      "step": 5683
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.333863377571106,
+      "learning_rate": 1.8151189604296727e-06,
+      "loss": 0.6105,
+      "step": 5684
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8466845750808716,
+      "learning_rate": 1.8135196298414498e-06,
+      "loss": 0.7075,
+      "step": 5685
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2560466527938843,
+      "learning_rate": 1.8119208480597033e-06,
+      "loss": 0.6165,
+      "step": 5686
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3292875289916992,
+      "learning_rate": 1.8103226153597919e-06,
+      "loss": 0.6225,
+      "step": 5687
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.3358840942382812,
+      "learning_rate": 1.8087249320169758e-06,
+      "loss": 0.6418,
+      "step": 5688
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.621479868888855,
+      "learning_rate": 1.807127798306425e-06,
+      "loss": 0.535,
+      "step": 5689
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5514147281646729,
+      "learning_rate": 1.8055312145032139e-06,
+      "loss": 0.6259,
+      "step": 5690
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.329399585723877,
+      "learning_rate": 1.8039351808823203e-06,
+      "loss": 0.56,
+      "step": 5691
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4586081504821777,
+      "learning_rate": 1.8023396977186297e-06,
+      "loss": 0.5567,
+      "step": 5692
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.481757402420044,
+      "learning_rate": 1.8007447652869314e-06,
+      "loss": 0.6491,
+      "step": 5693
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.280112624168396,
+      "learning_rate": 1.7991503838619219e-06,
+      "loss": 0.6109,
+      "step": 5694
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3548251390457153,
+      "learning_rate": 1.7975565537181971e-06,
+      "loss": 0.5796,
+      "step": 5695
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.166089415550232,
+      "learning_rate": 1.7959632751302636e-06,
+      "loss": 0.5353,
+      "step": 5696
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4079132080078125,
+      "learning_rate": 1.794370548372532e-06,
+      "loss": 0.5729,
+      "step": 5697
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 6.476421356201172,
+      "learning_rate": 1.7927783737193182e-06,
+      "loss": 0.6065,
+      "step": 5698
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1923269033432007,
+      "learning_rate": 1.7911867514448384e-06,
+      "loss": 0.614,
+      "step": 5699
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.501766324043274,
+      "learning_rate": 1.7895956818232191e-06,
+      "loss": 0.6037,
+      "step": 5700
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4375709295272827,
+      "learning_rate": 1.7880051651284897e-06,
+      "loss": 0.6155,
+      "step": 5701
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1795188188552856,
+      "learning_rate": 1.786415201634586e-06,
+      "loss": 0.606,
+      "step": 5702
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.210554599761963,
+      "learning_rate": 1.784825791615344e-06,
+      "loss": 0.5175,
+      "step": 5703
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2811566591262817,
+      "learning_rate": 1.7832369353445078e-06,
+      "loss": 0.5747,
+      "step": 5704
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2850680351257324,
+      "learning_rate": 1.7816486330957272e-06,
+      "loss": 0.6138,
+      "step": 5705
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2182974815368652,
+      "learning_rate": 1.780060885142555e-06,
+      "loss": 0.5599,
+      "step": 5706
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3909426927566528,
+      "learning_rate": 1.778473691758445e-06,
+      "loss": 0.5889,
+      "step": 5707
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6075366735458374,
+      "learning_rate": 1.7768870532167625e-06,
+      "loss": 0.5588,
+      "step": 5708
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3977075815200806,
+      "learning_rate": 1.7753009697907753e-06,
+      "loss": 0.5676,
+      "step": 5709
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.174627661705017,
+      "learning_rate": 1.7737154417536495e-06,
+      "loss": 0.5714,
+      "step": 5710
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6572093963623047,
+      "learning_rate": 1.7721304693784624e-06,
+      "loss": 0.6938,
+      "step": 5711
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2030808925628662,
+      "learning_rate": 1.770546052938193e-06,
+      "loss": 0.5375,
+      "step": 5712
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3561108112335205,
+      "learning_rate": 1.7689621927057265e-06,
+      "loss": 0.5343,
+      "step": 5713
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1456682682037354,
+      "learning_rate": 1.7673788889538473e-06,
+      "loss": 0.5621,
+      "step": 5714
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3912715911865234,
+      "learning_rate": 1.7657961419552488e-06,
+      "loss": 0.5994,
+      "step": 5715
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5939741134643555,
+      "learning_rate": 1.7642139519825274e-06,
+      "loss": 0.5317,
+      "step": 5716
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3256257772445679,
+      "learning_rate": 1.7626323193081852e-06,
+      "loss": 0.5304,
+      "step": 5717
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.559259057044983,
+      "learning_rate": 1.761051244204622e-06,
+      "loss": 0.6115,
+      "step": 5718
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6052910089492798,
+      "learning_rate": 1.7594707269441473e-06,
+      "loss": 0.5905,
+      "step": 5719
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3079288005828857,
+      "learning_rate": 1.757890767798976e-06,
+      "loss": 0.6405,
+      "step": 5720
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5783737897872925,
+      "learning_rate": 1.7563113670412197e-06,
+      "loss": 0.6213,
+      "step": 5721
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4406654834747314,
+      "learning_rate": 1.7547325249428999e-06,
+      "loss": 0.6082,
+      "step": 5722
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1159170866012573,
+      "learning_rate": 1.7531542417759401e-06,
+      "loss": 0.6584,
+      "step": 5723
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4534189701080322,
+      "learning_rate": 1.7515765178121696e-06,
+      "loss": 0.5921,
+      "step": 5724
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9966347217559814,
+      "learning_rate": 1.7499993533233134e-06,
+      "loss": 0.6042,
+      "step": 5725
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3243257999420166,
+      "learning_rate": 1.7484227485810119e-06,
+      "loss": 0.5952,
+      "step": 5726
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.409643530845642,
+      "learning_rate": 1.7468467038568033e-06,
+      "loss": 0.5421,
+      "step": 5727
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2723798751831055,
+      "learning_rate": 1.7452712194221243e-06,
+      "loss": 0.647,
+      "step": 5728
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3207684755325317,
+      "learning_rate": 1.743696295548324e-06,
+      "loss": 0.5333,
+      "step": 5729
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3573734760284424,
+      "learning_rate": 1.7421219325066495e-06,
+      "loss": 0.5481,
+      "step": 5730
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.479612946510315,
+      "learning_rate": 1.7405481305682547e-06,
+      "loss": 0.5887,
+      "step": 5731
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5791213512420654,
+      "learning_rate": 1.7389748900041926e-06,
+      "loss": 0.5848,
+      "step": 5732
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5573316812515259,
+      "learning_rate": 1.7374022110854222e-06,
+      "loss": 0.6232,
+      "step": 5733
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2642827033996582,
+      "learning_rate": 1.7358300940828067e-06,
+      "loss": 0.5412,
+      "step": 5734
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6646677255630493,
+      "learning_rate": 1.7342585392671117e-06,
+      "loss": 0.5998,
+      "step": 5735
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2101836204528809,
+      "learning_rate": 1.7326875469090037e-06,
+      "loss": 0.582,
+      "step": 5736
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.327369213104248,
+      "learning_rate": 1.7311171172790548e-06,
+      "loss": 0.5807,
+      "step": 5737
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1370022296905518,
+      "learning_rate": 1.72954725064774e-06,
+      "loss": 0.503,
+      "step": 5738
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4957923889160156,
+      "learning_rate": 1.7279779472854386e-06,
+      "loss": 0.5385,
+      "step": 5739
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2059375047683716,
+      "learning_rate": 1.7264092074624278e-06,
+      "loss": 0.5874,
+      "step": 5740
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1997199058532715,
+      "learning_rate": 1.7248410314488928e-06,
+      "loss": 0.5358,
+      "step": 5741
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4809540510177612,
+      "learning_rate": 1.7232734195149197e-06,
+      "loss": 0.5952,
+      "step": 5742
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9461658000946045,
+      "learning_rate": 1.7217063719304988e-06,
+      "loss": 0.6296,
+      "step": 5743
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.942325472831726,
+      "learning_rate": 1.7201398889655207e-06,
+      "loss": 0.6345,
+      "step": 5744
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3077179193496704,
+      "learning_rate": 1.718573970889782e-06,
+      "loss": 0.7035,
+      "step": 5745
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5481005907058716,
+      "learning_rate": 1.71700861797298e-06,
+      "loss": 0.5672,
+      "step": 5746
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3350780010223389,
+      "learning_rate": 1.7154438304847132e-06,
+      "loss": 0.5974,
+      "step": 5747
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4484832286834717,
+      "learning_rate": 1.7138796086944854e-06,
+      "loss": 0.6045,
+      "step": 5748
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4450817108154297,
+      "learning_rate": 1.712315952871702e-06,
+      "loss": 0.5568,
+      "step": 5749
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2990732192993164,
+      "learning_rate": 1.7107528632856729e-06,
+      "loss": 0.541,
+      "step": 5750
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9606150388717651,
+      "learning_rate": 1.7091903402056054e-06,
+      "loss": 0.584,
+      "step": 5751
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5782921314239502,
+      "learning_rate": 1.707628383900613e-06,
+      "loss": 0.5366,
+      "step": 5752
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3492310047149658,
+      "learning_rate": 1.7060669946397112e-06,
+      "loss": 0.6352,
+      "step": 5753
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2007477283477783,
+      "learning_rate": 1.7045061726918204e-06,
+      "loss": 0.5359,
+      "step": 5754
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.260575771331787,
+      "learning_rate": 1.7029459183257557e-06,
+      "loss": 0.5743,
+      "step": 5755
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4031670093536377,
+      "learning_rate": 1.7013862318102415e-06,
+      "loss": 0.5996,
+      "step": 5756
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4298655986785889,
+      "learning_rate": 1.6998271134139022e-06,
+      "loss": 0.627,
+      "step": 5757
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.1103055477142334,
+      "learning_rate": 1.698268563405266e-06,
+      "loss": 0.5912,
+      "step": 5758
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.184146761894226,
+      "learning_rate": 1.6967105820527558e-06,
+      "loss": 0.6081,
+      "step": 5759
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3072179555892944,
+      "learning_rate": 1.6951531696247075e-06,
+      "loss": 0.6095,
+      "step": 5760
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2722811698913574,
+      "learning_rate": 1.6935963263893546e-06,
+      "loss": 0.4975,
+      "step": 5761
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.443372368812561,
+      "learning_rate": 1.6920400526148272e-06,
+      "loss": 0.6409,
+      "step": 5762
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.277919054031372,
+      "learning_rate": 1.6904843485691635e-06,
+      "loss": 0.6582,
+      "step": 5763
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1344646215438843,
+      "learning_rate": 1.6889292145203028e-06,
+      "loss": 0.5916,
+      "step": 5764
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1989878416061401,
+      "learning_rate": 1.6873746507360865e-06,
+      "loss": 0.4848,
+      "step": 5765
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2986321449279785,
+      "learning_rate": 1.685820657484254e-06,
+      "loss": 0.6413,
+      "step": 5766
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6236834526062012,
+      "learning_rate": 1.684267235032449e-06,
+      "loss": 0.6122,
+      "step": 5767
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3555066585540771,
+      "learning_rate": 1.68271438364822e-06,
+      "loss": 0.5595,
+      "step": 5768
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2782316207885742,
+      "learning_rate": 1.6811621035990106e-06,
+      "loss": 0.7395,
+      "step": 5769
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4683347940444946,
+      "learning_rate": 1.6796103951521708e-06,
+      "loss": 0.5275,
+      "step": 5770
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3068945407867432,
+      "learning_rate": 1.678059258574951e-06,
+      "loss": 0.6087,
+      "step": 5771
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8507400751113892,
+      "learning_rate": 1.676508694134505e-06,
+      "loss": 0.5883,
+      "step": 5772
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1494866609573364,
+      "learning_rate": 1.6749587020978814e-06,
+      "loss": 0.5752,
+      "step": 5773
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5824859142303467,
+      "learning_rate": 1.673409282732038e-06,
+      "loss": 0.5977,
+      "step": 5774
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.35085129737854,
+      "learning_rate": 1.67186043630383e-06,
+      "loss": 0.5602,
+      "step": 5775
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6015706062316895,
+      "learning_rate": 1.670312163080015e-06,
+      "loss": 0.598,
+      "step": 5776
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6979427337646484,
+      "learning_rate": 1.6687644633272516e-06,
+      "loss": 0.6116,
+      "step": 5777
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.282360315322876,
+      "learning_rate": 1.6672173373120992e-06,
+      "loss": 0.5501,
+      "step": 5778
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.518515944480896,
+      "learning_rate": 1.6656707853010207e-06,
+      "loss": 0.564,
+      "step": 5779
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1478134393692017,
+      "learning_rate": 1.6641248075603756e-06,
+      "loss": 0.7633,
+      "step": 5780
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7522804737091064,
+      "learning_rate": 1.6625794043564275e-06,
+      "loss": 0.517,
+      "step": 5781
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3169957399368286,
+      "learning_rate": 1.661034575955342e-06,
+      "loss": 0.6518,
+      "step": 5782
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.0078976154327393,
+      "learning_rate": 1.659490322623185e-06,
+      "loss": 0.5437,
+      "step": 5783
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5404016971588135,
+      "learning_rate": 1.6579466446259201e-06,
+      "loss": 0.6104,
+      "step": 5784
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.531265377998352,
+      "learning_rate": 1.6564035422294156e-06,
+      "loss": 0.6512,
+      "step": 5785
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.444911241531372,
+      "learning_rate": 1.65486101569944e-06,
+      "loss": 0.5998,
+      "step": 5786
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3478665351867676,
+      "learning_rate": 1.653319065301664e-06,
+      "loss": 0.6803,
+      "step": 5787
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3380131721496582,
+      "learning_rate": 1.651777691301653e-06,
+      "loss": 0.6165,
+      "step": 5788
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.335648536682129,
+      "learning_rate": 1.6502368939648794e-06,
+      "loss": 0.6804,
+      "step": 5789
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.409081220626831,
+      "learning_rate": 1.6486966735567144e-06,
+      "loss": 0.5307,
+      "step": 5790
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3683109283447266,
+      "learning_rate": 1.6471570303424318e-06,
+      "loss": 0.5651,
+      "step": 5791
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4809428453445435,
+      "learning_rate": 1.6456179645871996e-06,
+      "loss": 0.5627,
+      "step": 5792
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6425578594207764,
+      "learning_rate": 1.6440794765560913e-06,
+      "loss": 0.6125,
+      "step": 5793
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5959701538085938,
+      "learning_rate": 1.6425415665140842e-06,
+      "loss": 0.507,
+      "step": 5794
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0912171602249146,
+      "learning_rate": 1.6410042347260486e-06,
+      "loss": 0.6124,
+      "step": 5795
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6113317012786865,
+      "learning_rate": 1.6394674814567585e-06,
+      "loss": 0.5317,
+      "step": 5796
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3795167207717896,
+      "learning_rate": 1.6379313069708896e-06,
+      "loss": 0.665,
+      "step": 5797
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1050100326538086,
+      "learning_rate": 1.6363957115330187e-06,
+      "loss": 0.6964,
+      "step": 5798
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3911715745925903,
+      "learning_rate": 1.6348606954076169e-06,
+      "loss": 0.6012,
+      "step": 5799
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1959513425827026,
+      "learning_rate": 1.6333262588590609e-06,
+      "loss": 0.7361,
+      "step": 5800
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2428171634674072,
+      "learning_rate": 1.631792402151627e-06,
+      "loss": 0.62,
+      "step": 5801
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.339919328689575,
+      "learning_rate": 1.6302591255494916e-06,
+      "loss": 0.6668,
+      "step": 5802
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3819876909255981,
+      "learning_rate": 1.6287264293167277e-06,
+      "loss": 0.6304,
+      "step": 5803
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.578447937965393,
+      "learning_rate": 1.627194313717313e-06,
+      "loss": 0.6149,
+      "step": 5804
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1103376150131226,
+      "learning_rate": 1.6256627790151224e-06,
+      "loss": 0.5619,
+      "step": 5805
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9362802505493164,
+      "learning_rate": 1.6241318254739346e-06,
+      "loss": 0.5938,
+      "step": 5806
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4348499774932861,
+      "learning_rate": 1.6226014533574208e-06,
+      "loss": 0.6009,
+      "step": 5807
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4850698709487915,
+      "learning_rate": 1.6210716629291584e-06,
+      "loss": 0.6183,
+      "step": 5808
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.44414484500885,
+      "learning_rate": 1.6195424544526228e-06,
+      "loss": 0.6009,
+      "step": 5809
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3758606910705566,
+      "learning_rate": 1.6180138281911889e-06,
+      "loss": 0.6444,
+      "step": 5810
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1687239408493042,
+      "learning_rate": 1.6164857844081316e-06,
+      "loss": 0.4735,
+      "step": 5811
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3271052837371826,
+      "learning_rate": 1.6149583233666254e-06,
+      "loss": 0.6004,
+      "step": 5812
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2638598680496216,
+      "learning_rate": 1.6134314453297467e-06,
+      "loss": 0.7543,
+      "step": 5813
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4526175260543823,
+      "learning_rate": 1.611905150560465e-06,
+      "loss": 0.5601,
+      "step": 5814
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.159395694732666,
+      "learning_rate": 1.6103794393216554e-06,
+      "loss": 0.5431,
+      "step": 5815
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4904710054397583,
+      "learning_rate": 1.6088543118760908e-06,
+      "loss": 0.5712,
+      "step": 5816
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4881311655044556,
+      "learning_rate": 1.6073297684864453e-06,
+      "loss": 0.5665,
+      "step": 5817
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.3589932918548584,
+      "learning_rate": 1.6058058094152862e-06,
+      "loss": 0.6032,
+      "step": 5818
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4021669626235962,
+      "learning_rate": 1.6042824349250873e-06,
+      "loss": 0.6179,
+      "step": 5819
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3111644983291626,
+      "learning_rate": 1.6027596452782202e-06,
+      "loss": 0.5615,
+      "step": 5820
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3525159358978271,
+      "learning_rate": 1.6012374407369514e-06,
+      "loss": 0.5584,
+      "step": 5821
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5664325952529907,
+      "learning_rate": 1.5997158215634506e-06,
+      "loss": 0.5334,
+      "step": 5822
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4455913305282593,
+      "learning_rate": 1.5981947880197862e-06,
+      "loss": 0.6552,
+      "step": 5823
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6153526306152344,
+      "learning_rate": 1.596674340367927e-06,
+      "loss": 0.5421,
+      "step": 5824
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.32578706741333,
+      "learning_rate": 1.5951544788697354e-06,
+      "loss": 0.621,
+      "step": 5825
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0666056871414185,
+      "learning_rate": 1.5936352037869796e-06,
+      "loss": 0.6811,
+      "step": 5826
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.09562349319458,
+      "learning_rate": 1.5921165153813222e-06,
+      "loss": 0.6983,
+      "step": 5827
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.482304573059082,
+      "learning_rate": 1.5905984139143277e-06,
+      "loss": 0.5927,
+      "step": 5828
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3417950868606567,
+      "learning_rate": 1.5890808996474576e-06,
+      "loss": 0.593,
+      "step": 5829
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6582002639770508,
+      "learning_rate": 1.5875639728420727e-06,
+      "loss": 0.5227,
+      "step": 5830
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.452853798866272,
+      "learning_rate": 1.586047633759435e-06,
+      "loss": 0.6726,
+      "step": 5831
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5506370067596436,
+      "learning_rate": 1.5845318826606997e-06,
+      "loss": 0.5796,
+      "step": 5832
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3531577587127686,
+      "learning_rate": 1.5830167198069256e-06,
+      "loss": 0.6651,
+      "step": 5833
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3325982093811035,
+      "learning_rate": 1.581502145459069e-06,
+      "loss": 0.6107,
+      "step": 5834
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6698253154754639,
+      "learning_rate": 1.5799881598779853e-06,
+      "loss": 0.6907,
+      "step": 5835
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4126887321472168,
+      "learning_rate": 1.5784747633244257e-06,
+      "loss": 0.6099,
+      "step": 5836
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.450058937072754,
+      "learning_rate": 1.5769619560590426e-06,
+      "loss": 0.6664,
+      "step": 5837
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.739638805389404,
+      "learning_rate": 1.5754497383423866e-06,
+      "loss": 0.5682,
+      "step": 5838
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2308710813522339,
+      "learning_rate": 1.573938110434909e-06,
+      "loss": 0.568,
+      "step": 5839
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7075868844985962,
+      "learning_rate": 1.5724270725969521e-06,
+      "loss": 0.5979,
+      "step": 5840
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3854655027389526,
+      "learning_rate": 1.570916625088764e-06,
+      "loss": 0.5529,
+      "step": 5841
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.426127314567566,
+      "learning_rate": 1.5694067681704888e-06,
+      "loss": 0.5614,
+      "step": 5842
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2803055047988892,
+      "learning_rate": 1.5678975021021703e-06,
+      "loss": 0.6027,
+      "step": 5843
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.465036392211914,
+      "learning_rate": 1.5663888271437434e-06,
+      "loss": 0.6433,
+      "step": 5844
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2361197471618652,
+      "learning_rate": 1.5648807435550518e-06,
+      "loss": 0.643,
+      "step": 5845
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3847153186798096,
+      "learning_rate": 1.5633732515958322e-06,
+      "loss": 0.6341,
+      "step": 5846
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.526636004447937,
+      "learning_rate": 1.5618663515257166e-06,
+      "loss": 0.5841,
+      "step": 5847
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0692555904388428,
+      "learning_rate": 1.5603600436042393e-06,
+      "loss": 0.6289,
+      "step": 5848
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4534244537353516,
+      "learning_rate": 1.5588543280908309e-06,
+      "loss": 0.5165,
+      "step": 5849
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.500308632850647,
+      "learning_rate": 1.5573492052448226e-06,
+      "loss": 0.6331,
+      "step": 5850
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.619732141494751,
+      "learning_rate": 1.5558446753254374e-06,
+      "loss": 0.6026,
+      "step": 5851
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5240439176559448,
+      "learning_rate": 1.5543407385918019e-06,
+      "loss": 0.5902,
+      "step": 5852
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.284821629524231,
+      "learning_rate": 1.5528373953029386e-06,
+      "loss": 0.5572,
+      "step": 5853
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4876408576965332,
+      "learning_rate": 1.5513346457177692e-06,
+      "loss": 0.55,
+      "step": 5854
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.397421956062317,
+      "learning_rate": 1.5498324900951083e-06,
+      "loss": 0.5822,
+      "step": 5855
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2312592267990112,
+      "learning_rate": 1.5483309286936743e-06,
+      "loss": 0.5442,
+      "step": 5856
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2165515422821045,
+      "learning_rate": 1.54682996177208e-06,
+      "loss": 0.6374,
+      "step": 5857
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3601298332214355,
+      "learning_rate": 1.5453295895888382e-06,
+      "loss": 0.5892,
+      "step": 5858
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.664817214012146,
+      "learning_rate": 1.5438298124023537e-06,
+      "loss": 0.5797,
+      "step": 5859
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.5244088172912598,
+      "learning_rate": 1.542330630470935e-06,
+      "loss": 0.4757,
+      "step": 5860
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2353243827819824,
+      "learning_rate": 1.5408320440527874e-06,
+      "loss": 0.5268,
+      "step": 5861
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2477408647537231,
+      "learning_rate": 1.5393340534060064e-06,
+      "loss": 0.6622,
+      "step": 5862
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1838089227676392,
+      "learning_rate": 1.5378366587885956e-06,
+      "loss": 0.5516,
+      "step": 5863
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4416934251785278,
+      "learning_rate": 1.5363398604584496e-06,
+      "loss": 0.5599,
+      "step": 5864
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2726426124572754,
+      "learning_rate": 1.5348436586733623e-06,
+      "loss": 0.5677,
+      "step": 5865
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5306282043457031,
+      "learning_rate": 1.5333480536910217e-06,
+      "loss": 0.6393,
+      "step": 5866
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.202629566192627,
+      "learning_rate": 1.5318530457690163e-06,
+      "loss": 0.6512,
+      "step": 5867
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.521677851676941,
+      "learning_rate": 1.5303586351648326e-06,
+      "loss": 0.6393,
+      "step": 5868
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.428450107574463,
+      "learning_rate": 1.5288648221358488e-06,
+      "loss": 0.6125,
+      "step": 5869
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4109379053115845,
+      "learning_rate": 1.5273716069393467e-06,
+      "loss": 0.5075,
+      "step": 5870
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4165418148040771,
+      "learning_rate": 1.525878989832501e-06,
+      "loss": 0.6775,
+      "step": 5871
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4802438020706177,
+      "learning_rate": 1.5243869710723875e-06,
+      "loss": 0.6267,
+      "step": 5872
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.338913917541504,
+      "learning_rate": 1.5228955509159715e-06,
+      "loss": 0.6234,
+      "step": 5873
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6580629348754883,
+      "learning_rate": 1.5214047296201218e-06,
+      "loss": 0.5374,
+      "step": 5874
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2051814794540405,
+      "learning_rate": 1.5199145074416028e-06,
+      "loss": 0.5343,
+      "step": 5875
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2296867370605469,
+      "learning_rate": 1.5184248846370763e-06,
+      "loss": 0.6521,
+      "step": 5876
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.1033120155334473,
+      "learning_rate": 1.5169358614630958e-06,
+      "loss": 0.66,
+      "step": 5877
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2728708982467651,
+      "learning_rate": 1.5154474381761175e-06,
+      "loss": 0.6069,
+      "step": 5878
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.303300380706787,
+      "learning_rate": 1.5139596150324915e-06,
+      "loss": 0.5893,
+      "step": 5879
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.422898530960083,
+      "learning_rate": 1.5124723922884655e-06,
+      "loss": 0.5823,
+      "step": 5880
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.917449712753296,
+      "learning_rate": 1.5109857702001834e-06,
+      "loss": 0.6907,
+      "step": 5881
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2992802858352661,
+      "learning_rate": 1.5094997490236857e-06,
+      "loss": 0.6293,
+      "step": 5882
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5122466087341309,
+      "learning_rate": 1.50801432901491e-06,
+      "loss": 0.6306,
+      "step": 5883
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4652949571609497,
+      "learning_rate": 1.5065295104296884e-06,
+      "loss": 0.5773,
+      "step": 5884
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2318650484085083,
+      "learning_rate": 1.5050452935237502e-06,
+      "loss": 0.5216,
+      "step": 5885
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5187735557556152,
+      "learning_rate": 1.5035616785527235e-06,
+      "loss": 0.662,
+      "step": 5886
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2845804691314697,
+      "learning_rate": 1.5020786657721309e-06,
+      "loss": 0.5735,
+      "step": 5887
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2593872547149658,
+      "learning_rate": 1.5005962554373887e-06,
+      "loss": 0.5358,
+      "step": 5888
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.33612060546875,
+      "learning_rate": 1.4991144478038133e-06,
+      "loss": 0.5702,
+      "step": 5889
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3789081573486328,
+      "learning_rate": 1.497633243126616e-06,
+      "loss": 0.5369,
+      "step": 5890
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.179154396057129,
+      "learning_rate": 1.4961526416609062e-06,
+      "loss": 0.6509,
+      "step": 5891
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2960028648376465,
+      "learning_rate": 1.4946726436616844e-06,
+      "loss": 0.4924,
+      "step": 5892
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3278669118881226,
+      "learning_rate": 1.493193249383851e-06,
+      "loss": 0.6282,
+      "step": 5893
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4293935298919678,
+      "learning_rate": 1.4917144590822015e-06,
+      "loss": 0.6909,
+      "step": 5894
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6118907928466797,
+      "learning_rate": 1.4902362730114296e-06,
+      "loss": 0.6674,
+      "step": 5895
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.550476312637329,
+      "learning_rate": 1.4887586914261175e-06,
+      "loss": 0.5613,
+      "step": 5896
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3202171325683594,
+      "learning_rate": 1.4872817145807545e-06,
+      "loss": 0.5946,
+      "step": 5897
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.505103588104248,
+      "learning_rate": 1.4858053427297186e-06,
+      "loss": 0.5916,
+      "step": 5898
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4744017124176025,
+      "learning_rate": 1.4843295761272824e-06,
+      "loss": 0.6262,
+      "step": 5899
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3989981412887573,
+      "learning_rate": 1.482854415027618e-06,
+      "loss": 0.6148,
+      "step": 5900
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4221875667572021,
+      "learning_rate": 1.4813798596847916e-06,
+      "loss": 0.57,
+      "step": 5901
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3431564569473267,
+      "learning_rate": 1.479905910352768e-06,
+      "loss": 0.5682,
+      "step": 5902
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2961328029632568,
+      "learning_rate": 1.4784325672854005e-06,
+      "loss": 0.6361,
+      "step": 5903
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5604289770126343,
+      "learning_rate": 1.4769598307364452e-06,
+      "loss": 0.6184,
+      "step": 5904
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.340232014656067,
+      "learning_rate": 1.4754877009595509e-06,
+      "loss": 0.5844,
+      "step": 5905
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2040913105010986,
+      "learning_rate": 1.4740161782082623e-06,
+      "loss": 0.6196,
+      "step": 5906
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2919450998306274,
+      "learning_rate": 1.4725452627360182e-06,
+      "loss": 0.623,
+      "step": 5907
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9808508157730103,
+      "learning_rate": 1.471074954796154e-06,
+      "loss": 0.6147,
+      "step": 5908
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3140724897384644,
+      "learning_rate": 1.4696052546419005e-06,
+      "loss": 0.5737,
+      "step": 5909
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.031933069229126,
+      "learning_rate": 1.4681361625263857e-06,
+      "loss": 0.6697,
+      "step": 5910
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5908561944961548,
+      "learning_rate": 1.4666676787026273e-06,
+      "loss": 0.6361,
+      "step": 5911
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4421895742416382,
+      "learning_rate": 1.4651998034235438e-06,
+      "loss": 0.6496,
+      "step": 5912
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0249710083007812,
+      "learning_rate": 1.463732536941947e-06,
+      "loss": 0.6466,
+      "step": 5913
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8768399953842163,
+      "learning_rate": 1.4622658795105427e-06,
+      "loss": 0.693,
+      "step": 5914
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4486182928085327,
+      "learning_rate": 1.460799831381934e-06,
+      "loss": 0.5236,
+      "step": 5915
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1291416883468628,
+      "learning_rate": 1.4593343928086185e-06,
+      "loss": 0.6452,
+      "step": 5916
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.273863434791565,
+      "learning_rate": 1.4578695640429863e-06,
+      "loss": 0.5713,
+      "step": 5917
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5281238555908203,
+      "learning_rate": 1.4564053453373245e-06,
+      "loss": 0.6862,
+      "step": 5918
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6069231033325195,
+      "learning_rate": 1.4549417369438162e-06,
+      "loss": 0.5982,
+      "step": 5919
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2721714973449707,
+      "learning_rate": 1.4534787391145388e-06,
+      "loss": 0.5969,
+      "step": 5920
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1203700304031372,
+      "learning_rate": 1.452016352101462e-06,
+      "loss": 0.5516,
+      "step": 5921
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6137348413467407,
+      "learning_rate": 1.4505545761564526e-06,
+      "loss": 0.6058,
+      "step": 5922
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4966486692428589,
+      "learning_rate": 1.4490934115312721e-06,
+      "loss": 0.6258,
+      "step": 5923
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3931201696395874,
+      "learning_rate": 1.4476328584775784e-06,
+      "loss": 0.5624,
+      "step": 5924
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4084396362304688,
+      "learning_rate": 1.4461729172469191e-06,
+      "loss": 0.52,
+      "step": 5925
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3359655141830444,
+      "learning_rate": 1.4447135880907397e-06,
+      "loss": 0.5936,
+      "step": 5926
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4408459663391113,
+      "learning_rate": 1.4432548712603816e-06,
+      "loss": 0.5981,
+      "step": 5927
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4925096035003662,
+      "learning_rate": 1.4417967670070799e-06,
+      "loss": 0.556,
+      "step": 5928
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.2388222217559814,
+      "learning_rate": 1.44033927558196e-06,
+      "loss": 0.5848,
+      "step": 5929
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3931220769882202,
+      "learning_rate": 1.4388823972360455e-06,
+      "loss": 0.5938,
+      "step": 5930
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.208591103553772,
+      "learning_rate": 1.4374261322202592e-06,
+      "loss": 0.6013,
+      "step": 5931
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.465299129486084,
+      "learning_rate": 1.435970480785408e-06,
+      "loss": 0.6509,
+      "step": 5932
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5342435836791992,
+      "learning_rate": 1.4345154431821995e-06,
+      "loss": 0.636,
+      "step": 5933
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.555898666381836,
+      "learning_rate": 1.4330610196612348e-06,
+      "loss": 0.5484,
+      "step": 5934
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.423935055732727,
+      "learning_rate": 1.4316072104730106e-06,
+      "loss": 0.5434,
+      "step": 5935
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2774029970169067,
+      "learning_rate": 1.4301540158679123e-06,
+      "loss": 0.5129,
+      "step": 5936
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.258390188217163,
+      "learning_rate": 1.4287014360962247e-06,
+      "loss": 0.5213,
+      "step": 5937
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3220731019973755,
+      "learning_rate": 1.427249471408126e-06,
+      "loss": 0.5337,
+      "step": 5938
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4194802045822144,
+      "learning_rate": 1.4257981220536883e-06,
+      "loss": 0.567,
+      "step": 5939
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.372889757156372,
+      "learning_rate": 1.424347388282874e-06,
+      "loss": 0.6144,
+      "step": 5940
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7539900541305542,
+      "learning_rate": 1.4228972703455441e-06,
+      "loss": 0.6644,
+      "step": 5941
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.544344186782837,
+      "learning_rate": 1.4214477684914524e-06,
+      "loss": 0.6441,
+      "step": 5942
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3730299472808838,
+      "learning_rate": 1.4199988829702472e-06,
+      "loss": 0.5675,
+      "step": 5943
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2083449363708496,
+      "learning_rate": 1.4185506140314665e-06,
+      "loss": 0.5938,
+      "step": 5944
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5624059438705444,
+      "learning_rate": 1.4171029619245468e-06,
+      "loss": 0.6969,
+      "step": 5945
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3644176721572876,
+      "learning_rate": 1.4156559268988168e-06,
+      "loss": 0.5768,
+      "step": 5946
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3280311822891235,
+      "learning_rate": 1.414209509203499e-06,
+      "loss": 0.6175,
+      "step": 5947
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4516396522521973,
+      "learning_rate": 1.4127637090877094e-06,
+      "loss": 0.5231,
+      "step": 5948
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.298318862915039,
+      "learning_rate": 1.4113185268004576e-06,
+      "loss": 0.5446,
+      "step": 5949
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2986234426498413,
+      "learning_rate": 1.4098739625906482e-06,
+      "loss": 0.5896,
+      "step": 5950
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3965665102005005,
+      "learning_rate": 1.4084300167070758e-06,
+      "loss": 0.6212,
+      "step": 5951
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5937730073928833,
+      "learning_rate": 1.4069866893984307e-06,
+      "loss": 0.5914,
+      "step": 5952
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.7656188011169434,
+      "learning_rate": 1.4055439809132987e-06,
+      "loss": 0.5545,
+      "step": 5953
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.46426260471344,
+      "learning_rate": 1.4041018915001564e-06,
+      "loss": 0.58,
+      "step": 5954
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5463647842407227,
+      "learning_rate": 1.4026604214073731e-06,
+      "loss": 0.5886,
+      "step": 5955
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4944566488265991,
+      "learning_rate": 1.4012195708832132e-06,
+      "loss": 0.5805,
+      "step": 5956
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.733450174331665,
+      "learning_rate": 1.3997793401758347e-06,
+      "loss": 0.5651,
+      "step": 5957
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2912808656692505,
+      "learning_rate": 1.398339729533289e-06,
+      "loss": 0.6123,
+      "step": 5958
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4565380811691284,
+      "learning_rate": 1.3969007392035166e-06,
+      "loss": 0.656,
+      "step": 5959
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2773964405059814,
+      "learning_rate": 1.3954623694343566e-06,
+      "loss": 0.5466,
+      "step": 5960
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2596515417099,
+      "learning_rate": 1.3940246204735398e-06,
+      "loss": 0.5493,
+      "step": 5961
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3928008079528809,
+      "learning_rate": 1.392587492568686e-06,
+      "loss": 0.5857,
+      "step": 5962
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3344323635101318,
+      "learning_rate": 1.3911509859673138e-06,
+      "loss": 0.5673,
+      "step": 5963
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.650927186012268,
+      "learning_rate": 1.3897151009168297e-06,
+      "loss": 0.5837,
+      "step": 5964
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2089368104934692,
+      "learning_rate": 1.3882798376645413e-06,
+      "loss": 0.5227,
+      "step": 5965
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2477643489837646,
+      "learning_rate": 1.386845196457638e-06,
+      "loss": 0.5825,
+      "step": 5966
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5391136407852173,
+      "learning_rate": 1.3854111775432093e-06,
+      "loss": 0.6396,
+      "step": 5967
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.506758213043213,
+      "learning_rate": 1.383977781168238e-06,
+      "loss": 0.5131,
+      "step": 5968
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.46280038356781,
+      "learning_rate": 1.3825450075795937e-06,
+      "loss": 0.5954,
+      "step": 5969
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0830283164978027,
+      "learning_rate": 1.3811128570240445e-06,
+      "loss": 0.6811,
+      "step": 5970
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7907627820968628,
+      "learning_rate": 1.3796813297482486e-06,
+      "loss": 0.6008,
+      "step": 5971
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3350781202316284,
+      "learning_rate": 1.3782504259987601e-06,
+      "loss": 0.538,
+      "step": 5972
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.362600326538086,
+      "learning_rate": 1.3768201460220187e-06,
+      "loss": 0.5904,
+      "step": 5973
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.149887204170227,
+      "learning_rate": 1.375390490064364e-06,
+      "loss": 0.5439,
+      "step": 5974
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1738979816436768,
+      "learning_rate": 1.373961458372024e-06,
+      "loss": 0.535,
+      "step": 5975
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3399025201797485,
+      "learning_rate": 1.3725330511911223e-06,
+      "loss": 0.5719,
+      "step": 5976
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3920693397521973,
+      "learning_rate": 1.3711052687676701e-06,
+      "loss": 0.6096,
+      "step": 5977
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3506783246994019,
+      "learning_rate": 1.3696781113475754e-06,
+      "loss": 0.594,
+      "step": 5978
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2706949710845947,
+      "learning_rate": 1.3682515791766372e-06,
+      "loss": 0.5891,
+      "step": 5979
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2764166593551636,
+      "learning_rate": 1.3668256725005475e-06,
+      "loss": 0.6208,
+      "step": 5980
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4557310342788696,
+      "learning_rate": 1.3654003915648873e-06,
+      "loss": 0.558,
+      "step": 5981
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1343717575073242,
+      "learning_rate": 1.3639757366151323e-06,
+      "loss": 0.5511,
+      "step": 5982
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4977566003799438,
+      "learning_rate": 1.3625517078966544e-06,
+      "loss": 0.5928,
+      "step": 5983
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3123750686645508,
+      "learning_rate": 1.3611283056547097e-06,
+      "loss": 0.5015,
+      "step": 5984
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.268973469734192,
+      "learning_rate": 1.3597055301344515e-06,
+      "loss": 0.6172,
+      "step": 5985
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3133175373077393,
+      "learning_rate": 1.3582833815809244e-06,
+      "loss": 0.6634,
+      "step": 5986
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5407078266143799,
+      "learning_rate": 1.356861860239065e-06,
+      "loss": 0.6416,
+      "step": 5987
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2365789413452148,
+      "learning_rate": 1.3554409663536993e-06,
+      "loss": 0.6078,
+      "step": 5988
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4890007972717285,
+      "learning_rate": 1.3540207001695489e-06,
+      "loss": 0.5228,
+      "step": 5989
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1936070919036865,
+      "learning_rate": 1.3526010619312252e-06,
+      "loss": 0.6939,
+      "step": 5990
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4938281774520874,
+      "learning_rate": 1.3511820518832342e-06,
+      "loss": 0.628,
+      "step": 5991
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7732173204421997,
+      "learning_rate": 1.3497636702699684e-06,
+      "loss": 0.6403,
+      "step": 5992
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5048422813415527,
+      "learning_rate": 1.3483459173357167e-06,
+      "loss": 0.6572,
+      "step": 5993
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2302240133285522,
+      "learning_rate": 1.3469287933246577e-06,
+      "loss": 0.562,
+      "step": 5994
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.197856068611145,
+      "learning_rate": 1.3455122984808644e-06,
+      "loss": 0.677,
+      "step": 5995
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3732661008834839,
+      "learning_rate": 1.3440964330482958e-06,
+      "loss": 0.6358,
+      "step": 5996
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2321995496749878,
+      "learning_rate": 1.3426811972708076e-06,
+      "loss": 0.6068,
+      "step": 5997
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1939703226089478,
+      "learning_rate": 1.3412665913921451e-06,
+      "loss": 0.5675,
+      "step": 5998
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3941285610198975,
+      "learning_rate": 1.339852615655946e-06,
+      "loss": 0.612,
+      "step": 5999
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4940359592437744,
+      "learning_rate": 1.3384392703057386e-06,
+      "loss": 0.653,
+      "step": 6000
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.482255220413208,
+      "learning_rate": 1.3370265555849427e-06,
+      "loss": 0.495,
+      "step": 6001
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4018694162368774,
+      "learning_rate": 1.335614471736872e-06,
+      "loss": 0.6908,
+      "step": 6002
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2565648555755615,
+      "learning_rate": 1.334203019004725e-06,
+      "loss": 0.5861,
+      "step": 6003
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4649759531021118,
+      "learning_rate": 1.3327921976315977e-06,
+      "loss": 0.5378,
+      "step": 6004
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.391389012336731,
+      "learning_rate": 1.3313820078604761e-06,
+      "loss": 0.6189,
+      "step": 6005
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4627043008804321,
+      "learning_rate": 1.3299724499342376e-06,
+      "loss": 0.6309,
+      "step": 6006
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.310528039932251,
+      "learning_rate": 1.3285635240956469e-06,
+      "loss": 0.6002,
+      "step": 6007
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.153878092765808,
+      "learning_rate": 1.3271552305873648e-06,
+      "loss": 0.6072,
+      "step": 6008
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3677489757537842,
+      "learning_rate": 1.3257475696519417e-06,
+      "loss": 0.6208,
+      "step": 6009
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3258711099624634,
+      "learning_rate": 1.3243405415318166e-06,
+      "loss": 0.5862,
+      "step": 6010
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.423606038093567,
+      "learning_rate": 1.322934146469323e-06,
+      "loss": 0.5903,
+      "step": 6011
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.480432152748108,
+      "learning_rate": 1.3215283847066828e-06,
+      "loss": 0.4928,
+      "step": 6012
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5109148025512695,
+      "learning_rate": 1.3201232564860128e-06,
+      "loss": 0.6378,
+      "step": 6013
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.531599998474121,
+      "learning_rate": 1.3187187620493137e-06,
+      "loss": 0.6011,
+      "step": 6014
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9274752140045166,
+      "learning_rate": 1.3173149016384829e-06,
+      "loss": 0.5542,
+      "step": 6015
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.15962553024292,
+      "learning_rate": 1.3159116754953073e-06,
+      "loss": 0.5979,
+      "step": 6016
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3300819396972656,
+      "learning_rate": 1.3145090838614633e-06,
+      "loss": 0.538,
+      "step": 6017
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3785098791122437,
+      "learning_rate": 1.3131071269785196e-06,
+      "loss": 0.6167,
+      "step": 6018
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.139225721359253,
+      "learning_rate": 1.3117058050879344e-06,
+      "loss": 0.7076,
+      "step": 6019
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5697603225708008,
+      "learning_rate": 1.3103051184310577e-06,
+      "loss": 0.614,
+      "step": 6020
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3997340202331543,
+      "learning_rate": 1.3089050672491278e-06,
+      "loss": 0.5386,
+      "step": 6021
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4354143142700195,
+      "learning_rate": 1.3075056517832758e-06,
+      "loss": 0.6331,
+      "step": 6022
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3406672477722168,
+      "learning_rate": 1.306106872274522e-06,
+      "loss": 0.5646,
+      "step": 6023
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3164324760437012,
+      "learning_rate": 1.30470872896378e-06,
+      "loss": 0.582,
+      "step": 6024
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4579137563705444,
+      "learning_rate": 1.3033112220918482e-06,
+      "loss": 0.6018,
+      "step": 6025
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4018653631210327,
+      "learning_rate": 1.3019143518994203e-06,
+      "loss": 0.6285,
+      "step": 6026
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6690030097961426,
+      "learning_rate": 1.3005181186270794e-06,
+      "loss": 0.6455,
+      "step": 6027
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.20956552028656,
+      "learning_rate": 1.2991225225152988e-06,
+      "loss": 0.5629,
+      "step": 6028
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3824257850646973,
+      "learning_rate": 1.297727563804439e-06,
+      "loss": 0.5708,
+      "step": 6029
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7339993715286255,
+      "learning_rate": 1.2963332427347552e-06,
+      "loss": 0.6303,
+      "step": 6030
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2081109285354614,
+      "learning_rate": 1.2949395595463904e-06,
+      "loss": 0.6148,
+      "step": 6031
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.271765947341919,
+      "learning_rate": 1.2935465144793797e-06,
+      "loss": 0.5637,
+      "step": 6032
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3769280910491943,
+      "learning_rate": 1.2921541077736428e-06,
+      "loss": 0.565,
+      "step": 6033
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.288026213645935,
+      "learning_rate": 1.2907623396689973e-06,
+      "loss": 0.5762,
+      "step": 6034
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.817467451095581,
+      "learning_rate": 1.289371210405147e-06,
+      "loss": 0.6836,
+      "step": 6035
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3072487115859985,
+      "learning_rate": 1.2879807202216826e-06,
+      "loss": 0.5799,
+      "step": 6036
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3482693433761597,
+      "learning_rate": 1.2865908693580903e-06,
+      "loss": 0.6028,
+      "step": 6037
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5620173215866089,
+      "learning_rate": 1.2852016580537424e-06,
+      "loss": 0.5214,
+      "step": 6038
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.521825909614563,
+      "learning_rate": 1.2838130865479049e-06,
+      "loss": 0.6226,
+      "step": 6039
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7323329448699951,
+      "learning_rate": 1.2824251550797268e-06,
+      "loss": 0.5745,
+      "step": 6040
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4529424905776978,
+      "learning_rate": 1.281037863888253e-06,
+      "loss": 0.6783,
+      "step": 6041
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3162099123001099,
+      "learning_rate": 1.2796512132124172e-06,
+      "loss": 0.6138,
+      "step": 6042
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3417941331863403,
+      "learning_rate": 1.278265203291042e-06,
+      "loss": 0.651,
+      "step": 6043
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2643450498580933,
+      "learning_rate": 1.2768798343628368e-06,
+      "loss": 0.5403,
+      "step": 6044
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5022494792938232,
+      "learning_rate": 1.2754951066664056e-06,
+      "loss": 0.592,
+      "step": 6045
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4565330743789673,
+      "learning_rate": 1.274111020440238e-06,
+      "loss": 0.5484,
+      "step": 6046
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3392274379730225,
+      "learning_rate": 1.2727275759227175e-06,
+      "loss": 0.6221,
+      "step": 6047
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.425985336303711,
+      "learning_rate": 1.271344773352111e-06,
+      "loss": 0.5557,
+      "step": 6048
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3691293001174927,
+      "learning_rate": 1.2699626129665798e-06,
+      "loss": 0.6133,
+      "step": 6049
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5187207460403442,
+      "learning_rate": 1.2685810950041722e-06,
+      "loss": 0.6545,
+      "step": 6050
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3241124153137207,
+      "learning_rate": 1.267200219702827e-06,
+      "loss": 0.5776,
+      "step": 6051
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4314247369766235,
+      "learning_rate": 1.2658199873003723e-06,
+      "loss": 0.5092,
+      "step": 6052
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4745780229568481,
+      "learning_rate": 1.264440398034525e-06,
+      "loss": 0.561,
+      "step": 6053
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2822953462600708,
+      "learning_rate": 1.2630614521428919e-06,
+      "loss": 0.601,
+      "step": 6054
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.345860242843628,
+      "learning_rate": 1.2616831498629668e-06,
+      "loss": 0.5509,
+      "step": 6055
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3719156980514526,
+      "learning_rate": 1.2603054914321343e-06,
+      "loss": 0.5773,
+      "step": 6056
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6491966247558594,
+      "learning_rate": 1.2589284770876692e-06,
+      "loss": 0.5813,
+      "step": 6057
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3380208015441895,
+      "learning_rate": 1.2575521070667352e-06,
+      "loss": 0.6066,
+      "step": 6058
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6886008977890015,
+      "learning_rate": 1.256176381606381e-06,
+      "loss": 0.6094,
+      "step": 6059
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4857572317123413,
+      "learning_rate": 1.2548013009435495e-06,
+      "loss": 0.5453,
+      "step": 6060
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2749769687652588,
+      "learning_rate": 1.2534268653150705e-06,
+      "loss": 0.5806,
+      "step": 6061
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.563644528388977,
+      "learning_rate": 1.2520530749576615e-06,
+      "loss": 0.6023,
+      "step": 6062
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2830228805541992,
+      "learning_rate": 1.2506799301079298e-06,
+      "loss": 0.5755,
+      "step": 6063
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7365049123764038,
+      "learning_rate": 1.2493074310023728e-06,
+      "loss": 0.5773,
+      "step": 6064
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2732226848602295,
+      "learning_rate": 1.2479355778773761e-06,
+      "loss": 0.5507,
+      "step": 6065
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5290393829345703,
+      "learning_rate": 1.2465643709692115e-06,
+      "loss": 0.6324,
+      "step": 6066
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.335749387741089,
+      "learning_rate": 1.245193810514041e-06,
+      "loss": 0.6078,
+      "step": 6067
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4709378480911255,
+      "learning_rate": 1.2438238967479205e-06,
+      "loss": 0.6015,
+      "step": 6068
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3493144512176514,
+      "learning_rate": 1.2424546299067846e-06,
+      "loss": 0.6496,
+      "step": 6069
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2655116319656372,
+      "learning_rate": 1.2410860102264644e-06,
+      "loss": 0.5568,
+      "step": 6070
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2244620323181152,
+      "learning_rate": 1.239718037942676e-06,
+      "loss": 0.6423,
+      "step": 6071
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.434319019317627,
+      "learning_rate": 1.238350713291026e-06,
+      "loss": 0.6144,
+      "step": 6072
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.324418067932129,
+      "learning_rate": 1.2369840365070063e-06,
+      "loss": 0.5737,
+      "step": 6073
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.387671709060669,
+      "learning_rate": 1.235618007826e-06,
+      "loss": 0.6031,
+      "step": 6074
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4040278196334839,
+      "learning_rate": 1.2342526274832778e-06,
+      "loss": 0.6111,
+      "step": 6075
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.373987078666687,
+      "learning_rate": 1.2328878957140006e-06,
+      "loss": 0.5173,
+      "step": 6076
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1661827564239502,
+      "learning_rate": 1.2315238127532126e-06,
+      "loss": 0.5731,
+      "step": 6077
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3571733236312866,
+      "learning_rate": 1.2301603788358501e-06,
+      "loss": 0.6028,
+      "step": 6078
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2105902433395386,
+      "learning_rate": 1.2287975941967377e-06,
+      "loss": 0.5522,
+      "step": 6079
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1787046194076538,
+      "learning_rate": 1.2274354590705888e-06,
+      "loss": 0.5634,
+      "step": 6080
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.513954520225525,
+      "learning_rate": 1.2260739736920002e-06,
+      "loss": 0.5923,
+      "step": 6081
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3857628107070923,
+      "learning_rate": 1.224713138295462e-06,
+      "loss": 0.5761,
+      "step": 6082
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.2354791164398193,
+      "learning_rate": 1.22335295311535e-06,
+      "loss": 0.5485,
+      "step": 6083
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4245437383651733,
+      "learning_rate": 1.2219934183859283e-06,
+      "loss": 0.6352,
+      "step": 6084
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.455147385597229,
+      "learning_rate": 1.22063453434135e-06,
+      "loss": 0.6363,
+      "step": 6085
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2811365127563477,
+      "learning_rate": 1.2192763012156544e-06,
+      "loss": 0.5653,
+      "step": 6086
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1058675050735474,
+      "learning_rate": 1.2179187192427716e-06,
+      "loss": 0.708,
+      "step": 6087
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4116437435150146,
+      "learning_rate": 1.216561788656514e-06,
+      "loss": 0.6089,
+      "step": 6088
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2147979736328125,
+      "learning_rate": 1.2152055096905868e-06,
+      "loss": 0.5706,
+      "step": 6089
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.238149881362915,
+      "learning_rate": 1.2138498825785822e-06,
+      "loss": 0.5763,
+      "step": 6090
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7066261768341064,
+      "learning_rate": 1.21249490755398e-06,
+      "loss": 0.5693,
+      "step": 6091
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3570530414581299,
+      "learning_rate": 1.2111405848501451e-06,
+      "loss": 0.6234,
+      "step": 6092
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.259835958480835,
+      "learning_rate": 1.2097869147003328e-06,
+      "loss": 0.5633,
+      "step": 6093
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6666969060897827,
+      "learning_rate": 1.2084338973376853e-06,
+      "loss": 0.575,
+      "step": 6094
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9863353967666626,
+      "learning_rate": 1.207081532995234e-06,
+      "loss": 0.6823,
+      "step": 6095
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.477820873260498,
+      "learning_rate": 1.2057298219058933e-06,
+      "loss": 0.6011,
+      "step": 6096
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3085917234420776,
+      "learning_rate": 1.204378764302469e-06,
+      "loss": 0.5428,
+      "step": 6097
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.393518090248108,
+      "learning_rate": 1.2030283604176545e-06,
+      "loss": 0.6187,
+      "step": 6098
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3972924947738647,
+      "learning_rate": 1.2016786104840296e-06,
+      "loss": 0.6148,
+      "step": 6099
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4566419124603271,
+      "learning_rate": 1.2003295147340587e-06,
+      "loss": 0.6794,
+      "step": 6100
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3721485137939453,
+      "learning_rate": 1.1989810734000962e-06,
+      "loss": 0.5104,
+      "step": 6101
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3800498247146606,
+      "learning_rate": 1.1976332867143887e-06,
+      "loss": 0.4953,
+      "step": 6102
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5365797281265259,
+      "learning_rate": 1.1962861549090593e-06,
+      "loss": 0.607,
+      "step": 6103
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.18382728099823,
+      "learning_rate": 1.1949396782161272e-06,
+      "loss": 0.725,
+      "step": 6104
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2551465034484863,
+      "learning_rate": 1.1935938568674943e-06,
+      "loss": 0.6576,
+      "step": 6105
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2952631711959839,
+      "learning_rate": 1.1922486910949527e-06,
+      "loss": 0.5513,
+      "step": 6106
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.188146948814392,
+      "learning_rate": 1.1909041811301775e-06,
+      "loss": 0.5895,
+      "step": 6107
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3294532299041748,
+      "learning_rate": 1.189560327204734e-06,
+      "loss": 0.5671,
+      "step": 6108
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3476665019989014,
+      "learning_rate": 1.188217129550076e-06,
+      "loss": 0.6287,
+      "step": 6109
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3214592933654785,
+      "learning_rate": 1.1868745883975386e-06,
+      "loss": 0.5985,
+      "step": 6110
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6524791717529297,
+      "learning_rate": 1.1855327039783481e-06,
+      "loss": 0.6008,
+      "step": 6111
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.334662914276123,
+      "learning_rate": 1.1841914765236178e-06,
+      "loss": 0.5857,
+      "step": 6112
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.622629642486572,
+      "learning_rate": 1.182850906264348e-06,
+      "loss": 0.663,
+      "step": 6113
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1818311214447021,
+      "learning_rate": 1.1815109934314218e-06,
+      "loss": 0.5183,
+      "step": 6114
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.55926513671875,
+      "learning_rate": 1.1801717382556133e-06,
+      "loss": 0.6356,
+      "step": 6115
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.465349793434143,
+      "learning_rate": 1.1788331409675812e-06,
+      "loss": 0.5734,
+      "step": 6116
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.333842396736145,
+      "learning_rate": 1.1774952017978748e-06,
+      "loss": 0.54,
+      "step": 6117
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6432085037231445,
+      "learning_rate": 1.1761579209769226e-06,
+      "loss": 0.6167,
+      "step": 6118
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2257115840911865,
+      "learning_rate": 1.174821298735045e-06,
+      "loss": 0.5268,
+      "step": 6119
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3372501134872437,
+      "learning_rate": 1.1734853353024517e-06,
+      "loss": 0.6231,
+      "step": 6120
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5983772277832031,
+      "learning_rate": 1.172150030909231e-06,
+      "loss": 0.6376,
+      "step": 6121
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.557028889656067,
+      "learning_rate": 1.1708153857853637e-06,
+      "loss": 0.661,
+      "step": 6122
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.520965576171875,
+      "learning_rate": 1.1694814001607152e-06,
+      "loss": 0.6603,
+      "step": 6123
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3232431411743164,
+      "learning_rate": 1.1681480742650387e-06,
+      "loss": 0.5791,
+      "step": 6124
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9522755146026611,
+      "learning_rate": 1.1668154083279692e-06,
+      "loss": 0.5742,
+      "step": 6125
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4064555168151855,
+      "learning_rate": 1.165483402579034e-06,
+      "loss": 0.6266,
+      "step": 6126
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2465460300445557,
+      "learning_rate": 1.1641520572476428e-06,
+      "loss": 0.5173,
+      "step": 6127
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.656630039215088,
+      "learning_rate": 1.162821372563095e-06,
+      "loss": 0.5574,
+      "step": 6128
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7703921794891357,
+      "learning_rate": 1.1614913487545704e-06,
+      "loss": 0.6125,
+      "step": 6129
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.557174801826477,
+      "learning_rate": 1.1601619860511403e-06,
+      "loss": 0.5712,
+      "step": 6130
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.303775429725647,
+      "learning_rate": 1.1588332846817608e-06,
+      "loss": 0.5705,
+      "step": 6131
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4587818384170532,
+      "learning_rate": 1.1575052448752744e-06,
+      "loss": 0.5648,
+      "step": 6132
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3065063953399658,
+      "learning_rate": 1.1561778668604068e-06,
+      "loss": 0.5471,
+      "step": 6133
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1926515102386475,
+      "learning_rate": 1.1548511508657733e-06,
+      "loss": 0.5889,
+      "step": 6134
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.834707260131836,
+      "learning_rate": 1.153525097119873e-06,
+      "loss": 0.5979,
+      "step": 6135
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8028165102005005,
+      "learning_rate": 1.1521997058510926e-06,
+      "loss": 0.6267,
+      "step": 6136
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2593283653259277,
+      "learning_rate": 1.1508749772877032e-06,
+      "loss": 0.604,
+      "step": 6137
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.511449933052063,
+      "learning_rate": 1.1495509116578628e-06,
+      "loss": 0.6545,
+      "step": 6138
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.41291344165802,
+      "learning_rate": 1.1482275091896155e-06,
+      "loss": 0.4631,
+      "step": 6139
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5056250095367432,
+      "learning_rate": 1.1469047701108888e-06,
+      "loss": 0.6018,
+      "step": 6140
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5309189558029175,
+      "learning_rate": 1.145582694649498e-06,
+      "loss": 0.6386,
+      "step": 6141
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.509054183959961,
+      "learning_rate": 1.144261283033144e-06,
+      "loss": 0.5576,
+      "step": 6142
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.0752272605895996,
+      "learning_rate": 1.1429405354894147e-06,
+      "loss": 0.6352,
+      "step": 6143
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3260945081710815,
+      "learning_rate": 1.1416204522457792e-06,
+      "loss": 0.5944,
+      "step": 6144
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1211093664169312,
+      "learning_rate": 1.140301033529596e-06,
+      "loss": 0.7431,
+      "step": 6145
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5786157846450806,
+      "learning_rate": 1.1389822795681088e-06,
+      "loss": 0.573,
+      "step": 6146
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4324467182159424,
+      "learning_rate": 1.1376641905884472e-06,
+      "loss": 0.5515,
+      "step": 6147
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2469350099563599,
+      "learning_rate": 1.1363467668176221e-06,
+      "loss": 0.532,
+      "step": 6148
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.782609462738037,
+      "learning_rate": 1.135030008482535e-06,
+      "loss": 0.5981,
+      "step": 6149
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5515788793563843,
+      "learning_rate": 1.1337139158099708e-06,
+      "loss": 0.6569,
+      "step": 6150
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6036688089370728,
+      "learning_rate": 1.1323984890266005e-06,
+      "loss": 0.6541,
+      "step": 6151
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3179755210876465,
+      "learning_rate": 1.1310837283589771e-06,
+      "loss": 0.6114,
+      "step": 6152
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.550906777381897,
+      "learning_rate": 1.1297696340335412e-06,
+      "loss": 0.614,
+      "step": 6153
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.434679388999939,
+      "learning_rate": 1.1284562062766236e-06,
+      "loss": 0.5633,
+      "step": 6154
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4772613048553467,
+      "learning_rate": 1.1271434453144308e-06,
+      "loss": 0.609,
+      "step": 6155
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4170340299606323,
+      "learning_rate": 1.1258313513730613e-06,
+      "loss": 0.6429,
+      "step": 6156
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5838710069656372,
+      "learning_rate": 1.1245199246784965e-06,
+      "loss": 0.6858,
+      "step": 6157
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.0961270332336426,
+      "learning_rate": 1.1232091654566024e-06,
+      "loss": 0.5789,
+      "step": 6158
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.436582326889038,
+      "learning_rate": 1.1218990739331297e-06,
+      "loss": 0.6391,
+      "step": 6159
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3351346254348755,
+      "learning_rate": 1.120589650333717e-06,
+      "loss": 0.5651,
+      "step": 6160
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1929622888565063,
+      "learning_rate": 1.1192808948838863e-06,
+      "loss": 0.552,
+      "step": 6161
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3724346160888672,
+      "learning_rate": 1.1179728078090413e-06,
+      "loss": 0.6234,
+      "step": 6162
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2918885946273804,
+      "learning_rate": 1.1166653893344753e-06,
+      "loss": 0.5676,
+      "step": 6163
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.776462435722351,
+      "learning_rate": 1.115358639685364e-06,
+      "loss": 0.5481,
+      "step": 6164
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4188241958618164,
+      "learning_rate": 1.11405255908677e-06,
+      "loss": 0.5834,
+      "step": 6165
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6242364645004272,
+      "learning_rate": 1.112747147763637e-06,
+      "loss": 0.6114,
+      "step": 6166
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0123103857040405,
+      "learning_rate": 1.1114424059407957e-06,
+      "loss": 0.7003,
+      "step": 6167
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5645592212677002,
+      "learning_rate": 1.1101383338429627e-06,
+      "loss": 0.6156,
+      "step": 6168
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3179627656936646,
+      "learning_rate": 1.1088349316947388e-06,
+      "loss": 0.5178,
+      "step": 6169
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7976934909820557,
+      "learning_rate": 1.107532199720604e-06,
+      "loss": 0.5481,
+      "step": 6170
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.527823567390442,
+      "learning_rate": 1.1062301381449313e-06,
+      "loss": 0.6192,
+      "step": 6171
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5737080574035645,
+      "learning_rate": 1.1049287471919756e-06,
+      "loss": 0.6406,
+      "step": 6172
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.55455482006073,
+      "learning_rate": 1.1036280270858712e-06,
+      "loss": 0.703,
+      "step": 6173
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3198282718658447,
+      "learning_rate": 1.1023279780506425e-06,
+      "loss": 0.6418,
+      "step": 6174
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.4118762016296387,
+      "learning_rate": 1.1010286003101967e-06,
+      "loss": 0.6689,
+      "step": 6175
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9782161712646484,
+      "learning_rate": 1.0997298940883261e-06,
+      "loss": 0.5931,
+      "step": 6176
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.633495807647705,
+      "learning_rate": 1.0984318596087035e-06,
+      "loss": 0.6079,
+      "step": 6177
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5618376731872559,
+      "learning_rate": 1.0971344970948911e-06,
+      "loss": 0.5518,
+      "step": 6178
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2417227029800415,
+      "learning_rate": 1.095837806770333e-06,
+      "loss": 0.5463,
+      "step": 6179
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.2190401554107666,
+      "learning_rate": 1.0945417888583593e-06,
+      "loss": 0.648,
+      "step": 6180
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2779728174209595,
+      "learning_rate": 1.0932464435821789e-06,
+      "loss": 0.5593,
+      "step": 6181
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9929521083831787,
+      "learning_rate": 1.0919517711648914e-06,
+      "loss": 0.5269,
+      "step": 6182
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4795068502426147,
+      "learning_rate": 1.090657771829477e-06,
+      "loss": 0.6332,
+      "step": 6183
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3366481065750122,
+      "learning_rate": 1.0893644457988029e-06,
+      "loss": 0.5933,
+      "step": 6184
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3400702476501465,
+      "learning_rate": 1.0880717932956148e-06,
+      "loss": 0.5539,
+      "step": 6185
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3519129753112793,
+      "learning_rate": 1.0867798145425468e-06,
+      "loss": 0.5936,
+      "step": 6186
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3224157094955444,
+      "learning_rate": 1.085488509762117e-06,
+      "loss": 0.5686,
+      "step": 6187
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4233057498931885,
+      "learning_rate": 1.084197879176726e-06,
+      "loss": 0.7274,
+      "step": 6188
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6368494033813477,
+      "learning_rate": 1.0829079230086587e-06,
+      "loss": 0.5718,
+      "step": 6189
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5278626680374146,
+      "learning_rate": 1.0816186414800838e-06,
+      "loss": 0.5603,
+      "step": 6190
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.375484824180603,
+      "learning_rate": 1.0803300348130552e-06,
+      "loss": 0.564,
+      "step": 6191
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.002624273300171,
+      "learning_rate": 1.0790421032295067e-06,
+      "loss": 0.5248,
+      "step": 6192
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4766088724136353,
+      "learning_rate": 1.0777548469512588e-06,
+      "loss": 0.5901,
+      "step": 6193
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3245599269866943,
+      "learning_rate": 1.0764682662000165e-06,
+      "loss": 0.5882,
+      "step": 6194
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1820677518844604,
+      "learning_rate": 1.0751823611973677e-06,
+      "loss": 0.5516,
+      "step": 6195
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.272409439086914,
+      "learning_rate": 1.0738971321647811e-06,
+      "loss": 0.6358,
+      "step": 6196
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.234097957611084,
+      "learning_rate": 1.0726125793236124e-06,
+      "loss": 0.5142,
+      "step": 6197
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3627111911773682,
+      "learning_rate": 1.0713287028950997e-06,
+      "loss": 0.5465,
+      "step": 6198
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6191880702972412,
+      "learning_rate": 1.0700455031003665e-06,
+      "loss": 0.6199,
+      "step": 6199
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6177563667297363,
+      "learning_rate": 1.0687629801604143e-06,
+      "loss": 0.6624,
+      "step": 6200
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1564007997512817,
+      "learning_rate": 1.0674811342961332e-06,
+      "loss": 0.5907,
+      "step": 6201
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.260568618774414,
+      "learning_rate": 1.0661999657282968e-06,
+      "loss": 0.5627,
+      "step": 6202
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5022335052490234,
+      "learning_rate": 1.0649194746775576e-06,
+      "loss": 0.6852,
+      "step": 6203
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.435604453086853,
+      "learning_rate": 1.063639661364454e-06,
+      "loss": 0.6222,
+      "step": 6204
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1896655559539795,
+      "learning_rate": 1.0623605260094105e-06,
+      "loss": 0.5527,
+      "step": 6205
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4578163623809814,
+      "learning_rate": 1.0610820688327323e-06,
+      "loss": 0.6038,
+      "step": 6206
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.404301643371582,
+      "learning_rate": 1.0598042900546051e-06,
+      "loss": 0.554,
+      "step": 6207
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2354761362075806,
+      "learning_rate": 1.058527189895101e-06,
+      "loss": 0.6951,
+      "step": 6208
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9446297883987427,
+      "learning_rate": 1.0572507685741767e-06,
+      "loss": 0.5824,
+      "step": 6209
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.273754596710205,
+      "learning_rate": 1.055975026311667e-06,
+      "loss": 0.6316,
+      "step": 6210
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.49468195438385,
+      "learning_rate": 1.0546999633272935e-06,
+      "loss": 0.6304,
+      "step": 6211
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2683830261230469,
+      "learning_rate": 1.0534255798406602e-06,
+      "loss": 0.5711,
+      "step": 6212
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4558610916137695,
+      "learning_rate": 1.0521518760712545e-06,
+      "loss": 0.5594,
+      "step": 6213
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4848997592926025,
+      "learning_rate": 1.0508788522384443e-06,
+      "loss": 0.5832,
+      "step": 6214
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1660897731781006,
+      "learning_rate": 1.0496065085614827e-06,
+      "loss": 0.5497,
+      "step": 6215
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3938348293304443,
+      "learning_rate": 1.0483348452595056e-06,
+      "loss": 0.6248,
+      "step": 6216
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2313079833984375,
+      "learning_rate": 1.0470638625515316e-06,
+      "loss": 0.549,
+      "step": 6217
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4544310569763184,
+      "learning_rate": 1.0457935606564596e-06,
+      "loss": 0.5853,
+      "step": 6218
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7911359071731567,
+      "learning_rate": 1.044523939793075e-06,
+      "loss": 0.6357,
+      "step": 6219
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5014184713363647,
+      "learning_rate": 1.0432550001800435e-06,
+      "loss": 0.6137,
+      "step": 6220
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3622461557388306,
+      "learning_rate": 1.0419867420359141e-06,
+      "loss": 0.5823,
+      "step": 6221
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2310843467712402,
+      "learning_rate": 1.0407191655791188e-06,
+      "loss": 0.5025,
+      "step": 6222
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.638079047203064,
+      "learning_rate": 1.039452271027972e-06,
+      "loss": 0.6757,
+      "step": 6223
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6458463668823242,
+      "learning_rate": 1.038186058600672e-06,
+      "loss": 0.4908,
+      "step": 6224
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4010100364685059,
+      "learning_rate": 1.036920528515295e-06,
+      "loss": 0.6098,
+      "step": 6225
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.987982153892517,
+      "learning_rate": 1.0356556809898044e-06,
+      "loss": 0.59,
+      "step": 6226
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2834595441818237,
+      "learning_rate": 1.0343915162420443e-06,
+      "loss": 0.5335,
+      "step": 6227
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4950449466705322,
+      "learning_rate": 1.0331280344897425e-06,
+      "loss": 0.5466,
+      "step": 6228
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2922565937042236,
+      "learning_rate": 1.0318652359505065e-06,
+      "loss": 0.5139,
+      "step": 6229
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3083853721618652,
+      "learning_rate": 1.0306031208418277e-06,
+      "loss": 0.5349,
+      "step": 6230
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3209140300750732,
+      "learning_rate": 1.0293416893810803e-06,
+      "loss": 0.5987,
+      "step": 6231
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6582406759262085,
+      "learning_rate": 1.028080941785522e-06,
+      "loss": 0.6124,
+      "step": 6232
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5571374893188477,
+      "learning_rate": 1.0268208782722876e-06,
+      "loss": 0.6194,
+      "step": 6233
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 6.914717197418213,
+      "learning_rate": 1.0255614990583995e-06,
+      "loss": 0.5848,
+      "step": 6234
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1963386535644531,
+      "learning_rate": 1.0243028043607588e-06,
+      "loss": 0.5973,
+      "step": 6235
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.909210205078125,
+      "learning_rate": 1.023044794396153e-06,
+      "loss": 0.6383,
+      "step": 6236
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.247591495513916,
+      "learning_rate": 1.0217874693812458e-06,
+      "loss": 0.643,
+      "step": 6237
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2099329233169556,
+      "learning_rate": 1.0205308295325844e-06,
+      "loss": 0.6963,
+      "step": 6238
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6623411178588867,
+      "learning_rate": 1.0192748750666053e-06,
+      "loss": 0.5453,
+      "step": 6239
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3051793575286865,
+      "learning_rate": 1.0180196061996162e-06,
+      "loss": 0.5942,
+      "step": 6240
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3945056200027466,
+      "learning_rate": 1.0167650231478132e-06,
+      "loss": 0.5567,
+      "step": 6241
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3888925313949585,
+      "learning_rate": 1.0155111261272726e-06,
+      "loss": 0.5803,
+      "step": 6242
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3073126077651978,
+      "learning_rate": 1.014257915353954e-06,
+      "loss": 0.5863,
+      "step": 6243
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.203535556793213,
+      "learning_rate": 1.0130053910436955e-06,
+      "loss": 0.5271,
+      "step": 6244
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2935106754302979,
+      "learning_rate": 1.0117535534122191e-06,
+      "loss": 0.5767,
+      "step": 6245
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3383389711380005,
+      "learning_rate": 1.0105024026751292e-06,
+      "loss": 0.6398,
+      "step": 6246
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4551408290863037,
+      "learning_rate": 1.0092519390479116e-06,
+      "loss": 0.6012,
+      "step": 6247
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4838967323303223,
+      "learning_rate": 1.0080021627459319e-06,
+      "loss": 0.6087,
+      "step": 6248
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7100906372070312,
+      "learning_rate": 1.0067530739844394e-06,
+      "loss": 0.4806,
+      "step": 6249
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.233081579208374,
+      "learning_rate": 1.005504672978564e-06,
+      "loss": 0.6828,
+      "step": 6250
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1398285627365112,
+      "learning_rate": 1.0042569599433183e-06,
+      "loss": 0.5084,
+      "step": 6251
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.179826021194458,
+      "learning_rate": 1.003009935093594e-06,
+      "loss": 0.5361,
+      "step": 6252
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7407220602035522,
+      "learning_rate": 1.0017635986441664e-06,
+      "loss": 0.5924,
+      "step": 6253
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.390802025794983,
+      "learning_rate": 1.0005179508096929e-06,
+      "loss": 0.6056,
+      "step": 6254
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1375401020050049,
+      "learning_rate": 9.992729918047085e-07,
+      "loss": 0.5618,
+      "step": 6255
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4758859872817993,
+      "learning_rate": 9.980287218436324e-07,
+      "loss": 0.6488,
+      "step": 6256
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.653335094451904,
+      "learning_rate": 9.967851411407685e-07,
+      "loss": 0.686,
+      "step": 6257
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4219043254852295,
+      "learning_rate": 9.955422499102941e-07,
+      "loss": 0.5705,
+      "step": 6258
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2233293056488037,
+      "learning_rate": 9.943000483662735e-07,
+      "loss": 0.4663,
+      "step": 6259
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3139796257019043,
+      "learning_rate": 9.930585367226508e-07,
+      "loss": 0.6691,
+      "step": 6260
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.184753179550171,
+      "learning_rate": 9.918177151932522e-07,
+      "loss": 0.5962,
+      "step": 6261
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4073246717453003,
+      "learning_rate": 9.90577583991782e-07,
+      "loss": 0.5851,
+      "step": 6262
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6594538688659668,
+      "learning_rate": 9.893381433318277e-07,
+      "loss": 0.6553,
+      "step": 6263
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5163381099700928,
+      "learning_rate": 9.88099393426859e-07,
+      "loss": 0.6235,
+      "step": 6264
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2041316032409668,
+      "learning_rate": 9.868613344902262e-07,
+      "loss": 0.5982,
+      "step": 6265
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3843533992767334,
+      "learning_rate": 9.856239667351569e-07,
+      "loss": 0.5743,
+      "step": 6266
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2988375425338745,
+      "learning_rate": 9.843872903747641e-07,
+      "loss": 0.6007,
+      "step": 6267
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8280378580093384,
+      "learning_rate": 9.831513056220405e-07,
+      "loss": 0.5599,
+      "step": 6268
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.389329433441162,
+      "learning_rate": 9.819160126898598e-07,
+      "loss": 0.568,
+      "step": 6269
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.785888195037842,
+      "learning_rate": 9.806814117909742e-07,
+      "loss": 0.6587,
+      "step": 6270
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6062169075012207,
+      "learning_rate": 9.794475031380195e-07,
+      "loss": 0.6054,
+      "step": 6271
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4643789529800415,
+      "learning_rate": 9.782142869435119e-07,
+      "loss": 0.5829,
+      "step": 6272
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3099743127822876,
+      "learning_rate": 9.769817634198475e-07,
+      "loss": 0.611,
+      "step": 6273
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5942202806472778,
+      "learning_rate": 9.757499327793035e-07,
+      "loss": 0.5887,
+      "step": 6274
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.8324639797210693,
+      "learning_rate": 9.745187952340374e-07,
+      "loss": 0.5849,
+      "step": 6275
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.438779592514038,
+      "learning_rate": 9.732883509960895e-07,
+      "loss": 0.5977,
+      "step": 6276
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2395045757293701,
+      "learning_rate": 9.720586002773757e-07,
+      "loss": 0.5873,
+      "step": 6277
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4779880046844482,
+      "learning_rate": 9.70829543289697e-07,
+      "loss": 0.5937,
+      "step": 6278
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.641114592552185,
+      "learning_rate": 9.696011802447337e-07,
+      "loss": 0.5943,
+      "step": 6279
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8073025941848755,
+      "learning_rate": 9.683735113540471e-07,
+      "loss": 0.5827,
+      "step": 6280
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2064876556396484,
+      "learning_rate": 9.671465368290761e-07,
+      "loss": 0.587,
+      "step": 6281
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4570554494857788,
+      "learning_rate": 9.659202568811437e-07,
+      "loss": 0.5751,
+      "step": 6282
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5656616687774658,
+      "learning_rate": 9.64694671721451e-07,
+      "loss": 0.5457,
+      "step": 6283
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6131380796432495,
+      "learning_rate": 9.634697815610815e-07,
+      "loss": 0.5962,
+      "step": 6284
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3118836879730225,
+      "learning_rate": 9.622455866109958e-07,
+      "loss": 0.5983,
+      "step": 6285
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4274793863296509,
+      "learning_rate": 9.610220870820374e-07,
+      "loss": 0.6167,
+      "step": 6286
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.510737419128418,
+      "learning_rate": 9.59799283184929e-07,
+      "loss": 0.6158,
+      "step": 6287
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8258140087127686,
+      "learning_rate": 9.585771751302753e-07,
+      "loss": 0.6055,
+      "step": 6288
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1926887035369873,
+      "learning_rate": 9.573557631285568e-07,
+      "loss": 0.687,
+      "step": 6289
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6286548376083374,
+      "learning_rate": 9.56135047390137e-07,
+      "loss": 0.6492,
+      "step": 6290
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3712109327316284,
+      "learning_rate": 9.549150281252633e-07,
+      "loss": 0.5782,
+      "step": 6291
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3473654985427856,
+      "learning_rate": 9.536957055440555e-07,
+      "loss": 0.6219,
+      "step": 6292
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4779305458068848,
+      "learning_rate": 9.524770798565174e-07,
+      "loss": 0.5425,
+      "step": 6293
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.329278826713562,
+      "learning_rate": 9.512591512725344e-07,
+      "loss": 0.6175,
+      "step": 6294
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6562693119049072,
+      "learning_rate": 9.500419200018695e-07,
+      "loss": 0.5724,
+      "step": 6295
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2669323682785034,
+      "learning_rate": 9.488253862541641e-07,
+      "loss": 0.5622,
+      "step": 6296
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.161300539970398,
+      "learning_rate": 9.476095502389432e-07,
+      "loss": 0.5179,
+      "step": 6297
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5029854774475098,
+      "learning_rate": 9.463944121656088e-07,
+      "loss": 0.6479,
+      "step": 6298
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.1170594692230225,
+      "learning_rate": 9.451799722434462e-07,
+      "loss": 0.5316,
+      "step": 6299
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3949368000030518,
+      "learning_rate": 9.439662306816144e-07,
+      "loss": 0.6038,
+      "step": 6300
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2971605062484741,
+      "learning_rate": 9.427531876891577e-07,
+      "loss": 0.618,
+      "step": 6301
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5163462162017822,
+      "learning_rate": 9.415408434749984e-07,
+      "loss": 0.6335,
+      "step": 6302
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2938801050186157,
+      "learning_rate": 9.403291982479368e-07,
+      "loss": 0.6646,
+      "step": 6303
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.369942545890808,
+      "learning_rate": 9.391182522166548e-07,
+      "loss": 0.5532,
+      "step": 6304
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.317028284072876,
+      "learning_rate": 9.379080055897133e-07,
+      "loss": 0.5684,
+      "step": 6305
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.504163146018982,
+      "learning_rate": 9.366984585755545e-07,
+      "loss": 0.6464,
+      "step": 6306
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.22713041305542,
+      "learning_rate": 9.354896113824935e-07,
+      "loss": 0.6153,
+      "step": 6307
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.341958999633789,
+      "learning_rate": 9.342814642187336e-07,
+      "loss": 0.5821,
+      "step": 6308
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6462088823318481,
+      "learning_rate": 9.330740172923547e-07,
+      "loss": 0.4702,
+      "step": 6309
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.349832534790039,
+      "learning_rate": 9.318672708113113e-07,
+      "loss": 0.6108,
+      "step": 6310
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3375132083892822,
+      "learning_rate": 9.306612249834429e-07,
+      "loss": 0.6344,
+      "step": 6311
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3548191785812378,
+      "learning_rate": 9.294558800164655e-07,
+      "loss": 0.5802,
+      "step": 6312
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3934367895126343,
+      "learning_rate": 9.282512361179779e-07,
+      "loss": 0.509,
+      "step": 6313
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4095515012741089,
+      "learning_rate": 9.270472934954516e-07,
+      "loss": 0.5798,
+      "step": 6314
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2284436225891113,
+      "learning_rate": 9.258440523562434e-07,
+      "loss": 0.6162,
+      "step": 6315
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4372409582138062,
+      "learning_rate": 9.246415129075869e-07,
+      "loss": 0.6458,
+      "step": 6316
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.468173861503601,
+      "learning_rate": 9.234396753565966e-07,
+      "loss": 0.5612,
+      "step": 6317
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2699594497680664,
+      "learning_rate": 9.222385399102613e-07,
+      "loss": 0.7351,
+      "step": 6318
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6350229978561401,
+      "learning_rate": 9.210381067754542e-07,
+      "loss": 0.5474,
+      "step": 6319
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.289797306060791,
+      "learning_rate": 9.198383761589247e-07,
+      "loss": 0.5446,
+      "step": 6320
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2473235130310059,
+      "learning_rate": 9.186393482673045e-07,
+      "loss": 0.5478,
+      "step": 6321
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.0618834495544434,
+      "learning_rate": 9.17441023307098e-07,
+      "loss": 0.6916,
+      "step": 6322
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.326453447341919,
+      "learning_rate": 9.162434014846944e-07,
+      "loss": 0.6769,
+      "step": 6323
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.895493745803833,
+      "learning_rate": 9.150464830063593e-07,
+      "loss": 0.5819,
+      "step": 6324
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2672816514968872,
+      "learning_rate": 9.13850268078238e-07,
+      "loss": 0.6084,
+      "step": 6325
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3241233825683594,
+      "learning_rate": 9.126547569063538e-07,
+      "loss": 0.5891,
+      "step": 6326
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.417624831199646,
+      "learning_rate": 9.114599496966093e-07,
+      "loss": 0.5614,
+      "step": 6327
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.364613652229309,
+      "learning_rate": 9.10265846654787e-07,
+      "loss": 0.5463,
+      "step": 6328
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3848798274993896,
+      "learning_rate": 9.090724479865443e-07,
+      "loss": 0.6403,
+      "step": 6329
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2594083547592163,
+      "learning_rate": 9.07879753897421e-07,
+      "loss": 0.5747,
+      "step": 6330
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2425966262817383,
+      "learning_rate": 9.066877645928351e-07,
+      "loss": 0.5759,
+      "step": 6331
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5962731838226318,
+      "learning_rate": 9.054964802780836e-07,
+      "loss": 0.6179,
+      "step": 6332
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6983554363250732,
+      "learning_rate": 9.043059011583378e-07,
+      "loss": 0.529,
+      "step": 6333
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4406030178070068,
+      "learning_rate": 9.031160274386524e-07,
+      "loss": 0.5584,
+      "step": 6334
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1120250225067139,
+      "learning_rate": 9.019268593239589e-07,
+      "loss": 0.5092,
+      "step": 6335
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3159717321395874,
+      "learning_rate": 9.007383970190692e-07,
+      "loss": 0.5571,
+      "step": 6336
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4577220678329468,
+      "learning_rate": 8.995506407286681e-07,
+      "loss": 0.5735,
+      "step": 6337
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.6757638454437256,
+      "learning_rate": 8.983635906573246e-07,
+      "loss": 0.6404,
+      "step": 6338
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3575419187545776,
+      "learning_rate": 8.971772470094841e-07,
+      "loss": 0.5763,
+      "step": 6339
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1019004583358765,
+      "learning_rate": 8.959916099894706e-07,
+      "loss": 0.6899,
+      "step": 6340
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2935214042663574,
+      "learning_rate": 8.948066798014832e-07,
+      "loss": 0.6393,
+      "step": 6341
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3130396604537964,
+      "learning_rate": 8.936224566496049e-07,
+      "loss": 0.5713,
+      "step": 6342
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.318345308303833,
+      "learning_rate": 8.924389407377948e-07,
+      "loss": 0.544,
+      "step": 6343
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3377659320831299,
+      "learning_rate": 8.912561322698859e-07,
+      "loss": 0.5473,
+      "step": 6344
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4414652585983276,
+      "learning_rate": 8.900740314495953e-07,
+      "loss": 0.596,
+      "step": 6345
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5321376323699951,
+      "learning_rate": 8.888926384805158e-07,
+      "loss": 0.5581,
+      "step": 6346
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2531845569610596,
+      "learning_rate": 8.877119535661189e-07,
+      "loss": 0.5722,
+      "step": 6347
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3081930875778198,
+      "learning_rate": 8.865319769097513e-07,
+      "loss": 0.5443,
+      "step": 6348
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.565206527709961,
+      "learning_rate": 8.853527087146413e-07,
+      "loss": 0.6366,
+      "step": 6349
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6505012512207031,
+      "learning_rate": 8.841741491838951e-07,
+      "loss": 0.6175,
+      "step": 6350
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3950464725494385,
+      "learning_rate": 8.829962985204931e-07,
+      "loss": 0.607,
+      "step": 6351
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5542408227920532,
+      "learning_rate": 8.818191569272977e-07,
+      "loss": 0.6109,
+      "step": 6352
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.618653416633606,
+      "learning_rate": 8.806427246070476e-07,
+      "loss": 0.5526,
+      "step": 6353
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.473083734512329,
+      "learning_rate": 8.794670017623602e-07,
+      "loss": 0.6048,
+      "step": 6354
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2885507345199585,
+      "learning_rate": 8.782919885957276e-07,
+      "loss": 0.6216,
+      "step": 6355
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6841661930084229,
+      "learning_rate": 8.771176853095226e-07,
+      "loss": 0.6373,
+      "step": 6356
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.770217180252075,
+      "learning_rate": 8.759440921059958e-07,
+      "loss": 0.6398,
+      "step": 6357
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.275274395942688,
+      "learning_rate": 8.747712091872751e-07,
+      "loss": 0.71,
+      "step": 6358
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.0190229415893555,
+      "learning_rate": 8.735990367553643e-07,
+      "loss": 0.5194,
+      "step": 6359
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.600679636001587,
+      "learning_rate": 8.72427575012148e-07,
+      "loss": 0.5969,
+      "step": 6360
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.613568663597107,
+      "learning_rate": 8.712568241593866e-07,
+      "loss": 0.5324,
+      "step": 6361
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3485932350158691,
+      "learning_rate": 8.700867843987165e-07,
+      "loss": 0.5907,
+      "step": 6362
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.446373462677002,
+      "learning_rate": 8.689174559316544e-07,
+      "loss": 0.5736,
+      "step": 6363
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2776132822036743,
+      "learning_rate": 8.677488389595923e-07,
+      "loss": 0.4848,
+      "step": 6364
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2803000211715698,
+      "learning_rate": 8.665809336838032e-07,
+      "loss": 0.5473,
+      "step": 6365
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1983240842819214,
+      "learning_rate": 8.654137403054324e-07,
+      "loss": 0.5795,
+      "step": 6366
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4812618494033813,
+      "learning_rate": 8.642472590255063e-07,
+      "loss": 0.6122,
+      "step": 6367
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6410510540008545,
+      "learning_rate": 8.630814900449269e-07,
+      "loss": 0.6678,
+      "step": 6368
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.134711742401123,
+      "learning_rate": 8.619164335644764e-07,
+      "loss": 0.4751,
+      "step": 6369
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3798121213912964,
+      "learning_rate": 8.607520897848093e-07,
+      "loss": 0.6079,
+      "step": 6370
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5486456155776978,
+      "learning_rate": 8.595884589064618e-07,
+      "loss": 0.5501,
+      "step": 6371
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4457107782363892,
+      "learning_rate": 8.584255411298448e-07,
+      "loss": 0.5237,
+      "step": 6372
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3767578601837158,
+      "learning_rate": 8.57263336655249e-07,
+      "loss": 0.5954,
+      "step": 6373
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.325913429260254,
+      "learning_rate": 8.561018456828379e-07,
+      "loss": 0.5758,
+      "step": 6374
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5465365648269653,
+      "learning_rate": 8.549410684126547e-07,
+      "loss": 0.5237,
+      "step": 6375
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5502859354019165,
+      "learning_rate": 8.537810050446238e-07,
+      "loss": 0.5968,
+      "step": 6376
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3833034038543701,
+      "learning_rate": 8.526216557785383e-07,
+      "loss": 0.4794,
+      "step": 6377
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4041199684143066,
+      "learning_rate": 8.51463020814074e-07,
+      "loss": 0.5775,
+      "step": 6378
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3076719045639038,
+      "learning_rate": 8.503051003507823e-07,
+      "loss": 0.5465,
+      "step": 6379
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.535282015800476,
+      "learning_rate": 8.49147894588092e-07,
+      "loss": 0.6229,
+      "step": 6380
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4696766138076782,
+      "learning_rate": 8.479914037253073e-07,
+      "loss": 0.5713,
+      "step": 6381
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4727602005004883,
+      "learning_rate": 8.468356279616102e-07,
+      "loss": 0.5585,
+      "step": 6382
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4697345495224,
+      "learning_rate": 8.456805674960594e-07,
+      "loss": 0.5728,
+      "step": 6383
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5066434144973755,
+      "learning_rate": 8.44526222527593e-07,
+      "loss": 0.6151,
+      "step": 6384
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.423194408416748,
+      "learning_rate": 8.433725932550207e-07,
+      "loss": 0.5928,
+      "step": 6385
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6643799543380737,
+      "learning_rate": 8.422196798770321e-07,
+      "loss": 0.6388,
+      "step": 6386
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3613812923431396,
+      "learning_rate": 8.410674825921938e-07,
+      "loss": 0.5786,
+      "step": 6387
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1162885427474976,
+      "learning_rate": 8.399160015989494e-07,
+      "loss": 0.5944,
+      "step": 6388
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5724189281463623,
+      "learning_rate": 8.387652370956151e-07,
+      "loss": 0.5654,
+      "step": 6389
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.359290361404419,
+      "learning_rate": 8.37615189280389e-07,
+      "loss": 0.5796,
+      "step": 6390
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3778443336486816,
+      "learning_rate": 8.36465858351343e-07,
+      "loss": 0.5938,
+      "step": 6391
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4014848470687866,
+      "learning_rate": 8.353172445064268e-07,
+      "loss": 0.5879,
+      "step": 6392
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.405988097190857,
+      "learning_rate": 8.341693479434626e-07,
+      "loss": 0.5998,
+      "step": 6393
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4378615617752075,
+      "learning_rate": 8.330221688601559e-07,
+      "loss": 0.6279,
+      "step": 6394
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.794307827949524,
+      "learning_rate": 8.318757074540845e-07,
+      "loss": 0.6281,
+      "step": 6395
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.450239896774292,
+      "learning_rate": 8.307299639227013e-07,
+      "loss": 0.5631,
+      "step": 6396
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.587395191192627,
+      "learning_rate": 8.295849384633381e-07,
+      "loss": 0.5966,
+      "step": 6397
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.633143663406372,
+      "learning_rate": 8.284406312732024e-07,
+      "loss": 0.587,
+      "step": 6398
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2260631322860718,
+      "learning_rate": 8.272970425493793e-07,
+      "loss": 0.5618,
+      "step": 6399
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3946280479431152,
+      "learning_rate": 8.261541724888256e-07,
+      "loss": 0.5446,
+      "step": 6400
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1478197574615479,
+      "learning_rate": 8.250120212883794e-07,
+      "loss": 0.5878,
+      "step": 6401
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6904059648513794,
+      "learning_rate": 8.23870589144754e-07,
+      "loss": 0.6042,
+      "step": 6402
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5287965536117554,
+      "learning_rate": 8.227298762545354e-07,
+      "loss": 0.6247,
+      "step": 6403
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4237464666366577,
+      "learning_rate": 8.215898828141894e-07,
+      "loss": 0.5744,
+      "step": 6404
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4170470237731934,
+      "learning_rate": 8.204506090200565e-07,
+      "loss": 0.5751,
+      "step": 6405
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3677936792373657,
+      "learning_rate": 8.193120550683553e-07,
+      "loss": 0.6154,
+      "step": 6406
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2820465564727783,
+      "learning_rate": 8.181742211551757e-07,
+      "loss": 0.5579,
+      "step": 6407
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6385856866836548,
+      "learning_rate": 8.170371074764872e-07,
+      "loss": 0.6446,
+      "step": 6408
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.253929615020752,
+      "learning_rate": 8.159007142281356e-07,
+      "loss": 0.55,
+      "step": 6409
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2222574949264526,
+      "learning_rate": 8.147650416058406e-07,
+      "loss": 0.5793,
+      "step": 6410
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3179538249969482,
+      "learning_rate": 8.136300898051996e-07,
+      "loss": 0.616,
+      "step": 6411
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3393512964248657,
+      "learning_rate": 8.124958590216841e-07,
+      "loss": 0.6454,
+      "step": 6412
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.482521414756775,
+      "learning_rate": 8.11362349450644e-07,
+      "loss": 0.5641,
+      "step": 6413
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3834075927734375,
+      "learning_rate": 8.102295612873007e-07,
+      "loss": 0.5252,
+      "step": 6414
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.378448724746704,
+      "learning_rate": 8.090974947267555e-07,
+      "loss": 0.5853,
+      "step": 6415
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5157331228256226,
+      "learning_rate": 8.07966149963983e-07,
+      "loss": 0.6003,
+      "step": 6416
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1425423622131348,
+      "learning_rate": 8.068355271938366e-07,
+      "loss": 0.7038,
+      "step": 6417
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4335917234420776,
+      "learning_rate": 8.057056266110397e-07,
+      "loss": 0.6066,
+      "step": 6418
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3131961822509766,
+      "learning_rate": 8.04576448410197e-07,
+      "loss": 0.4857,
+      "step": 6419
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3881715536117554,
+      "learning_rate": 8.034479927857852e-07,
+      "loss": 0.6013,
+      "step": 6420
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.620413064956665,
+      "learning_rate": 8.023202599321605e-07,
+      "loss": 0.6436,
+      "step": 6421
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1148796081542969,
+      "learning_rate": 8.011932500435482e-07,
+      "loss": 0.6773,
+      "step": 6422
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3660540580749512,
+      "learning_rate": 8.000669633140551e-07,
+      "loss": 0.6043,
+      "step": 6423
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3749091625213623,
+      "learning_rate": 7.989413999376605e-07,
+      "loss": 0.5466,
+      "step": 6424
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1317570209503174,
+      "learning_rate": 7.978165601082211e-07,
+      "loss": 0.4955,
+      "step": 6425
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1026514768600464,
+      "learning_rate": 7.966924440194657e-07,
+      "loss": 0.5888,
+      "step": 6426
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2896405458450317,
+      "learning_rate": 7.95569051865e-07,
+      "loss": 0.5944,
+      "step": 6427
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2608617544174194,
+      "learning_rate": 7.944463838383093e-07,
+      "loss": 0.5025,
+      "step": 6428
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1657538414001465,
+      "learning_rate": 7.933244401327472e-07,
+      "loss": 0.6009,
+      "step": 6429
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.1015689373016357,
+      "learning_rate": 7.922032209415459e-07,
+      "loss": 0.6016,
+      "step": 6430
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2412078380584717,
+      "learning_rate": 7.91082726457813e-07,
+      "loss": 0.5762,
+      "step": 6431
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2768174409866333,
+      "learning_rate": 7.899629568745327e-07,
+      "loss": 0.5765,
+      "step": 6432
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.256107211112976,
+      "learning_rate": 7.888439123845599e-07,
+      "loss": 0.6115,
+      "step": 6433
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.3772521018981934,
+      "learning_rate": 7.877255931806277e-07,
+      "loss": 0.546,
+      "step": 6434
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5032480955123901,
+      "learning_rate": 7.866079994553444e-07,
+      "loss": 0.6249,
+      "step": 6435
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3456975221633911,
+      "learning_rate": 7.854911314011942e-07,
+      "loss": 0.5095,
+      "step": 6436
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1659880876541138,
+      "learning_rate": 7.843749892105323e-07,
+      "loss": 0.5498,
+      "step": 6437
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2211744785308838,
+      "learning_rate": 7.832595730755927e-07,
+      "loss": 0.5963,
+      "step": 6438
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3915050029754639,
+      "learning_rate": 7.821448831884831e-07,
+      "loss": 0.6141,
+      "step": 6439
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.694148302078247,
+      "learning_rate": 7.810309197411875e-07,
+      "loss": 0.6192,
+      "step": 6440
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6715288162231445,
+      "learning_rate": 7.799176829255612e-07,
+      "loss": 0.6291,
+      "step": 6441
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2998833656311035,
+      "learning_rate": 7.788051729333373e-07,
+      "loss": 0.6008,
+      "step": 6442
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.360581874847412,
+      "learning_rate": 7.776933899561239e-07,
+      "loss": 0.5631,
+      "step": 6443
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3779425621032715,
+      "learning_rate": 7.765823341854017e-07,
+      "loss": 0.488,
+      "step": 6444
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4948135614395142,
+      "learning_rate": 7.754720058125293e-07,
+      "loss": 0.6333,
+      "step": 6445
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.061363935470581,
+      "learning_rate": 7.743624050287363e-07,
+      "loss": 0.532,
+      "step": 6446
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2560065984725952,
+      "learning_rate": 7.732535320251316e-07,
+      "loss": 0.5488,
+      "step": 6447
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.488149642944336,
+      "learning_rate": 7.721453869926926e-07,
+      "loss": 0.5659,
+      "step": 6448
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2953224182128906,
+      "learning_rate": 7.710379701222764e-07,
+      "loss": 0.6267,
+      "step": 6449
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.0666239261627197,
+      "learning_rate": 7.699312816046139e-07,
+      "loss": 0.6492,
+      "step": 6450
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.348408579826355,
+      "learning_rate": 7.688253216303082e-07,
+      "loss": 0.5643,
+      "step": 6451
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6481759548187256,
+      "learning_rate": 7.677200903898386e-07,
+      "loss": 0.6163,
+      "step": 6452
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.656936526298523,
+      "learning_rate": 7.666155880735593e-07,
+      "loss": 0.6407,
+      "step": 6453
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1746383905410767,
+      "learning_rate": 7.655118148716989e-07,
+      "loss": 0.75,
+      "step": 6454
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4573662281036377,
+      "learning_rate": 7.644087709743586e-07,
+      "loss": 0.5225,
+      "step": 6455
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2147821187973022,
+      "learning_rate": 7.633064565715159e-07,
+      "loss": 0.7242,
+      "step": 6456
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5926049947738647,
+      "learning_rate": 7.622048718530218e-07,
+      "loss": 0.5808,
+      "step": 6457
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6897194385528564,
+      "learning_rate": 7.611040170086032e-07,
+      "loss": 0.6736,
+      "step": 6458
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5272244215011597,
+      "learning_rate": 7.60003892227858e-07,
+      "loss": 0.6596,
+      "step": 6459
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4822163581848145,
+      "learning_rate": 7.589044977002607e-07,
+      "loss": 0.677,
+      "step": 6460
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.653491497039795,
+      "learning_rate": 7.578058336151611e-07,
+      "loss": 0.5817,
+      "step": 6461
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3933191299438477,
+      "learning_rate": 7.5670790016178e-07,
+      "loss": 0.6101,
+      "step": 6462
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2626433372497559,
+      "learning_rate": 7.556106975292155e-07,
+      "loss": 0.5215,
+      "step": 6463
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1545683145523071,
+      "learning_rate": 7.545142259064376e-07,
+      "loss": 0.5581,
+      "step": 6464
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.349684476852417,
+      "learning_rate": 7.534184854822929e-07,
+      "loss": 0.7176,
+      "step": 6465
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2531545162200928,
+      "learning_rate": 7.523234764454978e-07,
+      "loss": 0.5055,
+      "step": 6466
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6439186334609985,
+      "learning_rate": 7.512291989846465e-07,
+      "loss": 0.5796,
+      "step": 6467
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1926203966140747,
+      "learning_rate": 7.501356532882064e-07,
+      "loss": 0.468,
+      "step": 6468
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6254560947418213,
+      "learning_rate": 7.490428395445198e-07,
+      "loss": 0.6552,
+      "step": 6469
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2992578744888306,
+      "learning_rate": 7.479507579417989e-07,
+      "loss": 0.648,
+      "step": 6470
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4693169593811035,
+      "learning_rate": 7.46859408668133e-07,
+      "loss": 0.5684,
+      "step": 6471
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5550191402435303,
+      "learning_rate": 7.457687919114864e-07,
+      "loss": 0.5953,
+      "step": 6472
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1891635656356812,
+      "learning_rate": 7.446789078596961e-07,
+      "loss": 0.4812,
+      "step": 6473
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6289045810699463,
+      "learning_rate": 7.435897567004696e-07,
+      "loss": 0.6572,
+      "step": 6474
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2836540937423706,
+      "learning_rate": 7.425013386213931e-07,
+      "loss": 0.5618,
+      "step": 6475
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.0774881839752197,
+      "learning_rate": 7.414136538099242e-07,
+      "loss": 0.6091,
+      "step": 6476
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2852836847305298,
+      "learning_rate": 7.403267024533956e-07,
+      "loss": 0.584,
+      "step": 6477
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.08048677444458,
+      "learning_rate": 7.392404847390089e-07,
+      "loss": 0.7044,
+      "step": 6478
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3461830615997314,
+      "learning_rate": 7.381550008538468e-07,
+      "loss": 0.6609,
+      "step": 6479
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3070403337478638,
+      "learning_rate": 7.37070250984862e-07,
+      "loss": 0.5704,
+      "step": 6480
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1511784791946411,
+      "learning_rate": 7.359862353188774e-07,
+      "loss": 0.6785,
+      "step": 6481
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5748064517974854,
+      "learning_rate": 7.34902954042595e-07,
+      "loss": 0.5357,
+      "step": 6482
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5832569599151611,
+      "learning_rate": 7.33820407342587e-07,
+      "loss": 0.6216,
+      "step": 6483
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5341449975967407,
+      "learning_rate": 7.327385954053023e-07,
+      "loss": 0.6461,
+      "step": 6484
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2405881881713867,
+      "learning_rate": 7.316575184170577e-07,
+      "loss": 0.6008,
+      "step": 6485
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6158428192138672,
+      "learning_rate": 7.30577176564048e-07,
+      "loss": 0.6164,
+      "step": 6486
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4766799211502075,
+      "learning_rate": 7.294975700323404e-07,
+      "loss": 0.5562,
+      "step": 6487
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7121096849441528,
+      "learning_rate": 7.284186990078767e-07,
+      "loss": 0.5626,
+      "step": 6488
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9860489964485168,
+      "learning_rate": 7.273405636764675e-07,
+      "loss": 0.6356,
+      "step": 6489
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6255037784576416,
+      "learning_rate": 7.262631642238011e-07,
+      "loss": 0.5731,
+      "step": 6490
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4038243293762207,
+      "learning_rate": 7.25186500835438e-07,
+      "loss": 0.6349,
+      "step": 6491
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1845976114273071,
+      "learning_rate": 7.241105736968124e-07,
+      "loss": 0.5589,
+      "step": 6492
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3688479661941528,
+      "learning_rate": 7.230353829932285e-07,
+      "loss": 0.6446,
+      "step": 6493
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.30573570728302,
+      "learning_rate": 7.219609289098672e-07,
+      "loss": 0.6097,
+      "step": 6494
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4633636474609375,
+      "learning_rate": 7.208872116317822e-07,
+      "loss": 0.633,
+      "step": 6495
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5460313558578491,
+      "learning_rate": 7.198142313438983e-07,
+      "loss": 0.5651,
+      "step": 6496
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4103299379348755,
+      "learning_rate": 7.187419882310148e-07,
+      "loss": 0.5557,
+      "step": 6497
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.33799147605896,
+      "learning_rate": 7.176704824778052e-07,
+      "loss": 0.5859,
+      "step": 6498
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6135934591293335,
+      "learning_rate": 7.165997142688124e-07,
+      "loss": 0.5929,
+      "step": 6499
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3209842443466187,
+      "learning_rate": 7.155296837884557e-07,
+      "loss": 0.6129,
+      "step": 6500
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2033591270446777,
+      "learning_rate": 7.144603912210257e-07,
+      "loss": 0.6195,
+      "step": 6501
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1654304265975952,
+      "learning_rate": 7.133918367506876e-07,
+      "loss": 0.5069,
+      "step": 6502
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3715147972106934,
+      "learning_rate": 7.123240205614756e-07,
+      "loss": 0.5825,
+      "step": 6503
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4160212278366089,
+      "learning_rate": 7.112569428373012e-07,
+      "loss": 0.5936,
+      "step": 6504
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3100764751434326,
+      "learning_rate": 7.101906037619466e-07,
+      "loss": 0.5829,
+      "step": 6505
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.678920865058899,
+      "learning_rate": 7.091250035190678e-07,
+      "loss": 0.6412,
+      "step": 6506
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4690093994140625,
+      "learning_rate": 7.080601422921901e-07,
+      "loss": 0.5738,
+      "step": 6507
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3475110530853271,
+      "learning_rate": 7.069960202647169e-07,
+      "loss": 0.5587,
+      "step": 6508
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3803815841674805,
+      "learning_rate": 7.059326376199199e-07,
+      "loss": 0.6277,
+      "step": 6509
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3090627193450928,
+      "learning_rate": 7.048699945409477e-07,
+      "loss": 0.5846,
+      "step": 6510
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.225256323814392,
+      "learning_rate": 7.038080912108152e-07,
+      "loss": 0.5646,
+      "step": 6511
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3957878351211548,
+      "learning_rate": 7.027469278124155e-07,
+      "loss": 0.5509,
+      "step": 6512
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2957148551940918,
+      "learning_rate": 7.016865045285143e-07,
+      "loss": 0.6144,
+      "step": 6513
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1870754957199097,
+      "learning_rate": 7.00626821541745e-07,
+      "loss": 0.6008,
+      "step": 6514
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.483576774597168,
+      "learning_rate": 6.995678790346178e-07,
+      "loss": 0.6274,
+      "step": 6515
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2268551588058472,
+      "learning_rate": 6.985096771895139e-07,
+      "loss": 0.5665,
+      "step": 6516
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.424389362335205,
+      "learning_rate": 6.974522161886882e-07,
+      "loss": 0.5598,
+      "step": 6517
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3461567163467407,
+      "learning_rate": 6.963954962142644e-07,
+      "loss": 0.6131,
+      "step": 6518
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4819175004959106,
+      "learning_rate": 6.953395174482425e-07,
+      "loss": 0.6464,
+      "step": 6519
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.284936547279358,
+      "learning_rate": 6.94284280072493e-07,
+      "loss": 0.6178,
+      "step": 6520
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.185971736907959,
+      "learning_rate": 6.932297842687607e-07,
+      "loss": 0.6414,
+      "step": 6521
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4137840270996094,
+      "learning_rate": 6.921760302186587e-07,
+      "loss": 0.6293,
+      "step": 6522
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.347339391708374,
+      "learning_rate": 6.911230181036755e-07,
+      "loss": 0.5987,
+      "step": 6523
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.255189299583435,
+      "learning_rate": 6.90070748105171e-07,
+      "loss": 0.6005,
+      "step": 6524
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5423235893249512,
+      "learning_rate": 6.890192204043789e-07,
+      "loss": 0.6264,
+      "step": 6525
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6151106357574463,
+      "learning_rate": 6.879684351824012e-07,
+      "loss": 0.7098,
+      "step": 6526
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.403106451034546,
+      "learning_rate": 6.869183926202149e-07,
+      "loss": 0.5491,
+      "step": 6527
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.432671308517456,
+      "learning_rate": 6.858690928986689e-07,
+      "loss": 0.6377,
+      "step": 6528
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3803349733352661,
+      "learning_rate": 6.84820536198485e-07,
+      "loss": 0.5884,
+      "step": 6529
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3720000982284546,
+      "learning_rate": 6.837727227002522e-07,
+      "loss": 0.6173,
+      "step": 6530
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2312438488006592,
+      "learning_rate": 6.827256525844384e-07,
+      "loss": 0.5887,
+      "step": 6531
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3636536598205566,
+      "learning_rate": 6.816793260313798e-07,
+      "loss": 0.5294,
+      "step": 6532
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2737432718276978,
+      "learning_rate": 6.806337432212834e-07,
+      "loss": 0.5848,
+      "step": 6533
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3652080297470093,
+      "learning_rate": 6.795889043342302e-07,
+      "loss": 0.6028,
+      "step": 6534
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.309290885925293,
+      "learning_rate": 6.785448095501728e-07,
+      "loss": 0.5539,
+      "step": 6535
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4194649457931519,
+      "learning_rate": 6.775014590489359e-07,
+      "loss": 0.5934,
+      "step": 6536
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6991386413574219,
+      "learning_rate": 6.76458853010214e-07,
+      "loss": 0.6732,
+      "step": 6537
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4001078605651855,
+      "learning_rate": 6.75416991613575e-07,
+      "loss": 0.6728,
+      "step": 6538
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0708200931549072,
+      "learning_rate": 6.743758750384588e-07,
+      "loss": 0.6719,
+      "step": 6539
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.4985439777374268,
+      "learning_rate": 6.733355034641776e-07,
+      "loss": 0.5891,
+      "step": 6540
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2028065919876099,
+      "learning_rate": 6.722958770699123e-07,
+      "loss": 0.6056,
+      "step": 6541
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2086281776428223,
+      "learning_rate": 6.712569960347182e-07,
+      "loss": 0.5668,
+      "step": 6542
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2189396619796753,
+      "learning_rate": 6.702188605375226e-07,
+      "loss": 0.7119,
+      "step": 6543
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2520568370819092,
+      "learning_rate": 6.691814707571209e-07,
+      "loss": 0.5847,
+      "step": 6544
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4773744344711304,
+      "learning_rate": 6.681448268721841e-07,
+      "loss": 0.5806,
+      "step": 6545
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.519181728363037,
+      "learning_rate": 6.671089290612526e-07,
+      "loss": 0.5644,
+      "step": 6546
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.489691972732544,
+      "learning_rate": 6.660737775027381e-07,
+      "loss": 0.6373,
+      "step": 6547
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2815091609954834,
+      "learning_rate": 6.65039372374926e-07,
+      "loss": 0.5611,
+      "step": 6548
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2832820415496826,
+      "learning_rate": 6.640057138559702e-07,
+      "loss": 0.6027,
+      "step": 6549
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3316545486450195,
+      "learning_rate": 6.629728021238991e-07,
+      "loss": 0.5918,
+      "step": 6550
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3692222833633423,
+      "learning_rate": 6.619406373566079e-07,
+      "loss": 0.6008,
+      "step": 6551
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.561578392982483,
+      "learning_rate": 6.609092197318678e-07,
+      "loss": 0.5702,
+      "step": 6552
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8879069089889526,
+      "learning_rate": 6.598785494273197e-07,
+      "loss": 0.5556,
+      "step": 6553
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5902771949768066,
+      "learning_rate": 6.588486266204758e-07,
+      "loss": 0.6338,
+      "step": 6554
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5381138324737549,
+      "learning_rate": 6.578194514887176e-07,
+      "loss": 0.5805,
+      "step": 6555
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.417212963104248,
+      "learning_rate": 6.567910242093012e-07,
+      "loss": 0.5621,
+      "step": 6556
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.422182321548462,
+      "learning_rate": 6.557633449593515e-07,
+      "loss": 0.516,
+      "step": 6557
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2522732019424438,
+      "learning_rate": 6.547364139158674e-07,
+      "loss": 0.6348,
+      "step": 6558
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4380911588668823,
+      "learning_rate": 6.537102312557137e-07,
+      "loss": 0.6165,
+      "step": 6559
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3114643096923828,
+      "learning_rate": 6.52684797155631e-07,
+      "loss": 0.5736,
+      "step": 6560
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.367963433265686,
+      "learning_rate": 6.516601117922295e-07,
+      "loss": 0.5807,
+      "step": 6561
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5235767364501953,
+      "learning_rate": 6.506361753419916e-07,
+      "loss": 0.6142,
+      "step": 6562
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.021899700164795,
+      "learning_rate": 6.496129879812673e-07,
+      "loss": 0.5628,
+      "step": 6563
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5203709602355957,
+      "learning_rate": 6.485905498862799e-07,
+      "loss": 0.6086,
+      "step": 6564
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.330996036529541,
+      "learning_rate": 6.475688612331265e-07,
+      "loss": 0.5845,
+      "step": 6565
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3727091550827026,
+      "learning_rate": 6.465479221977694e-07,
+      "loss": 0.596,
+      "step": 6566
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5101768970489502,
+      "learning_rate": 6.455277329560456e-07,
+      "loss": 0.5664,
+      "step": 6567
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3237074613571167,
+      "learning_rate": 6.445082936836616e-07,
+      "loss": 0.6286,
+      "step": 6568
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.626060962677002,
+      "learning_rate": 6.434896045561967e-07,
+      "loss": 0.5562,
+      "step": 6569
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2659837007522583,
+      "learning_rate": 6.424716657490965e-07,
+      "loss": 0.5489,
+      "step": 6570
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7719166278839111,
+      "learning_rate": 6.414544774376819e-07,
+      "loss": 0.6006,
+      "step": 6571
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.414229393005371,
+      "learning_rate": 6.404380397971432e-07,
+      "loss": 0.5606,
+      "step": 6572
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.48387610912323,
+      "learning_rate": 6.394223530025418e-07,
+      "loss": 0.5362,
+      "step": 6573
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4343132972717285,
+      "learning_rate": 6.384074172288068e-07,
+      "loss": 0.6049,
+      "step": 6574
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.396812081336975,
+      "learning_rate": 6.373932326507415e-07,
+      "loss": 0.5969,
+      "step": 6575
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5124541521072388,
+      "learning_rate": 6.363797994430182e-07,
+      "loss": 0.6075,
+      "step": 6576
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2993794679641724,
+      "learning_rate": 6.353671177801824e-07,
+      "loss": 0.6173,
+      "step": 6577
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5706703662872314,
+      "learning_rate": 6.343551878366444e-07,
+      "loss": 0.6263,
+      "step": 6578
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3450409173965454,
+      "learning_rate": 6.333440097866905e-07,
+      "loss": 0.6767,
+      "step": 6579
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1720281839370728,
+      "learning_rate": 6.323335838044753e-07,
+      "loss": 0.5835,
+      "step": 6580
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2327227592468262,
+      "learning_rate": 6.31323910064024e-07,
+      "loss": 0.7767,
+      "step": 6581
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5149428844451904,
+      "learning_rate": 6.303149887392329e-07,
+      "loss": 0.6144,
+      "step": 6582
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.06614089012146,
+      "learning_rate": 6.293068200038677e-07,
+      "loss": 0.6909,
+      "step": 6583
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6291388273239136,
+      "learning_rate": 6.28299404031566e-07,
+      "loss": 0.6292,
+      "step": 6584
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6113803386688232,
+      "learning_rate": 6.272927409958323e-07,
+      "loss": 0.6151,
+      "step": 6585
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4304330348968506,
+      "learning_rate": 6.262868310700459e-07,
+      "loss": 0.6134,
+      "step": 6586
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3153834342956543,
+      "learning_rate": 6.252816744274542e-07,
+      "loss": 0.6329,
+      "step": 6587
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.266677737236023,
+      "learning_rate": 6.242772712411754e-07,
+      "loss": 0.5337,
+      "step": 6588
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8882852792739868,
+      "learning_rate": 6.232736216841956e-07,
+      "loss": 0.6046,
+      "step": 6589
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5751484632492065,
+      "learning_rate": 6.222707259293742e-07,
+      "loss": 0.59,
+      "step": 6590
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3150429725646973,
+      "learning_rate": 6.212685841494392e-07,
+      "loss": 0.6032,
+      "step": 6591
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3497982025146484,
+      "learning_rate": 6.202671965169909e-07,
+      "loss": 0.6491,
+      "step": 6592
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6363470554351807,
+      "learning_rate": 6.192665632044959e-07,
+      "loss": 0.6138,
+      "step": 6593
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4757647514343262,
+      "learning_rate": 6.182666843842933e-07,
+      "loss": 0.6082,
+      "step": 6594
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5103785991668701,
+      "learning_rate": 6.172675602285933e-07,
+      "loss": 0.6074,
+      "step": 6595
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3814870119094849,
+      "learning_rate": 6.162691909094726e-07,
+      "loss": 0.6029,
+      "step": 6596
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.476153016090393,
+      "learning_rate": 6.152715765988815e-07,
+      "loss": 0.6853,
+      "step": 6597
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8093366622924805,
+      "learning_rate": 6.142747174686381e-07,
+      "loss": 0.5474,
+      "step": 6598
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3409085273742676,
+      "learning_rate": 6.132786136904312e-07,
+      "loss": 0.6649,
+      "step": 6599
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.276571750640869,
+      "learning_rate": 6.122832654358196e-07,
+      "loss": 0.5591,
+      "step": 6600
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5459884405136108,
+      "learning_rate": 6.112886728762324e-07,
+      "loss": 0.6567,
+      "step": 6601
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4320536851882935,
+      "learning_rate": 6.10294836182968e-07,
+      "loss": 0.6105,
+      "step": 6602
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9960110783576965,
+      "learning_rate": 6.093017555271935e-07,
+      "loss": 0.6775,
+      "step": 6603
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5592219829559326,
+      "learning_rate": 6.08309431079947e-07,
+      "loss": 0.6259,
+      "step": 6604
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4093371629714966,
+      "learning_rate": 6.073178630121363e-07,
+      "loss": 0.6485,
+      "step": 6605
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1359326839447021,
+      "learning_rate": 6.063270514945402e-07,
+      "loss": 0.5621,
+      "step": 6606
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2226295471191406,
+      "learning_rate": 6.05336996697804e-07,
+      "loss": 0.5737,
+      "step": 6607
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3089041709899902,
+      "learning_rate": 6.043476987924452e-07,
+      "loss": 0.5701,
+      "step": 6608
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5044931173324585,
+      "learning_rate": 6.033591579488501e-07,
+      "loss": 0.7168,
+      "step": 6609
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2217152118682861,
+      "learning_rate": 6.023713743372761e-07,
+      "loss": 0.5639,
+      "step": 6610
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4304804801940918,
+      "learning_rate": 6.01384348127847e-07,
+      "loss": 0.6051,
+      "step": 6611
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.248270869255066,
+      "learning_rate": 6.003980794905584e-07,
+      "loss": 0.5676,
+      "step": 6612
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.317533016204834,
+      "learning_rate": 5.994125685952757e-07,
+      "loss": 0.5705,
+      "step": 6613
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.586898922920227,
+      "learning_rate": 5.984278156117335e-07,
+      "loss": 0.6188,
+      "step": 6614
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4527963399887085,
+      "learning_rate": 5.974438207095328e-07,
+      "loss": 0.5467,
+      "step": 6615
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3087095022201538,
+      "learning_rate": 5.964605840581494e-07,
+      "loss": 0.5782,
+      "step": 6616
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2765275239944458,
+      "learning_rate": 5.954781058269265e-07,
+      "loss": 0.5261,
+      "step": 6617
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2288613319396973,
+      "learning_rate": 5.944963861850738e-07,
+      "loss": 0.6082,
+      "step": 6618
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3720436096191406,
+      "learning_rate": 5.935154253016729e-07,
+      "loss": 0.6148,
+      "step": 6619
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.081778883934021,
+      "learning_rate": 5.925352233456749e-07,
+      "loss": 0.5368,
+      "step": 6620
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3894524574279785,
+      "learning_rate": 5.915557804859013e-07,
+      "loss": 0.5739,
+      "step": 6621
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4773305654525757,
+      "learning_rate": 5.905770968910379e-07,
+      "loss": 0.6667,
+      "step": 6622
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2038147449493408,
+      "learning_rate": 5.895991727296447e-07,
+      "loss": 0.5568,
+      "step": 6623
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.198391318321228,
+      "learning_rate": 5.886220081701494e-07,
+      "loss": 0.6375,
+      "step": 6624
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1705589294433594,
+      "learning_rate": 5.876456033808498e-07,
+      "loss": 0.5212,
+      "step": 6625
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.425979733467102,
+      "learning_rate": 5.86669958529909e-07,
+      "loss": 0.6197,
+      "step": 6626
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5084619522094727,
+      "learning_rate": 5.85695073785364e-07,
+      "loss": 0.6776,
+      "step": 6627
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.321138858795166,
+      "learning_rate": 5.847209493151185e-07,
+      "loss": 0.6305,
+      "step": 6628
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3939682245254517,
+      "learning_rate": 5.837475852869462e-07,
+      "loss": 0.6287,
+      "step": 6629
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 6.342435836791992,
+      "learning_rate": 5.82774981868488e-07,
+      "loss": 0.5973,
+      "step": 6630
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4819163084030151,
+      "learning_rate": 5.818031392272555e-07,
+      "loss": 0.6643,
+      "step": 6631
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5554163455963135,
+      "learning_rate": 5.808320575306292e-07,
+      "loss": 0.6145,
+      "step": 6632
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2572414875030518,
+      "learning_rate": 5.798617369458581e-07,
+      "loss": 0.6317,
+      "step": 6633
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3168387413024902,
+      "learning_rate": 5.788921776400597e-07,
+      "loss": 0.6413,
+      "step": 6634
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.401464819908142,
+      "learning_rate": 5.77923379780222e-07,
+      "loss": 0.6708,
+      "step": 6635
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3466784954071045,
+      "learning_rate": 5.769553435332009e-07,
+      "loss": 0.5765,
+      "step": 6636
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2899616956710815,
+      "learning_rate": 5.759880690657188e-07,
+      "loss": 0.5885,
+      "step": 6637
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3985425233840942,
+      "learning_rate": 5.750215565443707e-07,
+      "loss": 0.5156,
+      "step": 6638
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.468558669090271,
+      "learning_rate": 5.740558061356183e-07,
+      "loss": 0.6191,
+      "step": 6639
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5268968343734741,
+      "learning_rate": 5.730908180057937e-07,
+      "loss": 0.5815,
+      "step": 6640
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.1540591716766357,
+      "learning_rate": 5.721265923210944e-07,
+      "loss": 0.578,
+      "step": 6641
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5741490125656128,
+      "learning_rate": 5.711631292475894e-07,
+      "loss": 0.5626,
+      "step": 6642
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.359459638595581,
+      "learning_rate": 5.702004289512175e-07,
+      "loss": 0.6464,
+      "step": 6643
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2629036903381348,
+      "learning_rate": 5.692384915977811e-07,
+      "loss": 0.6634,
+      "step": 6644
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1385648250579834,
+      "learning_rate": 5.682773173529565e-07,
+      "loss": 0.6298,
+      "step": 6645
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4421610832214355,
+      "learning_rate": 5.673169063822853e-07,
+      "loss": 0.5755,
+      "step": 6646
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2397493124008179,
+      "learning_rate": 5.663572588511806e-07,
+      "loss": 0.5909,
+      "step": 6647
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.226677656173706,
+      "learning_rate": 5.653983749249198e-07,
+      "loss": 0.6041,
+      "step": 6648
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2758342027664185,
+      "learning_rate": 5.644402547686518e-07,
+      "loss": 0.6877,
+      "step": 6649
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6766910552978516,
+      "learning_rate": 5.63482898547395e-07,
+      "loss": 0.6041,
+      "step": 6650
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5834779739379883,
+      "learning_rate": 5.625263064260328e-07,
+      "loss": 0.5752,
+      "step": 6651
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.2436914443969727,
+      "learning_rate": 5.615704785693193e-07,
+      "loss": 0.6544,
+      "step": 6652
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5408368110656738,
+      "learning_rate": 5.606154151418763e-07,
+      "loss": 0.6088,
+      "step": 6653
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2919232845306396,
+      "learning_rate": 5.596611163081949e-07,
+      "loss": 0.5976,
+      "step": 6654
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.246882438659668,
+      "learning_rate": 5.587075822326326e-07,
+      "loss": 0.5889,
+      "step": 6655
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1017459630966187,
+      "learning_rate": 5.577548130794164e-07,
+      "loss": 0.5616,
+      "step": 6656
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.254428505897522,
+      "learning_rate": 5.568028090126415e-07,
+      "loss": 0.5466,
+      "step": 6657
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4765057563781738,
+      "learning_rate": 5.558515701962725e-07,
+      "loss": 0.609,
+      "step": 6658
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.194210171699524,
+      "learning_rate": 5.549010967941387e-07,
+      "loss": 0.5604,
+      "step": 6659
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3843317031860352,
+      "learning_rate": 5.539513889699411e-07,
+      "loss": 0.5851,
+      "step": 6660
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2867571115493774,
+      "learning_rate": 5.530024468872474e-07,
+      "loss": 0.6045,
+      "step": 6661
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4414231777191162,
+      "learning_rate": 5.520542707094945e-07,
+      "loss": 0.5784,
+      "step": 6662
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4258451461791992,
+      "learning_rate": 5.511068605999848e-07,
+      "loss": 0.6573,
+      "step": 6663
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4315848350524902,
+      "learning_rate": 5.501602167218912e-07,
+      "loss": 0.5564,
+      "step": 6664
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1932504177093506,
+      "learning_rate": 5.492143392382537e-07,
+      "loss": 0.5155,
+      "step": 6665
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.22544527053833,
+      "learning_rate": 5.482692283119817e-07,
+      "loss": 0.5858,
+      "step": 6666
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4700120687484741,
+      "learning_rate": 5.473248841058487e-07,
+      "loss": 0.6701,
+      "step": 6667
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4346530437469482,
+      "learning_rate": 5.463813067825008e-07,
+      "loss": 0.647,
+      "step": 6668
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.5285699367523193,
+      "learning_rate": 5.454384965044512e-07,
+      "loss": 0.5553,
+      "step": 6669
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.045205593109131,
+      "learning_rate": 5.444964534340768e-07,
+      "loss": 0.5509,
+      "step": 6670
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4717965126037598,
+      "learning_rate": 5.435551777336273e-07,
+      "loss": 0.6316,
+      "step": 6671
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3239675760269165,
+      "learning_rate": 5.426146695652173e-07,
+      "loss": 0.5174,
+      "step": 6672
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4008818864822388,
+      "learning_rate": 5.416749290908324e-07,
+      "loss": 0.5849,
+      "step": 6673
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3032972812652588,
+      "learning_rate": 5.407359564723202e-07,
+      "loss": 0.6331,
+      "step": 6674
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2396972179412842,
+      "learning_rate": 5.397977518714026e-07,
+      "loss": 0.5312,
+      "step": 6675
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1935837268829346,
+      "learning_rate": 5.388603154496647e-07,
+      "loss": 0.5171,
+      "step": 6676
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5318385362625122,
+      "learning_rate": 5.379236473685623e-07,
+      "loss": 0.567,
+      "step": 6677
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2079846858978271,
+      "learning_rate": 5.369877477894154e-07,
+      "loss": 0.7553,
+      "step": 6678
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4537231922149658,
+      "learning_rate": 5.360526168734154e-07,
+      "loss": 0.556,
+      "step": 6679
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2296948432922363,
+      "learning_rate": 5.351182547816186e-07,
+      "loss": 0.5602,
+      "step": 6680
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3176974058151245,
+      "learning_rate": 5.341846616749513e-07,
+      "loss": 0.6225,
+      "step": 6681
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0236971378326416,
+      "learning_rate": 5.332518377142043e-07,
+      "loss": 0.6221,
+      "step": 6682
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3097838163375854,
+      "learning_rate": 5.32319783060038e-07,
+      "loss": 0.4918,
+      "step": 6683
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4100576639175415,
+      "learning_rate": 5.3138849787298e-07,
+      "loss": 0.6335,
+      "step": 6684
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.343997836112976,
+      "learning_rate": 5.304579823134254e-07,
+      "loss": 0.5864,
+      "step": 6685
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2703300714492798,
+      "learning_rate": 5.295282365416365e-07,
+      "loss": 0.5433,
+      "step": 6686
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.460349440574646,
+      "learning_rate": 5.285992607177432e-07,
+      "loss": 0.6216,
+      "step": 6687
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8459396362304688,
+      "learning_rate": 5.276710550017433e-07,
+      "loss": 0.6032,
+      "step": 6688
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2343941926956177,
+      "learning_rate": 5.267436195535003e-07,
+      "loss": 0.7502,
+      "step": 6689
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4660110473632812,
+      "learning_rate": 5.258169545327462e-07,
+      "loss": 0.561,
+      "step": 6690
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.122137188911438,
+      "learning_rate": 5.248910600990814e-07,
+      "loss": 0.5409,
+      "step": 6691
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3979350328445435,
+      "learning_rate": 5.239659364119703e-07,
+      "loss": 0.6101,
+      "step": 6692
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6155316829681396,
+      "learning_rate": 5.230415836307485e-07,
+      "loss": 0.6467,
+      "step": 6693
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6388894319534302,
+      "learning_rate": 5.22118001914616e-07,
+      "loss": 0.6412,
+      "step": 6694
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4391796588897705,
+      "learning_rate": 5.21195191422642e-07,
+      "loss": 0.5887,
+      "step": 6695
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4248244762420654,
+      "learning_rate": 5.202731523137605e-07,
+      "loss": 0.6395,
+      "step": 6696
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8664724826812744,
+      "learning_rate": 5.193518847467749e-07,
+      "loss": 0.5536,
+      "step": 6697
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5709271430969238,
+      "learning_rate": 5.184313888803544e-07,
+      "loss": 0.6439,
+      "step": 6698
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2865567207336426,
+      "learning_rate": 5.175116648730366e-07,
+      "loss": 0.6003,
+      "step": 6699
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3031384944915771,
+      "learning_rate": 5.165927128832238e-07,
+      "loss": 0.6648,
+      "step": 6700
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2907569408416748,
+      "learning_rate": 5.156745330691871e-07,
+      "loss": 0.7093,
+      "step": 6701
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.391937017440796,
+      "learning_rate": 5.14757125589066e-07,
+      "loss": 0.6502,
+      "step": 6702
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4120694398880005,
+      "learning_rate": 5.138404906008631e-07,
+      "loss": 0.5444,
+      "step": 6703
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.370375156402588,
+      "learning_rate": 5.129246282624511e-07,
+      "loss": 0.5462,
+      "step": 6704
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3214823007583618,
+      "learning_rate": 5.120095387315688e-07,
+      "loss": 0.6326,
+      "step": 6705
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4014390707015991,
+      "learning_rate": 5.110952221658228e-07,
+      "loss": 0.5686,
+      "step": 6706
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2138499021530151,
+      "learning_rate": 5.101816787226832e-07,
+      "loss": 0.5041,
+      "step": 6707
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6902697086334229,
+      "learning_rate": 5.092689085594904e-07,
+      "loss": 0.5416,
+      "step": 6708
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6033650636672974,
+      "learning_rate": 5.083569118334508e-07,
+      "loss": 0.6395,
+      "step": 6709
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4746758937835693,
+      "learning_rate": 5.074456887016382e-07,
+      "loss": 0.5835,
+      "step": 6710
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.268338918685913,
+      "learning_rate": 5.065352393209899e-07,
+      "loss": 0.5185,
+      "step": 6711
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.275320291519165,
+      "learning_rate": 5.056255638483137e-07,
+      "loss": 0.6441,
+      "step": 6712
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6120179891586304,
+      "learning_rate": 5.047166624402822e-07,
+      "loss": 0.582,
+      "step": 6713
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3864774703979492,
+      "learning_rate": 5.038085352534367e-07,
+      "loss": 0.542,
+      "step": 6714
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0285900831222534,
+      "learning_rate": 5.029011824441821e-07,
+      "loss": 0.5212,
+      "step": 6715
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4503127336502075,
+      "learning_rate": 5.019946041687911e-07,
+      "loss": 0.5739,
+      "step": 6716
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4805421829223633,
+      "learning_rate": 5.010888005834047e-07,
+      "loss": 0.558,
+      "step": 6717
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2787694931030273,
+      "learning_rate": 5.001837718440283e-07,
+      "loss": 0.5401,
+      "step": 6718
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7053654193878174,
+      "learning_rate": 4.992795181065347e-07,
+      "loss": 0.711,
+      "step": 6719
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1280864477157593,
+      "learning_rate": 4.98376039526664e-07,
+      "loss": 0.5233,
+      "step": 6720
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7056137323379517,
+      "learning_rate": 4.974733362600225e-07,
+      "loss": 0.6195,
+      "step": 6721
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1686879396438599,
+      "learning_rate": 4.965714084620804e-07,
+      "loss": 0.6126,
+      "step": 6722
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1632872819900513,
+      "learning_rate": 4.956702562881783e-07,
+      "loss": 0.5181,
+      "step": 6723
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4708881378173828,
+      "learning_rate": 4.947698798935196e-07,
+      "loss": 0.6074,
+      "step": 6724
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4640129804611206,
+      "learning_rate": 4.938702794331785e-07,
+      "loss": 0.5477,
+      "step": 6725
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3392497301101685,
+      "learning_rate": 4.929714550620901e-07,
+      "loss": 0.5741,
+      "step": 6726
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.190356969833374,
+      "learning_rate": 4.920734069350597e-07,
+      "loss": 0.5884,
+      "step": 6727
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3186179399490356,
+      "learning_rate": 4.911761352067579e-07,
+      "loss": 0.6421,
+      "step": 6728
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3501472473144531,
+      "learning_rate": 4.902796400317228e-07,
+      "loss": 0.6491,
+      "step": 6729
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5455275774002075,
+      "learning_rate": 4.893839215643553e-07,
+      "loss": 0.6143,
+      "step": 6730
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5439045429229736,
+      "learning_rate": 4.884889799589254e-07,
+      "loss": 0.5792,
+      "step": 6731
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2626060247421265,
+      "learning_rate": 4.875948153695687e-07,
+      "loss": 0.5075,
+      "step": 6732
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0654497146606445,
+      "learning_rate": 4.86701427950288e-07,
+      "loss": 0.5002,
+      "step": 6733
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3448207378387451,
+      "learning_rate": 4.858088178549492e-07,
+      "loss": 0.5399,
+      "step": 6734
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4263006448745728,
+      "learning_rate": 4.849169852372864e-07,
+      "loss": 0.6053,
+      "step": 6735
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.585645079612732,
+      "learning_rate": 4.84025930250902e-07,
+      "loss": 0.6796,
+      "step": 6736
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.15276038646698,
+      "learning_rate": 4.831356530492598e-07,
+      "loss": 0.583,
+      "step": 6737
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6426639556884766,
+      "learning_rate": 4.822461537856927e-07,
+      "loss": 0.546,
+      "step": 6738
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0492527484893799,
+      "learning_rate": 4.813574326133985e-07,
+      "loss": 0.6961,
+      "step": 6739
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.360402226448059,
+      "learning_rate": 4.804694896854434e-07,
+      "loss": 0.6052,
+      "step": 6740
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.158557415008545,
+      "learning_rate": 4.795823251547544e-07,
+      "loss": 0.7179,
+      "step": 6741
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.3589587211608887,
+      "learning_rate": 4.786959391741286e-07,
+      "loss": 0.5791,
+      "step": 6742
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7391221523284912,
+      "learning_rate": 4.778103318962296e-07,
+      "loss": 0.606,
+      "step": 6743
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5412747859954834,
+      "learning_rate": 4.769255034735831e-07,
+      "loss": 0.6217,
+      "step": 6744
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2919334173202515,
+      "learning_rate": 4.760414540585839e-07,
+      "loss": 0.6057,
+      "step": 6745
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3662687540054321,
+      "learning_rate": 4.7515818380349074e-07,
+      "loss": 0.7293,
+      "step": 6746
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.417550802230835,
+      "learning_rate": 4.7427569286043086e-07,
+      "loss": 0.6307,
+      "step": 6747
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5828347206115723,
+      "learning_rate": 4.733939813813931e-07,
+      "loss": 0.5875,
+      "step": 6748
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5326488018035889,
+      "learning_rate": 4.725130495182356e-07,
+      "loss": 0.5814,
+      "step": 6749
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4294366836547852,
+      "learning_rate": 4.716328974226808e-07,
+      "loss": 0.5564,
+      "step": 6750
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.306718111038208,
+      "learning_rate": 4.707535252463175e-07,
+      "loss": 0.5206,
+      "step": 6751
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2117685079574585,
+      "learning_rate": 4.6987493314059716e-07,
+      "loss": 0.516,
+      "step": 6752
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3194915056228638,
+      "learning_rate": 4.689971212568428e-07,
+      "loss": 0.6414,
+      "step": 6753
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.456396222114563,
+      "learning_rate": 4.6812008974623845e-07,
+      "loss": 0.6447,
+      "step": 6754
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3069777488708496,
+      "learning_rate": 4.672438387598344e-07,
+      "loss": 0.6064,
+      "step": 6755
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.897487759590149,
+      "learning_rate": 4.6636836844854706e-07,
+      "loss": 0.5244,
+      "step": 6756
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.408294916152954,
+      "learning_rate": 4.6549367896315923e-07,
+      "loss": 0.5875,
+      "step": 6757
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1629550457000732,
+      "learning_rate": 4.646197704543187e-07,
+      "loss": 0.7416,
+      "step": 6758
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1940720081329346,
+      "learning_rate": 4.6374664307253625e-07,
+      "loss": 0.5178,
+      "step": 6759
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6622729301452637,
+      "learning_rate": 4.628742969681921e-07,
+      "loss": 0.6646,
+      "step": 6760
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.3263518810272217,
+      "learning_rate": 4.6200273229152994e-07,
+      "loss": 0.5751,
+      "step": 6761
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7037662267684937,
+      "learning_rate": 4.611319491926597e-07,
+      "loss": 0.6177,
+      "step": 6762
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4251806735992432,
+      "learning_rate": 4.602619478215542e-07,
+      "loss": 0.7138,
+      "step": 6763
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6694972515106201,
+      "learning_rate": 4.593927283280547e-07,
+      "loss": 0.6646,
+      "step": 6764
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3548812866210938,
+      "learning_rate": 4.5852429086186646e-07,
+      "loss": 0.5704,
+      "step": 6765
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6709963083267212,
+      "learning_rate": 4.576566355725609e-07,
+      "loss": 0.7045,
+      "step": 6766
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4117552042007446,
+      "learning_rate": 4.5678976260957243e-07,
+      "loss": 0.5623,
+      "step": 6767
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4504401683807373,
+      "learning_rate": 4.5592367212220324e-07,
+      "loss": 0.6227,
+      "step": 6768
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.266184687614441,
+      "learning_rate": 4.5505836425961956e-07,
+      "loss": 0.5448,
+      "step": 6769
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2798473834991455,
+      "learning_rate": 4.541938391708539e-07,
+      "loss": 0.5534,
+      "step": 6770
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5648812055587769,
+      "learning_rate": 4.533300970048016e-07,
+      "loss": 0.6557,
+      "step": 6771
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3745602369308472,
+      "learning_rate": 4.5246713791022633e-07,
+      "loss": 0.6003,
+      "step": 6772
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2130157947540283,
+      "learning_rate": 4.516049620357549e-07,
+      "loss": 0.7218,
+      "step": 6773
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1825891733169556,
+      "learning_rate": 4.507435695298784e-07,
+      "loss": 0.5922,
+      "step": 6774
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.352432131767273,
+      "learning_rate": 4.4988296054095494e-07,
+      "loss": 0.5628,
+      "step": 6775
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6166142225265503,
+      "learning_rate": 4.4902313521720696e-07,
+      "loss": 0.634,
+      "step": 6776
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.435806155204773,
+      "learning_rate": 4.4816409370672277e-07,
+      "loss": 0.5729,
+      "step": 6777
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5084896087646484,
+      "learning_rate": 4.473058361574534e-07,
+      "loss": 0.6539,
+      "step": 6778
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4006537199020386,
+      "learning_rate": 4.4644836271721617e-07,
+      "loss": 0.6189,
+      "step": 6779
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4999135732650757,
+      "learning_rate": 4.455916735336946e-07,
+      "loss": 0.6958,
+      "step": 6780
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1975650787353516,
+      "learning_rate": 4.4473576875443626e-07,
+      "loss": 0.6147,
+      "step": 6781
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.477052927017212,
+      "learning_rate": 4.438806485268515e-07,
+      "loss": 0.5882,
+      "step": 6782
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4066824913024902,
+      "learning_rate": 4.430263129982182e-07,
+      "loss": 0.5717,
+      "step": 6783
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.330893635749817,
+      "learning_rate": 4.421727623156796e-07,
+      "loss": 0.6369,
+      "step": 6784
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5608009099960327,
+      "learning_rate": 4.41319996626241e-07,
+      "loss": 0.5731,
+      "step": 6785
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.482822299003601,
+      "learning_rate": 4.404680160767727e-07,
+      "loss": 0.633,
+      "step": 6786
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.300209879875183,
+      "learning_rate": 4.3961682081401393e-07,
+      "loss": 0.615,
+      "step": 6787
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7004002332687378,
+      "learning_rate": 4.3876641098456574e-07,
+      "loss": 0.5625,
+      "step": 6788
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.4036426544189453,
+      "learning_rate": 4.379167867348916e-07,
+      "loss": 0.6332,
+      "step": 6789
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.266597032546997,
+      "learning_rate": 4.370679482113227e-07,
+      "loss": 0.6741,
+      "step": 6790
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4129186868667603,
+      "learning_rate": 4.362198955600561e-07,
+      "loss": 0.6334,
+      "step": 6791
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2967772483825684,
+      "learning_rate": 4.3537262892714926e-07,
+      "loss": 0.7026,
+      "step": 6792
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3166394233703613,
+      "learning_rate": 4.345261484585273e-07,
+      "loss": 0.5694,
+      "step": 6793
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5502707958221436,
+      "learning_rate": 4.3368045429997953e-07,
+      "loss": 0.5651,
+      "step": 6794
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5218241214752197,
+      "learning_rate": 4.3283554659716066e-07,
+      "loss": 0.6583,
+      "step": 6795
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3967206478118896,
+      "learning_rate": 4.3199142549558704e-07,
+      "loss": 0.5456,
+      "step": 6796
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2347934246063232,
+      "learning_rate": 4.3114809114064183e-07,
+      "loss": 0.5805,
+      "step": 6797
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6478691101074219,
+      "learning_rate": 4.303055436775727e-07,
+      "loss": 0.5988,
+      "step": 6798
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.3431971073150635,
+      "learning_rate": 4.2946378325149196e-07,
+      "loss": 0.5429,
+      "step": 6799
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5060168504714966,
+      "learning_rate": 4.286228100073742e-07,
+      "loss": 0.5957,
+      "step": 6800
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3222039937973022,
+      "learning_rate": 4.2778262409006064e-07,
+      "loss": 0.5746,
+      "step": 6801
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2762765884399414,
+      "learning_rate": 4.2694322564425626e-07,
+      "loss": 0.5936,
+      "step": 6802
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3855468034744263,
+      "learning_rate": 4.261046148145315e-07,
+      "loss": 0.6254,
+      "step": 6803
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.342413306236267,
+      "learning_rate": 4.2526679174531737e-07,
+      "loss": 0.6038,
+      "step": 6804
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.30988609790802,
+      "learning_rate": 4.244297565809136e-07,
+      "loss": 0.5995,
+      "step": 6805
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4815443754196167,
+      "learning_rate": 4.235935094654836e-07,
+      "loss": 0.6256,
+      "step": 6806
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2580639123916626,
+      "learning_rate": 4.2275805054305175e-07,
+      "loss": 0.5472,
+      "step": 6807
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2893187999725342,
+      "learning_rate": 4.219233799575101e-07,
+      "loss": 0.6253,
+      "step": 6808
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2105703353881836,
+      "learning_rate": 4.210894978526131e-07,
+      "loss": 0.5719,
+      "step": 6809
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4485310316085815,
+      "learning_rate": 4.202564043719809e-07,
+      "loss": 0.6083,
+      "step": 6810
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.28269624710083,
+      "learning_rate": 4.194240996590959e-07,
+      "loss": 0.5364,
+      "step": 6811
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2909283638000488,
+      "learning_rate": 4.185925838573057e-07,
+      "loss": 0.6054,
+      "step": 6812
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3592783212661743,
+      "learning_rate": 4.1776185710982244e-07,
+      "loss": 0.6188,
+      "step": 6813
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3486402034759521,
+      "learning_rate": 4.169319195597227e-07,
+      "loss": 0.6326,
+      "step": 6814
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3511604070663452,
+      "learning_rate": 4.161027713499444e-07,
+      "loss": 0.6333,
+      "step": 6815
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4477179050445557,
+      "learning_rate": 4.152744126232927e-07,
+      "loss": 0.5499,
+      "step": 6816
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2189664840698242,
+      "learning_rate": 4.1444684352243525e-07,
+      "loss": 0.5236,
+      "step": 6817
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2309170961380005,
+      "learning_rate": 4.136200641899052e-07,
+      "loss": 0.5513,
+      "step": 6818
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.251579761505127,
+      "learning_rate": 4.12794074768097e-07,
+      "loss": 0.5862,
+      "step": 6819
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.47301185131073,
+      "learning_rate": 4.119688753992707e-07,
+      "loss": 0.5965,
+      "step": 6820
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6270514726638794,
+      "learning_rate": 4.111444662255498e-07,
+      "loss": 0.5907,
+      "step": 6821
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4491541385650635,
+      "learning_rate": 4.103208473889231e-07,
+      "loss": 0.519,
+      "step": 6822
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2944505214691162,
+      "learning_rate": 4.09498019031242e-07,
+      "loss": 0.6209,
+      "step": 6823
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4557186365127563,
+      "learning_rate": 4.0867598129422146e-07,
+      "loss": 0.6311,
+      "step": 6824
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2103979587554932,
+      "learning_rate": 4.0785473431944165e-07,
+      "loss": 0.5814,
+      "step": 6825
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.428792953491211,
+      "learning_rate": 4.070342782483444e-07,
+      "loss": 0.5717,
+      "step": 6826
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3763489723205566,
+      "learning_rate": 4.062146132222372e-07,
+      "loss": 0.551,
+      "step": 6827
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3854894638061523,
+      "learning_rate": 4.053957393822905e-07,
+      "loss": 0.5466,
+      "step": 6828
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4803504943847656,
+      "learning_rate": 4.045776568695398e-07,
+      "loss": 0.6583,
+      "step": 6829
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1337047815322876,
+      "learning_rate": 4.037603658248812e-07,
+      "loss": 0.6015,
+      "step": 6830
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2642403841018677,
+      "learning_rate": 4.029438663890778e-07,
+      "loss": 0.6461,
+      "step": 6831
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5304993391036987,
+      "learning_rate": 4.021281587027548e-07,
+      "loss": 0.509,
+      "step": 6832
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3939540386199951,
+      "learning_rate": 4.0131324290640206e-07,
+      "loss": 0.5702,
+      "step": 6833
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4129104614257812,
+      "learning_rate": 4.0049911914037067e-07,
+      "loss": 0.6004,
+      "step": 6834
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4573888778686523,
+      "learning_rate": 3.9968578754487784e-07,
+      "loss": 0.6195,
+      "step": 6835
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4409502744674683,
+      "learning_rate": 3.988732482600033e-07,
+      "loss": 0.556,
+      "step": 6836
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4219567775726318,
+      "learning_rate": 3.9806150142569e-07,
+      "loss": 0.6177,
+      "step": 6837
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3022994995117188,
+      "learning_rate": 3.972505471817445e-07,
+      "loss": 0.5414,
+      "step": 6838
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4800045490264893,
+      "learning_rate": 3.9644038566783946e-07,
+      "loss": 0.6664,
+      "step": 6839
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3413270711898804,
+      "learning_rate": 3.9563101702350616e-07,
+      "loss": 0.5806,
+      "step": 6840
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4288663864135742,
+      "learning_rate": 3.9482244138814295e-07,
+      "loss": 0.5744,
+      "step": 6841
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5452122688293457,
+      "learning_rate": 3.940146589010108e-07,
+      "loss": 0.5657,
+      "step": 6842
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9402995109558105,
+      "learning_rate": 3.9320766970123383e-07,
+      "loss": 0.5839,
+      "step": 6843
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1517301797866821,
+      "learning_rate": 3.924014739277987e-07,
+      "loss": 0.5033,
+      "step": 6844
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.396001935005188,
+      "learning_rate": 3.915960717195566e-07,
+      "loss": 0.6478,
+      "step": 6845
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4981609582901,
+      "learning_rate": 3.907914632152215e-07,
+      "loss": 0.602,
+      "step": 6846
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3975633382797241,
+      "learning_rate": 3.8998764855337266e-07,
+      "loss": 0.5397,
+      "step": 6847
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1856523752212524,
+      "learning_rate": 3.8918462787244817e-07,
+      "loss": 0.5272,
+      "step": 6848
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2696727514266968,
+      "learning_rate": 3.8838240131075343e-07,
+      "loss": 0.5655,
+      "step": 6849
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.603298306465149,
+      "learning_rate": 3.8758096900645524e-07,
+      "loss": 0.522,
+      "step": 6850
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3075599670410156,
+      "learning_rate": 3.867803310975854e-07,
+      "loss": 0.5532,
+      "step": 6851
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4688489437103271,
+      "learning_rate": 3.859804877220352e-07,
+      "loss": 0.5474,
+      "step": 6852
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.316911220550537,
+      "learning_rate": 3.851814390175623e-07,
+      "loss": 0.6715,
+      "step": 6853
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.143972635269165,
+      "learning_rate": 3.843831851217872e-07,
+      "loss": 0.7013,
+      "step": 6854
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1683162450790405,
+      "learning_rate": 3.835857261721926e-07,
+      "loss": 0.5927,
+      "step": 6855
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6495614051818848,
+      "learning_rate": 3.827890623061242e-07,
+      "loss": 0.6061,
+      "step": 6856
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4121828079223633,
+      "learning_rate": 3.8199319366079177e-07,
+      "loss": 0.5579,
+      "step": 6857
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4523359537124634,
+      "learning_rate": 3.811981203732684e-07,
+      "loss": 0.5648,
+      "step": 6858
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.28249990940094,
+      "learning_rate": 3.8040384258048677e-07,
+      "loss": 0.6971,
+      "step": 6859
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5575233697891235,
+      "learning_rate": 3.7961036041924635e-07,
+      "loss": 0.5897,
+      "step": 6860
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.302139163017273,
+      "learning_rate": 3.788176740262089e-07,
+      "loss": 0.6636,
+      "step": 6861
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1842310428619385,
+      "learning_rate": 3.7802578353789864e-07,
+      "loss": 0.5368,
+      "step": 6862
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4447256326675415,
+      "learning_rate": 3.7723468909070136e-07,
+      "loss": 0.5266,
+      "step": 6863
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4478634595870972,
+      "learning_rate": 3.764443908208676e-07,
+      "loss": 0.5914,
+      "step": 6864
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.580125093460083,
+      "learning_rate": 3.7565488886451004e-07,
+      "loss": 0.5411,
+      "step": 6865
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5011149644851685,
+      "learning_rate": 3.748661833576056e-07,
+      "loss": 0.6147,
+      "step": 6866
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.561859369277954,
+      "learning_rate": 3.740782744359911e-07,
+      "loss": 0.6672,
+      "step": 6867
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3884916305541992,
+      "learning_rate": 3.7329116223536797e-07,
+      "loss": 0.6525,
+      "step": 6868
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4261103868484497,
+      "learning_rate": 3.7250484689130115e-07,
+      "loss": 0.5723,
+      "step": 6869
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4831106662750244,
+      "learning_rate": 3.717193285392179e-07,
+      "loss": 0.5859,
+      "step": 6870
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6929248571395874,
+      "learning_rate": 3.70934607314406e-07,
+      "loss": 0.6201,
+      "step": 6871
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2150589227676392,
+      "learning_rate": 3.70150683352018e-07,
+      "loss": 0.608,
+      "step": 6872
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2132083177566528,
+      "learning_rate": 3.693675567870714e-07,
+      "loss": 0.5384,
+      "step": 6873
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.33145010471344,
+      "learning_rate": 3.685852277544405e-07,
+      "loss": 0.6047,
+      "step": 6874
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2606712579727173,
+      "learning_rate": 3.678036963888676e-07,
+      "loss": 0.6038,
+      "step": 6875
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2835811376571655,
+      "learning_rate": 3.670229628249555e-07,
+      "loss": 0.5235,
+      "step": 6876
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5742946863174438,
+      "learning_rate": 3.662430271971695e-07,
+      "loss": 0.654,
+      "step": 6877
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2673202753067017,
+      "learning_rate": 3.6546388963983716e-07,
+      "loss": 0.529,
+      "step": 6878
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.531229019165039,
+      "learning_rate": 3.646855502871488e-07,
+      "loss": 0.5374,
+      "step": 6879
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3460601568222046,
+      "learning_rate": 3.639080092731584e-07,
+      "loss": 0.5097,
+      "step": 6880
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3324980735778809,
+      "learning_rate": 3.6313126673178213e-07,
+      "loss": 0.547,
+      "step": 6881
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4404997825622559,
+      "learning_rate": 3.623553227967963e-07,
+      "loss": 0.5873,
+      "step": 6882
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.44928777217865,
+      "learning_rate": 3.6158017760184237e-07,
+      "loss": 0.5764,
+      "step": 6883
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1881428956985474,
+      "learning_rate": 3.608058312804247e-07,
+      "loss": 0.6388,
+      "step": 6884
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5132038593292236,
+      "learning_rate": 3.600322839659065e-07,
+      "loss": 0.5741,
+      "step": 6885
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.413456439971924,
+      "learning_rate": 3.592595357915163e-07,
+      "loss": 0.6382,
+      "step": 6886
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2886861562728882,
+      "learning_rate": 3.584875868903448e-07,
+      "loss": 0.5972,
+      "step": 6887
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2081990242004395,
+      "learning_rate": 3.577164373953446e-07,
+      "loss": 0.5572,
+      "step": 6888
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4857827425003052,
+      "learning_rate": 3.569460874393288e-07,
+      "loss": 0.5606,
+      "step": 6889
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3529541492462158,
+      "learning_rate": 3.561765371549769e-07,
+      "loss": 0.5588,
+      "step": 6890
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1445674896240234,
+      "learning_rate": 3.554077866748279e-07,
+      "loss": 0.5352,
+      "step": 6891
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4179437160491943,
+      "learning_rate": 3.5463983613128136e-07,
+      "loss": 0.5433,
+      "step": 6892
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1696901321411133,
+      "learning_rate": 3.5387268565660324e-07,
+      "loss": 0.5855,
+      "step": 6893
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.259205937385559,
+      "learning_rate": 3.5310633538291894e-07,
+      "loss": 0.5802,
+      "step": 6894
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2908961772918701,
+      "learning_rate": 3.523407854422173e-07,
+      "loss": 0.6712,
+      "step": 6895
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3714510202407837,
+      "learning_rate": 3.5157603596634727e-07,
+      "loss": 0.6665,
+      "step": 6896
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2138036489486694,
+      "learning_rate": 3.508120870870224e-07,
+      "loss": 0.6244,
+      "step": 6897
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4784094095230103,
+      "learning_rate": 3.5004893893581746e-07,
+      "loss": 0.6091,
+      "step": 6898
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4526443481445312,
+      "learning_rate": 3.4928659164416956e-07,
+      "loss": 0.6336,
+      "step": 6899
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4262062311172485,
+      "learning_rate": 3.485250453433764e-07,
+      "loss": 0.5957,
+      "step": 6900
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7060856819152832,
+      "learning_rate": 3.4776430016459917e-07,
+      "loss": 0.5559,
+      "step": 6901
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9785739183425903,
+      "learning_rate": 3.4700435623886143e-07,
+      "loss": 0.6409,
+      "step": 6902
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.693240761756897,
+      "learning_rate": 3.462452136970479e-07,
+      "loss": 0.6472,
+      "step": 6903
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6329549551010132,
+      "learning_rate": 3.4548687266990453e-07,
+      "loss": 0.5792,
+      "step": 6904
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6811611652374268,
+      "learning_rate": 3.4472933328804134e-07,
+      "loss": 0.5417,
+      "step": 6905
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2126712799072266,
+      "learning_rate": 3.439725956819284e-07,
+      "loss": 0.5035,
+      "step": 6906
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4418469667434692,
+      "learning_rate": 3.4321665998189814e-07,
+      "loss": 0.5954,
+      "step": 6907
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.128082275390625,
+      "learning_rate": 3.424615263181458e-07,
+      "loss": 0.5779,
+      "step": 6908
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4310845136642456,
+      "learning_rate": 3.417071948207273e-07,
+      "loss": 0.505,
+      "step": 6909
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5503251552581787,
+      "learning_rate": 3.409536656195628e-07,
+      "loss": 0.6519,
+      "step": 6910
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8439539670944214,
+      "learning_rate": 3.4020093884442885e-07,
+      "loss": 0.5995,
+      "step": 6911
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.237624168395996,
+      "learning_rate": 3.3944901462497014e-07,
+      "loss": 0.4917,
+      "step": 6912
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3327163457870483,
+      "learning_rate": 3.3869789309068866e-07,
+      "loss": 0.5526,
+      "step": 6913
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.702526330947876,
+      "learning_rate": 3.3794757437095206e-07,
+      "loss": 0.6228,
+      "step": 6914
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.312321424484253,
+      "learning_rate": 3.3719805859498466e-07,
+      "loss": 0.6137,
+      "step": 6915
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5142812728881836,
+      "learning_rate": 3.364493458918772e-07,
+      "loss": 0.5638,
+      "step": 6916
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.370383858680725,
+      "learning_rate": 3.3570143639057916e-07,
+      "loss": 0.5413,
+      "step": 6917
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4188882112503052,
+      "learning_rate": 3.349543302199043e-07,
+      "loss": 0.5455,
+      "step": 6918
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3100755214691162,
+      "learning_rate": 3.342080275085252e-07,
+      "loss": 0.6094,
+      "step": 6919
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6879796981811523,
+      "learning_rate": 3.3346252838497683e-07,
+      "loss": 0.6627,
+      "step": 6920
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2003105878829956,
+      "learning_rate": 3.327178329776576e-07,
+      "loss": 0.6109,
+      "step": 6921
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4838978052139282,
+      "learning_rate": 3.31973941414826e-07,
+      "loss": 0.5857,
+      "step": 6922
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5608856678009033,
+      "learning_rate": 3.312308538246006e-07,
+      "loss": 0.6024,
+      "step": 6923
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.857037305831909,
+      "learning_rate": 3.3048857033496473e-07,
+      "loss": 0.5997,
+      "step": 6924
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4345684051513672,
+      "learning_rate": 3.2974709107376215e-07,
+      "loss": 0.6384,
+      "step": 6925
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6866978406906128,
+      "learning_rate": 3.290064161686962e-07,
+      "loss": 0.5539,
+      "step": 6926
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3979156017303467,
+      "learning_rate": 3.282665457473333e-07,
+      "loss": 0.6002,
+      "step": 6927
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.108978033065796,
+      "learning_rate": 3.2752747993710144e-07,
+      "loss": 0.7341,
+      "step": 6928
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3866137266159058,
+      "learning_rate": 3.267892188652905e-07,
+      "loss": 0.6452,
+      "step": 6929
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1434195041656494,
+      "learning_rate": 3.2605176265904925e-07,
+      "loss": 0.6161,
+      "step": 6930
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0713568925857544,
+      "learning_rate": 3.253151114453901e-07,
+      "loss": 0.6877,
+      "step": 6931
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3612641096115112,
+      "learning_rate": 3.245792653511876e-07,
+      "loss": 0.5224,
+      "step": 6932
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.527736783027649,
+      "learning_rate": 3.238442245031742e-07,
+      "loss": 0.6014,
+      "step": 6933
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1294968128204346,
+      "learning_rate": 3.2310998902794653e-07,
+      "loss": 0.7599,
+      "step": 6934
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4158496856689453,
+      "learning_rate": 3.223765590519623e-07,
+      "loss": 0.5286,
+      "step": 6935
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.530761957168579,
+      "learning_rate": 3.216439347015399e-07,
+      "loss": 0.5673,
+      "step": 6936
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.185805320739746,
+      "learning_rate": 3.209121161028583e-07,
+      "loss": 0.5242,
+      "step": 6937
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3359084129333496,
+      "learning_rate": 3.201811033819585e-07,
+      "loss": 0.5777,
+      "step": 6938
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.525627613067627,
+      "learning_rate": 3.194508966647425e-07,
+      "loss": 0.5719,
+      "step": 6939
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.422498345375061,
+      "learning_rate": 3.1872149607697466e-07,
+      "loss": 0.5689,
+      "step": 6940
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4468578100204468,
+      "learning_rate": 3.179929017442773e-07,
+      "loss": 0.5871,
+      "step": 6941
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.309499979019165,
+      "learning_rate": 3.1726511379213784e-07,
+      "loss": 0.6147,
+      "step": 6942
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.29915189743042,
+      "learning_rate": 3.1653813234590327e-07,
+      "loss": 0.6435,
+      "step": 6943
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3363467454910278,
+      "learning_rate": 3.1581195753078e-07,
+      "loss": 0.5975,
+      "step": 6944
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5139838457107544,
+      "learning_rate": 3.150865894718369e-07,
+      "loss": 0.6698,
+      "step": 6945
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.194031834602356,
+      "learning_rate": 3.143620282940046e-07,
+      "loss": 0.5815,
+      "step": 6946
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3560861349105835,
+      "learning_rate": 3.136382741220745e-07,
+      "loss": 0.4936,
+      "step": 6947
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2448467016220093,
+      "learning_rate": 3.1291532708069727e-07,
+      "loss": 0.575,
+      "step": 6948
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4894511699676514,
+      "learning_rate": 3.121931872943862e-07,
+      "loss": 0.6252,
+      "step": 6949
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2867622375488281,
+      "learning_rate": 3.1147185488751506e-07,
+      "loss": 0.641,
+      "step": 6950
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.371701955795288,
+      "learning_rate": 3.107513299843201e-07,
+      "loss": 0.6279,
+      "step": 6951
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.264610767364502,
+      "learning_rate": 3.100316127088954e-07,
+      "loss": 0.6091,
+      "step": 6952
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2578670978546143,
+      "learning_rate": 3.093127031851978e-07,
+      "loss": 0.5467,
+      "step": 6953
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3644834756851196,
+      "learning_rate": 3.0859460153704557e-07,
+      "loss": 0.5687,
+      "step": 6954
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1062878370285034,
+      "learning_rate": 3.0787730788811807e-07,
+      "loss": 0.5197,
+      "step": 6955
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.430008888244629,
+      "learning_rate": 3.0716082236195213e-07,
+      "loss": 0.5242,
+      "step": 6956
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4670532941818237,
+      "learning_rate": 3.064451450819489e-07,
+      "loss": 0.5455,
+      "step": 6957
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1277669668197632,
+      "learning_rate": 3.057302761713693e-07,
+      "loss": 0.7229,
+      "step": 6958
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.744103193283081,
+      "learning_rate": 3.050162157533354e-07,
+      "loss": 0.5868,
+      "step": 6959
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5596626996994019,
+      "learning_rate": 3.0430296395082883e-07,
+      "loss": 0.643,
+      "step": 6960
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3399211168289185,
+      "learning_rate": 3.035905208866935e-07,
+      "loss": 0.6114,
+      "step": 6961
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.7940516471862793,
+      "learning_rate": 3.028788866836335e-07,
+      "loss": 0.6472,
+      "step": 6962
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4219404458999634,
+      "learning_rate": 3.02168061464212e-07,
+      "loss": 0.5366,
+      "step": 6963
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.567854881286621,
+      "learning_rate": 3.0145804535085476e-07,
+      "loss": 0.5771,
+      "step": 6964
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4092823266983032,
+      "learning_rate": 3.00748838465848e-07,
+      "loss": 0.5056,
+      "step": 6965
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.33809494972229,
+      "learning_rate": 3.0004044093133844e-07,
+      "loss": 0.5536,
+      "step": 6966
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3685873746871948,
+      "learning_rate": 2.9933285286933177e-07,
+      "loss": 0.6435,
+      "step": 6967
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6785255670547485,
+      "learning_rate": 2.9862607440169724e-07,
+      "loss": 0.5749,
+      "step": 6968
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.203216314315796,
+      "learning_rate": 2.979201056501618e-07,
+      "loss": 0.5402,
+      "step": 6969
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2036579847335815,
+      "learning_rate": 2.972149467363161e-07,
+      "loss": 0.5646,
+      "step": 6970
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.28584623336792,
+      "learning_rate": 2.9651059778160685e-07,
+      "loss": 0.6238,
+      "step": 6971
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2839908599853516,
+      "learning_rate": 2.958070589073453e-07,
+      "loss": 0.6049,
+      "step": 6972
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.457728624343872,
+      "learning_rate": 2.9510433023470174e-07,
+      "loss": 0.5392,
+      "step": 6973
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3485701084136963,
+      "learning_rate": 2.9440241188470717e-07,
+      "loss": 0.6339,
+      "step": 6974
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3882330656051636,
+      "learning_rate": 2.93701303978251e-07,
+      "loss": 0.561,
+      "step": 6975
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0906554460525513,
+      "learning_rate": 2.930010066360872e-07,
+      "loss": 0.5579,
+      "step": 6976
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.292925238609314,
+      "learning_rate": 2.9230151997882703e-07,
+      "loss": 0.5328,
+      "step": 6977
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.570992350578308,
+      "learning_rate": 2.9160284412694195e-07,
+      "loss": 0.5696,
+      "step": 6978
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.475880742073059,
+      "learning_rate": 2.909049792007651e-07,
+      "loss": 0.6416,
+      "step": 6979
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4461641311645508,
+      "learning_rate": 2.902079253204898e-07,
+      "loss": 0.577,
+      "step": 6980
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4062169790267944,
+      "learning_rate": 2.8951168260617004e-07,
+      "loss": 0.5309,
+      "step": 6981
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2815173864364624,
+      "learning_rate": 2.888162511777176e-07,
+      "loss": 0.6191,
+      "step": 6982
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3789962530136108,
+      "learning_rate": 2.881216311549079e-07,
+      "loss": 0.5828,
+      "step": 6983
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1321500539779663,
+      "learning_rate": 2.8742782265737514e-07,
+      "loss": 0.7302,
+      "step": 6984
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0742570161819458,
+      "learning_rate": 2.8673482580461264e-07,
+      "loss": 0.5523,
+      "step": 6985
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1808911561965942,
+      "learning_rate": 2.8604264071597607e-07,
+      "loss": 0.5198,
+      "step": 6986
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1899702548980713,
+      "learning_rate": 2.8535126751067954e-07,
+      "loss": 0.7027,
+      "step": 6987
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4160841703414917,
+      "learning_rate": 2.846607063077994e-07,
+      "loss": 0.624,
+      "step": 6988
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3612381219863892,
+      "learning_rate": 2.8397095722626833e-07,
+      "loss": 0.6166,
+      "step": 6989
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.779995083808899,
+      "learning_rate": 2.832820203848835e-07,
+      "loss": 0.6595,
+      "step": 6990
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5072723627090454,
+      "learning_rate": 2.8259389590230003e-07,
+      "loss": 0.6803,
+      "step": 6991
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4251368045806885,
+      "learning_rate": 2.8190658389703304e-07,
+      "loss": 0.5835,
+      "step": 6992
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4748808145523071,
+      "learning_rate": 2.8122008448745795e-07,
+      "loss": 0.5532,
+      "step": 6993
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4076902866363525,
+      "learning_rate": 2.805343977918101e-07,
+      "loss": 0.5721,
+      "step": 6994
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2788536548614502,
+      "learning_rate": 2.798495239281868e-07,
+      "loss": 0.5865,
+      "step": 6995
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.204859972000122,
+      "learning_rate": 2.7916546301454185e-07,
+      "loss": 0.5728,
+      "step": 6996
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7112232446670532,
+      "learning_rate": 2.7848221516869114e-07,
+      "loss": 0.6384,
+      "step": 6997
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2741093635559082,
+      "learning_rate": 2.7779978050830993e-07,
+      "loss": 0.61,
+      "step": 6998
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2119300365447998,
+      "learning_rate": 2.771181591509353e-07,
+      "loss": 0.6146,
+      "step": 6999
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.110952615737915,
+      "learning_rate": 2.764373512139612e-07,
+      "loss": 0.6863,
+      "step": 7000
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2552884817123413,
+      "learning_rate": 2.757573568146432e-07,
+      "loss": 0.5632,
+      "step": 7001
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9371869564056396,
+      "learning_rate": 2.750781760700966e-07,
+      "loss": 0.6369,
+      "step": 7002
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.263620138168335,
+      "learning_rate": 2.7439980909729716e-07,
+      "loss": 0.6031,
+      "step": 7003
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5019482374191284,
+      "learning_rate": 2.737222560130787e-07,
+      "loss": 0.5737,
+      "step": 7004
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.525652289390564,
+      "learning_rate": 2.7304551693413616e-07,
+      "loss": 0.6343,
+      "step": 7005
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3184609413146973,
+      "learning_rate": 2.723695919770242e-07,
+      "loss": 0.6735,
+      "step": 7006
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2091474533081055,
+      "learning_rate": 2.7169448125815846e-07,
+      "loss": 0.5457,
+      "step": 7007
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3816118240356445,
+      "learning_rate": 2.7102018489381154e-07,
+      "loss": 0.6264,
+      "step": 7008
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2462316751480103,
+      "learning_rate": 2.7034670300011614e-07,
+      "loss": 0.6312,
+      "step": 7009
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.240803837776184,
+      "learning_rate": 2.696740356930694e-07,
+      "loss": 0.5673,
+      "step": 7010
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3565901517868042,
+      "learning_rate": 2.690021830885214e-07,
+      "loss": 0.6344,
+      "step": 7011
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1310712099075317,
+      "learning_rate": 2.6833114530218694e-07,
+      "loss": 0.5777,
+      "step": 7012
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.0522847175598145,
+      "learning_rate": 2.676609224496374e-07,
+      "loss": 0.6879,
+      "step": 7013
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6499149799346924,
+      "learning_rate": 2.6699151464630644e-07,
+      "loss": 0.6255,
+      "step": 7014
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2910633087158203,
+      "learning_rate": 2.663229220074842e-07,
+      "loss": 0.6169,
+      "step": 7015
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4426332712173462,
+      "learning_rate": 2.6565514464832354e-07,
+      "loss": 0.6411,
+      "step": 7016
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2526745796203613,
+      "learning_rate": 2.6498818268383465e-07,
+      "loss": 0.574,
+      "step": 7017
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5019805431365967,
+      "learning_rate": 2.6432203622888895e-07,
+      "loss": 0.5775,
+      "step": 7018
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2330554723739624,
+      "learning_rate": 2.636567053982164e-07,
+      "loss": 0.4841,
+      "step": 7019
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3986729383468628,
+      "learning_rate": 2.6299219030640586e-07,
+      "loss": 0.5777,
+      "step": 7020
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4264737367630005,
+      "learning_rate": 2.6232849106790745e-07,
+      "loss": 0.6921,
+      "step": 7021
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2543922662734985,
+      "learning_rate": 2.616656077970309e-07,
+      "loss": 0.5413,
+      "step": 7022
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8443100452423096,
+      "learning_rate": 2.6100354060794196e-07,
+      "loss": 0.6575,
+      "step": 7023
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.33084774017334,
+      "learning_rate": 2.603422896146696e-07,
+      "loss": 0.5558,
+      "step": 7024
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9156298637390137,
+      "learning_rate": 2.5968185493110034e-07,
+      "loss": 0.581,
+      "step": 7025
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.377859354019165,
+      "learning_rate": 2.590222366709816e-07,
+      "loss": 0.6449,
+      "step": 7026
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6291786432266235,
+      "learning_rate": 2.58363434947918e-07,
+      "loss": 0.5664,
+      "step": 7027
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.342464566230774,
+      "learning_rate": 2.5770544987537616e-07,
+      "loss": 0.6111,
+      "step": 7028
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6001750230789185,
+      "learning_rate": 2.570482815666797e-07,
+      "loss": 0.5381,
+      "step": 7029
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5080769062042236,
+      "learning_rate": 2.563919301350126e-07,
+      "loss": 0.6536,
+      "step": 7030
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3722752332687378,
+      "learning_rate": 2.557363956934178e-07,
+      "loss": 0.543,
+      "step": 7031
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3873956203460693,
+      "learning_rate": 2.550816783547988e-07,
+      "loss": 0.521,
+      "step": 7032
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.437137246131897,
+      "learning_rate": 2.54427778231916e-07,
+      "loss": 0.5755,
+      "step": 7033
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7044881582260132,
+      "learning_rate": 2.53774695437391e-07,
+      "loss": 0.6265,
+      "step": 7034
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2608447074890137,
+      "learning_rate": 2.531224300837043e-07,
+      "loss": 0.5938,
+      "step": 7035
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.528001308441162,
+      "learning_rate": 2.5247098228319557e-07,
+      "loss": 0.5902,
+      "step": 7036
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4568138122558594,
+      "learning_rate": 2.518203521480622e-07,
+      "loss": 0.5409,
+      "step": 7037
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.466330647468567,
+      "learning_rate": 2.51170539790363e-07,
+      "loss": 0.5988,
+      "step": 7038
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3123421669006348,
+      "learning_rate": 2.50521545322015e-07,
+      "loss": 0.6526,
+      "step": 7039
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.2476837635040283,
+      "learning_rate": 2.498733688547944e-07,
+      "loss": 0.6303,
+      "step": 7040
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.1309330463409424,
+      "learning_rate": 2.492260105003358e-07,
+      "loss": 0.606,
+      "step": 7041
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2924765348434448,
+      "learning_rate": 2.485794703701333e-07,
+      "loss": 0.5738,
+      "step": 7042
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3361096382141113,
+      "learning_rate": 2.479337485755412e-07,
+      "loss": 0.6349,
+      "step": 7043
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2557562589645386,
+      "learning_rate": 2.4728884522777106e-07,
+      "loss": 0.6163,
+      "step": 7044
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.16624116897583,
+      "learning_rate": 2.4664476043789523e-07,
+      "loss": 0.7221,
+      "step": 7045
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5108433961868286,
+      "learning_rate": 2.460014943168443e-07,
+      "loss": 0.6333,
+      "step": 7046
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4388067722320557,
+      "learning_rate": 2.4535904697540746e-07,
+      "loss": 0.5811,
+      "step": 7047
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2010633945465088,
+      "learning_rate": 2.447174185242324e-07,
+      "loss": 0.5747,
+      "step": 7048
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.1095924377441406,
+      "learning_rate": 2.4407660907382727e-07,
+      "loss": 0.5752,
+      "step": 7049
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2455929517745972,
+      "learning_rate": 2.434366187345588e-07,
+      "loss": 0.5444,
+      "step": 7050
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5187287330627441,
+      "learning_rate": 2.4279744761665225e-07,
+      "loss": 0.4958,
+      "step": 7051
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.457276463508606,
+      "learning_rate": 2.4215909583019117e-07,
+      "loss": 0.581,
+      "step": 7052
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4139649868011475,
+      "learning_rate": 2.4152156348511923e-07,
+      "loss": 0.5623,
+      "step": 7053
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4696898460388184,
+      "learning_rate": 2.408848506912381e-07,
+      "loss": 0.5736,
+      "step": 7054
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5024607181549072,
+      "learning_rate": 2.4024895755820956e-07,
+      "loss": 0.6737,
+      "step": 7055
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4502644538879395,
+      "learning_rate": 2.3961388419555145e-07,
+      "loss": 0.5557,
+      "step": 7056
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.718114972114563,
+      "learning_rate": 2.389796307126441e-07,
+      "loss": 0.5784,
+      "step": 7057
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5058298110961914,
+      "learning_rate": 2.3834619721872342e-07,
+      "loss": 0.5671,
+      "step": 7058
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4167766571044922,
+      "learning_rate": 2.3771358382288666e-07,
+      "loss": 0.5516,
+      "step": 7059
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5127261877059937,
+      "learning_rate": 2.370817906340872e-07,
+      "loss": 0.6122,
+      "step": 7060
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1658446788787842,
+      "learning_rate": 2.3645081776113965e-07,
+      "loss": 0.7223,
+      "step": 7061
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3336877822875977,
+      "learning_rate": 2.3582066531271708e-07,
+      "loss": 0.5896,
+      "step": 7062
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4371452331542969,
+      "learning_rate": 2.3519133339734823e-07,
+      "loss": 0.5797,
+      "step": 7063
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.455986738204956,
+      "learning_rate": 2.3456282212342363e-07,
+      "loss": 0.6198,
+      "step": 7064
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5969009399414062,
+      "learning_rate": 2.3393513159919223e-07,
+      "loss": 0.6366,
+      "step": 7065
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3942619562149048,
+      "learning_rate": 2.3330826193276145e-07,
+      "loss": 0.5663,
+      "step": 7066
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.943873405456543,
+      "learning_rate": 2.3268221323209494e-07,
+      "loss": 0.5417,
+      "step": 7067
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2764878273010254,
+      "learning_rate": 2.3205698560501755e-07,
+      "loss": 0.5953,
+      "step": 7068
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2757691144943237,
+      "learning_rate": 2.314325791592126e-07,
+      "loss": 0.5584,
+      "step": 7069
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3400564193725586,
+      "learning_rate": 2.3080899400222178e-07,
+      "loss": 0.5815,
+      "step": 7070
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1935240030288696,
+      "learning_rate": 2.301862302414437e-07,
+      "loss": 0.5787,
+      "step": 7071
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1886448860168457,
+      "learning_rate": 2.2956428798413755e-07,
+      "loss": 0.5357,
+      "step": 7072
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3621718883514404,
+      "learning_rate": 2.2894316733741985e-07,
+      "loss": 0.5345,
+      "step": 7073
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.316648006439209,
+      "learning_rate": 2.2832286840826667e-07,
+      "loss": 0.5749,
+      "step": 7074
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4001599550247192,
+      "learning_rate": 2.277033913035115e-07,
+      "loss": 0.6113,
+      "step": 7075
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.2548327445983887,
+      "learning_rate": 2.2708473612984617e-07,
+      "loss": 0.5458,
+      "step": 7076
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.441633701324463,
+      "learning_rate": 2.2646690299382212e-07,
+      "loss": 0.5897,
+      "step": 7077
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5713400840759277,
+      "learning_rate": 2.258498920018476e-07,
+      "loss": 0.5499,
+      "step": 7078
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.258521318435669,
+      "learning_rate": 2.2523370326019145e-07,
+      "loss": 0.5601,
+      "step": 7079
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3642133474349976,
+      "learning_rate": 2.246183368749799e-07,
+      "loss": 0.5553,
+      "step": 7080
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4287675619125366,
+      "learning_rate": 2.24003792952196e-07,
+      "loss": 0.6197,
+      "step": 7081
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6514580249786377,
+      "learning_rate": 2.233900715976828e-07,
+      "loss": 0.5636,
+      "step": 7082
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6768487691879272,
+      "learning_rate": 2.2277717291714184e-07,
+      "loss": 0.5855,
+      "step": 7083
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3704938888549805,
+      "learning_rate": 2.2216509701613265e-07,
+      "loss": 0.6517,
+      "step": 7084
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2699650526046753,
+      "learning_rate": 2.2155384400007196e-07,
+      "loss": 0.5112,
+      "step": 7085
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.198685884475708,
+      "learning_rate": 2.2094341397423558e-07,
+      "loss": 0.6796,
+      "step": 7086
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5341168642044067,
+      "learning_rate": 2.2033380704375829e-07,
+      "loss": 0.5189,
+      "step": 7087
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5040671825408936,
+      "learning_rate": 2.1972502331363332e-07,
+      "loss": 0.6402,
+      "step": 7088
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.403660774230957,
+      "learning_rate": 2.191170628887096e-07,
+      "loss": 0.5636,
+      "step": 7089
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2295106649398804,
+      "learning_rate": 2.1850992587369668e-07,
+      "loss": 0.5676,
+      "step": 7090
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3213706016540527,
+      "learning_rate": 2.1790361237316204e-07,
+      "loss": 0.5344,
+      "step": 7091
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5314936637878418,
+      "learning_rate": 2.1729812249153048e-07,
+      "loss": 0.6096,
+      "step": 7092
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.497307300567627,
+      "learning_rate": 2.1669345633308526e-07,
+      "loss": 0.5312,
+      "step": 7093
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2101197242736816,
+      "learning_rate": 2.1608961400196747e-07,
+      "loss": 0.6864,
+      "step": 7094
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2792714834213257,
+      "learning_rate": 2.1548659560217678e-07,
+      "loss": 0.529,
+      "step": 7095
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6750179529190063,
+      "learning_rate": 2.148844012375717e-07,
+      "loss": 0.6093,
+      "step": 7096
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2162268161773682,
+      "learning_rate": 2.1428303101186708e-07,
+      "loss": 0.6036,
+      "step": 7097
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2044456005096436,
+      "learning_rate": 2.1368248502863676e-07,
+      "loss": 0.5956,
+      "step": 7098
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3525432348251343,
+      "learning_rate": 2.1308276339131407e-07,
+      "loss": 0.6101,
+      "step": 7099
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3939049243927002,
+      "learning_rate": 2.124838662031864e-07,
+      "loss": 0.5844,
+      "step": 7100
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.0812456607818604,
+      "learning_rate": 2.1188579356740346e-07,
+      "loss": 0.5712,
+      "step": 7101
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.490818738937378,
+      "learning_rate": 2.1128854558697009e-07,
+      "loss": 0.597,
+      "step": 7102
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2523460388183594,
+      "learning_rate": 2.1069212236475177e-07,
+      "loss": 0.7394,
+      "step": 7103
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3945302963256836,
+      "learning_rate": 2.1009652400346802e-07,
+      "loss": 0.57,
+      "step": 7104
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7118395566940308,
+      "learning_rate": 2.0950175060569956e-07,
+      "loss": 0.5877,
+      "step": 7105
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2491495609283447,
+      "learning_rate": 2.0890780227388385e-07,
+      "loss": 0.5995,
+      "step": 7106
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2308311462402344,
+      "learning_rate": 2.08314679110318e-07,
+      "loss": 0.6299,
+      "step": 7107
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.270413875579834,
+      "learning_rate": 2.0772238121715248e-07,
+      "loss": 0.5532,
+      "step": 7108
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3391141891479492,
+      "learning_rate": 2.0713090869640072e-07,
+      "loss": 0.6113,
+      "step": 7109
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.320900559425354,
+      "learning_rate": 2.0654026164993123e-07,
+      "loss": 0.565,
+      "step": 7110
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3125171661376953,
+      "learning_rate": 2.0595044017947153e-07,
+      "loss": 0.621,
+      "step": 7111
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7887307405471802,
+      "learning_rate": 2.053614443866042e-07,
+      "loss": 0.6705,
+      "step": 7112
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.334878921508789,
+      "learning_rate": 2.0477327437277427e-07,
+      "loss": 0.6416,
+      "step": 7113
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0813583135604858,
+      "learning_rate": 2.041859302392818e-07,
+      "loss": 0.5807,
+      "step": 7114
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5186846256256104,
+      "learning_rate": 2.0359941208728363e-07,
+      "loss": 0.6429,
+      "step": 7115
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.855786919593811,
+      "learning_rate": 2.0301372001779673e-07,
+      "loss": 0.664,
+      "step": 7116
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.525266408920288,
+      "learning_rate": 2.0242885413169376e-07,
+      "loss": 0.618,
+      "step": 7117
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3486268520355225,
+      "learning_rate": 2.0184481452970694e-07,
+      "loss": 0.5815,
+      "step": 7118
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5850796699523926,
+      "learning_rate": 2.012616013124241e-07,
+      "loss": 0.5655,
+      "step": 7119
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.358275055885315,
+      "learning_rate": 2.0067921458029272e-07,
+      "loss": 0.5076,
+      "step": 7120
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3294881582260132,
+      "learning_rate": 2.000976544336164e-07,
+      "loss": 0.6009,
+      "step": 7121
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6353390216827393,
+      "learning_rate": 1.9951692097255836e-07,
+      "loss": 0.5333,
+      "step": 7122
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.490020990371704,
+      "learning_rate": 1.989370142971364e-07,
+      "loss": 0.6594,
+      "step": 7123
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7284151315689087,
+      "learning_rate": 1.983579345072284e-07,
+      "loss": 0.5152,
+      "step": 7124
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3690820932388306,
+      "learning_rate": 1.9777968170257012e-07,
+      "loss": 0.5566,
+      "step": 7125
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5744019746780396,
+      "learning_rate": 1.972022559827519e-07,
+      "loss": 0.6513,
+      "step": 7126
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.656482696533203,
+      "learning_rate": 1.9662565744722472e-07,
+      "loss": 0.523,
+      "step": 7127
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.3750336170196533,
+      "learning_rate": 1.9604988619529586e-07,
+      "loss": 0.5436,
+      "step": 7128
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3693758249282837,
+      "learning_rate": 1.954749423261304e-07,
+      "loss": 0.5961,
+      "step": 7129
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8211826086044312,
+      "learning_rate": 1.9490082593875026e-07,
+      "loss": 0.6507,
+      "step": 7130
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.833078384399414,
+      "learning_rate": 1.9432753713203524e-07,
+      "loss": 0.5015,
+      "step": 7131
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.487642526626587,
+      "learning_rate": 1.937550760047241e-07,
+      "loss": 0.5934,
+      "step": 7132
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3412609100341797,
+      "learning_rate": 1.9318344265540967e-07,
+      "loss": 0.564,
+      "step": 7133
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.496861219406128,
+      "learning_rate": 1.926126371825454e-07,
+      "loss": 0.5433,
+      "step": 7134
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3020457029342651,
+      "learning_rate": 1.9204265968444047e-07,
+      "loss": 0.4675,
+      "step": 7135
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3243365287780762,
+      "learning_rate": 1.9147351025926242e-07,
+      "loss": 0.583,
+      "step": 7136
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.173129677772522,
+      "learning_rate": 1.9090518900503508e-07,
+      "loss": 0.6188,
+      "step": 7137
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4048100709915161,
+      "learning_rate": 1.9033769601964013e-07,
+      "loss": 0.5896,
+      "step": 7138
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.206529140472412,
+      "learning_rate": 1.8977103140081664e-07,
+      "loss": 0.5598,
+      "step": 7139
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4039027690887451,
+      "learning_rate": 1.8920519524616265e-07,
+      "loss": 0.5779,
+      "step": 7140
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.485414743423462,
+      "learning_rate": 1.8864018765312963e-07,
+      "loss": 0.5744,
+      "step": 7141
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6506009101867676,
+      "learning_rate": 1.880760087190303e-07,
+      "loss": 0.5637,
+      "step": 7142
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.588201642036438,
+      "learning_rate": 1.8751265854103197e-07,
+      "loss": 0.4825,
+      "step": 7143
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8149200677871704,
+      "learning_rate": 1.869501372161614e-07,
+      "loss": 0.5513,
+      "step": 7144
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.2272982597351074,
+      "learning_rate": 1.8638844484130058e-07,
+      "loss": 0.5979,
+      "step": 7145
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5402497053146362,
+      "learning_rate": 1.858275815131888e-07,
+      "loss": 0.5783,
+      "step": 7146
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4437787532806396,
+      "learning_rate": 1.85267547328426e-07,
+      "loss": 0.5892,
+      "step": 7147
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.1562960147857666,
+      "learning_rate": 1.8470834238346448e-07,
+      "loss": 0.5223,
+      "step": 7148
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4266095161437988,
+      "learning_rate": 1.8414996677461605e-07,
+      "loss": 0.6465,
+      "step": 7149
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4627182483673096,
+      "learning_rate": 1.8359242059805048e-07,
+      "loss": 0.6431,
+      "step": 7150
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4455598592758179,
+      "learning_rate": 1.8303570394979375e-07,
+      "loss": 0.6584,
+      "step": 7151
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3495609760284424,
+      "learning_rate": 1.8247981692572802e-07,
+      "loss": 0.6074,
+      "step": 7152
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.1152602434158325,
+      "learning_rate": 1.8192475962159395e-07,
+      "loss": 0.4656,
+      "step": 7153
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3456722497940063,
+      "learning_rate": 1.8137053213298895e-07,
+      "loss": 0.5437,
+      "step": 7154
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8487489223480225,
+      "learning_rate": 1.808171345553683e-07,
+      "loss": 0.5297,
+      "step": 7155
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3239003419876099,
+      "learning_rate": 1.8026456698404192e-07,
+      "loss": 0.5854,
+      "step": 7156
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.531930923461914,
+      "learning_rate": 1.7971282951417923e-07,
+      "loss": 0.5994,
+      "step": 7157
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3221821784973145,
+      "learning_rate": 1.7916192224080586e-07,
+      "loss": 0.5434,
+      "step": 7158
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7776752710342407,
+      "learning_rate": 1.7861184525880425e-07,
+      "loss": 0.6325,
+      "step": 7159
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3057001829147339,
+      "learning_rate": 1.7806259866291366e-07,
+      "loss": 0.548,
+      "step": 7160
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2270419597625732,
+      "learning_rate": 1.775141825477311e-07,
+      "loss": 0.5853,
+      "step": 7161
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3529689311981201,
+      "learning_rate": 1.7696659700770948e-07,
+      "loss": 0.5753,
+      "step": 7162
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.1730376482009888,
+      "learning_rate": 1.764198421371599e-07,
+      "loss": 0.5227,
+      "step": 7163
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4664965867996216,
+      "learning_rate": 1.758739180302499e-07,
+      "loss": 0.5583,
+      "step": 7164
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7074847221374512,
+      "learning_rate": 1.753288247810031e-07,
+      "loss": 0.6546,
+      "step": 7165
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.511422872543335,
+      "learning_rate": 1.7478456248330166e-07,
+      "loss": 0.6486,
+      "step": 7166
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5393157005310059,
+      "learning_rate": 1.7424113123088282e-07,
+      "loss": 0.6168,
+      "step": 7167
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.8663549423217773,
+      "learning_rate": 1.736985311173417e-07,
+      "loss": 0.6082,
+      "step": 7168
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5243266820907593,
+      "learning_rate": 1.7315676223613077e-07,
+      "loss": 0.5553,
+      "step": 7169
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2957146167755127,
+      "learning_rate": 1.7261582468055872e-07,
+      "loss": 0.5589,
+      "step": 7170
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3966628313064575,
+      "learning_rate": 1.720757185437899e-07,
+      "loss": 0.5563,
+      "step": 7171
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2726695537567139,
+      "learning_rate": 1.7153644391884715e-07,
+      "loss": 0.6512,
+      "step": 7172
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4725221395492554,
+      "learning_rate": 1.7099800089861006e-07,
+      "loss": 0.5772,
+      "step": 7173
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5598526000976562,
+      "learning_rate": 1.7046038957581447e-07,
+      "loss": 0.6274,
+      "step": 7174
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.330385684967041,
+      "learning_rate": 1.6992361004305235e-07,
+      "loss": 0.5829,
+      "step": 7175
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.331067681312561,
+      "learning_rate": 1.693876623927737e-07,
+      "loss": 0.5583,
+      "step": 7176
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3554342985153198,
+      "learning_rate": 1.6885254671728468e-07,
+      "loss": 0.5674,
+      "step": 7177
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.254412055015564,
+      "learning_rate": 1.683182631087471e-07,
+      "loss": 0.6022,
+      "step": 7178
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.508400321006775,
+      "learning_rate": 1.677848116591807e-07,
+      "loss": 0.5477,
+      "step": 7179
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2750962972640991,
+      "learning_rate": 1.6725219246046254e-07,
+      "loss": 0.5805,
+      "step": 7180
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5220352411270142,
+      "learning_rate": 1.6672040560432533e-07,
+      "loss": 0.6261,
+      "step": 7181
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.256548523902893,
+      "learning_rate": 1.661894511823575e-07,
+      "loss": 0.5502,
+      "step": 7182
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.461651086807251,
+      "learning_rate": 1.6565932928600593e-07,
+      "loss": 0.4907,
+      "step": 7183
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.1890257596969604,
+      "learning_rate": 1.6513004000657418e-07,
+      "loss": 0.5885,
+      "step": 7184
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4478956460952759,
+      "learning_rate": 1.6460158343521993e-07,
+      "loss": 0.5186,
+      "step": 7185
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.0976368188858032,
+      "learning_rate": 1.640739596629598e-07,
+      "loss": 0.7179,
+      "step": 7186
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2835136651992798,
+      "learning_rate": 1.6354716878066612e-07,
+      "loss": 0.5308,
+      "step": 7187
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4048012495040894,
+      "learning_rate": 1.6302121087906854e-07,
+      "loss": 0.6106,
+      "step": 7188
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2641879320144653,
+      "learning_rate": 1.6249608604875177e-07,
+      "loss": 0.6223,
+      "step": 7189
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7437182664871216,
+      "learning_rate": 1.6197179438015798e-07,
+      "loss": 0.5893,
+      "step": 7190
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3747304677963257,
+      "learning_rate": 1.6144833596358656e-07,
+      "loss": 0.5148,
+      "step": 7191
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4542993307113647,
+      "learning_rate": 1.6092571088919205e-07,
+      "loss": 0.5563,
+      "step": 7192
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3954862356185913,
+      "learning_rate": 1.6040391924698584e-07,
+      "loss": 0.6165,
+      "step": 7193
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.301512598991394,
+      "learning_rate": 1.5988296112683598e-07,
+      "loss": 0.5522,
+      "step": 7194
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3332188129425049,
+      "learning_rate": 1.5936283661846686e-07,
+      "loss": 0.5854,
+      "step": 7195
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3937634229660034,
+      "learning_rate": 1.588435458114601e-07,
+      "loss": 0.6015,
+      "step": 7196
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4809768199920654,
+      "learning_rate": 1.5832508879525143e-07,
+      "loss": 0.4986,
+      "step": 7197
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.265257477760315,
+      "learning_rate": 1.5780746565913552e-07,
+      "loss": 0.6384,
+      "step": 7198
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.1740734577178955,
+      "learning_rate": 1.5729067649226327e-07,
+      "loss": 0.5811,
+      "step": 7199
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4921587705612183,
+      "learning_rate": 1.5677472138364014e-07,
+      "loss": 0.5796,
+      "step": 7200
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2573281526565552,
+      "learning_rate": 1.562596004221284e-07,
+      "loss": 0.5726,
+      "step": 7201
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4946744441986084,
+      "learning_rate": 1.557453136964482e-07,
+      "loss": 0.568,
+      "step": 7202
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.1617540121078491,
+      "learning_rate": 1.5523186129517475e-07,
+      "loss": 0.5592,
+      "step": 7203
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2968569993972778,
+      "learning_rate": 1.547192433067396e-07,
+      "loss": 0.5903,
+      "step": 7204
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3031530380249023,
+      "learning_rate": 1.5420745981943042e-07,
+      "loss": 0.5648,
+      "step": 7205
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2497785091400146,
+      "learning_rate": 1.5369651092139172e-07,
+      "loss": 0.5518,
+      "step": 7206
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.333309531211853,
+      "learning_rate": 1.531863967006253e-07,
+      "loss": 0.5962,
+      "step": 7207
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2219014167785645,
+      "learning_rate": 1.526771172449859e-07,
+      "loss": 0.6769,
+      "step": 7208
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6595335006713867,
+      "learning_rate": 1.5216867264218726e-07,
+      "loss": 0.5918,
+      "step": 7209
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4741884469985962,
+      "learning_rate": 1.516610629797993e-07,
+      "loss": 0.6679,
+      "step": 7210
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4178804159164429,
+      "learning_rate": 1.511542883452477e-07,
+      "loss": 0.592,
+      "step": 7211
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6573342084884644,
+      "learning_rate": 1.5064834882581314e-07,
+      "loss": 0.6468,
+      "step": 7212
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2453925609588623,
+      "learning_rate": 1.5014324450863316e-07,
+      "loss": 0.5826,
+      "step": 7213
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.3448381423950195,
+      "learning_rate": 1.496389754807026e-07,
+      "loss": 0.6063,
+      "step": 7214
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3461638689041138,
+      "learning_rate": 1.4913554182887147e-07,
+      "loss": 0.6598,
+      "step": 7215
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6426374912261963,
+      "learning_rate": 1.4863294363984536e-07,
+      "loss": 0.6191,
+      "step": 7216
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.748055100440979,
+      "learning_rate": 1.481311810001873e-07,
+      "loss": 0.6522,
+      "step": 7217
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.689281702041626,
+      "learning_rate": 1.4763025399631535e-07,
+      "loss": 0.6215,
+      "step": 7218
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.246228575706482,
+      "learning_rate": 1.4713016271450443e-07,
+      "loss": 0.5989,
+      "step": 7219
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.082564353942871,
+      "learning_rate": 1.466309072408839e-07,
+      "loss": 0.6292,
+      "step": 7220
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7110521793365479,
+      "learning_rate": 1.4613248766144172e-07,
+      "loss": 0.5804,
+      "step": 7221
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2282060384750366,
+      "learning_rate": 1.456349040620203e-07,
+      "loss": 0.6041,
+      "step": 7222
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.646616816520691,
+      "learning_rate": 1.4513815652831776e-07,
+      "loss": 0.604,
+      "step": 7223
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2719740867614746,
+      "learning_rate": 1.4464224514588842e-07,
+      "loss": 0.6009,
+      "step": 7224
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.63080632686615,
+      "learning_rate": 1.4414717000014456e-07,
+      "loss": 0.5949,
+      "step": 7225
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7271970510482788,
+      "learning_rate": 1.436529311763507e-07,
+      "loss": 0.5815,
+      "step": 7226
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3970266580581665,
+      "learning_rate": 1.43159528759631e-07,
+      "loss": 0.5571,
+      "step": 7227
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2534281015396118,
+      "learning_rate": 1.426669628349636e-07,
+      "loss": 0.6054,
+      "step": 7228
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.373757004737854,
+      "learning_rate": 1.4217523348718287e-07,
+      "loss": 0.6324,
+      "step": 7229
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6304547786712646,
+      "learning_rate": 1.4168434080097937e-07,
+      "loss": 0.6204,
+      "step": 7230
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.140513300895691,
+      "learning_rate": 1.411942848608988e-07,
+      "loss": 0.6375,
+      "step": 7231
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1663187742233276,
+      "learning_rate": 1.4070506575134367e-07,
+      "loss": 0.5516,
+      "step": 7232
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.328594446182251,
+      "learning_rate": 1.4021668355657215e-07,
+      "loss": 0.5821,
+      "step": 7233
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4575828313827515,
+      "learning_rate": 1.3972913836069857e-07,
+      "loss": 0.6476,
+      "step": 7234
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5077722072601318,
+      "learning_rate": 1.3924243024769135e-07,
+      "loss": 0.5528,
+      "step": 7235
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1910840272903442,
+      "learning_rate": 1.387565593013779e-07,
+      "loss": 0.5179,
+      "step": 7236
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.430280089378357,
+      "learning_rate": 1.3827152560543843e-07,
+      "loss": 0.6201,
+      "step": 7237
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1076302528381348,
+      "learning_rate": 1.377873292434101e-07,
+      "loss": 0.6917,
+      "step": 7238
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.003353476524353,
+      "learning_rate": 1.3730397029868615e-07,
+      "loss": 0.559,
+      "step": 7239
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5710481405258179,
+      "learning_rate": 1.3682144885451555e-07,
+      "loss": 0.5213,
+      "step": 7240
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5206317901611328,
+      "learning_rate": 1.3633976499400235e-07,
+      "loss": 0.59,
+      "step": 7241
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4319339990615845,
+      "learning_rate": 1.3585891880010747e-07,
+      "loss": 0.6111,
+      "step": 7242
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.376744031906128,
+      "learning_rate": 1.3537891035564576e-07,
+      "loss": 0.6414,
+      "step": 7243
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5035916566848755,
+      "learning_rate": 1.3489973974329053e-07,
+      "loss": 0.732,
+      "step": 7244
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.226749062538147,
+      "learning_rate": 1.3442140704556794e-07,
+      "loss": 0.5514,
+      "step": 7245
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.0988937616348267,
+      "learning_rate": 1.3394391234486104e-07,
+      "loss": 0.5946,
+      "step": 7246
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2332367897033691,
+      "learning_rate": 1.3346725572340903e-07,
+      "loss": 0.5686,
+      "step": 7247
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4782696962356567,
+      "learning_rate": 1.3299143726330676e-07,
+      "loss": 0.6448,
+      "step": 7248
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3674440383911133,
+      "learning_rate": 1.3251645704650262e-07,
+      "loss": 0.6015,
+      "step": 7249
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3126945495605469,
+      "learning_rate": 1.3204231515480447e-07,
+      "loss": 0.6499,
+      "step": 7250
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5984584093093872,
+      "learning_rate": 1.3156901166987258e-07,
+      "loss": 0.6032,
+      "step": 7251
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3169101476669312,
+      "learning_rate": 1.3109654667322337e-07,
+      "loss": 0.5563,
+      "step": 7252
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4062318801879883,
+      "learning_rate": 1.306249202462301e-07,
+      "loss": 0.6181,
+      "step": 7253
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.340091347694397,
+      "learning_rate": 1.3015413247012054e-07,
+      "loss": 0.592,
+      "step": 7254
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2546271085739136,
+      "learning_rate": 1.2968418342597876e-07,
+      "loss": 0.6095,
+      "step": 7255
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1093549728393555,
+      "learning_rate": 1.2921507319474337e-07,
+      "loss": 0.7472,
+      "step": 7256
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3403831720352173,
+      "learning_rate": 1.2874680185720857e-07,
+      "loss": 0.484,
+      "step": 7257
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4152884483337402,
+      "learning_rate": 1.2827936949402598e-07,
+      "loss": 0.6114,
+      "step": 7258
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3531930446624756,
+      "learning_rate": 1.2781277618570066e-07,
+      "loss": 0.5642,
+      "step": 7259
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5098471641540527,
+      "learning_rate": 1.273470220125933e-07,
+      "loss": 0.5642,
+      "step": 7260
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4057996273040771,
+      "learning_rate": 1.2688210705492088e-07,
+      "loss": 0.6116,
+      "step": 7261
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2736021280288696,
+      "learning_rate": 1.26418031392756e-07,
+      "loss": 0.6281,
+      "step": 7262
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.0702874660491943,
+      "learning_rate": 1.2595479510602583e-07,
+      "loss": 0.6903,
+      "step": 7263
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.287724494934082,
+      "learning_rate": 1.2549239827451375e-07,
+      "loss": 0.5868,
+      "step": 7264
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1175434589385986,
+      "learning_rate": 1.2503084097785777e-07,
+      "loss": 0.6007,
+      "step": 7265
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1822404861450195,
+      "learning_rate": 1.2457012329555206e-07,
+      "loss": 0.5906,
+      "step": 7266
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4445996284484863,
+      "learning_rate": 1.2411024530694537e-07,
+      "loss": 0.6044,
+      "step": 7267
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1158324480056763,
+      "learning_rate": 1.2365120709124322e-07,
+      "loss": 0.6941,
+      "step": 7268
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6972533464431763,
+      "learning_rate": 1.2319300872750518e-07,
+      "loss": 0.6134,
+      "step": 7269
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1953438520431519,
+      "learning_rate": 1.2273565029464694e-07,
+      "loss": 0.7026,
+      "step": 7270
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5305699110031128,
+      "learning_rate": 1.2227913187143835e-07,
+      "loss": 0.547,
+      "step": 7271
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.451676607131958,
+      "learning_rate": 1.2182345353650592e-07,
+      "loss": 0.6202,
+      "step": 7272
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5434238910675049,
+      "learning_rate": 1.2136861536833134e-07,
+      "loss": 0.6061,
+      "step": 7273
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2015151977539062,
+      "learning_rate": 1.209146174452508e-07,
+      "loss": 0.4739,
+      "step": 7274
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.211741328239441,
+      "learning_rate": 1.2046145984545565e-07,
+      "loss": 0.6157,
+      "step": 7275
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2866265773773193,
+      "learning_rate": 1.2000914264699403e-07,
+      "loss": 0.5505,
+      "step": 7276
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.437593698501587,
+      "learning_rate": 1.1955766592776863e-07,
+      "loss": 0.5961,
+      "step": 7277
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3555943965911865,
+      "learning_rate": 1.1910702976553612e-07,
+      "loss": 0.642,
+      "step": 7278
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1852178573608398,
+      "learning_rate": 1.1865723423790999e-07,
+      "loss": 0.5775,
+      "step": 7279
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3030401468276978,
+      "learning_rate": 1.1820827942235824e-07,
+      "loss": 0.5631,
+      "step": 7280
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.204147219657898,
+      "learning_rate": 1.177601653962046e-07,
+      "loss": 0.733,
+      "step": 7281
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1808563470840454,
+      "learning_rate": 1.1731289223662679e-07,
+      "loss": 0.5254,
+      "step": 7282
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5205473899841309,
+      "learning_rate": 1.1686646002065815e-07,
+      "loss": 0.586,
+      "step": 7283
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6285957098007202,
+      "learning_rate": 1.1642086882519e-07,
+      "loss": 0.6336,
+      "step": 7284
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5329946279525757,
+      "learning_rate": 1.1597611872696368e-07,
+      "loss": 0.6124,
+      "step": 7285
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4156322479248047,
+      "learning_rate": 1.1553220980257962e-07,
+      "loss": 0.597,
+      "step": 7286
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3494125604629517,
+      "learning_rate": 1.150891421284922e-07,
+      "loss": 0.5883,
+      "step": 7287
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2939672470092773,
+      "learning_rate": 1.146469157810104e-07,
+      "loss": 0.6399,
+      "step": 7288
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.1361777782440186,
+      "learning_rate": 1.1420553083629882e-07,
+      "loss": 0.6024,
+      "step": 7289
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2731001377105713,
+      "learning_rate": 1.1376498737037722e-07,
+      "loss": 0.5837,
+      "step": 7290
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1375939846038818,
+      "learning_rate": 1.1332528545911992e-07,
+      "loss": 0.6947,
+      "step": 7291
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.776005744934082,
+      "learning_rate": 1.1288642517825688e-07,
+      "loss": 0.5576,
+      "step": 7292
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6414631605148315,
+      "learning_rate": 1.1244840660337264e-07,
+      "loss": 0.5971,
+      "step": 7293
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1578058004379272,
+      "learning_rate": 1.1201122980990742e-07,
+      "loss": 0.7463,
+      "step": 7294
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6849491596221924,
+      "learning_rate": 1.1157489487315542e-07,
+      "loss": 0.581,
+      "step": 7295
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2874782085418701,
+      "learning_rate": 1.1113940186826767e-07,
+      "loss": 0.5576,
+      "step": 7296
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.0534873008728027,
+      "learning_rate": 1.1070475087024745e-07,
+      "loss": 0.6968,
+      "step": 7297
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.308283805847168,
+      "learning_rate": 1.1027094195395548e-07,
+      "loss": 0.5559,
+      "step": 7298
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.334275245666504,
+      "learning_rate": 1.0983797519410588e-07,
+      "loss": 0.6184,
+      "step": 7299
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2813701629638672,
+      "learning_rate": 1.0940585066526899e-07,
+      "loss": 0.587,
+      "step": 7300
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.355509638786316,
+      "learning_rate": 1.0897456844186971e-07,
+      "loss": 0.6175,
+      "step": 7301
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4116696119308472,
+      "learning_rate": 1.0854412859818642e-07,
+      "loss": 0.5073,
+      "step": 7302
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.340970516204834,
+      "learning_rate": 1.0811453120835535e-07,
+      "loss": 0.5772,
+      "step": 7303
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5687137842178345,
+      "learning_rate": 1.0768577634636457e-07,
+      "loss": 0.5782,
+      "step": 7304
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.321331262588501,
+      "learning_rate": 1.0725786408605887e-07,
+      "loss": 0.5902,
+      "step": 7305
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3391438722610474,
+      "learning_rate": 1.068307945011371e-07,
+      "loss": 0.6029,
+      "step": 7306
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2736539840698242,
+      "learning_rate": 1.0640456766515428e-07,
+      "loss": 0.6101,
+      "step": 7307
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5165783166885376,
+      "learning_rate": 1.0597918365151838e-07,
+      "loss": 0.5793,
+      "step": 7308
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2305033206939697,
+      "learning_rate": 1.0555464253349301e-07,
+      "loss": 0.5675,
+      "step": 7309
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2480313777923584,
+      "learning_rate": 1.0513094438419747e-07,
+      "loss": 0.518,
+      "step": 7310
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5531625747680664,
+      "learning_rate": 1.0470808927660503e-07,
+      "loss": 0.6093,
+      "step": 7311
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1417747735977173,
+      "learning_rate": 1.0428607728354301e-07,
+      "loss": 0.7112,
+      "step": 7312
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.267432689666748,
+      "learning_rate": 1.0386490847769547e-07,
+      "loss": 0.6231,
+      "step": 7313
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4412930011749268,
+      "learning_rate": 1.0344458293159998e-07,
+      "loss": 0.6256,
+      "step": 7314
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.665556788444519,
+      "learning_rate": 1.030251007176486e-07,
+      "loss": 0.5757,
+      "step": 7315
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5385949611663818,
+      "learning_rate": 1.026064619080891e-07,
+      "loss": 0.657,
+      "step": 7316
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.289075255393982,
+      "learning_rate": 1.0218866657502268e-07,
+      "loss": 0.6306,
+      "step": 7317
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4588640928268433,
+      "learning_rate": 1.0177171479040737e-07,
+      "loss": 0.5182,
+      "step": 7318
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3852373361587524,
+      "learning_rate": 1.0135560662605404e-07,
+      "loss": 0.5512,
+      "step": 7319
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2439759969711304,
+      "learning_rate": 1.0094034215362814e-07,
+      "loss": 0.639,
+      "step": 7320
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4485936164855957,
+      "learning_rate": 1.0052592144465189e-07,
+      "loss": 0.6265,
+      "step": 7321
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1735234260559082,
+      "learning_rate": 1.0011234457050045e-07,
+      "loss": 0.7429,
+      "step": 7322
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1867866516113281,
+      "learning_rate": 9.969961160240294e-08,
+      "loss": 0.563,
+      "step": 7323
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3140125274658203,
+      "learning_rate": 9.928772261144526e-08,
+      "loss": 0.5963,
+      "step": 7324
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2439604997634888,
+      "learning_rate": 9.887667766856678e-08,
+      "loss": 0.6702,
+      "step": 7325
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1434909105300903,
+      "learning_rate": 9.846647684456146e-08,
+      "loss": 0.5945,
+      "step": 7326
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1893295049667358,
+      "learning_rate": 9.805712021007774e-08,
+      "loss": 0.5411,
+      "step": 7327
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2865269184112549,
+      "learning_rate": 9.764860783561981e-08,
+      "loss": 0.676,
+      "step": 7328
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5296525955200195,
+      "learning_rate": 9.724093979154526e-08,
+      "loss": 0.5856,
+      "step": 7329
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.277660608291626,
+      "learning_rate": 9.683411614806626e-08,
+      "loss": 0.6107,
+      "step": 7330
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.141019344329834,
+      "learning_rate": 9.642813697525011e-08,
+      "loss": 0.5483,
+      "step": 7331
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5347297191619873,
+      "learning_rate": 9.602300234301865e-08,
+      "loss": 0.6413,
+      "step": 7332
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3887579441070557,
+      "learning_rate": 9.561871232114827e-08,
+      "loss": 0.5488,
+      "step": 7333
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2093347311019897,
+      "learning_rate": 9.521526697926887e-08,
+      "loss": 0.6469,
+      "step": 7334
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4027270078659058,
+      "learning_rate": 9.481266638686592e-08,
+      "loss": 0.5732,
+      "step": 7335
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4226722717285156,
+      "learning_rate": 9.441091061328067e-08,
+      "loss": 0.5628,
+      "step": 7336
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2715821266174316,
+      "learning_rate": 9.400999972770553e-08,
+      "loss": 0.5455,
+      "step": 7337
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5981942415237427,
+      "learning_rate": 9.360993379919026e-08,
+      "loss": 0.5994,
+      "step": 7338
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1296770572662354,
+      "learning_rate": 9.321071289663753e-08,
+      "loss": 0.6579,
+      "step": 7339
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.32581627368927,
+      "learning_rate": 9.281233708880566e-08,
+      "loss": 0.557,
+      "step": 7340
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3479461669921875,
+      "learning_rate": 9.241480644430644e-08,
+      "loss": 0.7366,
+      "step": 7341
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.926008701324463,
+      "learning_rate": 9.201812103160624e-08,
+      "loss": 0.6514,
+      "step": 7342
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4221993684768677,
+      "learning_rate": 9.162228091902593e-08,
+      "loss": 0.5706,
+      "step": 7343
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6545454263687134,
+      "learning_rate": 9.122728617474153e-08,
+      "loss": 0.6052,
+      "step": 7344
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3921095132827759,
+      "learning_rate": 9.083313686678252e-08,
+      "loss": 0.5529,
+      "step": 7345
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1098374128341675,
+      "learning_rate": 9.043983306303295e-08,
+      "loss": 0.7025,
+      "step": 7346
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4808984994888306,
+      "learning_rate": 9.004737483123194e-08,
+      "loss": 0.6095,
+      "step": 7347
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2280536890029907,
+      "learning_rate": 8.96557622389721e-08,
+      "loss": 0.6116,
+      "step": 7348
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2250851392745972,
+      "learning_rate": 8.926499535370058e-08,
+      "loss": 0.5939,
+      "step": 7349
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9256665706634521,
+      "learning_rate": 8.887507424271913e-08,
+      "loss": 0.5694,
+      "step": 7350
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2938241958618164,
+      "learning_rate": 8.8485998973184e-08,
+      "loss": 0.5579,
+      "step": 7351
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3634623289108276,
+      "learning_rate": 8.809776961210493e-08,
+      "loss": 0.5262,
+      "step": 7352
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.161146879196167,
+      "learning_rate": 8.771038622634731e-08,
+      "loss": 0.5316,
+      "step": 7353
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3339877128601074,
+      "learning_rate": 8.732384888262945e-08,
+      "loss": 0.5637,
+      "step": 7354
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.254198670387268,
+      "learning_rate": 8.693815764752477e-08,
+      "loss": 0.6377,
+      "step": 7355
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2894302606582642,
+      "learning_rate": 8.65533125874607e-08,
+      "loss": 0.5951,
+      "step": 7356
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4392236471176147,
+      "learning_rate": 8.616931376871917e-08,
+      "loss": 0.6323,
+      "step": 7357
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.222009539604187,
+      "learning_rate": 8.578616125743567e-08,
+      "loss": 0.6005,
+      "step": 7358
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.019447922706604,
+      "learning_rate": 8.540385511960126e-08,
+      "loss": 0.5459,
+      "step": 7359
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4560258388519287,
+      "learning_rate": 8.502239542105994e-08,
+      "loss": 0.6215,
+      "step": 7360
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.296736240386963,
+      "learning_rate": 8.464178222750974e-08,
+      "loss": 0.5495,
+      "step": 7361
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5001124143600464,
+      "learning_rate": 8.426201560450431e-08,
+      "loss": 0.6199,
+      "step": 7362
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5553319454193115,
+      "learning_rate": 8.388309561745134e-08,
+      "loss": 0.5854,
+      "step": 7363
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3460683822631836,
+      "learning_rate": 8.350502233161084e-08,
+      "loss": 0.5443,
+      "step": 7364
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.374433994293213,
+      "learning_rate": 8.312779581209852e-08,
+      "loss": 0.6094,
+      "step": 7365
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3948500156402588,
+      "learning_rate": 8.27514161238846e-08,
+      "loss": 0.5573,
+      "step": 7366
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.369402289390564,
+      "learning_rate": 8.237588333179225e-08,
+      "loss": 0.4974,
+      "step": 7367
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2202465534210205,
+      "learning_rate": 8.200119750049918e-08,
+      "loss": 0.5171,
+      "step": 7368
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5176337957382202,
+      "learning_rate": 8.162735869453764e-08,
+      "loss": 0.6837,
+      "step": 7369
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.0145180225372314,
+      "learning_rate": 8.125436697829503e-08,
+      "loss": 0.5776,
+      "step": 7370
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.281848669052124,
+      "learning_rate": 8.08822224160094e-08,
+      "loss": 0.6065,
+      "step": 7371
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6321043968200684,
+      "learning_rate": 8.051092507177672e-08,
+      "loss": 0.6207,
+      "step": 7372
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5435349941253662,
+      "learning_rate": 8.014047500954524e-08,
+      "loss": 0.6096,
+      "step": 7373
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.468544840812683,
+      "learning_rate": 7.97708722931162e-08,
+      "loss": 0.6066,
+      "step": 7374
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.739684581756592,
+      "learning_rate": 7.940211698614753e-08,
+      "loss": 0.5939,
+      "step": 7375
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.334362506866455,
+      "learning_rate": 7.903420915214899e-08,
+      "loss": 0.4706,
+      "step": 7376
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2078897953033447,
+      "learning_rate": 7.866714885448601e-08,
+      "loss": 0.5775,
+      "step": 7377
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5452244281768799,
+      "learning_rate": 7.830093615637635e-08,
+      "loss": 0.6114,
+      "step": 7378
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3688369989395142,
+      "learning_rate": 7.793557112089345e-08,
+      "loss": 0.5782,
+      "step": 7379
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3603882789611816,
+      "learning_rate": 7.757105381096363e-08,
+      "loss": 0.6609,
+      "step": 7380
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3486891984939575,
+      "learning_rate": 7.72073842893678e-08,
+      "loss": 0.5362,
+      "step": 7381
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3633257150650024,
+      "learning_rate": 7.684456261874085e-08,
+      "loss": 0.637,
+      "step": 7382
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2990694046020508,
+      "learning_rate": 7.648258886157056e-08,
+      "loss": 0.5077,
+      "step": 7383
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.217881202697754,
+      "learning_rate": 7.612146308019985e-08,
+      "loss": 0.5874,
+      "step": 7384
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1617398262023926,
+      "learning_rate": 7.57611853368262e-08,
+      "loss": 0.5252,
+      "step": 7385
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2338204383850098,
+      "learning_rate": 7.540175569349883e-08,
+      "loss": 0.5152,
+      "step": 7386
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3661803007125854,
+      "learning_rate": 7.504317421212326e-08,
+      "loss": 0.5535,
+      "step": 7387
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4716441631317139,
+      "learning_rate": 7.468544095445784e-08,
+      "loss": 0.6154,
+      "step": 7388
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3351925611495972,
+      "learning_rate": 7.432855598211386e-08,
+      "loss": 0.6322,
+      "step": 7389
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1938472986221313,
+      "learning_rate": 7.397251935655825e-08,
+      "loss": 0.5131,
+      "step": 7390
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2754592895507812,
+      "learning_rate": 7.36173311391114e-08,
+      "loss": 0.5288,
+      "step": 7391
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3025398254394531,
+      "learning_rate": 7.326299139094662e-08,
+      "loss": 0.6064,
+      "step": 7392
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4515681266784668,
+      "learning_rate": 7.290950017309173e-08,
+      "loss": 0.5584,
+      "step": 7393
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6512247323989868,
+      "learning_rate": 7.255685754642805e-08,
+      "loss": 0.5702,
+      "step": 7394
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3409143686294556,
+      "learning_rate": 7.220506357169255e-08,
+      "loss": 0.5642,
+      "step": 7395
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2338666915893555,
+      "learning_rate": 7.185411830947342e-08,
+      "loss": 0.5008,
+      "step": 7396
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2696747779846191,
+      "learning_rate": 7.150402182021399e-08,
+      "loss": 0.5333,
+      "step": 7397
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2724928855895996,
+      "learning_rate": 7.115477416421101e-08,
+      "loss": 0.6582,
+      "step": 7398
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1926487684249878,
+      "learning_rate": 7.080637540161583e-08,
+      "loss": 0.6197,
+      "step": 7399
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4144186973571777,
+      "learning_rate": 7.045882559243322e-08,
+      "loss": 0.6195,
+      "step": 7400
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3399546146392822,
+      "learning_rate": 7.011212479652141e-08,
+      "loss": 0.5682,
+      "step": 7401
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.146112084388733,
+      "learning_rate": 6.976627307359208e-08,
+      "loss": 0.5884,
+      "step": 7402
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5165796279907227,
+      "learning_rate": 6.94212704832109e-08,
+      "loss": 0.6208,
+      "step": 7403
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4220664501190186,
+      "learning_rate": 6.90771170847987e-08,
+      "loss": 0.6322,
+      "step": 7404
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4657684564590454,
+      "learning_rate": 6.873381293762859e-08,
+      "loss": 0.5636,
+      "step": 7405
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.468279480934143,
+      "learning_rate": 6.839135810082664e-08,
+      "loss": 0.5901,
+      "step": 7406
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.511237144470215,
+      "learning_rate": 6.804975263337565e-08,
+      "loss": 0.5578,
+      "step": 7407
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1953728199005127,
+      "learning_rate": 6.770899659410856e-08,
+      "loss": 0.5636,
+      "step": 7408
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.134719967842102,
+      "learning_rate": 6.736909004171399e-08,
+      "loss": 0.6885,
+      "step": 7409
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5010466575622559,
+      "learning_rate": 6.703003303473454e-08,
+      "loss": 0.5951,
+      "step": 7410
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.215299367904663,
+      "learning_rate": 6.669182563156574e-08,
+      "loss": 0.5923,
+      "step": 7411
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.380470633506775,
+      "learning_rate": 6.635446789045651e-08,
+      "loss": 0.6177,
+      "step": 7412
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3253923654556274,
+      "learning_rate": 6.601795986951044e-08,
+      "loss": 0.6369,
+      "step": 7413
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4012606143951416,
+      "learning_rate": 6.568230162668332e-08,
+      "loss": 0.6044,
+      "step": 7414
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2084659337997437,
+      "learning_rate": 6.534749321978617e-08,
+      "loss": 0.7131,
+      "step": 7415
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2429335117340088,
+      "learning_rate": 6.501353470648286e-08,
+      "loss": 0.5506,
+      "step": 7416
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1838606595993042,
+      "learning_rate": 6.46804261442907e-08,
+      "loss": 0.581,
+      "step": 7417
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5430114269256592,
+      "learning_rate": 6.43481675905816e-08,
+      "loss": 0.6428,
+      "step": 7418
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4366458654403687,
+      "learning_rate": 6.40167591025792e-08,
+      "loss": 0.5344,
+      "step": 7419
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7072436809539795,
+      "learning_rate": 6.368620073736176e-08,
+      "loss": 0.6056,
+      "step": 7420
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.645785927772522,
+      "learning_rate": 6.335649255186315e-08,
+      "loss": 0.6395,
+      "step": 7421
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3543944358825684,
+      "learning_rate": 6.302763460286687e-08,
+      "loss": 0.5263,
+      "step": 7422
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.483001232147217,
+      "learning_rate": 6.269962694701315e-08,
+      "loss": 0.5472,
+      "step": 7423
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2230137586593628,
+      "learning_rate": 6.237246964079458e-08,
+      "loss": 0.6058,
+      "step": 7424
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5250579118728638,
+      "learning_rate": 6.204616274055663e-08,
+      "loss": 0.6588,
+      "step": 7425
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.269244909286499,
+      "learning_rate": 6.172070630249993e-08,
+      "loss": 0.5417,
+      "step": 7426
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3151706457138062,
+      "learning_rate": 6.139610038267685e-08,
+      "loss": 0.548,
+      "step": 7427
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.407835602760315,
+      "learning_rate": 6.107234503699488e-08,
+      "loss": 0.6501,
+      "step": 7428
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3132225275039673,
+      "learning_rate": 6.074944032121388e-08,
+      "loss": 0.4961,
+      "step": 7429
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4748717546463013,
+      "learning_rate": 6.042738629094769e-08,
+      "loss": 0.5878,
+      "step": 7430
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.3013241291046143,
+      "learning_rate": 6.010618300166415e-08,
+      "loss": 0.5828,
+      "step": 7431
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3468983173370361,
+      "learning_rate": 5.978583050868292e-08,
+      "loss": 0.6095,
+      "step": 7432
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.144441604614258,
+      "learning_rate": 5.946632886717929e-08,
+      "loss": 0.5951,
+      "step": 7433
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5819791555404663,
+      "learning_rate": 5.914767813217981e-08,
+      "loss": 0.6179,
+      "step": 7434
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.079916000366211,
+      "learning_rate": 5.882987835856668e-08,
+      "loss": 0.6712,
+      "step": 7435
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3036378622055054,
+      "learning_rate": 5.851292960107391e-08,
+      "loss": 0.5571,
+      "step": 7436
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3435218334197998,
+      "learning_rate": 5.8196831914289485e-08,
+      "loss": 0.6113,
+      "step": 7437
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.429049015045166,
+      "learning_rate": 5.788158535265431e-08,
+      "loss": 0.5887,
+      "step": 7438
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3666672706604004,
+      "learning_rate": 5.75671899704644e-08,
+      "loss": 0.6005,
+      "step": 7439
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.053040623664856,
+      "learning_rate": 5.7253645821867546e-08,
+      "loss": 0.4651,
+      "step": 7440
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2168482542037964,
+      "learning_rate": 5.694095296086444e-08,
+      "loss": 0.5997,
+      "step": 7441
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2379217147827148,
+      "learning_rate": 5.662911144131145e-08,
+      "loss": 0.6519,
+      "step": 7442
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.175512671470642,
+      "learning_rate": 5.631812131691561e-08,
+      "loss": 0.5748,
+      "step": 7443
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4115569591522217,
+      "learning_rate": 5.600798264123963e-08,
+      "loss": 0.5639,
+      "step": 7444
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4109700918197632,
+      "learning_rate": 5.569869546769857e-08,
+      "loss": 0.58,
+      "step": 7445
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6279006004333496,
+      "learning_rate": 5.5390259849559813e-08,
+      "loss": 0.5886,
+      "step": 7446
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9307910203933716,
+      "learning_rate": 5.508267583994642e-08,
+      "loss": 0.5608,
+      "step": 7447
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3651392459869385,
+      "learning_rate": 5.4775943491832684e-08,
+      "loss": 0.6131,
+      "step": 7448
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3257839679718018,
+      "learning_rate": 5.4470062858047457e-08,
+      "loss": 0.6027,
+      "step": 7449
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4429939985275269,
+      "learning_rate": 5.4165033991271934e-08,
+      "loss": 0.5893,
+      "step": 7450
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.618682622909546,
+      "learning_rate": 5.3860856944041303e-08,
+      "loss": 0.5665,
+      "step": 7451
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1508351564407349,
+      "learning_rate": 5.3557531768744784e-08,
+      "loss": 0.6654,
+      "step": 7452
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1664789915084839,
+      "learning_rate": 5.32550585176228e-08,
+      "loss": 0.6164,
+      "step": 7453
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3211183547973633,
+      "learning_rate": 5.29534372427698e-08,
+      "loss": 0.61,
+      "step": 7454
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6111810207366943,
+      "learning_rate": 5.2652667996135884e-08,
+      "loss": 0.602,
+      "step": 7455
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.7178215980529785,
+      "learning_rate": 5.2352750829521294e-08,
+      "loss": 0.5855,
+      "step": 7456
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.387904167175293,
+      "learning_rate": 5.205368579458025e-08,
+      "loss": 0.6116,
+      "step": 7457
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 6.927826881408691,
+      "learning_rate": 5.175547294282102e-08,
+      "loss": 0.629,
+      "step": 7458
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.1902652978897095,
+      "learning_rate": 5.145811232560527e-08,
+      "loss": 0.5822,
+      "step": 7459
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.764687180519104,
+      "learning_rate": 5.116160399414649e-08,
+      "loss": 0.6164,
+      "step": 7460
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2136625051498413,
+      "learning_rate": 5.0865947999512166e-08,
+      "loss": 0.5099,
+      "step": 7461
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5262752771377563,
+      "learning_rate": 5.057114439262378e-08,
+      "loss": 0.5656,
+      "step": 7462
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.331688642501831,
+      "learning_rate": 5.0277193224255175e-08,
+      "loss": 0.5583,
+      "step": 7463
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4355744123458862,
+      "learning_rate": 4.9984094545033054e-08,
+      "loss": 0.5641,
+      "step": 7464
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4766184091567993,
+      "learning_rate": 4.969184840543706e-08,
+      "loss": 0.5645,
+      "step": 7465
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.319823980331421,
+      "learning_rate": 4.940045485580247e-08,
+      "loss": 0.5336,
+      "step": 7466
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2531654834747314,
+      "learning_rate": 4.9109913946314145e-08,
+      "loss": 0.6307,
+      "step": 7467
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4141623973846436,
+      "learning_rate": 4.882022572701261e-08,
+      "loss": 0.5306,
+      "step": 7468
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.1841273307800293,
+      "learning_rate": 4.8531390247790745e-08,
+      "loss": 0.5981,
+      "step": 7469
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.325745940208435,
+      "learning_rate": 4.824340755839485e-08,
+      "loss": 0.5935,
+      "step": 7470
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4514849185943604,
+      "learning_rate": 4.795627770842359e-08,
+      "loss": 0.6294,
+      "step": 7471
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2255972623825073,
+      "learning_rate": 4.767000074732908e-08,
+      "loss": 0.5396,
+      "step": 7472
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6118378639221191,
+      "learning_rate": 4.738457672441799e-08,
+      "loss": 0.6553,
+      "step": 7473
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5968915224075317,
+      "learning_rate": 4.710000568884709e-08,
+      "loss": 0.6712,
+      "step": 7474
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.486005425453186,
+      "learning_rate": 4.681628768962887e-08,
+      "loss": 0.6427,
+      "step": 7475
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2320314645767212,
+      "learning_rate": 4.653342277562811e-08,
+      "loss": 0.7184,
+      "step": 7476
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.1806683540344238,
+      "learning_rate": 4.625141099556252e-08,
+      "loss": 0.7453,
+      "step": 7477
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3961061239242554,
+      "learning_rate": 4.597025239800212e-08,
+      "loss": 0.5866,
+      "step": 7478
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8109081983566284,
+      "learning_rate": 4.5689947031371527e-08,
+      "loss": 0.592,
+      "step": 7479
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4340118169784546,
+      "learning_rate": 4.541049494394767e-08,
+      "loss": 0.5356,
+      "step": 7480
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.403652548789978,
+      "learning_rate": 4.513189618385983e-08,
+      "loss": 0.6765,
+      "step": 7481
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5432109832763672,
+      "learning_rate": 4.4854150799091836e-08,
+      "loss": 0.4902,
+      "step": 7482
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5097144842147827,
+      "learning_rate": 4.457725883747932e-08,
+      "loss": 0.6542,
+      "step": 7483
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4052143096923828,
+      "learning_rate": 4.4301220346710806e-08,
+      "loss": 0.5848,
+      "step": 7484
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4552314281463623,
+      "learning_rate": 4.402603537432937e-08,
+      "loss": 0.5903,
+      "step": 7485
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6251884698867798,
+      "learning_rate": 4.3751703967728765e-08,
+      "loss": 0.5735,
+      "step": 7486
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.1430637836456299,
+      "learning_rate": 4.347822617415787e-08,
+      "loss": 0.718,
+      "step": 7487
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0503935813903809,
+      "learning_rate": 4.320560204071733e-08,
+      "loss": 0.6921,
+      "step": 7488
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4668529033660889,
+      "learning_rate": 4.293383161436182e-08,
+      "loss": 0.6366,
+      "step": 7489
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.753491759300232,
+      "learning_rate": 4.2662914941897225e-08,
+      "loss": 0.5208,
+      "step": 7490
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3943778276443481,
+      "learning_rate": 4.239285206998345e-08,
+      "loss": 0.5593,
+      "step": 7491
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3868759870529175,
+      "learning_rate": 4.212364304513439e-08,
+      "loss": 0.6141,
+      "step": 7492
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3776869773864746,
+      "learning_rate": 4.185528791371518e-08,
+      "loss": 0.5684,
+      "step": 7493
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3493459224700928,
+      "learning_rate": 4.158778672194441e-08,
+      "loss": 0.5887,
+      "step": 7494
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.200941801071167,
+      "learning_rate": 4.1321139515894116e-08,
+      "loss": 0.5237,
+      "step": 7495
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.162359595298767,
+      "learning_rate": 4.105534634148867e-08,
+      "loss": 0.5887,
+      "step": 7496
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3749542236328125,
+      "learning_rate": 4.0790407244505335e-08,
+      "loss": 0.5458,
+      "step": 7497
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2350449562072754,
+      "learning_rate": 4.052632227057429e-08,
+      "loss": 0.5914,
+      "step": 7498
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3565806150436401,
+      "learning_rate": 4.02630914651797e-08,
+      "loss": 0.5873,
+      "step": 7499
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3490086793899536,
+      "learning_rate": 4.0000714873656976e-08,
+      "loss": 0.588,
+      "step": 7500
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4163496494293213,
+      "learning_rate": 3.973919254119551e-08,
+      "loss": 0.6271,
+      "step": 7501
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4125710725784302,
+      "learning_rate": 3.947852451283707e-08,
+      "loss": 0.6185,
+      "step": 7502
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3267148733139038,
+      "learning_rate": 3.921871083347628e-08,
+      "loss": 0.6035,
+      "step": 7503
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.564961314201355,
+      "learning_rate": 3.8959751547861244e-08,
+      "loss": 0.6344,
+      "step": 7504
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7320315837860107,
+      "learning_rate": 3.870164670059129e-08,
+      "loss": 0.6125,
+      "step": 7505
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.794987440109253,
+      "learning_rate": 3.844439633612085e-08,
+      "loss": 0.5969,
+      "step": 7506
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.538055181503296,
+      "learning_rate": 3.818800049875615e-08,
+      "loss": 0.582,
+      "step": 7507
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.460828185081482,
+      "learning_rate": 3.793245923265576e-08,
+      "loss": 0.6452,
+      "step": 7508
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2221049070358276,
+      "learning_rate": 3.767777258183114e-08,
+      "loss": 0.5363,
+      "step": 7509
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.714708685874939,
+      "learning_rate": 3.74239405901472e-08,
+      "loss": 0.5413,
+      "step": 7510
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3533588647842407,
+      "learning_rate": 3.717096330132175e-08,
+      "loss": 0.6549,
+      "step": 7511
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2847508192062378,
+      "learning_rate": 3.691884075892438e-08,
+      "loss": 0.5459,
+      "step": 7512
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5215178728103638,
+      "learning_rate": 3.666757300637813e-08,
+      "loss": 0.6188,
+      "step": 7513
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9925087690353394,
+      "learning_rate": 3.6417160086959505e-08,
+      "loss": 0.5738,
+      "step": 7514
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.384124755859375,
+      "learning_rate": 3.616760204379621e-08,
+      "loss": 0.6303,
+      "step": 7515
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3988687992095947,
+      "learning_rate": 3.5918898919869436e-08,
+      "loss": 0.5765,
+      "step": 7516
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3108985424041748,
+      "learning_rate": 3.5671050758013806e-08,
+      "loss": 0.5715,
+      "step": 7517
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2928988933563232,
+      "learning_rate": 3.542405760091627e-08,
+      "loss": 0.6436,
+      "step": 7518
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6511348485946655,
+      "learning_rate": 3.5177919491115595e-08,
+      "loss": 0.6539,
+      "step": 7519
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.209773063659668,
+      "learning_rate": 3.493263647100453e-08,
+      "loss": 0.5657,
+      "step": 7520
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4642279148101807,
+      "learning_rate": 3.468820858282818e-08,
+      "loss": 0.5891,
+      "step": 7521
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4938738346099854,
+      "learning_rate": 3.444463586868452e-08,
+      "loss": 0.5184,
+      "step": 7522
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.524194598197937,
+      "learning_rate": 3.420191837052278e-08,
+      "loss": 0.5887,
+      "step": 7523
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2606900930404663,
+      "learning_rate": 3.3960056130147855e-08,
+      "loss": 0.5932,
+      "step": 7524
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.350730538368225,
+      "learning_rate": 3.37190491892142e-08,
+      "loss": 0.5891,
+      "step": 7525
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3839001655578613,
+      "learning_rate": 3.34788975892314e-08,
+      "loss": 0.5905,
+      "step": 7526
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6060885190963745,
+      "learning_rate": 3.3239601371559685e-08,
+      "loss": 0.5926,
+      "step": 7527
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3335615396499634,
+      "learning_rate": 3.3001160577413873e-08,
+      "loss": 0.6283,
+      "step": 7528
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4750850200653076,
+      "learning_rate": 3.276357524786e-08,
+      "loss": 0.5586,
+      "step": 7529
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4482122659683228,
+      "learning_rate": 3.252684542381701e-08,
+      "loss": 0.589,
+      "step": 7530
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4635100364685059,
+      "learning_rate": 3.229097114605784e-08,
+      "loss": 0.6155,
+      "step": 7531
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.282564640045166,
+      "learning_rate": 3.20559524552061e-08,
+      "loss": 0.6131,
+      "step": 7532
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5637482404708862,
+      "learning_rate": 3.182178939173941e-08,
+      "loss": 0.6289,
+      "step": 7533
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3659104108810425,
+      "learning_rate": 3.1588481995987743e-08,
+      "loss": 0.5984,
+      "step": 7534
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.998304843902588,
+      "learning_rate": 3.1356030308132834e-08,
+      "loss": 0.5942,
+      "step": 7535
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9875431656837463,
+      "learning_rate": 3.1124434368210444e-08,
+      "loss": 0.6936,
+      "step": 7536
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2541671991348267,
+      "learning_rate": 3.089369421610866e-08,
+      "loss": 0.6214,
+      "step": 7537
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.374961018562317,
+      "learning_rate": 3.066380989156625e-08,
+      "loss": 0.5787,
+      "step": 7538
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4336551427841187,
+      "learning_rate": 3.043478143417766e-08,
+      "loss": 0.6287,
+      "step": 7539
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1535704135894775,
+      "learning_rate": 3.0206608883387444e-08,
+      "loss": 0.7096,
+      "step": 7540
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3666244745254517,
+      "learning_rate": 2.997929227849416e-08,
+      "loss": 0.5903,
+      "step": 7541
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4516652822494507,
+      "learning_rate": 2.9752831658648175e-08,
+      "loss": 0.6077,
+      "step": 7542
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3812227249145508,
+      "learning_rate": 2.952722706285327e-08,
+      "loss": 0.6462,
+      "step": 7543
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4420405626296997,
+      "learning_rate": 2.9302478529964484e-08,
+      "loss": 0.6442,
+      "step": 7544
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4087380170822144,
+      "learning_rate": 2.907858609869085e-08,
+      "loss": 0.5694,
+      "step": 7545
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8472365140914917,
+      "learning_rate": 2.8855549807592644e-08,
+      "loss": 0.6193,
+      "step": 7546
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2451483011245728,
+      "learning_rate": 2.8633369695084125e-08,
+      "loss": 0.5944,
+      "step": 7547
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3169342279434204,
+      "learning_rate": 2.84120457994308e-08,
+      "loss": 0.5901,
+      "step": 7548
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3715953826904297,
+      "learning_rate": 2.8191578158751064e-08,
+      "loss": 0.6093,
+      "step": 7549
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.386909008026123,
+      "learning_rate": 2.7971966811016772e-08,
+      "loss": 0.5776,
+      "step": 7550
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.138466715812683,
+      "learning_rate": 2.775321179405044e-08,
+      "loss": 0.59,
+      "step": 7551
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.033170223236084,
+      "learning_rate": 2.7535313145528597e-08,
+      "loss": 0.5274,
+      "step": 7552
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6652302742004395,
+      "learning_rate": 2.7318270902980116e-08,
+      "loss": 0.5159,
+      "step": 7553
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2595553398132324,
+      "learning_rate": 2.7102085103786203e-08,
+      "loss": 0.5573,
+      "step": 7554
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.474131464958191,
+      "learning_rate": 2.688675578517985e-08,
+      "loss": 0.6597,
+      "step": 7555
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.894659399986267,
+      "learning_rate": 2.6672282984248066e-08,
+      "loss": 0.5857,
+      "step": 7556
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4702141284942627,
+      "learning_rate": 2.6458666737927962e-08,
+      "loss": 0.6389,
+      "step": 7557
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2070083618164062,
+      "learning_rate": 2.624590708301178e-08,
+      "loss": 0.6223,
+      "step": 7558
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6800892353057861,
+      "learning_rate": 2.6034004056143535e-08,
+      "loss": 0.5541,
+      "step": 7559
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4756264686584473,
+      "learning_rate": 2.582295769381793e-08,
+      "loss": 0.5985,
+      "step": 7560
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7296569347381592,
+      "learning_rate": 2.5612768032383662e-08,
+      "loss": 0.5712,
+      "step": 7561
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2623050212860107,
+      "learning_rate": 2.540343510804233e-08,
+      "loss": 0.5563,
+      "step": 7562
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2337028980255127,
+      "learning_rate": 2.519495895684676e-08,
+      "loss": 0.6952,
+      "step": 7563
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9761078357696533,
+      "learning_rate": 2.498733961470268e-08,
+      "loss": 0.6188,
+      "step": 7564
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.36776864528656,
+      "learning_rate": 2.47805771173687e-08,
+      "loss": 0.5815,
+      "step": 7565
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5808404684066772,
+      "learning_rate": 2.4574671500455227e-08,
+      "loss": 0.6603,
+      "step": 7566
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1816710233688354,
+      "learning_rate": 2.4369622799425008e-08,
+      "loss": 0.6366,
+      "step": 7567
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2168693542480469,
+      "learning_rate": 2.416543104959368e-08,
+      "loss": 0.593,
+      "step": 7568
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2128366231918335,
+      "learning_rate": 2.3962096286129778e-08,
+      "loss": 0.6183,
+      "step": 7569
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4425259828567505,
+      "learning_rate": 2.375961854405251e-08,
+      "loss": 0.6046,
+      "step": 7570
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.308990716934204,
+      "learning_rate": 2.3557997858235647e-08,
+      "loss": 0.5935,
+      "step": 7571
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4226394891738892,
+      "learning_rate": 2.3357234263403083e-08,
+      "loss": 0.6166,
+      "step": 7572
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.415466070175171,
+      "learning_rate": 2.3157327794133267e-08,
+      "loss": 0.5839,
+      "step": 7573
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2484937906265259,
+      "learning_rate": 2.2958278484855323e-08,
+      "loss": 0.5463,
+      "step": 7574
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1190849542617798,
+      "learning_rate": 2.2760086369851832e-08,
+      "loss": 0.582,
+      "step": 7575
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3897758722305298,
+      "learning_rate": 2.2562751483257706e-08,
+      "loss": 0.5186,
+      "step": 7576
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5645219087600708,
+      "learning_rate": 2.2366273859059095e-08,
+      "loss": 0.5141,
+      "step": 7577
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2559103965759277,
+      "learning_rate": 2.21706535310956e-08,
+      "loss": 0.5594,
+      "step": 7578
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.294405460357666,
+      "learning_rate": 2.197589053305915e-08,
+      "loss": 0.6049,
+      "step": 7579
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4571852684020996,
+      "learning_rate": 2.1781984898492926e-08,
+      "loss": 0.6056,
+      "step": 7580
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2196654081344604,
+      "learning_rate": 2.158893666079409e-08,
+      "loss": 0.5471,
+      "step": 7581
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3833738565444946,
+      "learning_rate": 2.1396745853211054e-08,
+      "loss": 0.5686,
+      "step": 7582
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3073983192443848,
+      "learning_rate": 2.1205412508844557e-08,
+      "loss": 0.662,
+      "step": 7583
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2904523611068726,
+      "learning_rate": 2.1014936660647688e-08,
+      "loss": 0.6103,
+      "step": 7584
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7544934749603271,
+      "learning_rate": 2.0825318341426982e-08,
+      "loss": 0.5528,
+      "step": 7585
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.464415192604065,
+      "learning_rate": 2.063655758383909e-08,
+      "loss": 0.6172,
+      "step": 7586
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.500410795211792,
+      "learning_rate": 2.0448654420395232e-08,
+      "loss": 0.6548,
+      "step": 7587
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2668145895004272,
+      "learning_rate": 2.0261608883457297e-08,
+      "loss": 0.5142,
+      "step": 7588
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5883921384811401,
+      "learning_rate": 2.007542100524007e-08,
+      "loss": 0.5204,
+      "step": 7589
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5936336517333984,
+      "learning_rate": 1.9890090817811237e-08,
+      "loss": 0.5901,
+      "step": 7590
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3055000305175781,
+      "learning_rate": 1.970561835308915e-08,
+      "loss": 0.6495,
+      "step": 7591
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4675624370574951,
+      "learning_rate": 1.952200364284673e-08,
+      "loss": 0.6421,
+      "step": 7592
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.198490023612976,
+      "learning_rate": 1.933924671870646e-08,
+      "loss": 0.6123,
+      "step": 7593
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3107367753982544,
+      "learning_rate": 1.915734761214594e-08,
+      "loss": 0.5849,
+      "step": 7594
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3395153284072876,
+      "learning_rate": 1.8976306354492323e-08,
+      "loss": 0.6503,
+      "step": 7595
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2745283842086792,
+      "learning_rate": 1.8796122976927343e-08,
+      "loss": 0.5744,
+      "step": 7596
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3492122888565063,
+      "learning_rate": 1.8616797510483396e-08,
+      "loss": 0.6037,
+      "step": 7597
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5924208164215088,
+      "learning_rate": 1.8438329986045224e-08,
+      "loss": 0.6054,
+      "step": 7598
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8689661026000977,
+      "learning_rate": 1.8260720434351008e-08,
+      "loss": 0.6192,
+      "step": 7599
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1037673950195312,
+      "learning_rate": 1.8083968885990178e-08,
+      "loss": 0.657,
+      "step": 7600
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0584919452667236,
+      "learning_rate": 1.7908075371404487e-08,
+      "loss": 0.693,
+      "step": 7601
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1882449388504028,
+      "learning_rate": 1.7733039920887484e-08,
+      "loss": 0.5358,
+      "step": 7602
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2444685697555542,
+      "learning_rate": 1.755886256458672e-08,
+      "loss": 0.6336,
+      "step": 7603
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3037827014923096,
+      "learning_rate": 1.738554333249931e-08,
+      "loss": 0.6433,
+      "step": 7604
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2642031908035278,
+      "learning_rate": 1.721308225447693e-08,
+      "loss": 0.5269,
+      "step": 7605
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2133139371871948,
+      "learning_rate": 1.704147936022249e-08,
+      "loss": 0.6345,
+      "step": 7606
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2551854848861694,
+      "learning_rate": 1.687073467929068e-08,
+      "loss": 0.503,
+      "step": 7607
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7916251420974731,
+      "learning_rate": 1.670084824108853e-08,
+      "loss": 0.5754,
+      "step": 7608
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8876594305038452,
+      "learning_rate": 1.6531820074876526e-08,
+      "loss": 0.6057,
+      "step": 7609
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3388862609863281,
+      "learning_rate": 1.636365020976527e-08,
+      "loss": 0.5536,
+      "step": 7610
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.126447081565857,
+      "learning_rate": 1.619633867471937e-08,
+      "loss": 0.5904,
+      "step": 7611
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4953137636184692,
+      "learning_rate": 1.6029885498554108e-08,
+      "loss": 0.6016,
+      "step": 7612
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5743541717529297,
+      "learning_rate": 1.5864290709938778e-08,
+      "loss": 0.6243,
+      "step": 7613
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5609896183013916,
+      "learning_rate": 1.5699554337392786e-08,
+      "loss": 0.6461,
+      "step": 7614
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3598272800445557,
+      "learning_rate": 1.5535676409288435e-08,
+      "loss": 0.6131,
+      "step": 7615
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5744637250900269,
+      "learning_rate": 1.5372656953851482e-08,
+      "loss": 0.5648,
+      "step": 7616
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.973106622695923,
+      "learning_rate": 1.521049599915725e-08,
+      "loss": 0.6274,
+      "step": 7617
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3239871263504028,
+      "learning_rate": 1.5049193573136168e-08,
+      "loss": 0.6198,
+      "step": 7618
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3224722146987915,
+      "learning_rate": 1.4888749703568239e-08,
+      "loss": 0.521,
+      "step": 7619
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.16157865524292,
+      "learning_rate": 1.4729164418087472e-08,
+      "loss": 0.5791,
+      "step": 7620
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2755320072174072,
+      "learning_rate": 1.4570437744177989e-08,
+      "loss": 0.5609,
+      "step": 7621
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4971299171447754,
+      "learning_rate": 1.441256970917848e-08,
+      "loss": 0.5949,
+      "step": 7622
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2407373189926147,
+      "learning_rate": 1.4255560340277753e-08,
+      "loss": 0.5842,
+      "step": 7623
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3337568044662476,
+      "learning_rate": 1.409940966451806e-08,
+      "loss": 0.5259,
+      "step": 7624
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6187931299209595,
+      "learning_rate": 1.3944117708792338e-08,
+      "loss": 0.4729,
+      "step": 7625
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2470555305480957,
+      "learning_rate": 1.3789684499846967e-08,
+      "loss": 0.5533,
+      "step": 7626
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2147272825241089,
+      "learning_rate": 1.3636110064280117e-08,
+      "loss": 0.5493,
+      "step": 7627
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2246286869049072,
+      "learning_rate": 1.348339442854174e-08,
+      "loss": 0.5352,
+      "step": 7628
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2009773254394531,
+      "learning_rate": 1.3331537618934132e-08,
+      "loss": 0.6325,
+      "step": 7629
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.299921989440918,
+      "learning_rate": 1.318053966161137e-08,
+      "loss": 0.4974,
+      "step": 7630
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8579556941986084,
+      "learning_rate": 1.3030400582579317e-08,
+      "loss": 0.5998,
+      "step": 7631
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.709477424621582,
+      "learning_rate": 1.2881120407696734e-08,
+      "loss": 0.6101,
+      "step": 7632
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3852404356002808,
+      "learning_rate": 1.2732699162674721e-08,
+      "loss": 0.537,
+      "step": 7633
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1303093433380127,
+      "learning_rate": 1.2585136873075055e-08,
+      "loss": 0.6092,
+      "step": 7634
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6485886573791504,
+      "learning_rate": 1.2438433564312958e-08,
+      "loss": 0.5874,
+      "step": 7635
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3467371463775635,
+      "learning_rate": 1.2292589261654332e-08,
+      "loss": 0.5349,
+      "step": 7636
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4503722190856934,
+      "learning_rate": 1.214760399021797e-08,
+      "loss": 0.6611,
+      "step": 7637
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5155328512191772,
+      "learning_rate": 1.200347777497557e-08,
+      "loss": 0.489,
+      "step": 7638
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5880452394485474,
+      "learning_rate": 1.1860210640748936e-08,
+      "loss": 0.6368,
+      "step": 7639
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2064660787582397,
+      "learning_rate": 1.1717802612213336e-08,
+      "loss": 0.5961,
+      "step": 7640
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.406825304031372,
+      "learning_rate": 1.1576253713895258e-08,
+      "loss": 0.5817,
+      "step": 7641
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.6012203693389893,
+      "learning_rate": 1.1435563970174645e-08,
+      "loss": 0.594,
+      "step": 7642
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1902316808700562,
+      "learning_rate": 1.1295733405281006e-08,
+      "loss": 0.6312,
+      "step": 7643
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4329363107681274,
+      "learning_rate": 1.1156762043298963e-08,
+      "loss": 0.5915,
+      "step": 7644
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.167524814605713,
+      "learning_rate": 1.1018649908162149e-08,
+      "loss": 0.6717,
+      "step": 7645
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4185863733291626,
+      "learning_rate": 1.0881397023658202e-08,
+      "loss": 0.6306,
+      "step": 7646
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5840102434158325,
+      "learning_rate": 1.0745003413425992e-08,
+      "loss": 0.5838,
+      "step": 7647
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3226847648620605,
+      "learning_rate": 1.0609469100956171e-08,
+      "loss": 0.5167,
+      "step": 7648
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2246990203857422,
+      "learning_rate": 1.0474794109592845e-08,
+      "loss": 0.5422,
+      "step": 7649
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3924264907836914,
+      "learning_rate": 1.0340978462529682e-08,
+      "loss": 0.6216,
+      "step": 7650
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2300564050674438,
+      "learning_rate": 1.0208022182814914e-08,
+      "loss": 0.4941,
+      "step": 7651
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2392950057983398,
+      "learning_rate": 1.0075925293346888e-08,
+      "loss": 0.5609,
+      "step": 7652
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3816360235214233,
+      "learning_rate": 9.944687816876296e-09,
+      "loss": 0.592,
+      "step": 7653
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.271823763847351,
+      "learning_rate": 9.814309776006725e-09,
+      "loss": 0.5909,
+      "step": 7654
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.333675503730774,
+      "learning_rate": 9.684791193193543e-09,
+      "loss": 0.6167,
+      "step": 7655
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.711577296257019,
+      "learning_rate": 9.556132090742797e-09,
+      "loss": 0.5504,
+      "step": 7656
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2807025909423828,
+      "learning_rate": 9.42833249081343e-09,
+      "loss": 0.6422,
+      "step": 7657
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.3574330806732178,
+      "learning_rate": 9.301392415417276e-09,
+      "loss": 0.5771,
+      "step": 7658
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2664991617202759,
+      "learning_rate": 9.175311886416294e-09,
+      "loss": 0.5728,
+      "step": 7659
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3596203327178955,
+      "learning_rate": 9.05009092552478e-09,
+      "loss": 0.5903,
+      "step": 7660
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3723080158233643,
+      "learning_rate": 8.925729554311036e-09,
+      "loss": 0.638,
+      "step": 7661
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.689207673072815,
+      "learning_rate": 8.80222779419293e-09,
+      "loss": 0.6545,
+      "step": 7662
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2211661338806152,
+      "learning_rate": 8.679585666441226e-09,
+      "loss": 0.594,
+      "step": 7663
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2049944400787354,
+      "learning_rate": 8.557803192178471e-09,
+      "loss": 0.5486,
+      "step": 7664
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2144725322723389,
+      "learning_rate": 8.436880392379e-09,
+      "loss": 0.6328,
+      "step": 7665
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2572896480560303,
+      "learning_rate": 8.316817287870039e-09,
+      "loss": 0.6196,
+      "step": 7666
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.505215048789978,
+      "learning_rate": 8.197613899329493e-09,
+      "loss": 0.599,
+      "step": 7667
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4084683656692505,
+      "learning_rate": 8.079270247287607e-09,
+      "loss": 0.5462,
+      "step": 7668
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.439065933227539,
+      "learning_rate": 7.961786352126411e-09,
+      "loss": 0.5704,
+      "step": 7669
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5005910396575928,
+      "learning_rate": 7.845162234081382e-09,
+      "loss": 0.6232,
+      "step": 7670
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1550309658050537,
+      "learning_rate": 7.729397913237013e-09,
+      "loss": 0.5539,
+      "step": 7671
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.173595666885376,
+      "learning_rate": 7.614493409532909e-09,
+      "loss": 0.5558,
+      "step": 7672
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4196832180023193,
+      "learning_rate": 7.500448742757682e-09,
+      "loss": 0.566,
+      "step": 7673
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2495452165603638,
+      "learning_rate": 7.387263932554511e-09,
+      "loss": 0.6306,
+      "step": 7674
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6128349304199219,
+      "learning_rate": 7.274938998415581e-09,
+      "loss": 0.5693,
+      "step": 7675
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.603642702102661,
+      "learning_rate": 7.163473959688194e-09,
+      "loss": 0.5928,
+      "step": 7676
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5382452011108398,
+      "learning_rate": 7.052868835569215e-09,
+      "loss": 0.5689,
+      "step": 7677
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2661441564559937,
+      "learning_rate": 6.943123645107852e-09,
+      "loss": 0.5303,
+      "step": 7678
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5901883840560913,
+      "learning_rate": 6.8342384072056515e-09,
+      "loss": 0.5756,
+      "step": 7679
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.394724726676941,
+      "learning_rate": 6.7262131406165e-09,
+      "loss": 0.6474,
+      "step": 7680
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2976278066635132,
+      "learning_rate": 6.6190478639444056e-09,
+      "loss": 0.5856,
+      "step": 7681
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3308930397033691,
+      "learning_rate": 6.512742595647381e-09,
+      "loss": 0.5754,
+      "step": 7682
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4971197843551636,
+      "learning_rate": 6.407297354034115e-09,
+      "loss": 0.6073,
+      "step": 7683
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3769445419311523,
+      "learning_rate": 6.302712157265634e-09,
+      "loss": 0.584,
+      "step": 7684
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4821816682815552,
+      "learning_rate": 6.198987023353642e-09,
+      "loss": 0.585,
+      "step": 7685
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0962272882461548,
+      "learning_rate": 6.096121970163849e-09,
+      "loss": 0.5703,
+      "step": 7686
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3945462703704834,
+      "learning_rate": 5.994117015412082e-09,
+      "loss": 0.5628,
+      "step": 7687
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.610139012336731,
+      "learning_rate": 5.892972176666512e-09,
+      "loss": 0.6201,
+      "step": 7688
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2614445686340332,
+      "learning_rate": 5.792687471347646e-09,
+      "loss": 0.6168,
+      "step": 7689
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.807807207107544,
+      "learning_rate": 5.693262916726672e-09,
+      "loss": 0.5201,
+      "step": 7690
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3365129232406616,
+      "learning_rate": 5.594698529928777e-09,
+      "loss": 0.5822,
+      "step": 7691
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4841192960739136,
+      "learning_rate": 5.496994327928717e-09,
+      "loss": 0.5317,
+      "step": 7692
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2918846607208252,
+      "learning_rate": 5.4001503275546984e-09,
+      "loss": 0.651,
+      "step": 7693
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.43020761013031,
+      "learning_rate": 5.304166545485046e-09,
+      "loss": 0.6198,
+      "step": 7694
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3098740577697754,
+      "learning_rate": 5.209042998252645e-09,
+      "loss": 0.5555,
+      "step": 7695
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4694414138793945,
+      "learning_rate": 5.114779702238837e-09,
+      "loss": 0.6245,
+      "step": 7696
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2974458932876587,
+      "learning_rate": 5.021376673680078e-09,
+      "loss": 0.6221,
+      "step": 7697
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.057411789894104,
+      "learning_rate": 4.928833928661836e-09,
+      "loss": 0.6633,
+      "step": 7698
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.225907325744629,
+      "learning_rate": 4.837151483123581e-09,
+      "loss": 0.5786,
+      "step": 7699
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4693338871002197,
+      "learning_rate": 4.746329352856016e-09,
+      "loss": 0.5585,
+      "step": 7700
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.21501624584198,
+      "learning_rate": 4.6563675534999635e-09,
+      "loss": 0.5666,
+      "step": 7701
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2666512727737427,
+      "learning_rate": 4.567266100550805e-09,
+      "loss": 0.5707,
+      "step": 7702
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1822082996368408,
+      "learning_rate": 4.4790250093545984e-09,
+      "loss": 0.474,
+      "step": 7703
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4030053615570068,
+      "learning_rate": 4.3916442951080775e-09,
+      "loss": 0.5502,
+      "step": 7704
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.361227035522461,
+      "learning_rate": 4.305123972861424e-09,
+      "loss": 0.5675,
+      "step": 7705
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.310667634010315,
+      "learning_rate": 4.219464057516054e-09,
+      "loss": 0.6493,
+      "step": 7706
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2178518772125244,
+      "learning_rate": 4.1346645638246084e-09,
+      "loss": 0.5146,
+      "step": 7707
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.443044900894165,
+      "learning_rate": 4.0507255063926275e-09,
+      "loss": 0.6149,
+      "step": 7708
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3189752101898193,
+      "learning_rate": 3.96764689967688e-09,
+      "loss": 0.5632,
+      "step": 7709
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.260591983795166,
+      "learning_rate": 3.885428757985366e-09,
+      "loss": 0.5505,
+      "step": 7710
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4206947088241577,
+      "learning_rate": 3.804071095479533e-09,
+      "loss": 0.614,
+      "step": 7711
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2580530643463135,
+      "learning_rate": 3.7235739261703983e-09,
+      "loss": 0.5879,
+      "step": 7712
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.525988221168518,
+      "learning_rate": 3.643937263922981e-09,
+      "loss": 0.6059,
+      "step": 7713
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3162834644317627,
+      "learning_rate": 3.5651611224524207e-09,
+      "loss": 0.6687,
+      "step": 7714
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2362546920776367,
+      "learning_rate": 3.4872455153267536e-09,
+      "loss": 0.6078,
+      "step": 7715
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3531043529510498,
+      "learning_rate": 3.41019045596469e-09,
+      "loss": 0.5203,
+      "step": 7716
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5174033641815186,
+      "learning_rate": 3.3339959576378365e-09,
+      "loss": 0.5899,
+      "step": 7717
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8559000492095947,
+      "learning_rate": 3.258662033469584e-09,
+      "loss": 0.5911,
+      "step": 7718
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.32603919506073,
+      "learning_rate": 3.1841886964339987e-09,
+      "loss": 0.6394,
+      "step": 7719
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5879665613174438,
+      "learning_rate": 3.110575959358042e-09,
+      "loss": 0.59,
+      "step": 7720
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3852684497833252,
+      "learning_rate": 3.037823834919906e-09,
+      "loss": 0.5394,
+      "step": 7721
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2205983400344849,
+      "learning_rate": 2.9659323356490134e-09,
+      "loss": 0.7311,
+      "step": 7722
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.312951922416687,
+      "learning_rate": 2.8949014739282354e-09,
+      "loss": 0.4707,
+      "step": 7723
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5596814155578613,
+      "learning_rate": 2.8247312619905653e-09,
+      "loss": 0.606,
+      "step": 7724
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2269487380981445,
+      "learning_rate": 2.7554217119213354e-09,
+      "loss": 0.5177,
+      "step": 7725
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2579458951950073,
+      "learning_rate": 2.6869728356582192e-09,
+      "loss": 0.55,
+      "step": 7726
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6828994750976562,
+      "learning_rate": 2.6193846449901193e-09,
+      "loss": 0.6079,
+      "step": 7727
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.452315330505371,
+      "learning_rate": 2.5526571515571696e-09,
+      "loss": 0.5791,
+      "step": 7728
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2305513620376587,
+      "learning_rate": 2.4867903668518433e-09,
+      "loss": 0.5592,
+      "step": 7729
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.548684000968933,
+      "learning_rate": 2.4217843022189546e-09,
+      "loss": 0.4734,
+      "step": 7730
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.631661295890808,
+      "learning_rate": 2.3576389688539924e-09,
+      "loss": 0.5776,
+      "step": 7731
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.554509162902832,
+      "learning_rate": 2.294354377804786e-09,
+      "loss": 0.604,
+      "step": 7732
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5531501770019531,
+      "learning_rate": 2.2319305399715054e-09,
+      "loss": 0.5192,
+      "step": 7733
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3193473815917969,
+      "learning_rate": 2.170367466103884e-09,
+      "loss": 0.5963,
+      "step": 7734
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.141550898551941,
+      "learning_rate": 2.1096651668062184e-09,
+      "loss": 0.5749,
+      "step": 7735
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7103443145751953,
+      "learning_rate": 2.0498236525329228e-09,
+      "loss": 0.5586,
+      "step": 7736
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4387218952178955,
+      "learning_rate": 1.990842933590198e-09,
+      "loss": 0.5993,
+      "step": 7737
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.195109486579895,
+      "learning_rate": 1.932723020136584e-09,
+      "loss": 0.5556,
+      "step": 7738
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2314635515213013,
+      "learning_rate": 1.875463922181853e-09,
+      "loss": 0.5436,
+      "step": 7739
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4578646421432495,
+      "learning_rate": 1.8190656495875591e-09,
+      "loss": 0.6199,
+      "step": 7740
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4149136543273926,
+      "learning_rate": 1.7635282120676e-09,
+      "loss": 0.5495,
+      "step": 7741
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.971264362335205,
+      "learning_rate": 1.7088516191871018e-09,
+      "loss": 0.6472,
+      "step": 7742
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5405484437942505,
+      "learning_rate": 1.655035880362421e-09,
+      "loss": 0.5643,
+      "step": 7743
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.398697853088379,
+      "learning_rate": 1.602081004863365e-09,
+      "loss": 0.5151,
+      "step": 7744
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.603548288345337,
+      "learning_rate": 1.5499870018093054e-09,
+      "loss": 0.5618,
+      "step": 7745
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.4291934967041016,
+      "learning_rate": 1.49875388017251e-09,
+      "loss": 0.7038,
+      "step": 7746
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4780473709106445,
+      "learning_rate": 1.4483816487770309e-09,
+      "loss": 0.6409,
+      "step": 7747
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2052099704742432,
+      "learning_rate": 1.398870316298151e-09,
+      "loss": 0.5907,
+      "step": 7748
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4878346920013428,
+      "learning_rate": 1.3502198912640484e-09,
+      "loss": 0.649,
+      "step": 7749
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4148633480072021,
+      "learning_rate": 1.3024303820530216e-09,
+      "loss": 0.5942,
+      "step": 7750
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2521181106567383,
+      "learning_rate": 1.2555017968962635e-09,
+      "loss": 0.5044,
+      "step": 7751
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4545745849609375,
+      "learning_rate": 1.209434143875643e-09,
+      "loss": 0.5946,
+      "step": 7752
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1492974758148193,
+      "learning_rate": 1.1642274309259238e-09,
+      "loss": 0.6968,
+      "step": 7753
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2396249771118164,
+      "learning_rate": 1.1198816658331003e-09,
+      "loss": 0.4695,
+      "step": 7754
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2736849784851074,
+      "learning_rate": 1.0763968562343963e-09,
+      "loss": 0.5964,
+      "step": 7755
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.6767706871032715,
+      "learning_rate": 1.0337730096199316e-09,
+      "loss": 0.567,
+      "step": 7756
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1457759141921997,
+      "learning_rate": 9.920101333293907e-10,
+      "loss": 0.4591,
+      "step": 7757
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.395374059677124,
+      "learning_rate": 9.511082345570189e-10,
+      "loss": 0.5131,
+      "step": 7758
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.0349416732788086,
+      "learning_rate": 9.110673203471809e-10,
+      "loss": 0.7146,
+      "step": 7759
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1731163263320923,
+      "learning_rate": 8.718873975949171e-10,
+      "loss": 0.6224,
+      "step": 7760
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4639376401901245,
+      "learning_rate": 8.335684730492733e-10,
+      "loss": 0.6381,
+      "step": 7761
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3412548303604126,
+      "learning_rate": 7.9611055330997e-10,
+      "loss": 0.5763,
+      "step": 7762
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3731943368911743,
+      "learning_rate": 7.595136448274032e-10,
+      "loss": 0.5907,
+      "step": 7763
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.477068305015564,
+      "learning_rate": 7.237777539059743e-10,
+      "loss": 0.6016,
+      "step": 7764
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.240449070930481,
+      "learning_rate": 6.889028866990943e-10,
+      "loss": 0.5382,
+      "step": 7765
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.313112735748291,
+      "learning_rate": 6.548890492141802e-10,
+      "loss": 0.5249,
+      "step": 7766
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.2798494100570679,
+      "learning_rate": 6.217362473093235e-10,
+      "loss": 0.6216,
+      "step": 7767
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.7569220066070557,
+      "learning_rate": 5.894444866938465e-10,
+      "loss": 0.602,
+      "step": 7768
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4768586158752441,
+      "learning_rate": 5.580137729299661e-10,
+      "loss": 0.6307,
+      "step": 7769
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.2594913244247437,
+      "learning_rate": 5.274441114305751e-10,
+      "loss": 0.562,
+      "step": 7770
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.1919431686401367,
+      "learning_rate": 4.977355074609059e-10,
+      "loss": 0.5843,
+      "step": 7771
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5836091041564941,
+      "learning_rate": 4.688879661379764e-10,
+      "loss": 0.601,
+      "step": 7772
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3357963562011719,
+      "learning_rate": 4.409014924294797e-10,
+      "loss": 0.5608,
+      "step": 7773
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4678525924682617,
+      "learning_rate": 4.1377609115600403e-10,
+      "loss": 0.615,
+      "step": 7774
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.2490729093551636,
+      "learning_rate": 3.8751176698936797e-10,
+      "loss": 0.5467,
+      "step": 7775
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4261091947555542,
+      "learning_rate": 3.621085244526201e-10,
+      "loss": 0.5542,
+      "step": 7776
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5678194761276245,
+      "learning_rate": 3.375663679211494e-10,
+      "loss": 0.5603,
+      "step": 7777
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3209941387176514,
+      "learning_rate": 3.138853016221299e-10,
+      "loss": 0.5321,
+      "step": 7778
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.2347135543823242,
+      "learning_rate": 2.9106532963396605e-10,
+      "loss": 0.7119,
+      "step": 7779
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4404094219207764,
+      "learning_rate": 2.6910645588684726e-10,
+      "loss": 0.5572,
+      "step": 7780
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.1381596326828003,
+      "learning_rate": 2.480086841633034e-10,
+      "loss": 0.6961,
+      "step": 7781
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4806857109069824,
+      "learning_rate": 2.2777201809598415e-10,
+      "loss": 0.6237,
+      "step": 7782
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4877007007598877,
+      "learning_rate": 2.0839646117098988e-10,
+      "loss": 0.5864,
+      "step": 7783
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6828969717025757,
+      "learning_rate": 1.8988201672509587e-10,
+      "loss": 0.6237,
+      "step": 7784
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.460888385772705,
+      "learning_rate": 1.7222868794686264e-10,
+      "loss": 0.5588,
+      "step": 7785
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.2123541831970215,
+      "learning_rate": 1.5543647787719108e-10,
+      "loss": 0.5466,
+      "step": 7786
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.8442035913467407,
+      "learning_rate": 1.39505389407657e-10,
+      "loss": 0.5387,
+      "step": 7787
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.310922384262085,
+      "learning_rate": 1.2443542528273178e-10,
+      "loss": 0.5123,
+      "step": 7788
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.1582748889923096,
+      "learning_rate": 1.102265880975617e-10,
+      "loss": 0.6051,
+      "step": 7789
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.2169848680496216,
+      "learning_rate": 9.687888029907832e-11,
+      "loss": 0.5885,
+      "step": 7790
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.2619993686676025,
+      "learning_rate": 8.439230418655353e-11,
+      "loss": 0.5656,
+      "step": 7791
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3111398220062256,
+      "learning_rate": 7.27668619099342e-11,
+      "loss": 0.5407,
+      "step": 7792
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3473933935165405,
+      "learning_rate": 6.20025554726178e-11,
+      "loss": 0.575,
+      "step": 7793
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3455848693847656,
+      "learning_rate": 5.2099386727566535e-11,
+      "loss": 0.547,
+      "step": 7794
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5748672485351562,
+      "learning_rate": 4.305735738008299e-11,
+      "loss": 0.5743,
+      "step": 7795
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4658571481704712,
+      "learning_rate": 3.4876468988920275e-11,
+      "loss": 0.5293,
+      "step": 7796
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.108062505722046,
+      "learning_rate": 2.7556722961841197e-11,
+      "loss": 0.6937,
+      "step": 7797
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.1840686798095703,
+      "learning_rate": 2.1098120559503998e-11,
+      "loss": 0.476,
+      "step": 7798
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.324550747871399,
+      "learning_rate": 1.550066289490726e-11,
+      "loss": 0.5883,
+      "step": 7799
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7365845441818237,
+      "learning_rate": 1.0764350932279676e-11,
+      "loss": 0.5685,
+      "step": 7800
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3814128637313843,
+      "learning_rate": 6.889185486524952e-12,
+      "loss": 0.5706,
+      "step": 7801
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.1768162250518799,
+      "learning_rate": 3.875167225442234e-12,
+      "loss": 0.5444,
+      "step": 7802
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.977554440498352,
+      "learning_rate": 1.7222966680607856e-12,
+      "loss": 0.5997,
+      "step": 7803
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.1297414302825928,
+      "learning_rate": 4.3057418575020993e-13,
+      "loss": 0.6249,
+      "step": 7804
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.294604778289795,
+      "learning_rate": 0.0,
+      "loss": 0.5606,
+      "step": 7805
+    },
+    {
+      "epoch": 1.0,
+      "step": 7805,
+      "total_flos": 1.1237568065308721e+19,
+      "train_loss": 0.6154153387490028,
+      "train_runtime": 23859.4346,
+      "train_samples_per_second": 41.876,
+      "train_steps_per_second": 0.327
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 7805,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 1000,
+  "total_flos": 1.1237568065308721e+19,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}