diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,31173 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.9999438107546216,
+  "eval_steps": 500,
+  "global_step": 4449,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 9.353738797874001,
+      "learning_rate": 1.4925373134328358e-07,
+      "loss": 1.2766,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 11.183272720172308,
+      "learning_rate": 2.9850746268656716e-07,
+      "loss": 1.2559,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 11.528741582918375,
+      "learning_rate": 4.4776119402985074e-07,
+      "loss": 1.2454,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 15.994595200990059,
+      "learning_rate": 5.970149253731343e-07,
+      "loss": 1.7868,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 12.82650277228402,
+      "learning_rate": 7.462686567164179e-07,
+      "loss": 1.3556,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 15.418135582750546,
+      "learning_rate": 8.955223880597015e-07,
+      "loss": 1.814,
+      "step": 6
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 11.127359244076684,
+      "learning_rate": 1.044776119402985e-06,
+      "loss": 1.2734,
+      "step": 7
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 14.776442973948203,
+      "learning_rate": 1.1940298507462686e-06,
+      "loss": 1.8289,
+      "step": 8
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 15.282924117789555,
+      "learning_rate": 1.3432835820895524e-06,
+      "loss": 1.8412,
+      "step": 9
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 9.70332166038233,
+      "learning_rate": 1.4925373134328358e-06,
+      "loss": 1.3273,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 10.013780509725219,
+      "learning_rate": 1.6417910447761196e-06,
+      "loss": 1.198,
+      "step": 11
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 14.92248452278362,
+      "learning_rate": 1.791044776119403e-06,
+      "loss": 1.3771,
+      "step": 12
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 14.106753432474498,
+      "learning_rate": 1.9402985074626867e-06,
+      "loss": 1.7538,
+      "step": 13
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 9.269256398686611,
+      "learning_rate": 2.08955223880597e-06,
+      "loss": 1.2502,
+      "step": 14
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.434164246732801,
+      "learning_rate": 2.238805970149254e-06,
+      "loss": 1.2617,
+      "step": 15
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.631271892641452,
+      "learning_rate": 2.3880597014925373e-06,
+      "loss": 1.1964,
+      "step": 16
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.72954602230216,
+      "learning_rate": 2.537313432835821e-06,
+      "loss": 1.6419,
+      "step": 17
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.657338319624321,
+      "learning_rate": 2.686567164179105e-06,
+      "loss": 1.0472,
+      "step": 18
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.026569908685336,
+      "learning_rate": 2.835820895522388e-06,
+      "loss": 1.1737,
+      "step": 19
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.191607935888906,
+      "learning_rate": 2.9850746268656716e-06,
+      "loss": 1.1611,
+      "step": 20
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.0739886772598135,
+      "learning_rate": 3.1343283582089558e-06,
+      "loss": 1.5597,
+      "step": 21
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.1841655581156396,
+      "learning_rate": 3.283582089552239e-06,
+      "loss": 1.2026,
+      "step": 22
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.624809025375175,
+      "learning_rate": 3.4328358208955225e-06,
+      "loss": 1.6965,
+      "step": 23
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 11.33898906963253,
+      "learning_rate": 3.582089552238806e-06,
+      "loss": 1.7503,
+      "step": 24
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.122259779145581,
+      "learning_rate": 3.73134328358209e-06,
+      "loss": 1.0597,
+      "step": 25
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.124054841484952,
+      "learning_rate": 3.8805970149253735e-06,
+      "loss": 1.0722,
+      "step": 26
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.628511215374613,
+      "learning_rate": 4.029850746268657e-06,
+      "loss": 1.0317,
+      "step": 27
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.507735735914965,
+      "learning_rate": 4.17910447761194e-06,
+      "loss": 1.0988,
+      "step": 28
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.964717496774229,
+      "learning_rate": 4.3283582089552236e-06,
+      "loss": 1.1735,
+      "step": 29
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.764149418346652,
+      "learning_rate": 4.477611940298508e-06,
+      "loss": 1.0821,
+      "step": 30
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.3589005996535093,
+      "learning_rate": 4.626865671641791e-06,
+      "loss": 1.0154,
+      "step": 31
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.969542209983452,
+      "learning_rate": 4.7761194029850745e-06,
+      "loss": 0.9672,
+      "step": 32
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.385762664323885,
+      "learning_rate": 4.925373134328359e-06,
+      "loss": 1.5544,
+      "step": 33
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2514658238001197,
+      "learning_rate": 5.074626865671642e-06,
+      "loss": 0.9866,
+      "step": 34
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.109808313771725,
+      "learning_rate": 5.2238805970149255e-06,
+      "loss": 1.0059,
+      "step": 35
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.4176567119672274,
+      "learning_rate": 5.37313432835821e-06,
+      "loss": 1.005,
+      "step": 36
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.5748018501673107,
+      "learning_rate": 5.522388059701493e-06,
+      "loss": 1.6238,
+      "step": 37
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.401470736932171,
+      "learning_rate": 5.671641791044776e-06,
+      "loss": 1.1222,
+      "step": 38
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0746097452894663,
+      "learning_rate": 5.820895522388061e-06,
+      "loss": 1.0359,
+      "step": 39
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.550918766600462,
+      "learning_rate": 5.970149253731343e-06,
+      "loss": 1.1137,
+      "step": 40
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.508137881187459,
+      "learning_rate": 6.119402985074627e-06,
+      "loss": 1.0173,
+      "step": 41
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.1564070528416046,
+      "learning_rate": 6.2686567164179116e-06,
+      "loss": 1.5195,
+      "step": 42
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.6456048274435897,
+      "learning_rate": 6.417910447761194e-06,
+      "loss": 1.015,
+      "step": 43
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.6182736297857216,
+      "learning_rate": 6.567164179104478e-06,
+      "loss": 1.0014,
+      "step": 44
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.4623334789865337,
+      "learning_rate": 6.7164179104477625e-06,
+      "loss": 1.4785,
+      "step": 45
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.221632310789277,
+      "learning_rate": 6.865671641791045e-06,
+      "loss": 0.9625,
+      "step": 46
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1688934758008105,
+      "learning_rate": 7.014925373134329e-06,
+      "loss": 1.039,
+      "step": 47
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.8716217272625415,
+      "learning_rate": 7.164179104477612e-06,
+      "loss": 1.5037,
+      "step": 48
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7272082851053392,
+      "learning_rate": 7.313432835820896e-06,
+      "loss": 1.0647,
+      "step": 49
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.992642580454682,
+      "learning_rate": 7.46268656716418e-06,
+      "loss": 0.9781,
+      "step": 50
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9240694755080805,
+      "learning_rate": 7.611940298507463e-06,
+      "loss": 0.9374,
+      "step": 51
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.6326619399362063,
+      "learning_rate": 7.761194029850747e-06,
+      "loss": 1.5449,
+      "step": 52
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.7429887514346842,
+      "learning_rate": 7.91044776119403e-06,
+      "loss": 1.5259,
+      "step": 53
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9874767528259614,
+      "learning_rate": 8.059701492537314e-06,
+      "loss": 0.9578,
+      "step": 54
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.4181325641653313,
+      "learning_rate": 8.208955223880599e-06,
+      "loss": 1.4785,
+      "step": 55
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8707359426869592,
+      "learning_rate": 8.35820895522388e-06,
+      "loss": 0.8895,
+      "step": 56
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.545834287418945,
+      "learning_rate": 8.507462686567165e-06,
+      "loss": 1.6113,
+      "step": 57
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0316554347995317,
+      "learning_rate": 8.656716417910447e-06,
+      "loss": 0.9329,
+      "step": 58
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.425533327055374,
+      "learning_rate": 8.805970149253732e-06,
+      "loss": 1.5362,
+      "step": 59
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9175610340093707,
+      "learning_rate": 8.955223880597016e-06,
+      "loss": 0.9067,
+      "step": 60
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.893152957817381,
+      "learning_rate": 9.104477611940299e-06,
+      "loss": 0.9214,
+      "step": 61
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8060326852076096,
+      "learning_rate": 9.253731343283582e-06,
+      "loss": 0.9692,
+      "step": 62
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7486709585798619,
+      "learning_rate": 9.402985074626867e-06,
+      "loss": 0.9569,
+      "step": 63
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8538482074140394,
+      "learning_rate": 9.552238805970149e-06,
+      "loss": 0.99,
+      "step": 64
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.199828237051907,
+      "learning_rate": 9.701492537313434e-06,
+      "loss": 1.5114,
+      "step": 65
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3776375351428816,
+      "learning_rate": 9.850746268656717e-06,
+      "loss": 1.49,
+      "step": 66
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8363407100398825,
+      "learning_rate": 1e-05,
+      "loss": 0.9895,
+      "step": 67
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8916896029513484,
+      "learning_rate": 1.0149253731343284e-05,
+      "loss": 0.8786,
+      "step": 68
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.859427836878187,
+      "learning_rate": 1.029850746268657e-05,
+      "loss": 0.9429,
+      "step": 69
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7052830586759165,
+      "learning_rate": 1.0447761194029851e-05,
+      "loss": 0.9661,
+      "step": 70
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.4201982240994147,
+      "learning_rate": 1.0597014925373134e-05,
+      "loss": 1.4883,
+      "step": 71
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1037446260519985,
+      "learning_rate": 1.074626865671642e-05,
+      "loss": 1.4746,
+      "step": 72
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.904728898907631,
+      "learning_rate": 1.0895522388059703e-05,
+      "loss": 0.9234,
+      "step": 73
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6271200966246913,
+      "learning_rate": 1.1044776119402986e-05,
+      "loss": 0.9468,
+      "step": 74
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6764726306222137,
+      "learning_rate": 1.1194029850746268e-05,
+      "loss": 0.9025,
+      "step": 75
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8482782650135992,
+      "learning_rate": 1.1343283582089553e-05,
+      "loss": 1.0765,
+      "step": 76
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.205478039937708,
+      "learning_rate": 1.1492537313432836e-05,
+      "loss": 1.5861,
+      "step": 77
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7415825277638997,
+      "learning_rate": 1.1641791044776121e-05,
+      "loss": 0.924,
+      "step": 78
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.841673981319671,
+      "learning_rate": 1.1791044776119405e-05,
+      "loss": 0.9453,
+      "step": 79
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8153390054508702,
+      "learning_rate": 1.1940298507462686e-05,
+      "loss": 1.0236,
+      "step": 80
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7995520188309058,
+      "learning_rate": 1.2089552238805971e-05,
+      "loss": 1.05,
+      "step": 81
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5890752580584069,
+      "learning_rate": 1.2238805970149255e-05,
+      "loss": 0.9864,
+      "step": 82
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7086803775206472,
+      "learning_rate": 1.2388059701492538e-05,
+      "loss": 0.9468,
+      "step": 83
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6722898941514752,
+      "learning_rate": 1.2537313432835823e-05,
+      "loss": 0.8375,
+      "step": 84
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.670915115730752,
+      "learning_rate": 1.2686567164179107e-05,
+      "loss": 0.9129,
+      "step": 85
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5833078004277972,
+      "learning_rate": 1.2835820895522388e-05,
+      "loss": 0.8428,
+      "step": 86
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6971024913381227,
+      "learning_rate": 1.2985074626865673e-05,
+      "loss": 0.9603,
+      "step": 87
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.555401516986832,
+      "learning_rate": 1.3134328358208957e-05,
+      "loss": 0.8963,
+      "step": 88
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6591626256879006,
+      "learning_rate": 1.328358208955224e-05,
+      "loss": 0.814,
+      "step": 89
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.66876765748716,
+      "learning_rate": 1.3432835820895525e-05,
+      "loss": 0.8925,
+      "step": 90
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5504457756821457,
+      "learning_rate": 1.3582089552238807e-05,
+      "loss": 0.8955,
+      "step": 91
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6234379296645964,
+      "learning_rate": 1.373134328358209e-05,
+      "loss": 0.9132,
+      "step": 92
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6655690149019802,
+      "learning_rate": 1.3880597014925375e-05,
+      "loss": 0.8678,
+      "step": 93
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6271177002042638,
+      "learning_rate": 1.4029850746268658e-05,
+      "loss": 0.9021,
+      "step": 94
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.920313898124087,
+      "learning_rate": 1.4179104477611942e-05,
+      "loss": 1.5144,
+      "step": 95
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5306414225077378,
+      "learning_rate": 1.4328358208955224e-05,
+      "loss": 0.8273,
+      "step": 96
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.4822208404607204,
+      "learning_rate": 1.4477611940298509e-05,
+      "loss": 1.5335,
+      "step": 97
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6053810361879495,
+      "learning_rate": 1.4626865671641792e-05,
+      "loss": 0.8596,
+      "step": 98
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.669245865263158,
+      "learning_rate": 1.4776119402985077e-05,
+      "loss": 0.9033,
+      "step": 99
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8514240552317922,
+      "learning_rate": 1.492537313432836e-05,
+      "loss": 0.9357,
+      "step": 100
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1250782292222476,
+      "learning_rate": 1.5074626865671642e-05,
+      "loss": 1.5,
+      "step": 101
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2291206412944353,
+      "learning_rate": 1.5223880597014925e-05,
+      "loss": 1.5493,
+      "step": 102
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7828795585543737,
+      "learning_rate": 1.537313432835821e-05,
+      "loss": 0.944,
+      "step": 103
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1704659101256056,
+      "learning_rate": 1.5522388059701494e-05,
+      "loss": 1.3728,
+      "step": 104
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5242445507723519,
+      "learning_rate": 1.5671641791044777e-05,
+      "loss": 0.8809,
+      "step": 105
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.702335207248128,
+      "learning_rate": 1.582089552238806e-05,
+      "loss": 0.9362,
+      "step": 106
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5968553192783854,
+      "learning_rate": 1.5970149253731344e-05,
+      "loss": 0.897,
+      "step": 107
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8508732149977725,
+      "learning_rate": 1.6119402985074627e-05,
+      "loss": 0.9702,
+      "step": 108
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7373969107474254,
+      "learning_rate": 1.626865671641791e-05,
+      "loss": 0.9376,
+      "step": 109
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6186065995059449,
+      "learning_rate": 1.6417910447761197e-05,
+      "loss": 0.7478,
+      "step": 110
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8244183762501032,
+      "learning_rate": 1.6567164179104477e-05,
+      "loss": 0.9407,
+      "step": 111
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8435798185767667,
+      "learning_rate": 1.671641791044776e-05,
+      "loss": 0.9727,
+      "step": 112
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.679711688747193,
+      "learning_rate": 1.6865671641791048e-05,
+      "loss": 0.9599,
+      "step": 113
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6684655623186486,
+      "learning_rate": 1.701492537313433e-05,
+      "loss": 0.844,
+      "step": 114
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7602489892764615,
+      "learning_rate": 1.7164179104477614e-05,
+      "loss": 0.9091,
+      "step": 115
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.240674733646385,
+      "learning_rate": 1.7313432835820894e-05,
+      "loss": 1.5089,
+      "step": 116
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5339871903268083,
+      "learning_rate": 1.746268656716418e-05,
+      "loss": 0.8261,
+      "step": 117
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2000031028752023,
+      "learning_rate": 1.7611940298507464e-05,
+      "loss": 1.51,
+      "step": 118
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5978397443835444,
+      "learning_rate": 1.7761194029850748e-05,
+      "loss": 0.8534,
+      "step": 119
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2450332317433452,
+      "learning_rate": 1.791044776119403e-05,
+      "loss": 1.5038,
+      "step": 120
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5647891124425628,
+      "learning_rate": 1.8059701492537314e-05,
+      "loss": 0.8202,
+      "step": 121
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5859778183825153,
+      "learning_rate": 1.8208955223880598e-05,
+      "loss": 0.8519,
+      "step": 122
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.165975344764788,
+      "learning_rate": 1.835820895522388e-05,
+      "loss": 1.425,
+      "step": 123
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.3051725032099175,
+      "learning_rate": 1.8507462686567165e-05,
+      "loss": 1.5849,
+      "step": 124
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.167246701626067,
+      "learning_rate": 1.865671641791045e-05,
+      "loss": 1.4229,
+      "step": 125
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7911710602633961,
+      "learning_rate": 1.8805970149253735e-05,
+      "loss": 0.8374,
+      "step": 126
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8385498533793665,
+      "learning_rate": 1.8955223880597015e-05,
+      "loss": 0.9174,
+      "step": 127
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6771833072131288,
+      "learning_rate": 1.9104477611940298e-05,
+      "loss": 0.9129,
+      "step": 128
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7178183067382597,
+      "learning_rate": 1.9253731343283585e-05,
+      "loss": 0.841,
+      "step": 129
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.835644285977225,
+      "learning_rate": 1.9402985074626868e-05,
+      "loss": 0.8891,
+      "step": 130
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5161603388714724,
+      "learning_rate": 1.955223880597015e-05,
+      "loss": 0.7764,
+      "step": 131
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5847226644027719,
+      "learning_rate": 1.9701492537313435e-05,
+      "loss": 0.8483,
+      "step": 132
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.3883283494178036,
+      "learning_rate": 1.9850746268656718e-05,
+      "loss": 1.4191,
+      "step": 133
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7056140448600479,
+      "learning_rate": 2e-05,
+      "loss": 0.9219,
+      "step": 134
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.3335827399391977,
+      "learning_rate": 1.9999997349620096e-05,
+      "loss": 1.5551,
+      "step": 135
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6220855973878052,
+      "learning_rate": 1.999998939848178e-05,
+      "loss": 0.9057,
+      "step": 136
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6046509104524294,
+      "learning_rate": 1.9999976146589266e-05,
+      "loss": 0.8768,
+      "step": 137
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5956479713050493,
+      "learning_rate": 1.9999957593949586e-05,
+      "loss": 0.9073,
+      "step": 138
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6935803108447711,
+      "learning_rate": 1.999993374057257e-05,
+      "loss": 0.7542,
+      "step": 139
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6875811123362041,
+      "learning_rate": 1.9999904586470863e-05,
+      "loss": 0.8305,
+      "step": 140
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2630278894466374,
+      "learning_rate": 1.9999870131659914e-05,
+      "loss": 1.479,
+      "step": 141
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2672742262005916,
+      "learning_rate": 1.9999830376157994e-05,
+      "loss": 1.4951,
+      "step": 142
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6093841803315025,
+      "learning_rate": 1.9999785319986174e-05,
+      "loss": 0.8942,
+      "step": 143
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6187840634431574,
+      "learning_rate": 1.9999734963168334e-05,
+      "loss": 0.818,
+      "step": 144
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5763727015588564,
+      "learning_rate": 1.9999679305731172e-05,
+      "loss": 0.8739,
+      "step": 145
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9207268590364799,
+      "learning_rate": 1.9999618347704186e-05,
+      "loss": 0.8565,
+      "step": 146
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6045679276567968,
+      "learning_rate": 1.9999552089119693e-05,
+      "loss": 0.8164,
+      "step": 147
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.3262028470721234,
+      "learning_rate": 1.999948053001281e-05,
+      "loss": 1.5052,
+      "step": 148
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.3471173318987733,
+      "learning_rate": 1.9999403670421477e-05,
+      "loss": 1.4562,
+      "step": 149
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6259683853824725,
+      "learning_rate": 1.9999321510386425e-05,
+      "loss": 0.7892,
+      "step": 150
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.118985906942765,
+      "learning_rate": 1.999923404995121e-05,
+      "loss": 1.4208,
+      "step": 151
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6662933716122046,
+      "learning_rate": 1.9999141289162193e-05,
+      "loss": 0.8128,
+      "step": 152
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6253222619265304,
+      "learning_rate": 1.999904322806854e-05,
+      "loss": 0.8666,
+      "step": 153
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.71046946167768,
+      "learning_rate": 1.999893986672224e-05,
+      "loss": 0.8082,
+      "step": 154
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5419744094605234,
+      "learning_rate": 1.9998831205178075e-05,
+      "loss": 0.7544,
+      "step": 155
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5961300085537884,
+      "learning_rate": 1.999871724349365e-05,
+      "loss": 0.8825,
+      "step": 156
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6216142699276839,
+      "learning_rate": 1.999859798172936e-05,
+      "loss": 0.8792,
+      "step": 157
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5811441572555203,
+      "learning_rate": 1.999847341994844e-05,
+      "loss": 0.8385,
+      "step": 158
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.288893861820737,
+      "learning_rate": 1.9998343558216907e-05,
+      "loss": 1.4828,
+      "step": 159
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.678936918078145,
+      "learning_rate": 1.99982083966036e-05,
+      "loss": 0.8262,
+      "step": 160
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6117382528404862,
+      "learning_rate": 1.9998067935180167e-05,
+      "loss": 0.9976,
+      "step": 161
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.234543840530099,
+      "learning_rate": 1.9997922174021056e-05,
+      "loss": 1.3937,
+      "step": 162
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.167352457379188,
+      "learning_rate": 1.999777111320354e-05,
+      "loss": 1.4818,
+      "step": 163
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.373249498421002,
+      "learning_rate": 1.9997614752807686e-05,
+      "loss": 1.5131,
+      "step": 164
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8419129743627816,
+      "learning_rate": 1.9997453092916385e-05,
+      "loss": 0.8504,
+      "step": 165
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2160912992305746,
+      "learning_rate": 1.9997286133615318e-05,
+      "loss": 1.5205,
+      "step": 166
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.082321297141339,
+      "learning_rate": 1.9997113874992998e-05,
+      "loss": 1.5122,
+      "step": 167
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6053456872809597,
+      "learning_rate": 1.9996936317140726e-05,
+      "loss": 0.7418,
+      "step": 168
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.69580629773305,
+      "learning_rate": 1.9996753460152623e-05,
+      "loss": 0.8482,
+      "step": 169
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7855337395359314,
+      "learning_rate": 1.9996565304125623e-05,
+      "loss": 0.7716,
+      "step": 170
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6415453525963302,
+      "learning_rate": 1.999637184915946e-05,
+      "loss": 0.8524,
+      "step": 171
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7652221545602806,
+      "learning_rate": 1.999617309535667e-05,
+      "loss": 0.9337,
+      "step": 172
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5686907390589533,
+      "learning_rate": 1.999596904282262e-05,
+      "loss": 0.9378,
+      "step": 173
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.3443781073312273,
+      "learning_rate": 1.9995759691665472e-05,
+      "loss": 1.473,
+      "step": 174
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.19896133624504,
+      "learning_rate": 1.9995545041996193e-05,
+      "loss": 1.506,
+      "step": 175
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2418881373893322,
+      "learning_rate": 1.9995325093928568e-05,
+      "loss": 1.4112,
+      "step": 176
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.574782391682473,
+      "learning_rate": 1.9995099847579184e-05,
+      "loss": 0.8179,
+      "step": 177
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6306544284563094,
+      "learning_rate": 1.9994869303067436e-05,
+      "loss": 0.7738,
+      "step": 178
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6486594625939026,
+      "learning_rate": 1.9994633460515537e-05,
+      "loss": 0.8142,
+      "step": 179
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.593163602599557,
+      "learning_rate": 1.9994392320048493e-05,
+      "loss": 0.8313,
+      "step": 180
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.608178825040026,
+      "learning_rate": 1.9994145881794134e-05,
+      "loss": 1.5103,
+      "step": 181
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5921451710393688,
+      "learning_rate": 1.999389414588309e-05,
+      "loss": 0.8234,
+      "step": 182
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5711023366140981,
+      "learning_rate": 1.9993637112448796e-05,
+      "loss": 0.817,
+      "step": 183
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.3434888694015927,
+      "learning_rate": 1.9993374781627502e-05,
+      "loss": 1.4859,
+      "step": 184
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5993416108493899,
+      "learning_rate": 1.9993107153558263e-05,
+      "loss": 0.8618,
+      "step": 185
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.286772675218678,
+      "learning_rate": 1.9992834228382945e-05,
+      "loss": 1.4731,
+      "step": 186
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.16671732802512,
+      "learning_rate": 1.999255600624621e-05,
+      "loss": 1.4305,
+      "step": 187
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7388121469398992,
+      "learning_rate": 1.999227248729555e-05,
+      "loss": 0.814,
+      "step": 188
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6860286659680694,
+      "learning_rate": 1.9991983671681238e-05,
+      "loss": 0.9787,
+      "step": 189
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4725437001986743,
+      "learning_rate": 1.999168955955638e-05,
+      "loss": 0.7718,
+      "step": 190
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.675377585763303,
+      "learning_rate": 1.9991390151076873e-05,
+      "loss": 0.864,
+      "step": 191
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7172377475374727,
+      "learning_rate": 1.999108544640142e-05,
+      "loss": 0.8613,
+      "step": 192
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.463951548403978,
+      "learning_rate": 1.999077544569155e-05,
+      "loss": 0.8092,
+      "step": 193
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.673312227995075,
+      "learning_rate": 1.9990460149111577e-05,
+      "loss": 0.8181,
+      "step": 194
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.60967780700213,
+      "learning_rate": 1.9990139556828638e-05,
+      "loss": 0.77,
+      "step": 195
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.555237128576539,
+      "learning_rate": 1.9989813669012665e-05,
+      "loss": 0.8021,
+      "step": 196
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5289619054690635,
+      "learning_rate": 1.998948248583641e-05,
+      "loss": 0.7441,
+      "step": 197
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.559158382814449,
+      "learning_rate": 1.9989146007475426e-05,
+      "loss": 1.4548,
+      "step": 198
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6621639978072749,
+      "learning_rate": 1.9988804234108063e-05,
+      "loss": 0.7711,
+      "step": 199
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6866499125027081,
+      "learning_rate": 1.9988457165915495e-05,
+      "loss": 0.901,
+      "step": 200
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6261840223180264,
+      "learning_rate": 1.998810480308169e-05,
+      "loss": 0.8709,
+      "step": 201
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5387311849670695,
+      "learning_rate": 1.998774714579343e-05,
+      "loss": 0.7966,
+      "step": 202
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2468776475248,
+      "learning_rate": 1.99873841942403e-05,
+      "loss": 1.512,
+      "step": 203
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.275619739187718,
+      "learning_rate": 1.9987015948614693e-05,
+      "loss": 1.394,
+      "step": 204
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5708275997111598,
+      "learning_rate": 1.9986642409111802e-05,
+      "loss": 0.7688,
+      "step": 205
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6590787580237216,
+      "learning_rate": 1.9986263575929635e-05,
+      "loss": 0.8231,
+      "step": 206
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.937138825971171,
+      "learning_rate": 1.9985879449269006e-05,
+      "loss": 0.9039,
+      "step": 207
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6576305147140526,
+      "learning_rate": 1.9985490029333526e-05,
+      "loss": 0.8513,
+      "step": 208
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5965202188498848,
+      "learning_rate": 1.9985095316329617e-05,
+      "loss": 0.757,
+      "step": 209
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1999545903786566,
+      "learning_rate": 1.9984695310466508e-05,
+      "loss": 1.4352,
+      "step": 210
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5374582940592199,
+      "learning_rate": 1.9984290011956236e-05,
+      "loss": 0.8658,
+      "step": 211
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.65416334301326,
+      "learning_rate": 1.9983879421013636e-05,
+      "loss": 0.8571,
+      "step": 212
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2698595574781666,
+      "learning_rate": 1.9983463537856355e-05,
+      "loss": 1.4434,
+      "step": 213
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.07983827946468,
+      "learning_rate": 1.9983042362704837e-05,
+      "loss": 1.415,
+      "step": 214
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6532513018644899,
+      "learning_rate": 1.9982615895782348e-05,
+      "loss": 0.8322,
+      "step": 215
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5902277722968092,
+      "learning_rate": 1.9982184137314936e-05,
+      "loss": 0.8028,
+      "step": 216
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7813620533526588,
+      "learning_rate": 1.998174708753147e-05,
+      "loss": 0.7717,
+      "step": 217
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.869520721234094,
+      "learning_rate": 1.9981304746663625e-05,
+      "loss": 0.9091,
+      "step": 218
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2773144168470356,
+      "learning_rate": 1.9980857114945867e-05,
+      "loss": 1.4268,
+      "step": 219
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.559233631615229,
+      "learning_rate": 1.998040419261548e-05,
+      "loss": 1.4951,
+      "step": 220
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5786474488013897,
+      "learning_rate": 1.9979945979912543e-05,
+      "loss": 0.7944,
+      "step": 221
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.751446387503828,
+      "learning_rate": 1.997948247707995e-05,
+      "loss": 0.8889,
+      "step": 222
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.3471407993142477,
+      "learning_rate": 1.9979013684363385e-05,
+      "loss": 1.3845,
+      "step": 223
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5125393767843254,
+      "learning_rate": 1.997853960201135e-05,
+      "loss": 0.7587,
+      "step": 224
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4572138421692842,
+      "learning_rate": 1.997806023027514e-05,
+      "loss": 0.7147,
+      "step": 225
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6655493842229092,
+      "learning_rate": 1.997757556940886e-05,
+      "loss": 0.8711,
+      "step": 226
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6217869373446032,
+      "learning_rate": 1.9977085619669424e-05,
+      "loss": 0.8353,
+      "step": 227
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6805286325840894,
+      "learning_rate": 1.997659038131653e-05,
+      "loss": 0.784,
+      "step": 228
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5203187994156018,
+      "learning_rate": 1.9976089854612703e-05,
+      "loss": 0.7317,
+      "step": 229
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6505712856344315,
+      "learning_rate": 1.9975584039823253e-05,
+      "loss": 0.9192,
+      "step": 230
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.670644964977408,
+      "learning_rate": 1.9975072937216306e-05,
+      "loss": 0.8329,
+      "step": 231
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6150194199253087,
+      "learning_rate": 1.9974556547062778e-05,
+      "loss": 0.8779,
+      "step": 232
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6237200586527463,
+      "learning_rate": 1.99740348696364e-05,
+      "loss": 0.8531,
+      "step": 233
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5025184192383485,
+      "learning_rate": 1.9973507905213703e-05,
+      "loss": 0.8923,
+      "step": 234
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5970338767036156,
+      "learning_rate": 1.9972975654074012e-05,
+      "loss": 0.8886,
+      "step": 235
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6471771117220955,
+      "learning_rate": 1.9972438116499464e-05,
+      "loss": 0.8012,
+      "step": 236
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1181222472787997,
+      "learning_rate": 1.9971895292774994e-05,
+      "loss": 1.4352,
+      "step": 237
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5683183210595915,
+      "learning_rate": 1.9971347183188342e-05,
+      "loss": 0.7873,
+      "step": 238
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5433415182523071,
+      "learning_rate": 1.9970793788030044e-05,
+      "loss": 0.7785,
+      "step": 239
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6328442486064931,
+      "learning_rate": 1.997023510759344e-05,
+      "loss": 0.8225,
+      "step": 240
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.730794535485846,
+      "learning_rate": 1.9969671142174684e-05,
+      "loss": 0.8744,
+      "step": 241
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3621359076464667,
+      "learning_rate": 1.9969101892072706e-05,
+      "loss": 0.8223,
+      "step": 242
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2729006191726016,
+      "learning_rate": 1.9968527357589264e-05,
+      "loss": 1.4553,
+      "step": 243
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7564032518382,
+      "learning_rate": 1.9967947539028895e-05,
+      "loss": 0.9448,
+      "step": 244
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7300392941812839,
+      "learning_rate": 1.9967362436698956e-05,
+      "loss": 0.9174,
+      "step": 245
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.22581422181252,
+      "learning_rate": 1.9966772050909587e-05,
+      "loss": 1.4879,
+      "step": 246
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.650356200152166,
+      "learning_rate": 1.9966176381973746e-05,
+      "loss": 0.7951,
+      "step": 247
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5533378278699383,
+      "learning_rate": 1.9965575430207176e-05,
+      "loss": 0.8004,
+      "step": 248
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6813060887682667,
+      "learning_rate": 1.996496919592843e-05,
+      "loss": 0.8255,
+      "step": 249
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.3215588532615823,
+      "learning_rate": 1.9964357679458857e-05,
+      "loss": 1.3997,
+      "step": 250
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.186716659774522,
+      "learning_rate": 1.9963740881122608e-05,
+      "loss": 1.4503,
+      "step": 251
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.437264096122528,
+      "learning_rate": 1.996311880124663e-05,
+      "loss": 0.7897,
+      "step": 252
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6169218281082371,
+      "learning_rate": 1.9962491440160677e-05,
+      "loss": 0.8328,
+      "step": 253
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.601893041175427,
+      "learning_rate": 1.99618587981973e-05,
+      "loss": 0.7676,
+      "step": 254
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7454261943580274,
+      "learning_rate": 1.9961220875691842e-05,
+      "loss": 0.7449,
+      "step": 255
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6371372595136564,
+      "learning_rate": 1.996057767298245e-05,
+      "loss": 0.949,
+      "step": 256
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.2486575179027235,
+      "learning_rate": 1.9959929190410074e-05,
+      "loss": 1.4163,
+      "step": 257
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5574422620666701,
+      "learning_rate": 1.9959275428318458e-05,
+      "loss": 0.7942,
+      "step": 258
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.478649467689756,
+      "learning_rate": 1.9958616387054147e-05,
+      "loss": 0.8076,
+      "step": 259
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.269072568631651,
+      "learning_rate": 1.995795206696648e-05,
+      "loss": 1.4186,
+      "step": 260
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7098751534544372,
+      "learning_rate": 1.9957282468407594e-05,
+      "loss": 0.873,
+      "step": 261
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.674890615753832,
+      "learning_rate": 1.9956607591732437e-05,
+      "loss": 0.8465,
+      "step": 262
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5962022325473393,
+      "learning_rate": 1.9955927437298733e-05,
+      "loss": 0.8071,
+      "step": 263
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7180131562738048,
+      "learning_rate": 1.9955242005467025e-05,
+      "loss": 0.8489,
+      "step": 264
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5195169261031694,
+      "learning_rate": 1.995455129660064e-05,
+      "loss": 0.8755,
+      "step": 265
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6246896289827686,
+      "learning_rate": 1.995385531106571e-05,
+      "loss": 0.8367,
+      "step": 266
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.488308999127481,
+      "learning_rate": 1.9953154049231152e-05,
+      "loss": 0.8002,
+      "step": 267
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.556382195744101,
+      "learning_rate": 1.9952447511468696e-05,
+      "loss": 0.8224,
+      "step": 268
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4684940483216313,
+      "learning_rate": 1.995173569815286e-05,
+      "loss": 0.7511,
+      "step": 269
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.662196710035514,
+      "learning_rate": 1.9951018609660954e-05,
+      "loss": 0.8568,
+      "step": 270
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5117436572451728,
+      "learning_rate": 1.9950296246373096e-05,
+      "loss": 0.8119,
+      "step": 271
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.582589123245647,
+      "learning_rate": 1.9949568608672186e-05,
+      "loss": 0.8043,
+      "step": 272
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5746161260198477,
+      "learning_rate": 1.9948835696943934e-05,
+      "loss": 0.7904,
+      "step": 273
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3598356475872961,
+      "learning_rate": 1.9948097511576832e-05,
+      "loss": 0.6712,
+      "step": 274
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5002073247832104,
+      "learning_rate": 1.994735405296218e-05,
+      "loss": 0.7509,
+      "step": 275
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.30325562538523,
+      "learning_rate": 1.9946605321494066e-05,
+      "loss": 1.4487,
+      "step": 276
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.235664511276466,
+      "learning_rate": 1.9945851317569377e-05,
+      "loss": 1.4247,
+      "step": 277
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5274556390661786,
+      "learning_rate": 1.9945092041587788e-05,
+      "loss": 0.8486,
+      "step": 278
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5938736040599895,
+      "learning_rate": 1.9944327493951775e-05,
+      "loss": 0.7765,
+      "step": 279
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.3016442163519355,
+      "learning_rate": 1.9943557675066603e-05,
+      "loss": 1.4705,
+      "step": 280
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.75972302614624,
+      "learning_rate": 1.994278258534034e-05,
+      "loss": 0.8511,
+      "step": 281
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4922123889728967,
+      "learning_rate": 1.994200222518384e-05,
+      "loss": 0.7717,
+      "step": 282
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6411924373616222,
+      "learning_rate": 1.9941216595010754e-05,
+      "loss": 0.8053,
+      "step": 283
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6721157331690013,
+      "learning_rate": 1.9940425695237523e-05,
+      "loss": 0.8164,
+      "step": 284
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6485301485192154,
+      "learning_rate": 1.9939629526283387e-05,
+      "loss": 0.7927,
+      "step": 285
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.534101079245892,
+      "learning_rate": 1.9938828088570376e-05,
+      "loss": 0.7671,
+      "step": 286
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5751851190978725,
+      "learning_rate": 1.993802138252331e-05,
+      "loss": 0.8506,
+      "step": 287
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6857077769581443,
+      "learning_rate": 1.993720940856981e-05,
+      "loss": 0.8471,
+      "step": 288
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.523654625555751,
+      "learning_rate": 1.9936392167140277e-05,
+      "loss": 0.8535,
+      "step": 289
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5053242870765073,
+      "learning_rate": 1.9935569658667912e-05,
+      "loss": 0.781,
+      "step": 290
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.615157641823543,
+      "learning_rate": 1.9934741883588714e-05,
+      "loss": 0.9128,
+      "step": 291
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.603743732648274,
+      "learning_rate": 1.993390884234146e-05,
+      "loss": 0.7269,
+      "step": 292
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7329720224385576,
+      "learning_rate": 1.9933070535367725e-05,
+      "loss": 0.9468,
+      "step": 293
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.3739129572239737,
+      "learning_rate": 1.993222696311188e-05,
+      "loss": 1.4643,
+      "step": 294
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.508976160243827,
+      "learning_rate": 1.993137812602108e-05,
+      "loss": 0.8149,
+      "step": 295
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.168451327509016,
+      "learning_rate": 1.9930524024545274e-05,
+      "loss": 1.479,
+      "step": 296
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0927976468995664,
+      "learning_rate": 1.9929664659137192e-05,
+      "loss": 1.4713,
+      "step": 297
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6298841479485244,
+      "learning_rate": 1.992880003025238e-05,
+      "loss": 0.8701,
+      "step": 298
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.587793498598202,
+      "learning_rate": 1.9927930138349142e-05,
+      "loss": 0.723,
+      "step": 299
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6915203426694783,
+      "learning_rate": 1.9927054983888598e-05,
+      "loss": 0.7418,
+      "step": 300
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.714763003981645,
+      "learning_rate": 1.9926174567334636e-05,
+      "loss": 0.7242,
+      "step": 301
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.4136559301585705,
+      "learning_rate": 1.992528888915395e-05,
+      "loss": 1.3873,
+      "step": 302
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1054940112996348,
+      "learning_rate": 1.9924397949816015e-05,
+      "loss": 1.4566,
+      "step": 303
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5890747554546618,
+      "learning_rate": 1.9923501749793097e-05,
+      "loss": 0.8072,
+      "step": 304
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5833668973591908,
+      "learning_rate": 1.992260028956025e-05,
+      "loss": 0.8339,
+      "step": 305
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6444270618299326,
+      "learning_rate": 1.9921693569595315e-05,
+      "loss": 0.7753,
+      "step": 306
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5940612332198585,
+      "learning_rate": 1.9920781590378927e-05,
+      "loss": 0.899,
+      "step": 307
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6419646382936637,
+      "learning_rate": 1.99198643523945e-05,
+      "loss": 0.8106,
+      "step": 308
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5878993574437923,
+      "learning_rate": 1.9918941856128237e-05,
+      "loss": 0.828,
+      "step": 309
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5272866401707217,
+      "learning_rate": 1.9918014102069137e-05,
+      "loss": 0.8316,
+      "step": 310
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.090818748759643,
+      "learning_rate": 1.991708109070898e-05,
+      "loss": 0.8139,
+      "step": 311
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5755134217131335,
+      "learning_rate": 1.9916142822542327e-05,
+      "loss": 0.7927,
+      "step": 312
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4525653932135016,
+      "learning_rate": 1.991519929806654e-05,
+      "loss": 0.8035,
+      "step": 313
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4191478107063182,
+      "learning_rate": 1.9914250517781754e-05,
+      "loss": 0.7467,
+      "step": 314
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.557817744003753,
+      "learning_rate": 1.991329648219089e-05,
+      "loss": 0.8929,
+      "step": 315
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.524705327380714,
+      "learning_rate": 1.991233719179967e-05,
+      "loss": 0.7782,
+      "step": 316
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5173641385691985,
+      "learning_rate": 1.9911372647116584e-05,
+      "loss": 0.7583,
+      "step": 317
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6146711952256054,
+      "learning_rate": 1.9910402848652912e-05,
+      "loss": 0.8077,
+      "step": 318
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.5059388930400597,
+      "learning_rate": 1.9909427796922725e-05,
+      "loss": 1.4852,
+      "step": 319
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5305413274097992,
+      "learning_rate": 1.9908447492442874e-05,
+      "loss": 0.7789,
+      "step": 320
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4765505400798873,
+      "learning_rate": 1.9907461935732993e-05,
+      "loss": 0.7283,
+      "step": 321
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6113514633997215,
+      "learning_rate": 1.9906471127315504e-05,
+      "loss": 0.8513,
+      "step": 322
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3210655160208613,
+      "learning_rate": 1.9905475067715608e-05,
+      "loss": 0.7661,
+      "step": 323
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5548943846053953,
+      "learning_rate": 1.9904473757461295e-05,
+      "loss": 0.8532,
+      "step": 324
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4281382233300248,
+      "learning_rate": 1.9903467197083333e-05,
+      "loss": 0.7696,
+      "step": 325
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5298144204110449,
+      "learning_rate": 1.990245538711528e-05,
+      "loss": 0.8227,
+      "step": 326
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3643966417409483,
+      "learning_rate": 1.9901438328093468e-05,
+      "loss": 0.7271,
+      "step": 327
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5364497458539466,
+      "learning_rate": 1.9900416020557016e-05,
+      "loss": 0.7848,
+      "step": 328
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5513564121608612,
+      "learning_rate": 1.989938846504783e-05,
+      "loss": 0.8318,
+      "step": 329
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6128023479962763,
+      "learning_rate": 1.9898355662110584e-05,
+      "loss": 0.8863,
+      "step": 330
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7082071858865575,
+      "learning_rate": 1.9897317612292745e-05,
+      "loss": 0.8463,
+      "step": 331
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5807088006554506,
+      "learning_rate": 1.9896274316144558e-05,
+      "loss": 0.7967,
+      "step": 332
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6526777500591077,
+      "learning_rate": 1.9895225774219055e-05,
+      "loss": 0.9076,
+      "step": 333
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.53221756172183,
+      "learning_rate": 1.9894171987072038e-05,
+      "loss": 0.8089,
+      "step": 334
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5517326870792454,
+      "learning_rate": 1.989311295526209e-05,
+      "loss": 0.8094,
+      "step": 335
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4887150420217152,
+      "learning_rate": 1.9892048679350585e-05,
+      "loss": 0.8646,
+      "step": 336
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.535865443845866,
+      "learning_rate": 1.9890979159901673e-05,
+      "loss": 0.8882,
+      "step": 337
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5164550834515007,
+      "learning_rate": 1.988990439748227e-05,
+      "loss": 0.8025,
+      "step": 338
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4137009149388162,
+      "learning_rate": 1.988882439266209e-05,
+      "loss": 0.8078,
+      "step": 339
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6122665993484884,
+      "learning_rate": 1.9887739146013612e-05,
+      "loss": 0.9222,
+      "step": 340
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5435372551925122,
+      "learning_rate": 1.9886648658112103e-05,
+      "loss": 0.8332,
+      "step": 341
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6696756144915297,
+      "learning_rate": 1.9885552929535607e-05,
+      "loss": 0.9252,
+      "step": 342
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.3465898075450333,
+      "learning_rate": 1.9884451960864938e-05,
+      "loss": 1.3447,
+      "step": 343
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.214798775102066,
+      "learning_rate": 1.9883345752683695e-05,
+      "loss": 1.3641,
+      "step": 344
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5915825552765308,
+      "learning_rate": 1.9882234305578253e-05,
+      "loss": 0.7569,
+      "step": 345
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6821792531537183,
+      "learning_rate": 1.9881117620137762e-05,
+      "loss": 0.9235,
+      "step": 346
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.4157773360156325,
+      "learning_rate": 1.987999569695415e-05,
+      "loss": 1.4703,
+      "step": 347
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7011184425155454,
+      "learning_rate": 1.9878868536622124e-05,
+      "loss": 0.8797,
+      "step": 348
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.2684500170958017,
+      "learning_rate": 1.9877736139739162e-05,
+      "loss": 1.407,
+      "step": 349
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.581779221853985,
+      "learning_rate": 1.9876598506905522e-05,
+      "loss": 0.7796,
+      "step": 350
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.676953407316439,
+      "learning_rate": 1.987545563872424e-05,
+      "loss": 0.7866,
+      "step": 351
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.374704691364364,
+      "learning_rate": 1.9874307535801112e-05,
+      "loss": 0.7577,
+      "step": 352
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.5329682625986574,
+      "learning_rate": 1.987315419874473e-05,
+      "loss": 1.4558,
+      "step": 353
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.4017229640712814,
+      "learning_rate": 1.9871995628166442e-05,
+      "loss": 1.4882,
+      "step": 354
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1033503830342197,
+      "learning_rate": 1.987083182468039e-05,
+      "loss": 1.4573,
+      "step": 355
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6549726796972388,
+      "learning_rate": 1.9869662788903465e-05,
+      "loss": 0.877,
+      "step": 356
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7181664063142648,
+      "learning_rate": 1.9868488521455348e-05,
+      "loss": 0.8583,
+      "step": 357
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5618766051865554,
+      "learning_rate": 1.98673090229585e-05,
+      "loss": 0.759,
+      "step": 358
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5101248333403627,
+      "learning_rate": 1.9866124294038135e-05,
+      "loss": 0.8039,
+      "step": 359
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5488128538187185,
+      "learning_rate": 1.986493433532225e-05,
+      "loss": 0.8338,
+      "step": 360
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5731283413974633,
+      "learning_rate": 1.986373914744162e-05,
+      "loss": 0.7883,
+      "step": 361
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.644402593028274,
+      "learning_rate": 1.9862538731029778e-05,
+      "loss": 0.8572,
+      "step": 362
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.707355891733103,
+      "learning_rate": 1.986133308672304e-05,
+      "loss": 0.76,
+      "step": 363
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.9270349826016946,
+      "learning_rate": 1.986012221516049e-05,
+      "loss": 1.4477,
+      "step": 364
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4175667636092493,
+      "learning_rate": 1.9858906116983975e-05,
+      "loss": 0.7187,
+      "step": 365
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0864569112172213,
+      "learning_rate": 1.9857684792838127e-05,
+      "loss": 1.4565,
+      "step": 366
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6734254745085826,
+      "learning_rate": 1.9856458243370337e-05,
+      "loss": 0.7894,
+      "step": 367
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.718414311459475,
+      "learning_rate": 1.985522646923077e-05,
+      "loss": 0.7659,
+      "step": 368
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.673394632808843,
+      "learning_rate": 1.9853989471072362e-05,
+      "loss": 0.843,
+      "step": 369
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.580616673527355,
+      "learning_rate": 1.9852747249550813e-05,
+      "loss": 0.8454,
+      "step": 370
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.600015230578288,
+      "learning_rate": 1.98514998053246e-05,
+      "loss": 1.3936,
+      "step": 371
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5441213097358104,
+      "learning_rate": 1.9850247139054953e-05,
+      "loss": 0.7295,
+      "step": 372
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4903427734685464,
+      "learning_rate": 1.984898925140589e-05,
+      "loss": 0.889,
+      "step": 373
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4950368191309529,
+      "learning_rate": 1.9847726143044184e-05,
+      "loss": 0.8072,
+      "step": 374
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.538373410643168,
+      "learning_rate": 1.9846457814639378e-05,
+      "loss": 0.8319,
+      "step": 375
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5879382016565526,
+      "learning_rate": 1.984518426686378e-05,
+      "loss": 0.857,
+      "step": 376
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7682989840721284,
+      "learning_rate": 1.984390550039247e-05,
+      "loss": 0.801,
+      "step": 377
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5414422781511836,
+      "learning_rate": 1.9842621515903292e-05,
+      "loss": 0.8158,
+      "step": 378
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.624642454336686,
+      "learning_rate": 1.9841332314076856e-05,
+      "loss": 0.8742,
+      "step": 379
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5925706903107164,
+      "learning_rate": 1.9840037895596532e-05,
+      "loss": 0.8001,
+      "step": 380
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.799447224665979,
+      "learning_rate": 1.9838738261148465e-05,
+      "loss": 0.8416,
+      "step": 381
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5143787980189258,
+      "learning_rate": 1.9837433411421556e-05,
+      "loss": 0.8343,
+      "step": 382
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5465739699008163,
+      "learning_rate": 1.9836123347107482e-05,
+      "loss": 0.797,
+      "step": 383
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4942436999321522,
+      "learning_rate": 1.9834808068900666e-05,
+      "loss": 0.7359,
+      "step": 384
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5775685916022684,
+      "learning_rate": 1.9833487577498312e-05,
+      "loss": 0.8168,
+      "step": 385
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5132293673265838,
+      "learning_rate": 1.983216187360038e-05,
+      "loss": 0.7627,
+      "step": 386
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5353158723486429,
+      "learning_rate": 1.9830830957909596e-05,
+      "loss": 0.7941,
+      "step": 387
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5164723724233244,
+      "learning_rate": 1.982949483113144e-05,
+      "loss": 0.8096,
+      "step": 388
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.3297965516713735,
+      "learning_rate": 1.9828153493974164e-05,
+      "loss": 1.4707,
+      "step": 389
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5628402924247116,
+      "learning_rate": 1.9826806947148783e-05,
+      "loss": 0.6963,
+      "step": 390
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1178286446882475,
+      "learning_rate": 1.9825455191369065e-05,
+      "loss": 1.4313,
+      "step": 391
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6292976092785298,
+      "learning_rate": 1.982409822735154e-05,
+      "loss": 0.7499,
+      "step": 392
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5795837109519286,
+      "learning_rate": 1.982273605581551e-05,
+      "loss": 0.8666,
+      "step": 393
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4585555300803599,
+      "learning_rate": 1.982136867748302e-05,
+      "loss": 0.7908,
+      "step": 394
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.3215122207005168,
+      "learning_rate": 1.981999609307889e-05,
+      "loss": 1.4246,
+      "step": 395
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5721605255099451,
+      "learning_rate": 1.98186183033307e-05,
+      "loss": 0.8206,
+      "step": 396
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4949272493825847,
+      "learning_rate": 1.9817235308968772e-05,
+      "loss": 0.834,
+      "step": 397
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5638706811093301,
+      "learning_rate": 1.9815847110726203e-05,
+      "loss": 0.8222,
+      "step": 398
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.26166812586486,
+      "learning_rate": 1.9814453709338844e-05,
+      "loss": 1.4571,
+      "step": 399
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.584213638814795,
+      "learning_rate": 1.98130551055453e-05,
+      "loss": 0.8614,
+      "step": 400
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7615063716220298,
+      "learning_rate": 1.9811651300086946e-05,
+      "loss": 0.754,
+      "step": 401
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4594499352666774,
+      "learning_rate": 1.9810242293707896e-05,
+      "loss": 0.9057,
+      "step": 402
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5444578746517663,
+      "learning_rate": 1.9808828087155034e-05,
+      "loss": 0.8278,
+      "step": 403
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4539991002341044,
+      "learning_rate": 1.9807408681178e-05,
+      "loss": 0.7586,
+      "step": 404
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5442251680071133,
+      "learning_rate": 1.980598407652918e-05,
+      "loss": 0.8022,
+      "step": 405
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4771785050624275,
+      "learning_rate": 1.9804554273963728e-05,
+      "loss": 0.7698,
+      "step": 406
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.425188792758254,
+      "learning_rate": 1.9803119274239546e-05,
+      "loss": 0.7487,
+      "step": 407
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.309825771780373,
+      "learning_rate": 1.9801679078117298e-05,
+      "loss": 1.4345,
+      "step": 408
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.49266113927696,
+      "learning_rate": 1.980023368636039e-05,
+      "loss": 0.8441,
+      "step": 409
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.19709274036693,
+      "learning_rate": 1.9798783099734993e-05,
+      "loss": 1.4722,
+      "step": 410
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5613051886532363,
+      "learning_rate": 1.9797327319010028e-05,
+      "loss": 0.8768,
+      "step": 411
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.093470331282145,
+      "learning_rate": 1.9795866344957168e-05,
+      "loss": 1.3566,
+      "step": 412
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.267851184416205,
+      "learning_rate": 1.9794400178350842e-05,
+      "loss": 1.4939,
+      "step": 413
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6578706579781919,
+      "learning_rate": 1.979292881996823e-05,
+      "loss": 0.8378,
+      "step": 414
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.2731367131228937,
+      "learning_rate": 1.979145227058926e-05,
+      "loss": 1.4515,
+      "step": 415
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0258280015649315,
+      "learning_rate": 1.9789970530996623e-05,
+      "loss": 1.4534,
+      "step": 416
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.600859320743873,
+      "learning_rate": 1.9788483601975743e-05,
+      "loss": 0.7388,
+      "step": 417
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.2283644259547324,
+      "learning_rate": 1.9786991484314816e-05,
+      "loss": 1.4566,
+      "step": 418
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5431192192892154,
+      "learning_rate": 1.978549417880477e-05,
+      "loss": 0.7936,
+      "step": 419
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.621252976745805,
+      "learning_rate": 1.9783991686239295e-05,
+      "loss": 0.8512,
+      "step": 420
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.452274412188008,
+      "learning_rate": 1.9782484007414823e-05,
+      "loss": 0.8114,
+      "step": 421
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5330328858762625,
+      "learning_rate": 1.9780971143130543e-05,
+      "loss": 0.7127,
+      "step": 422
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.529637782683529,
+      "learning_rate": 1.9779453094188384e-05,
+      "loss": 0.7324,
+      "step": 423
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5804943240199387,
+      "learning_rate": 1.9777929861393027e-05,
+      "loss": 0.8769,
+      "step": 424
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4920419180396567,
+      "learning_rate": 1.9776401445551902e-05,
+      "loss": 0.7308,
+      "step": 425
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5069743944433995,
+      "learning_rate": 1.9774867847475186e-05,
+      "loss": 0.8329,
+      "step": 426
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5507937240929264,
+      "learning_rate": 1.9773329067975805e-05,
+      "loss": 0.8109,
+      "step": 427
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.418963668192508,
+      "learning_rate": 1.9771785107869427e-05,
+      "loss": 1.4591,
+      "step": 428
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.58443753819316,
+      "learning_rate": 1.977023596797446e-05,
+      "loss": 0.8754,
+      "step": 429
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4551911952104568,
+      "learning_rate": 1.9768681649112082e-05,
+      "loss": 0.7549,
+      "step": 430
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1313823950806836,
+      "learning_rate": 1.9767122152106188e-05,
+      "loss": 1.4442,
+      "step": 431
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5235586741271288,
+      "learning_rate": 1.9765557477783437e-05,
+      "loss": 0.773,
+      "step": 432
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5197535345278566,
+      "learning_rate": 1.976398762697322e-05,
+      "loss": 0.8051,
+      "step": 433
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3909070174961307,
+      "learning_rate": 1.9762412600507676e-05,
+      "loss": 0.7881,
+      "step": 434
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.572737658960169,
+      "learning_rate": 1.9760832399221698e-05,
+      "loss": 0.8125,
+      "step": 435
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5114315033180235,
+      "learning_rate": 1.97592470239529e-05,
+      "loss": 0.8334,
+      "step": 436
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.440458293950162,
+      "learning_rate": 1.9757656475541663e-05,
+      "loss": 0.7922,
+      "step": 437
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6692856060782524,
+      "learning_rate": 1.9756060754831093e-05,
+      "loss": 0.8077,
+      "step": 438
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.075795473987679,
+      "learning_rate": 1.975445986266704e-05,
+      "loss": 1.3636,
+      "step": 439
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.2630693738719496,
+      "learning_rate": 1.9752853799898105e-05,
+      "loss": 1.4693,
+      "step": 440
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6176722146358922,
+      "learning_rate": 1.9751242567375617e-05,
+      "loss": 0.8088,
+      "step": 441
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5941932996041412,
+      "learning_rate": 1.974962616595366e-05,
+      "loss": 0.8528,
+      "step": 442
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.091035957708569,
+      "learning_rate": 1.974800459648904e-05,
+      "loss": 1.4053,
+      "step": 443
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5144660819839653,
+      "learning_rate": 1.974637785984132e-05,
+      "loss": 0.7888,
+      "step": 444
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4645880502387654,
+      "learning_rate": 1.974474595687279e-05,
+      "loss": 0.727,
+      "step": 445
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4929235206162328,
+      "learning_rate": 1.974310888844848e-05,
+      "loss": 0.831,
+      "step": 446
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.426830297463986,
+      "learning_rate": 1.974146665543617e-05,
+      "loss": 1.474,
+      "step": 447
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.606413640453345,
+      "learning_rate": 1.9739819258706355e-05,
+      "loss": 0.742,
+      "step": 448
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4911819165327864,
+      "learning_rate": 1.9738166699132294e-05,
+      "loss": 0.7896,
+      "step": 449
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.291942130982717,
+      "learning_rate": 1.973650897758996e-05,
+      "loss": 1.5013,
+      "step": 450
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4823723577470485,
+      "learning_rate": 1.9734846094958074e-05,
+      "loss": 0.7029,
+      "step": 451
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4736382615995927,
+      "learning_rate": 1.973317805211809e-05,
+      "loss": 0.8637,
+      "step": 452
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4575702470157252,
+      "learning_rate": 1.9731504849954198e-05,
+      "loss": 0.85,
+      "step": 453
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.3157059758857286,
+      "learning_rate": 1.9729826489353322e-05,
+      "loss": 1.5112,
+      "step": 454
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5798365038004785,
+      "learning_rate": 1.972814297120512e-05,
+      "loss": 0.8345,
+      "step": 455
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5527270558686022,
+      "learning_rate": 1.9726454296401988e-05,
+      "loss": 0.9049,
+      "step": 456
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6158719337707246,
+      "learning_rate": 1.9724760465839046e-05,
+      "loss": 0.7431,
+      "step": 457
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4133919435767077,
+      "learning_rate": 1.9723061480414158e-05,
+      "loss": 0.7537,
+      "step": 458
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3960305910326667,
+      "learning_rate": 1.9721357341027912e-05,
+      "loss": 0.7424,
+      "step": 459
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.309260246583208,
+      "learning_rate": 1.971964804858363e-05,
+      "loss": 1.4954,
+      "step": 460
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5987141411932977,
+      "learning_rate": 1.9717933603987377e-05,
+      "loss": 0.7916,
+      "step": 461
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.525559153911604,
+      "learning_rate": 1.9716214008147926e-05,
+      "loss": 0.8017,
+      "step": 462
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4822091515687652,
+      "learning_rate": 1.97144892619768e-05,
+      "loss": 0.7969,
+      "step": 463
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5205678505190745,
+      "learning_rate": 1.9712759366388245e-05,
+      "loss": 0.8107,
+      "step": 464
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.3948544212350513,
+      "learning_rate": 1.971102432229924e-05,
+      "loss": 1.4545,
+      "step": 465
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.633226534518844,
+      "learning_rate": 1.970928413062948e-05,
+      "loss": 0.7573,
+      "step": 466
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.631317217243187,
+      "learning_rate": 1.9707538792301406e-05,
+      "loss": 0.7449,
+      "step": 467
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.636470752633967,
+      "learning_rate": 1.9705788308240183e-05,
+      "loss": 0.8131,
+      "step": 468
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9575259692446314,
+      "learning_rate": 1.9704032679373696e-05,
+      "loss": 1.4282,
+      "step": 469
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5985081652690063,
+      "learning_rate": 1.9702271906632564e-05,
+      "loss": 0.8816,
+      "step": 470
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5561543191686664,
+      "learning_rate": 1.9700505990950126e-05,
+      "loss": 0.8563,
+      "step": 471
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.3931279925906517,
+      "learning_rate": 1.969873493326246e-05,
+      "loss": 1.4858,
+      "step": 472
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4463188113695424,
+      "learning_rate": 1.969695873450835e-05,
+      "loss": 0.7832,
+      "step": 473
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5668418837498432,
+      "learning_rate": 1.9695177395629324e-05,
+      "loss": 0.7439,
+      "step": 474
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.217394679531263,
+      "learning_rate": 1.969339091756963e-05,
+      "loss": 1.3738,
+      "step": 475
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.54588446581502,
+      "learning_rate": 1.9691599301276226e-05,
+      "loss": 0.8581,
+      "step": 476
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1133443715786497,
+      "learning_rate": 1.968980254769881e-05,
+      "loss": 1.4643,
+      "step": 477
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1525273289069538,
+      "learning_rate": 1.96880006577898e-05,
+      "loss": 1.3602,
+      "step": 478
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.628463000218483,
+      "learning_rate": 1.968619363250434e-05,
+      "loss": 0.8281,
+      "step": 479
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6480010880049318,
+      "learning_rate": 1.9684381472800275e-05,
+      "loss": 0.8243,
+      "step": 480
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3587877224236686,
+      "learning_rate": 1.96825641796382e-05,
+      "loss": 0.7423,
+      "step": 481
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1692645202111382,
+      "learning_rate": 1.9680741753981416e-05,
+      "loss": 1.448,
+      "step": 482
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6087687287029866,
+      "learning_rate": 1.9678914196795942e-05,
+      "loss": 0.7476,
+      "step": 483
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6478789076508502,
+      "learning_rate": 1.967708150905053e-05,
+      "loss": 0.8951,
+      "step": 484
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.521566371984631,
+      "learning_rate": 1.967524369171664e-05,
+      "loss": 0.8171,
+      "step": 485
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5139034325547713,
+      "learning_rate": 1.9673400745768455e-05,
+      "loss": 0.8447,
+      "step": 486
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6954681025057368,
+      "learning_rate": 1.9671552672182872e-05,
+      "loss": 0.8224,
+      "step": 487
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5429504417953677,
+      "learning_rate": 1.9669699471939517e-05,
+      "loss": 0.7973,
+      "step": 488
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4820675558531284,
+      "learning_rate": 1.9667841146020725e-05,
+      "loss": 0.802,
+      "step": 489
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6394670454796723,
+      "learning_rate": 1.9665977695411547e-05,
+      "loss": 0.8359,
+      "step": 490
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.640196017986903,
+      "learning_rate": 1.966410912109976e-05,
+      "loss": 0.7677,
+      "step": 491
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6569400533056458,
+      "learning_rate": 1.9662235424075843e-05,
+      "loss": 0.9042,
+      "step": 492
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5820419619471737,
+      "learning_rate": 1.9660356605333e-05,
+      "loss": 0.8752,
+      "step": 493
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4305664484664413,
+      "learning_rate": 1.9658472665867147e-05,
+      "loss": 0.8447,
+      "step": 494
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.2811221874330894,
+      "learning_rate": 1.965658360667692e-05,
+      "loss": 1.4455,
+      "step": 495
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0513767086624686,
+      "learning_rate": 1.965468942876366e-05,
+      "loss": 1.4308,
+      "step": 496
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6107206140586445,
+      "learning_rate": 1.9652790133131426e-05,
+      "loss": 0.8167,
+      "step": 497
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1997793009681357,
+      "learning_rate": 1.965088572078699e-05,
+      "loss": 1.374,
+      "step": 498
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4446697035814815,
+      "learning_rate": 1.9648976192739832e-05,
+      "loss": 0.7732,
+      "step": 499
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1671543443084005,
+      "learning_rate": 1.9647061550002152e-05,
+      "loss": 1.3974,
+      "step": 500
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6228120340827634,
+      "learning_rate": 1.964514179358885e-05,
+      "loss": 0.7612,
+      "step": 501
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4802328854350606,
+      "learning_rate": 1.9643216924517545e-05,
+      "loss": 0.7391,
+      "step": 502
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.066353226433902,
+      "learning_rate": 1.9641286943808565e-05,
+      "loss": 1.4379,
+      "step": 503
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5879212865546426,
+      "learning_rate": 1.9639351852484946e-05,
+      "loss": 0.7692,
+      "step": 504
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.511985418536973,
+      "learning_rate": 1.9637411651572437e-05,
+      "loss": 0.715,
+      "step": 505
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.2497407235849995,
+      "learning_rate": 1.9635466342099486e-05,
+      "loss": 1.3749,
+      "step": 506
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.588502356148957,
+      "learning_rate": 1.9633515925097257e-05,
+      "loss": 0.761,
+      "step": 507
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4031040954812177,
+      "learning_rate": 1.9631560401599622e-05,
+      "loss": 0.7385,
+      "step": 508
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4921446082605672,
+      "learning_rate": 1.962959977264315e-05,
+      "loss": 0.8325,
+      "step": 509
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.557714926392037,
+      "learning_rate": 1.962763403926713e-05,
+      "loss": 0.7418,
+      "step": 510
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4887875571755544,
+      "learning_rate": 1.9625663202513553e-05,
+      "loss": 0.7873,
+      "step": 511
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5581359549962766,
+      "learning_rate": 1.9623687263427102e-05,
+      "loss": 0.8427,
+      "step": 512
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6268905285767286,
+      "learning_rate": 1.9621706223055178e-05,
+      "loss": 0.8379,
+      "step": 513
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5516115737706224,
+      "learning_rate": 1.961972008244789e-05,
+      "loss": 0.8165,
+      "step": 514
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5555635593935109,
+      "learning_rate": 1.9617728842658032e-05,
+      "loss": 0.7907,
+      "step": 515
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6136183310213907,
+      "learning_rate": 1.9615732504741123e-05,
+      "loss": 0.7495,
+      "step": 516
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4746382181665287,
+      "learning_rate": 1.9613731069755372e-05,
+      "loss": 0.697,
+      "step": 517
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5785169331793096,
+      "learning_rate": 1.9611724538761682e-05,
+      "loss": 0.8077,
+      "step": 518
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5483705063754283,
+      "learning_rate": 1.960971291282368e-05,
+      "loss": 0.7157,
+      "step": 519
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.196916244343295,
+      "learning_rate": 1.9607696193007673e-05,
+      "loss": 1.4481,
+      "step": 520
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.188149877771475,
+      "learning_rate": 1.9605674380382676e-05,
+      "loss": 1.4288,
+      "step": 521
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5824051087558242,
+      "learning_rate": 1.9603647476020406e-05,
+      "loss": 0.8368,
+      "step": 522
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5324813858937738,
+      "learning_rate": 1.9601615480995275e-05,
+      "loss": 0.8139,
+      "step": 523
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0151527590670524,
+      "learning_rate": 1.9599578396384393e-05,
+      "loss": 1.3518,
+      "step": 524
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5458403481633858,
+      "learning_rate": 1.9597536223267573e-05,
+      "loss": 0.7709,
+      "step": 525
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4792622950005796,
+      "learning_rate": 1.959548896272732e-05,
+      "loss": 0.7264,
+      "step": 526
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5494818270744555,
+      "learning_rate": 1.9593436615848836e-05,
+      "loss": 0.8086,
+      "step": 527
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.3934611462941127,
+      "learning_rate": 1.9591379183720026e-05,
+      "loss": 1.4683,
+      "step": 528
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.530221331040161,
+      "learning_rate": 1.9589316667431482e-05,
+      "loss": 0.7888,
+      "step": 529
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5599598352709865,
+      "learning_rate": 1.9587249068076494e-05,
+      "loss": 0.7889,
+      "step": 530
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.00350772041329,
+      "learning_rate": 1.958517638675104e-05,
+      "loss": 1.4594,
+      "step": 531
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.187819410428765,
+      "learning_rate": 1.9583098624553815e-05,
+      "loss": 1.4044,
+      "step": 532
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0302169951406874,
+      "learning_rate": 1.9581015782586177e-05,
+      "loss": 1.3692,
+      "step": 533
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4625875132233654,
+      "learning_rate": 1.9578927861952196e-05,
+      "loss": 0.7223,
+      "step": 534
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.673051863784586,
+      "learning_rate": 1.9576834863758628e-05,
+      "loss": 0.9088,
+      "step": 535
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6015384514403268,
+      "learning_rate": 1.9574736789114924e-05,
+      "loss": 0.7835,
+      "step": 536
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6146729095882764,
+      "learning_rate": 1.9572633639133213e-05,
+      "loss": 0.8577,
+      "step": 537
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5790129593183513,
+      "learning_rate": 1.9570525414928336e-05,
+      "loss": 0.8276,
+      "step": 538
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.551357458329538,
+      "learning_rate": 1.9568412117617806e-05,
+      "loss": 0.8256,
+      "step": 539
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5074865816008234,
+      "learning_rate": 1.956629374832183e-05,
+      "loss": 0.7734,
+      "step": 540
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5335335724668488,
+      "learning_rate": 1.9564170308163308e-05,
+      "loss": 0.8361,
+      "step": 541
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.456860619859961,
+      "learning_rate": 1.9562041798267826e-05,
+      "loss": 1.4445,
+      "step": 542
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5328456858926172,
+      "learning_rate": 1.9559908219763648e-05,
+      "loss": 0.7295,
+      "step": 543
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.245846160102314,
+      "learning_rate": 1.955776957378174e-05,
+      "loss": 1.3716,
+      "step": 544
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4947559110553863,
+      "learning_rate": 1.9555625861455748e-05,
+      "loss": 0.8178,
+      "step": 545
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7366791655436158,
+      "learning_rate": 1.9553477083922e-05,
+      "loss": 0.8287,
+      "step": 546
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.318304113381094,
+      "learning_rate": 1.9551323242319507e-05,
+      "loss": 1.4647,
+      "step": 547
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.4961730623157985,
+      "learning_rate": 1.9549164337789974e-05,
+      "loss": 1.4174,
+      "step": 548
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.603939335611788,
+      "learning_rate": 1.9547000371477782e-05,
+      "loss": 0.7875,
+      "step": 549
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5962726992748135,
+      "learning_rate": 1.954483134453e-05,
+      "loss": 0.8469,
+      "step": 550
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.4445120897647246,
+      "learning_rate": 1.9542657258096374e-05,
+      "loss": 1.4286,
+      "step": 551
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.526832648756402,
+      "learning_rate": 1.954047811332934e-05,
+      "loss": 0.7229,
+      "step": 552
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.3622709698154574,
+      "learning_rate": 1.9538293911384002e-05,
+      "loss": 1.4328,
+      "step": 553
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.2924193920315448,
+      "learning_rate": 1.953610465341816e-05,
+      "loss": 1.3705,
+      "step": 554
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6392025600604798,
+      "learning_rate": 1.9533910340592287e-05,
+      "loss": 0.8351,
+      "step": 555
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5799626749804807,
+      "learning_rate": 1.953171097406953e-05,
+      "loss": 0.7761,
+      "step": 556
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7404482176742522,
+      "learning_rate": 1.9529506555015726e-05,
+      "loss": 0.7936,
+      "step": 557
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5291615539042147,
+      "learning_rate": 1.9527297084599384e-05,
+      "loss": 0.7729,
+      "step": 558
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5129720962175532,
+      "learning_rate": 1.9525082563991685e-05,
+      "loss": 0.8419,
+      "step": 559
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.565571006755353,
+      "learning_rate": 1.9522862994366504e-05,
+      "loss": 0.8272,
+      "step": 560
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4853280875728994,
+      "learning_rate": 1.952063837690037e-05,
+      "loss": 0.86,
+      "step": 561
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.178257639824095,
+      "learning_rate": 1.9518408712772507e-05,
+      "loss": 1.4887,
+      "step": 562
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3970898290826845,
+      "learning_rate": 1.9516174003164803e-05,
+      "loss": 0.7918,
+      "step": 563
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4364050865406317,
+      "learning_rate": 1.9513934249261823e-05,
+      "loss": 0.6923,
+      "step": 564
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6307737066766599,
+      "learning_rate": 1.9511689452250814e-05,
+      "loss": 0.8583,
+      "step": 565
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.2703396936371556,
+      "learning_rate": 1.950943961332168e-05,
+      "loss": 1.4209,
+      "step": 566
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1176087067327214,
+      "learning_rate": 1.950718473366701e-05,
+      "loss": 1.404,
+      "step": 567
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7230607806491545,
+      "learning_rate": 1.9504924814482063e-05,
+      "loss": 0.7983,
+      "step": 568
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1595366916780936,
+      "learning_rate": 1.9502659856964764e-05,
+      "loss": 1.4143,
+      "step": 569
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4493584080132258,
+      "learning_rate": 1.950038986231572e-05,
+      "loss": 0.7153,
+      "step": 570
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.2417834198087565,
+      "learning_rate": 1.9498114831738192e-05,
+      "loss": 1.3818,
+      "step": 571
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.123872777738961,
+      "learning_rate": 1.9495834766438125e-05,
+      "loss": 1.4428,
+      "step": 572
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4931190444845848,
+      "learning_rate": 1.9493549667624125e-05,
+      "loss": 0.7516,
+      "step": 573
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4233418511881184,
+      "learning_rate": 1.949125953650747e-05,
+      "loss": 0.7289,
+      "step": 574
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.529841622953461,
+      "learning_rate": 1.9488964374302095e-05,
+      "loss": 0.8037,
+      "step": 575
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6523719233173317,
+      "learning_rate": 1.948666418222462e-05,
+      "loss": 0.8755,
+      "step": 576
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4888757537708721,
+      "learning_rate": 1.948435896149432e-05,
+      "loss": 0.7192,
+      "step": 577
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5203311272131874,
+      "learning_rate": 1.9482048713333132e-05,
+      "loss": 0.8238,
+      "step": 578
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.173340356468149,
+      "learning_rate": 1.9479733438965666e-05,
+      "loss": 1.4571,
+      "step": 579
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.504550374728338,
+      "learning_rate": 1.94774131396192e-05,
+      "loss": 0.7491,
+      "step": 580
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.2471915780371243,
+      "learning_rate": 1.9475087816523658e-05,
+      "loss": 1.4755,
+      "step": 581
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5506302095321849,
+      "learning_rate": 1.947275747091164e-05,
+      "loss": 0.8076,
+      "step": 582
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0791475445866907,
+      "learning_rate": 1.947042210401841e-05,
+      "loss": 1.4769,
+      "step": 583
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5706349352340407,
+      "learning_rate": 1.9468081717081893e-05,
+      "loss": 0.8216,
+      "step": 584
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.660378427957418,
+      "learning_rate": 1.9465736311342668e-05,
+      "loss": 0.8407,
+      "step": 585
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5268300471021108,
+      "learning_rate": 1.946338588804397e-05,
+      "loss": 0.8091,
+      "step": 586
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.2291231615779066,
+      "learning_rate": 1.946103044843172e-05,
+      "loss": 1.3135,
+      "step": 587
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.533548247633062,
+      "learning_rate": 1.9458669993754462e-05,
+      "loss": 0.777,
+      "step": 588
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5564526112157973,
+      "learning_rate": 1.9456304525263426e-05,
+      "loss": 0.8292,
+      "step": 589
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9522809862351784,
+      "learning_rate": 1.9453934044212494e-05,
+      "loss": 1.3646,
+      "step": 590
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.579869572348697,
+      "learning_rate": 1.9451558551858188e-05,
+      "loss": 0.8521,
+      "step": 591
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.137315674121054,
+      "learning_rate": 1.944917804945971e-05,
+      "loss": 1.3373,
+      "step": 592
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.487726543817893,
+      "learning_rate": 1.9446792538278904e-05,
+      "loss": 0.7187,
+      "step": 593
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0742235300083673,
+      "learning_rate": 1.944440201958027e-05,
+      "loss": 1.4409,
+      "step": 594
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4371663769846381,
+      "learning_rate": 1.944200649463097e-05,
+      "loss": 0.7942,
+      "step": 595
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6290498429427256,
+      "learning_rate": 1.9439605964700806e-05,
+      "loss": 0.8155,
+      "step": 596
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.197205325060275,
+      "learning_rate": 1.9437200431062247e-05,
+      "loss": 1.374,
+      "step": 597
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5848260233912583,
+      "learning_rate": 1.943478989499041e-05,
+      "loss": 0.8044,
+      "step": 598
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6028748920988958,
+      "learning_rate": 1.943237435776306e-05,
+      "loss": 0.8585,
+      "step": 599
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6004976216841338,
+      "learning_rate": 1.9429953820660614e-05,
+      "loss": 0.8072,
+      "step": 600
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4201749300450293,
+      "learning_rate": 1.9427528284966142e-05,
+      "loss": 0.8344,
+      "step": 601
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5342901806620661,
+      "learning_rate": 1.9425097751965364e-05,
+      "loss": 0.7742,
+      "step": 602
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.098956539322569,
+      "learning_rate": 1.9422662222946643e-05,
+      "loss": 1.3401,
+      "step": 603
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5807953348294717,
+      "learning_rate": 1.9420221699200997e-05,
+      "loss": 0.7888,
+      "step": 604
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.52446925438734,
+      "learning_rate": 1.9417776182022088e-05,
+      "loss": 0.7658,
+      "step": 605
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.2302891517318417,
+      "learning_rate": 1.9415325672706227e-05,
+      "loss": 1.4136,
+      "step": 606
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6643919969483996,
+      "learning_rate": 1.9412870172552368e-05,
+      "loss": 0.8689,
+      "step": 607
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5370088881177923,
+      "learning_rate": 1.9410409682862113e-05,
+      "loss": 0.8129,
+      "step": 608
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.1511257537493704,
+      "learning_rate": 1.9407944204939718e-05,
+      "loss": 1.4235,
+      "step": 609
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.478538624558788,
+      "learning_rate": 1.9405473740092058e-05,
+      "loss": 0.9157,
+      "step": 610
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.537242034476973,
+      "learning_rate": 1.9402998289628678e-05,
+      "loss": 0.8288,
+      "step": 611
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.1421162497291246,
+      "learning_rate": 1.940051785486175e-05,
+      "loss": 1.3834,
+      "step": 612
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.393264397928951,
+      "learning_rate": 1.9398032437106095e-05,
+      "loss": 0.7874,
+      "step": 613
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6251658305175587,
+      "learning_rate": 1.9395542037679172e-05,
+      "loss": 0.8178,
+      "step": 614
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.645773540058641,
+      "learning_rate": 1.939304665790108e-05,
+      "loss": 0.8456,
+      "step": 615
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4678638644987096,
+      "learning_rate": 1.9390546299094567e-05,
+      "loss": 0.747,
+      "step": 616
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6192198614565292,
+      "learning_rate": 1.9388040962585003e-05,
+      "loss": 0.8168,
+      "step": 617
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4959580078577437,
+      "learning_rate": 1.9385530649700418e-05,
+      "loss": 0.7502,
+      "step": 618
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4561942566165975,
+      "learning_rate": 1.938301536177146e-05,
+      "loss": 0.7974,
+      "step": 619
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.681820928363132,
+      "learning_rate": 1.9380495100131425e-05,
+      "loss": 1.5175,
+      "step": 620
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4695589531711677,
+      "learning_rate": 1.9377969866116242e-05,
+      "loss": 0.8136,
+      "step": 621
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5696909275390383,
+      "learning_rate": 1.9375439661064483e-05,
+      "loss": 0.7549,
+      "step": 622
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.397818631161825,
+      "learning_rate": 1.9372904486317343e-05,
+      "loss": 0.7784,
+      "step": 623
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5577266319165128,
+      "learning_rate": 1.9370364343218656e-05,
+      "loss": 0.8036,
+      "step": 624
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.573003280422721,
+      "learning_rate": 1.9367819233114895e-05,
+      "loss": 0.8219,
+      "step": 625
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.361237026393405,
+      "learning_rate": 1.936526915735516e-05,
+      "loss": 1.4322,
+      "step": 626
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4217050602699572,
+      "learning_rate": 1.9362714117291185e-05,
+      "loss": 0.779,
+      "step": 627
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.51830417022511,
+      "learning_rate": 1.9360154114277337e-05,
+      "loss": 0.703,
+      "step": 628
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5097391455050264,
+      "learning_rate": 1.935758914967061e-05,
+      "loss": 0.9269,
+      "step": 629
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5699052762270926,
+      "learning_rate": 1.935501922483063e-05,
+      "loss": 0.777,
+      "step": 630
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0879776485843626,
+      "learning_rate": 1.9352444341119655e-05,
+      "loss": 1.3736,
+      "step": 631
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4399571072385715,
+      "learning_rate": 1.9349864499902563e-05,
+      "loss": 0.7717,
+      "step": 632
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5333841091353977,
+      "learning_rate": 1.9347279702546876e-05,
+      "loss": 0.773,
+      "step": 633
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6551568177829998,
+      "learning_rate": 1.9344689950422722e-05,
+      "loss": 0.7805,
+      "step": 634
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4792879389910505,
+      "learning_rate": 1.9342095244902875e-05,
+      "loss": 0.7636,
+      "step": 635
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4932039881540562,
+      "learning_rate": 1.9339495587362722e-05,
+      "loss": 0.7283,
+      "step": 636
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5208195729856286,
+      "learning_rate": 1.933689097918028e-05,
+      "loss": 0.7706,
+      "step": 637
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4660122163757368,
+      "learning_rate": 1.9334281421736187e-05,
+      "loss": 0.8073,
+      "step": 638
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4729173289747284,
+      "learning_rate": 1.9331666916413708e-05,
+      "loss": 0.7411,
+      "step": 639
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.573373189434525,
+      "learning_rate": 1.932904746459873e-05,
+      "loss": 0.7697,
+      "step": 640
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.339263766244011,
+      "learning_rate": 1.9326423067679764e-05,
+      "loss": 1.3527,
+      "step": 641
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5507389062636263,
+      "learning_rate": 1.9323793727047935e-05,
+      "loss": 0.7633,
+      "step": 642
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5498313498602132,
+      "learning_rate": 1.9321159444096998e-05,
+      "loss": 0.8364,
+      "step": 643
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6412514649995613,
+      "learning_rate": 1.9318520220223318e-05,
+      "loss": 0.809,
+      "step": 644
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.497480723311617,
+      "learning_rate": 1.931587605682589e-05,
+      "loss": 0.761,
+      "step": 645
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.458626403061108,
+      "learning_rate": 1.9313226955306314e-05,
+      "loss": 0.7717,
+      "step": 646
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.299889010435605,
+      "learning_rate": 1.9310572917068824e-05,
+      "loss": 1.4404,
+      "step": 647
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.154190802148305,
+      "learning_rate": 1.9307913943520257e-05,
+      "loss": 1.4587,
+      "step": 648
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5099767468885867,
+      "learning_rate": 1.9305250036070067e-05,
+      "loss": 0.8117,
+      "step": 649
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.1107320698573147,
+      "learning_rate": 1.9302581196130335e-05,
+      "loss": 1.451,
+      "step": 650
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5654461118661485,
+      "learning_rate": 1.9299907425115747e-05,
+      "loss": 0.8323,
+      "step": 651
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4471382090936074,
+      "learning_rate": 1.92972287244436e-05,
+      "loss": 0.7668,
+      "step": 652
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.510992277134428,
+      "learning_rate": 1.9294545095533817e-05,
+      "loss": 0.8047,
+      "step": 653
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5797086225115085,
+      "learning_rate": 1.9291856539808917e-05,
+      "loss": 0.7539,
+      "step": 654
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6112298887232748,
+      "learning_rate": 1.9289163058694044e-05,
+      "loss": 0.9185,
+      "step": 655
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.3896432338433033,
+      "learning_rate": 1.9286464653616942e-05,
+      "loss": 1.4024,
+      "step": 656
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.274330089338529,
+      "learning_rate": 1.9283761326007982e-05,
+      "loss": 1.4174,
+      "step": 657
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.131987562475883,
+      "learning_rate": 1.9281053077300124e-05,
+      "loss": 1.4386,
+      "step": 658
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5608255300063383,
+      "learning_rate": 1.9278339908928943e-05,
+      "loss": 0.8242,
+      "step": 659
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6759407520231744,
+      "learning_rate": 1.927562182233263e-05,
+      "loss": 0.9003,
+      "step": 660
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.591721251000736,
+      "learning_rate": 1.927289881895198e-05,
+      "loss": 1.3947,
+      "step": 661
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.359697588871499,
+      "learning_rate": 1.9270170900230386e-05,
+      "loss": 1.422,
+      "step": 662
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4824595028807044,
+      "learning_rate": 1.926743806761385e-05,
+      "loss": 0.8228,
+      "step": 663
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5548115301991305,
+      "learning_rate": 1.9264700322550992e-05,
+      "loss": 0.8595,
+      "step": 664
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.391726209214158,
+      "learning_rate": 1.9261957666493013e-05,
+      "loss": 0.7194,
+      "step": 665
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6652415034943275,
+      "learning_rate": 1.9259210100893737e-05,
+      "loss": 0.8527,
+      "step": 666
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.474026310124931,
+      "learning_rate": 1.9256457627209578e-05,
+      "loss": 0.8035,
+      "step": 667
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4583818662119825,
+      "learning_rate": 1.9253700246899562e-05,
+      "loss": 0.8951,
+      "step": 668
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5488445245742501,
+      "learning_rate": 1.9250937961425303e-05,
+      "loss": 0.8568,
+      "step": 669
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6804776400415842,
+      "learning_rate": 1.924817077225102e-05,
+      "loss": 0.8966,
+      "step": 670
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.43115019769534,
+      "learning_rate": 1.9245398680843546e-05,
+      "loss": 0.8394,
+      "step": 671
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.9531267925044684,
+      "learning_rate": 1.924262168867229e-05,
+      "loss": 1.3899,
+      "step": 672
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.3682036224256713,
+      "learning_rate": 1.9239839797209268e-05,
+      "loss": 1.4674,
+      "step": 673
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.1263408244028748,
+      "learning_rate": 1.9237053007929103e-05,
+      "loss": 1.3945,
+      "step": 674
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.368497122822958,
+      "learning_rate": 1.9234261322309e-05,
+      "loss": 1.4748,
+      "step": 675
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.702679430987513,
+      "learning_rate": 1.9231464741828757e-05,
+      "loss": 0.775,
+      "step": 676
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7266126125009111,
+      "learning_rate": 1.9228663267970785e-05,
+      "loss": 0.7272,
+      "step": 677
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8557927221488664,
+      "learning_rate": 1.9225856902220076e-05,
+      "loss": 0.771,
+      "step": 678
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5897536227985263,
+      "learning_rate": 1.9223045646064215e-05,
+      "loss": 0.7901,
+      "step": 679
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6129775064374505,
+      "learning_rate": 1.9220229500993377e-05,
+      "loss": 0.8026,
+      "step": 680
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4728607334961727,
+      "learning_rate": 1.921740846850034e-05,
+      "loss": 0.757,
+      "step": 681
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.5333769306798675,
+      "learning_rate": 1.9214582550080463e-05,
+      "loss": 1.4001,
+      "step": 682
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6519661459609436,
+      "learning_rate": 1.9211751747231697e-05,
+      "loss": 0.8639,
+      "step": 683
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6083783620698617,
+      "learning_rate": 1.9208916061454584e-05,
+      "loss": 0.726,
+      "step": 684
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5912764508412136,
+      "learning_rate": 1.920607549425225e-05,
+      "loss": 0.7568,
+      "step": 685
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6286304034650065,
+      "learning_rate": 1.9203230047130416e-05,
+      "loss": 0.8731,
+      "step": 686
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5574660271871337,
+      "learning_rate": 1.920037972159738e-05,
+      "loss": 0.8769,
+      "step": 687
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5225433766558771,
+      "learning_rate": 1.9197524519164033e-05,
+      "loss": 0.8608,
+      "step": 688
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.274086658944761,
+      "learning_rate": 1.9194664441343853e-05,
+      "loss": 1.4411,
+      "step": 689
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.076322745431708,
+      "learning_rate": 1.9191799489652893e-05,
+      "loss": 1.389,
+      "step": 690
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3801156923855837,
+      "learning_rate": 1.9188929665609797e-05,
+      "loss": 0.7235,
+      "step": 691
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6501854190451384,
+      "learning_rate": 1.9186054970735797e-05,
+      "loss": 0.8154,
+      "step": 692
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.519751513075953,
+      "learning_rate": 1.9183175406554687e-05,
+      "loss": 0.7689,
+      "step": 693
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5305523478282306,
+      "learning_rate": 1.9180290974592863e-05,
+      "loss": 0.8155,
+      "step": 694
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5759264039975711,
+      "learning_rate": 1.9177401676379288e-05,
+      "loss": 0.8563,
+      "step": 695
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4658555239314546,
+      "learning_rate": 1.9174507513445514e-05,
+      "loss": 0.7974,
+      "step": 696
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4271124522677647,
+      "learning_rate": 1.9171608487325664e-05,
+      "loss": 0.7248,
+      "step": 697
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6093718840250788,
+      "learning_rate": 1.916870459955645e-05,
+      "loss": 0.7807,
+      "step": 698
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.527685026187284,
+      "learning_rate": 1.916579585167714e-05,
+      "loss": 0.8274,
+      "step": 699
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4779949915328645,
+      "learning_rate": 1.9162882245229604e-05,
+      "loss": 0.8011,
+      "step": 700
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5255493666774154,
+      "learning_rate": 1.9159963781758266e-05,
+      "loss": 0.7085,
+      "step": 701
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.516587206199579,
+      "learning_rate": 1.9157040462810138e-05,
+      "loss": 0.7354,
+      "step": 702
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.585744033123811,
+      "learning_rate": 1.91541122899348e-05,
+      "loss": 0.8259,
+      "step": 703
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.621032886448328,
+      "learning_rate": 1.9151179264684402e-05,
+      "loss": 0.8314,
+      "step": 704
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5650209829605444,
+      "learning_rate": 1.9148241388613678e-05,
+      "loss": 0.8562,
+      "step": 705
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.459406942983186,
+      "learning_rate": 1.9145298663279923e-05,
+      "loss": 0.779,
+      "step": 706
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.489599915404078,
+      "learning_rate": 1.9142351090243e-05,
+      "loss": 0.7944,
+      "step": 707
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.411778901690178,
+      "learning_rate": 1.9139398671065352e-05,
+      "loss": 1.3475,
+      "step": 708
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.485037809074314,
+      "learning_rate": 1.9136441407311985e-05,
+      "loss": 0.755,
+      "step": 709
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4763374027672462,
+      "learning_rate": 1.913347930055047e-05,
+      "loss": 0.864,
+      "step": 710
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6524850269143778,
+      "learning_rate": 1.9130512352350953e-05,
+      "loss": 0.8379,
+      "step": 711
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4549160759340631,
+      "learning_rate": 1.9127540564286142e-05,
+      "loss": 0.7075,
+      "step": 712
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5134133078105538,
+      "learning_rate": 1.91245639379313e-05,
+      "loss": 0.8183,
+      "step": 713
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.430364093322305,
+      "learning_rate": 1.9121582474864283e-05,
+      "loss": 0.7323,
+      "step": 714
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.003051058304768,
+      "learning_rate": 1.9118596176665478e-05,
+      "loss": 1.3992,
+      "step": 715
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6405914934519865,
+      "learning_rate": 1.9115605044917856e-05,
+      "loss": 0.8095,
+      "step": 716
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3726904901334942,
+      "learning_rate": 1.9112609081206948e-05,
+      "loss": 0.7824,
+      "step": 717
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6000827740217733,
+      "learning_rate": 1.9109608287120834e-05,
+      "loss": 0.8269,
+      "step": 718
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.562018511275804,
+      "learning_rate": 1.9106602664250166e-05,
+      "loss": 0.7848,
+      "step": 719
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3956020439759043,
+      "learning_rate": 1.9103592214188157e-05,
+      "loss": 0.7562,
+      "step": 720
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.234093034835445,
+      "learning_rate": 1.9100576938530567e-05,
+      "loss": 1.4559,
+      "step": 721
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4322519543349037,
+      "learning_rate": 1.9097556838875726e-05,
+      "loss": 0.7637,
+      "step": 722
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9392132468359944,
+      "learning_rate": 1.9094531916824518e-05,
+      "loss": 1.4052,
+      "step": 723
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.538108456213687,
+      "learning_rate": 1.9091502173980375e-05,
+      "loss": 0.7566,
+      "step": 724
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.445867399918639,
+      "learning_rate": 1.9088467611949295e-05,
+      "loss": 0.7948,
+      "step": 725
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0956005206926513,
+      "learning_rate": 1.9085428232339828e-05,
+      "loss": 1.4422,
+      "step": 726
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5068295999367205,
+      "learning_rate": 1.908238403676307e-05,
+      "loss": 0.8518,
+      "step": 727
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.50078000457749,
+      "learning_rate": 1.9079335026832687e-05,
+      "loss": 0.774,
+      "step": 728
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4908676851086686,
+      "learning_rate": 1.9076281204164875e-05,
+      "loss": 0.7935,
+      "step": 729
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4029901808538257,
+      "learning_rate": 1.9073222570378396e-05,
+      "loss": 0.8699,
+      "step": 730
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4537537639601077,
+      "learning_rate": 1.907015912709456e-05,
+      "loss": 0.7422,
+      "step": 731
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.1329030552549435,
+      "learning_rate": 1.906709087593722e-05,
+      "loss": 1.4746,
+      "step": 732
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4924320697181994,
+      "learning_rate": 1.9064017818532785e-05,
+      "loss": 0.8228,
+      "step": 733
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.573423306725373,
+      "learning_rate": 1.906093995651021e-05,
+      "loss": 0.7968,
+      "step": 734
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4868323582239646,
+      "learning_rate": 1.9057857291501e-05,
+      "loss": 0.8611,
+      "step": 735
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5514373650181483,
+      "learning_rate": 1.9054769825139193e-05,
+      "loss": 0.8889,
+      "step": 736
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.533875128743377,
+      "learning_rate": 1.9051677559061382e-05,
+      "loss": 0.8356,
+      "step": 737
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.571555482119612,
+      "learning_rate": 1.904858049490671e-05,
+      "loss": 0.8355,
+      "step": 738
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.400385786884392,
+      "learning_rate": 1.904547863431685e-05,
+      "loss": 0.7679,
+      "step": 739
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5400222488518271,
+      "learning_rate": 1.904237197893602e-05,
+      "loss": 0.8949,
+      "step": 740
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.449220133074422,
+      "learning_rate": 1.9039260530410998e-05,
+      "loss": 0.8014,
+      "step": 741
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5334564559868649,
+      "learning_rate": 1.9036144290391073e-05,
+      "loss": 0.8252,
+      "step": 742
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5104423135161806,
+      "learning_rate": 1.90330232605281e-05,
+      "loss": 0.8297,
+      "step": 743
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.09425233872391,
+      "learning_rate": 1.9029897442476453e-05,
+      "loss": 1.4049,
+      "step": 744
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5202676307833487,
+      "learning_rate": 1.9026766837893058e-05,
+      "loss": 0.8603,
+      "step": 745
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4828695152722084,
+      "learning_rate": 1.9023631448437376e-05,
+      "loss": 0.7746,
+      "step": 746
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.423932800072073,
+      "learning_rate": 1.902049127577139e-05,
+      "loss": 0.8847,
+      "step": 747
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4524164691146075,
+      "learning_rate": 1.9017346321559644e-05,
+      "loss": 0.7508,
+      "step": 748
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5448689605091743,
+      "learning_rate": 1.9014196587469195e-05,
+      "loss": 0.859,
+      "step": 749
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4996428298330202,
+      "learning_rate": 1.9011042075169644e-05,
+      "loss": 0.8148,
+      "step": 750
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4460120146452946,
+      "learning_rate": 1.9007882786333122e-05,
+      "loss": 0.8324,
+      "step": 751
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.2121302180320717,
+      "learning_rate": 1.9004718722634287e-05,
+      "loss": 1.3634,
+      "step": 752
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6244280823274657,
+      "learning_rate": 1.9001549885750344e-05,
+      "loss": 0.8916,
+      "step": 753
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5193055392242298,
+      "learning_rate": 1.8998376277361008e-05,
+      "loss": 0.8176,
+      "step": 754
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.1355318993753216,
+      "learning_rate": 1.8995197899148536e-05,
+      "loss": 1.4496,
+      "step": 755
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4553923428549957,
+      "learning_rate": 1.8992014752797708e-05,
+      "loss": 0.7893,
+      "step": 756
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7318384730035081,
+      "learning_rate": 1.8988826839995836e-05,
+      "loss": 0.9131,
+      "step": 757
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5700111647521744,
+      "learning_rate": 1.8985634162432757e-05,
+      "loss": 0.7777,
+      "step": 758
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5399914682788631,
+      "learning_rate": 1.898243672180083e-05,
+      "loss": 0.8524,
+      "step": 759
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4953454282902172,
+      "learning_rate": 1.897923451979494e-05,
+      "loss": 0.7535,
+      "step": 760
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.3352105356646646,
+      "learning_rate": 1.8976027558112502e-05,
+      "loss": 1.3857,
+      "step": 761
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5169563673684157,
+      "learning_rate": 1.8972815838453445e-05,
+      "loss": 0.8222,
+      "step": 762
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5705730299534537,
+      "learning_rate": 1.8969599362520224e-05,
+      "loss": 0.8599,
+      "step": 763
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.143223892023832,
+      "learning_rate": 1.896637813201782e-05,
+      "loss": 1.3781,
+      "step": 764
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4824902671748983,
+      "learning_rate": 1.896315214865373e-05,
+      "loss": 0.8247,
+      "step": 765
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.587233601019541,
+      "learning_rate": 1.895992141413797e-05,
+      "loss": 0.7611,
+      "step": 766
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.483108570309584,
+      "learning_rate": 1.8956685930183068e-05,
+      "loss": 0.8184,
+      "step": 767
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4201643667502835,
+      "learning_rate": 1.8953445698504084e-05,
+      "loss": 0.6864,
+      "step": 768
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4336893961290291,
+      "learning_rate": 1.8950200720818584e-05,
+      "loss": 0.7948,
+      "step": 769
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.3468094677629914,
+      "learning_rate": 1.8946950998846652e-05,
+      "loss": 1.4692,
+      "step": 770
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4612627182672215,
+      "learning_rate": 1.8943696534310892e-05,
+      "loss": 0.7565,
+      "step": 771
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.2382088386539496,
+      "learning_rate": 1.894043732893641e-05,
+      "loss": 1.4337,
+      "step": 772
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5121136044267613,
+      "learning_rate": 1.8937173384450838e-05,
+      "loss": 0.754,
+      "step": 773
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.2119660497753504,
+      "learning_rate": 1.8933904702584315e-05,
+      "loss": 1.4774,
+      "step": 774
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.144754041787048,
+      "learning_rate": 1.8930631285069487e-05,
+      "loss": 1.4124,
+      "step": 775
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.165169897607347,
+      "learning_rate": 1.8927353133641513e-05,
+      "loss": 1.3683,
+      "step": 776
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.08072361873932,
+      "learning_rate": 1.892407025003807e-05,
+      "loss": 1.4691,
+      "step": 777
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4717824060712765,
+      "learning_rate": 1.8920782635999328e-05,
+      "loss": 0.7597,
+      "step": 778
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4076440913314894,
+      "learning_rate": 1.8917490293267976e-05,
+      "loss": 0.8375,
+      "step": 779
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5565588375372155,
+      "learning_rate": 1.8914193223589207e-05,
+      "loss": 0.73,
+      "step": 780
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4994623943952936,
+      "learning_rate": 1.891089142871071e-05,
+      "loss": 0.718,
+      "step": 781
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5157365533413731,
+      "learning_rate": 1.8907584910382702e-05,
+      "loss": 0.727,
+      "step": 782
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4719729810259712,
+      "learning_rate": 1.8904273670357874e-05,
+      "loss": 0.8546,
+      "step": 783
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6121550615474436,
+      "learning_rate": 1.890095771039145e-05,
+      "loss": 0.8045,
+      "step": 784
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4474188894284357,
+      "learning_rate": 1.889763703224112e-05,
+      "loss": 0.8172,
+      "step": 785
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.910484808919811,
+      "learning_rate": 1.8894311637667114e-05,
+      "loss": 1.4608,
+      "step": 786
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5634775189964707,
+      "learning_rate": 1.889098152843214e-05,
+      "loss": 0.822,
+      "step": 787
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.698351808273585,
+      "learning_rate": 1.8887646706301406e-05,
+      "loss": 0.7988,
+      "step": 788
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.391611566216612,
+      "learning_rate": 1.888430717304262e-05,
+      "loss": 1.4351,
+      "step": 789
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.06919880905171,
+      "learning_rate": 1.8880962930425987e-05,
+      "loss": 1.4283,
+      "step": 790
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6783080361718008,
+      "learning_rate": 1.8877613980224215e-05,
+      "loss": 0.725,
+      "step": 791
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5463639871853796,
+      "learning_rate": 1.8874260324212503e-05,
+      "loss": 0.7682,
+      "step": 792
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.644454044155057,
+      "learning_rate": 1.887090196416854e-05,
+      "loss": 0.8427,
+      "step": 793
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5434917682980027,
+      "learning_rate": 1.886753890187251e-05,
+      "loss": 0.7512,
+      "step": 794
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.314895547537662,
+      "learning_rate": 1.886417113910709e-05,
+      "loss": 1.377,
+      "step": 795
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.1693160233532667,
+      "learning_rate": 1.8860798677657463e-05,
+      "loss": 1.4347,
+      "step": 796
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4478704315576947,
+      "learning_rate": 1.8857421519311273e-05,
+      "loss": 0.773,
+      "step": 797
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.2502981429670017,
+      "learning_rate": 1.885403966585868e-05,
+      "loss": 1.4486,
+      "step": 798
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5170656039799628,
+      "learning_rate": 1.8850653119092325e-05,
+      "loss": 0.8161,
+      "step": 799
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4544929281100707,
+      "learning_rate": 1.8847261880807327e-05,
+      "loss": 0.7777,
+      "step": 800
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5249538216428626,
+      "learning_rate": 1.884386595280131e-05,
+      "loss": 0.8039,
+      "step": 801
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.491110797276478,
+      "learning_rate": 1.8840465336874362e-05,
+      "loss": 0.7507,
+      "step": 802
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.734971076070025,
+      "learning_rate": 1.883706003482908e-05,
+      "loss": 0.7999,
+      "step": 803
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.5101330288197152,
+      "learning_rate": 1.8833650048470524e-05,
+      "loss": 1.4439,
+      "step": 804
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5742841328554997,
+      "learning_rate": 1.8830235379606252e-05,
+      "loss": 0.8487,
+      "step": 805
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5661500790768759,
+      "learning_rate": 1.8826816030046293e-05,
+      "loss": 0.8,
+      "step": 806
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5391532425266192,
+      "learning_rate": 1.8823392001603164e-05,
+      "loss": 0.8006,
+      "step": 807
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.538664627148793,
+      "learning_rate": 1.8819963296091857e-05,
+      "loss": 0.813,
+      "step": 808
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.100551998768598,
+      "learning_rate": 1.8816529915329855e-05,
+      "loss": 1.3874,
+      "step": 809
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4825746783805256,
+      "learning_rate": 1.8813091861137106e-05,
+      "loss": 0.7624,
+      "step": 810
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.2695750688040874,
+      "learning_rate": 1.8809649135336035e-05,
+      "loss": 1.4469,
+      "step": 811
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6278611793271576,
+      "learning_rate": 1.8806201739751555e-05,
+      "loss": 0.788,
+      "step": 812
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4120029338272833,
+      "learning_rate": 1.8802749676211046e-05,
+      "loss": 0.7545,
+      "step": 813
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4552483154792863,
+      "learning_rate": 1.8799292946544363e-05,
+      "loss": 0.8216,
+      "step": 814
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4686719948981424,
+      "learning_rate": 1.8795831552583832e-05,
+      "loss": 0.754,
+      "step": 815
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5640963389823765,
+      "learning_rate": 1.879236549616426e-05,
+      "loss": 0.7874,
+      "step": 816
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.1082494988207365,
+      "learning_rate": 1.878889477912292e-05,
+      "loss": 1.379,
+      "step": 817
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.058600074172232,
+      "learning_rate": 1.8785419403299556e-05,
+      "loss": 1.4755,
+      "step": 818
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0216782712961794,
+      "learning_rate": 1.8781939370536376e-05,
+      "loss": 1.4395,
+      "step": 819
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.136343484740886,
+      "learning_rate": 1.877845468267807e-05,
+      "loss": 1.4623,
+      "step": 820
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5911854114626123,
+      "learning_rate": 1.877496534157178e-05,
+      "loss": 0.8191,
+      "step": 821
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.3126383851382317,
+      "learning_rate": 1.8771471349067125e-05,
+      "loss": 1.4211,
+      "step": 822
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6545160012324291,
+      "learning_rate": 1.8767972707016186e-05,
+      "loss": 0.8502,
+      "step": 823
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5698361314402953,
+      "learning_rate": 1.8764469417273507e-05,
+      "loss": 0.8346,
+      "step": 824
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.740895054709823,
+      "learning_rate": 1.8760961481696103e-05,
+      "loss": 0.8106,
+      "step": 825
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6568598445551208,
+      "learning_rate": 1.8757448902143444e-05,
+      "loss": 0.7924,
+      "step": 826
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5074798706699875,
+      "learning_rate": 1.875393168047746e-05,
+      "loss": 0.7896,
+      "step": 827
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4927712724544862,
+      "learning_rate": 1.875040981856255e-05,
+      "loss": 0.693,
+      "step": 828
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.419708997120899,
+      "learning_rate": 1.874688331826557e-05,
+      "loss": 1.4273,
+      "step": 829
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.3483395293970384,
+      "learning_rate": 1.874335218145583e-05,
+      "loss": 1.3406,
+      "step": 830
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5374138377651085,
+      "learning_rate": 1.87398164100051e-05,
+      "loss": 0.785,
+      "step": 831
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6503579672182345,
+      "learning_rate": 1.873627600578761e-05,
+      "loss": 0.8359,
+      "step": 832
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5215269966429896,
+      "learning_rate": 1.8732730970680042e-05,
+      "loss": 0.7766,
+      "step": 833
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6886276912093645,
+      "learning_rate": 1.8729181306561534e-05,
+      "loss": 0.8645,
+      "step": 834
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5457781582012966,
+      "learning_rate": 1.872562701531368e-05,
+      "loss": 0.8036,
+      "step": 835
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.647608472120829,
+      "learning_rate": 1.872206809882052e-05,
+      "loss": 0.7768,
+      "step": 836
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.395010540802818,
+      "learning_rate": 1.8718504558968553e-05,
+      "loss": 0.7174,
+      "step": 837
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.510237059223901,
+      "learning_rate": 1.8714936397646727e-05,
+      "loss": 0.7563,
+      "step": 838
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.7087919530238604,
+      "learning_rate": 1.8711363616746438e-05,
+      "loss": 1.4746,
+      "step": 839
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.3939725505716623,
+      "learning_rate": 1.8707786218161527e-05,
+      "loss": 1.3864,
+      "step": 840
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6634244228554549,
+      "learning_rate": 1.870420420378829e-05,
+      "loss": 0.7779,
+      "step": 841
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5691330344902572,
+      "learning_rate": 1.870061757552547e-05,
+      "loss": 0.7763,
+      "step": 842
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.543810339277956,
+      "learning_rate": 1.869702633527425e-05,
+      "loss": 0.7837,
+      "step": 843
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4905202765214347,
+      "learning_rate": 1.8693430484938256e-05,
+      "loss": 0.6352,
+      "step": 844
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5462847869483423,
+      "learning_rate": 1.8689830026423566e-05,
+      "loss": 0.7405,
+      "step": 845
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.7135993184560885,
+      "learning_rate": 1.8686224961638704e-05,
+      "loss": 1.4087,
+      "step": 846
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.647679391615739,
+      "learning_rate": 1.8682615292494613e-05,
+      "loss": 0.8571,
+      "step": 847
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4468210569041369,
+      "learning_rate": 1.86790010209047e-05,
+      "loss": 0.7207,
+      "step": 848
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4961570266120165,
+      "learning_rate": 1.8675382148784805e-05,
+      "loss": 0.8207,
+      "step": 849
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.520606159729054,
+      "learning_rate": 1.8671758678053206e-05,
+      "loss": 0.7604,
+      "step": 850
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4779167344705053,
+      "learning_rate": 1.8668130610630613e-05,
+      "loss": 0.7462,
+      "step": 851
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.3361224637118316,
+      "learning_rate": 1.8664497948440178e-05,
+      "loss": 1.3573,
+      "step": 852
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.505804427492263,
+      "learning_rate": 1.866086069340749e-05,
+      "loss": 0.734,
+      "step": 853
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7157397437826059,
+      "learning_rate": 1.865721884746057e-05,
+      "loss": 0.7847,
+      "step": 854
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5849682105196845,
+      "learning_rate": 1.8653572412529874e-05,
+      "loss": 0.7925,
+      "step": 855
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4673793710616556,
+      "learning_rate": 1.8649921390548288e-05,
+      "loss": 0.7694,
+      "step": 856
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5463799272970784,
+      "learning_rate": 1.8646265783451133e-05,
+      "loss": 0.8153,
+      "step": 857
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.1379341218809724,
+      "learning_rate": 1.8642605593176157e-05,
+      "loss": 1.4229,
+      "step": 858
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5489587791598687,
+      "learning_rate": 1.8638940821663536e-05,
+      "loss": 0.7032,
+      "step": 859
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4191611828414816,
+      "learning_rate": 1.8635271470855883e-05,
+      "loss": 0.7096,
+      "step": 860
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5148669974809248,
+      "learning_rate": 1.8631597542698232e-05,
+      "loss": 0.8495,
+      "step": 861
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5463676219250369,
+      "learning_rate": 1.8627919039138042e-05,
+      "loss": 0.8946,
+      "step": 862
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5379759975575005,
+      "learning_rate": 1.8624235962125195e-05,
+      "loss": 0.7543,
+      "step": 863
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.595441536626647,
+      "learning_rate": 1.862054831361201e-05,
+      "loss": 0.8925,
+      "step": 864
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0060629025242602,
+      "learning_rate": 1.861685609555322e-05,
+      "loss": 1.4602,
+      "step": 865
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.1102471671483523,
+      "learning_rate": 1.861315930990597e-05,
+      "loss": 1.4319,
+      "step": 866
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3528200441750333,
+      "learning_rate": 1.8609457958629854e-05,
+      "loss": 0.7588,
+      "step": 867
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.459347656108802,
+      "learning_rate": 1.860575204368686e-05,
+      "loss": 0.8161,
+      "step": 868
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4528630527610797,
+      "learning_rate": 1.86020415670414e-05,
+      "loss": 0.831,
+      "step": 869
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5583713305799303,
+      "learning_rate": 1.859832653066032e-05,
+      "loss": 0.7716,
+      "step": 870
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.074360074973386,
+      "learning_rate": 1.8594606936512864e-05,
+      "loss": 1.4366,
+      "step": 871
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4590722504848146,
+      "learning_rate": 1.8590882786570698e-05,
+      "loss": 0.8079,
+      "step": 872
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4933741894908397,
+      "learning_rate": 1.858715408280791e-05,
+      "loss": 0.8236,
+      "step": 873
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4559425472937406,
+      "learning_rate": 1.8583420827200996e-05,
+      "loss": 0.7606,
+      "step": 874
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4351575373709171,
+      "learning_rate": 1.857968302172886e-05,
+      "loss": 0.72,
+      "step": 875
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4259778111575032,
+      "learning_rate": 1.857594066837283e-05,
+      "loss": 0.7729,
+      "step": 876
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.204752166356184,
+      "learning_rate": 1.8572193769116632e-05,
+      "loss": 1.4028,
+      "step": 877
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4256063096681568,
+      "learning_rate": 1.8568442325946408e-05,
+      "loss": 0.6881,
+      "step": 878
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4729870062849435,
+      "learning_rate": 1.8564686340850707e-05,
+      "loss": 0.7898,
+      "step": 879
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.967210674810177,
+      "learning_rate": 1.8560925815820493e-05,
+      "loss": 1.3615,
+      "step": 880
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5448483603422178,
+      "learning_rate": 1.855716075284912e-05,
+      "loss": 0.7892,
+      "step": 881
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5862711062602368,
+      "learning_rate": 1.8553391153932364e-05,
+      "loss": 0.8409,
+      "step": 882
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.163311188914557,
+      "learning_rate": 1.8549617021068402e-05,
+      "loss": 1.4066,
+      "step": 883
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.473851600053938,
+      "learning_rate": 1.8545838356257804e-05,
+      "loss": 0.8244,
+      "step": 884
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5589874355621813,
+      "learning_rate": 1.8542055161503552e-05,
+      "loss": 0.829,
+      "step": 885
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4632304186184197,
+      "learning_rate": 1.8538267438811027e-05,
+      "loss": 0.816,
+      "step": 886
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.506058199149983,
+      "learning_rate": 1.853447519018801e-05,
+      "loss": 0.769,
+      "step": 887
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4055345410812117,
+      "learning_rate": 1.8530678417644677e-05,
+      "loss": 0.7835,
+      "step": 888
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5634100494275163,
+      "learning_rate": 1.8526877123193615e-05,
+      "loss": 0.8123,
+      "step": 889
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.142263478007544,
+      "learning_rate": 1.852307130884979e-05,
+      "loss": 1.3656,
+      "step": 890
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.2174548228776993,
+      "learning_rate": 1.851926097663058e-05,
+      "loss": 1.352,
+      "step": 891
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.020732393748382,
+      "learning_rate": 1.8515446128555744e-05,
+      "loss": 1.405,
+      "step": 892
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.590541702004956,
+      "learning_rate": 1.8511626766647442e-05,
+      "loss": 0.8102,
+      "step": 893
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6956247900892891,
+      "learning_rate": 1.8507802892930232e-05,
+      "loss": 0.7486,
+      "step": 894
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4687423447053953,
+      "learning_rate": 1.8503974509431056e-05,
+      "loss": 0.7719,
+      "step": 895
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.516133727278507,
+      "learning_rate": 1.8500141618179242e-05,
+      "loss": 0.8076,
+      "step": 896
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5384571639965987,
+      "learning_rate": 1.849630422120652e-05,
+      "loss": 0.7706,
+      "step": 897
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.312032695567281,
+      "learning_rate": 1.8492462320546997e-05,
+      "loss": 1.3763,
+      "step": 898
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4412646295139926,
+      "learning_rate": 1.8488615918237175e-05,
+      "loss": 0.6415,
+      "step": 899
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3318924516625492,
+      "learning_rate": 1.848476501631594e-05,
+      "loss": 0.7237,
+      "step": 900
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.1663950833885828,
+      "learning_rate": 1.848090961682456e-05,
+      "loss": 1.4101,
+      "step": 901
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.394040273968732,
+      "learning_rate": 1.8477049721806695e-05,
+      "loss": 0.7772,
+      "step": 902
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.311387764891183,
+      "learning_rate": 1.847318533330838e-05,
+      "loss": 0.6988,
+      "step": 903
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4705804603858856,
+      "learning_rate": 1.846931645337803e-05,
+      "loss": 0.7605,
+      "step": 904
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3475047032921448,
+      "learning_rate": 1.846544308406645e-05,
+      "loss": 0.7193,
+      "step": 905
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.257789663046569,
+      "learning_rate": 1.846156522742682e-05,
+      "loss": 1.3707,
+      "step": 906
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.215654907696301,
+      "learning_rate": 1.8457682885514696e-05,
+      "loss": 1.4608,
+      "step": 907
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5641431976883646,
+      "learning_rate": 1.8453796060388016e-05,
+      "loss": 0.79,
+      "step": 908
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0619449311253675,
+      "learning_rate": 1.8449904754107092e-05,
+      "loss": 1.4745,
+      "step": 909
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4736000121163906,
+      "learning_rate": 1.8446008968734615e-05,
+      "loss": 0.8047,
+      "step": 910
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.4854415988520917,
+      "learning_rate": 1.8442108706335648e-05,
+      "loss": 1.3847,
+      "step": 911
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0196540111607812,
+      "learning_rate": 1.8438203968977618e-05,
+      "loss": 1.4,
+      "step": 912
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4286169109191527,
+      "learning_rate": 1.843429475873034e-05,
+      "loss": 0.8281,
+      "step": 913
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.0739232655437743,
+      "learning_rate": 1.843038107766599e-05,
+      "loss": 1.3625,
+      "step": 914
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5916408184454927,
+      "learning_rate": 1.8426462927859115e-05,
+      "loss": 0.7327,
+      "step": 915
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5584975174221876,
+      "learning_rate": 1.8422540311386633e-05,
+      "loss": 0.8351,
+      "step": 916
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4497176758662256,
+      "learning_rate": 1.841861323032783e-05,
+      "loss": 0.7892,
+      "step": 917
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4308200217179563,
+      "learning_rate": 1.841468168676436e-05,
+      "loss": 0.7443,
+      "step": 918
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6474227553092236,
+      "learning_rate": 1.8410745682780234e-05,
+      "loss": 0.8266,
+      "step": 919
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.6697094961497445,
+      "learning_rate": 1.8406805220461834e-05,
+      "loss": 1.3429,
+      "step": 920
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5468908046786194,
+      "learning_rate": 1.8402860301897908e-05,
+      "loss": 0.7868,
+      "step": 921
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.361218715823616,
+      "learning_rate": 1.839891092917956e-05,
+      "loss": 0.7382,
+      "step": 922
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5200446255188444,
+      "learning_rate": 1.8394957104400258e-05,
+      "loss": 0.7599,
+      "step": 923
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6463758892627796,
+      "learning_rate": 1.8390998829655834e-05,
+      "loss": 0.9007,
+      "step": 924
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.327777143446975,
+      "learning_rate": 1.8387036107044467e-05,
+      "loss": 1.4141,
+      "step": 925
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4124814514873345,
+      "learning_rate": 1.8383068938666702e-05,
+      "loss": 0.7469,
+      "step": 926
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5458445729716188,
+      "learning_rate": 1.837909732662545e-05,
+      "loss": 0.8213,
+      "step": 927
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4012907310688987,
+      "learning_rate": 1.837512127302595e-05,
+      "loss": 0.6784,
+      "step": 928
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.2276001437799633,
+      "learning_rate": 1.8371140779975826e-05,
+      "loss": 1.4139,
+      "step": 929
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5671066481276439,
+      "learning_rate": 1.836715584958504e-05,
+      "loss": 0.8136,
+      "step": 930
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.137876090802446,
+      "learning_rate": 1.83631664839659e-05,
+      "loss": 1.4301,
+      "step": 931
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.1744762683988332,
+      "learning_rate": 1.8359172685233078e-05,
+      "loss": 1.4241,
+      "step": 932
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4059369638991173,
+      "learning_rate": 1.8355174455503598e-05,
+      "loss": 0.7895,
+      "step": 933
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.462763505694101,
+      "learning_rate": 1.8351171796896815e-05,
+      "loss": 0.7197,
+      "step": 934
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4046582976408926,
+      "learning_rate": 1.8347164711534443e-05,
+      "loss": 0.746,
+      "step": 935
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.0714084675401727,
+      "learning_rate": 1.8343153201540545e-05,
+      "loss": 1.3904,
+      "step": 936
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5090345152284592,
+      "learning_rate": 1.833913726904153e-05,
+      "loss": 0.7268,
+      "step": 937
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.53655484807384,
+      "learning_rate": 1.833511691616614e-05,
+      "loss": 0.7696,
+      "step": 938
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5921240892334925,
+      "learning_rate": 1.833109214504547e-05,
+      "loss": 0.795,
+      "step": 939
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5523533977726567,
+      "learning_rate": 1.8327062957812956e-05,
+      "loss": 0.862,
+      "step": 940
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.459929157630895,
+      "learning_rate": 1.832302935660437e-05,
+      "loss": 0.716,
+      "step": 941
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3831493416082887,
+      "learning_rate": 1.831899134355783e-05,
+      "loss": 0.8506,
+      "step": 942
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.1782708818280496,
+      "learning_rate": 1.8314948920813794e-05,
+      "loss": 1.4747,
+      "step": 943
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5596012572926496,
+      "learning_rate": 1.8310902090515045e-05,
+      "loss": 0.8258,
+      "step": 944
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4229998235578203,
+      "learning_rate": 1.830685085480671e-05,
+      "loss": 0.8719,
+      "step": 945
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3529094297319564,
+      "learning_rate": 1.8302795215836263e-05,
+      "loss": 0.7283,
+      "step": 946
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5479390472757972,
+      "learning_rate": 1.829873517575349e-05,
+      "loss": 0.8571,
+      "step": 947
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.401097110399311,
+      "learning_rate": 1.8294670736710522e-05,
+      "loss": 0.7666,
+      "step": 948
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.1165982089420834,
+      "learning_rate": 1.8290601900861823e-05,
+      "loss": 1.4603,
+      "step": 949
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5288898878676935,
+      "learning_rate": 1.8286528670364187e-05,
+      "loss": 0.8096,
+      "step": 950
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.0561404883654824,
+      "learning_rate": 1.8282451047376734e-05,
+      "loss": 1.4163,
+      "step": 951
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5194524317849218,
+      "learning_rate": 1.8278369034060912e-05,
+      "loss": 0.7866,
+      "step": 952
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9674892372994175,
+      "learning_rate": 1.82742826325805e-05,
+      "loss": 1.3546,
+      "step": 953
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.442439027936151,
+      "learning_rate": 1.8270191845101602e-05,
+      "loss": 0.7841,
+      "step": 954
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4980032549768278,
+      "learning_rate": 1.8266096673792643e-05,
+      "loss": 0.7838,
+      "step": 955
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6225271736905915,
+      "learning_rate": 1.826199712082438e-05,
+      "loss": 0.7699,
+      "step": 956
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5321007192259313,
+      "learning_rate": 1.8257893188369883e-05,
+      "loss": 0.776,
+      "step": 957
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.516703186221938,
+      "learning_rate": 1.825378487860455e-05,
+      "loss": 0.7981,
+      "step": 958
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4460449822923997,
+      "learning_rate": 1.8249672193706097e-05,
+      "loss": 0.7259,
+      "step": 959
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4709396288187313,
+      "learning_rate": 1.8245555135854558e-05,
+      "loss": 0.8165,
+      "step": 960
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5230768782143649,
+      "learning_rate": 1.824143370723229e-05,
+      "loss": 0.7864,
+      "step": 961
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4498664897098508,
+      "learning_rate": 1.8237307910023963e-05,
+      "loss": 0.786,
+      "step": 962
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3436206471696441,
+      "learning_rate": 1.8233177746416557e-05,
+      "loss": 0.7168,
+      "step": 963
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4382303533437546,
+      "learning_rate": 1.8229043218599382e-05,
+      "loss": 0.7518,
+      "step": 964
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.22365865045598,
+      "learning_rate": 1.822490432876404e-05,
+      "loss": 1.3923,
+      "step": 965
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9642328632874824,
+      "learning_rate": 1.822076107910447e-05,
+      "loss": 1.3635,
+      "step": 966
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5229650486956017,
+      "learning_rate": 1.82166134718169e-05,
+      "loss": 0.7781,
+      "step": 967
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.1192183995116167,
+      "learning_rate": 1.8212461509099882e-05,
+      "loss": 1.3644,
+      "step": 968
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5072433640149494,
+      "learning_rate": 1.8208305193154266e-05,
+      "loss": 0.809,
+      "step": 969
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.415782975817706,
+      "learning_rate": 1.8204144526183224e-05,
+      "loss": 0.7907,
+      "step": 970
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.4238438963187106,
+      "learning_rate": 1.8199979510392216e-05,
+      "loss": 1.3328,
+      "step": 971
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.134413581941555,
+      "learning_rate": 1.8195810147989026e-05,
+      "loss": 1.4169,
+      "step": 972
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6025864856389342,
+      "learning_rate": 1.8191636441183727e-05,
+      "loss": 0.8259,
+      "step": 973
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6819476462376721,
+      "learning_rate": 1.81874583921887e-05,
+      "loss": 0.8751,
+      "step": 974
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.572233581667663,
+      "learning_rate": 1.818327600321863e-05,
+      "loss": 0.7627,
+      "step": 975
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5755557737678394,
+      "learning_rate": 1.8179089276490504e-05,
+      "loss": 0.8406,
+      "step": 976
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3889011483887361,
+      "learning_rate": 1.81748982142236e-05,
+      "loss": 0.7537,
+      "step": 977
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.6134938908276446,
+      "learning_rate": 1.8170702818639502e-05,
+      "loss": 1.3799,
+      "step": 978
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.531228078089025,
+      "learning_rate": 1.816650309196209e-05,
+      "loss": 0.7345,
+      "step": 979
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5617499987057513,
+      "learning_rate": 1.816229903641754e-05,
+      "loss": 0.8715,
+      "step": 980
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3754695773536159,
+      "learning_rate": 1.815809065423431e-05,
+      "loss": 0.7849,
+      "step": 981
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5475410581270381,
+      "learning_rate": 1.8153877947643178e-05,
+      "loss": 0.7826,
+      "step": 982
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.51575793674937,
+      "learning_rate": 1.8149660918877185e-05,
+      "loss": 0.7687,
+      "step": 983
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5332601123470222,
+      "learning_rate": 1.8145439570171684e-05,
+      "loss": 0.8147,
+      "step": 984
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5166143922994992,
+      "learning_rate": 1.814121390376431e-05,
+      "loss": 0.7853,
+      "step": 985
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0532345900114817,
+      "learning_rate": 1.8136983921894982e-05,
+      "loss": 1.4008,
+      "step": 986
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4346263634875025,
+      "learning_rate": 1.813274962680592e-05,
+      "loss": 0.858,
+      "step": 987
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4320879362919843,
+      "learning_rate": 1.812851102074162e-05,
+      "loss": 0.724,
+      "step": 988
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.475217967021458,
+      "learning_rate": 1.8124268105948857e-05,
+      "loss": 0.782,
+      "step": 989
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4393282814105803,
+      "learning_rate": 1.8120020884676708e-05,
+      "loss": 0.7927,
+      "step": 990
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5132683254544201,
+      "learning_rate": 1.8115769359176517e-05,
+      "loss": 0.8779,
+      "step": 991
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5034803001335912,
+      "learning_rate": 1.811151353170192e-05,
+      "loss": 0.7784,
+      "step": 992
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.374043156163993,
+      "learning_rate": 1.8107253404508827e-05,
+      "loss": 0.6762,
+      "step": 993
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0745385575310724,
+      "learning_rate": 1.8102988979855428e-05,
+      "loss": 1.4053,
+      "step": 994
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5051862251153545,
+      "learning_rate": 1.8098720260002192e-05,
+      "loss": 0.7603,
+      "step": 995
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6048578065138692,
+      "learning_rate": 1.8094447247211867e-05,
+      "loss": 0.8267,
+      "step": 996
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4377124498333023,
+      "learning_rate": 1.8090169943749477e-05,
+      "loss": 0.7695,
+      "step": 997
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4423715526791476,
+      "learning_rate": 1.808588835188231e-05,
+      "loss": 0.8448,
+      "step": 998
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0254905944089256,
+      "learning_rate": 1.808160247387994e-05,
+      "loss": 1.3555,
+      "step": 999
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9980911568529902,
+      "learning_rate": 1.807731231201421e-05,
+      "loss": 1.4687,
+      "step": 1000
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4104065590448913,
+      "learning_rate": 1.8073017868559223e-05,
+      "loss": 0.6788,
+      "step": 1001
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5449863113182067,
+      "learning_rate": 1.8068719145791373e-05,
+      "loss": 0.8094,
+      "step": 1002
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.486203507267181,
+      "learning_rate": 1.80644161459893e-05,
+      "loss": 0.7642,
+      "step": 1003
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4130780279471435,
+      "learning_rate": 1.8060108871433922e-05,
+      "loss": 0.7914,
+      "step": 1004
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4703013651933985,
+      "learning_rate": 1.8055797324408427e-05,
+      "loss": 0.7769,
+      "step": 1005
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0296075952806523,
+      "learning_rate": 1.8051481507198256e-05,
+      "loss": 1.4318,
+      "step": 1006
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.46681651714035,
+      "learning_rate": 1.8047161422091126e-05,
+      "loss": 0.7672,
+      "step": 1007
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.494343655500784,
+      "learning_rate": 1.8042837071377004e-05,
+      "loss": 0.8159,
+      "step": 1008
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4463865676228238,
+      "learning_rate": 1.803850845734813e-05,
+      "loss": 0.8049,
+      "step": 1009
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5476537560790846,
+      "learning_rate": 1.8034175582299e-05,
+      "loss": 0.8289,
+      "step": 1010
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4348986092663878,
+      "learning_rate": 1.8029838448526357e-05,
+      "loss": 0.7315,
+      "step": 1011
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5084300678366855,
+      "learning_rate": 1.8025497058329224e-05,
+      "loss": 0.8104,
+      "step": 1012
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6392383746766497,
+      "learning_rate": 1.8021151414008856e-05,
+      "loss": 0.8179,
+      "step": 1013
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0585155061445577,
+      "learning_rate": 1.8016801517868784e-05,
+      "loss": 1.4094,
+      "step": 1014
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.555443063160448,
+      "learning_rate": 1.8012447372214777e-05,
+      "loss": 0.7934,
+      "step": 1015
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9864077253707746,
+      "learning_rate": 1.8008088979354864e-05,
+      "loss": 1.4098,
+      "step": 1016
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.393476669926877,
+      "learning_rate": 1.800372634159933e-05,
+      "loss": 0.6049,
+      "step": 1017
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9636588885299693,
+      "learning_rate": 1.79993594612607e-05,
+      "loss": 1.3954,
+      "step": 1018
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9866683867049655,
+      "learning_rate": 1.799498834065375e-05,
+      "loss": 1.3897,
+      "step": 1019
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7532895516287246,
+      "learning_rate": 1.799061298209551e-05,
+      "loss": 0.8199,
+      "step": 1020
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5198313089663607,
+      "learning_rate": 1.7986233387905255e-05,
+      "loss": 0.7059,
+      "step": 1021
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5289626944841417,
+      "learning_rate": 1.7981849560404498e-05,
+      "loss": 0.8011,
+      "step": 1022
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3724418606110007,
+      "learning_rate": 1.7977461501916998e-05,
+      "loss": 0.8351,
+      "step": 1023
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5790881594779334,
+      "learning_rate": 1.7973069214768766e-05,
+      "loss": 0.7407,
+      "step": 1024
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5248268708022634,
+      "learning_rate": 1.796867270128804e-05,
+      "loss": 0.8008,
+      "step": 1025
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4506308945463213,
+      "learning_rate": 1.7964271963805317e-05,
+      "loss": 0.7025,
+      "step": 1026
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4364621967373585,
+      "learning_rate": 1.7959867004653315e-05,
+      "loss": 0.7546,
+      "step": 1027
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.295497817305776,
+      "learning_rate": 1.7955457826167e-05,
+      "loss": 1.3834,
+      "step": 1028
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4683412931414532,
+      "learning_rate": 1.7951044430683567e-05,
+      "loss": 0.8222,
+      "step": 1029
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.485002077087134,
+      "learning_rate": 1.7946626820542456e-05,
+      "loss": 0.7395,
+      "step": 1030
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4078531728392143,
+      "learning_rate": 1.7942204998085336e-05,
+      "loss": 0.7042,
+      "step": 1031
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5837753408659572,
+      "learning_rate": 1.7937778965656103e-05,
+      "loss": 0.8166,
+      "step": 1032
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4630759309216796,
+      "learning_rate": 1.7933348725600896e-05,
+      "loss": 0.7843,
+      "step": 1033
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6247367354613127,
+      "learning_rate": 1.792891428026808e-05,
+      "loss": 0.8136,
+      "step": 1034
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9999901336707246,
+      "learning_rate": 1.792447563200824e-05,
+      "loss": 1.3699,
+      "step": 1035
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5757129934851433,
+      "learning_rate": 1.7920032783174207e-05,
+      "loss": 0.7957,
+      "step": 1036
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3682539698651923,
+      "learning_rate": 1.791558573612102e-05,
+      "loss": 0.803,
+      "step": 1037
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.012241529583455,
+      "learning_rate": 1.791113449320596e-05,
+      "loss": 1.4441,
+      "step": 1038
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.338597793305356,
+      "learning_rate": 1.7906679056788516e-05,
+      "loss": 0.6914,
+      "step": 1039
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5110221008058635,
+      "learning_rate": 1.790221942923041e-05,
+      "loss": 0.7362,
+      "step": 1040
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6859359028097647,
+      "learning_rate": 1.7897755612895585e-05,
+      "loss": 0.8453,
+      "step": 1041
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0750850963043743,
+      "learning_rate": 1.7893287610150207e-05,
+      "loss": 1.46,
+      "step": 1042
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.588518342701017,
+      "learning_rate": 1.7888815423362647e-05,
+      "loss": 0.8501,
+      "step": 1043
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.499859638766234,
+      "learning_rate": 1.788433905490351e-05,
+      "loss": 0.847,
+      "step": 1044
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4618686917148283,
+      "learning_rate": 1.7879858507145615e-05,
+      "loss": 0.8772,
+      "step": 1045
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5241969014312047,
+      "learning_rate": 1.7875373782463986e-05,
+      "loss": 0.8616,
+      "step": 1046
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.284546483142614,
+      "learning_rate": 1.7870884883235872e-05,
+      "loss": 1.401,
+      "step": 1047
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4089422686933677,
+      "learning_rate": 1.7866391811840725e-05,
+      "loss": 0.7354,
+      "step": 1048
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5280680213587128,
+      "learning_rate": 1.7861894570660218e-05,
+      "loss": 0.7419,
+      "step": 1049
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3055705241160187,
+      "learning_rate": 1.7857393162078234e-05,
+      "loss": 0.7015,
+      "step": 1050
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5609313103630864,
+      "learning_rate": 1.7852887588480857e-05,
+      "loss": 0.9328,
+      "step": 1051
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.030109899677805,
+      "learning_rate": 1.7848377852256383e-05,
+      "loss": 1.3316,
+      "step": 1052
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.383099127479217,
+      "learning_rate": 1.784386395579532e-05,
+      "loss": 0.8145,
+      "step": 1053
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0457771114031904,
+      "learning_rate": 1.783934590149037e-05,
+      "loss": 1.3834,
+      "step": 1054
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4990513332289888,
+      "learning_rate": 1.7834823691736446e-05,
+      "loss": 0.7587,
+      "step": 1055
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.1657304307814087,
+      "learning_rate": 1.783029732893067e-05,
+      "loss": 1.4565,
+      "step": 1056
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3844405801339499,
+      "learning_rate": 1.7825766815472345e-05,
+      "loss": 0.768,
+      "step": 1057
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4520984382236597,
+      "learning_rate": 1.7821232153763e-05,
+      "loss": 0.7641,
+      "step": 1058
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4526050833059436,
+      "learning_rate": 1.7816693346206343e-05,
+      "loss": 0.7522,
+      "step": 1059
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0593569876734015,
+      "learning_rate": 1.781215039520829e-05,
+      "loss": 1.4671,
+      "step": 1060
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4353784465786168,
+      "learning_rate": 1.780760330317695e-05,
+      "loss": 0.7705,
+      "step": 1061
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.632531424036946,
+      "learning_rate": 1.7803052072522627e-05,
+      "loss": 0.777,
+      "step": 1062
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5325558835889745,
+      "learning_rate": 1.7798496705657817e-05,
+      "loss": 0.8638,
+      "step": 1063
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.152645465016432,
+      "learning_rate": 1.779393720499721e-05,
+      "loss": 1.405,
+      "step": 1064
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4549105694290365,
+      "learning_rate": 1.7789373572957696e-05,
+      "loss": 0.7628,
+      "step": 1065
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.421089301629109,
+      "learning_rate": 1.7784805811958336e-05,
+      "loss": 0.7562,
+      "step": 1066
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4271070655944575,
+      "learning_rate": 1.7780233924420395e-05,
+      "loss": 0.74,
+      "step": 1067
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4545907495611328,
+      "learning_rate": 1.7775657912767324e-05,
+      "loss": 0.7017,
+      "step": 1068
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.988980737657913,
+      "learning_rate": 1.777107777942475e-05,
+      "loss": 1.3993,
+      "step": 1069
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4422136921615276,
+      "learning_rate": 1.7766493526820498e-05,
+      "loss": 0.8418,
+      "step": 1070
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4382872671267062,
+      "learning_rate": 1.7761905157384567e-05,
+      "loss": 0.783,
+      "step": 1071
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4803186604342338,
+      "learning_rate": 1.7757312673549144e-05,
+      "loss": 0.7711,
+      "step": 1072
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4318448204172585,
+      "learning_rate": 1.775271607774859e-05,
+      "loss": 0.7269,
+      "step": 1073
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0320689245671506,
+      "learning_rate": 1.7748115372419453e-05,
+      "loss": 1.3865,
+      "step": 1074
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.499529207018485,
+      "learning_rate": 1.7743510560000456e-05,
+      "loss": 0.8554,
+      "step": 1075
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4502129037590112,
+      "learning_rate": 1.7738901642932495e-05,
+      "loss": 0.7619,
+      "step": 1076
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5135616842865642,
+      "learning_rate": 1.7734288623658654e-05,
+      "loss": 0.7943,
+      "step": 1077
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0348545142504015,
+      "learning_rate": 1.7729671504624183e-05,
+      "loss": 1.4178,
+      "step": 1078
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0783635334168507,
+      "learning_rate": 1.77250502882765e-05,
+      "loss": 1.3577,
+      "step": 1079
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5195960244248958,
+      "learning_rate": 1.7720424977065206e-05,
+      "loss": 0.7957,
+      "step": 1080
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4146210761385294,
+      "learning_rate": 1.771579557344206e-05,
+      "loss": 0.6771,
+      "step": 1081
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4670396440155355,
+      "learning_rate": 1.771116207986101e-05,
+      "loss": 0.8342,
+      "step": 1082
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4903206932725674,
+      "learning_rate": 1.770652449877815e-05,
+      "loss": 0.721,
+      "step": 1083
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5413404800430563,
+      "learning_rate": 1.7701882832651752e-05,
+      "loss": 0.8084,
+      "step": 1084
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4847949555861975,
+      "learning_rate": 1.7697237083942254e-05,
+      "loss": 0.8067,
+      "step": 1085
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.496094732113677,
+      "learning_rate": 1.769258725511225e-05,
+      "loss": 0.7728,
+      "step": 1086
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6028143880232006,
+      "learning_rate": 1.7687933348626517e-05,
+      "loss": 0.8634,
+      "step": 1087
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.2924168344064597,
+      "learning_rate": 1.768327536695196e-05,
+      "loss": 1.3265,
+      "step": 1088
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5209612518233788,
+      "learning_rate": 1.7678613312557674e-05,
+      "loss": 0.7011,
+      "step": 1089
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5664688735380696,
+      "learning_rate": 1.7673947187914905e-05,
+      "loss": 0.7945,
+      "step": 1090
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.450758617522419,
+      "learning_rate": 1.7669276995497045e-05,
+      "loss": 0.7889,
+      "step": 1091
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5999882186107592,
+      "learning_rate": 1.7664602737779655e-05,
+      "loss": 0.8589,
+      "step": 1092
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4404103848318306,
+      "learning_rate": 1.765992441724045e-05,
+      "loss": 0.8687,
+      "step": 1093
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3791714685840897,
+      "learning_rate": 1.7655242036359287e-05,
+      "loss": 0.7472,
+      "step": 1094
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.482329512168845,
+      "learning_rate": 1.765055559761819e-05,
+      "loss": 0.856,
+      "step": 1095
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4081856614398651,
+      "learning_rate": 1.764586510350133e-05,
+      "loss": 0.7741,
+      "step": 1096
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5088045294841121,
+      "learning_rate": 1.7641170556495016e-05,
+      "loss": 0.7399,
+      "step": 1097
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4288724597865992,
+      "learning_rate": 1.7636471959087722e-05,
+      "loss": 0.7742,
+      "step": 1098
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0484516142214715,
+      "learning_rate": 1.763176931377006e-05,
+      "loss": 1.3942,
+      "step": 1099
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.533756522149288,
+      "learning_rate": 1.762706262303479e-05,
+      "loss": 0.8618,
+      "step": 1100
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.498997915176529,
+      "learning_rate": 1.7622351889376817e-05,
+      "loss": 0.8156,
+      "step": 1101
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.016365497817233,
+      "learning_rate": 1.761763711529318e-05,
+      "loss": 1.4461,
+      "step": 1102
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.020986907967608,
+      "learning_rate": 1.7612918303283075e-05,
+      "loss": 1.3448,
+      "step": 1103
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.025123496508471,
+      "learning_rate": 1.760819545584783e-05,
+      "loss": 1.3186,
+      "step": 1104
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.508941001732434,
+      "learning_rate": 1.7603468575490912e-05,
+      "loss": 0.7803,
+      "step": 1105
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.082633606610517,
+      "learning_rate": 1.7598737664717927e-05,
+      "loss": 1.4156,
+      "step": 1106
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4554782607396153,
+      "learning_rate": 1.7594002726036615e-05,
+      "loss": 0.7406,
+      "step": 1107
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4125968308870858,
+      "learning_rate": 1.7589263761956856e-05,
+      "loss": 0.7514,
+      "step": 1108
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5739004735770472,
+      "learning_rate": 1.758452077499066e-05,
+      "loss": 0.7732,
+      "step": 1109
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0250768363383966,
+      "learning_rate": 1.7579773767652174e-05,
+      "loss": 1.4264,
+      "step": 1110
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4293254461942697,
+      "learning_rate": 1.7575022742457664e-05,
+      "loss": 0.7432,
+      "step": 1111
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.405291105466518,
+      "learning_rate": 1.757026770192554e-05,
+      "loss": 0.769,
+      "step": 1112
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.515542203041898,
+      "learning_rate": 1.7565508648576338e-05,
+      "loss": 0.8039,
+      "step": 1113
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4997244231853344,
+      "learning_rate": 1.7560745584932712e-05,
+      "loss": 0.7694,
+      "step": 1114
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5498503125105625,
+      "learning_rate": 1.755597851351945e-05,
+      "loss": 0.8261,
+      "step": 1115
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5016157075927437,
+      "learning_rate": 1.755120743686346e-05,
+      "loss": 0.7834,
+      "step": 1116
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4906750669022706,
+      "learning_rate": 1.7546432357493782e-05,
+      "loss": 0.833,
+      "step": 1117
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4948633163062288,
+      "learning_rate": 1.754165327794156e-05,
+      "loss": 0.8367,
+      "step": 1118
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5434266823045852,
+      "learning_rate": 1.7536870200740076e-05,
+      "loss": 0.8464,
+      "step": 1119
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4958976464944247,
+      "learning_rate": 1.7532083128424728e-05,
+      "loss": 0.7442,
+      "step": 1120
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4129727867412805,
+      "learning_rate": 1.752729206353302e-05,
+      "loss": 0.7595,
+      "step": 1121
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4773673673346397,
+      "learning_rate": 1.7522497008604586e-05,
+      "loss": 0.8099,
+      "step": 1122
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4916116034237727,
+      "learning_rate": 1.7517697966181167e-05,
+      "loss": 0.7496,
+      "step": 1123
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.179930705403237,
+      "learning_rate": 1.751289493880662e-05,
+      "loss": 1.3805,
+      "step": 1124
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3384919882378756,
+      "learning_rate": 1.7508087929026913e-05,
+      "loss": 0.7409,
+      "step": 1125
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4988381266775241,
+      "learning_rate": 1.7503276939390133e-05,
+      "loss": 0.8387,
+      "step": 1126
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4355844286787998,
+      "learning_rate": 1.7498461972446463e-05,
+      "loss": 0.8001,
+      "step": 1127
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.161293856584248,
+      "learning_rate": 1.7493643030748203e-05,
+      "loss": 1.3648,
+      "step": 1128
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4730843488635363,
+      "learning_rate": 1.748882011684976e-05,
+      "loss": 0.7958,
+      "step": 1129
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5239814102237283,
+      "learning_rate": 1.7483993233307644e-05,
+      "loss": 0.7684,
+      "step": 1130
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4464786784575003,
+      "learning_rate": 1.747916238268047e-05,
+      "loss": 0.7816,
+      "step": 1131
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3635854341200384,
+      "learning_rate": 1.7474327567528955e-05,
+      "loss": 0.813,
+      "step": 1132
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.099061106116985,
+      "learning_rate": 1.746948879041592e-05,
+      "loss": 1.4155,
+      "step": 1133
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.133372086073395,
+      "learning_rate": 1.7464646053906283e-05,
+      "loss": 1.3363,
+      "step": 1134
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4493624016910953,
+      "learning_rate": 1.745979936056706e-05,
+      "loss": 0.7735,
+      "step": 1135
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4345446235054014,
+      "learning_rate": 1.7454948712967377e-05,
+      "loss": 0.8072,
+      "step": 1136
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5548988534097994,
+      "learning_rate": 1.7450094113678436e-05,
+      "loss": 0.8433,
+      "step": 1137
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5711819652736905,
+      "learning_rate": 1.7445235565273542e-05,
+      "loss": 0.8005,
+      "step": 1138
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4934292046444937,
+      "learning_rate": 1.74403730703281e-05,
+      "loss": 0.71,
+      "step": 1139
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.5934473903461583,
+      "learning_rate": 1.7435506631419605e-05,
+      "loss": 1.3548,
+      "step": 1140
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.387551344034008,
+      "learning_rate": 1.7430636251127634e-05,
+      "loss": 0.7268,
+      "step": 1141
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9267723003570028,
+      "learning_rate": 1.7425761932033857e-05,
+      "loss": 1.403,
+      "step": 1142
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6171738589930178,
+      "learning_rate": 1.7420883676722036e-05,
+      "loss": 0.8103,
+      "step": 1143
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.134183723238342,
+      "learning_rate": 1.7416001487778017e-05,
+      "loss": 1.3867,
+      "step": 1144
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5668930768050215,
+      "learning_rate": 1.7411115367789733e-05,
+      "loss": 0.7565,
+      "step": 1145
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.142655070152206,
+      "learning_rate": 1.7406225319347197e-05,
+      "loss": 1.4247,
+      "step": 1146
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4388317834274948,
+      "learning_rate": 1.7401331345042504e-05,
+      "loss": 0.7439,
+      "step": 1147
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0974038008143374,
+      "learning_rate": 1.7396433447469836e-05,
+      "loss": 1.4326,
+      "step": 1148
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5436181824077233,
+      "learning_rate": 1.7391531629225448e-05,
+      "loss": 0.8179,
+      "step": 1149
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4989691490216914,
+      "learning_rate": 1.738662589290768e-05,
+      "loss": 0.7009,
+      "step": 1150
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5220814121471382,
+      "learning_rate": 1.7381716241116945e-05,
+      "loss": 0.7642,
+      "step": 1151
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5106658805342454,
+      "learning_rate": 1.7376802676455723e-05,
+      "loss": 0.7875,
+      "step": 1152
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4244863877491956,
+      "learning_rate": 1.7371885201528584e-05,
+      "loss": 0.7625,
+      "step": 1153
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3818180811825187,
+      "learning_rate": 1.736696381894216e-05,
+      "loss": 0.7785,
+      "step": 1154
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.4868286650325633,
+      "learning_rate": 1.736203853130516e-05,
+      "loss": 1.3639,
+      "step": 1155
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4538732554582317,
+      "learning_rate": 1.7357109341228366e-05,
+      "loss": 0.7207,
+      "step": 1156
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.1359864645074382,
+      "learning_rate": 1.7352176251324614e-05,
+      "loss": 1.3875,
+      "step": 1157
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4693516162683122,
+      "learning_rate": 1.7347239264208823e-05,
+      "loss": 0.7518,
+      "step": 1158
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9354826806501246,
+      "learning_rate": 1.7342298382497964e-05,
+      "loss": 1.3754,
+      "step": 1159
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.124331612037486,
+      "learning_rate": 1.7337353608811086e-05,
+      "loss": 1.3677,
+      "step": 1160
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5094988459529628,
+      "learning_rate": 1.7332404945769292e-05,
+      "loss": 0.8098,
+      "step": 1161
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5178209079540734,
+      "learning_rate": 1.7327452395995748e-05,
+      "loss": 0.8186,
+      "step": 1162
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4713216478121998,
+      "learning_rate": 1.7322495962115687e-05,
+      "loss": 0.8185,
+      "step": 1163
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4413288754322782,
+      "learning_rate": 1.7317535646756395e-05,
+      "loss": 0.842,
+      "step": 1164
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5224360576951677,
+      "learning_rate": 1.7312571452547208e-05,
+      "loss": 0.7289,
+      "step": 1165
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5415471312269842,
+      "learning_rate": 1.7307603382119532e-05,
+      "loss": 0.713,
+      "step": 1166
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5790333861891996,
+      "learning_rate": 1.7302631438106824e-05,
+      "loss": 0.7722,
+      "step": 1167
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.517771372165895,
+      "learning_rate": 1.729765562314459e-05,
+      "loss": 1.3931,
+      "step": 1168
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4132183037180654,
+      "learning_rate": 1.729267593987039e-05,
+      "loss": 0.778,
+      "step": 1169
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3296951360088927,
+      "learning_rate": 1.728769239092383e-05,
+      "loss": 0.6715,
+      "step": 1170
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4443011539435158,
+      "learning_rate": 1.728270497894658e-05,
+      "loss": 0.729,
+      "step": 1171
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5444509230623702,
+      "learning_rate": 1.7277713706582333e-05,
+      "loss": 0.7507,
+      "step": 1172
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4938587537577448,
+      "learning_rate": 1.7272718576476855e-05,
+      "loss": 0.6863,
+      "step": 1173
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.45431588518712,
+      "learning_rate": 1.726771959127794e-05,
+      "loss": 0.8281,
+      "step": 1174
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.538304086901393,
+      "learning_rate": 1.726271675363543e-05,
+      "loss": 0.7509,
+      "step": 1175
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.2729951734575,
+      "learning_rate": 1.7257710066201205e-05,
+      "loss": 1.3565,
+      "step": 1176
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3239534188862923,
+      "learning_rate": 1.7252699531629197e-05,
+      "loss": 0.6837,
+      "step": 1177
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9822654254740728,
+      "learning_rate": 1.7247685152575367e-05,
+      "loss": 1.3345,
+      "step": 1178
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4542945924762427,
+      "learning_rate": 1.724266693169772e-05,
+      "loss": 0.831,
+      "step": 1179
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.0555000700509454,
+      "learning_rate": 1.7237644871656286e-05,
+      "loss": 1.3796,
+      "step": 1180
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.1771555144144545,
+      "learning_rate": 1.7232618975113146e-05,
+      "loss": 1.3654,
+      "step": 1181
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4074880910874148,
+      "learning_rate": 1.7227589244732407e-05,
+      "loss": 0.7568,
+      "step": 1182
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4149048973170215,
+      "learning_rate": 1.72225556831802e-05,
+      "loss": 0.7798,
+      "step": 1183
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5131073124928105,
+      "learning_rate": 1.7217518293124705e-05,
+      "loss": 0.809,
+      "step": 1184
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.0062457770401148,
+      "learning_rate": 1.7212477077236115e-05,
+      "loss": 1.4542,
+      "step": 1185
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.197434027995912,
+      "learning_rate": 1.720743203818666e-05,
+      "loss": 1.3625,
+      "step": 1186
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5961478388547556,
+      "learning_rate": 1.7202383178650595e-05,
+      "loss": 0.7349,
+      "step": 1187
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3992250282836316,
+      "learning_rate": 1.7197330501304196e-05,
+      "loss": 0.8239,
+      "step": 1188
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.602105162233599,
+      "learning_rate": 1.7192274008825772e-05,
+      "loss": 0.8205,
+      "step": 1189
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.0897201927946982,
+      "learning_rate": 1.718721370389564e-05,
+      "loss": 1.3391,
+      "step": 1190
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4830633417633237,
+      "learning_rate": 1.7182149589196152e-05,
+      "loss": 0.7899,
+      "step": 1191
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.0283970650756067,
+      "learning_rate": 1.7177081667411672e-05,
+      "loss": 1.381,
+      "step": 1192
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4542922592637635,
+      "learning_rate": 1.7172009941228583e-05,
+      "loss": 0.7372,
+      "step": 1193
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5732208090354742,
+      "learning_rate": 1.7166934413335286e-05,
+      "loss": 0.7632,
+      "step": 1194
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.0653845599685354,
+      "learning_rate": 1.7161855086422195e-05,
+      "loss": 1.4074,
+      "step": 1195
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.0410428815690915,
+      "learning_rate": 1.7156771963181743e-05,
+      "loss": 1.3678,
+      "step": 1196
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.51451827299699,
+      "learning_rate": 1.7151685046308365e-05,
+      "loss": 0.7316,
+      "step": 1197
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6084206759500672,
+      "learning_rate": 1.7146594338498523e-05,
+      "loss": 0.8081,
+      "step": 1198
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5845785879411418,
+      "learning_rate": 1.7141499842450667e-05,
+      "loss": 0.7359,
+      "step": 1199
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3951904534911417,
+      "learning_rate": 1.7136401560865275e-05,
+      "loss": 0.6852,
+      "step": 1200
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5900069962512349,
+      "learning_rate": 1.7131299496444825e-05,
+      "loss": 0.8438,
+      "step": 1201
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4623640070941253,
+      "learning_rate": 1.7126193651893795e-05,
+      "loss": 0.8508,
+      "step": 1202
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4586466325898675,
+      "learning_rate": 1.7121084029918668e-05,
+      "loss": 0.8234,
+      "step": 1203
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3991338082144404,
+      "learning_rate": 1.7115970633227935e-05,
+      "loss": 0.7168,
+      "step": 1204
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.413900682173438,
+      "learning_rate": 1.7110853464532086e-05,
+      "loss": 1.3536,
+      "step": 1205
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.2016238325573987,
+      "learning_rate": 1.710573252654361e-05,
+      "loss": 1.4509,
+      "step": 1206
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4582345101483096,
+      "learning_rate": 1.710060782197699e-05,
+      "loss": 0.8473,
+      "step": 1207
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5271765237838748,
+      "learning_rate": 1.709547935354871e-05,
+      "loss": 0.7962,
+      "step": 1208
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5856353444819287,
+      "learning_rate": 1.7090347123977248e-05,
+      "loss": 0.9242,
+      "step": 1209
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4670806630399331,
+      "learning_rate": 1.7085211135983076e-05,
+      "loss": 0.8397,
+      "step": 1210
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5050341705273258,
+      "learning_rate": 1.7080071392288656e-05,
+      "loss": 0.6621,
+      "step": 1211
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.463229992097441,
+      "learning_rate": 1.7074927895618447e-05,
+      "loss": 0.7159,
+      "step": 1212
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.427584343354562,
+      "learning_rate": 1.7069780648698887e-05,
+      "loss": 0.7566,
+      "step": 1213
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5105900202051887,
+      "learning_rate": 1.7064629654258415e-05,
+      "loss": 0.746,
+      "step": 1214
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4734361518840766,
+      "learning_rate": 1.7059474915027443e-05,
+      "loss": 0.7556,
+      "step": 1215
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4984216212945323,
+      "learning_rate": 1.7054316433738377e-05,
+      "loss": 0.7976,
+      "step": 1216
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3660798319631857,
+      "learning_rate": 1.7049154213125606e-05,
+      "loss": 0.809,
+      "step": 1217
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.543633496131371,
+      "learning_rate": 1.7043988255925495e-05,
+      "loss": 0.8278,
+      "step": 1218
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.526003253274376,
+      "learning_rate": 1.7038818564876396e-05,
+      "loss": 1.4281,
+      "step": 1219
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4972002959374384,
+      "learning_rate": 1.703364514271864e-05,
+      "loss": 0.8442,
+      "step": 1220
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3311666150639991,
+      "learning_rate": 1.7028467992194534e-05,
+      "loss": 0.7005,
+      "step": 1221
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.1359482573249204,
+      "learning_rate": 1.7023287116048353e-05,
+      "loss": 1.408,
+      "step": 1222
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.0459028148122385,
+      "learning_rate": 1.7018102517026368e-05,
+      "loss": 1.3603,
+      "step": 1223
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6390144444151644,
+      "learning_rate": 1.70129141978768e-05,
+      "loss": 0.8244,
+      "step": 1224
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.533471709091363,
+      "learning_rate": 1.7007722161349856e-05,
+      "loss": 0.7923,
+      "step": 1225
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0406817011993703,
+      "learning_rate": 1.700252641019771e-05,
+      "loss": 1.4405,
+      "step": 1226
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4528766438533471,
+      "learning_rate": 1.6997326947174504e-05,
+      "loss": 0.7179,
+      "step": 1227
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0403887077120504,
+      "learning_rate": 1.6992123775036352e-05,
+      "loss": 1.4223,
+      "step": 1228
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4285693493830252,
+      "learning_rate": 1.6986916896541324e-05,
+      "loss": 0.8194,
+      "step": 1229
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4044863132442014,
+      "learning_rate": 1.6981706314449466e-05,
+      "loss": 0.793,
+      "step": 1230
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3839633015682995,
+      "learning_rate": 1.6976492031522786e-05,
+      "loss": 0.8584,
+      "step": 1231
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.671844296447469,
+      "learning_rate": 1.697127405052524e-05,
+      "loss": 0.8462,
+      "step": 1232
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.04701359752539,
+      "learning_rate": 1.696605237422276e-05,
+      "loss": 1.3762,
+      "step": 1233
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4689870808239858,
+      "learning_rate": 1.6960827005383236e-05,
+      "loss": 0.733,
+      "step": 1234
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.490708295925657,
+      "learning_rate": 1.69555979467765e-05,
+      "loss": 0.7311,
+      "step": 1235
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3261053766477942,
+      "learning_rate": 1.695036520117436e-05,
+      "loss": 0.7061,
+      "step": 1236
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.310798012049045,
+      "learning_rate": 1.694512877135056e-05,
+      "loss": 0.6938,
+      "step": 1237
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.025546327841953,
+      "learning_rate": 1.6939888660080813e-05,
+      "loss": 1.3701,
+      "step": 1238
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5841191629145066,
+      "learning_rate": 1.6934644870142777e-05,
+      "loss": 0.7854,
+      "step": 1239
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5218903082706097,
+      "learning_rate": 1.6929397404316052e-05,
+      "loss": 0.7695,
+      "step": 1240
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5658513086918364,
+      "learning_rate": 1.69241462653822e-05,
+      "loss": 0.8626,
+      "step": 1241
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4292397336325577,
+      "learning_rate": 1.6918891456124718e-05,
+      "loss": 0.7595,
+      "step": 1242
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.1396530836537333,
+      "learning_rate": 1.691363297932906e-05,
+      "loss": 1.4378,
+      "step": 1243
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4757409007239188,
+      "learning_rate": 1.690837083778261e-05,
+      "loss": 0.6789,
+      "step": 1244
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.475698105469845,
+      "learning_rate": 1.6903105034274715e-05,
+      "loss": 0.8661,
+      "step": 1245
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5028194901090202,
+      "learning_rate": 1.689783557159664e-05,
+      "loss": 0.7886,
+      "step": 1246
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5043107345979554,
+      "learning_rate": 1.6892562452541607e-05,
+      "loss": 0.8021,
+      "step": 1247
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3978452477751833,
+      "learning_rate": 1.6887285679904764e-05,
+      "loss": 0.6931,
+      "step": 1248
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.6265034576694783,
+      "learning_rate": 1.688200525648321e-05,
+      "loss": 1.3826,
+      "step": 1249
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3466472408558814,
+      "learning_rate": 1.687672118507596e-05,
+      "loss": 0.781,
+      "step": 1250
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4902400433572074,
+      "learning_rate": 1.687143346848398e-05,
+      "loss": 0.8637,
+      "step": 1251
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9484029193353107,
+      "learning_rate": 1.6866142109510158e-05,
+      "loss": 1.4208,
+      "step": 1252
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.57961905212225,
+      "learning_rate": 1.6860847110959325e-05,
+      "loss": 0.7765,
+      "step": 1253
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.092392856239657,
+      "learning_rate": 1.6855548475638225e-05,
+      "loss": 1.3616,
+      "step": 1254
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5212806466766375,
+      "learning_rate": 1.6850246206355538e-05,
+      "loss": 0.7467,
+      "step": 1255
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5763617916901789,
+      "learning_rate": 1.684494030592187e-05,
+      "loss": 0.7125,
+      "step": 1256
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0700528164117373,
+      "learning_rate": 1.683963077714975e-05,
+      "loss": 1.398,
+      "step": 1257
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.045400103368578,
+      "learning_rate": 1.6834317622853635e-05,
+      "loss": 1.4482,
+      "step": 1258
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4440995565274592,
+      "learning_rate": 1.6829000845849897e-05,
+      "loss": 0.7478,
+      "step": 1259
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4916014141981149,
+      "learning_rate": 1.6823680448956835e-05,
+      "loss": 0.8275,
+      "step": 1260
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3816804197186254,
+      "learning_rate": 1.6818356434994662e-05,
+      "loss": 0.7914,
+      "step": 1261
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0652217038563845,
+      "learning_rate": 1.6813028806785512e-05,
+      "loss": 1.3306,
+      "step": 1262
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4574776820081405,
+      "learning_rate": 1.6807697567153428e-05,
+      "loss": 0.7841,
+      "step": 1263
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4739775021877903,
+      "learning_rate": 1.6802362718924374e-05,
+      "loss": 0.7068,
+      "step": 1264
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.13945375630047,
+      "learning_rate": 1.6797024264926228e-05,
+      "loss": 1.3404,
+      "step": 1265
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.497736866786938,
+      "learning_rate": 1.6791682207988775e-05,
+      "loss": 0.7748,
+      "step": 1266
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6072964092774946,
+      "learning_rate": 1.6786336550943708e-05,
+      "loss": 0.8134,
+      "step": 1267
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.1651151345352644,
+      "learning_rate": 1.6780987296624635e-05,
+      "loss": 1.4051,
+      "step": 1268
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5387501161517236,
+      "learning_rate": 1.6775634447867064e-05,
+      "loss": 0.7095,
+      "step": 1269
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.192241302478756,
+      "learning_rate": 1.6770278007508416e-05,
+      "loss": 1.4666,
+      "step": 1270
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4138143223962816,
+      "learning_rate": 1.6764917978388004e-05,
+      "loss": 0.741,
+      "step": 1271
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5206226089985364,
+      "learning_rate": 1.675955436334706e-05,
+      "loss": 0.785,
+      "step": 1272
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5950130525770494,
+      "learning_rate": 1.6754187165228702e-05,
+      "loss": 0.8129,
+      "step": 1273
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.453209046696253,
+      "learning_rate": 1.6748816386877953e-05,
+      "loss": 0.6861,
+      "step": 1274
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.2583699379841438,
+      "learning_rate": 1.6743442031141733e-05,
+      "loss": 1.3722,
+      "step": 1275
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5267866471237228,
+      "learning_rate": 1.6738064100868864e-05,
+      "loss": 0.8004,
+      "step": 1276
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4319776083778208,
+      "learning_rate": 1.6732682598910052e-05,
+      "loss": 0.8169,
+      "step": 1277
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5437125711755726,
+      "learning_rate": 1.6727297528117897e-05,
+      "loss": 0.7885,
+      "step": 1278
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.421309293261037,
+      "learning_rate": 1.672190889134691e-05,
+      "loss": 0.9009,
+      "step": 1279
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3950965817732917,
+      "learning_rate": 1.671651669145347e-05,
+      "loss": 0.7245,
+      "step": 1280
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4457681316643107,
+      "learning_rate": 1.671112093129585e-05,
+      "loss": 0.8095,
+      "step": 1281
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.0048617983642294,
+      "learning_rate": 1.670572161373422e-05,
+      "loss": 1.4717,
+      "step": 1282
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3225616831398794,
+      "learning_rate": 1.6700318741630623e-05,
+      "loss": 0.7562,
+      "step": 1283
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4405468477476933,
+      "learning_rate": 1.6694912317848993e-05,
+      "loss": 0.7691,
+      "step": 1284
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5282146729798443,
+      "learning_rate": 1.6689502345255145e-05,
+      "loss": 0.7755,
+      "step": 1285
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.095523730708687,
+      "learning_rate": 1.6684088826716776e-05,
+      "loss": 1.386,
+      "step": 1286
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4514049428898184,
+      "learning_rate": 1.6678671765103466e-05,
+      "loss": 0.8594,
+      "step": 1287
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9839581362216496,
+      "learning_rate": 1.667325116328666e-05,
+      "loss": 1.4358,
+      "step": 1288
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9638625206434466,
+      "learning_rate": 1.6667827024139697e-05,
+      "loss": 1.3575,
+      "step": 1289
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5271262249975148,
+      "learning_rate": 1.666239935053778e-05,
+      "loss": 0.6772,
+      "step": 1290
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.1051695763461327,
+      "learning_rate": 1.6656968145357988e-05,
+      "loss": 1.4369,
+      "step": 1291
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6263443486602667,
+      "learning_rate": 1.6651533411479273e-05,
+      "loss": 0.8467,
+      "step": 1292
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9581435878853464,
+      "learning_rate": 1.664609515178246e-05,
+      "loss": 1.41,
+      "step": 1293
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3730829358504952,
+      "learning_rate": 1.6640653369150233e-05,
+      "loss": 0.7469,
+      "step": 1294
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.0895346797782137,
+      "learning_rate": 1.6635208066467158e-05,
+      "loss": 1.3305,
+      "step": 1295
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4026265316292394,
+      "learning_rate": 1.6629759246619652e-05,
+      "loss": 0.759,
+      "step": 1296
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4112035393584308,
+      "learning_rate": 1.6624306912496006e-05,
+      "loss": 0.7722,
+      "step": 1297
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4676552932651201,
+      "learning_rate": 1.6618851066986375e-05,
+      "loss": 0.7959,
+      "step": 1298
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.0705200868725244,
+      "learning_rate": 1.6613391712982765e-05,
+      "loss": 1.4172,
+      "step": 1299
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9887122567877322,
+      "learning_rate": 1.6607928853379058e-05,
+      "loss": 1.3717,
+      "step": 1300
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5050545619829478,
+      "learning_rate": 1.6602462491070975e-05,
+      "loss": 0.6674,
+      "step": 1301
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4832521972402661,
+      "learning_rate": 1.6596992628956104e-05,
+      "loss": 0.7172,
+      "step": 1302
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.541486842920162,
+      "learning_rate": 1.6591519269933897e-05,
+      "loss": 0.8245,
+      "step": 1303
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4564312471575716,
+      "learning_rate": 1.658604241690564e-05,
+      "loss": 0.7583,
+      "step": 1304
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.134753084928943,
+      "learning_rate": 1.658056207277449e-05,
+      "loss": 1.3113,
+      "step": 1305
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6179261504038016,
+      "learning_rate": 1.6575078240445437e-05,
+      "loss": 0.8303,
+      "step": 1306
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4282300477441123,
+      "learning_rate": 1.6569590922825334e-05,
+      "loss": 0.8116,
+      "step": 1307
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4082228187982648,
+      "learning_rate": 1.6564100122822874e-05,
+      "loss": 0.7572,
+      "step": 1308
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.192058627695944,
+      "learning_rate": 1.65586058433486e-05,
+      "loss": 1.3847,
+      "step": 1309
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.491700884847167,
+      "learning_rate": 1.6553108087314898e-05,
+      "loss": 0.8154,
+      "step": 1310
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4544310955894566,
+      "learning_rate": 1.6547606857635994e-05,
+      "loss": 0.78,
+      "step": 1311
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4372926499612535,
+      "learning_rate": 1.654210215722796e-05,
+      "loss": 0.7135,
+      "step": 1312
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3917170779547012,
+      "learning_rate": 1.65365939890087e-05,
+      "loss": 0.6651,
+      "step": 1313
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4910121525614621,
+      "learning_rate": 1.653108235589797e-05,
+      "loss": 0.7191,
+      "step": 1314
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5102081608209035,
+      "learning_rate": 1.652556726081735e-05,
+      "loss": 0.7647,
+      "step": 1315
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5187922414266963,
+      "learning_rate": 1.6520048706690263e-05,
+      "loss": 0.8559,
+      "step": 1316
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.571533298711159,
+      "learning_rate": 1.651452669644196e-05,
+      "loss": 0.7629,
+      "step": 1317
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3919896648088994,
+      "learning_rate": 1.6509001232999516e-05,
+      "loss": 0.7526,
+      "step": 1318
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.1923397790981225,
+      "learning_rate": 1.6503472319291864e-05,
+      "loss": 1.4567,
+      "step": 1319
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5192374985900736,
+      "learning_rate": 1.6497939958249732e-05,
+      "loss": 0.8241,
+      "step": 1320
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9861826981209318,
+      "learning_rate": 1.6492404152805705e-05,
+      "loss": 1.3974,
+      "step": 1321
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3233401159865035,
+      "learning_rate": 1.6486864905894173e-05,
+      "loss": 0.7536,
+      "step": 1322
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.433228235598881,
+      "learning_rate": 1.648132222045136e-05,
+      "loss": 0.7277,
+      "step": 1323
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5805184065738087,
+      "learning_rate": 1.6475776099415312e-05,
+      "loss": 0.8325,
+      "step": 1324
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4874483615680185,
+      "learning_rate": 1.647022654572589e-05,
+      "loss": 0.8037,
+      "step": 1325
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.404039731125953,
+      "learning_rate": 1.6464673562324782e-05,
+      "loss": 0.7652,
+      "step": 1326
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4973260979257166,
+      "learning_rate": 1.6459117152155496e-05,
+      "loss": 0.8063,
+      "step": 1327
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5579598141271689,
+      "learning_rate": 1.645355731816334e-05,
+      "loss": 0.8506,
+      "step": 1328
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.1387018919357326,
+      "learning_rate": 1.6447994063295454e-05,
+      "loss": 1.3961,
+      "step": 1329
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4978431912383903,
+      "learning_rate": 1.6442427390500794e-05,
+      "loss": 0.8098,
+      "step": 1330
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5441221715742002,
+      "learning_rate": 1.6436857302730105e-05,
+      "loss": 0.7933,
+      "step": 1331
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3426815677187522,
+      "learning_rate": 1.6431283802935968e-05,
+      "loss": 0.7394,
+      "step": 1332
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.36083592611761,
+      "learning_rate": 1.6425706894072753e-05,
+      "loss": 0.7121,
+      "step": 1333
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.090094491613444,
+      "learning_rate": 1.6420126579096656e-05,
+      "loss": 1.409,
+      "step": 1334
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4212006923573612,
+      "learning_rate": 1.6414542860965655e-05,
+      "loss": 0.7588,
+      "step": 1335
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5275258734332724,
+      "learning_rate": 1.640895574263956e-05,
+      "loss": 0.7547,
+      "step": 1336
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9993762242496231,
+      "learning_rate": 1.6403365227079955e-05,
+      "loss": 1.3237,
+      "step": 1337
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.390205726474753,
+      "learning_rate": 1.6397771317250242e-05,
+      "loss": 0.7495,
+      "step": 1338
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5286535203217975,
+      "learning_rate": 1.6392174016115623e-05,
+      "loss": 0.8242,
+      "step": 1339
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4593651152200924,
+      "learning_rate": 1.6386573326643083e-05,
+      "loss": 0.7361,
+      "step": 1340
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.115402595312582,
+      "learning_rate": 1.6380969251801424e-05,
+      "loss": 1.4587,
+      "step": 1341
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.0398590519744872,
+      "learning_rate": 1.6375361794561224e-05,
+      "loss": 1.4049,
+      "step": 1342
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4559619175195178,
+      "learning_rate": 1.6369750957894866e-05,
+      "loss": 0.8058,
+      "step": 1343
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4183688887074695,
+      "learning_rate": 1.6364136744776517e-05,
+      "loss": 0.7697,
+      "step": 1344
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4135703288503214,
+      "learning_rate": 1.6358519158182137e-05,
+      "loss": 0.7136,
+      "step": 1345
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.0393177976585255,
+      "learning_rate": 1.6352898201089472e-05,
+      "loss": 1.4673,
+      "step": 1346
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5159021096861,
+      "learning_rate": 1.634727387647806e-05,
+      "loss": 0.69,
+      "step": 1347
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.0741932964869862,
+      "learning_rate": 1.634164618732922e-05,
+      "loss": 1.4176,
+      "step": 1348
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.0089117489753012,
+      "learning_rate": 1.633601513662605e-05,
+      "loss": 1.3521,
+      "step": 1349
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4502794796238188,
+      "learning_rate": 1.633038072735344e-05,
+      "loss": 0.7927,
+      "step": 1350
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.386466417741092,
+      "learning_rate": 1.6324742962498054e-05,
+      "loss": 0.7806,
+      "step": 1351
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3819478343689364,
+      "learning_rate": 1.6319101845048333e-05,
+      "loss": 0.758,
+      "step": 1352
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.492665614956901,
+      "learning_rate": 1.63134573779945e-05,
+      "loss": 0.7761,
+      "step": 1353
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3692304194046143,
+      "learning_rate": 1.630780956432855e-05,
+      "loss": 0.7451,
+      "step": 1354
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.067261872076799,
+      "learning_rate": 1.6302158407044257e-05,
+      "loss": 1.3352,
+      "step": 1355
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4364610100047412,
+      "learning_rate": 1.629650390913716e-05,
+      "loss": 0.7377,
+      "step": 1356
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5373501194060075,
+      "learning_rate": 1.6290846073604577e-05,
+      "loss": 0.7614,
+      "step": 1357
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4676510899661588,
+      "learning_rate": 1.6285184903445586e-05,
+      "loss": 0.7631,
+      "step": 1358
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4544674161119389,
+      "learning_rate": 1.627952040166104e-05,
+      "loss": 0.7698,
+      "step": 1359
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4741634151348009,
+      "learning_rate": 1.627385257125355e-05,
+      "loss": 0.7736,
+      "step": 1360
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4509371297078995,
+      "learning_rate": 1.6268181415227503e-05,
+      "loss": 0.8361,
+      "step": 1361
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5291599104673474,
+      "learning_rate": 1.626250693658904e-05,
+      "loss": 0.7361,
+      "step": 1362
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.452572157103036,
+      "learning_rate": 1.625682913834607e-05,
+      "loss": 0.8226,
+      "step": 1363
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.477971435660668,
+      "learning_rate": 1.625114802350825e-05,
+      "loss": 0.7608,
+      "step": 1364
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5169859482331742,
+      "learning_rate": 1.6245463595087006e-05,
+      "loss": 0.7879,
+      "step": 1365
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6385323472882773,
+      "learning_rate": 1.623977585609552e-05,
+      "loss": 0.7494,
+      "step": 1366
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5972380142842248,
+      "learning_rate": 1.6234084809548724e-05,
+      "loss": 0.8503,
+      "step": 1367
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9714493289547854,
+      "learning_rate": 1.6228390458463303e-05,
+      "loss": 1.2981,
+      "step": 1368
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5094060619247012,
+      "learning_rate": 1.6222692805857698e-05,
+      "loss": 0.6752,
+      "step": 1369
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9853858397548405,
+      "learning_rate": 1.6216991854752097e-05,
+      "loss": 1.3791,
+      "step": 1370
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.43750914160143,
+      "learning_rate": 1.6211287608168436e-05,
+      "loss": 0.7418,
+      "step": 1371
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.589616225367494,
+      "learning_rate": 1.62055800691304e-05,
+      "loss": 0.8279,
+      "step": 1372
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5822302834492823,
+      "learning_rate": 1.6199869240663418e-05,
+      "loss": 0.8264,
+      "step": 1373
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.0833185410968023,
+      "learning_rate": 1.6194155125794664e-05,
+      "loss": 1.351,
+      "step": 1374
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5254108052948834,
+      "learning_rate": 1.618843772755305e-05,
+      "loss": 0.8467,
+      "step": 1375
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.374044468173295,
+      "learning_rate": 1.6182717048969238e-05,
+      "loss": 0.7684,
+      "step": 1376
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4768416172142793,
+      "learning_rate": 1.6176993093075618e-05,
+      "loss": 0.7246,
+      "step": 1377
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.459166339945056,
+      "learning_rate": 1.617126586290632e-05,
+      "loss": 0.8116,
+      "step": 1378
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5541167275355052,
+      "learning_rate": 1.6165535361497218e-05,
+      "loss": 0.83,
+      "step": 1379
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4582435666863034,
+      "learning_rate": 1.6159801591885905e-05,
+      "loss": 0.7138,
+      "step": 1380
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4334169841690836,
+      "learning_rate": 1.6154064557111716e-05,
+      "loss": 0.7401,
+      "step": 1381
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4409861034309042,
+      "learning_rate": 1.614832426021572e-05,
+      "loss": 0.75,
+      "step": 1382
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4371244257570421,
+      "learning_rate": 1.6142580704240708e-05,
+      "loss": 0.7452,
+      "step": 1383
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5365516711545413,
+      "learning_rate": 1.61368338922312e-05,
+      "loss": 0.7153,
+      "step": 1384
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.606024626577112,
+      "learning_rate": 1.6131083827233443e-05,
+      "loss": 0.8316,
+      "step": 1385
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.535974456214069,
+      "learning_rate": 1.6125330512295408e-05,
+      "loss": 0.8188,
+      "step": 1386
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5057041514956138,
+      "learning_rate": 1.611957395046679e-05,
+      "loss": 0.7801,
+      "step": 1387
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.1572984683969176,
+      "learning_rate": 1.6113814144799008e-05,
+      "loss": 1.3046,
+      "step": 1388
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.1502871061730904,
+      "learning_rate": 1.610805109834519e-05,
+      "loss": 1.3873,
+      "step": 1389
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5386359615926617,
+      "learning_rate": 1.6102284814160194e-05,
+      "loss": 0.7493,
+      "step": 1390
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5189137968069708,
+      "learning_rate": 1.6096515295300583e-05,
+      "loss": 0.7764,
+      "step": 1391
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4670371238764617,
+      "learning_rate": 1.6090742544824646e-05,
+      "loss": 0.7465,
+      "step": 1392
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5051793553582264,
+      "learning_rate": 1.6084966565792374e-05,
+      "loss": 0.7675,
+      "step": 1393
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.656702001013037,
+      "learning_rate": 1.607918736126548e-05,
+      "loss": 0.9,
+      "step": 1394
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4956349326428402,
+      "learning_rate": 1.607340493430738e-05,
+      "loss": 0.8337,
+      "step": 1395
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5380433420103998,
+      "learning_rate": 1.60676192879832e-05,
+      "loss": 0.8873,
+      "step": 1396
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5278022321394984,
+      "learning_rate": 1.6061830425359768e-05,
+      "loss": 0.8749,
+      "step": 1397
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5387676249960112,
+      "learning_rate": 1.6056038349505625e-05,
+      "loss": 0.7655,
+      "step": 1398
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5039421115789218,
+      "learning_rate": 1.605024306349101e-05,
+      "loss": 0.8106,
+      "step": 1399
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.383179034579015,
+      "learning_rate": 1.604444457038787e-05,
+      "loss": 1.4528,
+      "step": 1400
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.376365592876297,
+      "learning_rate": 1.6038642873269834e-05,
+      "loss": 0.677,
+      "step": 1401
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.102864525253633,
+      "learning_rate": 1.6032837975212256e-05,
+      "loss": 1.369,
+      "step": 1402
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5038514588068546,
+      "learning_rate": 1.6027029879292167e-05,
+      "loss": 0.8698,
+      "step": 1403
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4561525267441544,
+      "learning_rate": 1.60212185885883e-05,
+      "loss": 0.7931,
+      "step": 1404
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5585753726267342,
+      "learning_rate": 1.6015404106181076e-05,
+      "loss": 0.7482,
+      "step": 1405
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5302091120532526,
+      "learning_rate": 1.600958643515262e-05,
+      "loss": 0.8186,
+      "step": 1406
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4717702096102279,
+      "learning_rate": 1.6003765578586737e-05,
+      "loss": 0.7425,
+      "step": 1407
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.181562022116644,
+      "learning_rate": 1.5997941539568926e-05,
+      "loss": 1.4115,
+      "step": 1408
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4067812412438168,
+      "learning_rate": 1.599211432118636e-05,
+      "loss": 0.8,
+      "step": 1409
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.091148717435921,
+      "learning_rate": 1.5986283926527918e-05,
+      "loss": 1.4079,
+      "step": 1410
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5561361200832282,
+      "learning_rate": 1.598045035868415e-05,
+      "loss": 0.7845,
+      "step": 1411
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5804431066674387,
+      "learning_rate": 1.5974613620747286e-05,
+      "loss": 0.8062,
+      "step": 1412
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3945865768265684,
+      "learning_rate": 1.5968773715811245e-05,
+      "loss": 0.7464,
+      "step": 1413
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.419946903489252,
+      "learning_rate": 1.5962930646971617e-05,
+      "loss": 0.7513,
+      "step": 1414
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4844498842039335,
+      "learning_rate": 1.5957084417325675e-05,
+      "loss": 0.7535,
+      "step": 1415
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5140927795502452,
+      "learning_rate": 1.5951235029972366e-05,
+      "loss": 0.8059,
+      "step": 1416
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.374511458179032,
+      "learning_rate": 1.5945382488012306e-05,
+      "loss": 1.4428,
+      "step": 1417
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.508812781910841,
+      "learning_rate": 1.5939526794547787e-05,
+      "loss": 0.8728,
+      "step": 1418
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.0254631452053036,
+      "learning_rate": 1.5933667952682777e-05,
+      "loss": 1.3152,
+      "step": 1419
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4376591160028558,
+      "learning_rate": 1.59278059655229e-05,
+      "loss": 0.7658,
+      "step": 1420
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4766422539690232,
+      "learning_rate": 1.592194083617546e-05,
+      "loss": 0.7977,
+      "step": 1421
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.108137046700855,
+      "learning_rate": 1.5916072567749422e-05,
+      "loss": 1.4453,
+      "step": 1422
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5088082495841817,
+      "learning_rate": 1.5910201163355405e-05,
+      "loss": 0.7347,
+      "step": 1423
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9971568660609031,
+      "learning_rate": 1.590432662610571e-05,
+      "loss": 1.3522,
+      "step": 1424
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4553631354356904,
+      "learning_rate": 1.5898448959114286e-05,
+      "loss": 0.8396,
+      "step": 1425
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5800033343654303,
+      "learning_rate": 1.5892568165496736e-05,
+      "loss": 0.8353,
+      "step": 1426
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4596526337659368,
+      "learning_rate": 1.5886684248370336e-05,
+      "loss": 0.8677,
+      "step": 1427
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.397244074374043,
+      "learning_rate": 1.5880797210854003e-05,
+      "loss": 0.7295,
+      "step": 1428
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.218184189763992,
+      "learning_rate": 1.587490705606832e-05,
+      "loss": 1.3942,
+      "step": 1429
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4775262769311708,
+      "learning_rate": 1.5869013787135513e-05,
+      "loss": 0.8159,
+      "step": 1430
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4576997776438458,
+      "learning_rate": 1.5863117407179465e-05,
+      "loss": 0.7267,
+      "step": 1431
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3544004970930632,
+      "learning_rate": 1.5857217919325703e-05,
+      "loss": 0.78,
+      "step": 1432
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.484042678584917,
+      "learning_rate": 1.58513153267014e-05,
+      "loss": 0.7658,
+      "step": 1433
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.0331517470653253,
+      "learning_rate": 1.5845409632435385e-05,
+      "loss": 1.4092,
+      "step": 1434
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5831179647276592,
+      "learning_rate": 1.583950083965812e-05,
+      "loss": 0.8309,
+      "step": 1435
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3768903873921647,
+      "learning_rate": 1.583358895150172e-05,
+      "loss": 0.8194,
+      "step": 1436
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.429851202322633,
+      "learning_rate": 1.5827673971099927e-05,
+      "loss": 0.7381,
+      "step": 1437
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.50116263919019,
+      "learning_rate": 1.582175590158814e-05,
+      "loss": 0.7888,
+      "step": 1438
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.534494778075415,
+      "learning_rate": 1.5815834746103376e-05,
+      "loss": 0.8246,
+      "step": 1439
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.0839743493669767,
+      "learning_rate": 1.5809910507784302e-05,
+      "loss": 1.3694,
+      "step": 1440
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4660526588052942,
+      "learning_rate": 1.580398318977121e-05,
+      "loss": 0.7935,
+      "step": 1441
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5248836989642616,
+      "learning_rate": 1.579805279520604e-05,
+      "loss": 0.7609,
+      "step": 1442
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.384788030728724,
+      "learning_rate": 1.5792119327232337e-05,
+      "loss": 0.6992,
+      "step": 1443
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3983933223304525,
+      "learning_rate": 1.5786182788995302e-05,
+      "loss": 0.7525,
+      "step": 1444
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5389524463322244,
+      "learning_rate": 1.5780243183641747e-05,
+      "loss": 0.8493,
+      "step": 1445
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.480069041523941,
+      "learning_rate": 1.5774300514320112e-05,
+      "loss": 0.8219,
+      "step": 1446
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5054977849380127,
+      "learning_rate": 1.576835478418046e-05,
+      "loss": 0.8056,
+      "step": 1447
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4818346979112262,
+      "learning_rate": 1.576240599637449e-05,
+      "loss": 0.7208,
+      "step": 1448
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4165338226133157,
+      "learning_rate": 1.5756454154055508e-05,
+      "loss": 0.7103,
+      "step": 1449
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.3473336220462953,
+      "learning_rate": 1.5750499260378438e-05,
+      "loss": 1.3377,
+      "step": 1450
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.513223854482508,
+      "learning_rate": 1.5744541318499828e-05,
+      "loss": 0.7659,
+      "step": 1451
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4026528202320145,
+      "learning_rate": 1.573858033157784e-05,
+      "loss": 0.8452,
+      "step": 1452
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.402118811906913,
+      "learning_rate": 1.5732616302772255e-05,
+      "loss": 0.7226,
+      "step": 1453
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3972344555069909,
+      "learning_rate": 1.5726649235244455e-05,
+      "loss": 0.7496,
+      "step": 1454
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4463096090246215,
+      "learning_rate": 1.572067913215744e-05,
+      "loss": 0.7634,
+      "step": 1455
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5245959811079421,
+      "learning_rate": 1.571470599667582e-05,
+      "loss": 0.6822,
+      "step": 1456
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.734599194012061,
+      "learning_rate": 1.5708729831965814e-05,
+      "loss": 1.3991,
+      "step": 1457
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4710521765701143,
+      "learning_rate": 1.5702750641195235e-05,
+      "loss": 0.7946,
+      "step": 1458
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4615431552160003,
+      "learning_rate": 1.5696768427533515e-05,
+      "loss": 0.8229,
+      "step": 1459
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.053315345570868,
+      "learning_rate": 1.569078319415168e-05,
+      "loss": 1.4054,
+      "step": 1460
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.631327010508076,
+      "learning_rate": 1.5684794944222358e-05,
+      "loss": 0.8219,
+      "step": 1461
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5757491311636849,
+      "learning_rate": 1.5678803680919776e-05,
+      "loss": 0.833,
+      "step": 1462
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5417534011000396,
+      "learning_rate": 1.5672809407419762e-05,
+      "loss": 0.7755,
+      "step": 1463
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5168223199802275,
+      "learning_rate": 1.5666812126899732e-05,
+      "loss": 0.7207,
+      "step": 1464
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.472779077515762,
+      "learning_rate": 1.56608118425387e-05,
+      "loss": 1.4104,
+      "step": 1465
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.374155366334589,
+      "learning_rate": 1.565480855751728e-05,
+      "loss": 0.7493,
+      "step": 1466
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4438735140877124,
+      "learning_rate": 1.564880227501766e-05,
+      "loss": 0.7683,
+      "step": 1467
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5221312085563499,
+      "learning_rate": 1.5642792998223633e-05,
+      "loss": 0.7754,
+      "step": 1468
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.481320620060798,
+      "learning_rate": 1.563678073032057e-05,
+      "loss": 0.7219,
+      "step": 1469
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2864907316622523,
+      "learning_rate": 1.5630765474495426e-05,
+      "loss": 0.6924,
+      "step": 1470
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3609213283055532,
+      "learning_rate": 1.562474723393675e-05,
+      "loss": 0.7712,
+      "step": 1471
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5835504323635452,
+      "learning_rate": 1.561872601183466e-05,
+      "loss": 0.857,
+      "step": 1472
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5758677514137853,
+      "learning_rate": 1.561270181138087e-05,
+      "loss": 0.828,
+      "step": 1473
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5182411206225714,
+      "learning_rate": 1.5606674635768656e-05,
+      "loss": 0.7557,
+      "step": 1474
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.1978682571924257,
+      "learning_rate": 1.5600644488192882e-05,
+      "loss": 1.3565,
+      "step": 1475
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.571941959744457,
+      "learning_rate": 1.5594611371849986e-05,
+      "loss": 0.7538,
+      "step": 1476
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4798535846654468,
+      "learning_rate": 1.5588575289937975e-05,
+      "loss": 0.7496,
+      "step": 1477
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9553970293042906,
+      "learning_rate": 1.558253624565643e-05,
+      "loss": 1.3931,
+      "step": 1478
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4957615846753538,
+      "learning_rate": 1.557649424220651e-05,
+      "loss": 0.7773,
+      "step": 1479
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4085984580195228,
+      "learning_rate": 1.5570449282790928e-05,
+      "loss": 0.7361,
+      "step": 1480
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.496190651925664,
+      "learning_rate": 1.556440137061398e-05,
+      "loss": 0.7637,
+      "step": 1481
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3597180007090537,
+      "learning_rate": 1.555835050888151e-05,
+      "loss": 0.7513,
+      "step": 1482
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.441344976464899,
+      "learning_rate": 1.555229670080094e-05,
+      "loss": 0.7528,
+      "step": 1483
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3625551655404984,
+      "learning_rate": 1.554623994958125e-05,
+      "loss": 0.7925,
+      "step": 1484
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4882082301550343,
+      "learning_rate": 1.5540180258432974e-05,
+      "loss": 0.7523,
+      "step": 1485
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4115307781134205,
+      "learning_rate": 1.553411763056821e-05,
+      "loss": 0.8282,
+      "step": 1486
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.48552332782406,
+      "learning_rate": 1.5528052069200613e-05,
+      "loss": 0.8176,
+      "step": 1487
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3601805155577893,
+      "learning_rate": 1.5521983577545388e-05,
+      "loss": 0.8227,
+      "step": 1488
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6518433506876535,
+      "learning_rate": 1.55159121588193e-05,
+      "loss": 0.8142,
+      "step": 1489
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4988188557581437,
+      "learning_rate": 1.550983781624066e-05,
+      "loss": 0.8096,
+      "step": 1490
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4086526523376235,
+      "learning_rate": 1.5503760553029332e-05,
+      "loss": 0.6979,
+      "step": 1491
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.401023233990951,
+      "learning_rate": 1.5497680372406727e-05,
+      "loss": 0.8194,
+      "step": 1492
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3708081082565249,
+      "learning_rate": 1.5491597277595805e-05,
+      "loss": 0.7817,
+      "step": 1493
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.476696374452408,
+      "learning_rate": 1.5485511271821065e-05,
+      "loss": 0.7899,
+      "step": 1494
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3977279235434006,
+      "learning_rate": 1.5479422358308557e-05,
+      "loss": 0.714,
+      "step": 1495
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.395043423341745,
+      "learning_rate": 1.5473330540285863e-05,
+      "loss": 0.7283,
+      "step": 1496
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.403777427122053,
+      "learning_rate": 1.546723582098211e-05,
+      "loss": 1.42,
+      "step": 1497
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5869616249682876,
+      "learning_rate": 1.5461138203627965e-05,
+      "loss": 0.8424,
+      "step": 1498
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5017282702623007,
+      "learning_rate": 1.5455037691455626e-05,
+      "loss": 0.8115,
+      "step": 1499
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4571536161246603,
+      "learning_rate": 1.5448934287698827e-05,
+      "loss": 0.7111,
+      "step": 1500
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.165501111703767,
+      "learning_rate": 1.544282799559284e-05,
+      "loss": 1.3783,
+      "step": 1501
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3532194606158132,
+      "learning_rate": 1.543671881837446e-05,
+      "loss": 0.8379,
+      "step": 1502
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4121041591120025,
+      "learning_rate": 1.5430606759282022e-05,
+      "loss": 0.7198,
+      "step": 1503
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.0263065771455606,
+      "learning_rate": 1.542449182155537e-05,
+      "loss": 1.419,
+      "step": 1504
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4843460600152865,
+      "learning_rate": 1.5418374008435896e-05,
+      "loss": 0.7218,
+      "step": 1505
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.020558146203349,
+      "learning_rate": 1.54122533231665e-05,
+      "loss": 1.3829,
+      "step": 1506
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9121895902029453,
+      "learning_rate": 1.5406129768991615e-05,
+      "loss": 1.3714,
+      "step": 1507
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4599812300329966,
+      "learning_rate": 1.540000334915718e-05,
+      "loss": 0.764,
+      "step": 1508
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5610427815807486,
+      "learning_rate": 1.5393874066910677e-05,
+      "loss": 0.7791,
+      "step": 1509
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5060287582381713,
+      "learning_rate": 1.538774192550108e-05,
+      "loss": 0.8397,
+      "step": 1510
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5061751638765415,
+      "learning_rate": 1.53816069281789e-05,
+      "loss": 0.7274,
+      "step": 1511
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4622575406392002,
+      "learning_rate": 1.5375469078196138e-05,
+      "loss": 0.6903,
+      "step": 1512
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5170983418263915,
+      "learning_rate": 1.5369328378806333e-05,
+      "loss": 0.8586,
+      "step": 1513
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4459977494836607,
+      "learning_rate": 1.536318483326452e-05,
+      "loss": 0.7951,
+      "step": 1514
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5214838226514666,
+      "learning_rate": 1.5357038444827243e-05,
+      "loss": 0.7641,
+      "step": 1515
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4362774056570005,
+      "learning_rate": 1.535088921675255e-05,
+      "loss": 0.7615,
+      "step": 1516
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6005687486988212,
+      "learning_rate": 1.5344737152300004e-05,
+      "loss": 0.8079,
+      "step": 1517
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.495866212609544,
+      "learning_rate": 1.533858225473067e-05,
+      "loss": 0.7426,
+      "step": 1518
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.1961255303235854,
+      "learning_rate": 1.533242452730711e-05,
+      "loss": 1.409,
+      "step": 1519
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.085371972412003,
+      "learning_rate": 1.532626397329338e-05,
+      "loss": 1.4025,
+      "step": 1520
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6496930991189565,
+      "learning_rate": 1.532010059595505e-05,
+      "loss": 0.8976,
+      "step": 1521
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.060483391022416,
+      "learning_rate": 1.5313934398559175e-05,
+      "loss": 1.4089,
+      "step": 1522
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.504504601836301,
+      "learning_rate": 1.5307765384374306e-05,
+      "loss": 0.7632,
+      "step": 1523
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5769046944921683,
+      "learning_rate": 1.5301593556670497e-05,
+      "loss": 0.8353,
+      "step": 1524
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6725398448160942,
+      "learning_rate": 1.5295418918719275e-05,
+      "loss": 0.6902,
+      "step": 1525
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.536244869118302,
+      "learning_rate": 1.5289241473793678e-05,
+      "loss": 0.8668,
+      "step": 1526
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4444558420098932,
+      "learning_rate": 1.5283061225168214e-05,
+      "loss": 0.7137,
+      "step": 1527
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.316592924650983,
+      "learning_rate": 1.5276878176118884e-05,
+      "loss": 1.4027,
+      "step": 1528
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4123719797692373,
+      "learning_rate": 1.5270692329923176e-05,
+      "loss": 0.7267,
+      "step": 1529
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.1663350579939755,
+      "learning_rate": 1.5264503689860062e-05,
+      "loss": 1.364,
+      "step": 1530
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5942190792004534,
+      "learning_rate": 1.5258312259209983e-05,
+      "loss": 0.7423,
+      "step": 1531
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6019186443270352,
+      "learning_rate": 1.5252118041254877e-05,
+      "loss": 0.745,
+      "step": 1532
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.0584480150949487,
+      "learning_rate": 1.5245921039278143e-05,
+      "loss": 1.3334,
+      "step": 1533
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4117573766961888,
+      "learning_rate": 1.5239721256564665e-05,
+      "loss": 0.7498,
+      "step": 1534
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.169181694219209,
+      "learning_rate": 1.5233518696400802e-05,
+      "loss": 1.3468,
+      "step": 1535
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0829714076648416,
+      "learning_rate": 1.522731336207438e-05,
+      "loss": 1.3599,
+      "step": 1536
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.472143346834207,
+      "learning_rate": 1.5221105256874697e-05,
+      "loss": 0.686,
+      "step": 1537
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5087291878094617,
+      "learning_rate": 1.521489438409252e-05,
+      "loss": 0.8189,
+      "step": 1538
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.442726785968896,
+      "learning_rate": 1.5208680747020083e-05,
+      "loss": 0.8058,
+      "step": 1539
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4400429076439443,
+      "learning_rate": 1.5202464348951091e-05,
+      "loss": 0.664,
+      "step": 1540
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.1087768777730385,
+      "learning_rate": 1.5196245193180699e-05,
+      "loss": 1.3981,
+      "step": 1541
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.444206816700616,
+      "learning_rate": 1.5190023283005539e-05,
+      "loss": 0.7059,
+      "step": 1542
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3291481353294214,
+      "learning_rate": 1.5183798621723692e-05,
+      "loss": 0.7185,
+      "step": 1543
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3425259716362061,
+      "learning_rate": 1.5177571212634705e-05,
+      "loss": 0.7067,
+      "step": 1544
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5915063395998672,
+      "learning_rate": 1.5171341059039574e-05,
+      "loss": 0.8334,
+      "step": 1545
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4794733625849266,
+      "learning_rate": 1.5165108164240758e-05,
+      "loss": 0.7348,
+      "step": 1546
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4694553760940063,
+      "learning_rate": 1.5158872531542163e-05,
+      "loss": 0.7569,
+      "step": 1547
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.1859087021935566,
+      "learning_rate": 1.5152634164249147e-05,
+      "loss": 1.4729,
+      "step": 1548
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5477523995882956,
+      "learning_rate": 1.5146393065668518e-05,
+      "loss": 0.7554,
+      "step": 1549
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6399576971001322,
+      "learning_rate": 1.5140149239108538e-05,
+      "loss": 0.7657,
+      "step": 1550
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4807481743402544,
+      "learning_rate": 1.5133902687878904e-05,
+      "loss": 0.7099,
+      "step": 1551
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.1132828459665527,
+      "learning_rate": 1.5127653415290761e-05,
+      "loss": 1.3641,
+      "step": 1552
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.56492393213433,
+      "learning_rate": 1.5121401424656704e-05,
+      "loss": 0.7975,
+      "step": 1553
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.267417926610858,
+      "learning_rate": 1.5115146719290761e-05,
+      "loss": 1.2695,
+      "step": 1554
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3949199337608753,
+      "learning_rate": 1.5108889302508398e-05,
+      "loss": 0.7369,
+      "step": 1555
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.955561391638017,
+      "learning_rate": 1.5102629177626527e-05,
+      "loss": 1.3364,
+      "step": 1556
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.544605542881831,
+      "learning_rate": 1.5096366347963485e-05,
+      "loss": 0.808,
+      "step": 1557
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.499968318982523,
+      "learning_rate": 1.509010081683905e-05,
+      "loss": 0.7409,
+      "step": 1558
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3338377331209774,
+      "learning_rate": 1.508383258757443e-05,
+      "loss": 0.7344,
+      "step": 1559
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.571758550178254,
+      "learning_rate": 1.5077561663492259e-05,
+      "loss": 0.731,
+      "step": 1560
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.3116878972209585,
+      "learning_rate": 1.507128804791661e-05,
+      "loss": 1.3652,
+      "step": 1561
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.1737807836780596,
+      "learning_rate": 1.5065011744172968e-05,
+      "loss": 1.363,
+      "step": 1562
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4832827734094043,
+      "learning_rate": 1.5058732755588259e-05,
+      "loss": 0.7954,
+      "step": 1563
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.391627535856719,
+      "learning_rate": 1.5052451085490816e-05,
+      "loss": 0.7268,
+      "step": 1564
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3894639024300255,
+      "learning_rate": 1.5046166737210406e-05,
+      "loss": 0.764,
+      "step": 1565
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5644792200194604,
+      "learning_rate": 1.5039879714078211e-05,
+      "loss": 0.7965,
+      "step": 1566
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5622624485120777,
+      "learning_rate": 1.5033590019426832e-05,
+      "loss": 0.9435,
+      "step": 1567
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4297720223055657,
+      "learning_rate": 1.5027297656590282e-05,
+      "loss": 0.7599,
+      "step": 1568
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.520579683998602,
+      "learning_rate": 1.5021002628903994e-05,
+      "loss": 0.7659,
+      "step": 1569
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.2907048678057382,
+      "learning_rate": 1.5014704939704806e-05,
+      "loss": 1.3491,
+      "step": 1570
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.082048723161923,
+      "learning_rate": 1.5008404592330981e-05,
+      "loss": 1.3875,
+      "step": 1571
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4738289020363307,
+      "learning_rate": 1.5002101590122173e-05,
+      "loss": 0.9028,
+      "step": 1572
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5658773580460084,
+      "learning_rate": 1.4995795936419454e-05,
+      "loss": 0.7855,
+      "step": 1573
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9984685363146142,
+      "learning_rate": 1.4989487634565302e-05,
+      "loss": 1.3177,
+      "step": 1574
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4177312726426352,
+      "learning_rate": 1.4983176687903598e-05,
+      "loss": 0.6975,
+      "step": 1575
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4438217343366737,
+      "learning_rate": 1.4976863099779617e-05,
+      "loss": 0.7327,
+      "step": 1576
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4278666795710078,
+      "learning_rate": 1.4970546873540046e-05,
+      "loss": 0.812,
+      "step": 1577
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5118958620348815,
+      "learning_rate": 1.4964228012532963e-05,
+      "loss": 0.8372,
+      "step": 1578
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3632983126619886,
+      "learning_rate": 1.4957906520107846e-05,
+      "loss": 0.7979,
+      "step": 1579
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4161581967367003,
+      "learning_rate": 1.495158239961556e-05,
+      "loss": 0.6971,
+      "step": 1580
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.108374337587779,
+      "learning_rate": 1.4945255654408378e-05,
+      "loss": 1.3688,
+      "step": 1581
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.477008215983394,
+      "learning_rate": 1.493892628783995e-05,
+      "loss": 0.7904,
+      "step": 1582
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4378032034558623,
+      "learning_rate": 1.4932594303265324e-05,
+      "loss": 0.728,
+      "step": 1583
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4778716197084423,
+      "learning_rate": 1.492625970404093e-05,
+      "loss": 0.7358,
+      "step": 1584
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.451153998178174,
+      "learning_rate": 1.4919922493524588e-05,
+      "loss": 0.7667,
+      "step": 1585
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5211761810892175,
+      "learning_rate": 1.4913582675075504e-05,
+      "loss": 0.7621,
+      "step": 1586
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4840623349394977,
+      "learning_rate": 1.490724025205426e-05,
+      "loss": 0.7411,
+      "step": 1587
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3786768539352705,
+      "learning_rate": 1.4900895227822823e-05,
+      "loss": 0.7329,
+      "step": 1588
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.503122891186697,
+      "learning_rate": 1.4894547605744541e-05,
+      "loss": 0.7456,
+      "step": 1589
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0595510555883543,
+      "learning_rate": 1.4888197389184129e-05,
+      "loss": 1.3585,
+      "step": 1590
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4773474099556474,
+      "learning_rate": 1.4881844581507692e-05,
+      "loss": 0.8185,
+      "step": 1591
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5048516067627016,
+      "learning_rate": 1.4875489186082696e-05,
+      "loss": 0.8256,
+      "step": 1592
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5337089080687842,
+      "learning_rate": 1.4869131206277984e-05,
+      "loss": 0.7285,
+      "step": 1593
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5098149146251871,
+      "learning_rate": 1.486277064546377e-05,
+      "loss": 0.7968,
+      "step": 1594
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5068464231132532,
+      "learning_rate": 1.4856407507011635e-05,
+      "loss": 0.7852,
+      "step": 1595
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4001250010137243,
+      "learning_rate": 1.4850041794294523e-05,
+      "loss": 0.7734,
+      "step": 1596
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4407207752509155,
+      "learning_rate": 1.4843673510686746e-05,
+      "loss": 0.7763,
+      "step": 1597
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4668175180423224,
+      "learning_rate": 1.4837302659563981e-05,
+      "loss": 0.8086,
+      "step": 1598
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.479491976226391,
+      "learning_rate": 1.4830929244303262e-05,
+      "loss": 0.9206,
+      "step": 1599
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0222108177438,
+      "learning_rate": 1.4824553268282982e-05,
+      "loss": 1.4218,
+      "step": 1600
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4470152874604691,
+      "learning_rate": 1.481817473488289e-05,
+      "loss": 0.7225,
+      "step": 1601
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.003165900041584,
+      "learning_rate": 1.4811793647484103e-05,
+      "loss": 1.327,
+      "step": 1602
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.458512270825858,
+      "learning_rate": 1.4805410009469072e-05,
+      "loss": 0.8311,
+      "step": 1603
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9719150124839342,
+      "learning_rate": 1.4799023824221614e-05,
+      "loss": 1.4155,
+      "step": 1604
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5403581814036578,
+      "learning_rate": 1.4792635095126893e-05,
+      "loss": 0.8397,
+      "step": 1605
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3254405512306917,
+      "learning_rate": 1.478624382557142e-05,
+      "loss": 0.7159,
+      "step": 1606
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6360641610238371,
+      "learning_rate": 1.4779850018943054e-05,
+      "loss": 0.6865,
+      "step": 1607
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5332522944847082,
+      "learning_rate": 1.4773453678630998e-05,
+      "loss": 0.7343,
+      "step": 1608
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.230992481880025,
+      "learning_rate": 1.47670548080258e-05,
+      "loss": 1.331,
+      "step": 1609
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.1436650869533085,
+      "learning_rate": 1.4760653410519345e-05,
+      "loss": 1.4182,
+      "step": 1610
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4217323605376737,
+      "learning_rate": 1.475424948950486e-05,
+      "loss": 0.7656,
+      "step": 1611
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3071954037725342,
+      "learning_rate": 1.4747843048376914e-05,
+      "loss": 0.6925,
+      "step": 1612
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4212870565931068,
+      "learning_rate": 1.4741434090531405e-05,
+      "loss": 0.7773,
+      "step": 1613
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3778190443320844,
+      "learning_rate": 1.4735022619365564e-05,
+      "loss": 0.7545,
+      "step": 1614
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.469413834469898,
+      "learning_rate": 1.4728608638277964e-05,
+      "loss": 0.7762,
+      "step": 1615
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.519042575110362,
+      "learning_rate": 1.4722192150668495e-05,
+      "loss": 0.7666,
+      "step": 1616
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5012461395211136,
+      "learning_rate": 1.471577315993839e-05,
+      "loss": 0.785,
+      "step": 1617
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4626922003450449,
+      "learning_rate": 1.47093516694902e-05,
+      "loss": 0.7873,
+      "step": 1618
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5126174764629186,
+      "learning_rate": 1.4702927682727798e-05,
+      "loss": 0.7632,
+      "step": 1619
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.51444126639849,
+      "learning_rate": 1.4696501203056395e-05,
+      "loss": 0.8335,
+      "step": 1620
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.3412379377032946,
+      "learning_rate": 1.4690072233882502e-05,
+      "loss": 1.3368,
+      "step": 1621
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.2813586035440956,
+      "learning_rate": 1.468364077861397e-05,
+      "loss": 1.3387,
+      "step": 1622
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4252546445441354,
+      "learning_rate": 1.4677206840659956e-05,
+      "loss": 0.7415,
+      "step": 1623
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.0196740057169076,
+      "learning_rate": 1.4670770423430933e-05,
+      "loss": 1.377,
+      "step": 1624
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.106350832136813,
+      "learning_rate": 1.4664331530338696e-05,
+      "loss": 1.3766,
+      "step": 1625
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5189921029178868,
+      "learning_rate": 1.4657890164796345e-05,
+      "loss": 0.7638,
+      "step": 1626
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7255637373130195,
+      "learning_rate": 1.4651446330218292e-05,
+      "loss": 0.7659,
+      "step": 1627
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4920090578576208,
+      "learning_rate": 1.464500003002026e-05,
+      "loss": 0.7826,
+      "step": 1628
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.370204802588797,
+      "learning_rate": 1.463855126761928e-05,
+      "loss": 1.4086,
+      "step": 1629
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.3183289063076384,
+      "learning_rate": 1.4632100046433686e-05,
+      "loss": 1.3359,
+      "step": 1630
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5295867776732923,
+      "learning_rate": 1.4625646369883113e-05,
+      "loss": 0.7586,
+      "step": 1631
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.572454407146711,
+      "learning_rate": 1.4619190241388501e-05,
+      "loss": 0.7314,
+      "step": 1632
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5507283607780655,
+      "learning_rate": 1.461273166437209e-05,
+      "loss": 0.8406,
+      "step": 1633
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4564455422426408,
+      "learning_rate": 1.460627064225741e-05,
+      "loss": 0.8009,
+      "step": 1634
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5132896029198792,
+      "learning_rate": 1.4599807178469301e-05,
+      "loss": 0.8057,
+      "step": 1635
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3015041674604555,
+      "learning_rate": 1.4593341276433888e-05,
+      "loss": 0.7649,
+      "step": 1636
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.4421921053822686,
+      "learning_rate": 1.4586872939578589e-05,
+      "loss": 1.4031,
+      "step": 1637
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4434494331455063,
+      "learning_rate": 1.4580402171332115e-05,
+      "loss": 0.7132,
+      "step": 1638
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3924968045580508,
+      "learning_rate": 1.4573928975124466e-05,
+      "loss": 0.7701,
+      "step": 1639
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.34483686063842,
+      "learning_rate": 1.4567453354386925e-05,
+      "loss": 0.7104,
+      "step": 1640
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.0524819290188887,
+      "learning_rate": 1.4560975312552066e-05,
+      "loss": 1.3875,
+      "step": 1641
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5556640983690584,
+      "learning_rate": 1.4554494853053742e-05,
+      "loss": 0.7709,
+      "step": 1642
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4419097370660263,
+      "learning_rate": 1.4548011979327088e-05,
+      "loss": 0.6819,
+      "step": 1643
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4339521544918477,
+      "learning_rate": 1.4541526694808524e-05,
+      "loss": 0.7333,
+      "step": 1644
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4187831314239545,
+      "learning_rate": 1.4535039002935735e-05,
+      "loss": 0.7243,
+      "step": 1645
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.427501983298237,
+      "learning_rate": 1.4528548907147697e-05,
+      "loss": 0.7267,
+      "step": 1646
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5090227467660078,
+      "learning_rate": 1.4522056410884653e-05,
+      "loss": 0.7422,
+      "step": 1647
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.503830937244655,
+      "learning_rate": 1.451556151758812e-05,
+      "loss": 0.8345,
+      "step": 1648
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.1620465516770455,
+      "learning_rate": 1.4509064230700881e-05,
+      "loss": 1.3893,
+      "step": 1649
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.593821808910323,
+      "learning_rate": 1.4502564553666998e-05,
+      "loss": 0.8515,
+      "step": 1650
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5469931958526724,
+      "learning_rate": 1.4496062489931788e-05,
+      "loss": 0.8167,
+      "step": 1651
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5035820653894658,
+      "learning_rate": 1.4489558042941844e-05,
+      "loss": 0.7732,
+      "step": 1652
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3336067287107558,
+      "learning_rate": 1.4483051216145012e-05,
+      "loss": 0.714,
+      "step": 1653
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.989575792606994,
+      "learning_rate": 1.4476542012990412e-05,
+      "loss": 1.3951,
+      "step": 1654
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4061245934570377,
+      "learning_rate": 1.4470030436928404e-05,
+      "loss": 0.6997,
+      "step": 1655
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4591248293829044,
+      "learning_rate": 1.4463516491410627e-05,
+      "loss": 0.792,
+      "step": 1656
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.0846427955001356,
+      "learning_rate": 1.4457000179889966e-05,
+      "loss": 1.3413,
+      "step": 1657
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4190792212764236,
+      "learning_rate": 1.4450481505820562e-05,
+      "loss": 0.7247,
+      "step": 1658
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.1271095420336343,
+      "learning_rate": 1.4443960472657803e-05,
+      "loss": 1.4281,
+      "step": 1659
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5993945960018483,
+      "learning_rate": 1.4437437083858337e-05,
+      "loss": 0.8195,
+      "step": 1660
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4345353762069002,
+      "learning_rate": 1.4430911342880051e-05,
+      "loss": 0.748,
+      "step": 1661
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4487499053041963,
+      "learning_rate": 1.442438325318209e-05,
+      "loss": 0.7432,
+      "step": 1662
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.350664843973476,
+      "learning_rate": 1.4417852818224833e-05,
+      "loss": 0.7577,
+      "step": 1663
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3936412179088062,
+      "learning_rate": 1.441132004146991e-05,
+      "loss": 0.6759,
+      "step": 1664
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4780276205391607,
+      "learning_rate": 1.4404784926380183e-05,
+      "loss": 0.7177,
+      "step": 1665
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4404100251435608,
+      "learning_rate": 1.4398247476419764e-05,
+      "loss": 0.7495,
+      "step": 1666
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.305347607597455,
+      "learning_rate": 1.4391707695053999e-05,
+      "loss": 1.4029,
+      "step": 1667
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5468687656088058,
+      "learning_rate": 1.4385165585749463e-05,
+      "loss": 0.8167,
+      "step": 1668
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4567411507643622,
+      "learning_rate": 1.437862115197398e-05,
+      "loss": 0.6958,
+      "step": 1669
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4337860490834864,
+      "learning_rate": 1.437207439719659e-05,
+      "loss": 0.7473,
+      "step": 1670
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4804242635177682,
+      "learning_rate": 1.4365525324887573e-05,
+      "loss": 0.7134,
+      "step": 1671
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4636459588693944,
+      "learning_rate": 1.4358973938518435e-05,
+      "loss": 0.7772,
+      "step": 1672
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3713256928913269,
+      "learning_rate": 1.435242024156191e-05,
+      "loss": 0.7318,
+      "step": 1673
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.411405350528879,
+      "learning_rate": 1.4345864237491956e-05,
+      "loss": 0.7506,
+      "step": 1674
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4831463662903244,
+      "learning_rate": 1.4339305929783748e-05,
+      "loss": 0.6987,
+      "step": 1675
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.98236781683687,
+      "learning_rate": 1.4332745321913693e-05,
+      "loss": 1.3388,
+      "step": 1676
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.169724746937146,
+      "learning_rate": 1.4326182417359407e-05,
+      "loss": 1.3622,
+      "step": 1677
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5874778372115925,
+      "learning_rate": 1.4319617219599731e-05,
+      "loss": 0.8962,
+      "step": 1678
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9825863093585296,
+      "learning_rate": 1.4313049732114716e-05,
+      "loss": 1.3688,
+      "step": 1679
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.353800299835086,
+      "learning_rate": 1.4306479958385633e-05,
+      "loss": 0.664,
+      "step": 1680
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4337979344854246,
+      "learning_rate": 1.429990790189496e-05,
+      "loss": 0.8461,
+      "step": 1681
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4278197960274748,
+      "learning_rate": 1.4293333566126386e-05,
+      "loss": 0.7106,
+      "step": 1682
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.3242920789455606,
+      "learning_rate": 1.4286756954564808e-05,
+      "loss": 1.3091,
+      "step": 1683
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.467989039760585,
+      "learning_rate": 1.4280178070696332e-05,
+      "loss": 0.7872,
+      "step": 1684
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.500593200852713,
+      "learning_rate": 1.4273596918008263e-05,
+      "loss": 0.8443,
+      "step": 1685
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.0237162220898917,
+      "learning_rate": 1.4267013499989113e-05,
+      "loss": 1.3599,
+      "step": 1686
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9714492670011523,
+      "learning_rate": 1.4260427820128594e-05,
+      "loss": 1.3864,
+      "step": 1687
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3112589449947338,
+      "learning_rate": 1.4253839881917619e-05,
+      "loss": 0.754,
+      "step": 1688
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.35829324918678,
+      "learning_rate": 1.4247249688848293e-05,
+      "loss": 0.7201,
+      "step": 1689
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5318391478888067,
+      "learning_rate": 1.4240657244413918e-05,
+      "loss": 0.7659,
+      "step": 1690
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4858161195454844,
+      "learning_rate": 1.4234062552108993e-05,
+      "loss": 0.8298,
+      "step": 1691
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.485196659164032,
+      "learning_rate": 1.4227465615429206e-05,
+      "loss": 0.7286,
+      "step": 1692
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.329500969688324,
+      "learning_rate": 1.4220866437871432e-05,
+      "loss": 1.3721,
+      "step": 1693
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4142703919904938,
+      "learning_rate": 1.4214265022933744e-05,
+      "loss": 0.7271,
+      "step": 1694
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4388808574779364,
+      "learning_rate": 1.4207661374115384e-05,
+      "loss": 0.7106,
+      "step": 1695
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.37831240264064,
+      "learning_rate": 1.4201055494916793e-05,
+      "loss": 0.7397,
+      "step": 1696
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.453566055375399,
+      "learning_rate": 1.4194447388839585e-05,
+      "loss": 0.6944,
+      "step": 1697
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5535921709498124,
+      "learning_rate": 1.4187837059386564e-05,
+      "loss": 0.7562,
+      "step": 1698
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4815690651525477,
+      "learning_rate": 1.4181224510061702e-05,
+      "loss": 0.7562,
+      "step": 1699
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.555981709317436,
+      "learning_rate": 1.4174609744370153e-05,
+      "loss": 0.7842,
+      "step": 1700
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.528446795854986,
+      "learning_rate": 1.4167992765818248e-05,
+      "loss": 0.7722,
+      "step": 1701
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.510251107425476,
+      "learning_rate": 1.4161373577913486e-05,
+      "loss": 0.7346,
+      "step": 1702
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4707089007183178,
+      "learning_rate": 1.4154752184164542e-05,
+      "loss": 0.7346,
+      "step": 1703
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.051300700166994,
+      "learning_rate": 1.4148128588081258e-05,
+      "loss": 1.3235,
+      "step": 1704
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9530614830361952,
+      "learning_rate": 1.4141502793174639e-05,
+      "loss": 1.3759,
+      "step": 1705
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4928731616896445,
+      "learning_rate": 1.4134874802956866e-05,
+      "loss": 0.7434,
+      "step": 1706
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.526454622034699,
+      "learning_rate": 1.4128244620941271e-05,
+      "loss": 0.8539,
+      "step": 1707
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3980228195209377,
+      "learning_rate": 1.4121612250642357e-05,
+      "loss": 0.7992,
+      "step": 1708
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5707904116334626,
+      "learning_rate": 1.4114977695575784e-05,
+      "loss": 0.7713,
+      "step": 1709
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4533564866763427,
+      "learning_rate": 1.4108340959258371e-05,
+      "loss": 0.7952,
+      "step": 1710
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6435233960327582,
+      "learning_rate": 1.4101702045208093e-05,
+      "loss": 0.7881,
+      "step": 1711
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.436838625390399,
+      "learning_rate": 1.4095060956944075e-05,
+      "loss": 0.748,
+      "step": 1712
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4657100814450208,
+      "learning_rate": 1.4088417697986603e-05,
+      "loss": 0.7932,
+      "step": 1713
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5174371189193083,
+      "learning_rate": 1.4081772271857108e-05,
+      "loss": 0.8423,
+      "step": 1714
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4396586553256556,
+      "learning_rate": 1.4075124682078169e-05,
+      "loss": 0.8474,
+      "step": 1715
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.39754160188507,
+      "learning_rate": 1.4068474932173517e-05,
+      "loss": 0.7973,
+      "step": 1716
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4341433150869582,
+      "learning_rate": 1.406182302566802e-05,
+      "loss": 0.7165,
+      "step": 1717
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4863283717975742,
+      "learning_rate": 1.4055168966087698e-05,
+      "loss": 0.6917,
+      "step": 1718
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3453410502653338,
+      "learning_rate": 1.4048512756959705e-05,
+      "loss": 0.8428,
+      "step": 1719
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4996302272370716,
+      "learning_rate": 1.4041854401812338e-05,
+      "loss": 0.8704,
+      "step": 1720
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.399243764826014,
+      "learning_rate": 1.4035193904175033e-05,
+      "loss": 0.7361,
+      "step": 1721
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.091459016787623,
+      "learning_rate": 1.402853126757836e-05,
+      "loss": 1.3802,
+      "step": 1722
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3109526769178539,
+      "learning_rate": 1.4021866495554021e-05,
+      "loss": 0.5938,
+      "step": 1723
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.990126745324797,
+      "learning_rate": 1.4015199591634851e-05,
+      "loss": 1.3206,
+      "step": 1724
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.420165651684252,
+      "learning_rate": 1.4008530559354822e-05,
+      "loss": 0.6782,
+      "step": 1725
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4648277772837035,
+      "learning_rate": 1.4001859402249021e-05,
+      "loss": 0.7309,
+      "step": 1726
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4518643191061853,
+      "learning_rate": 1.3995186123853669e-05,
+      "loss": 0.7632,
+      "step": 1727
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9997275497366331,
+      "learning_rate": 1.3988510727706107e-05,
+      "loss": 1.3799,
+      "step": 1728
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4924754958905806,
+      "learning_rate": 1.3981833217344811e-05,
+      "loss": 0.7638,
+      "step": 1729
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6154244257800636,
+      "learning_rate": 1.3975153596309363e-05,
+      "loss": 0.6864,
+      "step": 1730
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5180143652920877,
+      "learning_rate": 1.3968471868140472e-05,
+      "loss": 0.7627,
+      "step": 1731
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4470301260248168,
+      "learning_rate": 1.396178803637996e-05,
+      "loss": 0.7506,
+      "step": 1732
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4760393184958163,
+      "learning_rate": 1.3955102104570766e-05,
+      "loss": 0.7485,
+      "step": 1733
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.30033624671771,
+      "learning_rate": 1.3948414076256945e-05,
+      "loss": 0.6608,
+      "step": 1734
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4426145734955431,
+      "learning_rate": 1.3941723954983656e-05,
+      "loss": 0.7511,
+      "step": 1735
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.39663023150164,
+      "learning_rate": 1.3935031744297174e-05,
+      "loss": 0.7725,
+      "step": 1736
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4742237394007935,
+      "learning_rate": 1.3928337447744877e-05,
+      "loss": 0.7416,
+      "step": 1737
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4844184484190612,
+      "learning_rate": 1.3921641068875254e-05,
+      "loss": 0.7528,
+      "step": 1738
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4034416150611952,
+      "learning_rate": 1.391494261123789e-05,
+      "loss": 0.6743,
+      "step": 1739
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4376525774381532,
+      "learning_rate": 1.3908242078383479e-05,
+      "loss": 0.7512,
+      "step": 1740
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5397334105364873,
+      "learning_rate": 1.3901539473863813e-05,
+      "loss": 0.7239,
+      "step": 1741
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3302579785817161,
+      "learning_rate": 1.389483480123178e-05,
+      "loss": 0.7514,
+      "step": 1742
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.441401367042654,
+      "learning_rate": 1.3888128064041371e-05,
+      "loss": 0.768,
+      "step": 1743
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4126571567862347,
+      "learning_rate": 1.388141926584766e-05,
+      "loss": 0.7609,
+      "step": 1744
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0830418063676697,
+      "learning_rate": 1.3874708410206824e-05,
+      "loss": 1.3762,
+      "step": 1745
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2969183249247858,
+      "learning_rate": 1.3867995500676123e-05,
+      "loss": 0.6722,
+      "step": 1746
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9739136672565392,
+      "learning_rate": 1.3861280540813916e-05,
+      "loss": 1.3404,
+      "step": 1747
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.002611189917482,
+      "learning_rate": 1.385456353417963e-05,
+      "loss": 1.3681,
+      "step": 1748
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.552031114320976,
+      "learning_rate": 1.3847844484333799e-05,
+      "loss": 0.7835,
+      "step": 1749
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4769718893096697,
+      "learning_rate": 1.3841123394838025e-05,
+      "loss": 0.7458,
+      "step": 1750
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4850209022289713,
+      "learning_rate": 1.3834400269255e-05,
+      "loss": 0.7815,
+      "step": 1751
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.46286686350256,
+      "learning_rate": 1.3827675111148486e-05,
+      "loss": 0.8118,
+      "step": 1752
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9176632472525077,
+      "learning_rate": 1.3820947924083332e-05,
+      "loss": 1.3497,
+      "step": 1753
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3921280458662273,
+      "learning_rate": 1.3814218711625457e-05,
+      "loss": 0.7339,
+      "step": 1754
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3945115748624084,
+      "learning_rate": 1.3807487477341857e-05,
+      "loss": 0.7596,
+      "step": 1755
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3973633376452204,
+      "learning_rate": 1.3800754224800597e-05,
+      "loss": 0.7859,
+      "step": 1756
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5513860382841784,
+      "learning_rate": 1.379401895757081e-05,
+      "loss": 0.7442,
+      "step": 1757
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.374605975590899,
+      "learning_rate": 1.3787281679222697e-05,
+      "loss": 0.7401,
+      "step": 1758
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.084889469460892,
+      "learning_rate": 1.3780542393327534e-05,
+      "loss": 1.343,
+      "step": 1759
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5224843973097624,
+      "learning_rate": 1.377380110345765e-05,
+      "loss": 0.7538,
+      "step": 1760
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4433964729887399,
+      "learning_rate": 1.3767057813186444e-05,
+      "loss": 0.7425,
+      "step": 1761
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.396675975334998,
+      "learning_rate": 1.376031252608837e-05,
+      "loss": 0.6759,
+      "step": 1762
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4749962413884665,
+      "learning_rate": 1.3753565245738943e-05,
+      "loss": 0.7332,
+      "step": 1763
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5540726621532943,
+      "learning_rate": 1.3746815975714734e-05,
+      "loss": 0.7892,
+      "step": 1764
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3212425356051951,
+      "learning_rate": 1.3740064719593372e-05,
+      "loss": 0.7835,
+      "step": 1765
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5657834140863192,
+      "learning_rate": 1.373331148095353e-05,
+      "loss": 0.9233,
+      "step": 1766
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.0560915874962054,
+      "learning_rate": 1.3726556263374944e-05,
+      "loss": 1.378,
+      "step": 1767
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3792657636941563,
+      "learning_rate": 1.3719799070438388e-05,
+      "loss": 0.6736,
+      "step": 1768
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5108340755231005,
+      "learning_rate": 1.3713039905725689e-05,
+      "loss": 0.7746,
+      "step": 1769
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4738731373195728,
+      "learning_rate": 1.3706278772819716e-05,
+      "loss": 0.7167,
+      "step": 1770
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4444864530059938,
+      "learning_rate": 1.3699515675304387e-05,
+      "loss": 0.6821,
+      "step": 1771
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5107391353358597,
+      "learning_rate": 1.3692750616764654e-05,
+      "loss": 0.7773,
+      "step": 1772
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.364089213408197,
+      "learning_rate": 1.3685983600786516e-05,
+      "loss": 0.6916,
+      "step": 1773
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4716062880831469,
+      "learning_rate": 1.3679214630957e-05,
+      "loss": 0.7841,
+      "step": 1774
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4599939988635808,
+      "learning_rate": 1.367244371086418e-05,
+      "loss": 0.7384,
+      "step": 1775
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5496758178372305,
+      "learning_rate": 1.3665670844097158e-05,
+      "loss": 0.7476,
+      "step": 1776
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4746945701641345,
+      "learning_rate": 1.3658896034246065e-05,
+      "loss": 0.7012,
+      "step": 1777
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4905714726336492,
+      "learning_rate": 1.3652119284902066e-05,
+      "loss": 0.78,
+      "step": 1778
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.0208858865233936,
+      "learning_rate": 1.364534059965735e-05,
+      "loss": 1.3335,
+      "step": 1779
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4634952537416965,
+      "learning_rate": 1.3638559982105139e-05,
+      "loss": 0.7887,
+      "step": 1780
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.0828731605171336,
+      "learning_rate": 1.3631777435839672e-05,
+      "loss": 1.3806,
+      "step": 1781
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.422022713102543,
+      "learning_rate": 1.362499296445622e-05,
+      "loss": 0.7694,
+      "step": 1782
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.397111934911135,
+      "learning_rate": 1.3618206571551064e-05,
+      "loss": 0.7873,
+      "step": 1783
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4400869510506638,
+      "learning_rate": 1.3611418260721507e-05,
+      "loss": 0.7531,
+      "step": 1784
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4195724923033148,
+      "learning_rate": 1.3604628035565872e-05,
+      "loss": 0.7486,
+      "step": 1785
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.0561721803146784,
+      "learning_rate": 1.3597835899683495e-05,
+      "loss": 1.3798,
+      "step": 1786
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5205643480080582,
+      "learning_rate": 1.3591041856674723e-05,
+      "loss": 0.7832,
+      "step": 1787
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.014326096176442,
+      "learning_rate": 1.3584245910140914e-05,
+      "loss": 1.3648,
+      "step": 1788
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3837666896283372,
+      "learning_rate": 1.3577448063684436e-05,
+      "loss": 0.7311,
+      "step": 1789
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4688412438613248,
+      "learning_rate": 1.3570648320908664e-05,
+      "loss": 0.7859,
+      "step": 1790
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3975435880802847,
+      "learning_rate": 1.3563846685417978e-05,
+      "loss": 0.7372,
+      "step": 1791
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4980177415367226,
+      "learning_rate": 1.3557043160817762e-05,
+      "loss": 0.7176,
+      "step": 1792
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.479317574051509,
+      "learning_rate": 1.3550237750714406e-05,
+      "loss": 0.7782,
+      "step": 1793
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5179734235162214,
+      "learning_rate": 1.3543430458715285e-05,
+      "loss": 0.7964,
+      "step": 1794
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.426947940520781,
+      "learning_rate": 1.3536621288428789e-05,
+      "loss": 0.6448,
+      "step": 1795
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4257182898162677,
+      "learning_rate": 1.3529810243464292e-05,
+      "loss": 0.7612,
+      "step": 1796
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3079708926713358,
+      "learning_rate": 1.3522997327432167e-05,
+      "loss": 0.73,
+      "step": 1797
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.530772499620138,
+      "learning_rate": 1.3516182543943774e-05,
+      "loss": 0.7286,
+      "step": 1798
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.371742755148677,
+      "learning_rate": 1.350936589661147e-05,
+      "loss": 0.7657,
+      "step": 1799
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4560008597273266,
+      "learning_rate": 1.3502547389048594e-05,
+      "loss": 0.7987,
+      "step": 1800
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.398473791981301,
+      "learning_rate": 1.3495727024869471e-05,
+      "loss": 0.7967,
+      "step": 1801
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.0655634683840196,
+      "learning_rate": 1.3488904807689419e-05,
+      "loss": 1.4023,
+      "step": 1802
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5215046463274662,
+      "learning_rate": 1.3482080741124723e-05,
+      "loss": 0.7629,
+      "step": 1803
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9489407316562077,
+      "learning_rate": 1.347525482879266e-05,
+      "loss": 1.365,
+      "step": 1804
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.0092724392106205,
+      "learning_rate": 1.3468427074311485e-05,
+      "loss": 1.3502,
+      "step": 1805
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3911411621344076,
+      "learning_rate": 1.3461597481300426e-05,
+      "loss": 0.6718,
+      "step": 1806
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5012996774607803,
+      "learning_rate": 1.3454766053379685e-05,
+      "loss": 0.7523,
+      "step": 1807
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.493923087549506,
+      "learning_rate": 1.3447932794170436e-05,
+      "loss": 0.7726,
+      "step": 1808
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.060965316753009,
+      "learning_rate": 1.3441097707294828e-05,
+      "loss": 1.4073,
+      "step": 1809
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4930503622616085,
+      "learning_rate": 1.3434260796375976e-05,
+      "loss": 0.7872,
+      "step": 1810
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4520131086177055,
+      "learning_rate": 1.3427422065037958e-05,
+      "loss": 0.7985,
+      "step": 1811
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5042006113440654,
+      "learning_rate": 1.3420581516905828e-05,
+      "loss": 0.7733,
+      "step": 1812
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3104749087590348,
+      "learning_rate": 1.3413739155605594e-05,
+      "loss": 0.669,
+      "step": 1813
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.385927318187178,
+      "learning_rate": 1.3406894984764227e-05,
+      "loss": 0.7372,
+      "step": 1814
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.163389508436242,
+      "learning_rate": 1.3400049008009655e-05,
+      "loss": 1.3282,
+      "step": 1815
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.574206387836042,
+      "learning_rate": 1.339320122897077e-05,
+      "loss": 0.865,
+      "step": 1816
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.473967993507167,
+      "learning_rate": 1.3386351651277415e-05,
+      "loss": 0.7801,
+      "step": 1817
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.394297505332185,
+      "learning_rate": 1.3379500278560383e-05,
+      "loss": 0.6753,
+      "step": 1818
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4072034699568907,
+      "learning_rate": 1.3372647114451422e-05,
+      "loss": 0.7578,
+      "step": 1819
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.497392763555299,
+      "learning_rate": 1.3365792162583237e-05,
+      "loss": 0.8091,
+      "step": 1820
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.532246497818356,
+      "learning_rate": 1.3358935426589466e-05,
+      "loss": 0.7646,
+      "step": 1821
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.1874916038033936,
+      "learning_rate": 1.33520769101047e-05,
+      "loss": 1.4521,
+      "step": 1822
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.575625048067617,
+      "learning_rate": 1.3345216616764478e-05,
+      "loss": 0.8253,
+      "step": 1823
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.554257069509092,
+      "learning_rate": 1.3338354550205274e-05,
+      "loss": 0.8033,
+      "step": 1824
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.033724486003009,
+      "learning_rate": 1.3331490714064506e-05,
+      "loss": 1.409,
+      "step": 1825
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3478966271661108,
+      "learning_rate": 1.332462511198053e-05,
+      "loss": 0.7226,
+      "step": 1826
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4695712047940357,
+      "learning_rate": 1.331775774759263e-05,
+      "loss": 0.8078,
+      "step": 1827
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4194091485283709,
+      "learning_rate": 1.331088862454104e-05,
+      "loss": 0.8263,
+      "step": 1828
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4251482425567743,
+      "learning_rate": 1.3304017746466912e-05,
+      "loss": 0.7881,
+      "step": 1829
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5058187030666572,
+      "learning_rate": 1.3297145117012336e-05,
+      "loss": 0.7403,
+      "step": 1830
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.051054635510419,
+      "learning_rate": 1.3290270739820326e-05,
+      "loss": 1.3806,
+      "step": 1831
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2774193362105533,
+      "learning_rate": 1.3283394618534821e-05,
+      "loss": 0.7674,
+      "step": 1832
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5357323509578547,
+      "learning_rate": 1.3276516756800692e-05,
+      "loss": 0.7107,
+      "step": 1833
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.095049797824425,
+      "learning_rate": 1.326963715826373e-05,
+      "loss": 1.3752,
+      "step": 1834
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3610579435849992,
+      "learning_rate": 1.3262755826570642e-05,
+      "loss": 0.7617,
+      "step": 1835
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.452263880918813,
+      "learning_rate": 1.3255872765369055e-05,
+      "loss": 0.7323,
+      "step": 1836
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4669069271031143,
+      "learning_rate": 1.3248987978307517e-05,
+      "loss": 0.7589,
+      "step": 1837
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.053473896729741,
+      "learning_rate": 1.3242101469035491e-05,
+      "loss": 1.4055,
+      "step": 1838
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.0069697778972695,
+      "learning_rate": 1.3235213241203344e-05,
+      "loss": 1.3444,
+      "step": 1839
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9618565147837737,
+      "learning_rate": 1.3228323298462362e-05,
+      "loss": 1.3684,
+      "step": 1840
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4699064768676766,
+      "learning_rate": 1.3221431644464743e-05,
+      "loss": 0.7602,
+      "step": 1841
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3803571327337933,
+      "learning_rate": 1.321453828286358e-05,
+      "loss": 0.7894,
+      "step": 1842
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4346184788208236,
+      "learning_rate": 1.3207643217312884e-05,
+      "loss": 0.6898,
+      "step": 1843
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5268167945324889,
+      "learning_rate": 1.3200746451467561e-05,
+      "loss": 0.8572,
+      "step": 1844
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.0746681081065,
+      "learning_rate": 1.3193847988983422e-05,
+      "loss": 1.3443,
+      "step": 1845
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4365498120650566,
+      "learning_rate": 1.3186947833517172e-05,
+      "loss": 0.8096,
+      "step": 1846
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.085939333005549,
+      "learning_rate": 1.3180045988726423e-05,
+      "loss": 1.4161,
+      "step": 1847
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.639923805718554,
+      "learning_rate": 1.3173142458269679e-05,
+      "loss": 0.761,
+      "step": 1848
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4648169211297204,
+      "learning_rate": 1.3166237245806333e-05,
+      "loss": 0.7953,
+      "step": 1849
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.416307477487777,
+      "learning_rate": 1.3159330354996667e-05,
+      "loss": 0.7333,
+      "step": 1850
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.16299671360716,
+      "learning_rate": 1.3152421789501865e-05,
+      "loss": 1.3617,
+      "step": 1851
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3542474035789334,
+      "learning_rate": 1.3145511552983989e-05,
+      "loss": 0.6295,
+      "step": 1852
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4401856341842367,
+      "learning_rate": 1.3138599649105988e-05,
+      "loss": 0.6897,
+      "step": 1853
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.42589808373227,
+      "learning_rate": 1.3131686081531698e-05,
+      "loss": 0.72,
+      "step": 1854
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.0389036331483603,
+      "learning_rate": 1.3124770853925835e-05,
+      "loss": 1.3521,
+      "step": 1855
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4158243788013907,
+      "learning_rate": 1.3117853969953996e-05,
+      "loss": 0.7847,
+      "step": 1856
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3748728057797066,
+      "learning_rate": 1.3110935433282653e-05,
+      "loss": 0.7906,
+      "step": 1857
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.1681987758656436,
+      "learning_rate": 1.3104015247579156e-05,
+      "loss": 1.381,
+      "step": 1858
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5651012999780234,
+      "learning_rate": 1.3097093416511733e-05,
+      "loss": 0.862,
+      "step": 1859
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.413990895825751,
+      "learning_rate": 1.3090169943749475e-05,
+      "loss": 0.6998,
+      "step": 1860
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3863249823391177,
+      "learning_rate": 1.3083244832962353e-05,
+      "loss": 0.7886,
+      "step": 1861
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4535276997726319,
+      "learning_rate": 1.3076318087821202e-05,
+      "loss": 0.7327,
+      "step": 1862
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5006411477580266,
+      "learning_rate": 1.3069389711997717e-05,
+      "loss": 0.7845,
+      "step": 1863
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.1350171882906785,
+      "learning_rate": 1.306245970916447e-05,
+      "loss": 1.3297,
+      "step": 1864
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.030593147668406,
+      "learning_rate": 1.3055528082994889e-05,
+      "loss": 1.3932,
+      "step": 1865
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.070965563370951,
+      "learning_rate": 1.3048594837163258e-05,
+      "loss": 1.4398,
+      "step": 1866
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3997210209258446,
+      "learning_rate": 1.3041659975344727e-05,
+      "loss": 0.7076,
+      "step": 1867
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4215401861251746,
+      "learning_rate": 1.3034723501215301e-05,
+      "loss": 0.8061,
+      "step": 1868
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9672516742543416,
+      "learning_rate": 1.3027785418451836e-05,
+      "loss": 1.3846,
+      "step": 1869
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.394635613142373,
+      "learning_rate": 1.3020845730732043e-05,
+      "loss": 0.6903,
+      "step": 1870
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.530214241871772,
+      "learning_rate": 1.3013904441734486e-05,
+      "loss": 0.7883,
+      "step": 1871
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5042630784300657,
+      "learning_rate": 1.3006961555138576e-05,
+      "loss": 0.7925,
+      "step": 1872
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4101808232609314,
+      "learning_rate": 1.3000017074624564e-05,
+      "loss": 0.8154,
+      "step": 1873
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4902342881959194,
+      "learning_rate": 1.2993071003873561e-05,
+      "loss": 0.7943,
+      "step": 1874
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4546048570640953,
+      "learning_rate": 1.2986123346567505e-05,
+      "loss": 0.7964,
+      "step": 1875
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.465055139575877,
+      "learning_rate": 1.2979174106389185e-05,
+      "loss": 0.7024,
+      "step": 1876
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4566402060094308,
+      "learning_rate": 1.2972223287022229e-05,
+      "loss": 0.773,
+      "step": 1877
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5246593952309633,
+      "learning_rate": 1.2965270892151096e-05,
+      "loss": 0.7928,
+      "step": 1878
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.114938512048669,
+      "learning_rate": 1.2958316925461084e-05,
+      "loss": 1.3722,
+      "step": 1879
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4646833173379508,
+      "learning_rate": 1.2951361390638328e-05,
+      "loss": 0.7895,
+      "step": 1880
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3989458967896418,
+      "learning_rate": 1.2944404291369783e-05,
+      "loss": 0.7033,
+      "step": 1881
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5072080615206485,
+      "learning_rate": 1.2937445631343243e-05,
+      "loss": 0.7335,
+      "step": 1882
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3461968708385073,
+      "learning_rate": 1.2930485414247331e-05,
+      "loss": 0.7514,
+      "step": 1883
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5159493091039808,
+      "learning_rate": 1.2923523643771485e-05,
+      "loss": 0.7308,
+      "step": 1884
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4518414359300527,
+      "learning_rate": 1.2916560323605973e-05,
+      "loss": 0.7198,
+      "step": 1885
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4325960356402943,
+      "learning_rate": 1.2909595457441885e-05,
+      "loss": 0.7422,
+      "step": 1886
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.415738034448475,
+      "learning_rate": 1.2902629048971132e-05,
+      "loss": 0.7795,
+      "step": 1887
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5158849026449928,
+      "learning_rate": 1.2895661101886435e-05,
+      "loss": 0.7565,
+      "step": 1888
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.083310161529156,
+      "learning_rate": 1.2888691619881338e-05,
+      "loss": 1.3368,
+      "step": 1889
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.388770784761867,
+      "learning_rate": 1.2881720606650196e-05,
+      "loss": 0.8039,
+      "step": 1890
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9624097259915636,
+      "learning_rate": 1.2874748065888174e-05,
+      "loss": 1.3615,
+      "step": 1891
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4717446694219778,
+      "learning_rate": 1.286777400129125e-05,
+      "loss": 0.8919,
+      "step": 1892
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4109274492351458,
+      "learning_rate": 1.2860798416556211e-05,
+      "loss": 0.7119,
+      "step": 1893
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4867566356430966,
+      "learning_rate": 1.285382131538064e-05,
+      "loss": 0.7262,
+      "step": 1894
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9107602150320713,
+      "learning_rate": 1.2846842701462936e-05,
+      "loss": 1.367,
+      "step": 1895
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4991762725170157,
+      "learning_rate": 1.2839862578502291e-05,
+      "loss": 0.8179,
+      "step": 1896
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4719557794538087,
+      "learning_rate": 1.2832880950198705e-05,
+      "loss": 0.7408,
+      "step": 1897
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5375597436689104,
+      "learning_rate": 1.2825897820252968e-05,
+      "loss": 0.7611,
+      "step": 1898
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.345373357125353,
+      "learning_rate": 1.2818913192366668e-05,
+      "loss": 0.7072,
+      "step": 1899
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4606193874396272,
+      "learning_rate": 1.2811927070242198e-05,
+      "loss": 0.761,
+      "step": 1900
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.457040139869362,
+      "learning_rate": 1.2804939457582721e-05,
+      "loss": 0.6829,
+      "step": 1901
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.408237496055506,
+      "learning_rate": 1.2797950358092211e-05,
+      "loss": 0.7435,
+      "step": 1902
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.581843080067121,
+      "learning_rate": 1.2790959775475415e-05,
+      "loss": 0.8793,
+      "step": 1903
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4226185614714981,
+      "learning_rate": 1.2783967713437882e-05,
+      "loss": 0.7566,
+      "step": 1904
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3807008331836605,
+      "learning_rate": 1.2776974175685928e-05,
+      "loss": 0.7476,
+      "step": 1905
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4685075531770593,
+      "learning_rate": 1.2769979165926662e-05,
+      "loss": 0.6873,
+      "step": 1906
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.429418250248808,
+      "learning_rate": 1.2762982687867972e-05,
+      "loss": 0.782,
+      "step": 1907
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3177718241828058,
+      "learning_rate": 1.2755984745218522e-05,
+      "loss": 0.739,
+      "step": 1908
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4184848778112225,
+      "learning_rate": 1.2748985341687755e-05,
+      "loss": 0.72,
+      "step": 1909
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.447668638983851,
+      "learning_rate": 1.2741984480985883e-05,
+      "loss": 0.8269,
+      "step": 1910
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.0523778817436407,
+      "learning_rate": 1.2734982166823898e-05,
+      "loss": 1.4027,
+      "step": 1911
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.150228813501627,
+      "learning_rate": 1.2727978402913555e-05,
+      "loss": 1.3176,
+      "step": 1912
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4383753148640095,
+      "learning_rate": 1.2720973192967384e-05,
+      "loss": 0.7423,
+      "step": 1913
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4313893166116947,
+      "learning_rate": 1.2713966540698678e-05,
+      "loss": 0.8044,
+      "step": 1914
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4766273558229872,
+      "learning_rate": 1.2706958449821492e-05,
+      "loss": 0.7947,
+      "step": 1915
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4615685009652164,
+      "learning_rate": 1.2699948924050651e-05,
+      "loss": 0.8249,
+      "step": 1916
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5014295321203321,
+      "learning_rate": 1.2692937967101732e-05,
+      "loss": 0.6704,
+      "step": 1917
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.1252367632518316,
+      "learning_rate": 1.268592558269108e-05,
+      "loss": 1.4222,
+      "step": 1918
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3837502749507515,
+      "learning_rate": 1.2678911774535787e-05,
+      "loss": 0.7174,
+      "step": 1919
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4497933620312748,
+      "learning_rate": 1.2671896546353705e-05,
+      "loss": 0.8297,
+      "step": 1920
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4195225641675802,
+      "learning_rate": 1.2664879901863442e-05,
+      "loss": 0.7783,
+      "step": 1921
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4441265259215115,
+      "learning_rate": 1.2657861844784347e-05,
+      "loss": 0.7423,
+      "step": 1922
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.476084544372996,
+      "learning_rate": 1.2650842378836528e-05,
+      "loss": 0.7366,
+      "step": 1923
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3894362044723405,
+      "learning_rate": 1.2643821507740832e-05,
+      "loss": 0.814,
+      "step": 1924
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.949803632598434,
+      "learning_rate": 1.2636799235218857e-05,
+      "loss": 1.3231,
+      "step": 1925
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.481622663529953,
+      "learning_rate": 1.2629775564992938e-05,
+      "loss": 0.8019,
+      "step": 1926
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4971885278477586,
+      "learning_rate": 1.2622750500786158e-05,
+      "loss": 0.8019,
+      "step": 1927
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4247819453450787,
+      "learning_rate": 1.2615724046322329e-05,
+      "loss": 0.677,
+      "step": 1928
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4858170409404485,
+      "learning_rate": 1.260869620532601e-05,
+      "loss": 0.8193,
+      "step": 1929
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.046430689612487,
+      "learning_rate": 1.2601666981522491e-05,
+      "loss": 1.3326,
+      "step": 1930
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4295816143433004,
+      "learning_rate": 1.2594636378637795e-05,
+      "loss": 0.7148,
+      "step": 1931
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5099137753513712,
+      "learning_rate": 1.258760440039867e-05,
+      "loss": 0.8043,
+      "step": 1932
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3814142952913213,
+      "learning_rate": 1.2580571050532607e-05,
+      "loss": 0.8115,
+      "step": 1933
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3789800039097928,
+      "learning_rate": 1.257353633276781e-05,
+      "loss": 0.7662,
+      "step": 1934
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4611000006849955,
+      "learning_rate": 1.2566500250833216e-05,
+      "loss": 0.7442,
+      "step": 1935
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.421403615939953,
+      "learning_rate": 1.255946280845848e-05,
+      "loss": 0.665,
+      "step": 1936
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4130632930835105,
+      "learning_rate": 1.2552424009373988e-05,
+      "loss": 0.7576,
+      "step": 1937
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4297282135121192,
+      "learning_rate": 1.2545383857310828e-05,
+      "loss": 0.7486,
+      "step": 1938
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.340351390395927,
+      "learning_rate": 1.2538342356000827e-05,
+      "loss": 0.7455,
+      "step": 1939
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.111395790906617,
+      "learning_rate": 1.253129950917651e-05,
+      "loss": 1.366,
+      "step": 1940
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5520236736222242,
+      "learning_rate": 1.252425532057112e-05,
+      "loss": 0.8196,
+      "step": 1941
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4375096398925946,
+      "learning_rate": 1.2517209793918617e-05,
+      "loss": 0.7414,
+      "step": 1942
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.432918857352485,
+      "learning_rate": 1.2510162932953657e-05,
+      "loss": 0.7752,
+      "step": 1943
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.468615657862205,
+      "learning_rate": 1.2503114741411619e-05,
+      "loss": 0.7689,
+      "step": 1944
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4132652254499538,
+      "learning_rate": 1.2496065223028579e-05,
+      "loss": 0.8152,
+      "step": 1945
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.97557176177674,
+      "learning_rate": 1.2489014381541317e-05,
+      "loss": 1.3018,
+      "step": 1946
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4397926995911916,
+      "learning_rate": 1.2481962220687312e-05,
+      "loss": 0.7712,
+      "step": 1947
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.968178955141677,
+      "learning_rate": 1.2474908744204748e-05,
+      "loss": 1.3475,
+      "step": 1948
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4694372626476422,
+      "learning_rate": 1.2467853955832501e-05,
+      "loss": 0.7067,
+      "step": 1949
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4133266184227489,
+      "learning_rate": 1.2460797859310147e-05,
+      "loss": 0.7689,
+      "step": 1950
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5491460518452511,
+      "learning_rate": 1.2453740458377953e-05,
+      "loss": 0.7995,
+      "step": 1951
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.42081530782939,
+      "learning_rate": 1.244668175677688e-05,
+      "loss": 0.7617,
+      "step": 1952
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.467510228682809,
+      "learning_rate": 1.2439621758248568e-05,
+      "loss": 0.7998,
+      "step": 1953
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.07718993081851,
+      "learning_rate": 1.2432560466535362e-05,
+      "loss": 1.3084,
+      "step": 1954
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3673689877399873,
+      "learning_rate": 1.2425497885380277e-05,
+      "loss": 0.7383,
+      "step": 1955
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2964096029602095,
+      "learning_rate": 1.2418434018527024e-05,
+      "loss": 0.7212,
+      "step": 1956
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4389686873024035,
+      "learning_rate": 1.2411368869719981e-05,
+      "loss": 0.7118,
+      "step": 1957
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3557423208383095,
+      "learning_rate": 1.2404302442704219e-05,
+      "loss": 0.7268,
+      "step": 1958
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5246262686951033,
+      "learning_rate": 1.2397234741225481e-05,
+      "loss": 0.8308,
+      "step": 1959
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5099920214447846,
+      "learning_rate": 1.2390165769030183e-05,
+      "loss": 0.8072,
+      "step": 1960
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4757383371968875,
+      "learning_rate": 1.238309552986542e-05,
+      "loss": 0.7232,
+      "step": 1961
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.1614296419146273,
+      "learning_rate": 1.2376024027478959e-05,
+      "loss": 1.4045,
+      "step": 1962
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0904181331865868,
+      "learning_rate": 1.2368951265619224e-05,
+      "loss": 1.3586,
+      "step": 1963
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0102809074988865,
+      "learning_rate": 1.2361877248035326e-05,
+      "loss": 1.376,
+      "step": 1964
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3013314128394167,
+      "learning_rate": 1.2354801978477024e-05,
+      "loss": 0.6388,
+      "step": 1965
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4405952979134076,
+      "learning_rate": 1.2347725460694754e-05,
+      "loss": 0.8042,
+      "step": 1966
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.117171027164783,
+      "learning_rate": 1.234064769843961e-05,
+      "loss": 1.2696,
+      "step": 1967
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5384602538002332,
+      "learning_rate": 1.2333568695463336e-05,
+      "loss": 0.7237,
+      "step": 1968
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.3670840609274975,
+      "learning_rate": 1.2326488455518346e-05,
+      "loss": 1.3529,
+      "step": 1969
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.443346239313561,
+      "learning_rate": 1.2319406982357705e-05,
+      "loss": 0.7893,
+      "step": 1970
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.1384541421502674,
+      "learning_rate": 1.2312324279735131e-05,
+      "loss": 1.3043,
+      "step": 1971
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0291638842787276,
+      "learning_rate": 1.2305240351404994e-05,
+      "loss": 1.3675,
+      "step": 1972
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.360256980565725,
+      "learning_rate": 1.2298155201122319e-05,
+      "loss": 0.6835,
+      "step": 1973
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.066819587578023,
+      "learning_rate": 1.2291068832642768e-05,
+      "loss": 1.3579,
+      "step": 1974
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0919598529794907,
+      "learning_rate": 1.2283981249722656e-05,
+      "loss": 1.3947,
+      "step": 1975
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9618390032752002,
+      "learning_rate": 1.2276892456118944e-05,
+      "loss": 1.3362,
+      "step": 1976
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4082298206434827,
+      "learning_rate": 1.2269802455589228e-05,
+      "loss": 0.7622,
+      "step": 1977
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4305849835817894,
+      "learning_rate": 1.2262711251891746e-05,
+      "loss": 0.7058,
+      "step": 1978
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.421510383367434,
+      "learning_rate": 1.2255618848785378e-05,
+      "loss": 0.7072,
+      "step": 1979
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4766867967516075,
+      "learning_rate": 1.2248525250029635e-05,
+      "loss": 0.8475,
+      "step": 1980
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5659330037777603,
+      "learning_rate": 1.2241430459384663e-05,
+      "loss": 0.8546,
+      "step": 1981
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.520200410735863,
+      "learning_rate": 1.2234334480611238e-05,
+      "loss": 0.7662,
+      "step": 1982
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5558773512995339,
+      "learning_rate": 1.2227237317470774e-05,
+      "loss": 0.7499,
+      "step": 1983
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.3971223910414126,
+      "learning_rate": 1.2220138973725304e-05,
+      "loss": 1.4049,
+      "step": 1984
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5054690468396394,
+      "learning_rate": 1.2213039453137485e-05,
+      "loss": 0.7932,
+      "step": 1985
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5554371516246024,
+      "learning_rate": 1.2205938759470608e-05,
+      "loss": 0.7796,
+      "step": 1986
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4198281853774506,
+      "learning_rate": 1.2198836896488578e-05,
+      "loss": 0.6396,
+      "step": 1987
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5997314335873865,
+      "learning_rate": 1.2191733867955922e-05,
+      "loss": 0.8976,
+      "step": 1988
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3792346449193802,
+      "learning_rate": 1.2184629677637785e-05,
+      "loss": 0.7956,
+      "step": 1989
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3275870738794375,
+      "learning_rate": 1.2177524329299927e-05,
+      "loss": 0.62,
+      "step": 1990
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5239278812432269,
+      "learning_rate": 1.2170417826708723e-05,
+      "loss": 0.801,
+      "step": 1991
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.021760373425191,
+      "learning_rate": 1.216331017363116e-05,
+      "loss": 1.3256,
+      "step": 1992
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.522356885743368,
+      "learning_rate": 1.2156201373834835e-05,
+      "loss": 0.6852,
+      "step": 1993
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.414365606532051,
+      "learning_rate": 1.214909143108795e-05,
+      "loss": 0.7566,
+      "step": 1994
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.524940691475542,
+      "learning_rate": 1.2141980349159312e-05,
+      "loss": 0.8111,
+      "step": 1995
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4890596980296054,
+      "learning_rate": 1.213486813181834e-05,
+      "loss": 0.7215,
+      "step": 1996
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4965008979064822,
+      "learning_rate": 1.2127754782835048e-05,
+      "loss": 0.8128,
+      "step": 1997
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3770666071062776,
+      "learning_rate": 1.2120640305980056e-05,
+      "loss": 0.6948,
+      "step": 1998
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.1393390116133997,
+      "learning_rate": 1.2113524705024566e-05,
+      "loss": 1.3805,
+      "step": 1999
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2624002805664791,
+      "learning_rate": 1.2106407983740396e-05,
+      "loss": 0.6542,
+      "step": 2000
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.906103237570778,
+      "learning_rate": 1.2099290145899947e-05,
+      "loss": 1.3482,
+      "step": 2001
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8959139360680968,
+      "learning_rate": 1.2092171195276214e-05,
+      "loss": 1.3478,
+      "step": 2002
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4152426898447148,
+      "learning_rate": 1.208505113564278e-05,
+      "loss": 0.7898,
+      "step": 2003
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4605762590352118,
+      "learning_rate": 1.2077929970773823e-05,
+      "loss": 0.8687,
+      "step": 2004
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4002495021445003,
+      "learning_rate": 1.2070807704444092e-05,
+      "loss": 0.7241,
+      "step": 2005
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.2125920797938177,
+      "learning_rate": 1.2063684340428939e-05,
+      "loss": 1.4155,
+      "step": 2006
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9931040262454485,
+      "learning_rate": 1.205655988250428e-05,
+      "loss": 1.3475,
+      "step": 2007
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.0158855271336606,
+      "learning_rate": 1.2049434334446626e-05,
+      "loss": 1.4318,
+      "step": 2008
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4436714104563892,
+      "learning_rate": 1.2042307700033055e-05,
+      "loss": 0.6921,
+      "step": 2009
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9879870716961099,
+      "learning_rate": 1.2035179983041225e-05,
+      "loss": 1.3259,
+      "step": 2010
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4495978189103669,
+      "learning_rate": 1.2028051187249367e-05,
+      "loss": 0.8072,
+      "step": 2011
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.105115905655739,
+      "learning_rate": 1.2020921316436289e-05,
+      "loss": 1.3542,
+      "step": 2012
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5032089756049123,
+      "learning_rate": 1.2013790374381357e-05,
+      "loss": 0.7506,
+      "step": 2013
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5087545924079406,
+      "learning_rate": 1.200665836486452e-05,
+      "loss": 0.7617,
+      "step": 2014
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.554096436016711,
+      "learning_rate": 1.1999525291666279e-05,
+      "loss": 0.7083,
+      "step": 2015
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.0434197454853584,
+      "learning_rate": 1.1992391158567704e-05,
+      "loss": 1.3489,
+      "step": 2016
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.427627310007113,
+      "learning_rate": 1.1985255969350433e-05,
+      "loss": 0.7212,
+      "step": 2017
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5498303299213998,
+      "learning_rate": 1.1978119727796655e-05,
+      "loss": 0.8358,
+      "step": 2018
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3956448693763517,
+      "learning_rate": 1.1970982437689122e-05,
+      "loss": 0.7288,
+      "step": 2019
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5199624617075123,
+      "learning_rate": 1.1963844102811139e-05,
+      "loss": 0.7684,
+      "step": 2020
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.1164905362341644,
+      "learning_rate": 1.1956704726946563e-05,
+      "loss": 1.3536,
+      "step": 2021
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4658472038626218,
+      "learning_rate": 1.1949564313879812e-05,
+      "loss": 0.7171,
+      "step": 2022
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.41233032805996,
+      "learning_rate": 1.1942422867395842e-05,
+      "loss": 0.7945,
+      "step": 2023
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5101783713884367,
+      "learning_rate": 1.1935280391280166e-05,
+      "loss": 0.7404,
+      "step": 2024
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3937584406537538,
+      "learning_rate": 1.1928136889318835e-05,
+      "loss": 0.6954,
+      "step": 2025
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5683439167976139,
+      "learning_rate": 1.1920992365298452e-05,
+      "loss": 0.7788,
+      "step": 2026
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.449595327310981,
+      "learning_rate": 1.1913846823006155e-05,
+      "loss": 0.7345,
+      "step": 2027
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.0888840098413364,
+      "learning_rate": 1.1906700266229622e-05,
+      "loss": 1.234,
+      "step": 2028
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4916975016922993,
+      "learning_rate": 1.1899552698757078e-05,
+      "loss": 0.7781,
+      "step": 2029
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.027791522017118,
+      "learning_rate": 1.1892404124377272e-05,
+      "loss": 1.3742,
+      "step": 2030
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4820752144546867,
+      "learning_rate": 1.1885254546879493e-05,
+      "loss": 0.7507,
+      "step": 2031
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5116550049306534,
+      "learning_rate": 1.1878103970053558e-05,
+      "loss": 0.7677,
+      "step": 2032
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5626015518597205,
+      "learning_rate": 1.1870952397689818e-05,
+      "loss": 0.7976,
+      "step": 2033
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3936556009019305,
+      "learning_rate": 1.1863799833579152e-05,
+      "loss": 0.7309,
+      "step": 2034
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5623921612827112,
+      "learning_rate": 1.185664628151296e-05,
+      "loss": 0.8456,
+      "step": 2035
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4820495390270791,
+      "learning_rate": 1.1849491745283166e-05,
+      "loss": 0.8636,
+      "step": 2036
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4708155268684926,
+      "learning_rate": 1.1842336228682221e-05,
+      "loss": 0.7512,
+      "step": 2037
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5314745369821468,
+      "learning_rate": 1.183517973550309e-05,
+      "loss": 0.783,
+      "step": 2038
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.0960114711000712,
+      "learning_rate": 1.1828022269539261e-05,
+      "loss": 1.434,
+      "step": 2039
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5083987340651106,
+      "learning_rate": 1.1820863834584735e-05,
+      "loss": 0.7244,
+      "step": 2040
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9993647022201828,
+      "learning_rate": 1.1813704434434023e-05,
+      "loss": 1.4091,
+      "step": 2041
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.008332430874572,
+      "learning_rate": 1.1806544072882154e-05,
+      "loss": 1.3714,
+      "step": 2042
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4580503288104374,
+      "learning_rate": 1.1799382753724662e-05,
+      "loss": 0.7819,
+      "step": 2043
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.0051858107506284,
+      "learning_rate": 1.1792220480757588e-05,
+      "loss": 1.4031,
+      "step": 2044
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4136828906354362,
+      "learning_rate": 1.178505725777749e-05,
+      "loss": 0.6905,
+      "step": 2045
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9994239832945293,
+      "learning_rate": 1.1777893088581408e-05,
+      "loss": 1.3497,
+      "step": 2046
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4635043353647017,
+      "learning_rate": 1.1770727976966905e-05,
+      "loss": 0.7975,
+      "step": 2047
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9621616595141975,
+      "learning_rate": 1.1763561926732033e-05,
+      "loss": 1.4092,
+      "step": 2048
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5053755855264694,
+      "learning_rate": 1.175639494167534e-05,
+      "loss": 0.7609,
+      "step": 2049
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.0014118216976553,
+      "learning_rate": 1.1749227025595873e-05,
+      "loss": 1.3603,
+      "step": 2050
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4364859425913534,
+      "learning_rate": 1.1742058182293178e-05,
+      "loss": 0.6988,
+      "step": 2051
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4124139601376728,
+      "learning_rate": 1.173488841556728e-05,
+      "loss": 0.7116,
+      "step": 2052
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9328156094749778,
+      "learning_rate": 1.1727717729218702e-05,
+      "loss": 1.3762,
+      "step": 2053
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4891631328019463,
+      "learning_rate": 1.1720546127048454e-05,
+      "loss": 0.7339,
+      "step": 2054
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4848872844549852,
+      "learning_rate": 1.171337361285803e-05,
+      "loss": 0.7912,
+      "step": 2055
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.422163674161754,
+      "learning_rate": 1.1706200190449403e-05,
+      "loss": 0.7564,
+      "step": 2056
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9368737431657383,
+      "learning_rate": 1.169902586362504e-05,
+      "loss": 1.3191,
+      "step": 2057
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4161354982256575,
+      "learning_rate": 1.1691850636187872e-05,
+      "loss": 0.7815,
+      "step": 2058
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4879129921306362,
+      "learning_rate": 1.1684674511941317e-05,
+      "loss": 0.7847,
+      "step": 2059
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5100507039871425,
+      "learning_rate": 1.1677497494689267e-05,
+      "loss": 0.7575,
+      "step": 2060
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.0333199461365625,
+      "learning_rate": 1.167031958823609e-05,
+      "loss": 1.3549,
+      "step": 2061
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3957444932536998,
+      "learning_rate": 1.1663140796386614e-05,
+      "loss": 0.7404,
+      "step": 2062
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.059627130925717,
+      "learning_rate": 1.165596112294615e-05,
+      "loss": 1.3106,
+      "step": 2063
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3464787940213234,
+      "learning_rate": 1.1648780571720466e-05,
+      "loss": 0.6993,
+      "step": 2064
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2856007606633806,
+      "learning_rate": 1.1641599146515805e-05,
+      "loss": 0.7348,
+      "step": 2065
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.427304996775436,
+      "learning_rate": 1.1634416851138865e-05,
+      "loss": 0.6728,
+      "step": 2066
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3912615754953954,
+      "learning_rate": 1.1627233689396809e-05,
+      "loss": 0.7792,
+      "step": 2067
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3770336052061372,
+      "learning_rate": 1.1620049665097258e-05,
+      "loss": 0.7718,
+      "step": 2068
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4577505769332122,
+      "learning_rate": 1.1612864782048288e-05,
+      "loss": 0.7058,
+      "step": 2069
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6751418402227107,
+      "learning_rate": 1.1605679044058438e-05,
+      "loss": 0.7826,
+      "step": 2070
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.418874521272601,
+      "learning_rate": 1.1598492454936692e-05,
+      "loss": 0.7501,
+      "step": 2071
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.0165986270839715,
+      "learning_rate": 1.1591305018492492e-05,
+      "loss": 1.3165,
+      "step": 2072
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9535274145963764,
+      "learning_rate": 1.158411673853572e-05,
+      "loss": 1.3435,
+      "step": 2073
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.451290338461931,
+      "learning_rate": 1.1576927618876711e-05,
+      "loss": 0.7155,
+      "step": 2074
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.203037823695548,
+      "learning_rate": 1.1569737663326251e-05,
+      "loss": 1.345,
+      "step": 2075
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3981566396582548,
+      "learning_rate": 1.1562546875695559e-05,
+      "loss": 0.6861,
+      "step": 2076
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4635174380307636,
+      "learning_rate": 1.1555355259796296e-05,
+      "loss": 0.7783,
+      "step": 2077
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4155631435877842,
+      "learning_rate": 1.1548162819440568e-05,
+      "loss": 0.7858,
+      "step": 2078
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.47446676947492,
+      "learning_rate": 1.1540969558440911e-05,
+      "loss": 0.7456,
+      "step": 2079
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3894894872452108,
+      "learning_rate": 1.1533775480610305e-05,
+      "loss": 0.8019,
+      "step": 2080
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.0576261193315872,
+      "learning_rate": 1.1526580589762157e-05,
+      "loss": 1.3205,
+      "step": 2081
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.499530029976506,
+      "learning_rate": 1.1519384889710306e-05,
+      "loss": 0.7937,
+      "step": 2082
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3816934813372768,
+      "learning_rate": 1.1512188384269016e-05,
+      "loss": 0.7864,
+      "step": 2083
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.476073808324487,
+      "learning_rate": 1.1504991077252986e-05,
+      "loss": 0.7371,
+      "step": 2084
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4837607246647073,
+      "learning_rate": 1.149779297247733e-05,
+      "loss": 0.749,
+      "step": 2085
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5110230972156693,
+      "learning_rate": 1.1490594073757598e-05,
+      "loss": 0.7655,
+      "step": 2086
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.145605596804879,
+      "learning_rate": 1.1483394384909748e-05,
+      "loss": 1.3786,
+      "step": 2087
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3737358323562512,
+      "learning_rate": 1.1476193909750161e-05,
+      "loss": 0.7703,
+      "step": 2088
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.438134965430536,
+      "learning_rate": 1.1468992652095643e-05,
+      "loss": 0.7037,
+      "step": 2089
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2951113375971421,
+      "learning_rate": 1.14617906157634e-05,
+      "loss": 0.6654,
+      "step": 2090
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4426802352488828,
+      "learning_rate": 1.1454587804571062e-05,
+      "loss": 0.7739,
+      "step": 2091
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.369589859969843,
+      "learning_rate": 1.1447384222336668e-05,
+      "loss": 0.6726,
+      "step": 2092
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.358235455563039,
+      "learning_rate": 1.1440179872878662e-05,
+      "loss": 0.7328,
+      "step": 2093
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4872462102478379,
+      "learning_rate": 1.1432974760015895e-05,
+      "loss": 0.7672,
+      "step": 2094
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.51314621730212,
+      "learning_rate": 1.1425768887567625e-05,
+      "loss": 0.8032,
+      "step": 2095
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.499193034353755,
+      "learning_rate": 1.1418562259353513e-05,
+      "loss": 0.6839,
+      "step": 2096
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4084225790716696,
+      "learning_rate": 1.141135487919362e-05,
+      "loss": 0.7497,
+      "step": 2097
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4603351787161187,
+      "learning_rate": 1.1404146750908407e-05,
+      "loss": 0.7265,
+      "step": 2098
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.966253602108733,
+      "learning_rate": 1.1396937878318723e-05,
+      "loss": 1.3235,
+      "step": 2099
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4233761291964273,
+      "learning_rate": 1.1389728265245823e-05,
+      "loss": 0.689,
+      "step": 2100
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3473965751376569,
+      "learning_rate": 1.138251791551135e-05,
+      "loss": 0.7828,
+      "step": 2101
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3534764929254477,
+      "learning_rate": 1.1375306832937334e-05,
+      "loss": 0.8086,
+      "step": 2102
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4840089932473914,
+      "learning_rate": 1.1368095021346201e-05,
+      "loss": 0.7595,
+      "step": 2103
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5125812604630873,
+      "learning_rate": 1.1360882484560756e-05,
+      "loss": 0.8718,
+      "step": 2104
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4114798543171811,
+      "learning_rate": 1.1353669226404192e-05,
+      "loss": 0.7608,
+      "step": 2105
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3255265773823972,
+      "learning_rate": 1.1346455250700084e-05,
+      "loss": 0.7087,
+      "step": 2106
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.1536627820151732,
+      "learning_rate": 1.1339240561272389e-05,
+      "loss": 1.4032,
+      "step": 2107
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5701851088551426,
+      "learning_rate": 1.133202516194544e-05,
+      "loss": 0.8028,
+      "step": 2108
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9814822370256027,
+      "learning_rate": 1.1324809056543942e-05,
+      "loss": 1.3066,
+      "step": 2109
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.558204600861426,
+      "learning_rate": 1.1317592248892986e-05,
+      "loss": 0.8558,
+      "step": 2110
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.452898307261567,
+      "learning_rate": 1.1310374742818025e-05,
+      "loss": 0.7479,
+      "step": 2111
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5454950133828236,
+      "learning_rate": 1.1303156542144885e-05,
+      "loss": 0.7694,
+      "step": 2112
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5272328071082366,
+      "learning_rate": 1.1295937650699762e-05,
+      "loss": 0.7661,
+      "step": 2113
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.476816675253822,
+      "learning_rate": 1.1288718072309219e-05,
+      "loss": 0.7738,
+      "step": 2114
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4519609057680498,
+      "learning_rate": 1.128149781080018e-05,
+      "loss": 0.6825,
+      "step": 2115
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4718527291311372,
+      "learning_rate": 1.1274276869999928e-05,
+      "loss": 0.7571,
+      "step": 2116
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4654394595057303,
+      "learning_rate": 1.1267055253736117e-05,
+      "loss": 0.8552,
+      "step": 2117
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4755646991203948,
+      "learning_rate": 1.1259832965836746e-05,
+      "loss": 0.8333,
+      "step": 2118
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.434753430263163,
+      "learning_rate": 1.125261001013018e-05,
+      "loss": 0.8181,
+      "step": 2119
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4369652433327729,
+      "learning_rate": 1.1245386390445132e-05,
+      "loss": 0.7488,
+      "step": 2120
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.562392024559366,
+      "learning_rate": 1.1238162110610673e-05,
+      "loss": 0.766,
+      "step": 2121
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4540107850540507,
+      "learning_rate": 1.1230937174456216e-05,
+      "loss": 0.7838,
+      "step": 2122
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3387621038862332,
+      "learning_rate": 1.122371158581153e-05,
+      "loss": 0.6713,
+      "step": 2123
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9970920676205997,
+      "learning_rate": 1.1216485348506725e-05,
+      "loss": 1.4061,
+      "step": 2124
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9858649016490086,
+      "learning_rate": 1.1209258466372254e-05,
+      "loss": 1.345,
+      "step": 2125
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.372385126189226,
+      "learning_rate": 1.1202030943238915e-05,
+      "loss": 0.741,
+      "step": 2126
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3887888277120253,
+      "learning_rate": 1.1194802782937841e-05,
+      "loss": 0.7016,
+      "step": 2127
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4300194419188617,
+      "learning_rate": 1.1187573989300515e-05,
+      "loss": 0.7021,
+      "step": 2128
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4364172495126608,
+      "learning_rate": 1.1180344566158739e-05,
+      "loss": 0.7518,
+      "step": 2129
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.0419359238739325,
+      "learning_rate": 1.1173114517344656e-05,
+      "loss": 1.3566,
+      "step": 2130
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9457469018799496,
+      "learning_rate": 1.1165883846690744e-05,
+      "loss": 1.3671,
+      "step": 2131
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4653785055795983,
+      "learning_rate": 1.115865255802981e-05,
+      "loss": 0.7291,
+      "step": 2132
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4415859582958872,
+      "learning_rate": 1.1151420655194982e-05,
+      "loss": 0.7363,
+      "step": 2133
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3857889385919004,
+      "learning_rate": 1.114418814201972e-05,
+      "loss": 0.7059,
+      "step": 2134
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.481722314188844,
+      "learning_rate": 1.113695502233781e-05,
+      "loss": 0.7404,
+      "step": 2135
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4483069313481753,
+      "learning_rate": 1.1129721299983349e-05,
+      "loss": 0.7606,
+      "step": 2136
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.1013417420930427,
+      "learning_rate": 1.112248697879076e-05,
+      "loss": 1.3715,
+      "step": 2137
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3551690429245657,
+      "learning_rate": 1.1115252062594785e-05,
+      "loss": 0.7585,
+      "step": 2138
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.297740274955257,
+      "learning_rate": 1.1108016555230476e-05,
+      "loss": 0.6486,
+      "step": 2139
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.483559929467288,
+      "learning_rate": 1.1100780460533207e-05,
+      "loss": 0.7782,
+      "step": 2140
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4227843658215187,
+      "learning_rate": 1.1093543782338653e-05,
+      "loss": 0.7935,
+      "step": 2141
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.434481749995292,
+      "learning_rate": 1.1086306524482807e-05,
+      "loss": 0.796,
+      "step": 2142
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4094448571192457,
+      "learning_rate": 1.1079068690801962e-05,
+      "loss": 0.755,
+      "step": 2143
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4426847552172095,
+      "learning_rate": 1.1071830285132722e-05,
+      "loss": 0.7332,
+      "step": 2144
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.081235297587269,
+      "learning_rate": 1.1064591311311994e-05,
+      "loss": 1.3944,
+      "step": 2145
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3694442732545762,
+      "learning_rate": 1.1057351773176981e-05,
+      "loss": 0.816,
+      "step": 2146
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.456492767174901,
+      "learning_rate": 1.1050111674565187e-05,
+      "loss": 0.7191,
+      "step": 2147
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.256382259810063,
+      "learning_rate": 1.104287101931442e-05,
+      "loss": 1.3012,
+      "step": 2148
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4377770392398903,
+      "learning_rate": 1.1035629811262772e-05,
+      "loss": 0.7564,
+      "step": 2149
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4878879247158945,
+      "learning_rate": 1.1028388054248633e-05,
+      "loss": 0.7373,
+      "step": 2150
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4672731926745803,
+      "learning_rate": 1.102114575211069e-05,
+      "loss": 0.8121,
+      "step": 2151
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2229030274896795,
+      "learning_rate": 1.1013902908687908e-05,
+      "loss": 0.7398,
+      "step": 2152
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4260270746403199,
+      "learning_rate": 1.1006659527819547e-05,
+      "loss": 0.7917,
+      "step": 2153
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4088178831245919,
+      "learning_rate": 1.099941561334515e-05,
+      "loss": 0.7604,
+      "step": 2154
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.076863387319099,
+      "learning_rate": 1.0992171169104541e-05,
+      "loss": 1.383,
+      "step": 2155
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3486150928483034,
+      "learning_rate": 1.0984926198937826e-05,
+      "loss": 0.7255,
+      "step": 2156
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5656709007929908,
+      "learning_rate": 1.097768070668539e-05,
+      "loss": 0.7535,
+      "step": 2157
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5622400709132314,
+      "learning_rate": 1.0970434696187893e-05,
+      "loss": 0.8684,
+      "step": 2158
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4131575638522706,
+      "learning_rate": 1.0963188171286275e-05,
+      "loss": 0.7429,
+      "step": 2159
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5307848106177664,
+      "learning_rate": 1.0955941135821737e-05,
+      "loss": 0.6986,
+      "step": 2160
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.463188144174011,
+      "learning_rate": 1.0948693593635763e-05,
+      "loss": 0.7376,
+      "step": 2161
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5544282158495715,
+      "learning_rate": 1.0941445548570104e-05,
+      "loss": 0.7431,
+      "step": 2162
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.094921162346549,
+      "learning_rate": 1.0934197004466772e-05,
+      "loss": 1.412,
+      "step": 2163
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.520071694406999,
+      "learning_rate": 1.0926947965168045e-05,
+      "loss": 0.7094,
+      "step": 2164
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3239075137932272,
+      "learning_rate": 1.0919698434516463e-05,
+      "loss": 0.6786,
+      "step": 2165
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4768392318145738,
+      "learning_rate": 1.0912448416354838e-05,
+      "loss": 0.7621,
+      "step": 2166
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.484124619979767,
+      "learning_rate": 1.0905197914526218e-05,
+      "loss": 0.8181,
+      "step": 2167
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3436431210913562,
+      "learning_rate": 1.0897946932873924e-05,
+      "loss": 0.7596,
+      "step": 2168
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9890925724857043,
+      "learning_rate": 1.0890695475241534e-05,
+      "loss": 1.3287,
+      "step": 2169
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.429633803446332,
+      "learning_rate": 1.0883443545472861e-05,
+      "loss": 0.7111,
+      "step": 2170
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9791515575393221,
+      "learning_rate": 1.0876191147411983e-05,
+      "loss": 1.369,
+      "step": 2171
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4220837392197956,
+      "learning_rate": 1.0868938284903224e-05,
+      "loss": 0.6945,
+      "step": 2172
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.579885901033223,
+      "learning_rate": 1.086168496179115e-05,
+      "loss": 0.8061,
+      "step": 2173
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.469023308333002,
+      "learning_rate": 1.0854431181920575e-05,
+      "loss": 0.7863,
+      "step": 2174
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4432203038680538,
+      "learning_rate": 1.0847176949136553e-05,
+      "loss": 0.7798,
+      "step": 2175
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.0915630384437796,
+      "learning_rate": 1.0839922267284377e-05,
+      "loss": 1.2843,
+      "step": 2176
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4963179483244131,
+      "learning_rate": 1.0832667140209586e-05,
+      "loss": 0.745,
+      "step": 2177
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.375254027649269,
+      "learning_rate": 1.0825411571757937e-05,
+      "loss": 0.6998,
+      "step": 2178
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3520163600399373,
+      "learning_rate": 1.0818155565775443e-05,
+      "loss": 0.7576,
+      "step": 2179
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5571775806471702,
+      "learning_rate": 1.0810899126108334e-05,
+      "loss": 0.7662,
+      "step": 2180
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4491207026559232,
+      "learning_rate": 1.0803642256603072e-05,
+      "loss": 0.7654,
+      "step": 2181
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.472303127081566,
+      "learning_rate": 1.0796384961106354e-05,
+      "loss": 0.7594,
+      "step": 2182
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4127511380012645,
+      "learning_rate": 1.0789127243465096e-05,
+      "loss": 0.776,
+      "step": 2183
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5468700802614292,
+      "learning_rate": 1.0781869107526441e-05,
+      "loss": 0.7555,
+      "step": 2184
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5118108680252698,
+      "learning_rate": 1.077461055713775e-05,
+      "loss": 0.7856,
+      "step": 2185
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.179400774468116,
+      "learning_rate": 1.0767351596146607e-05,
+      "loss": 1.3907,
+      "step": 2186
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.530850538964703,
+      "learning_rate": 1.0760092228400818e-05,
+      "loss": 0.7539,
+      "step": 2187
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4890274200102822,
+      "learning_rate": 1.075283245774839e-05,
+      "loss": 0.8412,
+      "step": 2188
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3615842602868398,
+      "learning_rate": 1.074557228803756e-05,
+      "loss": 0.7107,
+      "step": 2189
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.427447833698863,
+      "learning_rate": 1.073831172311677e-05,
+      "loss": 0.8045,
+      "step": 2190
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4833786847262793,
+      "learning_rate": 1.0731050766834668e-05,
+      "loss": 0.7427,
+      "step": 2191
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5082928646552092,
+      "learning_rate": 1.072378942304011e-05,
+      "loss": 0.7353,
+      "step": 2192
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5211594074782673,
+      "learning_rate": 1.0716527695582164e-05,
+      "loss": 0.7511,
+      "step": 2193
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5433642472668816,
+      "learning_rate": 1.0709265588310098e-05,
+      "loss": 0.8791,
+      "step": 2194
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.510424963252908,
+      "learning_rate": 1.070200310507338e-05,
+      "loss": 0.8024,
+      "step": 2195
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5887969187257474,
+      "learning_rate": 1.0694740249721673e-05,
+      "loss": 0.8054,
+      "step": 2196
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3988380874416508,
+      "learning_rate": 1.0687477026104852e-05,
+      "loss": 0.6998,
+      "step": 2197
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4654142471080607,
+      "learning_rate": 1.0680213438072969e-05,
+      "loss": 0.8137,
+      "step": 2198
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4335724281362774,
+      "learning_rate": 1.0672949489476281e-05,
+      "loss": 0.7669,
+      "step": 2199
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.1545252327267725,
+      "learning_rate": 1.0665685184165231e-05,
+      "loss": 1.3868,
+      "step": 2200
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.058841526372003,
+      "learning_rate": 1.0658420525990454e-05,
+      "loss": 1.3175,
+      "step": 2201
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4168570934894622,
+      "learning_rate": 1.065115551880277e-05,
+      "loss": 0.7241,
+      "step": 2202
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4415079497478942,
+      "learning_rate": 1.0643890166453186e-05,
+      "loss": 0.6916,
+      "step": 2203
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.55178767364398,
+      "learning_rate": 1.063662447279289e-05,
+      "loss": 0.7612,
+      "step": 2204
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.580533289700759,
+      "learning_rate": 1.062935844167325e-05,
+      "loss": 0.7629,
+      "step": 2205
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.0662035520103834,
+      "learning_rate": 1.0622092076945817e-05,
+      "loss": 1.3644,
+      "step": 2206
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.1061518103520798,
+      "learning_rate": 1.0614825382462317e-05,
+      "loss": 1.2846,
+      "step": 2207
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4517017556306602,
+      "learning_rate": 1.060755836207465e-05,
+      "loss": 0.8712,
+      "step": 2208
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.040638739757859,
+      "learning_rate": 1.0600291019634887e-05,
+      "loss": 1.3504,
+      "step": 2209
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9316801092743112,
+      "learning_rate": 1.0593023358995273e-05,
+      "loss": 1.2655,
+      "step": 2210
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.492051094297651,
+      "learning_rate": 1.058575538400822e-05,
+      "loss": 0.7065,
+      "step": 2211
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.383688176109523,
+      "learning_rate": 1.0578487098526305e-05,
+      "loss": 0.8092,
+      "step": 2212
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4155474235006313,
+      "learning_rate": 1.0571218506402275e-05,
+      "loss": 0.7721,
+      "step": 2213
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4060017575513668,
+      "learning_rate": 1.0563949611489032e-05,
+      "loss": 0.7071,
+      "step": 2214
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.527785954692571,
+      "learning_rate": 1.055668041763965e-05,
+      "loss": 0.8134,
+      "step": 2215
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4656460135157947,
+      "learning_rate": 1.0549410928707344e-05,
+      "loss": 0.6938,
+      "step": 2216
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4757766000350612,
+      "learning_rate": 1.0542141148545504e-05,
+      "loss": 0.8621,
+      "step": 2217
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.0485586506950115,
+      "learning_rate": 1.0534871081007663e-05,
+      "loss": 1.4011,
+      "step": 2218
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3351411406381828,
+      "learning_rate": 1.0527600729947509e-05,
+      "loss": 0.7707,
+      "step": 2219
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5064148421187453,
+      "learning_rate": 1.0520330099218878e-05,
+      "loss": 0.7428,
+      "step": 2220
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4336885263569843,
+      "learning_rate": 1.0513059192675762e-05,
+      "loss": 0.691,
+      "step": 2221
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4973568304367764,
+      "learning_rate": 1.0505788014172288e-05,
+      "loss": 0.7783,
+      "step": 2222
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9529529595993549,
+      "learning_rate": 1.0498516567562737e-05,
+      "loss": 1.331,
+      "step": 2223
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.463988074069333,
+      "learning_rate": 1.0491244856701528e-05,
+      "loss": 0.7624,
+      "step": 2224
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3913149388988788,
+      "learning_rate": 1.0483972885443218e-05,
+      "loss": 0.7085,
+      "step": 2225
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.050543374318307,
+      "learning_rate": 1.0476700657642507e-05,
+      "loss": 1.3065,
+      "step": 2226
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.0826283926934446,
+      "learning_rate": 1.0469428177154225e-05,
+      "loss": 1.4671,
+      "step": 2227
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3974313732287507,
+      "learning_rate": 1.0462155447833344e-05,
+      "loss": 0.7319,
+      "step": 2228
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.402449904348321,
+      "learning_rate": 1.0454882473534961e-05,
+      "loss": 0.6983,
+      "step": 2229
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.0228245880391023,
+      "learning_rate": 1.0447609258114301e-05,
+      "loss": 1.3876,
+      "step": 2230
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3626379609627852,
+      "learning_rate": 1.044033580542673e-05,
+      "loss": 0.7787,
+      "step": 2231
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.058120479263412,
+      "learning_rate": 1.043306211932772e-05,
+      "loss": 1.4138,
+      "step": 2232
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.428756750355869,
+      "learning_rate": 1.0425788203672882e-05,
+      "loss": 0.7513,
+      "step": 2233
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.0207993746733144,
+      "learning_rate": 1.0418514062317944e-05,
+      "loss": 1.3177,
+      "step": 2234
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2937300648887216,
+      "learning_rate": 1.0411239699118756e-05,
+      "loss": 0.7975,
+      "step": 2235
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9975234304260083,
+      "learning_rate": 1.0403965117931279e-05,
+      "loss": 1.3417,
+      "step": 2236
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3987945665477217,
+      "learning_rate": 1.0396690322611596e-05,
+      "loss": 0.7371,
+      "step": 2237
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.342468769197275,
+      "learning_rate": 1.03894153170159e-05,
+      "loss": 0.7749,
+      "step": 2238
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.433535657957225,
+      "learning_rate": 1.03821401050005e-05,
+      "loss": 0.7744,
+      "step": 2239
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8624140849444197,
+      "learning_rate": 1.0374864690421806e-05,
+      "loss": 1.3273,
+      "step": 2240
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.517823497311324,
+      "learning_rate": 1.0367589077136348e-05,
+      "loss": 0.7077,
+      "step": 2241
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.0714786030454273,
+      "learning_rate": 1.0360313269000746e-05,
+      "loss": 1.3657,
+      "step": 2242
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4552723778352339,
+      "learning_rate": 1.0353037269871733e-05,
+      "loss": 0.7454,
+      "step": 2243
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.951942583939817,
+      "learning_rate": 1.0345761083606145e-05,
+      "loss": 1.3187,
+      "step": 2244
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.062188518216045,
+      "learning_rate": 1.033848471406091e-05,
+      "loss": 1.3643,
+      "step": 2245
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5270898788812104,
+      "learning_rate": 1.0331208165093056e-05,
+      "loss": 0.778,
+      "step": 2246
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5155452189216998,
+      "learning_rate": 1.0323931440559711e-05,
+      "loss": 0.8537,
+      "step": 2247
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3293645791863202,
+      "learning_rate": 1.031665454431809e-05,
+      "loss": 0.6941,
+      "step": 2248
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.495470217133118,
+      "learning_rate": 1.03093774802255e-05,
+      "loss": 0.7417,
+      "step": 2249
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4400094965729004,
+      "learning_rate": 1.0302100252139343e-05,
+      "loss": 0.7449,
+      "step": 2250
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4446139235969246,
+      "learning_rate": 1.0294822863917094e-05,
+      "loss": 0.7516,
+      "step": 2251
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4170129499960764,
+      "learning_rate": 1.0287545319416328e-05,
+      "loss": 0.7142,
+      "step": 2252
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3565843672482254,
+      "learning_rate": 1.0280267622494691e-05,
+      "loss": 0.7397,
+      "step": 2253
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3924684798778861,
+      "learning_rate": 1.027298977700992e-05,
+      "loss": 0.771,
+      "step": 2254
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3788084061150598,
+      "learning_rate": 1.0265711786819827e-05,
+      "loss": 0.6932,
+      "step": 2255
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.407509725054194,
+      "learning_rate": 1.0258433655782296e-05,
+      "loss": 0.7075,
+      "step": 2256
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4028914737436067,
+      "learning_rate": 1.0251155387755291e-05,
+      "loss": 0.7841,
+      "step": 2257
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4313691348448254,
+      "learning_rate": 1.024387698659685e-05,
+      "loss": 0.76,
+      "step": 2258
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.535426016550671,
+      "learning_rate": 1.0236598456165075e-05,
+      "loss": 0.8195,
+      "step": 2259
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4867800197937737,
+      "learning_rate": 1.022931980031814e-05,
+      "loss": 0.8179,
+      "step": 2260
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.4806431568925253,
+      "learning_rate": 1.0222041022914288e-05,
+      "loss": 1.3416,
+      "step": 2261
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.422879267056148,
+      "learning_rate": 1.0214762127811824e-05,
+      "loss": 0.6784,
+      "step": 2262
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4008463677852372,
+      "learning_rate": 1.0207483118869113e-05,
+      "loss": 0.6997,
+      "step": 2263
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4715297769481994,
+      "learning_rate": 1.0200203999944582e-05,
+      "loss": 0.8253,
+      "step": 2264
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.072959629104732,
+      "learning_rate": 1.0192924774896722e-05,
+      "loss": 1.4097,
+      "step": 2265
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4625624020571213,
+      "learning_rate": 1.0185645447584072e-05,
+      "loss": 0.7644,
+      "step": 2266
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6101835208915156,
+      "learning_rate": 1.0178366021865226e-05,
+      "loss": 0.8347,
+      "step": 2267
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.969485334545715,
+      "learning_rate": 1.0171086501598839e-05,
+      "loss": 1.3495,
+      "step": 2268
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9688650876370304,
+      "learning_rate": 1.0163806890643606e-05,
+      "loss": 1.3387,
+      "step": 2269
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3617130519699259,
+      "learning_rate": 1.015652719285828e-05,
+      "loss": 0.6615,
+      "step": 2270
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4942708435918444,
+      "learning_rate": 1.0149247412101642e-05,
+      "loss": 0.7125,
+      "step": 2271
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4347604455757816,
+      "learning_rate": 1.0141967552232539e-05,
+      "loss": 0.6704,
+      "step": 2272
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4341737363311977,
+      "learning_rate": 1.0134687617109844e-05,
+      "loss": 0.7704,
+      "step": 2273
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9624686573510426,
+      "learning_rate": 1.0127407610592478e-05,
+      "loss": 1.3709,
+      "step": 2274
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6197588678727959,
+      "learning_rate": 1.0120127536539399e-05,
+      "loss": 0.8009,
+      "step": 2275
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.008891553505705,
+      "learning_rate": 1.0112847398809598e-05,
+      "loss": 1.3197,
+      "step": 2276
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4229683401171886,
+      "learning_rate": 1.0105567201262101e-05,
+      "loss": 0.7333,
+      "step": 2277
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3659446763970924,
+      "learning_rate": 1.0098286947755964e-05,
+      "loss": 0.7541,
+      "step": 2278
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.85998790602763,
+      "learning_rate": 1.009100664215028e-05,
+      "loss": 1.3105,
+      "step": 2279
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4653855318670965,
+      "learning_rate": 1.0083726288304158e-05,
+      "loss": 0.7696,
+      "step": 2280
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4502533739961547,
+      "learning_rate": 1.0076445890076746e-05,
+      "loss": 0.6699,
+      "step": 2281
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.181323837078593,
+      "learning_rate": 1.00691654513272e-05,
+      "loss": 1.3736,
+      "step": 2282
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4088848355476455,
+      "learning_rate": 1.0061884975914713e-05,
+      "loss": 0.7826,
+      "step": 2283
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.455457729692083,
+      "learning_rate": 1.0054604467698486e-05,
+      "loss": 0.7192,
+      "step": 2284
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9398881715034917,
+      "learning_rate": 1.004732393053774e-05,
+      "loss": 1.3859,
+      "step": 2285
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5171590135777657,
+      "learning_rate": 1.0040043368291714e-05,
+      "loss": 0.7507,
+      "step": 2286
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4210857733207531,
+      "learning_rate": 1.0032762784819662e-05,
+      "loss": 0.7268,
+      "step": 2287
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.483779836225133,
+      "learning_rate": 1.0025482183980847e-05,
+      "loss": 0.7569,
+      "step": 2288
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.482874403510123,
+      "learning_rate": 1.0018201569634534e-05,
+      "loss": 0.8381,
+      "step": 2289
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.419452680911026,
+      "learning_rate": 1.0010920945640007e-05,
+      "loss": 0.7778,
+      "step": 2290
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4979494273216019,
+      "learning_rate": 1.0003640315856553e-05,
+      "loss": 0.7838,
+      "step": 2291
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4446851939489618,
+      "learning_rate": 9.99635968414345e-06,
+      "loss": 0.787,
+      "step": 2292
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3770475076070823,
+      "learning_rate": 9.989079054359994e-06,
+      "loss": 0.7783,
+      "step": 2293
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3889395173772796,
+      "learning_rate": 9.98179843036547e-06,
+      "loss": 0.7374,
+      "step": 2294
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.441857521848306,
+      "learning_rate": 9.974517816019155e-06,
+      "loss": 0.7754,
+      "step": 2295
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.461996109662239,
+      "learning_rate": 9.967237215180341e-06,
+      "loss": 0.7447,
+      "step": 2296
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3619093398647437,
+      "learning_rate": 9.959956631708291e-06,
+      "loss": 0.7498,
+      "step": 2297
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4327845451242367,
+      "learning_rate": 9.952676069462263e-06,
+      "loss": 0.67,
+      "step": 2298
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4918009342861487,
+      "learning_rate": 9.94539553230152e-06,
+      "loss": 0.7683,
+      "step": 2299
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.0362427950911357,
+      "learning_rate": 9.938115024085289e-06,
+      "loss": 1.3667,
+      "step": 2300
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4949658643565507,
+      "learning_rate": 9.930834548672801e-06,
+      "loss": 0.7142,
+      "step": 2301
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5068981991467727,
+      "learning_rate": 9.923554109923257e-06,
+      "loss": 0.7633,
+      "step": 2302
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.404174661943443,
+      "learning_rate": 9.916273711695844e-06,
+      "loss": 0.755,
+      "step": 2303
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.03174110864918,
+      "learning_rate": 9.908993357849721e-06,
+      "loss": 1.3466,
+      "step": 2304
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3393801234679485,
+      "learning_rate": 9.901713052244038e-06,
+      "loss": 0.7864,
+      "step": 2305
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5270003622524044,
+      "learning_rate": 9.8944327987379e-06,
+      "loss": 0.7569,
+      "step": 2306
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.377671763138402,
+      "learning_rate": 9.887152601190406e-06,
+      "loss": 0.7183,
+      "step": 2307
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.111666046925258,
+      "learning_rate": 9.879872463460606e-06,
+      "loss": 1.4072,
+      "step": 2308
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.479172080208314,
+      "learning_rate": 9.872592389407525e-06,
+      "loss": 0.8456,
+      "step": 2309
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3832481544090194,
+      "learning_rate": 9.865312382890161e-06,
+      "loss": 0.7205,
+      "step": 2310
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3512644392017237,
+      "learning_rate": 9.858032447767464e-06,
+      "loss": 0.771,
+      "step": 2311
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4487361772957184,
+      "learning_rate": 9.850752587898363e-06,
+      "loss": 0.7727,
+      "step": 2312
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4259880291695939,
+      "learning_rate": 9.843472807141726e-06,
+      "loss": 0.9063,
+      "step": 2313
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.488761173085831,
+      "learning_rate": 9.836193109356397e-06,
+      "loss": 0.8009,
+      "step": 2314
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9638344233668117,
+      "learning_rate": 9.828913498401161e-06,
+      "loss": 1.2972,
+      "step": 2315
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.0170244570055567,
+      "learning_rate": 9.821633978134775e-06,
+      "loss": 1.3655,
+      "step": 2316
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5127413479591416,
+      "learning_rate": 9.81435455241593e-06,
+      "loss": 0.764,
+      "step": 2317
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5481940302516946,
+      "learning_rate": 9.807075225103281e-06,
+      "loss": 0.7601,
+      "step": 2318
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.432443748672784,
+      "learning_rate": 9.799796000055423e-06,
+      "loss": 0.7158,
+      "step": 2319
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.0126661635669305,
+      "learning_rate": 9.79251688113089e-06,
+      "loss": 1.3791,
+      "step": 2320
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3263806245145229,
+      "learning_rate": 9.785237872188181e-06,
+      "loss": 0.6711,
+      "step": 2321
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.484209735042234,
+      "learning_rate": 9.777958977085715e-06,
+      "loss": 0.7153,
+      "step": 2322
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4558431014620683,
+      "learning_rate": 9.770680199681863e-06,
+      "loss": 0.8552,
+      "step": 2323
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.396504762648972,
+      "learning_rate": 9.763401543834927e-06,
+      "loss": 0.7158,
+      "step": 2324
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4161067099496232,
+      "learning_rate": 9.756123013403154e-06,
+      "loss": 0.741,
+      "step": 2325
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9683986893217762,
+      "learning_rate": 9.748844612244709e-06,
+      "loss": 1.3259,
+      "step": 2326
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.1320668444050135,
+      "learning_rate": 9.741566344217707e-06,
+      "loss": 1.3345,
+      "step": 2327
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4464727029917517,
+      "learning_rate": 9.734288213180175e-06,
+      "loss": 0.6259,
+      "step": 2328
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3693042058741631,
+      "learning_rate": 9.727010222990083e-06,
+      "loss": 0.749,
+      "step": 2329
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5465881970765665,
+      "learning_rate": 9.719732377505314e-06,
+      "loss": 0.7281,
+      "step": 2330
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5494698973376029,
+      "learning_rate": 9.712454680583677e-06,
+      "loss": 0.8115,
+      "step": 2331
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.1144846513735964,
+      "learning_rate": 9.705177136082911e-06,
+      "loss": 1.3976,
+      "step": 2332
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.1088332309107174,
+      "learning_rate": 9.697899747860663e-06,
+      "loss": 1.3091,
+      "step": 2333
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4897787694987015,
+      "learning_rate": 9.690622519774502e-06,
+      "loss": 0.7909,
+      "step": 2334
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4174860688639979,
+      "learning_rate": 9.68334545568191e-06,
+      "loss": 0.7157,
+      "step": 2335
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.456424811559357,
+      "learning_rate": 9.67606855944029e-06,
+      "loss": 0.6401,
+      "step": 2336
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4530854037808498,
+      "learning_rate": 9.668791834906944e-06,
+      "loss": 0.7831,
+      "step": 2337
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.3714386129515863,
+      "learning_rate": 9.661515285939094e-06,
+      "loss": 1.2964,
+      "step": 2338
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3689621276700092,
+      "learning_rate": 9.654238916393862e-06,
+      "loss": 0.6971,
+      "step": 2339
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3323942621134435,
+      "learning_rate": 9.646962730128269e-06,
+      "loss": 0.6762,
+      "step": 2340
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.002929966073999,
+      "learning_rate": 9.63968673099926e-06,
+      "loss": 1.3458,
+      "step": 2341
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4578713968103902,
+      "learning_rate": 9.632410922863655e-06,
+      "loss": 0.7363,
+      "step": 2342
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3075444368488474,
+      "learning_rate": 9.625135309578195e-06,
+      "loss": 0.7258,
+      "step": 2343
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5746281854993776,
+      "learning_rate": 9.617859894999503e-06,
+      "loss": 0.773,
+      "step": 2344
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4208291649346718,
+      "learning_rate": 9.610584682984102e-06,
+      "loss": 0.7518,
+      "step": 2345
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3613233768326256,
+      "learning_rate": 9.603309677388406e-06,
+      "loss": 0.8213,
+      "step": 2346
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4734842828528836,
+      "learning_rate": 9.596034882068724e-06,
+      "loss": 0.7202,
+      "step": 2347
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0642882430894463,
+      "learning_rate": 9.588760300881245e-06,
+      "loss": 1.2492,
+      "step": 2348
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0345758531023885,
+      "learning_rate": 9.581485937682058e-06,
+      "loss": 1.3662,
+      "step": 2349
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4027586112624533,
+      "learning_rate": 9.574211796327123e-06,
+      "loss": 0.7213,
+      "step": 2350
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0077214701101047,
+      "learning_rate": 9.566937880672284e-06,
+      "loss": 1.3716,
+      "step": 2351
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3679172630923204,
+      "learning_rate": 9.559664194573276e-06,
+      "loss": 0.7269,
+      "step": 2352
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3250115271802732,
+      "learning_rate": 9.5523907418857e-06,
+      "loss": 0.7155,
+      "step": 2353
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.407848702083478,
+      "learning_rate": 9.545117526465042e-06,
+      "loss": 0.7797,
+      "step": 2354
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4625754612940598,
+      "learning_rate": 9.537844552166657e-06,
+      "loss": 0.7108,
+      "step": 2355
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.1803589382664916,
+      "learning_rate": 9.530571822845776e-06,
+      "loss": 1.3045,
+      "step": 2356
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9881343682935724,
+      "learning_rate": 9.523299342357493e-06,
+      "loss": 1.3484,
+      "step": 2357
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5567043244412608,
+      "learning_rate": 9.516027114556783e-06,
+      "loss": 0.7565,
+      "step": 2358
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5002921322416913,
+      "learning_rate": 9.508755143298472e-06,
+      "loss": 0.7587,
+      "step": 2359
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4322661023002594,
+      "learning_rate": 9.501483432437264e-06,
+      "loss": 0.7392,
+      "step": 2360
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4419792653223475,
+      "learning_rate": 9.494211985827717e-06,
+      "loss": 0.7847,
+      "step": 2361
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4994665620480807,
+      "learning_rate": 9.48694080732424e-06,
+      "loss": 0.6761,
+      "step": 2362
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.912837456468122,
+      "learning_rate": 9.479669900781125e-06,
+      "loss": 1.3666,
+      "step": 2363
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.961578006293662,
+      "learning_rate": 9.472399270052494e-06,
+      "loss": 1.2653,
+      "step": 2364
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4747853307181769,
+      "learning_rate": 9.465128918992342e-06,
+      "loss": 0.7148,
+      "step": 2365
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.1096920141424302,
+      "learning_rate": 9.457858851454496e-06,
+      "loss": 1.3696,
+      "step": 2366
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5188786969036847,
+      "learning_rate": 9.450589071292658e-06,
+      "loss": 0.6403,
+      "step": 2367
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0051547518565087,
+      "learning_rate": 9.443319582360352e-06,
+      "loss": 1.3614,
+      "step": 2368
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4642624052035984,
+      "learning_rate": 9.43605038851097e-06,
+      "loss": 0.8303,
+      "step": 2369
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4675858869856142,
+      "learning_rate": 9.428781493597727e-06,
+      "loss": 0.7426,
+      "step": 2370
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5076264703140732,
+      "learning_rate": 9.421512901473697e-06,
+      "loss": 0.8628,
+      "step": 2371
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9359039352797445,
+      "learning_rate": 9.414244615991786e-06,
+      "loss": 1.4152,
+      "step": 2372
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6284995826511979,
+      "learning_rate": 9.40697664100473e-06,
+      "loss": 0.7349,
+      "step": 2373
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3870255680557677,
+      "learning_rate": 9.399708980365116e-06,
+      "loss": 0.7433,
+      "step": 2374
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3724007713806399,
+      "learning_rate": 9.392441637925353e-06,
+      "loss": 0.7328,
+      "step": 2375
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0939959938140342,
+      "learning_rate": 9.385174617537686e-06,
+      "loss": 1.4032,
+      "step": 2376
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0492972484082266,
+      "learning_rate": 9.377907923054184e-06,
+      "loss": 1.3171,
+      "step": 2377
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.1899225366067196,
+      "learning_rate": 9.370641558326753e-06,
+      "loss": 1.2942,
+      "step": 2378
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4890626391394308,
+      "learning_rate": 9.363375527207112e-06,
+      "loss": 0.7683,
+      "step": 2379
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9554250433279228,
+      "learning_rate": 9.356109833546817e-06,
+      "loss": 1.3557,
+      "step": 2380
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4982301117225274,
+      "learning_rate": 9.348844481197235e-06,
+      "loss": 0.7193,
+      "step": 2381
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.371072421427961,
+      "learning_rate": 9.341579474009549e-06,
+      "loss": 0.6687,
+      "step": 2382
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.1492017465938975,
+      "learning_rate": 9.334314815834774e-06,
+      "loss": 1.3604,
+      "step": 2383
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.272560032367057,
+      "learning_rate": 9.327050510523722e-06,
+      "loss": 0.779,
+      "step": 2384
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.2274685816760407,
+      "learning_rate": 9.319786561927033e-06,
+      "loss": 1.3097,
+      "step": 2385
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4506952534644302,
+      "learning_rate": 9.31252297389515e-06,
+      "loss": 0.8027,
+      "step": 2386
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.476761251426625,
+      "learning_rate": 9.305259750278329e-06,
+      "loss": 0.6485,
+      "step": 2387
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4845861684341741,
+      "learning_rate": 9.297996894926622e-06,
+      "loss": 0.7057,
+      "step": 2388
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.473821654563341,
+      "learning_rate": 9.290734411689905e-06,
+      "loss": 0.7969,
+      "step": 2389
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9629408600902412,
+      "learning_rate": 9.283472304417836e-06,
+      "loss": 1.3541,
+      "step": 2390
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4689745962110135,
+      "learning_rate": 9.276210576959892e-06,
+      "loss": 0.7751,
+      "step": 2391
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9411849318807983,
+      "learning_rate": 9.268949233165339e-06,
+      "loss": 1.3289,
+      "step": 2392
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.523748420189359,
+      "learning_rate": 9.261688276883233e-06,
+      "loss": 0.7214,
+      "step": 2393
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.1149471616780358,
+      "learning_rate": 9.254427711962442e-06,
+      "loss": 1.3759,
+      "step": 2394
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.019100245121274,
+      "learning_rate": 9.247167542251612e-06,
+      "loss": 1.3301,
+      "step": 2395
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4897552039469049,
+      "learning_rate": 9.239907771599188e-06,
+      "loss": 0.7414,
+      "step": 2396
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5362111586855443,
+      "learning_rate": 9.232648403853393e-06,
+      "loss": 0.7743,
+      "step": 2397
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4213680729317308,
+      "learning_rate": 9.225389442862252e-06,
+      "loss": 0.7411,
+      "step": 2398
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4961448692893093,
+      "learning_rate": 9.21813089247356e-06,
+      "loss": 0.7407,
+      "step": 2399
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.481884960960149,
+      "learning_rate": 9.210872756534907e-06,
+      "loss": 0.8812,
+      "step": 2400
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.418743409847636,
+      "learning_rate": 9.203615038893646e-06,
+      "loss": 0.7341,
+      "step": 2401
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4491517219886294,
+      "learning_rate": 9.19635774339693e-06,
+      "loss": 0.7194,
+      "step": 2402
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4365622870000085,
+      "learning_rate": 9.18910087389167e-06,
+      "loss": 0.7111,
+      "step": 2403
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3453739802961966,
+      "learning_rate": 9.181844434224559e-06,
+      "loss": 0.8095,
+      "step": 2404
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5095968392139738,
+      "learning_rate": 9.174588428242064e-06,
+      "loss": 0.7014,
+      "step": 2405
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4878893142061107,
+      "learning_rate": 9.167332859790419e-06,
+      "loss": 0.8149,
+      "step": 2406
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4210816077193735,
+      "learning_rate": 9.160077732715625e-06,
+      "loss": 0.7212,
+      "step": 2407
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4155603974481195,
+      "learning_rate": 9.152823050863447e-06,
+      "loss": 0.7018,
+      "step": 2408
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.397212924155252,
+      "learning_rate": 9.145568818079428e-06,
+      "loss": 0.7221,
+      "step": 2409
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5287362751195046,
+      "learning_rate": 9.138315038208851e-06,
+      "loss": 0.7926,
+      "step": 2410
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4150596297309224,
+      "learning_rate": 9.131061715096778e-06,
+      "loss": 0.7699,
+      "step": 2411
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.105302678896134,
+      "learning_rate": 9.123808852588017e-06,
+      "loss": 1.2646,
+      "step": 2412
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7742032380231518,
+      "learning_rate": 9.116556454527142e-06,
+      "loss": 0.7652,
+      "step": 2413
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3632608349467001,
+      "learning_rate": 9.109304524758471e-06,
+      "loss": 0.6779,
+      "step": 2414
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.008414506812221,
+      "learning_rate": 9.102053067126079e-06,
+      "loss": 1.3122,
+      "step": 2415
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.007725511878899,
+      "learning_rate": 9.094802085473787e-06,
+      "loss": 1.3409,
+      "step": 2416
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3792210412547654,
+      "learning_rate": 9.087551583645167e-06,
+      "loss": 0.7787,
+      "step": 2417
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.296562404491596,
+      "learning_rate": 9.080301565483539e-06,
+      "loss": 0.7476,
+      "step": 2418
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4665359628430097,
+      "learning_rate": 9.073052034831956e-06,
+      "loss": 0.7301,
+      "step": 2419
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6788302427644726,
+      "learning_rate": 9.065802995533231e-06,
+      "loss": 0.714,
+      "step": 2420
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5153730391040237,
+      "learning_rate": 9.058554451429896e-06,
+      "loss": 0.7972,
+      "step": 2421
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4804710723060817,
+      "learning_rate": 9.051306406364238e-06,
+      "loss": 0.7977,
+      "step": 2422
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.1565428402404803,
+      "learning_rate": 9.044058864178268e-06,
+      "loss": 1.3528,
+      "step": 2423
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3531133128903405,
+      "learning_rate": 9.03681182871373e-06,
+      "loss": 0.7608,
+      "step": 2424
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.369123870873361,
+      "learning_rate": 9.02956530381211e-06,
+      "loss": 0.6564,
+      "step": 2425
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4624968782177872,
+      "learning_rate": 9.022319293314613e-06,
+      "loss": 0.7781,
+      "step": 2426
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4006069223306854,
+      "learning_rate": 9.015073801062178e-06,
+      "loss": 0.7752,
+      "step": 2427
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.451406190673438,
+      "learning_rate": 9.00782883089546e-06,
+      "loss": 0.6737,
+      "step": 2428
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.502841212294487,
+      "learning_rate": 9.000584386654853e-06,
+      "loss": 0.7175,
+      "step": 2429
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5365713866689426,
+      "learning_rate": 8.993340472180453e-06,
+      "loss": 0.7571,
+      "step": 2430
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4681524230850425,
+      "learning_rate": 8.986097091312093e-06,
+      "loss": 0.8365,
+      "step": 2431
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5161102279816412,
+      "learning_rate": 8.978854247889312e-06,
+      "loss": 0.8639,
+      "step": 2432
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.421353829092696,
+      "learning_rate": 8.971611945751368e-06,
+      "loss": 0.7753,
+      "step": 2433
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.09362427445093,
+      "learning_rate": 8.964370188737233e-06,
+      "loss": 1.2923,
+      "step": 2434
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.083585820579166,
+      "learning_rate": 8.957128980685582e-06,
+      "loss": 1.3515,
+      "step": 2435
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9550849404786583,
+      "learning_rate": 8.949888325434816e-06,
+      "loss": 1.3343,
+      "step": 2436
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5137793250524172,
+      "learning_rate": 8.942648226823024e-06,
+      "loss": 0.6922,
+      "step": 2437
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3728101380073359,
+      "learning_rate": 8.935408688688011e-06,
+      "loss": 0.7716,
+      "step": 2438
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4199654259788592,
+      "learning_rate": 8.928169714867278e-06,
+      "loss": 0.7532,
+      "step": 2439
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4292470679164597,
+      "learning_rate": 8.920931309198042e-06,
+      "loss": 0.7475,
+      "step": 2440
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.094174446116837,
+      "learning_rate": 8.913693475517195e-06,
+      "loss": 1.3866,
+      "step": 2441
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4542987211326353,
+      "learning_rate": 8.90645621766135e-06,
+      "loss": 0.7965,
+      "step": 2442
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4278100175887063,
+      "learning_rate": 8.899219539466793e-06,
+      "loss": 0.7686,
+      "step": 2443
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4538828553870087,
+      "learning_rate": 8.891983444769525e-06,
+      "loss": 0.7251,
+      "step": 2444
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3936343213115308,
+      "learning_rate": 8.88474793740522e-06,
+      "loss": 0.7637,
+      "step": 2445
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4053114201390244,
+      "learning_rate": 8.877513021209244e-06,
+      "loss": 0.826,
+      "step": 2446
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.474950925514181,
+      "learning_rate": 8.870278700016654e-06,
+      "loss": 0.721,
+      "step": 2447
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3171647619540885,
+      "learning_rate": 8.863044977662192e-06,
+      "loss": 0.6327,
+      "step": 2448
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4377287282568652,
+      "learning_rate": 8.855811857980282e-06,
+      "loss": 0.7684,
+      "step": 2449
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.475712621518193,
+      "learning_rate": 8.848579344805018e-06,
+      "loss": 0.7547,
+      "step": 2450
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0654809272154777,
+      "learning_rate": 8.841347441970194e-06,
+      "loss": 1.3652,
+      "step": 2451
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4114645244516197,
+      "learning_rate": 8.834116153309256e-06,
+      "loss": 0.741,
+      "step": 2452
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.983928097883752,
+      "learning_rate": 8.826885482655347e-06,
+      "loss": 1.3844,
+      "step": 2453
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.361212964048773,
+      "learning_rate": 8.819655433841263e-06,
+      "loss": 0.7715,
+      "step": 2454
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.1558823816681545,
+      "learning_rate": 8.812426010699488e-06,
+      "loss": 1.3089,
+      "step": 2455
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4355037403039765,
+      "learning_rate": 8.80519721706216e-06,
+      "loss": 0.8009,
+      "step": 2456
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.465106239853551,
+      "learning_rate": 8.797969056761089e-06,
+      "loss": 0.785,
+      "step": 2457
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4060588938666756,
+      "learning_rate": 8.79074153362775e-06,
+      "loss": 0.7079,
+      "step": 2458
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4342555865075204,
+      "learning_rate": 8.783514651493277e-06,
+      "loss": 0.7408,
+      "step": 2459
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.076765125287707,
+      "learning_rate": 8.776288414188472e-06,
+      "loss": 1.2579,
+      "step": 2460
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4572154195684297,
+      "learning_rate": 8.769062825543784e-06,
+      "loss": 0.7553,
+      "step": 2461
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0177586828902956,
+      "learning_rate": 8.76183788938933e-06,
+      "loss": 1.2874,
+      "step": 2462
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.043516530696666,
+      "learning_rate": 8.754613609554868e-06,
+      "loss": 1.2769,
+      "step": 2463
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5384456559683184,
+      "learning_rate": 8.747389989869823e-06,
+      "loss": 0.7135,
+      "step": 2464
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.929921757220168,
+      "learning_rate": 8.740167034163259e-06,
+      "loss": 1.2734,
+      "step": 2465
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0508413462914805,
+      "learning_rate": 8.732944746263887e-06,
+      "loss": 1.3132,
+      "step": 2466
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4224052720967195,
+      "learning_rate": 8.725723130000074e-06,
+      "loss": 0.7828,
+      "step": 2467
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4145497570223367,
+      "learning_rate": 8.718502189199825e-06,
+      "loss": 0.7473,
+      "step": 2468
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5128806708302724,
+      "learning_rate": 8.711281927690784e-06,
+      "loss": 0.7617,
+      "step": 2469
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.082671626335644,
+      "learning_rate": 8.70406234930024e-06,
+      "loss": 1.3107,
+      "step": 2470
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0450166068680433,
+      "learning_rate": 8.696843457855119e-06,
+      "loss": 1.367,
+      "step": 2471
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4813068227129023,
+      "learning_rate": 8.689625257181977e-06,
+      "loss": 0.7384,
+      "step": 2472
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.450227799021133,
+      "learning_rate": 8.682407751107017e-06,
+      "loss": 0.8042,
+      "step": 2473
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0104683250364603,
+      "learning_rate": 8.675190943456058e-06,
+      "loss": 1.3728,
+      "step": 2474
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4612070900707772,
+      "learning_rate": 8.667974838054565e-06,
+      "loss": 0.7433,
+      "step": 2475
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.465189388306478,
+      "learning_rate": 8.660759438727616e-06,
+      "loss": 0.803,
+      "step": 2476
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0458718584590367,
+      "learning_rate": 8.653544749299919e-06,
+      "loss": 1.404,
+      "step": 2477
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.011567462362421,
+      "learning_rate": 8.646330773595812e-06,
+      "loss": 1.3537,
+      "step": 2478
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3140394804466455,
+      "learning_rate": 8.639117515439249e-06,
+      "loss": 0.6428,
+      "step": 2479
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.445788920687675,
+      "learning_rate": 8.631904978653804e-06,
+      "loss": 0.7763,
+      "step": 2480
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4191848286621376,
+      "learning_rate": 8.624693167062668e-06,
+      "loss": 0.827,
+      "step": 2481
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4009134024578367,
+      "learning_rate": 8.617482084488656e-06,
+      "loss": 0.7455,
+      "step": 2482
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3312100598660228,
+      "learning_rate": 8.610271734754178e-06,
+      "loss": 0.7333,
+      "step": 2483
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.401019500620143,
+      "learning_rate": 8.60306212168128e-06,
+      "loss": 0.705,
+      "step": 2484
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.076967273071265,
+      "learning_rate": 8.595853249091595e-06,
+      "loss": 1.2537,
+      "step": 2485
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.346975984356479,
+      "learning_rate": 8.588645120806381e-06,
+      "loss": 0.7032,
+      "step": 2486
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5764983642983026,
+      "learning_rate": 8.581437740646488e-06,
+      "loss": 0.7467,
+      "step": 2487
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.399969915576486,
+      "learning_rate": 8.574231112432377e-06,
+      "loss": 0.8024,
+      "step": 2488
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5047332708194623,
+      "learning_rate": 8.56702523998411e-06,
+      "loss": 0.7636,
+      "step": 2489
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.104269652486922,
+      "learning_rate": 8.559820127121341e-06,
+      "loss": 1.3521,
+      "step": 2490
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5149855193197337,
+      "learning_rate": 8.552615777663335e-06,
+      "loss": 0.7832,
+      "step": 2491
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.378613885261769,
+      "learning_rate": 8.545412195428936e-06,
+      "loss": 0.7338,
+      "step": 2492
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4198947479827153,
+      "learning_rate": 8.538209384236602e-06,
+      "loss": 0.7357,
+      "step": 2493
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4365838794294383,
+      "learning_rate": 8.531007347904359e-06,
+      "loss": 0.696,
+      "step": 2494
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3769162028610926,
+      "learning_rate": 8.52380609024984e-06,
+      "loss": 0.731,
+      "step": 2495
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9763950792705194,
+      "learning_rate": 8.516605615090255e-06,
+      "loss": 1.413,
+      "step": 2496
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3511434551285395,
+      "learning_rate": 8.509405926242404e-06,
+      "loss": 0.7323,
+      "step": 2497
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0371108926434127,
+      "learning_rate": 8.502207027522673e-06,
+      "loss": 1.3088,
+      "step": 2498
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.04157357537549,
+      "learning_rate": 8.49500892274702e-06,
+      "loss": 1.318,
+      "step": 2499
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3243413969556164,
+      "learning_rate": 8.48781161573099e-06,
+      "loss": 0.7287,
+      "step": 2500
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4409031492818558,
+      "learning_rate": 8.480615110289698e-06,
+      "loss": 0.7165,
+      "step": 2501
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4075242390016738,
+      "learning_rate": 8.473419410237846e-06,
+      "loss": 0.6895,
+      "step": 2502
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5191426624798143,
+      "learning_rate": 8.466224519389694e-06,
+      "loss": 0.6975,
+      "step": 2503
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5295994751607738,
+      "learning_rate": 8.45903044155909e-06,
+      "loss": 0.7717,
+      "step": 2504
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3312577433912935,
+      "learning_rate": 8.451837180559434e-06,
+      "loss": 0.6949,
+      "step": 2505
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4043731350190098,
+      "learning_rate": 8.444644740203706e-06,
+      "loss": 0.7296,
+      "step": 2506
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.022535952268921,
+      "learning_rate": 8.437453124304446e-06,
+      "loss": 1.324,
+      "step": 2507
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.22134128807598,
+      "learning_rate": 8.43026233667375e-06,
+      "loss": 1.3395,
+      "step": 2508
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.53808187935062,
+      "learning_rate": 8.42307238112329e-06,
+      "loss": 0.7872,
+      "step": 2509
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9179209552639853,
+      "learning_rate": 8.415883261464282e-06,
+      "loss": 1.4424,
+      "step": 2510
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3597526750923692,
+      "learning_rate": 8.408694981507513e-06,
+      "loss": 0.7149,
+      "step": 2511
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0665362542701566,
+      "learning_rate": 8.40150754506331e-06,
+      "loss": 1.3898,
+      "step": 2512
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0681894394391547,
+      "learning_rate": 8.394320955941565e-06,
+      "loss": 1.3487,
+      "step": 2513
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.207681384054163,
+      "learning_rate": 8.387135217951712e-06,
+      "loss": 1.4304,
+      "step": 2514
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4780835945246578,
+      "learning_rate": 8.379950334902747e-06,
+      "loss": 0.7751,
+      "step": 2515
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4011805680828882,
+      "learning_rate": 8.372766310603191e-06,
+      "loss": 0.6985,
+      "step": 2516
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6407018287876556,
+      "learning_rate": 8.365583148861136e-06,
+      "loss": 0.7685,
+      "step": 2517
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4690162267460511,
+      "learning_rate": 8.358400853484197e-06,
+      "loss": 0.7676,
+      "step": 2518
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.441616745224504,
+      "learning_rate": 8.351219428279536e-06,
+      "loss": 0.8347,
+      "step": 2519
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1490506178142623,
+      "learning_rate": 8.344038877053856e-06,
+      "loss": 1.3524,
+      "step": 2520
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5127357373158252,
+      "learning_rate": 8.336859203613387e-06,
+      "loss": 0.7349,
+      "step": 2521
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4002590933673023,
+      "learning_rate": 8.329680411763915e-06,
+      "loss": 0.7092,
+      "step": 2522
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.147162040896921,
+      "learning_rate": 8.322502505310733e-06,
+      "loss": 1.3346,
+      "step": 2523
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3024624186209806,
+      "learning_rate": 8.315325488058685e-06,
+      "loss": 0.6263,
+      "step": 2524
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0058786950644105,
+      "learning_rate": 8.30814936381213e-06,
+      "loss": 1.3673,
+      "step": 2525
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4492069469125728,
+      "learning_rate": 8.300974136374964e-06,
+      "loss": 0.7481,
+      "step": 2526
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1038690945029765,
+      "learning_rate": 8.293799809550598e-06,
+      "loss": 1.3495,
+      "step": 2527
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.517195080781847,
+      "learning_rate": 8.286626387141974e-06,
+      "loss": 0.6818,
+      "step": 2528
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.005020699620835,
+      "learning_rate": 8.279453872951551e-06,
+      "loss": 1.4338,
+      "step": 2529
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4145167707915993,
+      "learning_rate": 8.272282270781301e-06,
+      "loss": 0.7281,
+      "step": 2530
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.035590524897147,
+      "learning_rate": 8.265111584432725e-06,
+      "loss": 1.3973,
+      "step": 2531
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4202492326234175,
+      "learning_rate": 8.257941817706825e-06,
+      "loss": 0.7819,
+      "step": 2532
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4654526547990627,
+      "learning_rate": 8.250772974404128e-06,
+      "loss": 0.7732,
+      "step": 2533
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4341656296035235,
+      "learning_rate": 8.243605058324661e-06,
+      "loss": 0.7064,
+      "step": 2534
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.507460462326525,
+      "learning_rate": 8.236438073267972e-06,
+      "loss": 0.796,
+      "step": 2535
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5465082856970946,
+      "learning_rate": 8.229272023033095e-06,
+      "loss": 0.713,
+      "step": 2536
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4845817400016308,
+      "learning_rate": 8.222106911418593e-06,
+      "loss": 0.7325,
+      "step": 2537
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5001677947869685,
+      "learning_rate": 8.214942742222513e-06,
+      "loss": 0.7385,
+      "step": 2538
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3975285705098612,
+      "learning_rate": 8.207779519242414e-06,
+      "loss": 0.7362,
+      "step": 2539
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.505116208322573,
+      "learning_rate": 8.200617246275342e-06,
+      "loss": 0.7317,
+      "step": 2540
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.380020798653761,
+      "learning_rate": 8.19345592711785e-06,
+      "loss": 0.7162,
+      "step": 2541
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4311954671242495,
+      "learning_rate": 8.186295565565982e-06,
+      "loss": 0.741,
+      "step": 2542
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.154500868823116,
+      "learning_rate": 8.179136165415266e-06,
+      "loss": 1.3283,
+      "step": 2543
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0985047794973193,
+      "learning_rate": 8.17197773046074e-06,
+      "loss": 1.4064,
+      "step": 2544
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.979919004203612,
+      "learning_rate": 8.16482026449691e-06,
+      "loss": 1.3627,
+      "step": 2545
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.648958893972136,
+      "learning_rate": 8.157663771317782e-06,
+      "loss": 0.7421,
+      "step": 2546
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4622244662300696,
+      "learning_rate": 8.150508254716834e-06,
+      "loss": 0.7606,
+      "step": 2547
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6161051336091405,
+      "learning_rate": 8.143353718487043e-06,
+      "loss": 0.7409,
+      "step": 2548
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.188425372954282,
+      "learning_rate": 8.136200166420851e-06,
+      "loss": 1.2779,
+      "step": 2549
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4308335663063707,
+      "learning_rate": 8.129047602310184e-06,
+      "loss": 0.7351,
+      "step": 2550
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4320033248771027,
+      "learning_rate": 8.121896029946447e-06,
+      "loss": 0.7325,
+      "step": 2551
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3400343647341209,
+      "learning_rate": 8.114745453120509e-06,
+      "loss": 0.7607,
+      "step": 2552
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5051579901366576,
+      "learning_rate": 8.107595875622733e-06,
+      "loss": 0.8304,
+      "step": 2553
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0818636121339926,
+      "learning_rate": 8.100447301242922e-06,
+      "loss": 1.3267,
+      "step": 2554
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4514869119827296,
+      "learning_rate": 8.09329973377038e-06,
+      "loss": 0.7219,
+      "step": 2555
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4904651668482398,
+      "learning_rate": 8.086153176993847e-06,
+      "loss": 0.6893,
+      "step": 2556
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4644291472634776,
+      "learning_rate": 8.079007634701553e-06,
+      "loss": 0.7706,
+      "step": 2557
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4210440318897704,
+      "learning_rate": 8.071863110681167e-06,
+      "loss": 0.7257,
+      "step": 2558
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.048585650664433,
+      "learning_rate": 8.064719608719836e-06,
+      "loss": 1.3589,
+      "step": 2559
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.1421311510938352,
+      "learning_rate": 8.057577132604163e-06,
+      "loss": 1.3043,
+      "step": 2560
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.359825912228062,
+      "learning_rate": 8.050435686120192e-06,
+      "loss": 0.7272,
+      "step": 2561
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.417150166447278,
+      "learning_rate": 8.04329527305344e-06,
+      "loss": 0.704,
+      "step": 2562
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4333835053222925,
+      "learning_rate": 8.036155897188864e-06,
+      "loss": 0.7135,
+      "step": 2563
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4508916912856602,
+      "learning_rate": 8.029017562310883e-06,
+      "loss": 0.7223,
+      "step": 2564
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5074289157830758,
+      "learning_rate": 8.021880272203347e-06,
+      "loss": 0.7918,
+      "step": 2565
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6109753986449384,
+      "learning_rate": 8.014744030649569e-06,
+      "loss": 0.7541,
+      "step": 2566
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4309704040624986,
+      "learning_rate": 8.007608841432296e-06,
+      "loss": 0.6475,
+      "step": 2567
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5019714032183813,
+      "learning_rate": 8.000474708333724e-06,
+      "loss": 0.8106,
+      "step": 2568
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5293775234805225,
+      "learning_rate": 7.993341635135483e-06,
+      "loss": 0.8009,
+      "step": 2569
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2926060485791322,
+      "learning_rate": 7.986209625618644e-06,
+      "loss": 0.6856,
+      "step": 2570
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5721733350429679,
+      "learning_rate": 7.979078683563714e-06,
+      "loss": 0.7881,
+      "step": 2571
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0842435476738785,
+      "learning_rate": 7.971948812750634e-06,
+      "loss": 1.3902,
+      "step": 2572
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.1036966571971574,
+      "learning_rate": 7.96482001695878e-06,
+      "loss": 1.3737,
+      "step": 2573
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9539299794495575,
+      "learning_rate": 7.957692299966947e-06,
+      "loss": 1.3347,
+      "step": 2574
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.618842357298111,
+      "learning_rate": 7.950565665553378e-06,
+      "loss": 0.784,
+      "step": 2575
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3616613534897717,
+      "learning_rate": 7.94344011749572e-06,
+      "loss": 0.7734,
+      "step": 2576
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.575262682459976,
+      "learning_rate": 7.936315659571066e-06,
+      "loss": 0.7195,
+      "step": 2577
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.358772876236496,
+      "learning_rate": 7.929192295555908e-06,
+      "loss": 0.7595,
+      "step": 2578
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.459274826765505,
+      "learning_rate": 7.92207002922618e-06,
+      "loss": 0.7164,
+      "step": 2579
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.374303669403954,
+      "learning_rate": 7.91494886435722e-06,
+      "loss": 0.7185,
+      "step": 2580
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5751457385116712,
+      "learning_rate": 7.90782880472379e-06,
+      "loss": 0.7547,
+      "step": 2581
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3115223318759193,
+      "learning_rate": 7.900709854100057e-06,
+      "loss": 0.6848,
+      "step": 2582
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.405155029945765,
+      "learning_rate": 7.893592016259605e-06,
+      "loss": 0.738,
+      "step": 2583
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4688536657539113,
+      "learning_rate": 7.886475294975437e-06,
+      "loss": 0.7082,
+      "step": 2584
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.070323206305518,
+      "learning_rate": 7.879359694019948e-06,
+      "loss": 1.3337,
+      "step": 2585
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4578206546913806,
+      "learning_rate": 7.872245217164954e-06,
+      "loss": 0.8121,
+      "step": 2586
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4454449017295357,
+      "learning_rate": 7.86513186818166e-06,
+      "loss": 0.8945,
+      "step": 2587
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9759554434021553,
+      "learning_rate": 7.858019650840691e-06,
+      "loss": 1.2697,
+      "step": 2588
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3596898182202668,
+      "learning_rate": 7.850908568912054e-06,
+      "loss": 0.725,
+      "step": 2589
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4772631018089568,
+      "learning_rate": 7.843798626165168e-06,
+      "loss": 0.775,
+      "step": 2590
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.435286939014759,
+      "learning_rate": 7.836689826368844e-06,
+      "loss": 0.7439,
+      "step": 2591
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4017835775339644,
+      "learning_rate": 7.829582173291279e-06,
+      "loss": 0.8193,
+      "step": 2592
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4240864996017188,
+      "learning_rate": 7.822475670700078e-06,
+      "loss": 0.7948,
+      "step": 2593
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4402724597173662,
+      "learning_rate": 7.815370322362218e-06,
+      "loss": 0.7037,
+      "step": 2594
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.565439759847493,
+      "learning_rate": 7.808266132044083e-06,
+      "loss": 0.6961,
+      "step": 2595
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3915708797004507,
+      "learning_rate": 7.801163103511424e-06,
+      "loss": 0.691,
+      "step": 2596
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9519426863374851,
+      "learning_rate": 7.794061240529395e-06,
+      "loss": 1.3204,
+      "step": 2597
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4692263993497514,
+      "learning_rate": 7.786960546862515e-06,
+      "loss": 0.6804,
+      "step": 2598
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4830164223820792,
+      "learning_rate": 7.7798610262747e-06,
+      "loss": 0.7745,
+      "step": 2599
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4884771241151615,
+      "learning_rate": 7.772762682529227e-06,
+      "loss": 0.6605,
+      "step": 2600
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4735384754754832,
+      "learning_rate": 7.765665519388764e-06,
+      "loss": 0.8248,
+      "step": 2601
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5145007848791223,
+      "learning_rate": 7.75856954061534e-06,
+      "loss": 0.7955,
+      "step": 2602
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0179510906632196,
+      "learning_rate": 7.751474749970369e-06,
+      "loss": 1.3367,
+      "step": 2603
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4895594420481757,
+      "learning_rate": 7.744381151214627e-06,
+      "loss": 0.8399,
+      "step": 2604
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.460690880666076,
+      "learning_rate": 7.737288748108256e-06,
+      "loss": 0.7941,
+      "step": 2605
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4602203730709085,
+      "learning_rate": 7.730197544410777e-06,
+      "loss": 0.682,
+      "step": 2606
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.951669130750824,
+      "learning_rate": 7.723107543881058e-06,
+      "loss": 1.4197,
+      "step": 2607
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3346916013549297,
+      "learning_rate": 7.716018750277347e-06,
+      "loss": 0.7119,
+      "step": 2608
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4930638688217852,
+      "learning_rate": 7.708931167357234e-06,
+      "loss": 0.7984,
+      "step": 2609
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4507559191234083,
+      "learning_rate": 7.701844798877684e-06,
+      "loss": 0.7765,
+      "step": 2610
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6559521642080193,
+      "learning_rate": 7.694759648595007e-06,
+      "loss": 0.7379,
+      "step": 2611
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4253216149247172,
+      "learning_rate": 7.687675720264874e-06,
+      "loss": 0.6924,
+      "step": 2612
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5362485419539609,
+      "learning_rate": 7.680593017642302e-06,
+      "loss": 0.8309,
+      "step": 2613
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.968779249142301,
+      "learning_rate": 7.673511544481657e-06,
+      "loss": 1.3405,
+      "step": 2614
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.371467653458409,
+      "learning_rate": 7.66643130453667e-06,
+      "loss": 0.7109,
+      "step": 2615
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.326754293561248,
+      "learning_rate": 7.659352301560394e-06,
+      "loss": 0.6913,
+      "step": 2616
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0227944379062093,
+      "learning_rate": 7.652274539305249e-06,
+      "loss": 1.3432,
+      "step": 2617
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5226704702742968,
+      "learning_rate": 7.645198021522976e-06,
+      "loss": 0.773,
+      "step": 2618
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4348199180861931,
+      "learning_rate": 7.638122751964679e-06,
+      "loss": 0.7618,
+      "step": 2619
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9290384900563011,
+      "learning_rate": 7.631048734380779e-06,
+      "loss": 1.3376,
+      "step": 2620
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0087692235281276,
+      "learning_rate": 7.6239759725210445e-06,
+      "loss": 1.3691,
+      "step": 2621
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.1088249664893173,
+      "learning_rate": 7.616904470134581e-06,
+      "loss": 1.2571,
+      "step": 2622
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4115493753205723,
+      "learning_rate": 7.609834230969819e-06,
+      "loss": 0.7451,
+      "step": 2623
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5414812306983432,
+      "learning_rate": 7.602765258774525e-06,
+      "loss": 0.8142,
+      "step": 2624
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4979686904506357,
+      "learning_rate": 7.5956975572957825e-06,
+      "loss": 0.7507,
+      "step": 2625
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3746461162577521,
+      "learning_rate": 7.588631130280023e-06,
+      "loss": 0.7,
+      "step": 2626
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3205903712529925,
+      "learning_rate": 7.5815659814729795e-06,
+      "loss": 0.6944,
+      "step": 2627
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5049660083746004,
+      "learning_rate": 7.574502114619725e-06,
+      "loss": 0.7404,
+      "step": 2628
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5044088626974053,
+      "learning_rate": 7.567439533464639e-06,
+      "loss": 0.7189,
+      "step": 2629
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3113852181074261,
+      "learning_rate": 7.560378241751434e-06,
+      "loss": 0.7725,
+      "step": 2630
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4872835702369713,
+      "learning_rate": 7.553318243223124e-06,
+      "loss": 0.6935,
+      "step": 2631
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3452093425990093,
+      "learning_rate": 7.5462595416220506e-06,
+      "loss": 0.7125,
+      "step": 2632
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.029453972328117,
+      "learning_rate": 7.539202140689856e-06,
+      "loss": 1.3395,
+      "step": 2633
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.541545849068975,
+      "learning_rate": 7.5321460441675025e-06,
+      "loss": 0.7654,
+      "step": 2634
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3932984225238285,
+      "learning_rate": 7.5250912557952575e-06,
+      "loss": 0.6913,
+      "step": 2635
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4425697308521395,
+      "learning_rate": 7.5180377793126905e-06,
+      "loss": 0.7334,
+      "step": 2636
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4400016622146163,
+      "learning_rate": 7.510985618458687e-06,
+      "loss": 0.7503,
+      "step": 2637
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4699258749566024,
+      "learning_rate": 7.503934776971421e-06,
+      "loss": 0.7275,
+      "step": 2638
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4825500364715345,
+      "learning_rate": 7.496885258588382e-06,
+      "loss": 0.7808,
+      "step": 2639
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4143582583381509,
+      "learning_rate": 7.489837067046342e-06,
+      "loss": 0.7344,
+      "step": 2640
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.495987563927895,
+      "learning_rate": 7.482790206081386e-06,
+      "loss": 0.7167,
+      "step": 2641
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4069615371206123,
+      "learning_rate": 7.47574467942888e-06,
+      "loss": 0.7624,
+      "step": 2642
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3903592956996371,
+      "learning_rate": 7.468700490823493e-06,
+      "loss": 0.7206,
+      "step": 2643
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.000264122814643,
+      "learning_rate": 7.461657643999177e-06,
+      "loss": 1.3853,
+      "step": 2644
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5521725536419089,
+      "learning_rate": 7.454616142689173e-06,
+      "loss": 0.7772,
+      "step": 2645
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2751052314848847,
+      "learning_rate": 7.447575990626019e-06,
+      "loss": 0.6962,
+      "step": 2646
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.36948290344254,
+      "learning_rate": 7.440537191541522e-06,
+      "loss": 0.6461,
+      "step": 2647
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5924837062322796,
+      "learning_rate": 7.43349974916679e-06,
+      "loss": 0.7058,
+      "step": 2648
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.450849246883768,
+      "learning_rate": 7.426463667232193e-06,
+      "loss": 0.7783,
+      "step": 2649
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4819544984804158,
+      "learning_rate": 7.419428949467397e-06,
+      "loss": 0.7508,
+      "step": 2650
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.447176510810315,
+      "learning_rate": 7.412395599601332e-06,
+      "loss": 0.7553,
+      "step": 2651
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4606054246333864,
+      "learning_rate": 7.405363621362209e-06,
+      "loss": 0.7544,
+      "step": 2652
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.417500020084587,
+      "learning_rate": 7.398333018477511e-06,
+      "loss": 0.7313,
+      "step": 2653
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.396306918498642,
+      "learning_rate": 7.391303794673992e-06,
+      "loss": 0.8264,
+      "step": 2654
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0466016112655874,
+      "learning_rate": 7.3842759536776756e-06,
+      "loss": 1.2533,
+      "step": 2655
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3873726071094314,
+      "learning_rate": 7.377249499213846e-06,
+      "loss": 0.7802,
+      "step": 2656
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.473444979057158,
+      "learning_rate": 7.370224435007065e-06,
+      "loss": 0.7624,
+      "step": 2657
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4340260819716053,
+      "learning_rate": 7.363200764781145e-06,
+      "loss": 0.7117,
+      "step": 2658
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3343674226443447,
+      "learning_rate": 7.356178492259172e-06,
+      "loss": 0.6953,
+      "step": 2659
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4788014924634159,
+      "learning_rate": 7.349157621163473e-06,
+      "loss": 0.7369,
+      "step": 2660
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.454675783202374,
+      "learning_rate": 7.342138155215655e-06,
+      "loss": 0.6815,
+      "step": 2661
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.416270391536268,
+      "learning_rate": 7.335120098136561e-06,
+      "loss": 0.7928,
+      "step": 2662
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5255716248107982,
+      "learning_rate": 7.328103453646298e-06,
+      "loss": 0.798,
+      "step": 2663
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4686359941271085,
+      "learning_rate": 7.321088225464215e-06,
+      "loss": 0.8229,
+      "step": 2664
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.654699740128114,
+      "learning_rate": 7.314074417308924e-06,
+      "loss": 0.7621,
+      "step": 2665
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9735148009252355,
+      "learning_rate": 7.307062032898274e-06,
+      "loss": 1.345,
+      "step": 2666
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.540598888025704,
+      "learning_rate": 7.300051075949352e-06,
+      "loss": 0.8739,
+      "step": 2667
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4497846868748079,
+      "learning_rate": 7.293041550178513e-06,
+      "loss": 0.735,
+      "step": 2668
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4128843107668867,
+      "learning_rate": 7.286033459301325e-06,
+      "loss": 0.7689,
+      "step": 2669
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4086966566272408,
+      "learning_rate": 7.27902680703262e-06,
+      "loss": 0.6823,
+      "step": 2670
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4705466745279399,
+      "learning_rate": 7.272021597086445e-06,
+      "loss": 0.7061,
+      "step": 2671
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.92258275650207,
+      "learning_rate": 7.265017833176105e-06,
+      "loss": 1.3559,
+      "step": 2672
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3789877523681282,
+      "learning_rate": 7.258015519014119e-06,
+      "loss": 0.7107,
+      "step": 2673
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0937715041164457,
+      "learning_rate": 7.2510146583122495e-06,
+      "loss": 1.2737,
+      "step": 2674
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5201984825944808,
+      "learning_rate": 7.244015254781482e-06,
+      "loss": 0.7638,
+      "step": 2675
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.131367755573301,
+      "learning_rate": 7.237017312132031e-06,
+      "loss": 1.4079,
+      "step": 2676
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4530265786913825,
+      "learning_rate": 7.230020834073341e-06,
+      "loss": 0.7592,
+      "step": 2677
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.961289517557446,
+      "learning_rate": 7.223025824314074e-06,
+      "loss": 1.4104,
+      "step": 2678
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9771173010929568,
+      "learning_rate": 7.216032286562122e-06,
+      "loss": 1.456,
+      "step": 2679
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4172242743902284,
+      "learning_rate": 7.209040224524585e-06,
+      "loss": 0.7651,
+      "step": 2680
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4232905553649933,
+      "learning_rate": 7.202049641907794e-06,
+      "loss": 0.6962,
+      "step": 2681
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.422975497424689,
+      "learning_rate": 7.195060542417282e-06,
+      "loss": 0.6254,
+      "step": 2682
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.512194497174519,
+      "learning_rate": 7.188072929757806e-06,
+      "loss": 0.8299,
+      "step": 2683
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3889026480502291,
+      "learning_rate": 7.181086807633331e-06,
+      "loss": 0.7863,
+      "step": 2684
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.936333890936938,
+      "learning_rate": 7.174102179747036e-06,
+      "loss": 1.3256,
+      "step": 2685
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4574783407062564,
+      "learning_rate": 7.167119049801301e-06,
+      "loss": 0.7398,
+      "step": 2686
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3817354715150476,
+      "learning_rate": 7.16013742149771e-06,
+      "loss": 0.7556,
+      "step": 2687
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5565652471353917,
+      "learning_rate": 7.15315729853707e-06,
+      "loss": 0.8323,
+      "step": 2688
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.366233472976688,
+      "learning_rate": 7.146178684619363e-06,
+      "loss": 0.7418,
+      "step": 2689
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.170364137728162,
+      "learning_rate": 7.139201583443793e-06,
+      "loss": 1.2956,
+      "step": 2690
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4058880047532865,
+      "learning_rate": 7.13222599870875e-06,
+      "loss": 0.7566,
+      "step": 2691
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0066667030700924,
+      "learning_rate": 7.1252519341118275e-06,
+      "loss": 1.3541,
+      "step": 2692
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.386055196920674,
+      "learning_rate": 7.118279393349807e-06,
+      "loss": 0.7425,
+      "step": 2693
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9486673655861793,
+      "learning_rate": 7.111308380118666e-06,
+      "loss": 1.3519,
+      "step": 2694
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.439215081698759,
+      "learning_rate": 7.104338898113566e-06,
+      "loss": 0.712,
+      "step": 2695
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5938286847650067,
+      "learning_rate": 7.097370951028872e-06,
+      "loss": 0.5935,
+      "step": 2696
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0382392096310014,
+      "learning_rate": 7.090404542558119e-06,
+      "loss": 1.3169,
+      "step": 2697
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.413894422867096,
+      "learning_rate": 7.08343967639403e-06,
+      "loss": 0.7452,
+      "step": 2698
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.360656735072184,
+      "learning_rate": 7.07647635622852e-06,
+      "loss": 0.7473,
+      "step": 2699
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3736337737230515,
+      "learning_rate": 7.069514585752671e-06,
+      "loss": 0.7169,
+      "step": 2700
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0080857714256166,
+      "learning_rate": 7.062554368656758e-06,
+      "loss": 1.3485,
+      "step": 2701
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.106479790846046,
+      "learning_rate": 7.055595708630217e-06,
+      "loss": 1.3082,
+      "step": 2702
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4807301424787036,
+      "learning_rate": 7.048638609361676e-06,
+      "loss": 0.8029,
+      "step": 2703
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3246113189312267,
+      "learning_rate": 7.041683074538916e-06,
+      "loss": 0.7217,
+      "step": 2704
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4623299142803317,
+      "learning_rate": 7.034729107848907e-06,
+      "loss": 0.7585,
+      "step": 2705
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4439309925025092,
+      "learning_rate": 7.027776712977772e-06,
+      "loss": 0.7527,
+      "step": 2706
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.090639255614897,
+      "learning_rate": 7.020825893610816e-06,
+      "loss": 1.3276,
+      "step": 2707
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3151014500836598,
+      "learning_rate": 7.013876653432501e-06,
+      "loss": 0.7688,
+      "step": 2708
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4419463504430816,
+      "learning_rate": 7.006928996126443e-06,
+      "loss": 0.7044,
+      "step": 2709
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9325020126163603,
+      "learning_rate": 6.999982925375439e-06,
+      "loss": 1.384,
+      "step": 2710
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5332114626209408,
+      "learning_rate": 6.9930384448614265e-06,
+      "loss": 0.7694,
+      "step": 2711
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.453068161760663,
+      "learning_rate": 6.986095558265516e-06,
+      "loss": 0.758,
+      "step": 2712
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4076429913328736,
+      "learning_rate": 6.979154269267958e-06,
+      "loss": 0.7705,
+      "step": 2713
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3486896934111638,
+      "learning_rate": 6.972214581548166e-06,
+      "loss": 0.7225,
+      "step": 2714
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5817199760504448,
+      "learning_rate": 6.965276498784701e-06,
+      "loss": 0.8586,
+      "step": 2715
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3168538459504686,
+      "learning_rate": 6.958340024655274e-06,
+      "loss": 0.6858,
+      "step": 2716
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9656056363865335,
+      "learning_rate": 6.9514051628367465e-06,
+      "loss": 1.383,
+      "step": 2717
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3936148788853786,
+      "learning_rate": 6.944471917005114e-06,
+      "loss": 0.7654,
+      "step": 2718
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.44164712860213,
+      "learning_rate": 6.937540290835534e-06,
+      "loss": 0.7433,
+      "step": 2719
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4125172866884907,
+      "learning_rate": 6.930610288002284e-06,
+      "loss": 0.746,
+      "step": 2720
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.417872101147799,
+      "learning_rate": 6.9236819121788035e-06,
+      "loss": 0.7527,
+      "step": 2721
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.453685953537016,
+      "learning_rate": 6.916755167037647e-06,
+      "loss": 0.8209,
+      "step": 2722
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3132920352596211,
+      "learning_rate": 6.909830056250527e-06,
+      "loss": 0.666,
+      "step": 2723
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4311727284283782,
+      "learning_rate": 6.902906583488269e-06,
+      "loss": 0.7898,
+      "step": 2724
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3677159688707528,
+      "learning_rate": 6.895984752420848e-06,
+      "loss": 0.6967,
+      "step": 2725
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4716517115229362,
+      "learning_rate": 6.889064566717349e-06,
+      "loss": 0.6872,
+      "step": 2726
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.896816106794466,
+      "learning_rate": 6.882146030046008e-06,
+      "loss": 1.3332,
+      "step": 2727
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3693483175237946,
+      "learning_rate": 6.875229146074169e-06,
+      "loss": 0.8516,
+      "step": 2728
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5397474095886672,
+      "learning_rate": 6.868313918468305e-06,
+      "loss": 0.8269,
+      "step": 2729
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4269710613748123,
+      "learning_rate": 6.861400350894016e-06,
+      "loss": 0.7334,
+      "step": 2730
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3997286110698446,
+      "learning_rate": 6.854488447016014e-06,
+      "loss": 0.7216,
+      "step": 2731
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4657291640685122,
+      "learning_rate": 6.847578210498139e-06,
+      "loss": 0.7256,
+      "step": 2732
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4428388676880062,
+      "learning_rate": 6.840669645003334e-06,
+      "loss": 0.7375,
+      "step": 2733
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3592776655190886,
+      "learning_rate": 6.8337627541936714e-06,
+      "loss": 0.7262,
+      "step": 2734
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9557035016585462,
+      "learning_rate": 6.826857541730323e-06,
+      "loss": 1.2947,
+      "step": 2735
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4183229674664481,
+      "learning_rate": 6.819954011273577e-06,
+      "loss": 0.7251,
+      "step": 2736
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8898362570029634,
+      "learning_rate": 6.813052166482828e-06,
+      "loss": 1.2954,
+      "step": 2737
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3678455277560708,
+      "learning_rate": 6.806152011016582e-06,
+      "loss": 0.6782,
+      "step": 2738
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3926107660328004,
+      "learning_rate": 6.799253548532445e-06,
+      "loss": 0.7841,
+      "step": 2739
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3501431588666497,
+      "learning_rate": 6.79235678268712e-06,
+      "loss": 0.7485,
+      "step": 2740
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4914313886495942,
+      "learning_rate": 6.7854617171364236e-06,
+      "loss": 0.7626,
+      "step": 2741
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3730266430616487,
+      "learning_rate": 6.77856835553526e-06,
+      "loss": 0.7309,
+      "step": 2742
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.396953629668719,
+      "learning_rate": 6.77167670153764e-06,
+      "loss": 0.7387,
+      "step": 2743
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1932184571546491,
+      "learning_rate": 6.7647867587966595e-06,
+      "loss": 0.7456,
+      "step": 2744
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.011660253798577,
+      "learning_rate": 6.757898530964513e-06,
+      "loss": 1.3244,
+      "step": 2745
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4437591825276659,
+      "learning_rate": 6.751012021692484e-06,
+      "loss": 0.644,
+      "step": 2746
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.455882341223061,
+      "learning_rate": 6.744127234630948e-06,
+      "loss": 0.7506,
+      "step": 2747
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0648752810341855,
+      "learning_rate": 6.737244173429361e-06,
+      "loss": 1.3627,
+      "step": 2748
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9043898763578764,
+      "learning_rate": 6.730362841736273e-06,
+      "loss": 1.3326,
+      "step": 2749
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4946390401254546,
+      "learning_rate": 6.7234832431993115e-06,
+      "loss": 0.7787,
+      "step": 2750
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4150495047968568,
+      "learning_rate": 6.716605381465181e-06,
+      "loss": 0.71,
+      "step": 2751
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3480805838556156,
+      "learning_rate": 6.70972926017968e-06,
+      "loss": 0.707,
+      "step": 2752
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5634860082260897,
+      "learning_rate": 6.702854882987666e-06,
+      "loss": 0.7228,
+      "step": 2753
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6236599477556817,
+      "learning_rate": 6.69598225353309e-06,
+      "loss": 0.7802,
+      "step": 2754
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.479077581349084,
+      "learning_rate": 6.689111375458961e-06,
+      "loss": 0.7898,
+      "step": 2755
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5025721622325976,
+      "learning_rate": 6.682242252407372e-06,
+      "loss": 0.788,
+      "step": 2756
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9977876442910294,
+      "learning_rate": 6.6753748880194725e-06,
+      "loss": 1.3676,
+      "step": 2757
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.357432385253729,
+      "learning_rate": 6.668509285935496e-06,
+      "loss": 0.5967,
+      "step": 2758
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.391169798479651,
+      "learning_rate": 6.66164544979473e-06,
+      "loss": 0.7282,
+      "step": 2759
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4652935077496183,
+      "learning_rate": 6.654783383235525e-06,
+      "loss": 0.7851,
+      "step": 2760
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3531368386214244,
+      "learning_rate": 6.647923089895304e-06,
+      "loss": 0.7231,
+      "step": 2761
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3657427542594376,
+      "learning_rate": 6.641064573410537e-06,
+      "loss": 0.6762,
+      "step": 2762
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4096302494453132,
+      "learning_rate": 6.634207837416768e-06,
+      "loss": 0.7646,
+      "step": 2763
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3545176957274367,
+      "learning_rate": 6.627352885548577e-06,
+      "loss": 0.7595,
+      "step": 2764
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.027444943562145,
+      "learning_rate": 6.620499721439619e-06,
+      "loss": 1.3137,
+      "step": 2765
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3788896520747724,
+      "learning_rate": 6.613648348722588e-06,
+      "loss": 0.7061,
+      "step": 2766
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.027441817973188,
+      "learning_rate": 6.606798771029233e-06,
+      "loss": 1.3803,
+      "step": 2767
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3471427461741114,
+      "learning_rate": 6.599950991990345e-06,
+      "loss": 0.6819,
+      "step": 2768
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3641741687427122,
+      "learning_rate": 6.593105015235776e-06,
+      "loss": 0.6884,
+      "step": 2769
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4253607141480025,
+      "learning_rate": 6.586260844394411e-06,
+      "loss": 0.6734,
+      "step": 2770
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.005084811828053,
+      "learning_rate": 6.579418483094173e-06,
+      "loss": 1.3324,
+      "step": 2771
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4171457337113227,
+      "learning_rate": 6.572577934962046e-06,
+      "loss": 0.7322,
+      "step": 2772
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.37892074219273,
+      "learning_rate": 6.565739203624028e-06,
+      "loss": 0.7175,
+      "step": 2773
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4806530431776468,
+      "learning_rate": 6.558902292705176e-06,
+      "loss": 0.82,
+      "step": 2774
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3202149723992127,
+      "learning_rate": 6.552067205829568e-06,
+      "loss": 0.7035,
+      "step": 2775
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4770886955487286,
+      "learning_rate": 6.545233946620318e-06,
+      "loss": 0.8004,
+      "step": 2776
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4153562935859163,
+      "learning_rate": 6.5384025186995755e-06,
+      "loss": 0.7536,
+      "step": 2777
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.52510891435155,
+      "learning_rate": 6.531572925688516e-06,
+      "loss": 0.792,
+      "step": 2778
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.945747929832291,
+      "learning_rate": 6.524745171207339e-06,
+      "loss": 1.3494,
+      "step": 2779
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2950257014665558,
+      "learning_rate": 6.517919258875281e-06,
+      "loss": 0.7318,
+      "step": 2780
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3881371772759805,
+      "learning_rate": 6.511095192310587e-06,
+      "loss": 0.6277,
+      "step": 2781
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9143045577917057,
+      "learning_rate": 6.50427297513053e-06,
+      "loss": 1.3286,
+      "step": 2782
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5200162435398168,
+      "learning_rate": 6.49745261095141e-06,
+      "loss": 0.7075,
+      "step": 2783
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4769056967729977,
+      "learning_rate": 6.490634103388531e-06,
+      "loss": 0.5985,
+      "step": 2784
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5123902783894145,
+      "learning_rate": 6.483817456056228e-06,
+      "loss": 0.7766,
+      "step": 2785
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4432393489748645,
+      "learning_rate": 6.477002672567836e-06,
+      "loss": 0.7563,
+      "step": 2786
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3712134335710977,
+      "learning_rate": 6.470189756535712e-06,
+      "loss": 0.7374,
+      "step": 2787
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.475186077225775,
+      "learning_rate": 6.463378711571212e-06,
+      "loss": 0.7343,
+      "step": 2788
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.072366231398787,
+      "learning_rate": 6.456569541284717e-06,
+      "loss": 1.3442,
+      "step": 2789
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4597868707687458,
+      "learning_rate": 6.449762249285596e-06,
+      "loss": 0.7185,
+      "step": 2790
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0053879641550907,
+      "learning_rate": 6.44295683918224e-06,
+      "loss": 1.3246,
+      "step": 2791
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3985682829142723,
+      "learning_rate": 6.436153314582026e-06,
+      "loss": 0.7765,
+      "step": 2792
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5187432013656836,
+      "learning_rate": 6.42935167909134e-06,
+      "loss": 0.7373,
+      "step": 2793
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5099852034993846,
+      "learning_rate": 6.422551936315568e-06,
+      "loss": 0.716,
+      "step": 2794
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.065644323742865,
+      "learning_rate": 6.415754089859087e-06,
+      "loss": 1.3238,
+      "step": 2795
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4999713500781986,
+      "learning_rate": 6.40895814332528e-06,
+      "loss": 0.7849,
+      "step": 2796
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3861658255717662,
+      "learning_rate": 6.402164100316506e-06,
+      "loss": 0.728,
+      "step": 2797
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3882487157846113,
+      "learning_rate": 6.395371964434129e-06,
+      "loss": 0.7631,
+      "step": 2798
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4705359577143904,
+      "learning_rate": 6.388581739278492e-06,
+      "loss": 0.7773,
+      "step": 2799
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3546670736869149,
+      "learning_rate": 6.381793428448939e-06,
+      "loss": 0.6778,
+      "step": 2800
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0278981931718807,
+      "learning_rate": 6.375007035543784e-06,
+      "loss": 1.419,
+      "step": 2801
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.948270065774976,
+      "learning_rate": 6.368222564160329e-06,
+      "loss": 1.324,
+      "step": 2802
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.975839870087413,
+      "learning_rate": 6.361440017894866e-06,
+      "loss": 1.299,
+      "step": 2803
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3246467128362558,
+      "learning_rate": 6.354659400342653e-06,
+      "loss": 0.6717,
+      "step": 2804
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5520830586842496,
+      "learning_rate": 6.34788071509794e-06,
+      "loss": 0.8363,
+      "step": 2805
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3319285164155028,
+      "learning_rate": 6.341103965753939e-06,
+      "loss": 0.6913,
+      "step": 2806
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9612465761360203,
+      "learning_rate": 6.334329155902843e-06,
+      "loss": 1.2805,
+      "step": 2807
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0142904988425565,
+      "learning_rate": 6.327556289135819e-06,
+      "loss": 1.4107,
+      "step": 2808
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.485011718750147,
+      "learning_rate": 6.320785369043001e-06,
+      "loss": 0.7547,
+      "step": 2809
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6430819847198261,
+      "learning_rate": 6.314016399213485e-06,
+      "loss": 0.7552,
+      "step": 2810
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.98395410153276,
+      "learning_rate": 6.307249383235348e-06,
+      "loss": 1.2755,
+      "step": 2811
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3611304088582314,
+      "learning_rate": 6.300484324695617e-06,
+      "loss": 0.7658,
+      "step": 2812
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.553411314358259,
+      "learning_rate": 6.293721227180286e-06,
+      "loss": 0.6901,
+      "step": 2813
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9095014915124735,
+      "learning_rate": 6.286960094274317e-06,
+      "loss": 1.3688,
+      "step": 2814
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3426379260168455,
+      "learning_rate": 6.280200929561615e-06,
+      "loss": 0.7285,
+      "step": 2815
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5113881988336426,
+      "learning_rate": 6.27344373662506e-06,
+      "loss": 0.7738,
+      "step": 2816
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3844354289623648,
+      "learning_rate": 6.266688519046471e-06,
+      "loss": 0.7597,
+      "step": 2817
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3556900270381387,
+      "learning_rate": 6.259935280406633e-06,
+      "loss": 0.7196,
+      "step": 2818
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4112589237299649,
+      "learning_rate": 6.253184024285267e-06,
+      "loss": 0.7498,
+      "step": 2819
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.437808566393213,
+      "learning_rate": 6.24643475426106e-06,
+      "loss": 0.7844,
+      "step": 2820
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4683255577975345,
+      "learning_rate": 6.2396874739116306e-06,
+      "loss": 0.7335,
+      "step": 2821
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5123736652198125,
+      "learning_rate": 6.232942186813559e-06,
+      "loss": 0.7358,
+      "step": 2822
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4406171437343624,
+      "learning_rate": 6.2261988965423535e-06,
+      "loss": 0.8374,
+      "step": 2823
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4433979493342366,
+      "learning_rate": 6.219457606672468e-06,
+      "loss": 0.7604,
+      "step": 2824
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.401808155314247,
+      "learning_rate": 6.212718320777307e-06,
+      "loss": 0.7711,
+      "step": 2825
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5048775538381318,
+      "learning_rate": 6.205981042429193e-06,
+      "loss": 0.8224,
+      "step": 2826
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4362466273741257,
+      "learning_rate": 6.199245775199407e-06,
+      "loss": 0.7655,
+      "step": 2827
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4873207144525047,
+      "learning_rate": 6.192512522658144e-06,
+      "loss": 0.7371,
+      "step": 2828
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3700623887258079,
+      "learning_rate": 6.185781288374545e-06,
+      "loss": 0.746,
+      "step": 2829
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4191401540023527,
+      "learning_rate": 6.179052075916668e-06,
+      "loss": 0.7375,
+      "step": 2830
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4771438328789126,
+      "learning_rate": 6.172324888851516e-06,
+      "loss": 0.6858,
+      "step": 2831
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9771975427696118,
+      "learning_rate": 6.165599730745002e-06,
+      "loss": 1.3802,
+      "step": 2832
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3736676582577243,
+      "learning_rate": 6.1588766051619774e-06,
+      "loss": 0.7631,
+      "step": 2833
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5953093403729985,
+      "learning_rate": 6.152155515666206e-06,
+      "loss": 0.8109,
+      "step": 2834
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.58212105068959,
+      "learning_rate": 6.145436465820371e-06,
+      "loss": 0.7889,
+      "step": 2835
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4708274452161905,
+      "learning_rate": 6.13871945918609e-06,
+      "loss": 0.746,
+      "step": 2836
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.443907553777772,
+      "learning_rate": 6.132004499323878e-06,
+      "loss": 0.7729,
+      "step": 2837
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.377475412501214,
+      "learning_rate": 6.125291589793179e-06,
+      "loss": 0.7201,
+      "step": 2838
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4738697434605366,
+      "learning_rate": 6.1185807341523405e-06,
+      "loss": 0.7597,
+      "step": 2839
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.329945088247552,
+      "learning_rate": 6.111871935958633e-06,
+      "loss": 0.7484,
+      "step": 2840
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9470075353260838,
+      "learning_rate": 6.105165198768219e-06,
+      "loss": 1.342,
+      "step": 2841
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.011339945254417,
+      "learning_rate": 6.09846052613619e-06,
+      "loss": 1.3593,
+      "step": 2842
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3687092937180583,
+      "learning_rate": 6.091757921616526e-06,
+      "loss": 0.6984,
+      "step": 2843
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9953760236134066,
+      "learning_rate": 6.085057388762114e-06,
+      "loss": 1.3634,
+      "step": 2844
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.012181513646272,
+      "learning_rate": 6.078358931124753e-06,
+      "loss": 1.275,
+      "step": 2845
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3819435950714707,
+      "learning_rate": 6.071662552255125e-06,
+      "loss": 0.6769,
+      "step": 2846
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.445314513550757,
+      "learning_rate": 6.06496825570283e-06,
+      "loss": 0.7373,
+      "step": 2847
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5173649368265199,
+      "learning_rate": 6.058276045016347e-06,
+      "loss": 0.7275,
+      "step": 2848
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4000080264793855,
+      "learning_rate": 6.051585923743059e-06,
+      "loss": 0.7859,
+      "step": 2849
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4799044870044478,
+      "learning_rate": 6.044897895429234e-06,
+      "loss": 0.8021,
+      "step": 2850
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5104544587667434,
+      "learning_rate": 6.038211963620042e-06,
+      "loss": 0.7543,
+      "step": 2851
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3045028501611406,
+      "learning_rate": 6.031528131859529e-06,
+      "loss": 0.6803,
+      "step": 2852
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5336732160920619,
+      "learning_rate": 6.024846403690639e-06,
+      "loss": 0.6866,
+      "step": 2853
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9694759224157992,
+      "learning_rate": 6.018166782655193e-06,
+      "loss": 1.3203,
+      "step": 2854
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4554764851902902,
+      "learning_rate": 6.011489272293894e-06,
+      "loss": 0.7369,
+      "step": 2855
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4884896462257269,
+      "learning_rate": 6.004813876146338e-06,
+      "loss": 0.7719,
+      "step": 2856
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0163208210533266,
+      "learning_rate": 5.998140597750982e-06,
+      "loss": 1.4005,
+      "step": 2857
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3533326416677187,
+      "learning_rate": 5.991469440645181e-06,
+      "loss": 0.6975,
+      "step": 2858
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9942138694590268,
+      "learning_rate": 5.984800408365148e-06,
+      "loss": 1.3713,
+      "step": 2859
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9445905441177789,
+      "learning_rate": 5.978133504445982e-06,
+      "loss": 1.3492,
+      "step": 2860
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5679300645197745,
+      "learning_rate": 5.971468732421641e-06,
+      "loss": 0.7487,
+      "step": 2861
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4862048590813852,
+      "learning_rate": 5.964806095824968e-06,
+      "loss": 0.7663,
+      "step": 2862
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.360705364360144,
+      "learning_rate": 5.958145598187662e-06,
+      "loss": 0.776,
+      "step": 2863
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.075520447816987,
+      "learning_rate": 5.951487243040299e-06,
+      "loss": 1.4108,
+      "step": 2864
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4641324423302635,
+      "learning_rate": 5.944831033912308e-06,
+      "loss": 0.7828,
+      "step": 2865
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.457121193711552,
+      "learning_rate": 5.938176974331982e-06,
+      "loss": 0.7719,
+      "step": 2866
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.021293918535668,
+      "learning_rate": 5.931525067826488e-06,
+      "loss": 1.3023,
+      "step": 2867
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5076416679903661,
+      "learning_rate": 5.924875317921834e-06,
+      "loss": 0.7352,
+      "step": 2868
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0586376676039904,
+      "learning_rate": 5.918227728142895e-06,
+      "loss": 1.3112,
+      "step": 2869
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3848046187126628,
+      "learning_rate": 5.911582302013399e-06,
+      "loss": 0.6592,
+      "step": 2870
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.380688012318203,
+      "learning_rate": 5.904939043055928e-06,
+      "loss": 0.7561,
+      "step": 2871
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.444708141798832,
+      "learning_rate": 5.89829795479191e-06,
+      "loss": 0.7575,
+      "step": 2872
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3806967706758042,
+      "learning_rate": 5.891659040741633e-06,
+      "loss": 0.7777,
+      "step": 2873
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9873176799556618,
+      "learning_rate": 5.885022304424217e-06,
+      "loss": 1.3197,
+      "step": 2874
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9702957682983846,
+      "learning_rate": 5.878387749357646e-06,
+      "loss": 1.3308,
+      "step": 2875
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9844721131767948,
+      "learning_rate": 5.871755379058734e-06,
+      "loss": 1.3967,
+      "step": 2876
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.367379364510473,
+      "learning_rate": 5.865125197043137e-06,
+      "loss": 0.7681,
+      "step": 2877
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4223025900772792,
+      "learning_rate": 5.858497206825362e-06,
+      "loss": 0.7188,
+      "step": 2878
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9589959990072958,
+      "learning_rate": 5.851871411918744e-06,
+      "loss": 1.342,
+      "step": 2879
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4177018754839823,
+      "learning_rate": 5.84524781583546e-06,
+      "loss": 0.7489,
+      "step": 2880
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4798196425962438,
+      "learning_rate": 5.838626422086514e-06,
+      "loss": 0.7802,
+      "step": 2881
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4961036642325363,
+      "learning_rate": 5.832007234181754e-06,
+      "loss": 0.8219,
+      "step": 2882
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0603346001072365,
+      "learning_rate": 5.825390255629848e-06,
+      "loss": 1.362,
+      "step": 2883
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4442585567449107,
+      "learning_rate": 5.8187754899383046e-06,
+      "loss": 0.682,
+      "step": 2884
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4919356332470695,
+      "learning_rate": 5.8121629406134414e-06,
+      "loss": 0.7069,
+      "step": 2885
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4210781413322524,
+      "learning_rate": 5.805552611160419e-06,
+      "loss": 0.7586,
+      "step": 2886
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4228784804169625,
+      "learning_rate": 5.798944505083213e-06,
+      "loss": 0.7557,
+      "step": 2887
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.108127307467997,
+      "learning_rate": 5.7923386258846194e-06,
+      "loss": 1.347,
+      "step": 2888
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4307251018476157,
+      "learning_rate": 5.78573497706626e-06,
+      "loss": 0.7464,
+      "step": 2889
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4136480819172486,
+      "learning_rate": 5.779133562128567e-06,
+      "loss": 0.7822,
+      "step": 2890
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6365360737839125,
+      "learning_rate": 5.772534384570796e-06,
+      "loss": 0.7649,
+      "step": 2891
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5223255340897024,
+      "learning_rate": 5.765937447891008e-06,
+      "loss": 0.7365,
+      "step": 2892
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.333476142390147,
+      "learning_rate": 5.7593427555860835e-06,
+      "loss": 0.7074,
+      "step": 2893
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4532722699774996,
+      "learning_rate": 5.752750311151709e-06,
+      "loss": 0.727,
+      "step": 2894
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5627837934135618,
+      "learning_rate": 5.746160118082386e-06,
+      "loss": 0.828,
+      "step": 2895
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.955216316922757,
+      "learning_rate": 5.739572179871409e-06,
+      "loss": 1.3239,
+      "step": 2896
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5028882145046911,
+      "learning_rate": 5.732986500010891e-06,
+      "loss": 0.7826,
+      "step": 2897
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4675819925771159,
+      "learning_rate": 5.7264030819917405e-06,
+      "loss": 0.7333,
+      "step": 2898
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4821027296318827,
+      "learning_rate": 5.719821929303671e-06,
+      "loss": 0.7806,
+      "step": 2899
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0080495257211286,
+      "learning_rate": 5.713243045435193e-06,
+      "loss": 1.3342,
+      "step": 2900
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9012532551546852,
+      "learning_rate": 5.706666433873616e-06,
+      "loss": 1.3662,
+      "step": 2901
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3886695937923579,
+      "learning_rate": 5.700092098105041e-06,
+      "loss": 0.6924,
+      "step": 2902
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.031693992178242,
+      "learning_rate": 5.6935200416143665e-06,
+      "loss": 1.3262,
+      "step": 2903
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.002789893370712,
+      "learning_rate": 5.686950267885284e-06,
+      "loss": 1.3747,
+      "step": 2904
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4022999726158076,
+      "learning_rate": 5.680382780400271e-06,
+      "loss": 0.7415,
+      "step": 2905
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9448909688902343,
+      "learning_rate": 5.673817582640597e-06,
+      "loss": 1.3739,
+      "step": 2906
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9579326592443602,
+      "learning_rate": 5.667254678086313e-06,
+      "loss": 1.3306,
+      "step": 2907
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.544361355907107,
+      "learning_rate": 5.660694070216255e-06,
+      "loss": 0.7106,
+      "step": 2908
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4237147039876905,
+      "learning_rate": 5.654135762508048e-06,
+      "loss": 0.7295,
+      "step": 2909
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.389752846219808,
+      "learning_rate": 5.647579758438091e-06,
+      "loss": 0.7204,
+      "step": 2910
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4229813865418333,
+      "learning_rate": 5.641026061481565e-06,
+      "loss": 0.6881,
+      "step": 2911
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3946228778332455,
+      "learning_rate": 5.634474675112428e-06,
+      "loss": 0.7734,
+      "step": 2912
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0955409197374038,
+      "learning_rate": 5.627925602803413e-06,
+      "loss": 1.3412,
+      "step": 2913
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.1140283527243353,
+      "learning_rate": 5.621378848026022e-06,
+      "loss": 1.3535,
+      "step": 2914
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4303826758080667,
+      "learning_rate": 5.614834414250541e-06,
+      "loss": 0.7137,
+      "step": 2915
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.000628362703011,
+      "learning_rate": 5.608292304946004e-06,
+      "loss": 1.3149,
+      "step": 2916
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4494728584015266,
+      "learning_rate": 5.601752523580241e-06,
+      "loss": 0.8383,
+      "step": 2917
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4390254050691604,
+      "learning_rate": 5.595215073619822e-06,
+      "loss": 0.7899,
+      "step": 2918
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9545765811444444,
+      "learning_rate": 5.588679958530095e-06,
+      "loss": 1.3586,
+      "step": 2919
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.384959189949584,
+      "learning_rate": 5.582147181775169e-06,
+      "loss": 0.7213,
+      "step": 2920
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5323279659827738,
+      "learning_rate": 5.575616746817911e-06,
+      "loss": 0.768,
+      "step": 2921
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0073733470506916,
+      "learning_rate": 5.569088657119949e-06,
+      "loss": 1.317,
+      "step": 2922
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3708111362505926,
+      "learning_rate": 5.562562916141665e-06,
+      "loss": 0.7733,
+      "step": 2923
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.512704850093893,
+      "learning_rate": 5.5560395273421985e-06,
+      "loss": 0.701,
+      "step": 2924
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4229133589544785,
+      "learning_rate": 5.54951849417944e-06,
+      "loss": 0.7195,
+      "step": 2925
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4013853120380677,
+      "learning_rate": 5.542999820110038e-06,
+      "loss": 0.8007,
+      "step": 2926
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4361269351973682,
+      "learning_rate": 5.536483508589377e-06,
+      "loss": 0.7268,
+      "step": 2927
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4086823670401958,
+      "learning_rate": 5.5299695630716e-06,
+      "loss": 0.7276,
+      "step": 2928
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.271095172664951,
+      "learning_rate": 5.523457987009595e-06,
+      "loss": 0.6646,
+      "step": 2929
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.438800613773805,
+      "learning_rate": 5.516948783854989e-06,
+      "loss": 0.6965,
+      "step": 2930
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3954741653418359,
+      "learning_rate": 5.510441957058158e-06,
+      "loss": 0.7074,
+      "step": 2931
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.015501271520496,
+      "learning_rate": 5.503937510068211e-06,
+      "loss": 1.2737,
+      "step": 2932
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.987450235426671,
+      "learning_rate": 5.497435446333003e-06,
+      "loss": 1.329,
+      "step": 2933
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5357321531096857,
+      "learning_rate": 5.4909357692991185e-06,
+      "loss": 0.8126,
+      "step": 2934
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3157733538749936,
+      "learning_rate": 5.484438482411882e-06,
+      "loss": 0.6894,
+      "step": 2935
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9695447432085171,
+      "learning_rate": 5.477943589115347e-06,
+      "loss": 1.3326,
+      "step": 2936
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5019995047850194,
+      "learning_rate": 5.471451092852307e-06,
+      "loss": 0.7785,
+      "step": 2937
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3342735535631085,
+      "learning_rate": 5.46496099706427e-06,
+      "loss": 0.6708,
+      "step": 2938
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4947776769317271,
+      "learning_rate": 5.458473305191483e-06,
+      "loss": 0.8277,
+      "step": 2939
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3718135479652047,
+      "learning_rate": 5.451988020672916e-06,
+      "loss": 0.6855,
+      "step": 2940
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4494324094294686,
+      "learning_rate": 5.445505146946262e-06,
+      "loss": 0.6943,
+      "step": 2941
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.467450408403217,
+      "learning_rate": 5.439024687447936e-06,
+      "loss": 0.8019,
+      "step": 2942
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4048652622334876,
+      "learning_rate": 5.432546645613077e-06,
+      "loss": 0.7068,
+      "step": 2943
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4447065622297648,
+      "learning_rate": 5.426071024875537e-06,
+      "loss": 0.7315,
+      "step": 2944
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3708067471455463,
+      "learning_rate": 5.419597828667886e-06,
+      "loss": 0.6425,
+      "step": 2945
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4519805218943063,
+      "learning_rate": 5.413127060421416e-06,
+      "loss": 0.7059,
+      "step": 2946
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9481199476976807,
+      "learning_rate": 5.406658723566114e-06,
+      "loss": 1.3133,
+      "step": 2947
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5547809574368652,
+      "learning_rate": 5.400192821530703e-06,
+      "loss": 0.7033,
+      "step": 2948
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0536470026849845,
+      "learning_rate": 5.393729357742594e-06,
+      "loss": 1.3077,
+      "step": 2949
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4650221488727155,
+      "learning_rate": 5.387268335627917e-06,
+      "loss": 0.8182,
+      "step": 2950
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9426492444977728,
+      "learning_rate": 5.380809758611503e-06,
+      "loss": 1.3496,
+      "step": 2951
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4687117920305006,
+      "learning_rate": 5.37435363011689e-06,
+      "loss": 0.6996,
+      "step": 2952
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3811859601696612,
+      "learning_rate": 5.367899953566315e-06,
+      "loss": 0.7059,
+      "step": 2953
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.967222179804127,
+      "learning_rate": 5.3614487323807195e-06,
+      "loss": 1.3293,
+      "step": 2954
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.385072512446408,
+      "learning_rate": 5.35499996997974e-06,
+      "loss": 0.7336,
+      "step": 2955
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9537429436447207,
+      "learning_rate": 5.348553669781708e-06,
+      "loss": 1.2902,
+      "step": 2956
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.00853716978786,
+      "learning_rate": 5.34210983520366e-06,
+      "loss": 1.2717,
+      "step": 2957
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4043635321537542,
+      "learning_rate": 5.335668469661305e-06,
+      "loss": 0.7052,
+      "step": 2958
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5290869362833825,
+      "learning_rate": 5.32922957656907e-06,
+      "loss": 0.8432,
+      "step": 2959
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3833964008973727,
+      "learning_rate": 5.322793159340049e-06,
+      "loss": 0.7576,
+      "step": 2960
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4508499393788463,
+      "learning_rate": 5.316359221386033e-06,
+      "loss": 0.7973,
+      "step": 2961
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.479033372398701,
+      "learning_rate": 5.3099277661174995e-06,
+      "loss": 0.8298,
+      "step": 2962
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4544450979310488,
+      "learning_rate": 5.3034987969436095e-06,
+      "loss": 0.7611,
+      "step": 2963
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3994528408423181,
+      "learning_rate": 5.2970723172722024e-06,
+      "loss": 0.7533,
+      "step": 2964
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.475271555887298,
+      "learning_rate": 5.2906483305098024e-06,
+      "loss": 0.7738,
+      "step": 2965
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4118296367050283,
+      "learning_rate": 5.284226840061611e-06,
+      "loss": 0.7325,
+      "step": 2966
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9721213014658223,
+      "learning_rate": 5.2778078493315055e-06,
+      "loss": 1.371,
+      "step": 2967
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4460437757806848,
+      "learning_rate": 5.271391361722043e-06,
+      "loss": 0.7624,
+      "step": 2968
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.0331928398458503,
+      "learning_rate": 5.26497738063444e-06,
+      "loss": 1.3338,
+      "step": 2969
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3566481536170063,
+      "learning_rate": 5.2585659094686005e-06,
+      "loss": 0.7747,
+      "step": 2970
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.499782974243569,
+      "learning_rate": 5.252156951623089e-06,
+      "loss": 0.7074,
+      "step": 2971
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.06992553660031,
+      "learning_rate": 5.24575051049514e-06,
+      "loss": 1.3883,
+      "step": 2972
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9351431896582,
+      "learning_rate": 5.239346589480656e-06,
+      "loss": 1.2816,
+      "step": 2973
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.0239207703076136,
+      "learning_rate": 5.232945191974201e-06,
+      "loss": 1.3854,
+      "step": 2974
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4730889188239766,
+      "learning_rate": 5.226546321369003e-06,
+      "loss": 0.7219,
+      "step": 2975
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4501939164859126,
+      "learning_rate": 5.220149981056947e-06,
+      "loss": 0.7163,
+      "step": 2976
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.012648407931203,
+      "learning_rate": 5.213756174428585e-06,
+      "loss": 1.2925,
+      "step": 2977
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3917034608271523,
+      "learning_rate": 5.207364904873108e-06,
+      "loss": 0.7199,
+      "step": 2978
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.937447147002946,
+      "learning_rate": 5.200976175778391e-06,
+      "loss": 1.4022,
+      "step": 2979
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4115510081054263,
+      "learning_rate": 5.194589990530933e-06,
+      "loss": 0.8263,
+      "step": 2980
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4933033167616867,
+      "learning_rate": 5.188206352515902e-06,
+      "loss": 0.7343,
+      "step": 2981
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4436545001462886,
+      "learning_rate": 5.1818252651171105e-06,
+      "loss": 0.7063,
+      "step": 2982
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3240424339166976,
+      "learning_rate": 5.175446731717022e-06,
+      "loss": 0.7289,
+      "step": 2983
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5982952348537185,
+      "learning_rate": 5.1690707556967414e-06,
+      "loss": 0.7694,
+      "step": 2984
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3232104277711625,
+      "learning_rate": 5.16269734043602e-06,
+      "loss": 0.7257,
+      "step": 2985
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3673617394978719,
+      "learning_rate": 5.156326489313255e-06,
+      "loss": 0.7365,
+      "step": 2986
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.102054556137277,
+      "learning_rate": 5.149958205705479e-06,
+      "loss": 1.3433,
+      "step": 2987
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.475225602397934,
+      "learning_rate": 5.143592492988369e-06,
+      "loss": 0.7875,
+      "step": 2988
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.300535907456855,
+      "learning_rate": 5.1372293545362295e-06,
+      "loss": 0.764,
+      "step": 2989
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.025619639081872,
+      "learning_rate": 5.130868793722019e-06,
+      "loss": 1.3577,
+      "step": 2990
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4033219853544592,
+      "learning_rate": 5.124510813917307e-06,
+      "loss": 0.7368,
+      "step": 2991
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9091841788078319,
+      "learning_rate": 5.118155418492312e-06,
+      "loss": 1.3463,
+      "step": 2992
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4067263437685062,
+      "learning_rate": 5.111802610815872e-06,
+      "loss": 0.7249,
+      "step": 2993
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9082331069751823,
+      "learning_rate": 5.105452394255464e-06,
+      "loss": 1.2833,
+      "step": 2994
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.032778864947353,
+      "learning_rate": 5.099104772177179e-06,
+      "loss": 1.344,
+      "step": 2995
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3856873905019067,
+      "learning_rate": 5.0927597479457415e-06,
+      "loss": 0.8199,
+      "step": 2996
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.492350477566081,
+      "learning_rate": 5.086417324924497e-06,
+      "loss": 0.74,
+      "step": 2997
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3692625477116083,
+      "learning_rate": 5.080077506475412e-06,
+      "loss": 0.7396,
+      "step": 2998
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.068783460858655,
+      "learning_rate": 5.0737402959590755e-06,
+      "loss": 1.3301,
+      "step": 2999
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4051459578285694,
+      "learning_rate": 5.067405696734678e-06,
+      "loss": 0.7834,
+      "step": 3000
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3279252540084927,
+      "learning_rate": 5.061073712160054e-06,
+      "loss": 0.6746,
+      "step": 3001
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4682156713754657,
+      "learning_rate": 5.054744345591626e-06,
+      "loss": 0.6565,
+      "step": 3002
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5164942727582742,
+      "learning_rate": 5.048417600384442e-06,
+      "loss": 0.736,
+      "step": 3003
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4598014549965217,
+      "learning_rate": 5.042093479892159e-06,
+      "loss": 0.7897,
+      "step": 3004
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4069628964248762,
+      "learning_rate": 5.035771987467038e-06,
+      "loss": 0.7285,
+      "step": 3005
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.406696149537592,
+      "learning_rate": 5.029453126459955e-06,
+      "loss": 0.7487,
+      "step": 3006
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5428828909770675,
+      "learning_rate": 5.023136900220383e-06,
+      "loss": 0.7584,
+      "step": 3007
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.465120085827652,
+      "learning_rate": 5.016823312096408e-06,
+      "loss": 0.7009,
+      "step": 3008
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4257278654378251,
+      "learning_rate": 5.010512365434698e-06,
+      "loss": 0.7047,
+      "step": 3009
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4417596264333077,
+      "learning_rate": 5.00420406358055e-06,
+      "loss": 0.681,
+      "step": 3010
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0045554148902993,
+      "learning_rate": 4.9978984098778324e-06,
+      "loss": 1.3236,
+      "step": 3011
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4519964849651967,
+      "learning_rate": 4.9915954076690245e-06,
+      "loss": 0.6738,
+      "step": 3012
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9115126710024448,
+      "learning_rate": 4.985295060295197e-06,
+      "loss": 1.2833,
+      "step": 3013
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9373557900234357,
+      "learning_rate": 4.97899737109601e-06,
+      "loss": 1.2813,
+      "step": 3014
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4192870380032256,
+      "learning_rate": 4.972702343409721e-06,
+      "loss": 0.7736,
+      "step": 3015
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.012439081678838,
+      "learning_rate": 4.96640998057317e-06,
+      "loss": 1.3867,
+      "step": 3016
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3314158373640792,
+      "learning_rate": 4.96012028592179e-06,
+      "loss": 0.77,
+      "step": 3017
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.392898723210078,
+      "learning_rate": 4.9538332627895944e-06,
+      "loss": 0.7694,
+      "step": 3018
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.1254526831260816,
+      "learning_rate": 4.947548914509188e-06,
+      "loss": 1.3059,
+      "step": 3019
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4230215841944007,
+      "learning_rate": 4.9412672444117425e-06,
+      "loss": 0.7528,
+      "step": 3020
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5760753190839745,
+      "learning_rate": 4.934988255827036e-06,
+      "loss": 0.794,
+      "step": 3021
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0459366078294816,
+      "learning_rate": 4.928711952083395e-06,
+      "loss": 1.2758,
+      "step": 3022
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5301555127701918,
+      "learning_rate": 4.9224383365077435e-06,
+      "loss": 0.799,
+      "step": 3023
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4983886276995437,
+      "learning_rate": 4.9161674124255735e-06,
+      "loss": 0.7382,
+      "step": 3024
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0070889554049725,
+      "learning_rate": 4.909899183160952e-06,
+      "loss": 1.3917,
+      "step": 3025
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4339788203456088,
+      "learning_rate": 4.903633652036517e-06,
+      "loss": 0.7569,
+      "step": 3026
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3327521717906679,
+      "learning_rate": 4.897370822373475e-06,
+      "loss": 0.6756,
+      "step": 3027
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5166645982580365,
+      "learning_rate": 4.891110697491603e-06,
+      "loss": 0.7631,
+      "step": 3028
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.027212148245818,
+      "learning_rate": 4.884853280709242e-06,
+      "loss": 1.368,
+      "step": 3029
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9935214180276708,
+      "learning_rate": 4.8785985753433005e-06,
+      "loss": 1.2738,
+      "step": 3030
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.419290635370468,
+      "learning_rate": 4.872346584709239e-06,
+      "loss": 0.7845,
+      "step": 3031
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4827042160061275,
+      "learning_rate": 4.866097312121102e-06,
+      "loss": 0.7144,
+      "step": 3032
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4803534653305095,
+      "learning_rate": 4.859850760891466e-06,
+      "loss": 0.7214,
+      "step": 3033
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.421315245591924,
+      "learning_rate": 4.8536069343314825e-06,
+      "loss": 0.7651,
+      "step": 3034
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3470121992246742,
+      "learning_rate": 4.847365835750856e-06,
+      "loss": 0.7483,
+      "step": 3035
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3981828507156382,
+      "learning_rate": 4.841127468457839e-06,
+      "loss": 0.6726,
+      "step": 3036
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5563387057556937,
+      "learning_rate": 4.8348918357592424e-06,
+      "loss": 0.814,
+      "step": 3037
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.975578032075408,
+      "learning_rate": 4.828658940960426e-06,
+      "loss": 1.4219,
+      "step": 3038
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6233994832693384,
+      "learning_rate": 4.8224287873653e-06,
+      "loss": 0.8328,
+      "step": 3039
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4560459860249115,
+      "learning_rate": 4.8162013782763075e-06,
+      "loss": 0.7646,
+      "step": 3040
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.423056672114662,
+      "learning_rate": 4.809976716994465e-06,
+      "loss": 0.7165,
+      "step": 3041
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4154189112035096,
+      "learning_rate": 4.8037548068193005e-06,
+      "loss": 0.7188,
+      "step": 3042
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8783199093472216,
+      "learning_rate": 4.7975356510489145e-06,
+      "loss": 1.331,
+      "step": 3043
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5444760306310779,
+      "learning_rate": 4.79131925297992e-06,
+      "loss": 0.7654,
+      "step": 3044
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.570482297748785,
+      "learning_rate": 4.785105615907484e-06,
+      "loss": 0.7198,
+      "step": 3045
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4777353280486358,
+      "learning_rate": 4.778894743125306e-06,
+      "loss": 0.7357,
+      "step": 3046
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4920571024659415,
+      "learning_rate": 4.772686637925622e-06,
+      "loss": 0.7279,
+      "step": 3047
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9165654845443556,
+      "learning_rate": 4.766481303599198e-06,
+      "loss": 1.335,
+      "step": 3048
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.2623494114837746,
+      "learning_rate": 4.760278743435335e-06,
+      "loss": 1.2736,
+      "step": 3049
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4143417564190288,
+      "learning_rate": 4.754078960721862e-06,
+      "loss": 0.7467,
+      "step": 3050
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4108713814941982,
+      "learning_rate": 4.747881958745124e-06,
+      "loss": 0.7452,
+      "step": 3051
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3566280432254358,
+      "learning_rate": 4.74168774079002e-06,
+      "loss": 0.7919,
+      "step": 3052
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9556367632937754,
+      "learning_rate": 4.735496310139943e-06,
+      "loss": 1.2991,
+      "step": 3053
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.372966503171842,
+      "learning_rate": 4.729307670076826e-06,
+      "loss": 0.7675,
+      "step": 3054
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5046007664592724,
+      "learning_rate": 4.723121823881119e-06,
+      "loss": 0.6812,
+      "step": 3055
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.991562871666889,
+      "learning_rate": 4.71693877483179e-06,
+      "loss": 1.364,
+      "step": 3056
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8689023786041612,
+      "learning_rate": 4.710758526206325e-06,
+      "loss": 1.3367,
+      "step": 3057
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4395603021785406,
+      "learning_rate": 4.704581081280725e-06,
+      "loss": 0.6981,
+      "step": 3058
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.429745787667873,
+      "learning_rate": 4.698406443329504e-06,
+      "loss": 0.6664,
+      "step": 3059
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4531706403696438,
+      "learning_rate": 4.692234615625693e-06,
+      "loss": 0.7389,
+      "step": 3060
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8969552740538915,
+      "learning_rate": 4.686065601440829e-06,
+      "loss": 1.2872,
+      "step": 3061
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3623956210874146,
+      "learning_rate": 4.679899404044949e-06,
+      "loss": 0.6922,
+      "step": 3062
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3217877464875465,
+      "learning_rate": 4.6737360267066235e-06,
+      "loss": 0.6568,
+      "step": 3063
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0526904078368395,
+      "learning_rate": 4.667575472692895e-06,
+      "loss": 1.2652,
+      "step": 3064
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0038476284378977,
+      "learning_rate": 4.661417745269332e-06,
+      "loss": 1.3394,
+      "step": 3065
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9588124536733567,
+      "learning_rate": 4.655262847699997e-06,
+      "loss": 1.3762,
+      "step": 3066
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2997131833783746,
+      "learning_rate": 4.649110783247452e-06,
+      "loss": 0.7321,
+      "step": 3067
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4658754059116306,
+      "learning_rate": 4.642961555172762e-06,
+      "loss": 0.7763,
+      "step": 3068
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6034538505888176,
+      "learning_rate": 4.636815166735483e-06,
+      "loss": 0.7064,
+      "step": 3069
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4494447040055796,
+      "learning_rate": 4.6306716211936705e-06,
+      "loss": 0.7274,
+      "step": 3070
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4367430423642378,
+      "learning_rate": 4.6245309218038616e-06,
+      "loss": 0.7332,
+      "step": 3071
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5789356615555818,
+      "learning_rate": 4.618393071821107e-06,
+      "loss": 0.761,
+      "step": 3072
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3683549434812412,
+      "learning_rate": 4.6122580744989185e-06,
+      "loss": 0.7558,
+      "step": 3073
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.983553175228309,
+      "learning_rate": 4.606125933089327e-06,
+      "loss": 1.3261,
+      "step": 3074
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4039399938950372,
+      "learning_rate": 4.599996650842822e-06,
+      "loss": 0.7062,
+      "step": 3075
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3711718199708949,
+      "learning_rate": 4.59387023100839e-06,
+      "loss": 0.7306,
+      "step": 3076
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5379951264883969,
+      "learning_rate": 4.587746676833502e-06,
+      "loss": 0.6746,
+      "step": 3077
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4271661846165604,
+      "learning_rate": 4.581625991564107e-06,
+      "loss": 0.8244,
+      "step": 3078
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9298405712808009,
+      "learning_rate": 4.575508178444631e-06,
+      "loss": 1.3473,
+      "step": 3079
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3473050134944964,
+      "learning_rate": 4.5693932407179806e-06,
+      "loss": 0.7804,
+      "step": 3080
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4242360447415918,
+      "learning_rate": 4.5632811816255425e-06,
+      "loss": 0.7181,
+      "step": 3081
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4583694912905223,
+      "learning_rate": 4.55717200440716e-06,
+      "loss": 0.6882,
+      "step": 3082
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0320031294582415,
+      "learning_rate": 4.551065712301177e-06,
+      "loss": 1.2527,
+      "step": 3083
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0508991879437874,
+      "learning_rate": 4.544962308544376e-06,
+      "loss": 1.3835,
+      "step": 3084
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4258824092102746,
+      "learning_rate": 4.53886179637204e-06,
+      "loss": 0.6954,
+      "step": 3085
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5649890427699316,
+      "learning_rate": 4.532764179017894e-06,
+      "loss": 0.6871,
+      "step": 3086
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3613995440854614,
+      "learning_rate": 4.526669459714142e-06,
+      "loss": 0.7297,
+      "step": 3087
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.375982865063245,
+      "learning_rate": 4.5205776416914475e-06,
+      "loss": 0.718,
+      "step": 3088
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.381886377160224,
+      "learning_rate": 4.514488728178936e-06,
+      "loss": 0.7488,
+      "step": 3089
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3922008755608983,
+      "learning_rate": 4.508402722404196e-06,
+      "loss": 0.7369,
+      "step": 3090
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9660854301441553,
+      "learning_rate": 4.502319627593272e-06,
+      "loss": 1.3167,
+      "step": 3091
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4780820061866384,
+      "learning_rate": 4.496239446970671e-06,
+      "loss": 0.7785,
+      "step": 3092
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8811347114858101,
+      "learning_rate": 4.49016218375934e-06,
+      "loss": 1.3233,
+      "step": 3093
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3601466489307432,
+      "learning_rate": 4.4840878411807045e-06,
+      "loss": 0.7199,
+      "step": 3094
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8934716639603228,
+      "learning_rate": 4.478016422454615e-06,
+      "loss": 1.3677,
+      "step": 3095
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.931761859697252,
+      "learning_rate": 4.471947930799391e-06,
+      "loss": 1.3178,
+      "step": 3096
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4998000681602113,
+      "learning_rate": 4.465882369431792e-06,
+      "loss": 0.6937,
+      "step": 3097
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9581480298166818,
+      "learning_rate": 4.459819741567027e-06,
+      "loss": 1.3154,
+      "step": 3098
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4241424819986934,
+      "learning_rate": 4.453760050418752e-06,
+      "loss": 0.653,
+      "step": 3099
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4705226702107481,
+      "learning_rate": 4.447703299199059e-06,
+      "loss": 0.7426,
+      "step": 3100
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3487914271179373,
+      "learning_rate": 4.441649491118494e-06,
+      "loss": 0.6837,
+      "step": 3101
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4785342369955878,
+      "learning_rate": 4.435598629386021e-06,
+      "loss": 0.7499,
+      "step": 3102
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4455614182244394,
+      "learning_rate": 4.429550717209075e-06,
+      "loss": 0.7312,
+      "step": 3103
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4175076637970505,
+      "learning_rate": 4.42350575779349e-06,
+      "loss": 0.7874,
+      "step": 3104
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0220191132896157,
+      "learning_rate": 4.417463754343572e-06,
+      "loss": 1.3682,
+      "step": 3105
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.046631287597472,
+      "learning_rate": 4.411424710062029e-06,
+      "loss": 1.2275,
+      "step": 3106
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.416508258071913,
+      "learning_rate": 4.405388628150018e-06,
+      "loss": 0.7401,
+      "step": 3107
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4915988086694743,
+      "learning_rate": 4.399355511807121e-06,
+      "loss": 0.7492,
+      "step": 3108
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.360642737036299,
+      "learning_rate": 4.393325364231347e-06,
+      "loss": 0.7556,
+      "step": 3109
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3880140790014925,
+      "learning_rate": 4.387298188619132e-06,
+      "loss": 0.7005,
+      "step": 3110
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.47519177720547,
+      "learning_rate": 4.38127398816534e-06,
+      "loss": 0.753,
+      "step": 3111
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4055226601481048,
+      "learning_rate": 4.375252766063256e-06,
+      "loss": 0.7465,
+      "step": 3112
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4225078073052617,
+      "learning_rate": 4.369234525504574e-06,
+      "loss": 0.679,
+      "step": 3113
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.304955965181913,
+      "learning_rate": 4.363219269679435e-06,
+      "loss": 0.7411,
+      "step": 3114
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8505809922483574,
+      "learning_rate": 4.3572070017763665e-06,
+      "loss": 1.283,
+      "step": 3115
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3644632954991165,
+      "learning_rate": 4.351197724982342e-06,
+      "loss": 0.7042,
+      "step": 3116
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3541020519381521,
+      "learning_rate": 4.345191442482724e-06,
+      "loss": 0.8305,
+      "step": 3117
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.224283332414356,
+      "learning_rate": 4.3391881574613e-06,
+      "loss": 1.3192,
+      "step": 3118
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3404599328238995,
+      "learning_rate": 4.333187873100271e-06,
+      "loss": 0.7182,
+      "step": 3119
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5022193636406609,
+      "learning_rate": 4.327190592580242e-06,
+      "loss": 0.7124,
+      "step": 3120
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9736316677113421,
+      "learning_rate": 4.321196319080225e-06,
+      "loss": 1.3108,
+      "step": 3121
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3663848579077795,
+      "learning_rate": 4.315205055777644e-06,
+      "loss": 0.6438,
+      "step": 3122
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4799722273729954,
+      "learning_rate": 4.3092168058483256e-06,
+      "loss": 0.762,
+      "step": 3123
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4259769733014118,
+      "learning_rate": 4.303231572466485e-06,
+      "loss": 0.7041,
+      "step": 3124
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3707881983216406,
+      "learning_rate": 4.297249358804769e-06,
+      "loss": 0.7653,
+      "step": 3125
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8558678021192239,
+      "learning_rate": 4.291270168034187e-06,
+      "loss": 1.4157,
+      "step": 3126
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5489245557319011,
+      "learning_rate": 4.285294003324182e-06,
+      "loss": 0.7864,
+      "step": 3127
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4614873971526956,
+      "learning_rate": 4.2793208678425626e-06,
+      "loss": 0.7815,
+      "step": 3128
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4255987215882782,
+      "learning_rate": 4.2733507647555485e-06,
+      "loss": 0.7819,
+      "step": 3129
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5659690868568896,
+      "learning_rate": 4.267383697227747e-06,
+      "loss": 0.7394,
+      "step": 3130
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4435909550262558,
+      "learning_rate": 4.26141966842216e-06,
+      "loss": 0.6823,
+      "step": 3131
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3372618924279163,
+      "learning_rate": 4.255458681500176e-06,
+      "loss": 0.8371,
+      "step": 3132
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6086776995543126,
+      "learning_rate": 4.249500739621564e-06,
+      "loss": 0.7537,
+      "step": 3133
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.388617854519949,
+      "learning_rate": 4.243545845944497e-06,
+      "loss": 1.2579,
+      "step": 3134
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.488703886384289,
+      "learning_rate": 4.237594003625508e-06,
+      "loss": 0.8021,
+      "step": 3135
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8876651444462804,
+      "learning_rate": 4.231645215819541e-06,
+      "loss": 1.3322,
+      "step": 3136
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.069896403559514,
+      "learning_rate": 4.225699485679893e-06,
+      "loss": 1.2467,
+      "step": 3137
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5111970530951802,
+      "learning_rate": 4.219756816358257e-06,
+      "loss": 0.7132,
+      "step": 3138
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.449134648765044,
+      "learning_rate": 4.213817211004699e-06,
+      "loss": 0.6726,
+      "step": 3139
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9194369576496388,
+      "learning_rate": 4.207880672767663e-06,
+      "loss": 1.3307,
+      "step": 3140
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4691157235525085,
+      "learning_rate": 4.201947204793964e-06,
+      "loss": 0.7603,
+      "step": 3141
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.1668908015778503,
+      "learning_rate": 4.1960168102287895e-06,
+      "loss": 1.4196,
+      "step": 3142
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4585401280771753,
+      "learning_rate": 4.190089492215704e-06,
+      "loss": 0.6641,
+      "step": 3143
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.30216182098523,
+      "learning_rate": 4.184165253896626e-06,
+      "loss": 0.7476,
+      "step": 3144
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4517081007074308,
+      "learning_rate": 4.178244098411865e-06,
+      "loss": 0.7095,
+      "step": 3145
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.142896608598938,
+      "learning_rate": 4.172326028900072e-06,
+      "loss": 1.2818,
+      "step": 3146
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.400195781873027,
+      "learning_rate": 4.166411048498285e-06,
+      "loss": 0.689,
+      "step": 3147
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6269218278933772,
+      "learning_rate": 4.160499160341882e-06,
+      "loss": 0.847,
+      "step": 3148
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.370511396621948,
+      "learning_rate": 4.154590367564619e-06,
+      "loss": 0.7098,
+      "step": 3149
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4415787761294863,
+      "learning_rate": 4.148684673298603e-06,
+      "loss": 0.7485,
+      "step": 3150
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4734769767041462,
+      "learning_rate": 4.1427820806743015e-06,
+      "loss": 0.7875,
+      "step": 3151
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3089010419355993,
+      "learning_rate": 4.136882592820537e-06,
+      "loss": 0.6995,
+      "step": 3152
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3919430122093341,
+      "learning_rate": 4.130986212864486e-06,
+      "loss": 0.6803,
+      "step": 3153
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.40823885916791,
+      "learning_rate": 4.125092943931682e-06,
+      "loss": 0.6756,
+      "step": 3154
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4054672032325757,
+      "learning_rate": 4.119202789145994e-06,
+      "loss": 0.7728,
+      "step": 3155
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5020930614870969,
+      "learning_rate": 4.1133157516296675e-06,
+      "loss": 0.7357,
+      "step": 3156
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.484673250660756,
+      "learning_rate": 4.107431834503263e-06,
+      "loss": 0.7015,
+      "step": 3157
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3536832356879092,
+      "learning_rate": 4.101551040885719e-06,
+      "loss": 0.6841,
+      "step": 3158
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4183130440197473,
+      "learning_rate": 4.095673373894293e-06,
+      "loss": 0.7774,
+      "step": 3159
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3246067992725448,
+      "learning_rate": 4.089798836644596e-06,
+      "loss": 0.7784,
+      "step": 3160
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.50761646965602,
+      "learning_rate": 4.083927432250583e-06,
+      "loss": 0.7313,
+      "step": 3161
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4553784378529069,
+      "learning_rate": 4.078059163824543e-06,
+      "loss": 0.7775,
+      "step": 3162
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4088024947140836,
+      "learning_rate": 4.072194034477105e-06,
+      "loss": 0.7099,
+      "step": 3163
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4159318818500235,
+      "learning_rate": 4.066332047317226e-06,
+      "loss": 0.6737,
+      "step": 3164
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.358298930822538,
+      "learning_rate": 4.060473205452218e-06,
+      "loss": 0.7218,
+      "step": 3165
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4220604456723134,
+      "learning_rate": 4.054617511987696e-06,
+      "loss": 0.6903,
+      "step": 3166
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4496850923381064,
+      "learning_rate": 4.04876497002764e-06,
+      "loss": 0.8234,
+      "step": 3167
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8703480599417803,
+      "learning_rate": 4.042915582674325e-06,
+      "loss": 1.3229,
+      "step": 3168
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5398531221236005,
+      "learning_rate": 4.0370693530283866e-06,
+      "loss": 0.8339,
+      "step": 3169
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.364920565149042,
+      "learning_rate": 4.031226284188759e-06,
+      "loss": 0.7184,
+      "step": 3170
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.078054164063379,
+      "learning_rate": 4.0253863792527185e-06,
+      "loss": 1.2971,
+      "step": 3171
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4348020426903925,
+      "learning_rate": 4.019549641315854e-06,
+      "loss": 0.8273,
+      "step": 3172
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4559406413490343,
+      "learning_rate": 4.013716073472084e-06,
+      "loss": 0.7635,
+      "step": 3173
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3900739936320525,
+      "learning_rate": 4.007885678813645e-06,
+      "loss": 0.7343,
+      "step": 3174
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5395040785216787,
+      "learning_rate": 4.002058460431079e-06,
+      "loss": 0.702,
+      "step": 3175
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4793316316364629,
+      "learning_rate": 3.996234421413267e-06,
+      "loss": 0.7377,
+      "step": 3176
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.068645012786346,
+      "learning_rate": 3.99041356484738e-06,
+      "loss": 1.2966,
+      "step": 3177
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3699198639887673,
+      "learning_rate": 3.984595893818927e-06,
+      "loss": 0.6417,
+      "step": 3178
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4741550358186173,
+      "learning_rate": 3.978781411411705e-06,
+      "loss": 0.7429,
+      "step": 3179
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4961002525632445,
+      "learning_rate": 3.9729701207078366e-06,
+      "loss": 0.6509,
+      "step": 3180
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.379876974144094,
+      "learning_rate": 3.967162024787746e-06,
+      "loss": 0.8178,
+      "step": 3181
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5523815175163493,
+      "learning_rate": 3.9613571267301665e-06,
+      "loss": 0.7196,
+      "step": 3182
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3060959099969798,
+      "learning_rate": 3.9555554296121345e-06,
+      "loss": 0.6693,
+      "step": 3183
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4151242915522686,
+      "learning_rate": 3.949756936508991e-06,
+      "loss": 0.7902,
+      "step": 3184
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3033292672738566,
+      "learning_rate": 3.943961650494379e-06,
+      "loss": 0.7884,
+      "step": 3185
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3396340497614787,
+      "learning_rate": 3.938169574640233e-06,
+      "loss": 0.7345,
+      "step": 3186
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4990742327964313,
+      "learning_rate": 3.9323807120168044e-06,
+      "loss": 0.7638,
+      "step": 3187
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9110803793090532,
+      "learning_rate": 3.926595065692619e-06,
+      "loss": 1.3557,
+      "step": 3188
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4443304721700947,
+      "learning_rate": 3.920812638734522e-06,
+      "loss": 0.6821,
+      "step": 3189
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4532471892574117,
+      "learning_rate": 3.915033434207627e-06,
+      "loss": 0.7457,
+      "step": 3190
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0376233870388796,
+      "learning_rate": 3.909257455175357e-06,
+      "loss": 1.3728,
+      "step": 3191
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3963169402125655,
+      "learning_rate": 3.903484704699418e-06,
+      "loss": 0.7139,
+      "step": 3192
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4305997879542909,
+      "learning_rate": 3.897715185839809e-06,
+      "loss": 0.7724,
+      "step": 3193
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9308249279596061,
+      "learning_rate": 3.891948901654815e-06,
+      "loss": 1.3234,
+      "step": 3194
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5367122919498604,
+      "learning_rate": 3.8861858552009935e-06,
+      "loss": 0.7114,
+      "step": 3195
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5190928228215952,
+      "learning_rate": 3.880426049533212e-06,
+      "loss": 0.6952,
+      "step": 3196
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9628112377480573,
+      "learning_rate": 3.874669487704592e-06,
+      "loss": 1.3605,
+      "step": 3197
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4669005048220105,
+      "learning_rate": 3.868916172766562e-06,
+      "loss": 0.741,
+      "step": 3198
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4429253867024314,
+      "learning_rate": 3.863166107768801e-06,
+      "loss": 0.7646,
+      "step": 3199
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4360935278259155,
+      "learning_rate": 3.857419295759296e-06,
+      "loss": 0.7544,
+      "step": 3200
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5318687190119649,
+      "learning_rate": 3.851675739784282e-06,
+      "loss": 0.739,
+      "step": 3201
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4335127793092164,
+      "learning_rate": 3.8459354428882855e-06,
+      "loss": 0.7509,
+      "step": 3202
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9454411638429103,
+      "learning_rate": 3.840198408114097e-06,
+      "loss": 1.3189,
+      "step": 3203
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.493604520613072,
+      "learning_rate": 3.834464638502784e-06,
+      "loss": 0.813,
+      "step": 3204
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9759315944653197,
+      "learning_rate": 3.828734137093683e-06,
+      "loss": 1.3652,
+      "step": 3205
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4850111946055655,
+      "learning_rate": 3.823006906924381e-06,
+      "loss": 0.8061,
+      "step": 3206
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4477863747474848,
+      "learning_rate": 3.817282951030764e-06,
+      "loss": 0.8021,
+      "step": 3207
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9257801441769409,
+      "learning_rate": 3.811562272446948e-06,
+      "loss": 1.381,
+      "step": 3208
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4074031701893792,
+      "learning_rate": 3.80584487420534e-06,
+      "loss": 0.6994,
+      "step": 3209
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.394072700790333,
+      "learning_rate": 3.8001307593365822e-06,
+      "loss": 0.659,
+      "step": 3210
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4486976205697246,
+      "learning_rate": 3.7944199308696052e-06,
+      "loss": 0.8613,
+      "step": 3211
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3785914063411528,
+      "learning_rate": 3.7887123918315682e-06,
+      "loss": 0.6086,
+      "step": 3212
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9416374505638854,
+      "learning_rate": 3.7830081452479073e-06,
+      "loss": 1.3224,
+      "step": 3213
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3921488696974138,
+      "learning_rate": 3.7773071941423044e-06,
+      "loss": 0.7064,
+      "step": 3214
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5196429476765498,
+      "learning_rate": 3.771609541536698e-06,
+      "loss": 0.7531,
+      "step": 3215
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9474709648794801,
+      "learning_rate": 3.76591519045128e-06,
+      "loss": 1.332,
+      "step": 3216
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4515281912279545,
+      "learning_rate": 3.7602241439044796e-06,
+      "loss": 0.8079,
+      "step": 3217
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4186095095932032,
+      "learning_rate": 3.754536404912996e-06,
+      "loss": 0.7221,
+      "step": 3218
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9447741072992593,
+      "learning_rate": 3.7488519764917497e-06,
+      "loss": 1.3043,
+      "step": 3219
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9795766864955413,
+      "learning_rate": 3.743170861653934e-06,
+      "loss": 1.3006,
+      "step": 3220
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.404686085434942,
+      "learning_rate": 3.737493063410962e-06,
+      "loss": 0.7056,
+      "step": 3221
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4120682613798985,
+      "learning_rate": 3.7318185847724996e-06,
+      "loss": 0.6719,
+      "step": 3222
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.339243572527197,
+      "learning_rate": 3.7261474287464527e-06,
+      "loss": 0.7347,
+      "step": 3223
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3663323295380916,
+      "learning_rate": 3.720479598338964e-06,
+      "loss": 0.7004,
+      "step": 3224
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4853390723340452,
+      "learning_rate": 3.7148150965544205e-06,
+      "loss": 0.7335,
+      "step": 3225
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9630622628811005,
+      "learning_rate": 3.7091539263954246e-06,
+      "loss": 1.3683,
+      "step": 3226
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5571430187594235,
+      "learning_rate": 3.7034960908628426e-06,
+      "loss": 0.7766,
+      "step": 3227
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3607309028232835,
+      "learning_rate": 3.697841592955742e-06,
+      "loss": 0.6848,
+      "step": 3228
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4804522484010303,
+      "learning_rate": 3.692190435671452e-06,
+      "loss": 0.7022,
+      "step": 3229
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4083779927602083,
+      "learning_rate": 3.6865426220055002e-06,
+      "loss": 0.7441,
+      "step": 3230
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.363046144817507,
+      "learning_rate": 3.680898154951671e-06,
+      "loss": 0.718,
+      "step": 3231
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3327753597030554,
+      "learning_rate": 3.67525703750195e-06,
+      "loss": 0.6599,
+      "step": 3232
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0183891487576195,
+      "learning_rate": 3.669619272646563e-06,
+      "loss": 1.2894,
+      "step": 3233
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9401568913269773,
+      "learning_rate": 3.6639848633739526e-06,
+      "loss": 1.3642,
+      "step": 3234
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.362576442856578,
+      "learning_rate": 3.658353812670784e-06,
+      "loss": 0.7393,
+      "step": 3235
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3977828281730067,
+      "learning_rate": 3.652726123521945e-06,
+      "loss": 0.7409,
+      "step": 3236
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.55586702616328,
+      "learning_rate": 3.6471017989105284e-06,
+      "loss": 0.7858,
+      "step": 3237
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5515527295015552,
+      "learning_rate": 3.6414808418178684e-06,
+      "loss": 0.7658,
+      "step": 3238
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4070674116849053,
+      "learning_rate": 3.635863255223484e-06,
+      "loss": 0.7389,
+      "step": 3239
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.401171336521111,
+      "learning_rate": 3.6302490421051373e-06,
+      "loss": 0.794,
+      "step": 3240
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.03669814597537,
+      "learning_rate": 3.624638205438774e-06,
+      "loss": 1.3374,
+      "step": 3241
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9506655345104642,
+      "learning_rate": 3.6190307481985786e-06,
+      "loss": 1.3177,
+      "step": 3242
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9800759403491033,
+      "learning_rate": 3.6134266733569177e-06,
+      "loss": 1.3222,
+      "step": 3243
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5663279929963305,
+      "learning_rate": 3.607825983884381e-06,
+      "loss": 0.7612,
+      "step": 3244
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4024337826387379,
+      "learning_rate": 3.602228682749759e-06,
+      "loss": 0.7828,
+      "step": 3245
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3231346973754585,
+      "learning_rate": 3.5966347729200467e-06,
+      "loss": 0.6983,
+      "step": 3246
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5511819607842345,
+      "learning_rate": 3.5910442573604455e-06,
+      "loss": 0.7134,
+      "step": 3247
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9966717879715365,
+      "learning_rate": 3.585457139034342e-06,
+      "loss": 1.3294,
+      "step": 3248
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3759462552747503,
+      "learning_rate": 3.5798734209033483e-06,
+      "loss": 0.8102,
+      "step": 3249
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0007463785698913,
+      "learning_rate": 3.5742931059272445e-06,
+      "loss": 1.3176,
+      "step": 3250
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2822248862572612,
+      "learning_rate": 3.568716197064036e-06,
+      "loss": 0.6479,
+      "step": 3251
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3488320262970699,
+      "learning_rate": 3.5631426972698948e-06,
+      "loss": 0.7164,
+      "step": 3252
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5182895773219924,
+      "learning_rate": 3.5575726094992112e-06,
+      "loss": 0.7264,
+      "step": 3253
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4035704714595176,
+      "learning_rate": 3.5520059367045467e-06,
+      "loss": 0.7056,
+      "step": 3254
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.450183589538764,
+      "learning_rate": 3.5464426818366638e-06,
+      "loss": 0.8016,
+      "step": 3255
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3636690573996664,
+      "learning_rate": 3.540882847844512e-06,
+      "loss": 0.7149,
+      "step": 3256
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2917959733651827,
+      "learning_rate": 3.535326437675218e-06,
+      "loss": 0.8056,
+      "step": 3257
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3155895938534432,
+      "learning_rate": 3.529773454274115e-06,
+      "loss": 0.6949,
+      "step": 3258
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4259081545338919,
+      "learning_rate": 3.524223900584689e-06,
+      "loss": 0.7091,
+      "step": 3259
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.437091760138754,
+      "learning_rate": 3.5186777795486426e-06,
+      "loss": 0.6645,
+      "step": 3260
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4696977513537033,
+      "learning_rate": 3.513135094105826e-06,
+      "loss": 0.7581,
+      "step": 3261
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6111218696595337,
+      "learning_rate": 3.507595847194297e-06,
+      "loss": 0.7734,
+      "step": 3262
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2554871527083549,
+      "learning_rate": 3.5020600417502693e-06,
+      "loss": 0.7013,
+      "step": 3263
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0048375636392874,
+      "learning_rate": 3.4965276807081407e-06,
+      "loss": 1.3346,
+      "step": 3264
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9192315946737195,
+      "learning_rate": 3.4909987670004853e-06,
+      "loss": 1.4158,
+      "step": 3265
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5543193398254627,
+      "learning_rate": 3.485473303558046e-06,
+      "loss": 0.764,
+      "step": 3266
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4626620809124615,
+      "learning_rate": 3.4799512933097413e-06,
+      "loss": 0.7949,
+      "step": 3267
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8881465863915974,
+      "learning_rate": 3.4744327391826483e-06,
+      "loss": 1.3037,
+      "step": 3268
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3597931926582956,
+      "learning_rate": 3.468917644102031e-06,
+      "loss": 0.6461,
+      "step": 3269
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8898955146543914,
+      "learning_rate": 3.463406010991297e-06,
+      "loss": 1.3725,
+      "step": 3270
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5572355459653848,
+      "learning_rate": 3.4578978427720442e-06,
+      "loss": 0.7079,
+      "step": 3271
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5486630077302452,
+      "learning_rate": 3.4523931423640057e-06,
+      "loss": 0.6164,
+      "step": 3272
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3501344750082824,
+      "learning_rate": 3.446891912685105e-06,
+      "loss": 0.7271,
+      "step": 3273
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3642887724283166,
+      "learning_rate": 3.441394156651402e-06,
+      "loss": 0.7458,
+      "step": 3274
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4262445469197538,
+      "learning_rate": 3.4358998771771267e-06,
+      "loss": 0.7435,
+      "step": 3275
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4866077486273988,
+      "learning_rate": 3.430409077174668e-06,
+      "loss": 0.6846,
+      "step": 3276
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3339953079900415,
+      "learning_rate": 3.4249217595545638e-06,
+      "loss": 0.7198,
+      "step": 3277
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9838728565815174,
+      "learning_rate": 3.419437927225515e-06,
+      "loss": 1.3807,
+      "step": 3278
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.32386545487484,
+      "learning_rate": 3.4139575830943584e-06,
+      "loss": 0.6834,
+      "step": 3279
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3950328350725147,
+      "learning_rate": 3.408480730066106e-06,
+      "loss": 0.7457,
+      "step": 3280
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3445292783879934,
+      "learning_rate": 3.403007371043893e-06,
+      "loss": 0.8037,
+      "step": 3281
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4628609325488349,
+      "learning_rate": 3.3975375089290297e-06,
+      "loss": 0.7999,
+      "step": 3282
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4824264580784563,
+      "learning_rate": 3.392071146620943e-06,
+      "loss": 0.706,
+      "step": 3283
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.457500938631722,
+      "learning_rate": 3.386608287017236e-06,
+      "loss": 0.7659,
+      "step": 3284
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4259757148398342,
+      "learning_rate": 3.3811489330136284e-06,
+      "loss": 0.7239,
+      "step": 3285
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.470175824381347,
+      "learning_rate": 3.375693087503996e-06,
+      "loss": 0.784,
+      "step": 3286
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0543515216740116,
+      "learning_rate": 3.3702407533803536e-06,
+      "loss": 1.3096,
+      "step": 3287
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.499134951609572,
+      "learning_rate": 3.3647919335328462e-06,
+      "loss": 0.738,
+      "step": 3288
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.991045281585451,
+      "learning_rate": 3.359346630849771e-06,
+      "loss": 1.2931,
+      "step": 3289
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4213519464043527,
+      "learning_rate": 3.3539048482175418e-06,
+      "loss": 0.7566,
+      "step": 3290
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.520207304408493,
+      "learning_rate": 3.34846658852073e-06,
+      "loss": 0.7507,
+      "step": 3291
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.559920286634534,
+      "learning_rate": 3.3430318546420113e-06,
+      "loss": 0.7922,
+      "step": 3292
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.461316593181742,
+      "learning_rate": 3.337600649462224e-06,
+      "loss": 0.8576,
+      "step": 3293
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4743737479142314,
+      "learning_rate": 3.332172975860306e-06,
+      "loss": 0.7539,
+      "step": 3294
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.520901686921927,
+      "learning_rate": 3.3267488367133427e-06,
+      "loss": 0.7386,
+      "step": 3295
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5671072296458974,
+      "learning_rate": 3.3213282348965382e-06,
+      "loss": 0.8266,
+      "step": 3296
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3969821788853642,
+      "learning_rate": 3.315911173283225e-06,
+      "loss": 0.7253,
+      "step": 3297
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4358448717062737,
+      "learning_rate": 3.3104976547448597e-06,
+      "loss": 0.699,
+      "step": 3298
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4872863163417216,
+      "learning_rate": 3.305087682151008e-06,
+      "loss": 0.7175,
+      "step": 3299
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4543637650598562,
+      "learning_rate": 3.2996812583693804e-06,
+      "loss": 0.7225,
+      "step": 3300
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.343346566855804,
+      "learning_rate": 3.29427838626578e-06,
+      "loss": 0.7024,
+      "step": 3301
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.944848824266303,
+      "learning_rate": 3.2888790687041506e-06,
+      "loss": 1.2959,
+      "step": 3302
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4764006093957622,
+      "learning_rate": 3.283483308546529e-06,
+      "loss": 0.6557,
+      "step": 3303
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4331052215258382,
+      "learning_rate": 3.2780911086530907e-06,
+      "loss": 0.7487,
+      "step": 3304
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3837441828025545,
+      "learning_rate": 3.272702471882102e-06,
+      "loss": 0.7516,
+      "step": 3305
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3754877031756876,
+      "learning_rate": 3.267317401089953e-06,
+      "loss": 0.7007,
+      "step": 3306
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4389538879662473,
+      "learning_rate": 3.26193589913114e-06,
+      "loss": 0.646,
+      "step": 3307
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4949274533597832,
+      "learning_rate": 3.2565579688582683e-06,
+      "loss": 0.7317,
+      "step": 3308
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4370859219202676,
+      "learning_rate": 3.2511836131220533e-06,
+      "loss": 0.6443,
+      "step": 3309
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.476003868223906,
+      "learning_rate": 3.2458128347713002e-06,
+      "loss": 0.6772,
+      "step": 3310
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.417167611903852,
+      "learning_rate": 3.2404456366529445e-06,
+      "loss": 0.7243,
+      "step": 3311
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0432533802023913,
+      "learning_rate": 3.235082021611995e-06,
+      "loss": 1.392,
+      "step": 3312
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3515675090828358,
+      "learning_rate": 3.229721992491589e-06,
+      "loss": 0.6566,
+      "step": 3313
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.396166714183865,
+      "learning_rate": 3.2243655521329354e-06,
+      "loss": 0.6879,
+      "step": 3314
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0270990853368533,
+      "learning_rate": 3.219012703375368e-06,
+      "loss": 1.3163,
+      "step": 3315
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4635923887745685,
+      "learning_rate": 3.2136634490562948e-06,
+      "loss": 0.7058,
+      "step": 3316
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4419807700226794,
+      "learning_rate": 3.2083177920112274e-06,
+      "loss": 0.7449,
+      "step": 3317
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3553791911933024,
+      "learning_rate": 3.202975735073777e-06,
+      "loss": 0.7339,
+      "step": 3318
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4825457168544687,
+      "learning_rate": 3.197637281075626e-06,
+      "loss": 0.6533,
+      "step": 3319
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0213299932895854,
+      "learning_rate": 3.1923024328465767e-06,
+      "loss": 1.3701,
+      "step": 3320
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9752769681796238,
+      "learning_rate": 3.1869711932144896e-06,
+      "loss": 1.3648,
+      "step": 3321
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4833483660166649,
+      "learning_rate": 3.181643565005341e-06,
+      "loss": 0.8168,
+      "step": 3322
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4397103101928468,
+      "learning_rate": 3.1763195510431642e-06,
+      "loss": 0.7654,
+      "step": 3323
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9756497487730547,
+      "learning_rate": 3.170999154150105e-06,
+      "loss": 1.3141,
+      "step": 3324
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3407887551607838,
+      "learning_rate": 3.1656823771463672e-06,
+      "loss": 0.7083,
+      "step": 3325
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4872080693792837,
+      "learning_rate": 3.1603692228502515e-06,
+      "loss": 0.7738,
+      "step": 3326
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5287968923550617,
+      "learning_rate": 3.155059694078133e-06,
+      "loss": 0.7676,
+      "step": 3327
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3196302110664198,
+      "learning_rate": 3.1497537936444645e-06,
+      "loss": 0.6711,
+      "step": 3328
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9351232733810422,
+      "learning_rate": 3.144451524361779e-06,
+      "loss": 1.3402,
+      "step": 3329
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.909095587206969,
+      "learning_rate": 3.1391528890406754e-06,
+      "loss": 1.2647,
+      "step": 3330
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3810387621348088,
+      "learning_rate": 3.1338578904898422e-06,
+      "loss": 0.6764,
+      "step": 3331
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.058048923540104,
+      "learning_rate": 3.1285665315160208e-06,
+      "loss": 1.3652,
+      "step": 3332
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3730665597161742,
+      "learning_rate": 3.1232788149240446e-06,
+      "loss": 0.6297,
+      "step": 3333
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3991806517740488,
+      "learning_rate": 3.1179947435167936e-06,
+      "loss": 0.6882,
+      "step": 3334
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.936964500509936,
+      "learning_rate": 3.1127143200952393e-06,
+      "loss": 1.3412,
+      "step": 3335
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0518756742242026,
+      "learning_rate": 3.1074375474583986e-06,
+      "loss": 1.3854,
+      "step": 3336
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4655094002200262,
+      "learning_rate": 3.1021644284033637e-06,
+      "loss": 0.8288,
+      "step": 3337
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3874271639678422,
+      "learning_rate": 3.096894965725289e-06,
+      "loss": 0.6961,
+      "step": 3338
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4907093409979362,
+      "learning_rate": 3.09162916221739e-06,
+      "loss": 0.776,
+      "step": 3339
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4209965456804408,
+      "learning_rate": 3.0863670206709462e-06,
+      "loss": 0.7652,
+      "step": 3340
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4185670166404156,
+      "learning_rate": 3.081108543875283e-06,
+      "loss": 0.8611,
+      "step": 3341
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4188983940412858,
+      "learning_rate": 3.0758537346178052e-06,
+      "loss": 0.7793,
+      "step": 3342
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.930934708297618,
+      "learning_rate": 3.070602595683948e-06,
+      "loss": 1.3524,
+      "step": 3343
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3855967695714597,
+      "learning_rate": 3.065355129857227e-06,
+      "loss": 0.7012,
+      "step": 3344
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.424711982268728,
+      "learning_rate": 3.060111339919184e-06,
+      "loss": 0.7034,
+      "step": 3345
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.382521638240503,
+      "learning_rate": 3.054871228649441e-06,
+      "loss": 0.7252,
+      "step": 3346
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9200139920504649,
+      "learning_rate": 3.0496347988256435e-06,
+      "loss": 1.3721,
+      "step": 3347
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.107982018284323,
+      "learning_rate": 3.044402053223502e-06,
+      "loss": 1.3062,
+      "step": 3348
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.481731387595528,
+      "learning_rate": 3.0391729946167714e-06,
+      "loss": 0.7344,
+      "step": 3349
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.411169126599975,
+      "learning_rate": 3.0339476257772405e-06,
+      "loss": 0.7611,
+      "step": 3350
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9406385312898136,
+      "learning_rate": 3.0287259494747657e-06,
+      "loss": 1.2374,
+      "step": 3351
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.368678766233009,
+      "learning_rate": 3.023507968477217e-06,
+      "loss": 0.7808,
+      "step": 3352
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5036486699340768,
+      "learning_rate": 3.018293685550536e-06,
+      "loss": 0.7618,
+      "step": 3353
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3628291694628292,
+      "learning_rate": 3.013083103458676e-06,
+      "loss": 0.7317,
+      "step": 3354
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3931917727699812,
+      "learning_rate": 3.0078762249636526e-06,
+      "loss": 0.7187,
+      "step": 3355
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9965013356399257,
+      "learning_rate": 3.002673052825499e-06,
+      "loss": 1.3119,
+      "step": 3356
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4188708161841774,
+      "learning_rate": 2.9974735898022934e-06,
+      "loss": 0.7877,
+      "step": 3357
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.507521437502076,
+      "learning_rate": 2.9922778386501473e-06,
+      "loss": 0.7857,
+      "step": 3358
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4246740339379902,
+      "learning_rate": 2.987085802123203e-06,
+      "loss": 0.7429,
+      "step": 3359
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4023901020908809,
+      "learning_rate": 2.981897482973638e-06,
+      "loss": 0.6836,
+      "step": 3360
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3804870377280696,
+      "learning_rate": 2.9767128839516467e-06,
+      "loss": 0.6521,
+      "step": 3361
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4164477502113566,
+      "learning_rate": 2.971532007805472e-06,
+      "loss": 0.7963,
+      "step": 3362
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4669206518568623,
+      "learning_rate": 2.9663548572813603e-06,
+      "loss": 0.7867,
+      "step": 3363
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4481355410929955,
+      "learning_rate": 2.961181435123607e-06,
+      "loss": 0.6494,
+      "step": 3364
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.385914668080777,
+      "learning_rate": 2.956011744074505e-06,
+      "loss": 0.6614,
+      "step": 3365
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.052219092166001,
+      "learning_rate": 2.9508457868743977e-06,
+      "loss": 1.3654,
+      "step": 3366
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3711073756712908,
+      "learning_rate": 2.9456835662616255e-06,
+      "loss": 0.7444,
+      "step": 3367
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.519248666607185,
+      "learning_rate": 2.94052508497256e-06,
+      "loss": 0.8235,
+      "step": 3368
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3138451358920862,
+      "learning_rate": 2.935370345741588e-06,
+      "loss": 0.7619,
+      "step": 3369
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3719932190857869,
+      "learning_rate": 2.9302193513011133e-06,
+      "loss": 0.7799,
+      "step": 3370
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.401610615009723,
+      "learning_rate": 2.9250721043815576e-06,
+      "loss": 0.7119,
+      "step": 3371
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5002131835733021,
+      "learning_rate": 2.919928607711344e-06,
+      "loss": 0.7396,
+      "step": 3372
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.44750616211563,
+      "learning_rate": 2.9147888640169287e-06,
+      "loss": 0.8691,
+      "step": 3373
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5035845819751483,
+      "learning_rate": 2.9096528760227526e-06,
+      "loss": 0.7611,
+      "step": 3374
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4147083174338406,
+      "learning_rate": 2.9045206464512933e-06,
+      "loss": 0.775,
+      "step": 3375
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4012832709498513,
+      "learning_rate": 2.8993921780230096e-06,
+      "loss": 0.7242,
+      "step": 3376
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4716623280695396,
+      "learning_rate": 2.8942674734563924e-06,
+      "loss": 0.6714,
+      "step": 3377
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5198967546852855,
+      "learning_rate": 2.889146535467915e-06,
+      "loss": 0.7914,
+      "step": 3378
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.473955103188922,
+      "learning_rate": 2.8840293667720653e-06,
+      "loss": 0.7181,
+      "step": 3379
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4136979744336693,
+      "learning_rate": 2.8789159700813374e-06,
+      "loss": 0.7458,
+      "step": 3380
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3746079515666156,
+      "learning_rate": 2.8738063481062083e-06,
+      "loss": 0.6586,
+      "step": 3381
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3356512456939962,
+      "learning_rate": 2.8687005035551797e-06,
+      "loss": 0.6712,
+      "step": 3382
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9314875528364581,
+      "learning_rate": 2.863598439134724e-06,
+      "loss": 1.3683,
+      "step": 3383
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4681849448664022,
+      "learning_rate": 2.8585001575493355e-06,
+      "loss": 0.7339,
+      "step": 3384
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3907599232348444,
+      "learning_rate": 2.8534056615014784e-06,
+      "loss": 0.6602,
+      "step": 3385
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5875986981710708,
+      "learning_rate": 2.8483149536916353e-06,
+      "loss": 0.7653,
+      "step": 3386
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4771039318243684,
+      "learning_rate": 2.8432280368182594e-06,
+      "loss": 0.8006,
+      "step": 3387
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5175167683463906,
+      "learning_rate": 2.838144913577805e-06,
+      "loss": 0.6621,
+      "step": 3388
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9137862173286653,
+      "learning_rate": 2.833065586664715e-06,
+      "loss": 1.2909,
+      "step": 3389
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3917304520794485,
+      "learning_rate": 2.827990058771417e-06,
+      "loss": 0.7296,
+      "step": 3390
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.356240688190666,
+      "learning_rate": 2.8229183325883314e-06,
+      "loss": 0.6755,
+      "step": 3391
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.901675341870836,
+      "learning_rate": 2.8178504108038483e-06,
+      "loss": 1.3228,
+      "step": 3392
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4584227389952986,
+      "learning_rate": 2.8127862961043628e-06,
+      "loss": 0.7703,
+      "step": 3393
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.443297450979637,
+      "learning_rate": 2.8077259911742294e-06,
+      "loss": 0.7188,
+      "step": 3394
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.467633531552784,
+      "learning_rate": 2.8026694986958058e-06,
+      "loss": 0.7923,
+      "step": 3395
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4806855428769496,
+      "learning_rate": 2.797616821349405e-06,
+      "loss": 0.732,
+      "step": 3396
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4613688155575766,
+      "learning_rate": 2.792567961813343e-06,
+      "loss": 0.7326,
+      "step": 3397
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3996538330669666,
+      "learning_rate": 2.7875229227638877e-06,
+      "loss": 0.7589,
+      "step": 3398
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5127774447220643,
+      "learning_rate": 2.7824817068752984e-06,
+      "loss": 0.662,
+      "step": 3399
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.977424363548746,
+      "learning_rate": 2.777444316819802e-06,
+      "loss": 1.3404,
+      "step": 3400
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3749106327932696,
+      "learning_rate": 2.7724107552675973e-06,
+      "loss": 0.7381,
+      "step": 3401
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4815950959037725,
+      "learning_rate": 2.767381024886858e-06,
+      "loss": 0.7435,
+      "step": 3402
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.299392233414523,
+      "learning_rate": 2.7623551283437144e-06,
+      "loss": 0.7854,
+      "step": 3403
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.936940845075208,
+      "learning_rate": 2.757333068302285e-06,
+      "loss": 1.3047,
+      "step": 3404
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3467518508738394,
+      "learning_rate": 2.752314847424631e-06,
+      "loss": 0.686,
+      "step": 3405
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4767232141036095,
+      "learning_rate": 2.747300468370804e-06,
+      "loss": 0.7235,
+      "step": 3406
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.934152991061837,
+      "learning_rate": 2.742289933798793e-06,
+      "loss": 1.3125,
+      "step": 3407
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3938071101296947,
+      "learning_rate": 2.7372832463645736e-06,
+      "loss": 0.7565,
+      "step": 3408
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4242094394421079,
+      "learning_rate": 2.7322804087220623e-06,
+      "loss": 0.7644,
+      "step": 3409
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3324289615034552,
+      "learning_rate": 2.7272814235231472e-06,
+      "loss": 0.7255,
+      "step": 3410
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.895920632535077,
+      "learning_rate": 2.722286293417671e-06,
+      "loss": 1.3703,
+      "step": 3411
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4879318426030685,
+      "learning_rate": 2.7172950210534243e-06,
+      "loss": 0.7281,
+      "step": 3412
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.346933079277136,
+      "learning_rate": 2.7123076090761735e-06,
+      "loss": 0.7566,
+      "step": 3413
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.442478847614152,
+      "learning_rate": 2.7073240601296125e-06,
+      "loss": 0.7356,
+      "step": 3414
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3613844473190084,
+      "learning_rate": 2.7023443768554135e-06,
+      "loss": 0.7056,
+      "step": 3415
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.388956665781414,
+      "learning_rate": 2.6973685618931746e-06,
+      "loss": 0.8056,
+      "step": 3416
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.514828823038458,
+      "learning_rate": 2.6923966178804685e-06,
+      "loss": 0.8011,
+      "step": 3417
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4919098535395974,
+      "learning_rate": 2.687428547452794e-06,
+      "loss": 0.721,
+      "step": 3418
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4563747916121468,
+      "learning_rate": 2.682464353243609e-06,
+      "loss": 0.7402,
+      "step": 3419
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3885401660887258,
+      "learning_rate": 2.6775040378843133e-06,
+      "loss": 0.6783,
+      "step": 3420
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3414649296748595,
+      "learning_rate": 2.672547604004252e-06,
+      "loss": 0.6824,
+      "step": 3421
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5423042502393982,
+      "learning_rate": 2.667595054230713e-06,
+      "loss": 0.7421,
+      "step": 3422
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.0090282595375992,
+      "learning_rate": 2.6626463911889154e-06,
+      "loss": 1.3199,
+      "step": 3423
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.0576828134726526,
+      "learning_rate": 2.6577016175020396e-06,
+      "loss": 1.3391,
+      "step": 3424
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.438178390073961,
+      "learning_rate": 2.652760735791179e-06,
+      "loss": 0.7505,
+      "step": 3425
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3707442507970222,
+      "learning_rate": 2.6478237486753877e-06,
+      "loss": 0.7673,
+      "step": 3426
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.848997854558212,
+      "learning_rate": 2.642890658771633e-06,
+      "loss": 1.3758,
+      "step": 3427
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5183433821422314,
+      "learning_rate": 2.637961468694837e-06,
+      "loss": 0.8219,
+      "step": 3428
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5154860337514293,
+      "learning_rate": 2.63303618105784e-06,
+      "loss": 0.8204,
+      "step": 3429
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4807171366272307,
+      "learning_rate": 2.6281147984714183e-06,
+      "loss": 0.779,
+      "step": 3430
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4614385965551655,
+      "learning_rate": 2.62319732354428e-06,
+      "loss": 0.7661,
+      "step": 3431
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4958130768124516,
+      "learning_rate": 2.618283758883059e-06,
+      "loss": 0.78,
+      "step": 3432
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5006885425198209,
+      "learning_rate": 2.6133741070923237e-06,
+      "loss": 0.7321,
+      "step": 3433
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2695472888425243,
+      "learning_rate": 2.6084683707745506e-06,
+      "loss": 0.7331,
+      "step": 3434
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4536296518889533,
+      "learning_rate": 2.603566552530167e-06,
+      "loss": 0.6942,
+      "step": 3435
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4359231460589141,
+      "learning_rate": 2.5986686549574956e-06,
+      "loss": 0.7536,
+      "step": 3436
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3197359138959779,
+      "learning_rate": 2.5937746806528076e-06,
+      "loss": 0.7474,
+      "step": 3437
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4387055011638405,
+      "learning_rate": 2.5888846322102667e-06,
+      "loss": 0.8339,
+      "step": 3438
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4196402397172274,
+      "learning_rate": 2.5839985122219847e-06,
+      "loss": 0.6654,
+      "step": 3439
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3722052935755928,
+      "learning_rate": 2.579116323277967e-06,
+      "loss": 0.7429,
+      "step": 3440
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4624082646752612,
+      "learning_rate": 2.574238067966146e-06,
+      "loss": 0.6721,
+      "step": 3441
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4533903833539843,
+      "learning_rate": 2.5693637488723733e-06,
+      "loss": 0.6445,
+      "step": 3442
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.0503641634344114,
+      "learning_rate": 2.5644933685803974e-06,
+      "loss": 1.3409,
+      "step": 3443
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6046451020303572,
+      "learning_rate": 2.5596269296719023e-06,
+      "loss": 0.6855,
+      "step": 3444
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4449905830187377,
+      "learning_rate": 2.5547644347264587e-06,
+      "loss": 0.7798,
+      "step": 3445
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4459147483350872,
+      "learning_rate": 2.5499058863215696e-06,
+      "loss": 0.7308,
+      "step": 3446
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9035823446047282,
+      "learning_rate": 2.5450512870326247e-06,
+      "loss": 1.3806,
+      "step": 3447
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5394709555966217,
+      "learning_rate": 2.5402006394329403e-06,
+      "loss": 0.7215,
+      "step": 3448
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3697586715844658,
+      "learning_rate": 2.5353539460937203e-06,
+      "loss": 0.7914,
+      "step": 3449
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8599471655353619,
+      "learning_rate": 2.5305112095840835e-06,
+      "loss": 1.3711,
+      "step": 3450
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5873408296434761,
+      "learning_rate": 2.525672432471048e-06,
+      "loss": 0.7859,
+      "step": 3451
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9224628118733982,
+      "learning_rate": 2.5208376173195335e-06,
+      "loss": 1.2484,
+      "step": 3452
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.014327514621207,
+      "learning_rate": 2.5160067666923617e-06,
+      "loss": 1.3371,
+      "step": 3453
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8844883416449714,
+      "learning_rate": 2.5111798831502422e-06,
+      "loss": 1.3233,
+      "step": 3454
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.502375662684232,
+      "learning_rate": 2.506356969251802e-06,
+      "loss": 0.6713,
+      "step": 3455
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4876122852880311,
+      "learning_rate": 2.5015380275535395e-06,
+      "loss": 0.7616,
+      "step": 3456
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5935511672621414,
+      "learning_rate": 2.4967230606098726e-06,
+      "loss": 0.8457,
+      "step": 3457
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5105785459151624,
+      "learning_rate": 2.491912070973086e-06,
+      "loss": 0.707,
+      "step": 3458
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.561806445664145,
+      "learning_rate": 2.4871050611933835e-06,
+      "loss": 0.8281,
+      "step": 3459
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4709576838833573,
+      "learning_rate": 2.4823020338188365e-06,
+      "loss": 0.6979,
+      "step": 3460
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4325190616336,
+      "learning_rate": 2.477502991395416e-06,
+      "loss": 0.7074,
+      "step": 3461
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.931766958445536,
+      "learning_rate": 2.4727079364669816e-06,
+      "loss": 1.3313,
+      "step": 3462
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9866576549907744,
+      "learning_rate": 2.4679168715752743e-06,
+      "loss": 1.2637,
+      "step": 3463
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5079779351486056,
+      "learning_rate": 2.4631297992599256e-06,
+      "loss": 0.7715,
+      "step": 3464
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4084575370940984,
+      "learning_rate": 2.458346722058441e-06,
+      "loss": 0.7046,
+      "step": 3465
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3149711810796803,
+      "learning_rate": 2.453567642506224e-06,
+      "loss": 0.6368,
+      "step": 3466
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3812912135697988,
+      "learning_rate": 2.4487925631365396e-06,
+      "loss": 0.7087,
+      "step": 3467
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3477290776989843,
+      "learning_rate": 2.444021486480553e-06,
+      "loss": 0.6805,
+      "step": 3468
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.359884910483441,
+      "learning_rate": 2.4392544150672883e-06,
+      "loss": 0.7629,
+      "step": 3469
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.0152693535257056,
+      "learning_rate": 2.4344913514236658e-06,
+      "loss": 1.2674,
+      "step": 3470
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3292989648389726,
+      "learning_rate": 2.429732298074461e-06,
+      "loss": 0.6922,
+      "step": 3471
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8829528397933901,
+      "learning_rate": 2.424977257542338e-06,
+      "loss": 1.3492,
+      "step": 3472
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3550497712247005,
+      "learning_rate": 2.4202262323478332e-06,
+      "loss": 0.7391,
+      "step": 3473
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.463825907694828,
+      "learning_rate": 2.4154792250093406e-06,
+      "loss": 0.7404,
+      "step": 3474
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.0259659948709956,
+      "learning_rate": 2.4107362380431474e-06,
+      "loss": 1.3251,
+      "step": 3475
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3338882523228077,
+      "learning_rate": 2.4059972739633853e-06,
+      "loss": 0.6909,
+      "step": 3476
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9837917456585579,
+      "learning_rate": 2.4012623352820773e-06,
+      "loss": 1.3379,
+      "step": 3477
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9940957570241005,
+      "learning_rate": 2.3965314245090887e-06,
+      "loss": 1.3664,
+      "step": 3478
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.415995146528969,
+      "learning_rate": 2.391804544152172e-06,
+      "loss": 0.7127,
+      "step": 3479
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4009654838845278,
+      "learning_rate": 2.3870816967169262e-06,
+      "loss": 0.7471,
+      "step": 3480
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.539398347277995,
+      "learning_rate": 2.3823628847068226e-06,
+      "loss": 0.8242,
+      "step": 3481
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3381165249751785,
+      "learning_rate": 2.3776481106231873e-06,
+      "loss": 0.7168,
+      "step": 3482
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5041189898501968,
+      "learning_rate": 2.372937376965211e-06,
+      "loss": 0.78,
+      "step": 3483
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4000854964751208,
+      "learning_rate": 2.368230686229943e-06,
+      "loss": 0.7273,
+      "step": 3484
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4579671126190101,
+      "learning_rate": 2.3635280409122786e-06,
+      "loss": 0.6591,
+      "step": 3485
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.438006875086645,
+      "learning_rate": 2.3588294435049874e-06,
+      "loss": 0.7258,
+      "step": 3486
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.419226491948402,
+      "learning_rate": 2.3541348964986733e-06,
+      "loss": 0.7378,
+      "step": 3487
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4397053238058037,
+      "learning_rate": 2.349444402381813e-06,
+      "loss": 0.7087,
+      "step": 3488
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.406586299402203,
+      "learning_rate": 2.3447579636407136e-06,
+      "loss": 0.6172,
+      "step": 3489
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.0395106303034476,
+      "learning_rate": 2.3400755827595556e-06,
+      "loss": 1.344,
+      "step": 3490
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9588871436618442,
+      "learning_rate": 2.335397262220347e-06,
+      "loss": 1.363,
+      "step": 3491
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9233225581429603,
+      "learning_rate": 2.330723004502957e-06,
+      "loss": 1.3073,
+      "step": 3492
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.495446367820794,
+      "learning_rate": 2.326052812085098e-06,
+      "loss": 0.725,
+      "step": 3493
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4084343559245551,
+      "learning_rate": 2.321386687442326e-06,
+      "loss": 0.7509,
+      "step": 3494
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3909841397519203,
+      "learning_rate": 2.3167246330480442e-06,
+      "loss": 0.7106,
+      "step": 3495
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.438760077564209,
+      "learning_rate": 2.312066651373487e-06,
+      "loss": 0.7297,
+      "step": 3496
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4148550200606878,
+      "learning_rate": 2.3074127448877504e-06,
+      "loss": 0.7438,
+      "step": 3497
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.295593956265939,
+      "learning_rate": 2.302762916057747e-06,
+      "loss": 0.7947,
+      "step": 3498
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6383611318193305,
+      "learning_rate": 2.2981171673482514e-06,
+      "loss": 0.7617,
+      "step": 3499
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.401164870689358,
+      "learning_rate": 2.2934755012218503e-06,
+      "loss": 0.8138,
+      "step": 3500
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.417880397954858,
+      "learning_rate": 2.2888379201389934e-06,
+      "loss": 0.7031,
+      "step": 3501
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4393603378950148,
+      "learning_rate": 2.28420442655794e-06,
+      "loss": 0.6979,
+      "step": 3502
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4038052758864523,
+      "learning_rate": 2.279575022934797e-06,
+      "loss": 0.6766,
+      "step": 3503
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4768631171700082,
+      "learning_rate": 2.274949711723502e-06,
+      "loss": 0.6762,
+      "step": 3504
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.0032373418428766,
+      "learning_rate": 2.2703284953758186e-06,
+      "loss": 1.3104,
+      "step": 3505
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5169145322330537,
+      "learning_rate": 2.2657113763413487e-06,
+      "loss": 0.8673,
+      "step": 3506
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4122969676570214,
+      "learning_rate": 2.261098357067505e-06,
+      "loss": 0.7663,
+      "step": 3507
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3962537141674725,
+      "learning_rate": 2.2564894399995495e-06,
+      "loss": 0.8201,
+      "step": 3508
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4454232252400319,
+      "learning_rate": 2.2518846275805494e-06,
+      "loss": 0.678,
+      "step": 3509
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3653227313779193,
+      "learning_rate": 2.247283922251414e-06,
+      "loss": 0.7898,
+      "step": 3510
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9138117268062067,
+      "learning_rate": 2.242687326450861e-06,
+      "loss": 1.3484,
+      "step": 3511
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5663279787816455,
+      "learning_rate": 2.2380948426154346e-06,
+      "loss": 0.8394,
+      "step": 3512
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4668618884642868,
+      "learning_rate": 2.233506473179504e-06,
+      "loss": 0.7272,
+      "step": 3513
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3737054646751905,
+      "learning_rate": 2.2289222205752502e-06,
+      "loss": 0.7334,
+      "step": 3514
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.422333213912197,
+      "learning_rate": 2.2243420872326816e-06,
+      "loss": 0.7685,
+      "step": 3515
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.454563481416395,
+      "learning_rate": 2.2197660755796058e-06,
+      "loss": 0.7616,
+      "step": 3516
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3786041819858978,
+      "learning_rate": 2.215194188041668e-06,
+      "loss": 0.6817,
+      "step": 3517
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.391876572733245,
+      "learning_rate": 2.2106264270423062e-06,
+      "loss": 0.6418,
+      "step": 3518
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3999936452040904,
+      "learning_rate": 2.206062795002791e-06,
+      "loss": 0.7358,
+      "step": 3519
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.0084878255134675,
+      "learning_rate": 2.2015032943421843e-06,
+      "loss": 1.3035,
+      "step": 3520
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9894650754539145,
+      "learning_rate": 2.1969479274773763e-06,
+      "loss": 1.3028,
+      "step": 3521
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.908359059852855,
+      "learning_rate": 2.1923966968230524e-06,
+      "loss": 1.3357,
+      "step": 3522
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4810177030753435,
+      "learning_rate": 2.187849604791712e-06,
+      "loss": 0.6923,
+      "step": 3523
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4965938919969102,
+      "learning_rate": 2.1833066537936587e-06,
+      "loss": 0.6984,
+      "step": 3524
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4400555680477876,
+      "learning_rate": 2.178767846237002e-06,
+      "loss": 0.719,
+      "step": 3525
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4150456093661037,
+      "learning_rate": 2.1742331845276574e-06,
+      "loss": 0.6961,
+      "step": 3526
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4102906593649553,
+      "learning_rate": 2.1697026710693337e-06,
+      "loss": 0.6802,
+      "step": 3527
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3840618181946847,
+      "learning_rate": 2.165176308263557e-06,
+      "loss": 0.7143,
+      "step": 3528
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.0123671107537557,
+      "learning_rate": 2.1606540985096314e-06,
+      "loss": 1.3564,
+      "step": 3529
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.0475121030703187,
+      "learning_rate": 2.156136044204684e-06,
+      "loss": 1.3307,
+      "step": 3530
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.404923729745584,
+      "learning_rate": 2.151622147743616e-06,
+      "loss": 0.7602,
+      "step": 3531
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4108962731092214,
+      "learning_rate": 2.1471124115191456e-06,
+      "loss": 0.7256,
+      "step": 3532
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4057476702868883,
+      "learning_rate": 2.142606837921768e-06,
+      "loss": 0.7877,
+      "step": 3533
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4017664361601891,
+      "learning_rate": 2.138105429339783e-06,
+      "loss": 0.7488,
+      "step": 3534
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.413471895080426,
+      "learning_rate": 2.133608188159277e-06,
+      "loss": 0.6457,
+      "step": 3535
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5241949900996352,
+      "learning_rate": 2.129115116764132e-06,
+      "loss": 0.8189,
+      "step": 3536
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4747425687789812,
+      "learning_rate": 2.1246262175360178e-06,
+      "loss": 0.7063,
+      "step": 3537
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.979821639100524,
+      "learning_rate": 2.120141492854386e-06,
+      "loss": 1.2811,
+      "step": 3538
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5716674525175611,
+      "learning_rate": 2.1156609450964906e-06,
+      "loss": 0.7381,
+      "step": 3539
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.0027346935178243,
+      "learning_rate": 2.111184576637352e-06,
+      "loss": 1.3081,
+      "step": 3540
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4155015897551038,
+      "learning_rate": 2.1067123898497964e-06,
+      "loss": 0.6417,
+      "step": 3541
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4837139419900358,
+      "learning_rate": 2.102244387104415e-06,
+      "loss": 0.7419,
+      "step": 3542
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.388412507985895,
+      "learning_rate": 2.0977805707695918e-06,
+      "loss": 0.7446,
+      "step": 3543
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4359113681414746,
+      "learning_rate": 2.093320943211488e-06,
+      "loss": 0.7081,
+      "step": 3544
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4001766084141039,
+      "learning_rate": 2.0888655067940435e-06,
+      "loss": 0.7172,
+      "step": 3545
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3758774399313567,
+      "learning_rate": 2.084414263878981e-06,
+      "loss": 0.7784,
+      "step": 3546
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.885683955025129,
+      "learning_rate": 2.079967216825796e-06,
+      "loss": 1.343,
+      "step": 3547
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3703794801462093,
+      "learning_rate": 2.0755243679917623e-06,
+      "loss": 0.7454,
+      "step": 3548
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3650248744504863,
+      "learning_rate": 2.0710857197319224e-06,
+      "loss": 0.7561,
+      "step": 3549
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4766177408125842,
+      "learning_rate": 2.0666512743991072e-06,
+      "loss": 0.725,
+      "step": 3550
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4599804156836733,
+      "learning_rate": 2.0622210343438977e-06,
+      "loss": 0.7999,
+      "step": 3551
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3792689469352812,
+      "learning_rate": 2.0577950019146685e-06,
+      "loss": 0.7453,
+      "step": 3552
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.523988592609364,
+      "learning_rate": 2.0533731794575464e-06,
+      "loss": 0.7405,
+      "step": 3553
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.467989557698611,
+      "learning_rate": 2.0489555693164344e-06,
+      "loss": 0.7376,
+      "step": 3554
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5376858742594475,
+      "learning_rate": 2.0445421738330028e-06,
+      "loss": 0.7752,
+      "step": 3555
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8787392238657297,
+      "learning_rate": 2.0401329953466865e-06,
+      "loss": 1.3449,
+      "step": 3556
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.083120527229625,
+      "learning_rate": 2.035728036194686e-06,
+      "loss": 1.2908,
+      "step": 3557
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.576893188479935,
+      "learning_rate": 2.031327298711959e-06,
+      "loss": 0.7548,
+      "step": 3558
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4018615867062993,
+      "learning_rate": 2.0269307852312394e-06,
+      "loss": 0.6973,
+      "step": 3559
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.0090396785941333,
+      "learning_rate": 2.0225384980830044e-06,
+      "loss": 1.365,
+      "step": 3560
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3463685711384104,
+      "learning_rate": 2.0181504395955077e-06,
+      "loss": 0.7094,
+      "step": 3561
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4070563369539688,
+      "learning_rate": 2.0137666120947463e-06,
+      "loss": 0.6903,
+      "step": 3562
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.456891512090408,
+      "learning_rate": 2.0093870179044906e-06,
+      "loss": 0.7265,
+      "step": 3563
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4851888300254856,
+      "learning_rate": 2.005011659346251e-06,
+      "loss": 0.7978,
+      "step": 3564
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3506419000684997,
+      "learning_rate": 2.0006405387393025e-06,
+      "loss": 0.7853,
+      "step": 3565
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5034435951421696,
+      "learning_rate": 1.9962736584006714e-06,
+      "loss": 0.7497,
+      "step": 3566
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3359975989889292,
+      "learning_rate": 1.9919110206451364e-06,
+      "loss": 0.7987,
+      "step": 3567
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4419020121410135,
+      "learning_rate": 1.9875526277852275e-06,
+      "loss": 0.7292,
+      "step": 3568
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9335180544629313,
+      "learning_rate": 1.983198482131219e-06,
+      "loss": 1.341,
+      "step": 3569
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.457806213561725,
+      "learning_rate": 1.9788485859911475e-06,
+      "loss": 0.716,
+      "step": 3570
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3476800724520086,
+      "learning_rate": 1.9745029416707784e-06,
+      "loss": 0.7598,
+      "step": 3571
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3444465844025881,
+      "learning_rate": 1.970161551473645e-06,
+      "loss": 0.7913,
+      "step": 3572
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3636654773021208,
+      "learning_rate": 1.9658244177010043e-06,
+      "loss": 0.6433,
+      "step": 3573
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3695522179057131,
+      "learning_rate": 1.961491542651871e-06,
+      "loss": 0.7736,
+      "step": 3574
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.488167646162848,
+      "learning_rate": 1.9571629286229977e-06,
+      "loss": 0.7474,
+      "step": 3575
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4931854452089213,
+      "learning_rate": 1.9528385779088776e-06,
+      "loss": 0.6996,
+      "step": 3576
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.421927619872752,
+      "learning_rate": 1.9485184928017453e-06,
+      "loss": 0.7463,
+      "step": 3577
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4165273872107067,
+      "learning_rate": 1.9442026755915767e-06,
+      "loss": 0.7274,
+      "step": 3578
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4509305080823287,
+      "learning_rate": 1.9398911285660817e-06,
+      "loss": 0.7962,
+      "step": 3579
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.35062423484359,
+      "learning_rate": 1.935583854010703e-06,
+      "loss": 0.6928,
+      "step": 3580
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5165136419226968,
+      "learning_rate": 1.9312808542086325e-06,
+      "loss": 0.7103,
+      "step": 3581
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4372882601127477,
+      "learning_rate": 1.9269821314407765e-06,
+      "loss": 0.7408,
+      "step": 3582
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3177957870149168,
+      "learning_rate": 1.922687687985795e-06,
+      "loss": 0.7479,
+      "step": 3583
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.332666796986121,
+      "learning_rate": 1.9183975261200625e-06,
+      "loss": 0.7004,
+      "step": 3584
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.21398156517953,
+      "learning_rate": 1.9141116481176926e-06,
+      "loss": 0.6697,
+      "step": 3585
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4932723390725322,
+      "learning_rate": 1.9098300562505266e-06,
+      "loss": 0.7364,
+      "step": 3586
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.530849714231259,
+      "learning_rate": 1.9055527527881334e-06,
+      "loss": 0.7666,
+      "step": 3587
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4241734246926347,
+      "learning_rate": 1.9012797399978088e-06,
+      "loss": 0.7806,
+      "step": 3588
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5256391524178785,
+      "learning_rate": 1.8970110201445734e-06,
+      "loss": 0.7765,
+      "step": 3589
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.249787234647971,
+      "learning_rate": 1.8927465954911773e-06,
+      "loss": 0.7017,
+      "step": 3590
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3431270228732608,
+      "learning_rate": 1.8884864682980809e-06,
+      "loss": 0.8148,
+      "step": 3591
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8697881557944567,
+      "learning_rate": 1.8842306408234855e-06,
+      "loss": 1.379,
+      "step": 3592
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.058320997353357,
+      "learning_rate": 1.8799791153232938e-06,
+      "loss": 1.3242,
+      "step": 3593
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9277419164421636,
+      "learning_rate": 1.875731894051146e-06,
+      "loss": 1.3581,
+      "step": 3594
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3943217174224423,
+      "learning_rate": 1.8714889792583856e-06,
+      "loss": 0.7135,
+      "step": 3595
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4889260380962523,
+      "learning_rate": 1.8672503731940827e-06,
+      "loss": 0.7924,
+      "step": 3596
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.930467066671124,
+      "learning_rate": 1.8630160781050188e-06,
+      "loss": 1.3103,
+      "step": 3597
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3691998993046235,
+      "learning_rate": 1.858786096235693e-06,
+      "loss": 0.713,
+      "step": 3598
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2997568774140609,
+      "learning_rate": 1.8545604298283204e-06,
+      "loss": 0.763,
+      "step": 3599
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.276894829747784,
+      "learning_rate": 1.850339081122816e-06,
+      "loss": 0.6758,
+      "step": 3600
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9621798367721293,
+      "learning_rate": 1.846122052356828e-06,
+      "loss": 1.2642,
+      "step": 3601
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8928514871750344,
+      "learning_rate": 1.841909345765689e-06,
+      "loss": 1.2774,
+      "step": 3602
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9694388571265882,
+      "learning_rate": 1.837700963582466e-06,
+      "loss": 1.4,
+      "step": 3603
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4402294522095311,
+      "learning_rate": 1.8334969080379116e-06,
+      "loss": 0.6985,
+      "step": 3604
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3294064524240738,
+      "learning_rate": 1.8292971813604998e-06,
+      "loss": 0.666,
+      "step": 3605
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3380689993815462,
+      "learning_rate": 1.825101785776403e-06,
+      "loss": 0.7351,
+      "step": 3606
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4140055432896457,
+      "learning_rate": 1.8209107235095003e-06,
+      "loss": 0.7365,
+      "step": 3607
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3911921724756473,
+      "learning_rate": 1.8167239967813721e-06,
+      "loss": 0.7109,
+      "step": 3608
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4380936448054904,
+      "learning_rate": 1.8125416078113035e-06,
+      "loss": 0.7395,
+      "step": 3609
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3761796745444892,
+      "learning_rate": 1.8083635588162796e-06,
+      "loss": 0.7857,
+      "step": 3610
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9716437371853546,
+      "learning_rate": 1.8041898520109757e-06,
+      "loss": 1.2982,
+      "step": 3611
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3368901195043792,
+      "learning_rate": 1.8000204896077855e-06,
+      "loss": 0.7285,
+      "step": 3612
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5138271102911434,
+      "learning_rate": 1.7958554738167778e-06,
+      "loss": 0.7489,
+      "step": 3613
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.38746420009045,
+      "learning_rate": 1.7916948068457352e-06,
+      "loss": 0.7511,
+      "step": 3614
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3924415604322595,
+      "learning_rate": 1.7875384909001215e-06,
+      "loss": 0.7653,
+      "step": 3615
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.468927936406242,
+      "learning_rate": 1.783386528183102e-06,
+      "loss": 0.7595,
+      "step": 3616
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9907055281335413,
+      "learning_rate": 1.779238920895532e-06,
+      "loss": 1.3929,
+      "step": 3617
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.329307826339525,
+      "learning_rate": 1.7750956712359601e-06,
+      "loss": 0.7301,
+      "step": 3618
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4340998791528448,
+      "learning_rate": 1.7709567814006213e-06,
+      "loss": 0.7496,
+      "step": 3619
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3685549744098382,
+      "learning_rate": 1.7668222535834434e-06,
+      "loss": 0.7668,
+      "step": 3620
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9717426107026916,
+      "learning_rate": 1.7626920899760425e-06,
+      "loss": 1.3493,
+      "step": 3621
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4554772877562168,
+      "learning_rate": 1.7585662927677105e-06,
+      "loss": 0.7384,
+      "step": 3622
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4029054379849772,
+      "learning_rate": 1.7544448641454436e-06,
+      "loss": 0.7461,
+      "step": 3623
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9792426702518993,
+      "learning_rate": 1.7503278062939044e-06,
+      "loss": 1.2154,
+      "step": 3624
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3931579708954036,
+      "learning_rate": 1.7462151213954536e-06,
+      "loss": 0.764,
+      "step": 3625
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9806673984124283,
+      "learning_rate": 1.74210681163012e-06,
+      "loss": 1.3373,
+      "step": 3626
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4707506670452184,
+      "learning_rate": 1.738002879175623e-06,
+      "loss": 0.6774,
+      "step": 3627
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.90875078314018,
+      "learning_rate": 1.733903326207358e-06,
+      "loss": 1.3303,
+      "step": 3628
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5575723797980499,
+      "learning_rate": 1.7298081548984003e-06,
+      "loss": 0.728,
+      "step": 3629
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3636192260975508,
+      "learning_rate": 1.7257173674195016e-06,
+      "loss": 0.7327,
+      "step": 3630
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4358102736814227,
+      "learning_rate": 1.7216309659390895e-06,
+      "loss": 0.7335,
+      "step": 3631
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5429257400692509,
+      "learning_rate": 1.7175489526232703e-06,
+      "loss": 0.7435,
+      "step": 3632
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3172980192887982,
+      "learning_rate": 1.713471329635813e-06,
+      "loss": 0.6926,
+      "step": 3633
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3716737709375881,
+      "learning_rate": 1.7093980991381787e-06,
+      "loss": 0.6925,
+      "step": 3634
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9802730440716345,
+      "learning_rate": 1.7053292632894813e-06,
+      "loss": 1.3088,
+      "step": 3635
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.903687724597961,
+      "learning_rate": 1.7012648242465145e-06,
+      "loss": 1.4019,
+      "step": 3636
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.978609497354336,
+      "learning_rate": 1.6972047841637419e-06,
+      "loss": 1.3486,
+      "step": 3637
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5579482284837798,
+      "learning_rate": 1.6931491451932902e-06,
+      "loss": 0.7191,
+      "step": 3638
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.509389240904482,
+      "learning_rate": 1.6890979094849592e-06,
+      "loss": 0.7399,
+      "step": 3639
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.930583237040469,
+      "learning_rate": 1.6850510791862097e-06,
+      "loss": 1.3186,
+      "step": 3640
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9598946561213404,
+      "learning_rate": 1.6810086564421723e-06,
+      "loss": 1.3323,
+      "step": 3641
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4378287563987926,
+      "learning_rate": 1.6769706433956302e-06,
+      "loss": 0.7621,
+      "step": 3642
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.333193557964428,
+      "learning_rate": 1.6729370421870484e-06,
+      "loss": 0.6887,
+      "step": 3643
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4383054314735895,
+      "learning_rate": 1.6689078549545313e-06,
+      "loss": 0.8007,
+      "step": 3644
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9502634757696697,
+      "learning_rate": 1.6648830838338636e-06,
+      "loss": 1.3535,
+      "step": 3645
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.358678927945857,
+      "learning_rate": 1.6608627309584734e-06,
+      "loss": 0.7789,
+      "step": 3646
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.346041577954463,
+      "learning_rate": 1.656846798459456e-06,
+      "loss": 0.7288,
+      "step": 3647
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8944474760365537,
+      "learning_rate": 1.6528352884655596e-06,
+      "loss": 1.3027,
+      "step": 3648
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.300555752554688,
+      "learning_rate": 1.6488282031031889e-06,
+      "loss": 0.6551,
+      "step": 3649
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3455565858694878,
+      "learning_rate": 1.6448255444964056e-06,
+      "loss": 0.76,
+      "step": 3650
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4570735887147606,
+      "learning_rate": 1.6408273147669208e-06,
+      "loss": 0.7372,
+      "step": 3651
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.039629879898904,
+      "learning_rate": 1.6368335160341032e-06,
+      "loss": 1.3736,
+      "step": 3652
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4366567215522352,
+      "learning_rate": 1.632844150414964e-06,
+      "loss": 0.688,
+      "step": 3653
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.499286274706798,
+      "learning_rate": 1.6288592200241771e-06,
+      "loss": 0.7408,
+      "step": 3654
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8474416818089447,
+      "learning_rate": 1.6248787269740496e-06,
+      "loss": 1.3391,
+      "step": 3655
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4042376263428242,
+      "learning_rate": 1.620902673374556e-06,
+      "loss": 0.6485,
+      "step": 3656
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9195950479926625,
+      "learning_rate": 1.6169310613332988e-06,
+      "loss": 1.3332,
+      "step": 3657
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4130943885476643,
+      "learning_rate": 1.6129638929555357e-06,
+      "loss": 0.753,
+      "step": 3658
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4719677988730062,
+      "learning_rate": 1.6090011703441687e-06,
+      "loss": 0.8454,
+      "step": 3659
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9966625909811415,
+      "learning_rate": 1.605042895599742e-06,
+      "loss": 1.3249,
+      "step": 3660
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4291869856624126,
+      "learning_rate": 1.6010890708204407e-06,
+      "loss": 0.7741,
+      "step": 3661
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4518024951140005,
+      "learning_rate": 1.5971396981020937e-06,
+      "loss": 0.6723,
+      "step": 3662
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4886678999662082,
+      "learning_rate": 1.5931947795381696e-06,
+      "loss": 0.7614,
+      "step": 3663
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4428622452201398,
+      "learning_rate": 1.5892543172197694e-06,
+      "loss": 0.7653,
+      "step": 3664
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5421502110552519,
+      "learning_rate": 1.5853183132356453e-06,
+      "loss": 0.8592,
+      "step": 3665
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5171228398116505,
+      "learning_rate": 1.581386769672173e-06,
+      "loss": 0.7167,
+      "step": 3666
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9548371603962689,
+      "learning_rate": 1.5774596886133698e-06,
+      "loss": 1.3418,
+      "step": 3667
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4249940700661983,
+      "learning_rate": 1.5735370721408894e-06,
+      "loss": 0.7231,
+      "step": 3668
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9516563529084001,
+      "learning_rate": 1.5696189223340152e-06,
+      "loss": 1.3147,
+      "step": 3669
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.440462503214756,
+      "learning_rate": 1.565705241269664e-06,
+      "loss": 0.7463,
+      "step": 3670
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3405750714340432,
+      "learning_rate": 1.5617960310223845e-06,
+      "loss": 0.745,
+      "step": 3671
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.0442971091078053,
+      "learning_rate": 1.5578912936643553e-06,
+      "loss": 1.3437,
+      "step": 3672
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4659276577932212,
+      "learning_rate": 1.553991031265385e-06,
+      "loss": 0.7171,
+      "step": 3673
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.467619894529933,
+      "learning_rate": 1.5500952458929087e-06,
+      "loss": 0.7054,
+      "step": 3674
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4430778758796665,
+      "learning_rate": 1.5462039396119844e-06,
+      "loss": 0.7003,
+      "step": 3675
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4313675286430205,
+      "learning_rate": 1.5423171144853078e-06,
+      "loss": 0.697,
+      "step": 3676
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.938557961565526,
+      "learning_rate": 1.5384347725731841e-06,
+      "loss": 1.2764,
+      "step": 3677
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3286707477805273,
+      "learning_rate": 1.5345569159335527e-06,
+      "loss": 0.788,
+      "step": 3678
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9741058721602802,
+      "learning_rate": 1.5306835466219738e-06,
+      "loss": 1.2992,
+      "step": 3679
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.390914304814906,
+      "learning_rate": 1.526814666691624e-06,
+      "loss": 0.6882,
+      "step": 3680
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5001577496836158,
+      "learning_rate": 1.522950278193307e-06,
+      "loss": 0.7052,
+      "step": 3681
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.323646597067388,
+      "learning_rate": 1.51909038317544e-06,
+      "loss": 0.7516,
+      "step": 3682
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.042196452180918,
+      "learning_rate": 1.515234983684064e-06,
+      "loss": 1.3256,
+      "step": 3683
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.482676607985081,
+      "learning_rate": 1.5113840817628267e-06,
+      "loss": 0.7033,
+      "step": 3684
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5484448054924302,
+      "learning_rate": 1.5075376794530072e-06,
+      "loss": 0.7517,
+      "step": 3685
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9748851447418176,
+      "learning_rate": 1.5036957787934825e-06,
+      "loss": 1.3182,
+      "step": 3686
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8899881696172247,
+      "learning_rate": 1.4998583818207613e-06,
+      "loss": 1.2961,
+      "step": 3687
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4664252969463367,
+      "learning_rate": 1.4960254905689475e-06,
+      "loss": 0.8109,
+      "step": 3688
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.46899650216028,
+      "learning_rate": 1.4921971070697683e-06,
+      "loss": 0.7305,
+      "step": 3689
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3085153164830363,
+      "learning_rate": 1.488373233352557e-06,
+      "loss": 0.6936,
+      "step": 3690
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2174437395716826,
+      "learning_rate": 1.4845538714442576e-06,
+      "loss": 0.5601,
+      "step": 3691
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4863294807701368,
+      "learning_rate": 1.4807390233694218e-06,
+      "loss": 0.7025,
+      "step": 3692
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3442970959691942,
+      "learning_rate": 1.47692869115021e-06,
+      "loss": 0.7691,
+      "step": 3693
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4604850136253251,
+      "learning_rate": 1.4731228768063888e-06,
+      "loss": 0.7667,
+      "step": 3694
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3651103362883152,
+      "learning_rate": 1.469321582355322e-06,
+      "loss": 0.6802,
+      "step": 3695
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4955812918054148,
+      "learning_rate": 1.4655248098119945e-06,
+      "loss": 0.722,
+      "step": 3696
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4876716460409034,
+      "learning_rate": 1.4617325611889766e-06,
+      "loss": 0.7685,
+      "step": 3697
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.493545061095729,
+      "learning_rate": 1.4579448384964522e-06,
+      "loss": 0.6904,
+      "step": 3698
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3651072862120566,
+      "learning_rate": 1.4541616437422002e-06,
+      "loss": 0.7436,
+      "step": 3699
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.0287985709101535,
+      "learning_rate": 1.4503829789316016e-06,
+      "loss": 1.3436,
+      "step": 3700
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5305980438293598,
+      "learning_rate": 1.4466088460676364e-06,
+      "loss": 0.7261,
+      "step": 3701
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3713998225193185,
+      "learning_rate": 1.4428392471508813e-06,
+      "loss": 0.6281,
+      "step": 3702
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5227355750103857,
+      "learning_rate": 1.439074184179511e-06,
+      "loss": 0.7017,
+      "step": 3703
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.0515313541220257,
+      "learning_rate": 1.4353136591492933e-06,
+      "loss": 1.3155,
+      "step": 3704
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8712207036767041,
+      "learning_rate": 1.4315576740535964e-06,
+      "loss": 1.3107,
+      "step": 3705
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5088557775083193,
+      "learning_rate": 1.4278062308833707e-06,
+      "loss": 0.7148,
+      "step": 3706
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9094277746087163,
+      "learning_rate": 1.4240593316271733e-06,
+      "loss": 1.3438,
+      "step": 3707
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.0180709964642003,
+      "learning_rate": 1.4203169782711413e-06,
+      "loss": 1.3346,
+      "step": 3708
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5071100142081384,
+      "learning_rate": 1.4165791727990063e-06,
+      "loss": 0.7941,
+      "step": 3709
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4452273013803185,
+      "learning_rate": 1.4128459171920906e-06,
+      "loss": 0.7446,
+      "step": 3710
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9522096102514719,
+      "learning_rate": 1.409117213429303e-06,
+      "loss": 1.2676,
+      "step": 3711
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3946286679541213,
+      "learning_rate": 1.4053930634871393e-06,
+      "loss": 0.6581,
+      "step": 3712
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.0401387816684475,
+      "learning_rate": 1.4016734693396827e-06,
+      "loss": 1.3528,
+      "step": 3713
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9630703911874245,
+      "learning_rate": 1.3979584329586005e-06,
+      "loss": 1.3427,
+      "step": 3714
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.435559803341805,
+      "learning_rate": 1.3942479563131439e-06,
+      "loss": 0.6718,
+      "step": 3715
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.0053280595347247,
+      "learning_rate": 1.3905420413701498e-06,
+      "loss": 1.3662,
+      "step": 3716
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4181736118545833,
+      "learning_rate": 1.3868406900940279e-06,
+      "loss": 0.7847,
+      "step": 3717
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8617218498451282,
+      "learning_rate": 1.3831439044467853e-06,
+      "loss": 1.3507,
+      "step": 3718
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3881485948599397,
+      "learning_rate": 1.3794516863879925e-06,
+      "loss": 0.7441,
+      "step": 3719
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4381419419610797,
+      "learning_rate": 1.375764037874807e-06,
+      "loss": 0.7385,
+      "step": 3720
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4268592141484162,
+      "learning_rate": 1.3720809608619623e-06,
+      "loss": 0.6306,
+      "step": 3721
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5157883897389237,
+      "learning_rate": 1.3684024573017708e-06,
+      "loss": 0.8085,
+      "step": 3722
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2762075972145945,
+      "learning_rate": 1.364728529144118e-06,
+      "loss": 0.7142,
+      "step": 3723
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9357646997901492,
+      "learning_rate": 1.3610591783364647e-06,
+      "loss": 1.3347,
+      "step": 3724
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.826149661133602,
+      "learning_rate": 1.3573944068238477e-06,
+      "loss": 1.3248,
+      "step": 3725
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3595643151147312,
+      "learning_rate": 1.3537342165488688e-06,
+      "loss": 0.7169,
+      "step": 3726
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4405335095426202,
+      "learning_rate": 1.3500786094517148e-06,
+      "loss": 0.7378,
+      "step": 3727
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4920118629715846,
+      "learning_rate": 1.3464275874701282e-06,
+      "loss": 0.8166,
+      "step": 3728
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.41553877883717,
+      "learning_rate": 1.3427811525394318e-06,
+      "loss": 0.704,
+      "step": 3729
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4255149938688585,
+      "learning_rate": 1.3391393065925117e-06,
+      "loss": 0.6783,
+      "step": 3730
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.434793961188979,
+      "learning_rate": 1.3355020515598239e-06,
+      "loss": 0.7164,
+      "step": 3731
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3950533328311288,
+      "learning_rate": 1.3318693893693901e-06,
+      "loss": 0.6722,
+      "step": 3732
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3709719289967865,
+      "learning_rate": 1.328241321946796e-06,
+      "loss": 0.7598,
+      "step": 3733
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3720469916688551,
+      "learning_rate": 1.324617851215194e-06,
+      "loss": 0.7274,
+      "step": 3734
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4254763589117505,
+      "learning_rate": 1.3209989790952992e-06,
+      "loss": 0.672,
+      "step": 3735
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3649761532351021,
+      "learning_rate": 1.3173847075053902e-06,
+      "loss": 0.6474,
+      "step": 3736
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4348714750538556,
+      "learning_rate": 1.313775038361299e-06,
+      "loss": 0.7489,
+      "step": 3737
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4997296532204771,
+      "learning_rate": 1.3101699735764339e-06,
+      "loss": 0.759,
+      "step": 3738
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8681067409623693,
+      "learning_rate": 1.3065695150617463e-06,
+      "loss": 1.3165,
+      "step": 3739
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9424051776085627,
+      "learning_rate": 1.302973664725754e-06,
+      "loss": 1.3261,
+      "step": 3740
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.953438941494434,
+      "learning_rate": 1.2993824244745324e-06,
+      "loss": 1.3538,
+      "step": 3741
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4381148012480232,
+      "learning_rate": 1.2957957962117107e-06,
+      "loss": 0.8421,
+      "step": 3742
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4004265421313522,
+      "learning_rate": 1.2922137818384762e-06,
+      "loss": 0.7239,
+      "step": 3743
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4429753450459468,
+      "learning_rate": 1.2886363832535652e-06,
+      "loss": 0.7593,
+      "step": 3744
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4234477359931774,
+      "learning_rate": 1.2850636023532747e-06,
+      "loss": 0.6983,
+      "step": 3745
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9819076404534333,
+      "learning_rate": 1.2814954410314474e-06,
+      "loss": 1.3213,
+      "step": 3746
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4607781274965177,
+      "learning_rate": 1.2779319011794821e-06,
+      "loss": 0.6757,
+      "step": 3747
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.395583951299115,
+      "learning_rate": 1.2743729846863206e-06,
+      "loss": 0.735,
+      "step": 3748
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3306465947493562,
+      "learning_rate": 1.270818693438468e-06,
+      "loss": 0.7331,
+      "step": 3749
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5113317470488965,
+      "learning_rate": 1.2672690293199596e-06,
+      "loss": 0.7974,
+      "step": 3750
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4772398448156465,
+      "learning_rate": 1.2637239942123925e-06,
+      "loss": 0.7537,
+      "step": 3751
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.438482908394203,
+      "learning_rate": 1.2601835899949021e-06,
+      "loss": 0.674,
+      "step": 3752
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.417925022533067,
+      "learning_rate": 1.2566478185441733e-06,
+      "loss": 0.6526,
+      "step": 3753
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3918556190647704,
+      "learning_rate": 1.253116681734432e-06,
+      "loss": 0.7777,
+      "step": 3754
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.453376297113425,
+      "learning_rate": 1.2495901814374511e-06,
+      "loss": 0.7123,
+      "step": 3755
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4071887425913434,
+      "learning_rate": 1.246068319522542e-06,
+      "loss": 0.6763,
+      "step": 3756
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3711051211985967,
+      "learning_rate": 1.2425510978565592e-06,
+      "loss": 0.6955,
+      "step": 3757
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3921041017782687,
+      "learning_rate": 1.2390385183039011e-06,
+      "loss": 0.6339,
+      "step": 3758
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5007073319367612,
+      "learning_rate": 1.2355305827264951e-06,
+      "loss": 0.8086,
+      "step": 3759
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4650501617845657,
+      "learning_rate": 1.2320272929838183e-06,
+      "loss": 0.7789,
+      "step": 3760
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9201999150980023,
+      "learning_rate": 1.2285286509328786e-06,
+      "loss": 1.2905,
+      "step": 3761
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.418152953050618,
+      "learning_rate": 1.2250346584282235e-06,
+      "loss": 0.7418,
+      "step": 3762
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3632099254028134,
+      "learning_rate": 1.2215453173219327e-06,
+      "loss": 0.7643,
+      "step": 3763
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3974478315183494,
+      "learning_rate": 1.2180606294636244e-06,
+      "loss": 0.7548,
+      "step": 3764
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4382172111242704,
+      "learning_rate": 1.2145805967004453e-06,
+      "loss": 0.7726,
+      "step": 3765
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.417416376339436,
+      "learning_rate": 1.2111052208770791e-06,
+      "loss": 0.6957,
+      "step": 3766
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3581194114371196,
+      "learning_rate": 1.2076345038357407e-06,
+      "loss": 0.6653,
+      "step": 3767
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4753426214353362,
+      "learning_rate": 1.2041684474161675e-06,
+      "loss": 0.7248,
+      "step": 3768
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8791749169985648,
+      "learning_rate": 1.200707053455642e-06,
+      "loss": 1.3197,
+      "step": 3769
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.366513705162778,
+      "learning_rate": 1.1972503237889577e-06,
+      "loss": 0.7507,
+      "step": 3770
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4104206898575633,
+      "learning_rate": 1.1937982602484476e-06,
+      "loss": 0.7214,
+      "step": 3771
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4751794556082551,
+      "learning_rate": 1.1903508646639672e-06,
+      "loss": 0.767,
+      "step": 3772
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4896098218759555,
+      "learning_rate": 1.1869081388628978e-06,
+      "loss": 0.7722,
+      "step": 3773
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3960117755554582,
+      "learning_rate": 1.1834700846701463e-06,
+      "loss": 0.7918,
+      "step": 3774
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9503325981716901,
+      "learning_rate": 1.1800367039081427e-06,
+      "loss": 1.3243,
+      "step": 3775
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9114186872086052,
+      "learning_rate": 1.1766079983968382e-06,
+      "loss": 1.3297,
+      "step": 3776
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9195026606644496,
+      "learning_rate": 1.1731839699537096e-06,
+      "loss": 1.3122,
+      "step": 3777
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3973162019402234,
+      "learning_rate": 1.169764620393753e-06,
+      "loss": 0.6847,
+      "step": 3778
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9415576256355263,
+      "learning_rate": 1.1663499515294763e-06,
+      "loss": 1.3372,
+      "step": 3779
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3924290417458776,
+      "learning_rate": 1.162939965170924e-06,
+      "loss": 0.7493,
+      "step": 3780
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.907101401039195,
+      "learning_rate": 1.1595346631256388e-06,
+      "loss": 1.3201,
+      "step": 3781
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.433567231859795,
+      "learning_rate": 1.1561340471986949e-06,
+      "loss": 0.6741,
+      "step": 3782
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3673652088295445,
+      "learning_rate": 1.1527381191926745e-06,
+      "loss": 0.73,
+      "step": 3783
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.295794731600275,
+      "learning_rate": 1.149346880907678e-06,
+      "loss": 0.667,
+      "step": 3784
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8541201275082442,
+      "learning_rate": 1.1459603341413205e-06,
+      "loss": 1.351,
+      "step": 3785
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4442419266438102,
+      "learning_rate": 1.1425784806887286e-06,
+      "loss": 0.7394,
+      "step": 3786
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5521465562988095,
+      "learning_rate": 1.1392013223425414e-06,
+      "loss": 0.8081,
+      "step": 3787
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.323938136405012,
+      "learning_rate": 1.1358288608929092e-06,
+      "loss": 0.7445,
+      "step": 3788
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.848040522454922,
+      "learning_rate": 1.1324610981274953e-06,
+      "loss": 1.3266,
+      "step": 3789
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3607801910952426,
+      "learning_rate": 1.1290980358314652e-06,
+      "loss": 0.7289,
+      "step": 3790
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.0444394486970316,
+      "learning_rate": 1.1257396757874993e-06,
+      "loss": 1.2206,
+      "step": 3791
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2793721296768785,
+      "learning_rate": 1.1223860197757842e-06,
+      "loss": 0.694,
+      "step": 3792
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3649004309901334,
+      "learning_rate": 1.1190370695740127e-06,
+      "loss": 0.7062,
+      "step": 3793
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.488841299277983,
+      "learning_rate": 1.1156928269573831e-06,
+      "loss": 0.787,
+      "step": 3794
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9299635168434457,
+      "learning_rate": 1.112353293698597e-06,
+      "loss": 1.3197,
+      "step": 3795
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.1049470708482625,
+      "learning_rate": 1.109018471567861e-06,
+      "loss": 1.2607,
+      "step": 3796
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4330850891480997,
+      "learning_rate": 1.1056883623328852e-06,
+      "loss": 0.7639,
+      "step": 3797
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3819058138589722,
+      "learning_rate": 1.1023629677588787e-06,
+      "loss": 0.7602,
+      "step": 3798
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.341714840742556,
+      "learning_rate": 1.099042289608555e-06,
+      "loss": 0.8307,
+      "step": 3799
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.0174991562034177,
+      "learning_rate": 1.0957263296421273e-06,
+      "loss": 1.3675,
+      "step": 3800
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.443881264057751,
+      "learning_rate": 1.0924150896173014e-06,
+      "loss": 0.8193,
+      "step": 3801
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.460415908371837,
+      "learning_rate": 1.08910857128929e-06,
+      "loss": 0.7385,
+      "step": 3802
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3942530224736176,
+      "learning_rate": 1.085806776410797e-06,
+      "loss": 0.6376,
+      "step": 3803
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.493308782059089,
+      "learning_rate": 1.0825097067320267e-06,
+      "loss": 0.7036,
+      "step": 3804
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5088336525689092,
+      "learning_rate": 1.0792173640006743e-06,
+      "loss": 0.7167,
+      "step": 3805
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.366715974949383,
+      "learning_rate": 1.0759297499619325e-06,
+      "loss": 0.6942,
+      "step": 3806
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4551680919646781,
+      "learning_rate": 1.0726468663584877e-06,
+      "loss": 0.6708,
+      "step": 3807
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3448662113747603,
+      "learning_rate": 1.0693687149305166e-06,
+      "loss": 0.8051,
+      "step": 3808
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.53144809776173,
+      "learning_rate": 1.0660952974156901e-06,
+      "loss": 0.7779,
+      "step": 3809
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4297413973269084,
+      "learning_rate": 1.062826615549163e-06,
+      "loss": 0.7077,
+      "step": 3810
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.322450520579539,
+      "learning_rate": 1.0595626710635931e-06,
+      "loss": 0.7272,
+      "step": 3811
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3143918571496318,
+      "learning_rate": 1.0563034656891125e-06,
+      "loss": 0.7298,
+      "step": 3812
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.457183739846222,
+      "learning_rate": 1.0530490011533489e-06,
+      "loss": 0.7449,
+      "step": 3813
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9424724519450207,
+      "learning_rate": 1.0497992791814183e-06,
+      "loss": 1.2801,
+      "step": 3814
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.504292984510082,
+      "learning_rate": 1.0465543014959168e-06,
+      "loss": 0.794,
+      "step": 3815
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2951154726335918,
+      "learning_rate": 1.0433140698169319e-06,
+      "loss": 0.6874,
+      "step": 3816
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5151457227203022,
+      "learning_rate": 1.0400785858620321e-06,
+      "loss": 0.736,
+      "step": 3817
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4887250513819263,
+      "learning_rate": 1.0368478513462698e-06,
+      "loss": 0.7903,
+      "step": 3818
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4612509835854697,
+      "learning_rate": 1.0336218679821785e-06,
+      "loss": 0.7055,
+      "step": 3819
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3919699958752432,
+      "learning_rate": 1.030400637479777e-06,
+      "loss": 0.7139,
+      "step": 3820
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4700355115272663,
+      "learning_rate": 1.0271841615465583e-06,
+      "loss": 0.7247,
+      "step": 3821
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8798064642786838,
+      "learning_rate": 1.0239724418875018e-06,
+      "loss": 1.273,
+      "step": 3822
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4238222537669345,
+      "learning_rate": 1.0207654802050626e-06,
+      "loss": 0.734,
+      "step": 3823
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3722557576092431,
+      "learning_rate": 1.017563278199174e-06,
+      "loss": 0.7083,
+      "step": 3824
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.413642183610429,
+      "learning_rate": 1.0143658375672461e-06,
+      "loss": 0.7309,
+      "step": 3825
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9240066242646376,
+      "learning_rate": 1.0111731600041653e-06,
+      "loss": 1.3368,
+      "step": 3826
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4387131504260784,
+      "learning_rate": 1.0079852472022934e-06,
+      "loss": 0.7382,
+      "step": 3827
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9541943233841759,
+      "learning_rate": 1.0048021008514664e-06,
+      "loss": 1.3122,
+      "step": 3828
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9171044382794675,
+      "learning_rate": 1.0016237226389946e-06,
+      "loss": 1.3346,
+      "step": 3829
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3099583368798076,
+      "learning_rate": 9.984501142496582e-07,
+      "loss": 0.7945,
+      "step": 3830
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3167824035624713,
+      "learning_rate": 9.952812773657139e-07,
+      "loss": 0.7597,
+      "step": 3831
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.899298957095748,
+      "learning_rate": 9.921172136668822e-07,
+      "loss": 1.271,
+      "step": 3832
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.033023422892756,
+      "learning_rate": 9.889579248303593e-07,
+      "loss": 1.2882,
+      "step": 3833
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4549613583713572,
+      "learning_rate": 9.85803412530808e-07,
+      "loss": 0.7366,
+      "step": 3834
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3808222699640318,
+      "learning_rate": 9.826536784403595e-07,
+      "loss": 0.6856,
+      "step": 3835
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9198538623100003,
+      "learning_rate": 9.795087242286116e-07,
+      "loss": 1.3125,
+      "step": 3836
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4783700800854214,
+      "learning_rate": 9.763685515626298e-07,
+      "loss": 0.7598,
+      "step": 3837
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5254160137917216,
+      "learning_rate": 9.732331621069435e-07,
+      "loss": 0.8414,
+      "step": 3838
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4834220020479223,
+      "learning_rate": 9.70102557523549e-07,
+      "loss": 0.8292,
+      "step": 3839
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3780226429144982,
+      "learning_rate": 9.669767394719033e-07,
+      "loss": 0.7274,
+      "step": 3840
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3510783989440138,
+      "learning_rate": 9.638557096089262e-07,
+      "loss": 0.7836,
+      "step": 3841
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4519184043209983,
+      "learning_rate": 9.607394695890048e-07,
+      "loss": 0.7381,
+      "step": 3842
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3382747775821147,
+      "learning_rate": 9.576280210639788e-07,
+      "loss": 0.782,
+      "step": 3843
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4503781200522474,
+      "learning_rate": 9.545213656831543e-07,
+      "loss": 0.7201,
+      "step": 3844
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.414986035843107,
+      "learning_rate": 9.514195050932929e-07,
+      "loss": 0.7438,
+      "step": 3845
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4408915301839398,
+      "learning_rate": 9.483224409386182e-07,
+      "loss": 0.7197,
+      "step": 3846
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3942352307403283,
+      "learning_rate": 9.452301748608095e-07,
+      "loss": 0.7407,
+      "step": 3847
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3345577354341174,
+      "learning_rate": 9.42142708499002e-07,
+      "loss": 0.7317,
+      "step": 3848
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4459279334124588,
+      "learning_rate": 9.39060043489789e-07,
+      "loss": 0.8003,
+      "step": 3849
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.459989855318109,
+      "learning_rate": 9.359821814672143e-07,
+      "loss": 0.7169,
+      "step": 3850
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8959631228190672,
+      "learning_rate": 9.32909124062783e-07,
+      "loss": 1.3405,
+      "step": 3851
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3191880793746502,
+      "learning_rate": 9.298408729054453e-07,
+      "loss": 0.7134,
+      "step": 3852
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4185295402736067,
+      "learning_rate": 9.26777429621607e-07,
+      "loss": 0.76,
+      "step": 3853
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.941099061119808,
+      "learning_rate": 9.237187958351279e-07,
+      "loss": 1.2904,
+      "step": 3854
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5023393202918984,
+      "learning_rate": 9.206649731673168e-07,
+      "loss": 0.7321,
+      "step": 3855
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2546276946937702,
+      "learning_rate": 9.176159632369297e-07,
+      "loss": 0.6725,
+      "step": 3856
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9871395692234182,
+      "learning_rate": 9.14571767660174e-07,
+      "loss": 1.3499,
+      "step": 3857
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.0244325749454686,
+      "learning_rate": 9.115323880507054e-07,
+      "loss": 1.4171,
+      "step": 3858
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4670629407052702,
+      "learning_rate": 9.084978260196264e-07,
+      "loss": 0.725,
+      "step": 3859
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8680334258314466,
+      "learning_rate": 9.054680831754847e-07,
+      "loss": 1.2716,
+      "step": 3860
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4819713453134928,
+      "learning_rate": 9.024431611242746e-07,
+      "loss": 0.7281,
+      "step": 3861
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3260310582108106,
+      "learning_rate": 8.994230614694354e-07,
+      "loss": 0.7208,
+      "step": 3862
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4080540553414842,
+      "learning_rate": 8.96407785811847e-07,
+      "loss": 0.6834,
+      "step": 3863
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.305516678548659,
+      "learning_rate": 8.93397335749836e-07,
+      "loss": 0.6103,
+      "step": 3864
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.419362795875117,
+      "learning_rate": 8.903917128791695e-07,
+      "loss": 0.7331,
+      "step": 3865
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4368216552098743,
+      "learning_rate": 8.873909187930563e-07,
+      "loss": 0.8005,
+      "step": 3866
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4018700406893219,
+      "learning_rate": 8.843949550821451e-07,
+      "loss": 0.783,
+      "step": 3867
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4360358608270447,
+      "learning_rate": 8.814038233345234e-07,
+      "loss": 0.7722,
+      "step": 3868
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3879260327354666,
+      "learning_rate": 8.784175251357208e-07,
+      "loss": 0.7059,
+      "step": 3869
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4077074635594737,
+      "learning_rate": 8.754360620686997e-07,
+      "loss": 0.7272,
+      "step": 3870
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9332854327276225,
+      "learning_rate": 8.724594357138627e-07,
+      "loss": 1.3143,
+      "step": 3871
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.471282104485297,
+      "learning_rate": 8.694876476490488e-07,
+      "loss": 0.7331,
+      "step": 3872
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9077872193556025,
+      "learning_rate": 8.665206994495324e-07,
+      "loss": 1.3406,
+      "step": 3873
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3536903508007574,
+      "learning_rate": 8.635585926880185e-07,
+      "loss": 0.6327,
+      "step": 3874
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3981806604235603,
+      "learning_rate": 8.606013289346505e-07,
+      "loss": 0.7879,
+      "step": 3875
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.334011952163594,
+      "learning_rate": 8.576489097570017e-07,
+      "loss": 0.7617,
+      "step": 3876
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4033958475680128,
+      "learning_rate": 8.547013367200807e-07,
+      "loss": 0.704,
+      "step": 3877
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3911605052153708,
+      "learning_rate": 8.51758611386323e-07,
+      "loss": 0.6938,
+      "step": 3878
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8947683675881706,
+      "learning_rate": 8.488207353155986e-07,
+      "loss": 1.2745,
+      "step": 3879
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4293735295560976,
+      "learning_rate": 8.45887710065203e-07,
+      "loss": 0.7995,
+      "step": 3880
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.0005149871614414,
+      "learning_rate": 8.429595371898646e-07,
+      "loss": 1.2513,
+      "step": 3881
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4192439619742836,
+      "learning_rate": 8.400362182417355e-07,
+      "loss": 0.6645,
+      "step": 3882
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9127027286166352,
+      "learning_rate": 8.371177547704001e-07,
+      "loss": 1.2938,
+      "step": 3883
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5605220782931057,
+      "learning_rate": 8.342041483228613e-07,
+      "loss": 0.7566,
+      "step": 3884
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3070933790422647,
+      "learning_rate": 8.312954004435536e-07,
+      "loss": 0.6577,
+      "step": 3885
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.470323653422562,
+      "learning_rate": 8.283915126743358e-07,
+      "loss": 0.7387,
+      "step": 3886
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3392108865908368,
+      "learning_rate": 8.254924865544878e-07,
+      "loss": 0.7581,
+      "step": 3887
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4409129157928844,
+      "learning_rate": 8.225983236207136e-07,
+      "loss": 0.708,
+      "step": 3888
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4140080899427303,
+      "learning_rate": 8.197090254071405e-07,
+      "loss": 0.7374,
+      "step": 3889
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4140724078368525,
+      "learning_rate": 8.168245934453156e-07,
+      "loss": 0.6868,
+      "step": 3890
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4389594982904987,
+      "learning_rate": 8.139450292642059e-07,
+      "loss": 0.8256,
+      "step": 3891
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2901968676853466,
+      "learning_rate": 8.110703343902005e-07,
+      "loss": 0.7034,
+      "step": 3892
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9508376503173392,
+      "learning_rate": 8.08200510347108e-07,
+      "loss": 1.3152,
+      "step": 3893
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3857598536063833,
+      "learning_rate": 8.053355586561496e-07,
+      "loss": 0.5961,
+      "step": 3894
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.0035472721934036,
+      "learning_rate": 8.024754808359681e-07,
+      "loss": 1.3203,
+      "step": 3895
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4692615246833847,
+      "learning_rate": 7.996202784026219e-07,
+      "loss": 0.7342,
+      "step": 3896
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3274688017450504,
+      "learning_rate": 7.967699528695871e-07,
+      "loss": 0.7,
+      "step": 3897
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4788597346069297,
+      "learning_rate": 7.939245057477518e-07,
+      "loss": 0.7541,
+      "step": 3898
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4774681104871654,
+      "learning_rate": 7.910839385454183e-07,
+      "loss": 0.7155,
+      "step": 3899
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4226806789228474,
+      "learning_rate": 7.882482527683055e-07,
+      "loss": 0.7562,
+      "step": 3900
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.409321958755319,
+      "learning_rate": 7.854174499195389e-07,
+      "loss": 0.7118,
+      "step": 3901
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.392477408152208,
+      "learning_rate": 7.825915314996624e-07,
+      "loss": 0.7048,
+      "step": 3902
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4615291520277374,
+      "learning_rate": 7.797704990066246e-07,
+      "loss": 0.7786,
+      "step": 3903
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3309335668066762,
+      "learning_rate": 7.769543539357904e-07,
+      "loss": 0.7385,
+      "step": 3904
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4799230660921132,
+      "learning_rate": 7.74143097779928e-07,
+      "loss": 0.8486,
+      "step": 3905
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.039372609967088,
+      "learning_rate": 7.713367320292165e-07,
+      "loss": 1.2751,
+      "step": 3906
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3753027296713576,
+      "learning_rate": 7.685352581712447e-07,
+      "loss": 0.7088,
+      "step": 3907
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3484541584855585,
+      "learning_rate": 7.657386776910048e-07,
+      "loss": 0.7895,
+      "step": 3908
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4498098099860868,
+      "learning_rate": 7.629469920708987e-07,
+      "loss": 0.7603,
+      "step": 3909
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4771643136140429,
+      "learning_rate": 7.601602027907318e-07,
+      "loss": 0.669,
+      "step": 3910
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5517241424145656,
+      "learning_rate": 7.573783113277122e-07,
+      "loss": 0.8067,
+      "step": 3911
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5043769445042308,
+      "learning_rate": 7.54601319156456e-07,
+      "loss": 0.7554,
+      "step": 3912
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.0158630953828363,
+      "learning_rate": 7.518292277489791e-07,
+      "loss": 1.2569,
+      "step": 3913
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2869504120145723,
+      "learning_rate": 7.490620385747016e-07,
+      "loss": 0.7011,
+      "step": 3914
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4032992588652227,
+      "learning_rate": 7.462997531004423e-07,
+      "loss": 0.7394,
+      "step": 3915
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.0901523702209386,
+      "learning_rate": 7.435423727904223e-07,
+      "loss": 1.3412,
+      "step": 3916
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9665179299494584,
+      "learning_rate": 7.407898991062646e-07,
+      "loss": 1.3745,
+      "step": 3917
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.296272110848553,
+      "learning_rate": 7.380423335069875e-07,
+      "loss": 0.7126,
+      "step": 3918
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3777827179775926,
+      "learning_rate": 7.352996774490096e-07,
+      "loss": 0.7526,
+      "step": 3919
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9579421108191486,
+      "learning_rate": 7.325619323861488e-07,
+      "loss": 1.3005,
+      "step": 3920
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.329438418686927,
+      "learning_rate": 7.298290997696167e-07,
+      "loss": 0.7989,
+      "step": 3921
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3907521838611125,
+      "learning_rate": 7.271011810480222e-07,
+      "loss": 0.7036,
+      "step": 3922
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4385298875891925,
+      "learning_rate": 7.243781776673698e-07,
+      "loss": 0.7816,
+      "step": 3923
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4666336845438248,
+      "learning_rate": 7.216600910710581e-07,
+      "loss": 0.842,
+      "step": 3924
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5316703698308274,
+      "learning_rate": 7.189469226998813e-07,
+      "loss": 0.8227,
+      "step": 3925
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4216231962008683,
+      "learning_rate": 7.162386739920225e-07,
+      "loss": 0.612,
+      "step": 3926
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4388158614813042,
+      "learning_rate": 7.135353463830585e-07,
+      "loss": 0.7062,
+      "step": 3927
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.399354783680509,
+      "learning_rate": 7.108369413059602e-07,
+      "loss": 0.8169,
+      "step": 3928
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4746246810069052,
+      "learning_rate": 7.081434601910863e-07,
+      "loss": 0.7691,
+      "step": 3929
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9964447341788742,
+      "learning_rate": 7.054549044661874e-07,
+      "loss": 1.3503,
+      "step": 3930
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.975263264586766,
+      "learning_rate": 7.027712755564009e-07,
+      "loss": 1.3549,
+      "step": 3931
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5048847136712058,
+      "learning_rate": 7.000925748842557e-07,
+      "loss": 0.8224,
+      "step": 3932
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3982047640847692,
+      "learning_rate": 6.974188038696661e-07,
+      "loss": 0.7634,
+      "step": 3933
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4838881811609579,
+      "learning_rate": 6.947499639299337e-07,
+      "loss": 0.6942,
+      "step": 3934
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9297703596933233,
+      "learning_rate": 6.920860564797483e-07,
+      "loss": 1.3236,
+      "step": 3935
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.567750695892792,
+      "learning_rate": 6.8942708293118e-07,
+      "loss": 0.7815,
+      "step": 3936
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4752003688078295,
+      "learning_rate": 6.867730446936871e-07,
+      "loss": 0.7243,
+      "step": 3937
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4292816256949246,
+      "learning_rate": 6.841239431741143e-07,
+      "loss": 0.682,
+      "step": 3938
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3203772558582025,
+      "learning_rate": 6.814797797766837e-07,
+      "loss": 0.67,
+      "step": 3939
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9647166179024789,
+      "learning_rate": 6.788405559030053e-07,
+      "loss": 1.3116,
+      "step": 3940
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.0508720909074287,
+      "learning_rate": 6.762062729520669e-07,
+      "loss": 1.3368,
+      "step": 3941
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5154089103778314,
+      "learning_rate": 6.735769323202379e-07,
+      "loss": 0.7277,
+      "step": 3942
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9393810479240834,
+      "learning_rate": 6.709525354012714e-07,
+      "loss": 1.2724,
+      "step": 3943
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2941609022273652,
+      "learning_rate": 6.683330835862933e-07,
+      "loss": 0.6916,
+      "step": 3944
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9070428713690828,
+      "learning_rate": 6.657185782638176e-07,
+      "loss": 1.2409,
+      "step": 3945
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4391028459862527,
+      "learning_rate": 6.631090208197243e-07,
+      "loss": 0.7278,
+      "step": 3946
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.0003071918174156,
+      "learning_rate": 6.605044126372817e-07,
+      "loss": 1.3399,
+      "step": 3947
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4117836371851844,
+      "learning_rate": 6.579047550971273e-07,
+      "loss": 0.6743,
+      "step": 3948
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.456923877189213,
+      "learning_rate": 6.55310049577279e-07,
+      "loss": 0.7331,
+      "step": 3949
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3060705338284886,
+      "learning_rate": 6.527202974531277e-07,
+      "loss": 0.641,
+      "step": 3950
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4892367932441812,
+      "learning_rate": 6.501355000974374e-07,
+      "loss": 0.7625,
+      "step": 3951
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9369708430706076,
+      "learning_rate": 6.475556588803477e-07,
+      "loss": 1.3095,
+      "step": 3952
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4148685487236217,
+      "learning_rate": 6.449807751693715e-07,
+      "loss": 0.6768,
+      "step": 3953
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6037389538122222,
+      "learning_rate": 6.424108503293924e-07,
+      "loss": 0.7628,
+      "step": 3954
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.320369752479383,
+      "learning_rate": 6.398458857226642e-07,
+      "loss": 0.65,
+      "step": 3955
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9693799574643436,
+      "learning_rate": 6.372858827088168e-07,
+      "loss": 1.3639,
+      "step": 3956
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.914246319985174,
+      "learning_rate": 6.347308426448429e-07,
+      "loss": 1.3313,
+      "step": 3957
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4748524436653943,
+      "learning_rate": 6.321807668851077e-07,
+      "loss": 0.7004,
+      "step": 3958
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.894166109187003,
+      "learning_rate": 6.296356567813466e-07,
+      "loss": 1.2989,
+      "step": 3959
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4058346888939193,
+      "learning_rate": 6.270955136826618e-07,
+      "loss": 0.7364,
+      "step": 3960
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2639196724282986,
+      "learning_rate": 6.245603389355204e-07,
+      "loss": 0.6719,
+      "step": 3961
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3607455162213755,
+      "learning_rate": 6.220301338837587e-07,
+      "loss": 0.6985,
+      "step": 3962
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3748901956656374,
+      "learning_rate": 6.195048998685771e-07,
+      "loss": 0.7502,
+      "step": 3963
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4847903512764475,
+      "learning_rate": 6.169846382285427e-07,
+      "loss": 0.6897,
+      "step": 3964
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.452329100669913,
+      "learning_rate": 6.144693502995846e-07,
+      "loss": 0.7747,
+      "step": 3965
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4464528953101123,
+      "learning_rate": 6.119590374149964e-07,
+      "loss": 0.733,
+      "step": 3966
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9905956777316707,
+      "learning_rate": 6.094537009054368e-07,
+      "loss": 1.2512,
+      "step": 3967
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.449528274004076,
+      "learning_rate": 6.069533420989215e-07,
+      "loss": 0.7561,
+      "step": 3968
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5360410255254802,
+      "learning_rate": 6.044579623208313e-07,
+      "loss": 0.749,
+      "step": 3969
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.375969544668793,
+      "learning_rate": 6.019675628939082e-07,
+      "loss": 0.7141,
+      "step": 3970
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5077147535225492,
+      "learning_rate": 5.994821451382527e-07,
+      "loss": 0.6902,
+      "step": 3971
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.585537485007743,
+      "learning_rate": 5.970017103713243e-07,
+      "loss": 0.8036,
+      "step": 3972
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4042510783402398,
+      "learning_rate": 5.945262599079438e-07,
+      "loss": 0.7112,
+      "step": 3973
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4104200904752073,
+      "learning_rate": 5.920557950602856e-07,
+      "loss": 0.7803,
+      "step": 3974
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3861188174733337,
+      "learning_rate": 5.895903171378858e-07,
+      "loss": 0.7373,
+      "step": 3975
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.437162825685855,
+      "learning_rate": 5.871298274476345e-07,
+      "loss": 0.7394,
+      "step": 3976
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.542883577843603,
+      "learning_rate": 5.846743272937771e-07,
+      "loss": 0.6835,
+      "step": 3977
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3414378473696926,
+      "learning_rate": 5.822238179779161e-07,
+      "loss": 0.6654,
+      "step": 3978
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4090138757938504,
+      "learning_rate": 5.797783007990077e-07,
+      "loss": 0.7064,
+      "step": 3979
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3960026781308121,
+      "learning_rate": 5.773377770533617e-07,
+      "loss": 0.814,
+      "step": 3980
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4432125354503473,
+      "learning_rate": 5.7490224803464e-07,
+      "loss": 0.7233,
+      "step": 3981
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.386284943175363,
+      "learning_rate": 5.7247171503386e-07,
+      "loss": 0.6979,
+      "step": 3982
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.968204524301752,
+      "learning_rate": 5.700461793393886e-07,
+      "loss": 1.2475,
+      "step": 3983
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3353288669689227,
+      "learning_rate": 5.676256422369419e-07,
+      "loss": 0.6811,
+      "step": 3984
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9277102596407227,
+      "learning_rate": 5.652101050095915e-07,
+      "loss": 1.3397,
+      "step": 3985
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4947897229730223,
+      "learning_rate": 5.62799568937753e-07,
+      "loss": 0.701,
+      "step": 3986
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3820555871010485,
+      "learning_rate": 5.603940352991977e-07,
+      "loss": 0.7332,
+      "step": 3987
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4907012470387626,
+      "learning_rate": 5.579935053690355e-07,
+      "loss": 0.7648,
+      "step": 3988
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3368971852930651,
+      "learning_rate": 5.555979804197331e-07,
+      "loss": 0.7379,
+      "step": 3989
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.480368007733484,
+      "learning_rate": 5.532074617211003e-07,
+      "loss": 0.6888,
+      "step": 3990
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3727081371965033,
+      "learning_rate": 5.508219505402934e-07,
+      "loss": 0.6699,
+      "step": 3991
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4644770079626264,
+      "learning_rate": 5.484414481418143e-07,
+      "loss": 0.7055,
+      "step": 3992
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.852485645255184,
+      "learning_rate": 5.460659557875104e-07,
+      "loss": 1.301,
+      "step": 3993
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5604528034469358,
+      "learning_rate": 5.43695474736573e-07,
+      "loss": 0.7513,
+      "step": 3994
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.451099257412905,
+      "learning_rate": 5.413300062455385e-07,
+      "loss": 0.7132,
+      "step": 3995
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.001689557370865,
+      "learning_rate": 5.389695515682835e-07,
+      "loss": 1.297,
+      "step": 3996
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4333892847124239,
+      "learning_rate": 5.366141119560286e-07,
+      "loss": 0.669,
+      "step": 3997
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8685937111377846,
+      "learning_rate": 5.342636886573372e-07,
+      "loss": 1.2473,
+      "step": 3998
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4720610434531292,
+      "learning_rate": 5.31918282918109e-07,
+      "loss": 0.7526,
+      "step": 3999
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5026518549340053,
+      "learning_rate": 5.2957789598159e-07,
+      "loss": 0.7477,
+      "step": 4000
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9940204068605123,
+      "learning_rate": 5.272425290883609e-07,
+      "loss": 1.3343,
+      "step": 4001
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3299167517779256,
+      "learning_rate": 5.24912183476346e-07,
+      "loss": 0.7103,
+      "step": 4002
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4633669051270688,
+      "learning_rate": 5.225868603808049e-07,
+      "loss": 0.6855,
+      "step": 4003
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3559217088758013,
+      "learning_rate": 5.202665610343338e-07,
+      "loss": 0.7908,
+      "step": 4004
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8816265178354519,
+      "learning_rate": 5.179512866668702e-07,
+      "loss": 1.2607,
+      "step": 4005
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.038302316044815,
+      "learning_rate": 5.156410385056831e-07,
+      "loss": 1.3467,
+      "step": 4006
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4646575986249564,
+      "learning_rate": 5.133358177753823e-07,
+      "loss": 0.7267,
+      "step": 4007
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.469503885398861,
+      "learning_rate": 5.110356256979066e-07,
+      "loss": 0.7739,
+      "step": 4008
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2967664985646652,
+      "learning_rate": 5.087404634925353e-07,
+      "loss": 0.7412,
+      "step": 4009
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8976637462647765,
+      "learning_rate": 5.064503323758774e-07,
+      "loss": 1.2229,
+      "step": 4010
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.366666998952394,
+      "learning_rate": 5.041652335618763e-07,
+      "loss": 0.7552,
+      "step": 4011
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3865147338747053,
+      "learning_rate": 5.018851682618087e-07,
+      "loss": 0.667,
+      "step": 4012
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.412011512775996,
+      "learning_rate": 4.996101376842821e-07,
+      "loss": 0.7497,
+      "step": 4013
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.811234069767589,
+      "learning_rate": 4.973401430352353e-07,
+      "loss": 1.3353,
+      "step": 4014
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.560495283768061,
+      "learning_rate": 4.950751855179381e-07,
+      "loss": 0.7292,
+      "step": 4015
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9704551954543283,
+      "learning_rate": 4.928152663329911e-07,
+      "loss": 1.3748,
+      "step": 4016
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.0020462712667766,
+      "learning_rate": 4.905603866783216e-07,
+      "loss": 1.3227,
+      "step": 4017
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5312629364438746,
+      "learning_rate": 4.8831054774919e-07,
+      "loss": 0.767,
+      "step": 4018
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.052309430640051,
+      "learning_rate": 4.860657507381772e-07,
+      "loss": 1.3064,
+      "step": 4019
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3073514159130968,
+      "learning_rate": 4.838259968351999e-07,
+      "loss": 0.7652,
+      "step": 4020
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3452555863312519,
+      "learning_rate": 4.815912872274963e-07,
+      "loss": 0.7317,
+      "step": 4021
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.495636667622426,
+      "learning_rate": 4.793616230996323e-07,
+      "loss": 0.8139,
+      "step": 4022
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.0184900068938756,
+      "learning_rate": 4.771370056335001e-07,
+      "loss": 1.3064,
+      "step": 4023
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4817504487980557,
+      "learning_rate": 4.749174360083153e-07,
+      "loss": 0.7736,
+      "step": 4024
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3647058911739807,
+      "learning_rate": 4.727029154006191e-07,
+      "loss": 0.7452,
+      "step": 4025
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4499475457724524,
+      "learning_rate": 4.7049344498427465e-07,
+      "loss": 0.7779,
+      "step": 4026
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4170483295181817,
+      "learning_rate": 4.6828902593046953e-07,
+      "loss": 0.6846,
+      "step": 4027
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.490991940411522,
+      "learning_rate": 4.6608965940771464e-07,
+      "loss": 0.7074,
+      "step": 4028
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4001899158091047,
+      "learning_rate": 4.6389534658184077e-07,
+      "loss": 0.7409,
+      "step": 4029
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9589922211759532,
+      "learning_rate": 4.6170608861599965e-07,
+      "loss": 1.3176,
+      "step": 4030
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4380921905222326,
+      "learning_rate": 4.59521886670663e-07,
+      "loss": 0.711,
+      "step": 4031
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4675977397164441,
+      "learning_rate": 4.573427419036269e-07,
+      "loss": 0.7376,
+      "step": 4032
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5178887767285332,
+      "learning_rate": 4.5516865547000186e-07,
+      "loss": 0.8117,
+      "step": 4033
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4135315926976468,
+      "learning_rate": 4.5299962852221934e-07,
+      "loss": 0.7166,
+      "step": 4034
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.543243224298964,
+      "learning_rate": 4.508356622100274e-07,
+      "loss": 0.7809,
+      "step": 4035
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8853825777027486,
+      "learning_rate": 4.486767576804951e-07,
+      "loss": 1.3517,
+      "step": 4036
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.371462681069368,
+      "learning_rate": 4.465229160780027e-07,
+      "loss": 0.6976,
+      "step": 4037
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3214871848350345,
+      "learning_rate": 4.443741385442535e-07,
+      "loss": 0.6489,
+      "step": 4038
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.512212531130697,
+      "learning_rate": 4.422304262182586e-07,
+      "loss": 0.7148,
+      "step": 4039
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5620141268782897,
+      "learning_rate": 4.4009178023635355e-07,
+      "loss": 0.7788,
+      "step": 4040
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.913840520145674,
+      "learning_rate": 4.3795820173217816e-07,
+      "loss": 1.3618,
+      "step": 4041
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.352194824902192,
+      "learning_rate": 4.3582969183669443e-07,
+      "loss": 0.783,
+      "step": 4042
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4103094511451362,
+      "learning_rate": 4.3370625167817205e-07,
+      "loss": 0.7011,
+      "step": 4043
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.999915582101265,
+      "learning_rate": 4.3158788238219726e-07,
+      "loss": 1.4179,
+      "step": 4044
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4590428731864935,
+      "learning_rate": 4.294745850716664e-07,
+      "loss": 0.7149,
+      "step": 4045
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.419100089881493,
+      "learning_rate": 4.2736636086678774e-07,
+      "loss": 0.6658,
+      "step": 4046
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2841283170429996,
+      "learning_rate": 4.252632108850796e-07,
+      "loss": 0.645,
+      "step": 4047
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5180063590535209,
+      "learning_rate": 4.2316513624137245e-07,
+      "loss": 0.8015,
+      "step": 4048
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.0551407808854054,
+      "learning_rate": 4.2107213804780556e-07,
+      "loss": 1.3641,
+      "step": 4049
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.374954087772604,
+      "learning_rate": 4.1898421741382367e-07,
+      "loss": 0.7074,
+      "step": 4050
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.365316510199859,
+      "learning_rate": 4.1690137544618815e-07,
+      "loss": 0.7445,
+      "step": 4051
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4435046429334715,
+      "learning_rate": 4.1482361324895916e-07,
+      "loss": 0.7479,
+      "step": 4052
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.46729618944633,
+      "learning_rate": 4.127509319235112e-07,
+      "loss": 0.7445,
+      "step": 4053
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4579493819751672,
+      "learning_rate": 4.106833325685222e-07,
+      "loss": 0.7148,
+      "step": 4054
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3940019916338309,
+      "learning_rate": 4.0862081627997653e-07,
+      "loss": 0.767,
+      "step": 4055
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9261940681426564,
+      "learning_rate": 4.0656338415116404e-07,
+      "loss": 1.3545,
+      "step": 4056
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.379080065216703,
+      "learning_rate": 4.0451103727268237e-07,
+      "loss": 0.6872,
+      "step": 4057
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.007669772567066,
+      "learning_rate": 4.024637767324291e-07,
+      "loss": 1.2822,
+      "step": 4058
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4501021577389248,
+      "learning_rate": 4.004216036156083e-07,
+      "loss": 0.7225,
+      "step": 4059
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4719977134770392,
+      "learning_rate": 3.9838451900472865e-07,
+      "loss": 0.7475,
+      "step": 4060
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4599201010717917,
+      "learning_rate": 3.963525239795962e-07,
+      "loss": 0.6909,
+      "step": 4061
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6157705915245508,
+      "learning_rate": 3.943256196173262e-07,
+      "loss": 0.8298,
+      "step": 4062
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9849836171829935,
+      "learning_rate": 3.9230380699233015e-07,
+      "loss": 1.3666,
+      "step": 4063
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3127359608207725,
+      "learning_rate": 3.9028708717632204e-07,
+      "loss": 0.6912,
+      "step": 4064
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3506395864396439,
+      "learning_rate": 3.8827546123831773e-07,
+      "loss": 0.7285,
+      "step": 4065
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5023920554353747,
+      "learning_rate": 3.862689302446321e-07,
+      "loss": 0.7151,
+      "step": 4066
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4445991484689102,
+      "learning_rate": 3.842674952588776e-07,
+      "loss": 0.7514,
+      "step": 4067
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3534254184214947,
+      "learning_rate": 3.8227115734196796e-07,
+      "loss": 0.7824,
+      "step": 4068
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2872464681250273,
+      "learning_rate": 3.802799175521155e-07,
+      "loss": 0.7179,
+      "step": 4069
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5088164342284744,
+      "learning_rate": 3.782937769448225e-07,
+      "loss": 0.7317,
+      "step": 4070
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3714434857165043,
+      "learning_rate": 3.7631273657290247e-07,
+      "loss": 0.8092,
+      "step": 4071
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.497045524204096,
+      "learning_rate": 3.74336797486452e-07,
+      "loss": 0.6952,
+      "step": 4072
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.396058536591429,
+      "learning_rate": 3.7236596073287e-07,
+      "loss": 0.6919,
+      "step": 4073
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.502407172473421,
+      "learning_rate": 3.7040022735685076e-07,
+      "loss": 0.6912,
+      "step": 4074
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3691242486517796,
+      "learning_rate": 3.684395984003808e-07,
+      "loss": 0.7674,
+      "step": 4075
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.422210945625025,
+      "learning_rate": 3.664840749027443e-07,
+      "loss": 0.7317,
+      "step": 4076
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4300654462130906,
+      "learning_rate": 3.6453365790051543e-07,
+      "loss": 0.6763,
+      "step": 4077
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4332652559812014,
+      "learning_rate": 3.6258834842756387e-07,
+      "loss": 0.7541,
+      "step": 4078
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4382137857372168,
+      "learning_rate": 3.606481475150536e-07,
+      "loss": 0.7305,
+      "step": 4079
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3667836646217688,
+      "learning_rate": 3.5871305619143647e-07,
+      "loss": 0.6924,
+      "step": 4080
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4051573846746865,
+      "learning_rate": 3.567830754824564e-07,
+      "loss": 0.7337,
+      "step": 4081
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3326569512914945,
+      "learning_rate": 3.5485820641115386e-07,
+      "loss": 0.6989,
+      "step": 4082
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.302779972436697,
+      "learning_rate": 3.5293844999785163e-07,
+      "loss": 0.6947,
+      "step": 4083
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4920503794050826,
+      "learning_rate": 3.5102380726016905e-07,
+      "loss": 0.6716,
+      "step": 4084
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4328662647423391,
+      "learning_rate": 3.49114279213012e-07,
+      "loss": 0.7391,
+      "step": 4085
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5086808296321172,
+      "learning_rate": 3.4720986686857415e-07,
+      "loss": 0.8015,
+      "step": 4086
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9926267651091636,
+      "learning_rate": 3.4531057123634026e-07,
+      "loss": 1.2514,
+      "step": 4087
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4265056668082885,
+      "learning_rate": 3.434163933230794e-07,
+      "loss": 0.7732,
+      "step": 4088
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4134896674987298,
+      "learning_rate": 3.4152733413285175e-07,
+      "loss": 0.6489,
+      "step": 4089
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4756507631923106,
+      "learning_rate": 3.3964339466700194e-07,
+      "loss": 0.6922,
+      "step": 4090
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.358921412770631,
+      "learning_rate": 3.37764575924161e-07,
+      "loss": 0.7475,
+      "step": 4091
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4477224462090275,
+      "learning_rate": 3.3589087890024354e-07,
+      "loss": 0.6886,
+      "step": 4092
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0079292663362946,
+      "learning_rate": 3.34022304588455e-07,
+      "loss": 1.2895,
+      "step": 4093
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5279755876435728,
+      "learning_rate": 3.321588539792775e-07,
+      "loss": 0.7517,
+      "step": 4094
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4339265424381908,
+      "learning_rate": 3.303005280604854e-07,
+      "loss": 0.8114,
+      "step": 4095
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3451837878812025,
+      "learning_rate": 3.284473278171296e-07,
+      "loss": 0.6492,
+      "step": 4096
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4863505754059392,
+      "learning_rate": 3.265992542315488e-07,
+      "loss": 0.7321,
+      "step": 4097
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.416771412096149,
+      "learning_rate": 3.247563082833627e-07,
+      "loss": 0.7295,
+      "step": 4098
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4105897199517,
+      "learning_rate": 3.229184909494709e-07,
+      "loss": 0.7682,
+      "step": 4099
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4310526409852213,
+      "learning_rate": 3.210858032040587e-07,
+      "loss": 0.7284,
+      "step": 4100
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.958756488506208,
+      "learning_rate": 3.1925824601858555e-07,
+      "loss": 1.3103,
+      "step": 4101
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5088044587633476,
+      "learning_rate": 3.1743582036180223e-07,
+      "loss": 0.8866,
+      "step": 4102
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4052524550103105,
+      "learning_rate": 3.1561852719972695e-07,
+      "loss": 0.7265,
+      "step": 4103
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5057824838602274,
+      "learning_rate": 3.1380636749566487e-07,
+      "loss": 0.725,
+      "step": 4104
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4887418699088784,
+      "learning_rate": 3.119993422101997e-07,
+      "loss": 0.7875,
+      "step": 4105
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.896151369245305,
+      "learning_rate": 3.101974523011908e-07,
+      "loss": 1.2934,
+      "step": 4106
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9516426536722216,
+      "learning_rate": 3.0840069872377756e-07,
+      "loss": 1.3781,
+      "step": 4107
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4054010643217583,
+      "learning_rate": 3.0660908243037466e-07,
+      "loss": 0.8004,
+      "step": 4108
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.426256819302936,
+      "learning_rate": 3.048226043706759e-07,
+      "loss": 0.6709,
+      "step": 4109
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3449841763288783,
+      "learning_rate": 3.0304126549165035e-07,
+      "loss": 0.7204,
+      "step": 4110
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5207375643220975,
+      "learning_rate": 3.0126506673754387e-07,
+      "loss": 0.7287,
+      "step": 4111
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3576465991334705,
+      "learning_rate": 2.994940090498732e-07,
+      "loss": 0.6598,
+      "step": 4112
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.40025639806177,
+      "learning_rate": 2.977280933674387e-07,
+      "loss": 0.7395,
+      "step": 4113
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9000728191672538,
+      "learning_rate": 2.9596732062630584e-07,
+      "loss": 1.2547,
+      "step": 4114
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4023183974763398,
+      "learning_rate": 2.9421169175981924e-07,
+      "loss": 0.7166,
+      "step": 4115
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.915429023775152,
+      "learning_rate": 2.9246120769859444e-07,
+      "loss": 1.1803,
+      "step": 4116
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.520960672769277,
+      "learning_rate": 2.9071586937052365e-07,
+      "loss": 0.7403,
+      "step": 4117
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3950127997366475,
+      "learning_rate": 2.8897567770076685e-07,
+      "loss": 0.7713,
+      "step": 4118
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3735397888557808,
+      "learning_rate": 2.8724063361175725e-07,
+      "loss": 0.7852,
+      "step": 4119
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.034869803326848,
+      "learning_rate": 2.855107380232025e-07,
+      "loss": 1.3507,
+      "step": 4120
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9481092687566717,
+      "learning_rate": 2.837859918520758e-07,
+      "loss": 1.3058,
+      "step": 4121
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4030947762015402,
+      "learning_rate": 2.82066396012628e-07,
+      "loss": 0.7673,
+      "step": 4122
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3817770052077738,
+      "learning_rate": 2.803519514163688e-07,
+      "loss": 0.7605,
+      "step": 4123
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.332979617362398,
+      "learning_rate": 2.786426589720914e-07,
+      "loss": 0.7341,
+      "step": 4124
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4373371930425687,
+      "learning_rate": 2.769385195858454e-07,
+      "loss": 0.7626,
+      "step": 4125
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4382223527038542,
+      "learning_rate": 2.752395341609559e-07,
+      "loss": 0.6837,
+      "step": 4126
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5228359510073477,
+      "learning_rate": 2.7354570359801494e-07,
+      "loss": 0.6884,
+      "step": 4127
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9207293594178723,
+      "learning_rate": 2.718570287948807e-07,
+      "loss": 1.3645,
+      "step": 4128
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5058678740499603,
+      "learning_rate": 2.701735106466796e-07,
+      "loss": 0.6284,
+      "step": 4129
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9662404940573934,
+      "learning_rate": 2.684951500458033e-07,
+      "loss": 1.4029,
+      "step": 4130
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3955091878959913,
+      "learning_rate": 2.668219478819134e-07,
+      "loss": 0.7532,
+      "step": 4131
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5029486957384623,
+      "learning_rate": 2.651539050419283e-07,
+      "loss": 0.7462,
+      "step": 4132
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5309438553199681,
+      "learning_rate": 2.6349102241004287e-07,
+      "loss": 0.6857,
+      "step": 4133
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9187014615599087,
+      "learning_rate": 2.6183330086770766e-07,
+      "loss": 1.2544,
+      "step": 4134
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5019648395288272,
+      "learning_rate": 2.6018074129364523e-07,
+      "loss": 0.7941,
+      "step": 4135
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4567513616932468,
+      "learning_rate": 2.585333445638338e-07,
+      "loss": 0.6826,
+      "step": 4136
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4174486734124923,
+      "learning_rate": 2.568911115515205e-07,
+      "loss": 0.7761,
+      "step": 4137
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.487607363036803,
+      "learning_rate": 2.5525404312721325e-07,
+      "loss": 0.6812,
+      "step": 4138
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9579648347878162,
+      "learning_rate": 2.5362214015868246e-07,
+      "loss": 1.3354,
+      "step": 4139
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4426650614194014,
+      "learning_rate": 2.5199540351096064e-07,
+      "loss": 0.7644,
+      "step": 4140
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4060577153687228,
+      "learning_rate": 2.503738340463424e-07,
+      "loss": 0.7557,
+      "step": 4141
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5078366106112415,
+      "learning_rate": 2.487574326243847e-07,
+      "loss": 0.7379,
+      "step": 4142
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.01208206838324,
+      "learning_rate": 2.471462001018976e-07,
+      "loss": 1.3243,
+      "step": 4143
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9487140128779592,
+      "learning_rate": 2.4554013733296244e-07,
+      "loss": 1.29,
+      "step": 4144
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4053941324670502,
+      "learning_rate": 2.4393924516891157e-07,
+      "loss": 0.6943,
+      "step": 4145
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3449879552992052,
+      "learning_rate": 2.4234352445833944e-07,
+      "loss": 0.7627,
+      "step": 4146
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.408369743181901,
+      "learning_rate": 2.4075297604710057e-07,
+      "loss": 0.7401,
+      "step": 4147
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8678613860175668,
+      "learning_rate": 2.3916760077830503e-07,
+      "loss": 1.3265,
+      "step": 4148
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3739418745506127,
+      "learning_rate": 2.375873994923239e-07,
+      "loss": 0.678,
+      "step": 4149
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.408369837176442,
+      "learning_rate": 2.360123730267827e-07,
+      "loss": 0.7515,
+      "step": 4150
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3918272936230058,
+      "learning_rate": 2.3444252221656472e-07,
+      "loss": 0.7046,
+      "step": 4151
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5040102911394677,
+      "learning_rate": 2.3287784789381095e-07,
+      "loss": 0.775,
+      "step": 4152
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5483886824347681,
+      "learning_rate": 2.313183508879191e-07,
+      "loss": 0.777,
+      "step": 4153
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4379625311364264,
+      "learning_rate": 2.2976403202553789e-07,
+      "loss": 0.7522,
+      "step": 4154
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.489949556711756,
+      "learning_rate": 2.2821489213057824e-07,
+      "loss": 0.6947,
+      "step": 4155
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3970620846885162,
+      "learning_rate": 2.2667093202419777e-07,
+      "loss": 0.7195,
+      "step": 4156
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8982922895588505,
+      "learning_rate": 2.2513215252481513e-07,
+      "loss": 1.2924,
+      "step": 4157
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3290404959564135,
+      "learning_rate": 2.2359855444810008e-07,
+      "loss": 0.6717,
+      "step": 4158
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4057015853994168,
+      "learning_rate": 2.220701386069768e-07,
+      "loss": 0.7187,
+      "step": 4159
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3783747920937355,
+      "learning_rate": 2.2054690581161942e-07,
+      "loss": 0.6497,
+      "step": 4160
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5108785611052047,
+      "learning_rate": 2.1902885686945984e-07,
+      "loss": 0.7386,
+      "step": 4161
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.982290902524895,
+      "learning_rate": 2.1751599258517885e-07,
+      "loss": 1.3142,
+      "step": 4162
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9310071020766437,
+      "learning_rate": 2.160083137607072e-07,
+      "loss": 1.3332,
+      "step": 4163
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4826778017860454,
+      "learning_rate": 2.1450582119523221e-07,
+      "loss": 0.8307,
+      "step": 4164
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4836086321551822,
+      "learning_rate": 2.1300851568518577e-07,
+      "loss": 0.74,
+      "step": 4165
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4100017043237678,
+      "learning_rate": 2.115163980242574e-07,
+      "loss": 0.7535,
+      "step": 4166
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.410887016514663,
+      "learning_rate": 2.1002946900338105e-07,
+      "loss": 0.6605,
+      "step": 4167
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4748956086051823,
+      "learning_rate": 2.0854772941074075e-07,
+      "loss": 0.8431,
+      "step": 4168
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.913640572269238,
+      "learning_rate": 2.0707118003177264e-07,
+      "loss": 1.3216,
+      "step": 4169
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.449399498399552,
+      "learning_rate": 2.055998216491595e-07,
+      "loss": 0.7932,
+      "step": 4170
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4610825319449556,
+      "learning_rate": 2.041336550428341e-07,
+      "loss": 0.6607,
+      "step": 4171
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9633112520251683,
+      "learning_rate": 2.0267268098997485e-07,
+      "loss": 1.289,
+      "step": 4172
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3793383405850206,
+      "learning_rate": 2.0121690026500996e-07,
+      "loss": 0.8532,
+      "step": 4173
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5532880778925757,
+      "learning_rate": 1.997663136396122e-07,
+      "loss": 0.6674,
+      "step": 4174
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8295468103778871,
+      "learning_rate": 1.9832092188270534e-07,
+      "loss": 1.3233,
+      "step": 4175
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3497419703926228,
+      "learning_rate": 1.9688072576045436e-07,
+      "loss": 0.7522,
+      "step": 4176
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9188822008879514,
+      "learning_rate": 1.954457260362741e-07,
+      "loss": 1.3341,
+      "step": 4177
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.969214207272168,
+      "learning_rate": 1.9401592347082275e-07,
+      "loss": 1.2954,
+      "step": 4178
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4034191940263019,
+      "learning_rate": 1.925913188220052e-07,
+      "loss": 0.7836,
+      "step": 4179
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4735976351787794,
+      "learning_rate": 1.911719128449685e-07,
+      "loss": 0.6632,
+      "step": 4180
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4039129042957097,
+      "learning_rate": 1.8975770629210632e-07,
+      "loss": 0.7296,
+      "step": 4181
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3578573090648032,
+      "learning_rate": 1.8834869991305682e-07,
+      "loss": 0.6653,
+      "step": 4182
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5457361752379362,
+      "learning_rate": 1.8694489445469922e-07,
+      "loss": 0.6476,
+      "step": 4183
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3717874343246348,
+      "learning_rate": 1.8554629066115827e-07,
+      "loss": 0.6536,
+      "step": 4184
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9109029367777683,
+      "learning_rate": 1.841528892737987e-07,
+      "loss": 1.3379,
+      "step": 4185
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4320380019925338,
+      "learning_rate": 1.8276469103122973e-07,
+      "loss": 0.7013,
+      "step": 4186
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9741589788899594,
+      "learning_rate": 1.813816966693027e-07,
+      "loss": 1.4064,
+      "step": 4187
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.472937647082961,
+      "learning_rate": 1.8000390692110792e-07,
+      "loss": 0.7932,
+      "step": 4188
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.0279200288845427,
+      "learning_rate": 1.7863132251698e-07,
+      "loss": 1.3139,
+      "step": 4189
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.488210485473795,
+      "learning_rate": 1.7726394418449256e-07,
+      "loss": 0.7729,
+      "step": 4190
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4055815763339548,
+      "learning_rate": 1.759017726484602e-07,
+      "loss": 0.7019,
+      "step": 4191
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3959306767675772,
+      "learning_rate": 1.7454480863093758e-07,
+      "loss": 0.7201,
+      "step": 4192
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.385658116620952,
+      "learning_rate": 1.7319305285121823e-07,
+      "loss": 0.6654,
+      "step": 4193
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4546905413842,
+      "learning_rate": 1.7184650602583564e-07,
+      "loss": 0.813,
+      "step": 4194
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9242550952593462,
+      "learning_rate": 1.7050516886856217e-07,
+      "loss": 1.3654,
+      "step": 4195
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9726964994762157,
+      "learning_rate": 1.6916904209040686e-07,
+      "loss": 1.3585,
+      "step": 4196
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.422913414129238,
+      "learning_rate": 1.6783812639962094e-07,
+      "loss": 0.7147,
+      "step": 4197
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5471375361805044,
+      "learning_rate": 1.665124225016901e-07,
+      "loss": 0.735,
+      "step": 4198
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4178782623689157,
+      "learning_rate": 1.651919310993366e-07,
+      "loss": 0.7178,
+      "step": 4199
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9366263778399333,
+      "learning_rate": 1.6387665289252175e-07,
+      "loss": 1.3165,
+      "step": 4200
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3754369880111779,
+      "learning_rate": 1.6256658857844443e-07,
+      "loss": 0.7472,
+      "step": 4201
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3930466500554683,
+      "learning_rate": 1.6126173885153695e-07,
+      "loss": 0.7986,
+      "step": 4202
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4949326959585005,
+      "learning_rate": 1.5996210440346937e-07,
+      "loss": 0.7666,
+      "step": 4203
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4350568416728389,
+      "learning_rate": 1.586676859231473e-07,
+      "loss": 0.7184,
+      "step": 4204
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4273343106217682,
+      "learning_rate": 1.5737848409670853e-07,
+      "loss": 0.7396,
+      "step": 4205
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3839968094047927,
+      "learning_rate": 1.5609449960753087e-07,
+      "loss": 0.7425,
+      "step": 4206
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9366543173152326,
+      "learning_rate": 1.5481573313622102e-07,
+      "loss": 1.3466,
+      "step": 4207
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9332812024792247,
+      "learning_rate": 1.535421853606256e-07,
+      "loss": 1.3552,
+      "step": 4208
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4941360107994424,
+      "learning_rate": 1.5227385695581798e-07,
+      "loss": 0.7722,
+      "step": 4209
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.38339205990637,
+      "learning_rate": 1.5101074859411146e-07,
+      "loss": 0.6983,
+      "step": 4210
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4637251576195398,
+      "learning_rate": 1.4975286094504715e-07,
+      "loss": 0.7352,
+      "step": 4211
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4601919003554502,
+      "learning_rate": 1.4850019467540277e-07,
+      "loss": 0.7275,
+      "step": 4212
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9367826772286885,
+      "learning_rate": 1.4725275044918607e-07,
+      "loss": 1.3956,
+      "step": 4213
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4170090544711218,
+      "learning_rate": 1.4601052892763812e-07,
+      "loss": 0.8352,
+      "step": 4214
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4607330204444227,
+      "learning_rate": 1.4477353076922997e-07,
+      "loss": 0.7719,
+      "step": 4215
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4181925395731794,
+      "learning_rate": 1.4354175662966374e-07,
+      "loss": 0.7489,
+      "step": 4216
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.339588798367399,
+      "learning_rate": 1.4231520716187498e-07,
+      "loss": 0.6931,
+      "step": 4217
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9478502796452926,
+      "learning_rate": 1.410938830160258e-07,
+      "loss": 1.3597,
+      "step": 4218
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3904397283837042,
+      "learning_rate": 1.3987778483951388e-07,
+      "loss": 0.7182,
+      "step": 4219
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3575417410320607,
+      "learning_rate": 1.3866691327696136e-07,
+      "loss": 0.7082,
+      "step": 4220
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3481822762400437,
+      "learning_rate": 1.374612689702226e-07,
+      "loss": 0.6723,
+      "step": 4221
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4518482390629024,
+      "learning_rate": 1.3626085255838196e-07,
+      "loss": 0.7478,
+      "step": 4222
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.483056805795826,
+      "learning_rate": 1.3506566467774928e-07,
+      "loss": 0.6581,
+      "step": 4223
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4003296107851828,
+      "learning_rate": 1.3387570596186673e-07,
+      "loss": 0.7533,
+      "step": 4224
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.38683754206465,
+      "learning_rate": 1.3269097704150192e-07,
+      "loss": 0.7142,
+      "step": 4225
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5387006378255328,
+      "learning_rate": 1.315114785446514e-07,
+      "loss": 0.7811,
+      "step": 4226
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.397663088128719,
+      "learning_rate": 1.3033721109653842e-07,
+      "loss": 0.7148,
+      "step": 4227
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3375982868975698,
+      "learning_rate": 1.291681753196161e-07,
+      "loss": 0.7254,
+      "step": 4228
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.459388329246823,
+      "learning_rate": 1.2800437183355886e-07,
+      "loss": 0.8078,
+      "step": 4229
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3917044658462414,
+      "learning_rate": 1.2684580125527423e-07,
+      "loss": 0.796,
+      "step": 4230
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3569296495153345,
+      "learning_rate": 1.2569246419888993e-07,
+      "loss": 0.7857,
+      "step": 4231
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.411791663553017,
+      "learning_rate": 1.2454436127576464e-07,
+      "loss": 0.657,
+      "step": 4232
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.928546231867017,
+      "learning_rate": 1.2340149309447824e-07,
+      "loss": 1.3072,
+      "step": 4233
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.970097859715844,
+      "learning_rate": 1.2226386026083835e-07,
+      "loss": 1.2748,
+      "step": 4234
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4092873759735722,
+      "learning_rate": 1.2113146337787817e-07,
+      "loss": 0.7793,
+      "step": 4235
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6021221202768638,
+      "learning_rate": 1.2000430304585088e-07,
+      "loss": 0.7095,
+      "step": 4236
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.543502365807497,
+      "learning_rate": 1.1888237986224071e-07,
+      "loss": 0.7413,
+      "step": 4237
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4615202627595647,
+      "learning_rate": 1.1776569442174868e-07,
+      "loss": 0.7052,
+      "step": 4238
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9917455290125934,
+      "learning_rate": 1.1665424731630681e-07,
+      "loss": 1.2831,
+      "step": 4239
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.0645838617554166,
+      "learning_rate": 1.1554803913506386e-07,
+      "loss": 1.275,
+      "step": 4240
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3899080690930816,
+      "learning_rate": 1.1444707046439407e-07,
+      "loss": 0.7049,
+      "step": 4241
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3906737520302732,
+      "learning_rate": 1.133513418878962e-07,
+      "loss": 0.7383,
+      "step": 4242
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.407635738474132,
+      "learning_rate": 1.1226085398638898e-07,
+      "loss": 0.7845,
+      "step": 4243
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3373792008109522,
+      "learning_rate": 1.1117560733791222e-07,
+      "loss": 0.7337,
+      "step": 4244
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5149447666662885,
+      "learning_rate": 1.1009560251773132e-07,
+      "loss": 0.7628,
+      "step": 4245
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4346592902850641,
+      "learning_rate": 1.0902084009833058e-07,
+      "loss": 0.7895,
+      "step": 4246
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.444354375438625,
+      "learning_rate": 1.0795132064941427e-07,
+      "loss": 0.711,
+      "step": 4247
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4473688968992058,
+      "learning_rate": 1.0688704473791112e-07,
+      "loss": 0.7324,
+      "step": 4248
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4218304223836788,
+      "learning_rate": 1.058280129279654e-07,
+      "loss": 0.6839,
+      "step": 4249
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.309762927535086,
+      "learning_rate": 1.0477422578094698e-07,
+      "loss": 0.7388,
+      "step": 4250
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4935100400273735,
+      "learning_rate": 1.0372568385544234e-07,
+      "loss": 0.8414,
+      "step": 4251
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3905558734330699,
+      "learning_rate": 1.0268238770725691e-07,
+      "loss": 0.6557,
+      "step": 4252
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.388461104847511,
+      "learning_rate": 1.0164433788941941e-07,
+      "loss": 0.768,
+      "step": 4253
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3565909789662165,
+      "learning_rate": 1.0061153495217413e-07,
+      "loss": 0.8106,
+      "step": 4254
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9586770054035063,
+      "learning_rate": 9.958397944298426e-08,
+      "loss": 1.2866,
+      "step": 4255
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4720090119385971,
+      "learning_rate": 9.856167190653298e-08,
+      "loss": 0.7734,
+      "step": 4256
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5111190417882079,
+      "learning_rate": 9.754461288472128e-08,
+      "loss": 0.7011,
+      "step": 4257
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4230086346234736,
+      "learning_rate": 9.653280291666678e-08,
+      "loss": 0.6665,
+      "step": 4258
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4100505517988968,
+      "learning_rate": 9.552624253870713e-08,
+      "loss": 0.707,
+      "step": 4259
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3910943952091568,
+      "learning_rate": 9.45249322843933e-08,
+      "loss": 0.6316,
+      "step": 4260
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2631858382208252,
+      "learning_rate": 9.352887268449962e-08,
+      "loss": 0.746,
+      "step": 4261
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5282524140904232,
+      "learning_rate": 9.253806426700928e-08,
+      "loss": 0.7444,
+      "step": 4262
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9940054414781845,
+      "learning_rate": 9.155250755712885e-08,
+      "loss": 1.2958,
+      "step": 4263
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4792244026954677,
+      "learning_rate": 9.05722030772771e-08,
+      "loss": 0.7731,
+      "step": 4264
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3152658662025256,
+      "learning_rate": 8.959715134708946e-08,
+      "loss": 0.6997,
+      "step": 4265
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4888305632365264,
+      "learning_rate": 8.862735288341918e-08,
+      "loss": 0.8072,
+      "step": 4266
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3837349693176884,
+      "learning_rate": 8.766280820033168e-08,
+      "loss": 0.6851,
+      "step": 4267
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4935005679763056,
+      "learning_rate": 8.670351780910913e-08,
+      "loss": 0.7783,
+      "step": 4268
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.888524469493043,
+      "learning_rate": 8.574948221824808e-08,
+      "loss": 1.3583,
+      "step": 4269
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3589925268452934,
+      "learning_rate": 8.480070193346069e-08,
+      "loss": 0.7043,
+      "step": 4270
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3923317303103435,
+      "learning_rate": 8.385717745767242e-08,
+      "loss": 0.7625,
+      "step": 4271
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9383850818304864,
+      "learning_rate": 8.291890929102209e-08,
+      "loss": 1.3063,
+      "step": 4272
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.005668073902393,
+      "learning_rate": 8.198589793086409e-08,
+      "loss": 1.2967,
+      "step": 4273
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4415811615699587,
+      "learning_rate": 8.105814387176503e-08,
+      "loss": 0.7918,
+      "step": 4274
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4111245841344042,
+      "learning_rate": 8.013564760550374e-08,
+      "loss": 0.6682,
+      "step": 4275
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.984318043954892,
+      "learning_rate": 7.921840962107574e-08,
+      "loss": 1.3006,
+      "step": 4276
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3711594000779275,
+      "learning_rate": 7.830643040468655e-08,
+      "loss": 0.7892,
+      "step": 4277
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.389011807287051,
+      "learning_rate": 7.73997104397517e-08,
+      "loss": 0.7305,
+      "step": 4278
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.96277214295514,
+      "learning_rate": 7.64982502069056e-08,
+      "loss": 1.312,
+      "step": 4279
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9769230976044365,
+      "learning_rate": 7.560205018398714e-08,
+      "loss": 1.2875,
+      "step": 4280
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3860360268832712,
+      "learning_rate": 7.471111084605187e-08,
+      "loss": 0.7117,
+      "step": 4281
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4470133047640021,
+      "learning_rate": 7.382543266536646e-08,
+      "loss": 0.7977,
+      "step": 4282
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4561707038265939,
+      "learning_rate": 7.294501611140536e-08,
+      "loss": 0.6574,
+      "step": 4283
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4142209008706554,
+      "learning_rate": 7.20698616508586e-08,
+      "loss": 0.7198,
+      "step": 4284
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.595173640937556,
+      "learning_rate": 7.119996974762178e-08,
+      "loss": 0.8037,
+      "step": 4285
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.527788581321623,
+      "learning_rate": 7.033534086280714e-08,
+      "loss": 0.7966,
+      "step": 4286
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3729919527969672,
+      "learning_rate": 6.947597545473028e-08,
+      "loss": 0.6988,
+      "step": 4287
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5023847495740166,
+      "learning_rate": 6.86218739789235e-08,
+      "loss": 0.6946,
+      "step": 4288
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5753130061267462,
+      "learning_rate": 6.777303688812132e-08,
+      "loss": 0.8376,
+      "step": 4289
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9366045118502793,
+      "learning_rate": 6.692946463227601e-08,
+      "loss": 1.3027,
+      "step": 4290
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9364508289507885,
+      "learning_rate": 6.609115765854324e-08,
+      "loss": 1.3066,
+      "step": 4291
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.305239795818099,
+      "learning_rate": 6.525811641128865e-08,
+      "loss": 0.7753,
+      "step": 4292
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4617262712266739,
+      "learning_rate": 6.443034133208903e-08,
+      "loss": 0.6852,
+      "step": 4293
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3891758663784415,
+      "learning_rate": 6.360783285972671e-08,
+      "loss": 0.7499,
+      "step": 4294
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9251831428881925,
+      "learning_rate": 6.279059143019295e-08,
+      "loss": 1.3597,
+      "step": 4295
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3439277534524694,
+      "learning_rate": 6.19786174766901e-08,
+      "loss": 0.6857,
+      "step": 4296
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.46793520143216,
+      "learning_rate": 6.117191142962609e-08,
+      "loss": 0.717,
+      "step": 4297
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.916971436290364,
+      "learning_rate": 6.037047371661331e-08,
+      "loss": 1.323,
+      "step": 4298
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4156955204443664,
+      "learning_rate": 5.95743047624775e-08,
+      "loss": 0.8153,
+      "step": 4299
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3955662986353343,
+      "learning_rate": 5.878340498924773e-08,
+      "loss": 0.762,
+      "step": 4300
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3766383386681103,
+      "learning_rate": 5.799777481616087e-08,
+      "loss": 0.6855,
+      "step": 4301
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9340757079739288,
+      "learning_rate": 5.7217414659660466e-08,
+      "loss": 1.2758,
+      "step": 4302
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.32533261522436,
+      "learning_rate": 5.644232493339785e-08,
+      "loss": 0.735,
+      "step": 4303
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5155698289189787,
+      "learning_rate": 5.5672506048228825e-08,
+      "loss": 0.7763,
+      "step": 4304
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.423643931472469,
+      "learning_rate": 5.490795841221475e-08,
+      "loss": 0.7793,
+      "step": 4305
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9753492741171954,
+      "learning_rate": 5.4148682430625874e-08,
+      "loss": 1.2932,
+      "step": 4306
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.454832731136037,
+      "learning_rate": 5.339467850593472e-08,
+      "loss": 0.7897,
+      "step": 4307
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4456978535119616,
+      "learning_rate": 5.2645947037820445e-08,
+      "loss": 0.7624,
+      "step": 4308
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4278422692790385,
+      "learning_rate": 5.190248842316892e-08,
+      "loss": 0.673,
+      "step": 4309
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4517535496338048,
+      "learning_rate": 5.116430305606934e-08,
+      "loss": 0.7405,
+      "step": 4310
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.420602694694665,
+      "learning_rate": 5.043139132781538e-08,
+      "loss": 0.7187,
+      "step": 4311
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4666791921617979,
+      "learning_rate": 4.9703753626906273e-08,
+      "loss": 0.796,
+      "step": 4312
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4085761314794736,
+      "learning_rate": 4.8981390339045695e-08,
+      "loss": 0.6641,
+      "step": 4313
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4122167175725555,
+      "learning_rate": 4.826430184714071e-08,
+      "loss": 0.7548,
+      "step": 4314
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4427638131901377,
+      "learning_rate": 4.755248853130279e-08,
+      "loss": 0.7391,
+      "step": 4315
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4037085790666375,
+      "learning_rate": 4.6845950768846796e-08,
+      "loss": 0.8092,
+      "step": 4316
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4875663527251757,
+      "learning_rate": 4.6144688934292025e-08,
+      "loss": 0.8155,
+      "step": 4317
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.393704528722436,
+      "learning_rate": 4.544870339936003e-08,
+      "loss": 0.7205,
+      "step": 4318
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.931422437131606,
+      "learning_rate": 4.475799453297569e-08,
+      "loss": 1.3225,
+      "step": 4319
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4141094757819859,
+      "learning_rate": 4.407256270126725e-08,
+      "loss": 0.6703,
+      "step": 4320
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4523581460454733,
+      "learning_rate": 4.339240826756741e-08,
+      "loss": 0.8022,
+      "step": 4321
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4583930855794276,
+      "learning_rate": 4.271753159240777e-08,
+      "loss": 0.7261,
+      "step": 4322
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2801764188974825,
+      "learning_rate": 4.20479330335255e-08,
+      "loss": 0.6574,
+      "step": 4323
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4916575992471908,
+      "learning_rate": 4.138361294585669e-08,
+      "loss": 0.7888,
+      "step": 4324
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4862073769797035,
+      "learning_rate": 4.072457168154298e-08,
+      "loss": 0.7607,
+      "step": 4325
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8845914583596834,
+      "learning_rate": 4.007080958992715e-08,
+      "loss": 1.349,
+      "step": 4326
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4809430295594275,
+      "learning_rate": 3.942232701755089e-08,
+      "loss": 0.7445,
+      "step": 4327
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4188012336777733,
+      "learning_rate": 3.8779124308161445e-08,
+      "loss": 0.6726,
+      "step": 4328
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3909630750068291,
+      "learning_rate": 3.8141201802701643e-08,
+      "loss": 0.7382,
+      "step": 4329
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4004572955986005,
+      "learning_rate": 3.7508559839323224e-08,
+      "loss": 0.642,
+      "step": 4330
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.091179175788726,
+      "learning_rate": 3.688119875337126e-08,
+      "loss": 1.3544,
+      "step": 4331
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5272622825310385,
+      "learning_rate": 3.6259118877395305e-08,
+      "loss": 0.7819,
+      "step": 4332
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8242389928446412,
+      "learning_rate": 3.564232054114602e-08,
+      "loss": 1.3278,
+      "step": 4333
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3950802561140705,
+      "learning_rate": 3.503080407157411e-08,
+      "loss": 0.6392,
+      "step": 4334
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4351770236354655,
+      "learning_rate": 3.442456979282693e-08,
+      "loss": 0.6778,
+      "step": 4335
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3860309072469172,
+      "learning_rate": 3.382361802625633e-08,
+      "loss": 0.7127,
+      "step": 4336
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.01658163367919,
+      "learning_rate": 3.322794909041305e-08,
+      "loss": 1.3104,
+      "step": 4337
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3887853310035483,
+      "learning_rate": 3.2637563301045615e-08,
+      "loss": 0.703,
+      "step": 4338
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5457320395897751,
+      "learning_rate": 3.2052460971104814e-08,
+      "loss": 0.7006,
+      "step": 4339
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.376717346458425,
+      "learning_rate": 3.1472642410738106e-08,
+      "loss": 0.7089,
+      "step": 4340
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4343138611783552,
+      "learning_rate": 3.0898107927294084e-08,
+      "loss": 0.6987,
+      "step": 4341
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3507573273334086,
+      "learning_rate": 3.0328857825318024e-08,
+      "loss": 0.7414,
+      "step": 4342
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.428576411386499,
+      "learning_rate": 2.9764892406558554e-08,
+      "loss": 0.7112,
+      "step": 4343
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.488549788038982,
+      "learning_rate": 2.9206211969958764e-08,
+      "loss": 0.788,
+      "step": 4344
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.402679859502115,
+      "learning_rate": 2.8652816811660656e-08,
+      "loss": 0.7077,
+      "step": 4345
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5306214522010282,
+      "learning_rate": 2.8104707225007356e-08,
+      "loss": 0.769,
+      "step": 4346
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4417426419354544,
+      "learning_rate": 2.7561883500537568e-08,
+      "loss": 0.7278,
+      "step": 4347
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4167966411201016,
+      "learning_rate": 2.7024345925990015e-08,
+      "loss": 0.7135,
+      "step": 4348
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3265380561118865,
+      "learning_rate": 2.6492094786300103e-08,
+      "loss": 0.7184,
+      "step": 4349
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.0566929370529796,
+      "learning_rate": 2.5965130363601042e-08,
+      "loss": 1.2739,
+      "step": 4350
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8620787710409006,
+      "learning_rate": 2.5443452937222723e-08,
+      "loss": 1.3904,
+      "step": 4351
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5228197447795067,
+      "learning_rate": 2.4927062783697275e-08,
+      "loss": 0.7845,
+      "step": 4352
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5054696341908322,
+      "learning_rate": 2.4415960176747965e-08,
+      "loss": 0.7043,
+      "step": 4353
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4604798706557043,
+      "learning_rate": 2.3910145387299188e-08,
+      "loss": 0.7572,
+      "step": 4354
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9366454378630193,
+      "learning_rate": 2.3409618683469803e-08,
+      "loss": 1.2655,
+      "step": 4355
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4620076684489034,
+      "learning_rate": 2.291438033057869e-08,
+      "loss": 0.746,
+      "step": 4356
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3649678501450464,
+      "learning_rate": 2.242443059113919e-08,
+      "loss": 0.6974,
+      "step": 4357
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3979670665338662,
+      "learning_rate": 2.1939769724861338e-08,
+      "loss": 0.6982,
+      "step": 4358
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4145229684894365,
+      "learning_rate": 2.146039798865407e-08,
+      "loss": 0.7367,
+      "step": 4359
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4470592306487995,
+      "learning_rate": 2.098631563661746e-08,
+      "loss": 0.7246,
+      "step": 4360
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.375570607476308,
+      "learning_rate": 2.0517522920054932e-08,
+      "loss": 0.6921,
+      "step": 4361
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9814583984585215,
+      "learning_rate": 2.0054020087458824e-08,
+      "loss": 1.2884,
+      "step": 4362
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8851916044673527,
+      "learning_rate": 1.9595807384523713e-08,
+      "loss": 1.2928,
+      "step": 4363
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3776714442777607,
+      "learning_rate": 1.9142885054136418e-08,
+      "loss": 0.7537,
+      "step": 4364
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4766061990062622,
+      "learning_rate": 1.8695253336378228e-08,
+      "loss": 0.762,
+      "step": 4365
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4974428025313113,
+      "learning_rate": 1.8252912468530448e-08,
+      "loss": 0.7621,
+      "step": 4366
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3409201317264512,
+      "learning_rate": 1.781586268506663e-08,
+      "loss": 0.7629,
+      "step": 4367
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3682566597266057,
+      "learning_rate": 1.7384104217654796e-08,
+      "loss": 0.7448,
+      "step": 4368
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4094046755857796,
+      "learning_rate": 1.6957637295161867e-08,
+      "loss": 0.7375,
+      "step": 4369
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4830788231080352,
+      "learning_rate": 1.6536462143647016e-08,
+      "loss": 0.685,
+      "step": 4370
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8619444555808804,
+      "learning_rate": 1.6120578986364987e-08,
+      "loss": 1.2934,
+      "step": 4371
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3665320719235048,
+      "learning_rate": 1.5709988043765e-08,
+      "loss": 0.7642,
+      "step": 4372
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.42915944572531,
+      "learning_rate": 1.5304689533491844e-08,
+      "loss": 0.7394,
+      "step": 4373
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9486510385263365,
+      "learning_rate": 1.4904683670384777e-08,
+      "loss": 1.3655,
+      "step": 4374
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9593249897595066,
+      "learning_rate": 1.4509970666477524e-08,
+      "loss": 1.3745,
+      "step": 4375
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4323857257066044,
+      "learning_rate": 1.4120550730997163e-08,
+      "loss": 0.7342,
+      "step": 4376
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4944861645921865,
+      "learning_rate": 1.373642407036635e-08,
+      "loss": 0.7573,
+      "step": 4377
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4800803366106596,
+      "learning_rate": 1.3357590888199989e-08,
+      "loss": 0.7613,
+      "step": 4378
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2970132212066432,
+      "learning_rate": 1.2984051385310782e-08,
+      "loss": 0.7023,
+      "step": 4379
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.012664619651251,
+      "learning_rate": 1.261580575970256e-08,
+      "loss": 1.3676,
+      "step": 4380
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9508611913495877,
+      "learning_rate": 1.2252854206572518e-08,
+      "loss": 1.3523,
+      "step": 4381
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5027596590226264,
+      "learning_rate": 1.1895196918312313e-08,
+      "loss": 0.7902,
+      "step": 4382
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.500975811169588,
+      "learning_rate": 1.154283408450807e-08,
+      "loss": 0.702,
+      "step": 4383
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4693755545029195,
+      "learning_rate": 1.119576589193927e-08,
+      "loss": 0.7299,
+      "step": 4384
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4453882668437799,
+      "learning_rate": 1.0853992524577639e-08,
+      "loss": 0.7826,
+      "step": 4385
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3634439376285423,
+      "learning_rate": 1.0517514163590482e-08,
+      "loss": 0.7454,
+      "step": 4386
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.446370348416553,
+      "learning_rate": 1.0186330987335124e-08,
+      "loss": 0.7813,
+      "step": 4387
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4987324808866234,
+      "learning_rate": 9.860443171364476e-09,
+      "loss": 0.7343,
+      "step": 4388
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2937806287956553,
+      "learning_rate": 9.539850888423685e-09,
+      "loss": 0.5843,
+      "step": 4389
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4001643888883486,
+      "learning_rate": 9.224554308452371e-09,
+      "loss": 0.7628,
+      "step": 4390
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.356610913968446,
+      "learning_rate": 8.914553598580178e-09,
+      "loss": 0.6998,
+      "step": 4391
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.425008347560418,
+      "learning_rate": 8.609848923131215e-09,
+      "loss": 0.7597,
+      "step": 4392
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4438856364811463,
+      "learning_rate": 8.310440443621837e-09,
+      "loss": 0.7578,
+      "step": 4393
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9382761947155085,
+      "learning_rate": 8.016328318761757e-09,
+      "loss": 1.3621,
+      "step": 4394
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.350698477534222,
+      "learning_rate": 7.727512704452933e-09,
+      "loss": 0.6597,
+      "step": 4395
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4445809004359822,
+      "learning_rate": 7.443993753789569e-09,
+      "loss": 0.7621,
+      "step": 4396
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4109740861798559,
+      "learning_rate": 7.165771617058115e-09,
+      "loss": 0.7155,
+      "step": 4397
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4933959625403852,
+      "learning_rate": 6.892846441738376e-09,
+      "loss": 0.7325,
+      "step": 4398
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4084943150818996,
+      "learning_rate": 6.625218372499076e-09,
+      "loss": 0.7425,
+      "step": 4399
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9859552808171073,
+      "learning_rate": 6.362887551205621e-09,
+      "loss": 1.3192,
+      "step": 4400
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.511499677609229,
+      "learning_rate": 6.1058541169123355e-09,
+      "loss": 0.7769,
+      "step": 4401
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9711586056232593,
+      "learning_rate": 5.854118205865789e-09,
+      "loss": 1.3205,
+      "step": 4402
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4087390635753199,
+      "learning_rate": 5.6076799515070166e-09,
+      "loss": 0.6987,
+      "step": 4403
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9327744921389751,
+      "learning_rate": 5.366539484464861e-09,
+      "loss": 1.3149,
+      "step": 4404
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5117279068171705,
+      "learning_rate": 5.130696932564849e-09,
+      "loss": 0.8493,
+      "step": 4405
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4434357693465236,
+      "learning_rate": 4.900152420818094e-09,
+      "loss": 0.7254,
+      "step": 4406
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.0361204062827007,
+      "learning_rate": 4.674906071433505e-09,
+      "loss": 1.3232,
+      "step": 4407
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2837066811383164,
+      "learning_rate": 4.454958003807797e-09,
+      "loss": 0.6814,
+      "step": 4408
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9840724711015116,
+      "learning_rate": 4.24030833452882e-09,
+      "loss": 1.3752,
+      "step": 4409
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4622859259205097,
+      "learning_rate": 4.030957177380002e-09,
+      "loss": 0.7738,
+      "step": 4410
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.332170466761563,
+      "learning_rate": 3.826904643330354e-09,
+      "loss": 0.7049,
+      "step": 4411
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4575037554338257,
+      "learning_rate": 3.6281508405455746e-09,
+      "loss": 0.7267,
+      "step": 4412
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.911674855881431,
+      "learning_rate": 3.434695874379168e-09,
+      "loss": 1.3773,
+      "step": 4413
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5220246809493954,
+      "learning_rate": 3.246539847376884e-09,
+      "loss": 0.6493,
+      "step": 4414
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3144277372657855,
+      "learning_rate": 3.0636828592756076e-09,
+      "loss": 0.6822,
+      "step": 4415
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9179495766853225,
+      "learning_rate": 2.8861250070044698e-09,
+      "loss": 1.2883,
+      "step": 4416
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2784917067854624,
+      "learning_rate": 2.7138663846826284e-09,
+      "loss": 0.6762,
+      "step": 4417
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.363244712287765,
+      "learning_rate": 2.5469070836181555e-09,
+      "loss": 0.6742,
+      "step": 4418
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5103468373793862,
+      "learning_rate": 2.385247192314699e-09,
+      "loss": 0.7639,
+      "step": 4419
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3792218029523906,
+      "learning_rate": 2.228886796462604e-09,
+      "loss": 0.6825,
+      "step": 4420
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9826722075802985,
+      "learning_rate": 2.0778259789455692e-09,
+      "loss": 1.2576,
+      "step": 4421
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4606910783938551,
+      "learning_rate": 1.9320648198362104e-09,
+      "loss": 0.7729,
+      "step": 4422
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5130043275502547,
+      "learning_rate": 1.7916033964016088e-09,
+      "loss": 0.7988,
+      "step": 4423
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.251325804599792,
+      "learning_rate": 1.656441783093321e-09,
+      "loss": 0.6763,
+      "step": 4424
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5456674212393084,
+      "learning_rate": 1.5265800515607e-09,
+      "loss": 0.7589,
+      "step": 4425
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9612649178978119,
+      "learning_rate": 1.4020182706386831e-09,
+      "loss": 1.3185,
+      "step": 4426
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4687323702053499,
+      "learning_rate": 1.2827565063533443e-09,
+      "loss": 0.7364,
+      "step": 4427
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.425155590731795,
+      "learning_rate": 1.1687948219252232e-09,
+      "loss": 0.7031,
+      "step": 4428
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.0025539507422185,
+      "learning_rate": 1.0601332777604445e-09,
+      "loss": 1.3378,
+      "step": 4429
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4088031336613853,
+      "learning_rate": 9.567719314595992e-10,
+      "loss": 0.6329,
+      "step": 4430
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3261920770917788,
+      "learning_rate": 8.587108378099729e-10,
+      "loss": 0.6751,
+      "step": 4431
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.019116605962925,
+      "learning_rate": 7.659500487933181e-10,
+      "loss": 1.2844,
+      "step": 4432
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.360555582136866,
+      "learning_rate": 6.784896135780816e-10,
+      "loss": 0.6803,
+      "step": 4433
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.496172883913503,
+      "learning_rate": 5.963295785271772e-10,
+      "loss": 0.7573,
+      "step": 4434
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.378918506523109,
+      "learning_rate": 5.194699871902132e-10,
+      "loss": 0.653,
+      "step": 4435
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5387305068977315,
+      "learning_rate": 4.479108803079335e-10,
+      "loss": 0.7254,
+      "step": 4436
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.8772185302807445,
+      "learning_rate": 3.8165229581443863e-10,
+      "loss": 1.2749,
+      "step": 4437
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3684830745647678,
+      "learning_rate": 3.2069426882941345e-10,
+      "loss": 0.7185,
+      "step": 4438
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4258663640400022,
+      "learning_rate": 2.650368316658991e-10,
+      "loss": 0.7222,
+      "step": 4439
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.9523389211570774,
+      "learning_rate": 2.1468001382807247e-10,
+      "loss": 1.2961,
+      "step": 4440
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.9972617781398996,
+      "learning_rate": 1.6962384200680525e-10,
+      "loss": 1.3418,
+      "step": 4441
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.9282634919649675,
+      "learning_rate": 1.2986834008632543e-10,
+      "loss": 1.2606,
+      "step": 4442
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3498429187424918,
+      "learning_rate": 9.541352914088642e-11,
+      "loss": 0.741,
+      "step": 4443
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.407150723587064,
+      "learning_rate": 6.625942743254676e-11,
+      "loss": 0.7901,
+      "step": 4444
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.388569573432166,
+      "learning_rate": 4.2406050415610966e-11,
+      "loss": 0.6817,
+      "step": 4445
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3667779237941506,
+      "learning_rate": 2.3853410734409054e-11,
+      "loss": 0.7386,
+      "step": 4446
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3604032877777263,
+      "learning_rate": 1.060151822329658e-11,
+      "loss": 0.6758,
+      "step": 4447
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3407601423873798,
+      "learning_rate": 2.650379906654621e-12,
+      "loss": 0.68,
+      "step": 4448
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6715519150997455,
+      "learning_rate": 0.0,
+      "loss": 0.8032,
+      "step": 4449
+    },
+    {
+      "epoch": 1.0,
+      "step": 4449,
+      "total_flos": 3059311801892864.0,
+      "train_loss": 0.9097134597410269,
+      "train_runtime": 23240.9777,
+      "train_samples_per_second": 49.007,
+      "train_steps_per_second": 0.191
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 4449,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 2000,
+  "total_flos": 3059311801892864.0,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}