diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,43472 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.9999194392975107,
+  "eval_steps": 500,
+  "global_step": 6206,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.53196797006627,
+      "learning_rate": 1.0695187165775401e-07,
+      "loss": 1.541,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.311062312027166,
+      "learning_rate": 2.1390374331550802e-07,
+      "loss": 1.6299,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.436112959736087,
+      "learning_rate": 3.208556149732621e-07,
+      "loss": 1.5385,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.335068617726015,
+      "learning_rate": 4.2780748663101604e-07,
+      "loss": 1.6662,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.909720869133421,
+      "learning_rate": 5.347593582887701e-07,
+      "loss": 1.5632,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.666190490609123,
+      "learning_rate": 6.417112299465242e-07,
+      "loss": 1.7323,
+      "step": 6
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.435249428878192,
+      "learning_rate": 7.486631016042781e-07,
+      "loss": 1.5217,
+      "step": 7
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.6538306097949063,
+      "learning_rate": 8.556149732620321e-07,
+      "loss": 0.283,
+      "step": 8
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.340246462572206,
+      "learning_rate": 9.625668449197862e-07,
+      "loss": 1.6522,
+      "step": 9
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.123407602103352,
+      "learning_rate": 1.0695187165775401e-06,
+      "loss": 1.5194,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.011151051186081,
+      "learning_rate": 1.1764705882352942e-06,
+      "loss": 1.7549,
+      "step": 11
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.7412442035271506,
+      "learning_rate": 1.2834224598930483e-06,
+      "loss": 0.2853,
+      "step": 12
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.479597583899963,
+      "learning_rate": 1.3903743315508022e-06,
+      "loss": 1.6564,
+      "step": 13
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.287917688005597,
+      "learning_rate": 1.4973262032085562e-06,
+      "loss": 1.4943,
+      "step": 14
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.45375430027414,
+      "learning_rate": 1.6042780748663103e-06,
+      "loss": 1.552,
+      "step": 15
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.844408782956599,
+      "learning_rate": 1.7112299465240642e-06,
+      "loss": 1.4518,
+      "step": 16
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.2620626062752733,
+      "learning_rate": 1.8181818181818183e-06,
+      "loss": 1.4307,
+      "step": 17
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.722124968461394,
+      "learning_rate": 1.9251336898395724e-06,
+      "loss": 1.431,
+      "step": 18
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.1641531001176832,
+      "learning_rate": 2.0320855614973265e-06,
+      "loss": 1.3461,
+      "step": 19
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.6808611887113956,
+      "learning_rate": 2.1390374331550802e-06,
+      "loss": 1.583,
+      "step": 20
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.298234965142126,
+      "learning_rate": 2.2459893048128343e-06,
+      "loss": 1.3332,
+      "step": 21
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.4592974966390613,
+      "learning_rate": 2.3529411764705885e-06,
+      "loss": 1.4626,
+      "step": 22
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.7742613771782929,
+      "learning_rate": 2.4598930481283426e-06,
+      "loss": 0.2514,
+      "step": 23
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.038969709014357,
+      "learning_rate": 2.5668449197860967e-06,
+      "loss": 1.4968,
+      "step": 24
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.2174413710426824,
+      "learning_rate": 2.673796791443851e-06,
+      "loss": 1.3723,
+      "step": 25
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.7424892182432326,
+      "learning_rate": 2.7807486631016045e-06,
+      "loss": 0.3019,
+      "step": 26
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.9059395919149864,
+      "learning_rate": 2.8877005347593586e-06,
+      "loss": 1.3785,
+      "step": 27
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.2769614466398784,
+      "learning_rate": 2.9946524064171123e-06,
+      "loss": 1.3185,
+      "step": 28
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.225934634441806,
+      "learning_rate": 3.101604278074867e-06,
+      "loss": 1.3484,
+      "step": 29
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.2311105656532804,
+      "learning_rate": 3.2085561497326205e-06,
+      "loss": 1.2698,
+      "step": 30
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.7313827859365953,
+      "learning_rate": 3.3155080213903747e-06,
+      "loss": 1.2595,
+      "step": 31
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.4180644252133177,
+      "learning_rate": 3.4224598930481284e-06,
+      "loss": 1.2453,
+      "step": 32
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9719030566418552,
+      "learning_rate": 3.529411764705883e-06,
+      "loss": 1.2943,
+      "step": 33
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.281268512474351,
+      "learning_rate": 3.6363636363636366e-06,
+      "loss": 1.2718,
+      "step": 34
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1712818913925984,
+      "learning_rate": 3.7433155080213907e-06,
+      "loss": 1.1555,
+      "step": 35
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1235626246995163,
+      "learning_rate": 3.850267379679145e-06,
+      "loss": 1.2219,
+      "step": 36
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.222512457575872,
+      "learning_rate": 3.957219251336899e-06,
+      "loss": 1.27,
+      "step": 37
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9888253675939873,
+      "learning_rate": 4.064171122994653e-06,
+      "loss": 1.26,
+      "step": 38
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8742132473427955,
+      "learning_rate": 4.171122994652407e-06,
+      "loss": 1.2675,
+      "step": 39
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7767172264587079,
+      "learning_rate": 4.2780748663101604e-06,
+      "loss": 1.1135,
+      "step": 40
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.194225907803369,
+      "learning_rate": 4.385026737967915e-06,
+      "loss": 1.2028,
+      "step": 41
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.021768498379485,
+      "learning_rate": 4.491978609625669e-06,
+      "loss": 1.1143,
+      "step": 42
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.951684289763473,
+      "learning_rate": 4.598930481283423e-06,
+      "loss": 1.137,
+      "step": 43
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3684429555053232,
+      "learning_rate": 4.705882352941177e-06,
+      "loss": 1.2509,
+      "step": 44
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0715721967418115,
+      "learning_rate": 4.812834224598931e-06,
+      "loss": 1.2642,
+      "step": 45
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9425322320312466,
+      "learning_rate": 4.919786096256685e-06,
+      "loss": 1.1564,
+      "step": 46
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.860974752620202,
+      "learning_rate": 5.026737967914439e-06,
+      "loss": 1.1736,
+      "step": 47
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.062549334896917,
+      "learning_rate": 5.133689839572193e-06,
+      "loss": 1.2111,
+      "step": 48
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.925396788988329,
+      "learning_rate": 5.240641711229947e-06,
+      "loss": 1.2041,
+      "step": 49
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1103645440814014,
+      "learning_rate": 5.347593582887702e-06,
+      "loss": 1.1769,
+      "step": 50
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.15020264679653,
+      "learning_rate": 5.4545454545454545e-06,
+      "loss": 1.3617,
+      "step": 51
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6850801586450304,
+      "learning_rate": 5.561497326203209e-06,
+      "loss": 1.1572,
+      "step": 52
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8924022333001187,
+      "learning_rate": 5.6684491978609635e-06,
+      "loss": 1.1743,
+      "step": 53
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.485608070791642,
+      "learning_rate": 5.775401069518717e-06,
+      "loss": 1.1586,
+      "step": 54
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.888615712905799,
+      "learning_rate": 5.882352941176471e-06,
+      "loss": 1.126,
+      "step": 55
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7415196492224712,
+      "learning_rate": 5.989304812834225e-06,
+      "loss": 1.1537,
+      "step": 56
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8693679209978158,
+      "learning_rate": 6.096256684491979e-06,
+      "loss": 1.2311,
+      "step": 57
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9081308771900998,
+      "learning_rate": 6.203208556149734e-06,
+      "loss": 1.1317,
+      "step": 58
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1073178503402636,
+      "learning_rate": 6.3101604278074865e-06,
+      "loss": 1.2236,
+      "step": 59
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.893703437144612,
+      "learning_rate": 6.417112299465241e-06,
+      "loss": 1.0762,
+      "step": 60
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.142937542267072,
+      "learning_rate": 6.524064171122996e-06,
+      "loss": 1.247,
+      "step": 61
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7095661502780057,
+      "learning_rate": 6.631016042780749e-06,
+      "loss": 1.1433,
+      "step": 62
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7205317814630048,
+      "learning_rate": 6.737967914438504e-06,
+      "loss": 1.2264,
+      "step": 63
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9025266939275545,
+      "learning_rate": 6.844919786096257e-06,
+      "loss": 1.2017,
+      "step": 64
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7742469406575907,
+      "learning_rate": 6.951871657754011e-06,
+      "loss": 0.9904,
+      "step": 65
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8717123434598966,
+      "learning_rate": 7.058823529411766e-06,
+      "loss": 1.0933,
+      "step": 66
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.827996217336415,
+      "learning_rate": 7.1657754010695195e-06,
+      "loss": 1.168,
+      "step": 67
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2054898964695444,
+      "learning_rate": 7.272727272727273e-06,
+      "loss": 1.1252,
+      "step": 68
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8097542866352285,
+      "learning_rate": 7.379679144385027e-06,
+      "loss": 1.0983,
+      "step": 69
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6601209715292462,
+      "learning_rate": 7.486631016042781e-06,
+      "loss": 1.0822,
+      "step": 70
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9329582051308762,
+      "learning_rate": 7.593582887700536e-06,
+      "loss": 1.1667,
+      "step": 71
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8737046903488233,
+      "learning_rate": 7.70053475935829e-06,
+      "loss": 1.2134,
+      "step": 72
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8726308139765264,
+      "learning_rate": 7.807486631016043e-06,
+      "loss": 1.139,
+      "step": 73
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8082208279847156,
+      "learning_rate": 7.914438502673799e-06,
+      "loss": 1.1306,
+      "step": 74
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7999342511093201,
+      "learning_rate": 8.02139037433155e-06,
+      "loss": 1.0592,
+      "step": 75
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.925622009789158,
+      "learning_rate": 8.128342245989306e-06,
+      "loss": 1.05,
+      "step": 76
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.0099471147044214,
+      "learning_rate": 8.23529411764706e-06,
+      "loss": 1.1996,
+      "step": 77
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6907884755838152,
+      "learning_rate": 8.342245989304813e-06,
+      "loss": 1.1032,
+      "step": 78
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.076613101666246,
+      "learning_rate": 8.449197860962567e-06,
+      "loss": 0.3197,
+      "step": 79
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.045408360454748,
+      "learning_rate": 8.556149732620321e-06,
+      "loss": 1.1921,
+      "step": 80
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9764200920701724,
+      "learning_rate": 8.663101604278076e-06,
+      "loss": 1.133,
+      "step": 81
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.024395421370614,
+      "learning_rate": 8.77005347593583e-06,
+      "loss": 1.1514,
+      "step": 82
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.845005434362296,
+      "learning_rate": 8.877005347593584e-06,
+      "loss": 1.1382,
+      "step": 83
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9149695801567612,
+      "learning_rate": 8.983957219251337e-06,
+      "loss": 1.135,
+      "step": 84
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8194727956227852,
+      "learning_rate": 9.090909090909091e-06,
+      "loss": 1.1115,
+      "step": 85
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.938362929715415,
+      "learning_rate": 9.197860962566846e-06,
+      "loss": 1.0397,
+      "step": 86
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7911833005517368,
+      "learning_rate": 9.3048128342246e-06,
+      "loss": 1.1121,
+      "step": 87
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.661275803670195,
+      "learning_rate": 9.411764705882354e-06,
+      "loss": 1.1139,
+      "step": 88
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.035726729042163,
+      "learning_rate": 9.518716577540108e-06,
+      "loss": 1.0997,
+      "step": 89
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.866391987105394,
+      "learning_rate": 9.625668449197861e-06,
+      "loss": 1.0734,
+      "step": 90
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.039139150589232,
+      "learning_rate": 9.732620320855617e-06,
+      "loss": 1.095,
+      "step": 91
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8804583453725374,
+      "learning_rate": 9.83957219251337e-06,
+      "loss": 1.1043,
+      "step": 92
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7630157929624992,
+      "learning_rate": 9.946524064171124e-06,
+      "loss": 1.1254,
+      "step": 93
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.176711001947502,
+      "learning_rate": 1.0053475935828878e-05,
+      "loss": 1.1558,
+      "step": 94
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6176395648249458,
+      "learning_rate": 1.0160427807486633e-05,
+      "loss": 1.1931,
+      "step": 95
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.663288501156598,
+      "learning_rate": 1.0267379679144387e-05,
+      "loss": 1.1311,
+      "step": 96
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7754888444428265,
+      "learning_rate": 1.0374331550802139e-05,
+      "loss": 1.1876,
+      "step": 97
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7860000367033118,
+      "learning_rate": 1.0481283422459894e-05,
+      "loss": 1.1278,
+      "step": 98
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0241960238016596,
+      "learning_rate": 1.0588235294117648e-05,
+      "loss": 0.312,
+      "step": 99
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8151561916827776,
+      "learning_rate": 1.0695187165775403e-05,
+      "loss": 1.1143,
+      "step": 100
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9800457173425494,
+      "learning_rate": 1.0802139037433157e-05,
+      "loss": 1.1773,
+      "step": 101
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7866228879808546,
+      "learning_rate": 1.0909090909090909e-05,
+      "loss": 1.1064,
+      "step": 102
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8476968790026995,
+      "learning_rate": 1.1016042780748664e-05,
+      "loss": 1.1745,
+      "step": 103
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9918767346656621,
+      "learning_rate": 1.1122994652406418e-05,
+      "loss": 0.2951,
+      "step": 104
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7922576744014986,
+      "learning_rate": 1.1229946524064172e-05,
+      "loss": 1.1567,
+      "step": 105
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9851329868909744,
+      "learning_rate": 1.1336898395721927e-05,
+      "loss": 1.0187,
+      "step": 106
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8704732102826962,
+      "learning_rate": 1.1443850267379679e-05,
+      "loss": 1.1149,
+      "step": 107
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3819123224989367,
+      "learning_rate": 1.1550802139037434e-05,
+      "loss": 1.1869,
+      "step": 108
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7911645791981734,
+      "learning_rate": 1.1657754010695188e-05,
+      "loss": 1.1098,
+      "step": 109
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8259111310069072,
+      "learning_rate": 1.1764705882352942e-05,
+      "loss": 1.1297,
+      "step": 110
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.631045496479277,
+      "learning_rate": 1.1871657754010697e-05,
+      "loss": 1.1629,
+      "step": 111
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9198604255691587,
+      "learning_rate": 1.197860962566845e-05,
+      "loss": 1.0011,
+      "step": 112
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8552065076721744,
+      "learning_rate": 1.2085561497326203e-05,
+      "loss": 1.1413,
+      "step": 113
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5587142769860194,
+      "learning_rate": 1.2192513368983958e-05,
+      "loss": 0.9748,
+      "step": 114
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8968289196884078,
+      "learning_rate": 1.2299465240641712e-05,
+      "loss": 1.0986,
+      "step": 115
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.73161972015209,
+      "learning_rate": 1.2406417112299467e-05,
+      "loss": 1.0496,
+      "step": 116
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.993942806695572,
+      "learning_rate": 1.251336898395722e-05,
+      "loss": 1.2119,
+      "step": 117
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7452115127867245,
+      "learning_rate": 1.2620320855614973e-05,
+      "loss": 1.0531,
+      "step": 118
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8448890818407966,
+      "learning_rate": 1.2727272727272728e-05,
+      "loss": 1.0764,
+      "step": 119
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8865129672544962,
+      "learning_rate": 1.2834224598930482e-05,
+      "loss": 1.038,
+      "step": 120
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0099952740785687,
+      "learning_rate": 1.2941176470588238e-05,
+      "loss": 1.083,
+      "step": 121
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.893496895626763,
+      "learning_rate": 1.3048128342245991e-05,
+      "loss": 1.1138,
+      "step": 122
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7035437843358887,
+      "learning_rate": 1.3155080213903743e-05,
+      "loss": 1.067,
+      "step": 123
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7522601174420176,
+      "learning_rate": 1.3262032085561499e-05,
+      "loss": 1.0795,
+      "step": 124
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2574278300025763,
+      "learning_rate": 1.3368983957219252e-05,
+      "loss": 1.2813,
+      "step": 125
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.180472958805692,
+      "learning_rate": 1.3475935828877008e-05,
+      "loss": 1.1244,
+      "step": 126
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9639436649054822,
+      "learning_rate": 1.3582887700534761e-05,
+      "loss": 1.0696,
+      "step": 127
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8280256592829247,
+      "learning_rate": 1.3689839572192513e-05,
+      "loss": 1.1231,
+      "step": 128
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7034101581049972,
+      "learning_rate": 1.3796791443850269e-05,
+      "loss": 1.1078,
+      "step": 129
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8470609353073417,
+      "learning_rate": 1.3903743315508022e-05,
+      "loss": 1.0944,
+      "step": 130
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6983980246626533,
+      "learning_rate": 1.4010695187165778e-05,
+      "loss": 1.1211,
+      "step": 131
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8081912499390137,
+      "learning_rate": 1.4117647058823532e-05,
+      "loss": 0.2879,
+      "step": 132
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9252928439475747,
+      "learning_rate": 1.4224598930481284e-05,
+      "loss": 1.1393,
+      "step": 133
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0287409251251614,
+      "learning_rate": 1.4331550802139039e-05,
+      "loss": 1.1213,
+      "step": 134
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5189652455235887,
+      "learning_rate": 1.4438502673796793e-05,
+      "loss": 0.9858,
+      "step": 135
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8087314796323413,
+      "learning_rate": 1.4545454545454546e-05,
+      "loss": 1.0565,
+      "step": 136
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8380488024596233,
+      "learning_rate": 1.4652406417112302e-05,
+      "loss": 1.0444,
+      "step": 137
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.842529761897001,
+      "learning_rate": 1.4759358288770054e-05,
+      "loss": 0.9803,
+      "step": 138
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.695847818462051,
+      "learning_rate": 1.4866310160427807e-05,
+      "loss": 1.0971,
+      "step": 139
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.658080892359949,
+      "learning_rate": 1.4973262032085563e-05,
+      "loss": 1.0766,
+      "step": 140
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.534511371260837,
+      "learning_rate": 1.5080213903743316e-05,
+      "loss": 1.146,
+      "step": 141
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.687273172525131,
+      "learning_rate": 1.5187165775401072e-05,
+      "loss": 1.0289,
+      "step": 142
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9564170273263932,
+      "learning_rate": 1.5294117647058822e-05,
+      "loss": 1.0891,
+      "step": 143
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.016508270211656,
+      "learning_rate": 1.540106951871658e-05,
+      "loss": 1.1391,
+      "step": 144
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.053799951976019,
+      "learning_rate": 1.5508021390374333e-05,
+      "loss": 0.9739,
+      "step": 145
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.830280776187756,
+      "learning_rate": 1.5614973262032087e-05,
+      "loss": 1.1195,
+      "step": 146
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0163638993212434,
+      "learning_rate": 1.572192513368984e-05,
+      "loss": 1.1383,
+      "step": 147
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7667897747824024,
+      "learning_rate": 1.5828877005347597e-05,
+      "loss": 1.1388,
+      "step": 148
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6899698338271028,
+      "learning_rate": 1.5935828877005348e-05,
+      "loss": 1.0119,
+      "step": 149
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8653969329380322,
+      "learning_rate": 1.60427807486631e-05,
+      "loss": 1.0592,
+      "step": 150
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4913781661875962,
+      "learning_rate": 1.614973262032086e-05,
+      "loss": 0.9858,
+      "step": 151
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.857071966913271,
+      "learning_rate": 1.6256684491978612e-05,
+      "loss": 1.0336,
+      "step": 152
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7681772791047755,
+      "learning_rate": 1.6363636363636366e-05,
+      "loss": 0.3091,
+      "step": 153
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7973483037596603,
+      "learning_rate": 1.647058823529412e-05,
+      "loss": 1.0741,
+      "step": 154
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7031188503206849,
+      "learning_rate": 1.6577540106951873e-05,
+      "loss": 1.1208,
+      "step": 155
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5884058815477005,
+      "learning_rate": 1.6684491978609627e-05,
+      "loss": 1.031,
+      "step": 156
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.942411030314576,
+      "learning_rate": 1.679144385026738e-05,
+      "loss": 1.1414,
+      "step": 157
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8014056205268636,
+      "learning_rate": 1.6898395721925134e-05,
+      "loss": 1.0525,
+      "step": 158
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.019216987094288,
+      "learning_rate": 1.7005347593582888e-05,
+      "loss": 1.0838,
+      "step": 159
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7706589097372759,
+      "learning_rate": 1.7112299465240642e-05,
+      "loss": 0.3062,
+      "step": 160
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.008390867483698,
+      "learning_rate": 1.7219251336898395e-05,
+      "loss": 1.1059,
+      "step": 161
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5253240769849044,
+      "learning_rate": 1.7326203208556153e-05,
+      "loss": 0.9575,
+      "step": 162
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8312856767551149,
+      "learning_rate": 1.7433155080213906e-05,
+      "loss": 1.1092,
+      "step": 163
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.728249129263797,
+      "learning_rate": 1.754010695187166e-05,
+      "loss": 1.1073,
+      "step": 164
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7704008841273098,
+      "learning_rate": 1.7647058823529414e-05,
+      "loss": 1.0628,
+      "step": 165
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6661351505952953,
+      "learning_rate": 1.7754010695187167e-05,
+      "loss": 1.0585,
+      "step": 166
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8485500470317957,
+      "learning_rate": 1.786096256684492e-05,
+      "loss": 1.0741,
+      "step": 167
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6544345091365844,
+      "learning_rate": 1.7967914438502675e-05,
+      "loss": 1.026,
+      "step": 168
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7496030966221727,
+      "learning_rate": 1.807486631016043e-05,
+      "loss": 1.095,
+      "step": 169
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.664799656814396,
+      "learning_rate": 1.8181818181818182e-05,
+      "loss": 1.0104,
+      "step": 170
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.591765179932876,
+      "learning_rate": 1.8288770053475936e-05,
+      "loss": 1.0212,
+      "step": 171
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6665906143097107,
+      "learning_rate": 1.8395721925133693e-05,
+      "loss": 1.0319,
+      "step": 172
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.775528084120115,
+      "learning_rate": 1.8502673796791447e-05,
+      "loss": 1.115,
+      "step": 173
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6831446986970393,
+      "learning_rate": 1.86096256684492e-05,
+      "loss": 1.0538,
+      "step": 174
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7067504958551498,
+      "learning_rate": 1.8716577540106954e-05,
+      "loss": 1.0346,
+      "step": 175
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8000110915623775,
+      "learning_rate": 1.8823529411764708e-05,
+      "loss": 1.0941,
+      "step": 176
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.631768825070943,
+      "learning_rate": 1.893048128342246e-05,
+      "loss": 1.086,
+      "step": 177
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.591403918529108,
+      "learning_rate": 1.9037433155080215e-05,
+      "loss": 0.9815,
+      "step": 178
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8318516443947612,
+      "learning_rate": 1.9144385026737972e-05,
+      "loss": 1.033,
+      "step": 179
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8872040870712627,
+      "learning_rate": 1.9251336898395722e-05,
+      "loss": 1.1848,
+      "step": 180
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8544692188224614,
+      "learning_rate": 1.9358288770053476e-05,
+      "loss": 1.0957,
+      "step": 181
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8651969753044279,
+      "learning_rate": 1.9465240641711233e-05,
+      "loss": 0.2902,
+      "step": 182
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7281646587725787,
+      "learning_rate": 1.9572192513368987e-05,
+      "loss": 0.9997,
+      "step": 183
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.748033940677634,
+      "learning_rate": 1.967914438502674e-05,
+      "loss": 1.1354,
+      "step": 184
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7433533045349984,
+      "learning_rate": 1.9786096256684494e-05,
+      "loss": 0.2895,
+      "step": 185
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9279365114715972,
+      "learning_rate": 1.9893048128342248e-05,
+      "loss": 1.1446,
+      "step": 186
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5954366961135809,
+      "learning_rate": 2e-05,
+      "loss": 1.0631,
+      "step": 187
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7074825859927216,
+      "learning_rate": 1.9999998637862175e-05,
+      "loss": 1.0039,
+      "step": 188
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.86033997018962,
+      "learning_rate": 1.999999455144907e-05,
+      "loss": 1.1297,
+      "step": 189
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0091468258782172,
+      "learning_rate": 1.9999987740761794e-05,
+      "loss": 1.0665,
+      "step": 190
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7261058140821122,
+      "learning_rate": 1.999997820580221e-05,
+      "loss": 1.0258,
+      "step": 191
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9504220111566979,
+      "learning_rate": 1.999996594657291e-05,
+      "loss": 1.0149,
+      "step": 192
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.004685784760631,
+      "learning_rate": 1.9999950963077235e-05,
+      "loss": 1.072,
+      "step": 193
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9309316979287292,
+      "learning_rate": 1.999993325531927e-05,
+      "loss": 1.0467,
+      "step": 194
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6356637207113613,
+      "learning_rate": 1.9999912823303832e-05,
+      "loss": 0.9829,
+      "step": 195
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.045396696853703,
+      "learning_rate": 1.9999889667036496e-05,
+      "loss": 1.0942,
+      "step": 196
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9932046640345547,
+      "learning_rate": 1.9999863786523567e-05,
+      "loss": 1.1781,
+      "step": 197
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6717669015323524,
+      "learning_rate": 1.999983518177209e-05,
+      "loss": 1.0575,
+      "step": 198
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7365123932062148,
+      "learning_rate": 1.9999803852789864e-05,
+      "loss": 1.0566,
+      "step": 199
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.062516656721653,
+      "learning_rate": 1.999976979958542e-05,
+      "loss": 1.027,
+      "step": 200
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7495776362860862,
+      "learning_rate": 1.9999733022168043e-05,
+      "loss": 1.054,
+      "step": 201
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6099227533297795,
+      "learning_rate": 1.9999693520547745e-05,
+      "loss": 1.0498,
+      "step": 202
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6277078827555203,
+      "learning_rate": 1.9999651294735285e-05,
+      "loss": 1.0575,
+      "step": 203
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6762913736664313,
+      "learning_rate": 1.9999606344742176e-05,
+      "loss": 1.1026,
+      "step": 204
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0819182123700086,
+      "learning_rate": 1.9999558670580656e-05,
+      "loss": 1.1238,
+      "step": 205
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.933011872225037,
+      "learning_rate": 1.999950827226371e-05,
+      "loss": 0.3296,
+      "step": 206
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7318182671057278,
+      "learning_rate": 1.9999455149805076e-05,
+      "loss": 1.0823,
+      "step": 207
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4232312529086144,
+      "learning_rate": 1.9999399303219222e-05,
+      "loss": 0.9811,
+      "step": 208
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7046086876602637,
+      "learning_rate": 1.9999340732521363e-05,
+      "loss": 1.0608,
+      "step": 209
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.793007597782919,
+      "learning_rate": 1.9999279437727456e-05,
+      "loss": 1.0736,
+      "step": 210
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5795068073409793,
+      "learning_rate": 1.99992154188542e-05,
+      "loss": 1.0054,
+      "step": 211
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.751789952443813,
+      "learning_rate": 1.999914867591903e-05,
+      "loss": 1.0345,
+      "step": 212
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7243277880402155,
+      "learning_rate": 1.9999079208940137e-05,
+      "loss": 1.1349,
+      "step": 213
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6695632873035346,
+      "learning_rate": 1.9999007017936436e-05,
+      "loss": 1.0211,
+      "step": 214
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7941020451680676,
+      "learning_rate": 1.99989321029276e-05,
+      "loss": 0.982,
+      "step": 215
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9803340995881384,
+      "learning_rate": 1.999885446393404e-05,
+      "loss": 1.07,
+      "step": 216
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8949564240388055,
+      "learning_rate": 1.9998774100976903e-05,
+      "loss": 0.9726,
+      "step": 217
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6204232041265776,
+      "learning_rate": 1.999869101407808e-05,
+      "loss": 1.0177,
+      "step": 218
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.437885558599507,
+      "learning_rate": 1.999860520326021e-05,
+      "loss": 1.0564,
+      "step": 219
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6407564383035635,
+      "learning_rate": 1.9998516668546675e-05,
+      "loss": 0.973,
+      "step": 220
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.887030025916089,
+      "learning_rate": 1.9998425409961585e-05,
+      "loss": 0.2966,
+      "step": 221
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8936230193188892,
+      "learning_rate": 1.9998331427529803e-05,
+      "loss": 1.0326,
+      "step": 222
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.682042121032617,
+      "learning_rate": 1.9998234721276938e-05,
+      "loss": 0.9614,
+      "step": 223
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6461869488164882,
+      "learning_rate": 1.999813529122933e-05,
+      "loss": 1.1369,
+      "step": 224
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6307905816513593,
+      "learning_rate": 1.999803313741407e-05,
+      "loss": 1.0643,
+      "step": 225
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7634064025143267,
+      "learning_rate": 1.9997928259858985e-05,
+      "loss": 0.314,
+      "step": 226
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7318951519404067,
+      "learning_rate": 1.9997820658592645e-05,
+      "loss": 1.094,
+      "step": 227
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4966028781453264,
+      "learning_rate": 1.999771033364437e-05,
+      "loss": 1.0218,
+      "step": 228
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8623967950584346,
+      "learning_rate": 1.999759728504421e-05,
+      "loss": 1.0368,
+      "step": 229
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8766268765177092,
+      "learning_rate": 1.9997481512822966e-05,
+      "loss": 1.1044,
+      "step": 230
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6312093604760378,
+      "learning_rate": 1.9997363017012174e-05,
+      "loss": 0.983,
+      "step": 231
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.501956967595695,
+      "learning_rate": 1.9997241797644117e-05,
+      "loss": 0.9985,
+      "step": 232
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0000772545094674,
+      "learning_rate": 1.9997117854751818e-05,
+      "loss": 1.0484,
+      "step": 233
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.607509648300964,
+      "learning_rate": 1.9996991188369045e-05,
+      "loss": 0.9913,
+      "step": 234
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5917588609708224,
+      "learning_rate": 1.9996861798530304e-05,
+      "loss": 0.9898,
+      "step": 235
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7948878807028363,
+      "learning_rate": 1.9996729685270844e-05,
+      "loss": 1.0135,
+      "step": 236
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5680855462110115,
+      "learning_rate": 1.9996594848626655e-05,
+      "loss": 1.0215,
+      "step": 237
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0745048954174523,
+      "learning_rate": 1.999645728863447e-05,
+      "loss": 1.0933,
+      "step": 238
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7530365154261676,
+      "learning_rate": 1.9996317005331768e-05,
+      "loss": 1.0567,
+      "step": 239
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6240767758637327,
+      "learning_rate": 1.999617399875676e-05,
+      "loss": 1.0015,
+      "step": 240
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8086770545196882,
+      "learning_rate": 1.9996028268948414e-05,
+      "loss": 1.0575,
+      "step": 241
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.938130289662243,
+      "learning_rate": 1.9995879815946423e-05,
+      "loss": 1.0534,
+      "step": 242
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5878139852137902,
+      "learning_rate": 1.999572863979123e-05,
+      "loss": 1.0343,
+      "step": 243
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9034428631931116,
+      "learning_rate": 1.9995574740524024e-05,
+      "loss": 0.2794,
+      "step": 244
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8347427663141224,
+      "learning_rate": 1.9995418118186728e-05,
+      "loss": 0.259,
+      "step": 245
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7036518766055384,
+      "learning_rate": 1.9995258772822012e-05,
+      "loss": 1.0136,
+      "step": 246
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.638416518183063,
+      "learning_rate": 1.9995096704473287e-05,
+      "loss": 1.0135,
+      "step": 247
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5603252338237057,
+      "learning_rate": 1.99949319131847e-05,
+      "loss": 1.0414,
+      "step": 248
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7747313662404303,
+      "learning_rate": 1.9994764399001153e-05,
+      "loss": 1.1135,
+      "step": 249
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5724468043882305,
+      "learning_rate": 1.999459416196827e-05,
+      "loss": 1.0623,
+      "step": 250
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8867847177373578,
+      "learning_rate": 1.9994421202132436e-05,
+      "loss": 0.9757,
+      "step": 251
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.762573509551815,
+      "learning_rate": 1.9994245519540772e-05,
+      "loss": 1.0376,
+      "step": 252
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5064456518314533,
+      "learning_rate": 1.9994067114241135e-05,
+      "loss": 1.0614,
+      "step": 253
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7505906238725135,
+      "learning_rate": 1.9993885986282125e-05,
+      "loss": 1.0021,
+      "step": 254
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4459820987644794,
+      "learning_rate": 1.9993702135713093e-05,
+      "loss": 0.3405,
+      "step": 255
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3961162375339864,
+      "learning_rate": 1.9993515562584117e-05,
+      "loss": 1.0088,
+      "step": 256
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7188801748657532,
+      "learning_rate": 1.9993326266946033e-05,
+      "loss": 0.9858,
+      "step": 257
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.65953326373045,
+      "learning_rate": 1.9993134248850402e-05,
+      "loss": 0.9896,
+      "step": 258
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7575140535464575,
+      "learning_rate": 1.9992939508349544e-05,
+      "loss": 1.037,
+      "step": 259
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6528061147711812,
+      "learning_rate": 1.9992742045496502e-05,
+      "loss": 0.9862,
+      "step": 260
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.326822861609363,
+      "learning_rate": 1.999254186034508e-05,
+      "loss": 0.988,
+      "step": 261
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.641734343088275,
+      "learning_rate": 1.9992338952949805e-05,
+      "loss": 1.0615,
+      "step": 262
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6920213770140218,
+      "learning_rate": 1.9992133323365963e-05,
+      "loss": 1.0513,
+      "step": 263
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.74228701927351,
+      "learning_rate": 1.9991924971649566e-05,
+      "loss": 1.0678,
+      "step": 264
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.548480057299955,
+      "learning_rate": 1.9991713897857376e-05,
+      "loss": 0.9962,
+      "step": 265
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.508814252888038,
+      "learning_rate": 1.99915001020469e-05,
+      "loss": 0.9918,
+      "step": 266
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5128546957737885,
+      "learning_rate": 1.999128358427638e-05,
+      "loss": 1.0114,
+      "step": 267
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7263825649686189,
+      "learning_rate": 1.9991064344604798e-05,
+      "loss": 1.0112,
+      "step": 268
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8627291717203835,
+      "learning_rate": 1.9990842383091884e-05,
+      "loss": 0.9972,
+      "step": 269
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9641887063953032,
+      "learning_rate": 1.9990617699798104e-05,
+      "loss": 1.1887,
+      "step": 270
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.671211556682236,
+      "learning_rate": 1.999039029478467e-05,
+      "loss": 1.0209,
+      "step": 271
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9177823850207563,
+      "learning_rate": 1.9990160168113534e-05,
+      "loss": 1.1049,
+      "step": 272
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7138496283911413,
+      "learning_rate": 1.998992731984739e-05,
+      "loss": 0.9027,
+      "step": 273
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.68064719107991,
+      "learning_rate": 1.998969175004967e-05,
+      "loss": 1.0262,
+      "step": 274
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.582945199395315,
+      "learning_rate": 1.9989453458784544e-05,
+      "loss": 1.0366,
+      "step": 275
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4032919568281048,
+      "learning_rate": 1.998921244611694e-05,
+      "loss": 0.3259,
+      "step": 276
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.575045969364126,
+      "learning_rate": 1.9988968712112512e-05,
+      "loss": 1.0383,
+      "step": 277
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.160030215253682,
+      "learning_rate": 1.9988722256837656e-05,
+      "loss": 0.3398,
+      "step": 278
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.517575500859136,
+      "learning_rate": 1.998847308035952e-05,
+      "loss": 0.9764,
+      "step": 279
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.735991152407227,
+      "learning_rate": 1.998822118274598e-05,
+      "loss": 1.1008,
+      "step": 280
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.043477594962616,
+      "learning_rate": 1.9987966564065663e-05,
+      "loss": 1.108,
+      "step": 281
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0173223338875044,
+      "learning_rate": 1.9987709224387935e-05,
+      "loss": 1.0837,
+      "step": 282
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6558523476719926,
+      "learning_rate": 1.9987449163782902e-05,
+      "loss": 1.051,
+      "step": 283
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.909934445923109,
+      "learning_rate": 1.9987186382321408e-05,
+      "loss": 1.1229,
+      "step": 284
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5780926862798184,
+      "learning_rate": 1.998692088007505e-05,
+      "loss": 1.0195,
+      "step": 285
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6962813639095509,
+      "learning_rate": 1.998665265711615e-05,
+      "loss": 1.0278,
+      "step": 286
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1093561242283343,
+      "learning_rate": 1.9986381713517783e-05,
+      "loss": 1.0229,
+      "step": 287
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4771685082044144,
+      "learning_rate": 1.998610804935376e-05,
+      "loss": 1.0514,
+      "step": 288
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.3456627789091917,
+      "learning_rate": 1.998583166469864e-05,
+      "loss": 1.0326,
+      "step": 289
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6743407986083125,
+      "learning_rate": 1.9985552559627708e-05,
+      "loss": 1.0177,
+      "step": 290
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9978865622552715,
+      "learning_rate": 1.998527073421701e-05,
+      "loss": 1.0525,
+      "step": 291
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4441850756252455,
+      "learning_rate": 1.9984986188543314e-05,
+      "loss": 0.9472,
+      "step": 292
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8749097494418914,
+      "learning_rate": 1.9984698922684146e-05,
+      "loss": 1.045,
+      "step": 293
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.656638989338678,
+      "learning_rate": 1.9984408936717758e-05,
+      "loss": 1.0225,
+      "step": 294
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6307851877299302,
+      "learning_rate": 1.9984116230723157e-05,
+      "loss": 1.0178,
+      "step": 295
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8527458952212414,
+      "learning_rate": 1.9983820804780082e-05,
+      "loss": 1.1196,
+      "step": 296
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.50788174665385,
+      "learning_rate": 1.998352265896901e-05,
+      "loss": 0.9815,
+      "step": 297
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.637605150546744,
+      "learning_rate": 1.9983221793371173e-05,
+      "loss": 0.995,
+      "step": 298
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6395445388038,
+      "learning_rate": 1.9982918208068525e-05,
+      "loss": 0.3293,
+      "step": 299
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0625504790977898,
+      "learning_rate": 1.9982611903143782e-05,
+      "loss": 1.1525,
+      "step": 300
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.56115777539679,
+      "learning_rate": 1.998230287868038e-05,
+      "loss": 1.0189,
+      "step": 301
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5393848263201955,
+      "learning_rate": 1.998199113476251e-05,
+      "loss": 1.0209,
+      "step": 302
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.939003223778072,
+      "learning_rate": 1.9981676671475103e-05,
+      "loss": 1.0647,
+      "step": 303
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7285409785421004,
+      "learning_rate": 1.9981359488903818e-05,
+      "loss": 0.9428,
+      "step": 304
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6750264212851753,
+      "learning_rate": 1.9981039587135078e-05,
+      "loss": 1.0533,
+      "step": 305
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.650964779271924,
+      "learning_rate": 1.9980716966256023e-05,
+      "loss": 1.0834,
+      "step": 306
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9692782552394792,
+      "learning_rate": 1.9980391626354543e-05,
+      "loss": 1.0232,
+      "step": 307
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6910486944626006,
+      "learning_rate": 1.998006356751928e-05,
+      "loss": 1.0881,
+      "step": 308
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.402373648300866,
+      "learning_rate": 1.997973278983959e-05,
+      "loss": 0.9751,
+      "step": 309
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6707533277844913,
+      "learning_rate": 1.99793992934056e-05,
+      "loss": 1.0244,
+      "step": 310
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3672112820464226,
+      "learning_rate": 1.997906307830816e-05,
+      "loss": 1.0024,
+      "step": 311
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.677283721679435,
+      "learning_rate": 1.9978724144638863e-05,
+      "loss": 0.9786,
+      "step": 312
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5901802641775895,
+      "learning_rate": 1.997838249249004e-05,
+      "loss": 1.0446,
+      "step": 313
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9036247426850756,
+      "learning_rate": 1.9978038121954775e-05,
+      "loss": 1.0828,
+      "step": 314
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6039006692071103,
+      "learning_rate": 1.9977691033126875e-05,
+      "loss": 1.0634,
+      "step": 315
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5195992045174704,
+      "learning_rate": 1.9977341226100905e-05,
+      "loss": 1.0296,
+      "step": 316
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7189598502667076,
+      "learning_rate": 1.9976988700972154e-05,
+      "loss": 1.0893,
+      "step": 317
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7745096855871998,
+      "learning_rate": 1.9976633457836664e-05,
+      "loss": 1.0324,
+      "step": 318
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7552055662842703,
+      "learning_rate": 1.9976275496791216e-05,
+      "loss": 0.9983,
+      "step": 319
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8826269210673037,
+      "learning_rate": 1.9975914817933325e-05,
+      "loss": 1.1194,
+      "step": 320
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7413744756051104,
+      "learning_rate": 1.9975551421361244e-05,
+      "loss": 1.0757,
+      "step": 321
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6086285095942818,
+      "learning_rate": 1.9975185307173985e-05,
+      "loss": 1.0564,
+      "step": 322
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4523290535940954,
+      "learning_rate": 1.9974816475471277e-05,
+      "loss": 1.0273,
+      "step": 323
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4488141846005422,
+      "learning_rate": 1.9974444926353605e-05,
+      "loss": 0.9917,
+      "step": 324
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.483359612904228,
+      "learning_rate": 1.9974070659922184e-05,
+      "loss": 1.0363,
+      "step": 325
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.181937241172379,
+      "learning_rate": 1.9973693676278985e-05,
+      "loss": 0.3275,
+      "step": 326
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7873387330257697,
+      "learning_rate": 1.9973313975526696e-05,
+      "loss": 1.0662,
+      "step": 327
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4729436282844879,
+      "learning_rate": 1.9972931557768768e-05,
+      "loss": 1.0659,
+      "step": 328
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8189004703508875,
+      "learning_rate": 1.9972546423109378e-05,
+      "loss": 1.0371,
+      "step": 329
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8560641588276288,
+      "learning_rate": 1.9972158571653442e-05,
+      "loss": 1.062,
+      "step": 330
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7930752705947612,
+      "learning_rate": 1.9971768003506635e-05,
+      "loss": 1.0352,
+      "step": 331
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.720750424864003,
+      "learning_rate": 1.9971374718775346e-05,
+      "loss": 0.9813,
+      "step": 332
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3731088142749825,
+      "learning_rate": 1.997097871756672e-05,
+      "loss": 1.0148,
+      "step": 333
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5444373873369466,
+      "learning_rate": 1.9970579999988643e-05,
+      "loss": 0.906,
+      "step": 334
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.765355201379636,
+      "learning_rate": 1.9970178566149734e-05,
+      "loss": 1.0247,
+      "step": 335
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5319686220315032,
+      "learning_rate": 1.996977441615935e-05,
+      "loss": 1.0515,
+      "step": 336
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5419664326426117,
+      "learning_rate": 1.99693675501276e-05,
+      "loss": 1.0184,
+      "step": 337
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5518617205103495,
+      "learning_rate": 1.996895796816532e-05,
+      "loss": 0.9798,
+      "step": 338
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.630601807773335,
+      "learning_rate": 1.9968545670384094e-05,
+      "loss": 0.9609,
+      "step": 339
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0892124042353273,
+      "learning_rate": 1.9968130656896244e-05,
+      "loss": 1.1176,
+      "step": 340
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8048389658180481,
+      "learning_rate": 1.996771292781483e-05,
+      "loss": 1.0576,
+      "step": 341
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6254413451284853,
+      "learning_rate": 1.996729248325365e-05,
+      "loss": 1.0423,
+      "step": 342
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5440792373104453,
+      "learning_rate": 1.996686932332725e-05,
+      "loss": 0.994,
+      "step": 343
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.62651917679758,
+      "learning_rate": 1.9966443448150906e-05,
+      "loss": 1.0093,
+      "step": 344
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5464446222891104,
+      "learning_rate": 1.9966014857840644e-05,
+      "loss": 0.9879,
+      "step": 345
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6410926016242666,
+      "learning_rate": 1.9965583552513216e-05,
+      "loss": 0.9956,
+      "step": 346
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3818818202674275,
+      "learning_rate": 1.9965149532286126e-05,
+      "loss": 1.0365,
+      "step": 347
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6038971364616648,
+      "learning_rate": 1.9964712797277614e-05,
+      "loss": 1.0633,
+      "step": 348
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8897259166516216,
+      "learning_rate": 1.9964273347606656e-05,
+      "loss": 1.0792,
+      "step": 349
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4591939123300526,
+      "learning_rate": 1.996383118339297e-05,
+      "loss": 0.986,
+      "step": 350
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.563494940726578,
+      "learning_rate": 1.9963386304757018e-05,
+      "loss": 1.0064,
+      "step": 351
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3998777043409887,
+      "learning_rate": 1.9962938711819993e-05,
+      "loss": 1.0882,
+      "step": 352
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8929350165109857,
+      "learning_rate": 1.9962488404703832e-05,
+      "loss": 1.139,
+      "step": 353
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4728290806799589,
+      "learning_rate": 1.996203538353121e-05,
+      "loss": 1.0492,
+      "step": 354
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.639981639829432,
+      "learning_rate": 1.9961579648425552e-05,
+      "loss": 0.9444,
+      "step": 355
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7431463267388456,
+      "learning_rate": 1.9961121199510996e-05,
+      "loss": 1.0705,
+      "step": 356
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.411548324313841,
+      "learning_rate": 1.9960660036912453e-05,
+      "loss": 1.0136,
+      "step": 357
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7627777735476993,
+      "learning_rate": 1.9960196160755542e-05,
+      "loss": 1.0714,
+      "step": 358
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3695222792709723,
+      "learning_rate": 1.995972957116665e-05,
+      "loss": 0.9472,
+      "step": 359
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8009094463650925,
+      "learning_rate": 1.9959260268272876e-05,
+      "loss": 1.0052,
+      "step": 360
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.564560276938673,
+      "learning_rate": 1.9958788252202078e-05,
+      "loss": 1.1055,
+      "step": 361
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7573095566418,
+      "learning_rate": 1.9958313523082842e-05,
+      "loss": 1.0904,
+      "step": 362
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4765610903021726,
+      "learning_rate": 1.9957836081044498e-05,
+      "loss": 0.9864,
+      "step": 363
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.68714992520169,
+      "learning_rate": 1.995735592621712e-05,
+      "loss": 1.0252,
+      "step": 364
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6182491697953183,
+      "learning_rate": 1.9956873058731514e-05,
+      "loss": 1.1404,
+      "step": 365
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7733148089161146,
+      "learning_rate": 1.9956387478719222e-05,
+      "loss": 1.0286,
+      "step": 366
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8230162778386978,
+      "learning_rate": 1.9955899186312527e-05,
+      "loss": 1.0189,
+      "step": 367
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5368141632441858,
+      "learning_rate": 1.9955408181644464e-05,
+      "loss": 1.0836,
+      "step": 368
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6499475064832012,
+      "learning_rate": 1.9954914464848787e-05,
+      "loss": 1.0254,
+      "step": 369
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7510700594464543,
+      "learning_rate": 1.995441803606e-05,
+      "loss": 1.0241,
+      "step": 370
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9730704557263021,
+      "learning_rate": 1.9953918895413346e-05,
+      "loss": 1.0729,
+      "step": 371
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.614765088108355,
+      "learning_rate": 1.9953417043044806e-05,
+      "loss": 1.0195,
+      "step": 372
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.888963527902822,
+      "learning_rate": 1.9952912479091094e-05,
+      "loss": 1.0155,
+      "step": 373
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7263120571016581,
+      "learning_rate": 1.9952405203689668e-05,
+      "loss": 1.0115,
+      "step": 374
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7561186608571346,
+      "learning_rate": 1.9951895216978725e-05,
+      "loss": 0.9807,
+      "step": 375
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5541686445435405,
+      "learning_rate": 1.9951382519097197e-05,
+      "loss": 1.0311,
+      "step": 376
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.594049936242053,
+      "learning_rate": 1.9950867110184765e-05,
+      "loss": 1.0424,
+      "step": 377
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6909264110736675,
+      "learning_rate": 1.995034899038183e-05,
+      "loss": 0.9985,
+      "step": 378
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.824913138309026,
+      "learning_rate": 1.994982815982955e-05,
+      "loss": 0.9497,
+      "step": 379
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6141206621915138,
+      "learning_rate": 1.994930461866981e-05,
+      "loss": 1.0562,
+      "step": 380
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.455954081298042,
+      "learning_rate": 1.9948778367045235e-05,
+      "loss": 0.968,
+      "step": 381
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5789371527564335,
+      "learning_rate": 1.99482494050992e-05,
+      "loss": 1.0081,
+      "step": 382
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9240866879437077,
+      "learning_rate": 1.9947717732975795e-05,
+      "loss": 1.0633,
+      "step": 383
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.686415790353978,
+      "learning_rate": 1.994718335081987e-05,
+      "loss": 1.0187,
+      "step": 384
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5507195584159947,
+      "learning_rate": 1.9946646258777008e-05,
+      "loss": 1.0076,
+      "step": 385
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4315745063907792,
+      "learning_rate": 1.994610645699352e-05,
+      "loss": 0.9508,
+      "step": 386
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3875781277566939,
+      "learning_rate": 1.994556394561647e-05,
+      "loss": 0.952,
+      "step": 387
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7184657533726364,
+      "learning_rate": 1.994501872479365e-05,
+      "loss": 1.0146,
+      "step": 388
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8936008587422144,
+      "learning_rate": 1.9944470794673592e-05,
+      "loss": 0.945,
+      "step": 389
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4647632761149083,
+      "learning_rate": 1.994392015540557e-05,
+      "loss": 1.0143,
+      "step": 390
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7073484018130995,
+      "learning_rate": 1.994336680713959e-05,
+      "loss": 1.0782,
+      "step": 391
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4760453400861175,
+      "learning_rate": 1.9942810750026403e-05,
+      "loss": 0.9985,
+      "step": 392
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8949264271527535,
+      "learning_rate": 1.9942251984217492e-05,
+      "loss": 0.9794,
+      "step": 393
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.908689342272236,
+      "learning_rate": 1.994169050986508e-05,
+      "loss": 1.0248,
+      "step": 394
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3058824948504468,
+      "learning_rate": 1.9941126327122128e-05,
+      "loss": 0.9788,
+      "step": 395
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8176228099850074,
+      "learning_rate": 1.9940559436142338e-05,
+      "loss": 1.0306,
+      "step": 396
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.476718204982747,
+      "learning_rate": 1.9939989837080143e-05,
+      "loss": 0.9584,
+      "step": 397
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.060066210472057,
+      "learning_rate": 1.9939417530090722e-05,
+      "loss": 1.0027,
+      "step": 398
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.702479137220705,
+      "learning_rate": 1.993884251532998e-05,
+      "loss": 1.0553,
+      "step": 399
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6689142645039947,
+      "learning_rate": 1.9938264792954573e-05,
+      "loss": 1.0583,
+      "step": 400
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.524097043100204,
+      "learning_rate": 1.9937684363121886e-05,
+      "loss": 0.9643,
+      "step": 401
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0548985399170636,
+      "learning_rate": 1.9937101225990046e-05,
+      "loss": 0.9941,
+      "step": 402
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9681639215261673,
+      "learning_rate": 1.993651538171791e-05,
+      "loss": 1.046,
+      "step": 403
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5211915443206312,
+      "learning_rate": 1.993592683046509e-05,
+      "loss": 1.0181,
+      "step": 404
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8378544676180313,
+      "learning_rate": 1.993533557239191e-05,
+      "loss": 0.9782,
+      "step": 405
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.525910033862189,
+      "learning_rate": 1.993474160765945e-05,
+      "loss": 0.9425,
+      "step": 406
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.433443295434048,
+      "learning_rate": 1.9934144936429526e-05,
+      "loss": 0.9982,
+      "step": 407
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8243129854884879,
+      "learning_rate": 1.9933545558864686e-05,
+      "loss": 0.3481,
+      "step": 408
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7822853266813,
+      "learning_rate": 1.9932943475128215e-05,
+      "loss": 0.9145,
+      "step": 409
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.609621489731567,
+      "learning_rate": 1.993233868538414e-05,
+      "loss": 1.0196,
+      "step": 410
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8076276784022587,
+      "learning_rate": 1.9931731189797216e-05,
+      "loss": 1.0013,
+      "step": 411
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3161724523931488,
+      "learning_rate": 1.993112098853295e-05,
+      "loss": 1.0404,
+      "step": 412
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7723900625807714,
+      "learning_rate": 1.9930508081757572e-05,
+      "loss": 1.0119,
+      "step": 413
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4812569338147326,
+      "learning_rate": 1.9929892469638056e-05,
+      "loss": 1.0798,
+      "step": 414
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6181961544448267,
+      "learning_rate": 1.9929274152342113e-05,
+      "loss": 1.0017,
+      "step": 415
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7371539248301142,
+      "learning_rate": 1.9928653130038188e-05,
+      "loss": 1.0232,
+      "step": 416
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4399499242924538,
+      "learning_rate": 1.9928029402895466e-05,
+      "loss": 1.0,
+      "step": 417
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8199510913530488,
+      "learning_rate": 1.992740297108387e-05,
+      "loss": 0.3309,
+      "step": 418
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5022318361924163,
+      "learning_rate": 1.992677383477405e-05,
+      "loss": 1.0469,
+      "step": 419
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4357756270239264,
+      "learning_rate": 1.9926141994137404e-05,
+      "loss": 1.0193,
+      "step": 420
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7176110254294175,
+      "learning_rate": 1.9925507449346066e-05,
+      "loss": 0.9908,
+      "step": 421
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5735301802556574,
+      "learning_rate": 1.99248702005729e-05,
+      "loss": 1.0161,
+      "step": 422
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.965738007688637,
+      "learning_rate": 1.9924230247991508e-05,
+      "loss": 1.0884,
+      "step": 423
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7346010264901173,
+      "learning_rate": 1.9923587591776236e-05,
+      "loss": 0.9484,
+      "step": 424
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4528084007557365,
+      "learning_rate": 1.992294223210216e-05,
+      "loss": 0.9744,
+      "step": 425
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4905042816359426,
+      "learning_rate": 1.9922294169145088e-05,
+      "loss": 0.9992,
+      "step": 426
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3913613620721677,
+      "learning_rate": 1.992164340308158e-05,
+      "loss": 1.0361,
+      "step": 427
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.375308021945723,
+      "learning_rate": 1.9920989934088914e-05,
+      "loss": 0.9669,
+      "step": 428
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5424953205182184,
+      "learning_rate": 1.9920333762345116e-05,
+      "loss": 0.8922,
+      "step": 429
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4799823823901987,
+      "learning_rate": 1.9919674888028946e-05,
+      "loss": 0.9667,
+      "step": 430
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7058697045041662,
+      "learning_rate": 1.99190133113199e-05,
+      "loss": 1.0567,
+      "step": 431
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.644740411908266,
+      "learning_rate": 1.991834903239821e-05,
+      "loss": 0.9518,
+      "step": 432
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7007928600142808,
+      "learning_rate": 1.991768205144484e-05,
+      "loss": 1.0071,
+      "step": 433
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.385453195392088,
+      "learning_rate": 1.9917012368641497e-05,
+      "loss": 0.9523,
+      "step": 434
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.627875518812952,
+      "learning_rate": 1.991633998417062e-05,
+      "loss": 0.9832,
+      "step": 435
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9107051420111736,
+      "learning_rate": 1.991566489821539e-05,
+      "loss": 0.3308,
+      "step": 436
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4563285754252062,
+      "learning_rate": 1.9914987110959713e-05,
+      "loss": 0.9231,
+      "step": 437
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9196050106916664,
+      "learning_rate": 1.9914306622588237e-05,
+      "loss": 1.0093,
+      "step": 438
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7538293904496163,
+      "learning_rate": 1.9913623433286346e-05,
+      "loss": 1.0578,
+      "step": 439
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8652342337248711,
+      "learning_rate": 1.9912937543240164e-05,
+      "loss": 0.3396,
+      "step": 440
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6572881146556147,
+      "learning_rate": 1.9912248952636543e-05,
+      "loss": 1.0178,
+      "step": 441
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7109232068009135,
+      "learning_rate": 1.9911557661663073e-05,
+      "loss": 0.9887,
+      "step": 442
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.595434251402399,
+      "learning_rate": 1.9910863670508088e-05,
+      "loss": 1.1132,
+      "step": 443
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5232395569267112,
+      "learning_rate": 1.991016697936064e-05,
+      "loss": 1.0143,
+      "step": 444
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.651354793071197,
+      "learning_rate": 1.990946758841053e-05,
+      "loss": 1.0009,
+      "step": 445
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6310819331229927,
+      "learning_rate": 1.9908765497848296e-05,
+      "loss": 0.9179,
+      "step": 446
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7366726107695452,
+      "learning_rate": 1.99080607078652e-05,
+      "loss": 1.06,
+      "step": 447
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.815619264362159,
+      "learning_rate": 1.9907353218653254e-05,
+      "loss": 1.0394,
+      "step": 448
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5390582350850919,
+      "learning_rate": 1.9906643030405194e-05,
+      "loss": 0.9819,
+      "step": 449
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5205153836874603,
+      "learning_rate": 1.990593014331449e-05,
+      "loss": 1.0491,
+      "step": 450
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6056494263108478,
+      "learning_rate": 1.9905214557575357e-05,
+      "loss": 1.0196,
+      "step": 451
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5559826127831224,
+      "learning_rate": 1.990449627338274e-05,
+      "loss": 0.9372,
+      "step": 452
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4761872388681696,
+      "learning_rate": 1.990377529093232e-05,
+      "loss": 1.0121,
+      "step": 453
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.703525297670604,
+      "learning_rate": 1.9903051610420513e-05,
+      "loss": 1.0269,
+      "step": 454
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5054716713238074,
+      "learning_rate": 1.990232523204447e-05,
+      "loss": 1.0465,
+      "step": 455
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6121865869864453,
+      "learning_rate": 1.9901596156002068e-05,
+      "loss": 1.0375,
+      "step": 456
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7807917397564925,
+      "learning_rate": 1.9900864382491936e-05,
+      "loss": 1.0955,
+      "step": 457
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4097293434748794,
+      "learning_rate": 1.9900129911713432e-05,
+      "loss": 0.9582,
+      "step": 458
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7116592226250393,
+      "learning_rate": 1.9899392743866638e-05,
+      "loss": 1.04,
+      "step": 459
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7387317555884088,
+      "learning_rate": 1.989865287915238e-05,
+      "loss": 0.977,
+      "step": 460
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1775498709759833,
+      "learning_rate": 1.9897910317772225e-05,
+      "loss": 1.0005,
+      "step": 461
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8294968857577099,
+      "learning_rate": 1.989716505992846e-05,
+      "loss": 0.9827,
+      "step": 462
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8810275201714084,
+      "learning_rate": 1.9896417105824113e-05,
+      "loss": 0.9714,
+      "step": 463
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.36651019008579,
+      "learning_rate": 1.9895666455662953e-05,
+      "loss": 0.9835,
+      "step": 464
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3585447617215967,
+      "learning_rate": 1.9894913109649473e-05,
+      "loss": 1.0176,
+      "step": 465
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6100825309998545,
+      "learning_rate": 1.9894157067988908e-05,
+      "loss": 0.9368,
+      "step": 466
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3114512239743978,
+      "learning_rate": 1.989339833088722e-05,
+      "loss": 1.0122,
+      "step": 467
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5010144995819839,
+      "learning_rate": 1.989263689855112e-05,
+      "loss": 0.9749,
+      "step": 468
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5059774695126955,
+      "learning_rate": 1.9891872771188033e-05,
+      "loss": 0.9067,
+      "step": 469
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6094562165127992,
+      "learning_rate": 1.989110594900613e-05,
+      "loss": 1.0768,
+      "step": 470
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4925052644007377,
+      "learning_rate": 1.989033643221432e-05,
+      "loss": 1.0608,
+      "step": 471
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.544481324366754,
+      "learning_rate": 1.9889564221022238e-05,
+      "loss": 0.9554,
+      "step": 472
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5266762681060964,
+      "learning_rate": 1.9888789315640253e-05,
+      "loss": 1.0018,
+      "step": 473
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4751214603497247,
+      "learning_rate": 1.9888011716279473e-05,
+      "loss": 0.943,
+      "step": 474
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3785953866989327,
+      "learning_rate": 1.9887231423151734e-05,
+      "loss": 0.9564,
+      "step": 475
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3859412827864828,
+      "learning_rate": 1.9886448436469618e-05,
+      "loss": 0.9552,
+      "step": 476
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8289578636784593,
+      "learning_rate": 1.988566275644642e-05,
+      "loss": 0.3187,
+      "step": 477
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3707416308466214,
+      "learning_rate": 1.988487438329619e-05,
+      "loss": 0.9648,
+      "step": 478
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6446717801385633,
+      "learning_rate": 1.98840833172337e-05,
+      "loss": 0.9258,
+      "step": 479
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2461160662276098,
+      "learning_rate": 1.988328955847446e-05,
+      "loss": 0.9164,
+      "step": 480
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6122334555239015,
+      "learning_rate": 1.9882493107234706e-05,
+      "loss": 0.8933,
+      "step": 481
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4534250252179526,
+      "learning_rate": 1.9881693963731418e-05,
+      "loss": 1.0058,
+      "step": 482
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3428309716123157,
+      "learning_rate": 1.9880892128182302e-05,
+      "loss": 1.0055,
+      "step": 483
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4970362703567461,
+      "learning_rate": 1.9880087600805807e-05,
+      "loss": 0.981,
+      "step": 484
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4071957461062348,
+      "learning_rate": 1.9879280381821104e-05,
+      "loss": 0.9707,
+      "step": 485
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4701499465382966,
+      "learning_rate": 1.9878470471448094e-05,
+      "loss": 1.0165,
+      "step": 486
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6010401333283664,
+      "learning_rate": 1.987765786990743e-05,
+      "loss": 0.9935,
+      "step": 487
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3917991517985413,
+      "learning_rate": 1.9876842577420484e-05,
+      "loss": 1.069,
+      "step": 488
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4724889693126333,
+      "learning_rate": 1.987602459420936e-05,
+      "loss": 1.0009,
+      "step": 489
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8976833241275014,
+      "learning_rate": 1.9875203920496905e-05,
+      "loss": 0.3302,
+      "step": 490
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6949833141619641,
+      "learning_rate": 1.987438055650669e-05,
+      "loss": 1.0055,
+      "step": 491
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.5229104062877368,
+      "learning_rate": 1.987355450246302e-05,
+      "loss": 1.038,
+      "step": 492
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4779530075832674,
+      "learning_rate": 1.9872725758590943e-05,
+      "loss": 0.9973,
+      "step": 493
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.421245705880415,
+      "learning_rate": 1.987189432511622e-05,
+      "loss": 0.9712,
+      "step": 494
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.469867318722636,
+      "learning_rate": 1.9871060202265367e-05,
+      "loss": 0.9934,
+      "step": 495
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7104782006206596,
+      "learning_rate": 1.9870223390265614e-05,
+      "loss": 1.0077,
+      "step": 496
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.706796633268622,
+      "learning_rate": 1.9869383889344937e-05,
+      "loss": 1.0335,
+      "step": 497
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.480755832863949,
+      "learning_rate": 1.9868541699732037e-05,
+      "loss": 0.9532,
+      "step": 498
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6526716957477727,
+      "learning_rate": 1.986769682165635e-05,
+      "loss": 0.9847,
+      "step": 499
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1786649935095883,
+      "learning_rate": 1.9866849255348045e-05,
+      "loss": 1.0877,
+      "step": 500
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4980741993228723,
+      "learning_rate": 1.9865999001038022e-05,
+      "loss": 0.951,
+      "step": 501
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.575089665729132,
+      "learning_rate": 1.986514605895791e-05,
+      "loss": 1.0699,
+      "step": 502
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7843754122845494,
+      "learning_rate": 1.9864290429340084e-05,
+      "loss": 1.003,
+      "step": 503
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4165192622628238,
+      "learning_rate": 1.9863432112417628e-05,
+      "loss": 0.9954,
+      "step": 504
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.593616671688619,
+      "learning_rate": 1.9862571108424377e-05,
+      "loss": 0.934,
+      "step": 505
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5011571931828283,
+      "learning_rate": 1.9861707417594896e-05,
+      "loss": 1.0165,
+      "step": 506
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7520585979918581,
+      "learning_rate": 1.9860841040164476e-05,
+      "loss": 0.9835,
+      "step": 507
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7396596527894448,
+      "learning_rate": 1.9859971976369136e-05,
+      "loss": 1.0211,
+      "step": 508
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7306438081944724,
+      "learning_rate": 1.9859100226445643e-05,
+      "loss": 1.1035,
+      "step": 509
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7422858229719738,
+      "learning_rate": 1.9858225790631477e-05,
+      "loss": 1.0117,
+      "step": 510
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7122130353245015,
+      "learning_rate": 1.9857348669164863e-05,
+      "loss": 1.0893,
+      "step": 511
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5510761829707955,
+      "learning_rate": 1.9856468862284752e-05,
+      "loss": 0.9938,
+      "step": 512
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.600675357534195,
+      "learning_rate": 1.9855586370230832e-05,
+      "loss": 0.9905,
+      "step": 513
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8417931222500095,
+      "learning_rate": 1.9854701193243507e-05,
+      "loss": 0.3402,
+      "step": 514
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4444928667550752,
+      "learning_rate": 1.9853813331563934e-05,
+      "loss": 1.0186,
+      "step": 515
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8910820355383855,
+      "learning_rate": 1.9852922785433985e-05,
+      "loss": 1.0379,
+      "step": 516
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6887345136293097,
+      "learning_rate": 1.9852029555096278e-05,
+      "loss": 1.0209,
+      "step": 517
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5834507190227096,
+      "learning_rate": 1.985113364079414e-05,
+      "loss": 0.9821,
+      "step": 518
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2275269601842889,
+      "learning_rate": 1.9850235042771655e-05,
+      "loss": 0.9621,
+      "step": 519
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8640374742348473,
+      "learning_rate": 1.984933376127362e-05,
+      "loss": 1.0059,
+      "step": 520
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8888832694970744,
+      "learning_rate": 1.9848429796545566e-05,
+      "loss": 0.3355,
+      "step": 521
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.814925186786731,
+      "learning_rate": 1.9847523148833767e-05,
+      "loss": 1.0618,
+      "step": 522
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.549381263161665,
+      "learning_rate": 1.9846613818385215e-05,
+      "loss": 0.9817,
+      "step": 523
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7808394402329957,
+      "learning_rate": 1.984570180544763e-05,
+      "loss": 0.3423,
+      "step": 524
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.530479577938019,
+      "learning_rate": 1.9844787110269478e-05,
+      "loss": 0.9335,
+      "step": 525
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5625682839731763,
+      "learning_rate": 1.984386973309994e-05,
+      "loss": 0.9963,
+      "step": 526
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5507215445144904,
+      "learning_rate": 1.9842949674188946e-05,
+      "loss": 0.9877,
+      "step": 527
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8920167261577927,
+      "learning_rate": 1.9842026933787134e-05,
+      "loss": 1.016,
+      "step": 528
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6458307679423931,
+      "learning_rate": 1.984110151214589e-05,
+      "loss": 0.9544,
+      "step": 529
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5886101119090976,
+      "learning_rate": 1.984017340951732e-05,
+      "loss": 1.0105,
+      "step": 530
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.592595595085504,
+      "learning_rate": 1.983924262615427e-05,
+      "loss": 0.997,
+      "step": 531
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6143874314245261,
+      "learning_rate": 1.9838309162310304e-05,
+      "loss": 0.9618,
+      "step": 532
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6251791227176355,
+      "learning_rate": 1.9837373018239733e-05,
+      "loss": 1.0344,
+      "step": 533
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6484005955547039,
+      "learning_rate": 1.983643419419758e-05,
+      "loss": 1.0121,
+      "step": 534
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6276407380127966,
+      "learning_rate": 1.983549269043961e-05,
+      "loss": 1.0624,
+      "step": 535
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5911812042159124,
+      "learning_rate": 1.9834548507222312e-05,
+      "loss": 0.9862,
+      "step": 536
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5357598975532294,
+      "learning_rate": 1.9833601644802915e-05,
+      "loss": 0.9959,
+      "step": 537
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4626722669422325,
+      "learning_rate": 1.983265210343936e-05,
+      "loss": 1.0341,
+      "step": 538
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.563432996072597,
+      "learning_rate": 1.9831699883390335e-05,
+      "loss": 0.926,
+      "step": 539
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.541617320144639,
+      "learning_rate": 1.9830744984915247e-05,
+      "loss": 0.9362,
+      "step": 540
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6672714895545047,
+      "learning_rate": 1.9829787408274247e-05,
+      "loss": 0.9223,
+      "step": 541
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.610094210069431,
+      "learning_rate": 1.982882715372819e-05,
+      "loss": 0.9344,
+      "step": 542
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7564744911497732,
+      "learning_rate": 1.9827864221538684e-05,
+      "loss": 1.0191,
+      "step": 543
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5489001948845225,
+      "learning_rate": 1.982689861196806e-05,
+      "loss": 0.9263,
+      "step": 544
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.512339729347485,
+      "learning_rate": 1.9825930325279373e-05,
+      "loss": 0.9847,
+      "step": 545
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4154611172666631,
+      "learning_rate": 1.982495936173641e-05,
+      "loss": 1.0316,
+      "step": 546
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7988069400513216,
+      "learning_rate": 1.9823985721603693e-05,
+      "loss": 0.9575,
+      "step": 547
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5701253516142373,
+      "learning_rate": 1.9823009405146465e-05,
+      "loss": 1.0409,
+      "step": 548
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4442892215336347,
+      "learning_rate": 1.98220304126307e-05,
+      "loss": 0.8872,
+      "step": 549
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8217126021464274,
+      "learning_rate": 1.9821048744323108e-05,
+      "loss": 0.9411,
+      "step": 550
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5172658286588774,
+      "learning_rate": 1.9820064400491118e-05,
+      "loss": 0.9895,
+      "step": 551
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5446771221060194,
+      "learning_rate": 1.9819077381402895e-05,
+      "loss": 1.0491,
+      "step": 552
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7391716680339055,
+      "learning_rate": 1.9818087687327328e-05,
+      "loss": 1.058,
+      "step": 553
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4993407829935261,
+      "learning_rate": 1.9817095318534038e-05,
+      "loss": 0.9445,
+      "step": 554
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.685212541974343,
+      "learning_rate": 1.981610027529337e-05,
+      "loss": 1.0597,
+      "step": 555
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4486640179510566,
+      "learning_rate": 1.9815102557876406e-05,
+      "loss": 0.9561,
+      "step": 556
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4420148642321307,
+      "learning_rate": 1.9814102166554954e-05,
+      "loss": 0.9668,
+      "step": 557
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5418215156501733,
+      "learning_rate": 1.981309910160154e-05,
+      "loss": 1.017,
+      "step": 558
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.61033671252527,
+      "learning_rate": 1.9812093363289433e-05,
+      "loss": 1.0394,
+      "step": 559
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6622225675803144,
+      "learning_rate": 1.9811084951892625e-05,
+      "loss": 0.8988,
+      "step": 560
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6713045686798444,
+      "learning_rate": 1.9810073867685828e-05,
+      "loss": 0.9534,
+      "step": 561
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0464720757135413,
+      "learning_rate": 1.980906011094449e-05,
+      "loss": 1.0253,
+      "step": 562
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6212485078684926,
+      "learning_rate": 1.9808043681944794e-05,
+      "loss": 1.0683,
+      "step": 563
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5537987635669974,
+      "learning_rate": 1.980702458096364e-05,
+      "loss": 1.0626,
+      "step": 564
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4552820328435718,
+      "learning_rate": 1.9806002808278658e-05,
+      "loss": 0.9823,
+      "step": 565
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6850626278621434,
+      "learning_rate": 1.9804978364168203e-05,
+      "loss": 1.0844,
+      "step": 566
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3830419941974257,
+      "learning_rate": 1.980395124891137e-05,
+      "loss": 1.0151,
+      "step": 567
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6673025569701518,
+      "learning_rate": 1.9802921462787968e-05,
+      "loss": 0.9543,
+      "step": 568
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.846863848915581,
+      "learning_rate": 1.980188900607854e-05,
+      "loss": 1.0018,
+      "step": 569
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7474502493652029,
+      "learning_rate": 1.9800853879064356e-05,
+      "loss": 0.9577,
+      "step": 570
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5046378300788403,
+      "learning_rate": 1.9799816082027413e-05,
+      "loss": 1.0407,
+      "step": 571
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5528864468175123,
+      "learning_rate": 1.9798775615250434e-05,
+      "loss": 0.9954,
+      "step": 572
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.825977782272403,
+      "learning_rate": 1.9797732479016874e-05,
+      "loss": 1.0272,
+      "step": 573
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.637205952713108,
+      "learning_rate": 1.979668667361091e-05,
+      "loss": 1.0378,
+      "step": 574
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5489515265026832,
+      "learning_rate": 1.9795638199317452e-05,
+      "loss": 0.9048,
+      "step": 575
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.440307956013016,
+      "learning_rate": 1.9794587056422125e-05,
+      "loss": 0.9603,
+      "step": 576
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.502972004265578,
+      "learning_rate": 1.9793533245211298e-05,
+      "loss": 0.9497,
+      "step": 577
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6354326991618577,
+      "learning_rate": 1.9792476765972055e-05,
+      "loss": 1.0276,
+      "step": 578
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4253277612656148,
+      "learning_rate": 1.979141761899221e-05,
+      "loss": 0.9528,
+      "step": 579
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9072753643670058,
+      "learning_rate": 1.9790355804560303e-05,
+      "loss": 0.3403,
+      "step": 580
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5730365640109556,
+      "learning_rate": 1.97892913229656e-05,
+      "loss": 1.0069,
+      "step": 581
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.153647820009763,
+      "learning_rate": 1.9788224174498098e-05,
+      "loss": 1.0298,
+      "step": 582
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3392441269046116,
+      "learning_rate": 1.9787154359448518e-05,
+      "loss": 0.8837,
+      "step": 583
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.749464771152944,
+      "learning_rate": 1.9786081878108304e-05,
+      "loss": 0.9364,
+      "step": 584
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5865814036377155,
+      "learning_rate": 1.9785006730769636e-05,
+      "loss": 0.96,
+      "step": 585
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5561946184480493,
+      "learning_rate": 1.9783928917725404e-05,
+      "loss": 1.0065,
+      "step": 586
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5148461346265685,
+      "learning_rate": 1.9782848439269244e-05,
+      "loss": 0.914,
+      "step": 587
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.661197143623179,
+      "learning_rate": 1.97817652956955e-05,
+      "loss": 0.9168,
+      "step": 588
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7360151521247759,
+      "learning_rate": 1.9780679487299255e-05,
+      "loss": 0.337,
+      "step": 589
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.526903974532986,
+      "learning_rate": 1.9779591014376312e-05,
+      "loss": 0.9939,
+      "step": 590
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3157470935434137,
+      "learning_rate": 1.9778499877223198e-05,
+      "loss": 0.9608,
+      "step": 591
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8192849490062599,
+      "learning_rate": 1.9777406076137174e-05,
+      "loss": 0.3163,
+      "step": 592
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4728078296169678,
+      "learning_rate": 1.977630961141622e-05,
+      "loss": 0.9406,
+      "step": 593
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7115210992278456,
+      "learning_rate": 1.9775210483359037e-05,
+      "loss": 1.017,
+      "step": 594
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.387427566598701,
+      "learning_rate": 1.977410869226507e-05,
+      "loss": 1.0243,
+      "step": 595
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5437284941082634,
+      "learning_rate": 1.9773004238434465e-05,
+      "loss": 0.9284,
+      "step": 596
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4392059151420191,
+      "learning_rate": 1.9771897122168112e-05,
+      "loss": 0.9605,
+      "step": 597
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7448877099396356,
+      "learning_rate": 1.9770787343767622e-05,
+      "loss": 0.9585,
+      "step": 598
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6424659504728774,
+      "learning_rate": 1.9769674903535324e-05,
+      "loss": 0.9871,
+      "step": 599
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4184094557276807,
+      "learning_rate": 1.976855980177428e-05,
+      "loss": 0.9796,
+      "step": 600
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8428809416753293,
+      "learning_rate": 1.9767442038788273e-05,
+      "loss": 0.3396,
+      "step": 601
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4714204677621197,
+      "learning_rate": 1.9766321614881814e-05,
+      "loss": 0.9283,
+      "step": 602
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9803664690802016,
+      "learning_rate": 1.976519853036014e-05,
+      "loss": 0.9631,
+      "step": 603
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.95296992926116,
+      "learning_rate": 1.9764072785529203e-05,
+      "loss": 1.034,
+      "step": 604
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.67264133898393,
+      "learning_rate": 1.9762944380695692e-05,
+      "loss": 1.0424,
+      "step": 605
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5533802161818493,
+      "learning_rate": 1.9761813316167014e-05,
+      "loss": 0.9876,
+      "step": 606
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6730692427326996,
+      "learning_rate": 1.9760679592251306e-05,
+      "loss": 0.9586,
+      "step": 607
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.834201012837671,
+      "learning_rate": 1.975954320925742e-05,
+      "loss": 0.9773,
+      "step": 608
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7643256712935411,
+      "learning_rate": 1.975840416749494e-05,
+      "loss": 1.0029,
+      "step": 609
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.447756337232945,
+      "learning_rate": 1.9757262467274173e-05,
+      "loss": 0.9421,
+      "step": 610
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6828710792886568,
+      "learning_rate": 1.975611810890615e-05,
+      "loss": 0.863,
+      "step": 611
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5781019328338877,
+      "learning_rate": 1.9754971092702623e-05,
+      "loss": 0.9429,
+      "step": 612
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.616119128613189,
+      "learning_rate": 1.9753821418976077e-05,
+      "loss": 0.9497,
+      "step": 613
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8413753957676127,
+      "learning_rate": 1.975266908803971e-05,
+      "loss": 1.0221,
+      "step": 614
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7576390000655862,
+      "learning_rate": 1.9751514100207444e-05,
+      "loss": 0.998,
+      "step": 615
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.532347846351968,
+      "learning_rate": 1.975035645579394e-05,
+      "loss": 0.9582,
+      "step": 616
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7654309962669035,
+      "learning_rate": 1.9749196155114568e-05,
+      "loss": 0.9294,
+      "step": 617
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8417837389237854,
+      "learning_rate": 1.9748033198485422e-05,
+      "loss": 0.939,
+      "step": 618
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8983221332990239,
+      "learning_rate": 1.974686758622333e-05,
+      "loss": 0.9404,
+      "step": 619
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.41671136601195,
+      "learning_rate": 1.9745699318645833e-05,
+      "loss": 0.9284,
+      "step": 620
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5361361670290725,
+      "learning_rate": 1.9744528396071197e-05,
+      "loss": 1.0013,
+      "step": 621
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4955227922733323,
+      "learning_rate": 1.9743354818818418e-05,
+      "loss": 1.071,
+      "step": 622
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9223903191922407,
+      "learning_rate": 1.974217858720721e-05,
+      "loss": 1.0495,
+      "step": 623
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6179983817389056,
+      "learning_rate": 1.9740999701558007e-05,
+      "loss": 0.97,
+      "step": 624
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.39945336690626,
+      "learning_rate": 1.9739818162191976e-05,
+      "loss": 0.9472,
+      "step": 625
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5573252263780237,
+      "learning_rate": 1.9738633969431e-05,
+      "loss": 0.9569,
+      "step": 626
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8919730759552575,
+      "learning_rate": 1.973744712359768e-05,
+      "loss": 1.0172,
+      "step": 627
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6220258997211678,
+      "learning_rate": 1.973625762501535e-05,
+      "loss": 0.9754,
+      "step": 628
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4493628562627832,
+      "learning_rate": 1.973506547400806e-05,
+      "loss": 0.8455,
+      "step": 629
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4652016348754395,
+      "learning_rate": 1.9733870670900586e-05,
+      "loss": 1.0017,
+      "step": 630
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.198366364166625,
+      "learning_rate": 1.973267321601843e-05,
+      "loss": 0.8924,
+      "step": 631
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.539601500583414,
+      "learning_rate": 1.9731473109687807e-05,
+      "loss": 1.0371,
+      "step": 632
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4713246272224831,
+      "learning_rate": 1.9730270352235657e-05,
+      "loss": 1.0249,
+      "step": 633
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4570848983061602,
+      "learning_rate": 1.9729064943989646e-05,
+      "loss": 0.9311,
+      "step": 634
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.470201882463915,
+      "learning_rate": 1.972785688527816e-05,
+      "loss": 0.9737,
+      "step": 635
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.567526958052572,
+      "learning_rate": 1.972664617643031e-05,
+      "loss": 0.9863,
+      "step": 636
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4042792538000235,
+      "learning_rate": 1.9725432817775925e-05,
+      "loss": 0.9627,
+      "step": 637
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4887018003786663,
+      "learning_rate": 1.9724216809645557e-05,
+      "loss": 0.9179,
+      "step": 638
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4677996445915789,
+      "learning_rate": 1.9722998152370482e-05,
+      "loss": 0.9562,
+      "step": 639
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4153372724811504,
+      "learning_rate": 1.9721776846282692e-05,
+      "loss": 0.9961,
+      "step": 640
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7276058119433573,
+      "learning_rate": 1.9720552891714912e-05,
+      "loss": 1.0181,
+      "step": 641
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4730948988886279,
+      "learning_rate": 1.9719326289000568e-05,
+      "loss": 1.0313,
+      "step": 642
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.750053732762348,
+      "learning_rate": 1.971809703847383e-05,
+      "loss": 0.9964,
+      "step": 643
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7496420279162306,
+      "learning_rate": 1.971686514046958e-05,
+      "loss": 0.9402,
+      "step": 644
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6719026819504315,
+      "learning_rate": 1.971563059532342e-05,
+      "loss": 0.9999,
+      "step": 645
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.367447775861517,
+      "learning_rate": 1.971439340337167e-05,
+      "loss": 1.0441,
+      "step": 646
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7507052985343017,
+      "learning_rate": 1.971315356495138e-05,
+      "loss": 0.9752,
+      "step": 647
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5748322063049078,
+      "learning_rate": 1.9711911080400313e-05,
+      "loss": 0.9808,
+      "step": 648
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7030050679509858,
+      "learning_rate": 1.971066595005696e-05,
+      "loss": 1.05,
+      "step": 649
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4900469662127558,
+      "learning_rate": 1.9709418174260523e-05,
+      "loss": 1.0449,
+      "step": 650
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.457156969787931,
+      "learning_rate": 1.9708167753350932e-05,
+      "loss": 0.9481,
+      "step": 651
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3808206009175872,
+      "learning_rate": 1.9706914687668842e-05,
+      "loss": 1.0495,
+      "step": 652
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8923930729142374,
+      "learning_rate": 1.9705658977555617e-05,
+      "loss": 0.3003,
+      "step": 653
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.610521852316017,
+      "learning_rate": 1.970440062335335e-05,
+      "loss": 0.9797,
+      "step": 654
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.399962063710708,
+      "learning_rate": 1.9703139625404847e-05,
+      "loss": 0.9548,
+      "step": 655
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.463369223098361,
+      "learning_rate": 1.9701875984053642e-05,
+      "loss": 0.9197,
+      "step": 656
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.2312646170813215,
+      "learning_rate": 1.9700609699643984e-05,
+      "loss": 1.0428,
+      "step": 657
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.597291799915292,
+      "learning_rate": 1.9699340772520847e-05,
+      "loss": 1.016,
+      "step": 658
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7694565907785076,
+      "learning_rate": 1.969806920302992e-05,
+      "loss": 0.9757,
+      "step": 659
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7935228250027269,
+      "learning_rate": 1.9696794991517613e-05,
+      "loss": 0.9329,
+      "step": 660
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6505922019468575,
+      "learning_rate": 1.9695518138331055e-05,
+      "loss": 0.9741,
+      "step": 661
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5367578628034648,
+      "learning_rate": 1.9694238643818097e-05,
+      "loss": 0.9973,
+      "step": 662
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4502875305165264,
+      "learning_rate": 1.9692956508327313e-05,
+      "loss": 0.9538,
+      "step": 663
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4146351004012245,
+      "learning_rate": 1.9691671732207986e-05,
+      "loss": 0.9769,
+      "step": 664
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7567028951603778,
+      "learning_rate": 1.9690384315810126e-05,
+      "loss": 0.9966,
+      "step": 665
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2693071152830517,
+      "learning_rate": 1.9689094259484462e-05,
+      "loss": 0.9405,
+      "step": 666
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.568509474729593,
+      "learning_rate": 1.968780156358244e-05,
+      "loss": 1.045,
+      "step": 667
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6338375811717487,
+      "learning_rate": 1.9686506228456226e-05,
+      "loss": 0.9617,
+      "step": 668
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8935986755243135,
+      "learning_rate": 1.9685208254458707e-05,
+      "loss": 1.0117,
+      "step": 669
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4740464405465192,
+      "learning_rate": 1.9683907641943484e-05,
+      "loss": 0.9596,
+      "step": 670
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7310962777025045,
+      "learning_rate": 1.9682604391264882e-05,
+      "loss": 1.0062,
+      "step": 671
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7158998786832855,
+      "learning_rate": 1.968129850277794e-05,
+      "loss": 0.9914,
+      "step": 672
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3717880158855513,
+      "learning_rate": 1.9679989976838417e-05,
+      "loss": 0.9737,
+      "step": 673
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3973638032947824,
+      "learning_rate": 1.9678678813802796e-05,
+      "loss": 0.9382,
+      "step": 674
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6606824500968425,
+      "learning_rate": 1.9677365014028275e-05,
+      "loss": 1.0297,
+      "step": 675
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8757940823127864,
+      "learning_rate": 1.9676048577872762e-05,
+      "loss": 0.3257,
+      "step": 676
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6892297122950004,
+      "learning_rate": 1.9674729505694894e-05,
+      "loss": 1.0191,
+      "step": 677
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3807853317373082,
+      "learning_rate": 1.9673407797854024e-05,
+      "loss": 0.9823,
+      "step": 678
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.721324456665953,
+      "learning_rate": 1.967208345471022e-05,
+      "loss": 1.0126,
+      "step": 679
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5577845785428022,
+      "learning_rate": 1.967075647662427e-05,
+      "loss": 0.9194,
+      "step": 680
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.727788862903958,
+      "learning_rate": 1.9669426863957685e-05,
+      "loss": 0.9571,
+      "step": 681
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7814221275968994,
+      "learning_rate": 1.9668094617072683e-05,
+      "loss": 0.3445,
+      "step": 682
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8131501957442242,
+      "learning_rate": 1.96667597363322e-05,
+      "loss": 1.0327,
+      "step": 683
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.632343488391787,
+      "learning_rate": 1.96654222220999e-05,
+      "loss": 0.9829,
+      "step": 684
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4787056830991403,
+      "learning_rate": 1.966408207474016e-05,
+      "loss": 0.9649,
+      "step": 685
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3966674731337554,
+      "learning_rate": 1.9662739294618067e-05,
+      "loss": 0.9872,
+      "step": 686
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.39018185391193,
+      "learning_rate": 1.9661393882099436e-05,
+      "loss": 0.9799,
+      "step": 687
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.475626628726188,
+      "learning_rate": 1.9660045837550798e-05,
+      "loss": 0.9701,
+      "step": 688
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5287454080941583,
+      "learning_rate": 1.965869516133939e-05,
+      "loss": 0.9643,
+      "step": 689
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7054515661264253,
+      "learning_rate": 1.9657341853833176e-05,
+      "loss": 1.087,
+      "step": 690
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.657946627428576,
+      "learning_rate": 1.9655985915400834e-05,
+      "loss": 0.9404,
+      "step": 691
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6907505621177301,
+      "learning_rate": 1.965462734641176e-05,
+      "loss": 0.9408,
+      "step": 692
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8749806015680115,
+      "learning_rate": 1.965326614723607e-05,
+      "loss": 0.9964,
+      "step": 693
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.45616433608959,
+      "learning_rate": 1.9651902318244582e-05,
+      "loss": 0.9314,
+      "step": 694
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0767157063071906,
+      "learning_rate": 1.9650535859808847e-05,
+      "loss": 0.999,
+      "step": 695
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5365009800331833,
+      "learning_rate": 1.9649166772301126e-05,
+      "loss": 1.0027,
+      "step": 696
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.441424532195454,
+      "learning_rate": 1.96477950560944e-05,
+      "loss": 0.9286,
+      "step": 697
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.349589205879623,
+      "learning_rate": 1.9646420711562354e-05,
+      "loss": 0.9229,
+      "step": 698
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4226175611955258,
+      "learning_rate": 1.9645043739079398e-05,
+      "loss": 1.0558,
+      "step": 699
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3311118518059941,
+      "learning_rate": 1.9643664139020666e-05,
+      "loss": 0.9723,
+      "step": 700
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7226782157258724,
+      "learning_rate": 1.964228191176199e-05,
+      "loss": 0.9137,
+      "step": 701
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5117564480046164,
+      "learning_rate": 1.964089705767993e-05,
+      "loss": 1.0138,
+      "step": 702
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.416227656352775,
+      "learning_rate": 1.963950957715176e-05,
+      "loss": 0.9068,
+      "step": 703
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3735480994446339,
+      "learning_rate": 1.9638119470555462e-05,
+      "loss": 0.9429,
+      "step": 704
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7373350161156125,
+      "learning_rate": 1.9636726738269752e-05,
+      "loss": 0.9379,
+      "step": 705
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7882266538742742,
+      "learning_rate": 1.9635331380674035e-05,
+      "loss": 0.9882,
+      "step": 706
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4977221056090533,
+      "learning_rate": 1.9633933398148452e-05,
+      "loss": 1.0493,
+      "step": 707
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4260295886797731,
+      "learning_rate": 1.9632532791073847e-05,
+      "loss": 0.895,
+      "step": 708
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5192417669480027,
+      "learning_rate": 1.963112955983179e-05,
+      "loss": 0.9807,
+      "step": 709
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1667505531833817,
+      "learning_rate": 1.962972370480456e-05,
+      "loss": 0.9853,
+      "step": 710
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6719722719505665,
+      "learning_rate": 1.9628315226375146e-05,
+      "loss": 0.9816,
+      "step": 711
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7255617700254204,
+      "learning_rate": 1.962690412492726e-05,
+      "loss": 1.0424,
+      "step": 712
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5500255732580126,
+      "learning_rate": 1.9625490400845318e-05,
+      "loss": 0.9391,
+      "step": 713
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6285249929417152,
+      "learning_rate": 1.9624074054514467e-05,
+      "loss": 0.9907,
+      "step": 714
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5597101601498649,
+      "learning_rate": 1.9622655086320553e-05,
+      "loss": 1.0135,
+      "step": 715
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5207713302428703,
+      "learning_rate": 1.9621233496650143e-05,
+      "loss": 0.9534,
+      "step": 716
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5493225881391293,
+      "learning_rate": 1.9619809285890522e-05,
+      "loss": 0.962,
+      "step": 717
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.50810602955461,
+      "learning_rate": 1.9618382454429675e-05,
+      "loss": 0.9718,
+      "step": 718
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2724014690640397,
+      "learning_rate": 1.9616953002656316e-05,
+      "loss": 0.9657,
+      "step": 719
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9781501948064185,
+      "learning_rate": 1.961552093095987e-05,
+      "loss": 0.3045,
+      "step": 720
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8744044167605414,
+      "learning_rate": 1.9614086239730464e-05,
+      "loss": 1.0012,
+      "step": 721
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.473644649642615,
+      "learning_rate": 1.9612648929358953e-05,
+      "loss": 0.9605,
+      "step": 722
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.468918031100101,
+      "learning_rate": 1.96112090002369e-05,
+      "loss": 1.0273,
+      "step": 723
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6835241369738312,
+      "learning_rate": 1.960976645275658e-05,
+      "loss": 0.9245,
+      "step": 724
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.560987440184807,
+      "learning_rate": 1.9608321287310988e-05,
+      "loss": 0.9835,
+      "step": 725
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7013735366972578,
+      "learning_rate": 1.9606873504293818e-05,
+      "loss": 0.9607,
+      "step": 726
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.771160337932077,
+      "learning_rate": 1.9605423104099492e-05,
+      "loss": 0.9985,
+      "step": 727
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5019205620637528,
+      "learning_rate": 1.9603970087123138e-05,
+      "loss": 0.9475,
+      "step": 728
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.365805532363745,
+      "learning_rate": 1.9602514453760594e-05,
+      "loss": 0.979,
+      "step": 729
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5650361804561912,
+      "learning_rate": 1.960105620440842e-05,
+      "loss": 0.9291,
+      "step": 730
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5180372000381994,
+      "learning_rate": 1.959959533946388e-05,
+      "loss": 1.0143,
+      "step": 731
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5059946484298883,
+      "learning_rate": 1.9598131859324957e-05,
+      "loss": 0.9948,
+      "step": 732
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3796741695019268,
+      "learning_rate": 1.9596665764390344e-05,
+      "loss": 0.9525,
+      "step": 733
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5811880931920612,
+      "learning_rate": 1.959519705505944e-05,
+      "loss": 0.9811,
+      "step": 734
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4000327628662808,
+      "learning_rate": 1.9593725731732366e-05,
+      "loss": 0.9445,
+      "step": 735
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.541705488401343,
+      "learning_rate": 1.9592251794809945e-05,
+      "loss": 0.8818,
+      "step": 736
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.546617510586694,
+      "learning_rate": 1.959077524469373e-05,
+      "loss": 0.9014,
+      "step": 737
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.424102952254737,
+      "learning_rate": 1.9589296081785966e-05,
+      "loss": 0.9278,
+      "step": 738
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4411367291040114,
+      "learning_rate": 1.9587814306489616e-05,
+      "loss": 1.0,
+      "step": 739
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8873381590103104,
+      "learning_rate": 1.958632991920836e-05,
+      "loss": 0.9418,
+      "step": 740
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5136138763391123,
+      "learning_rate": 1.9584842920346587e-05,
+      "loss": 1.0045,
+      "step": 741
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8201100754642396,
+      "learning_rate": 1.9583353310309393e-05,
+      "loss": 0.9731,
+      "step": 742
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8966083357799485,
+      "learning_rate": 1.9581861089502593e-05,
+      "loss": 0.3393,
+      "step": 743
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5898697626015028,
+      "learning_rate": 1.9580366258332703e-05,
+      "loss": 0.9036,
+      "step": 744
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3720072633368825,
+      "learning_rate": 1.957886881720696e-05,
+      "loss": 1.0097,
+      "step": 745
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.771442218308349,
+      "learning_rate": 1.957736876653331e-05,
+      "loss": 0.9831,
+      "step": 746
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3587493124522976,
+      "learning_rate": 1.9575866106720405e-05,
+      "loss": 0.9454,
+      "step": 747
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5740061828229877,
+      "learning_rate": 1.957436083817762e-05,
+      "loss": 0.9369,
+      "step": 748
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5381501859536975,
+      "learning_rate": 1.9572852961315014e-05,
+      "loss": 0.9945,
+      "step": 749
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7111310544553027,
+      "learning_rate": 1.9571342476543387e-05,
+      "loss": 0.9366,
+      "step": 750
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4992931831532665,
+      "learning_rate": 1.9569829384274236e-05,
+      "loss": 1.0093,
+      "step": 751
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5998138359943668,
+      "learning_rate": 1.9568313684919765e-05,
+      "loss": 1.0145,
+      "step": 752
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6760071210983765,
+      "learning_rate": 1.9566795378892894e-05,
+      "loss": 1.0251,
+      "step": 753
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.506149623967082,
+      "learning_rate": 1.9565274466607248e-05,
+      "loss": 1.0222,
+      "step": 754
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4014235854015034,
+      "learning_rate": 1.9563750948477174e-05,
+      "loss": 0.951,
+      "step": 755
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5764428500224703,
+      "learning_rate": 1.9562224824917716e-05,
+      "loss": 0.9965,
+      "step": 756
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5551635121890086,
+      "learning_rate": 1.956069609634463e-05,
+      "loss": 0.9541,
+      "step": 757
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.334932481145727,
+      "learning_rate": 1.9559164763174384e-05,
+      "loss": 0.9184,
+      "step": 758
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7232799136344867,
+      "learning_rate": 1.9557630825824156e-05,
+      "loss": 0.9821,
+      "step": 759
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4842009015631046,
+      "learning_rate": 1.9556094284711834e-05,
+      "loss": 0.9316,
+      "step": 760
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4021308001397177,
+      "learning_rate": 1.9554555140256016e-05,
+      "loss": 0.9371,
+      "step": 761
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6581601428245059,
+      "learning_rate": 1.9553013392876005e-05,
+      "loss": 1.0304,
+      "step": 762
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5336689939807528,
+      "learning_rate": 1.955146904299181e-05,
+      "loss": 0.9504,
+      "step": 763
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4200238023350953,
+      "learning_rate": 1.9549922091024164e-05,
+      "loss": 0.9908,
+      "step": 764
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.600408378979318,
+      "learning_rate": 1.9548372537394494e-05,
+      "loss": 0.9249,
+      "step": 765
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3351951775922273,
+      "learning_rate": 1.9546820382524943e-05,
+      "loss": 0.9492,
+      "step": 766
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3801701760115814,
+      "learning_rate": 1.9545265626838362e-05,
+      "loss": 0.9564,
+      "step": 767
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4167002923448229,
+      "learning_rate": 1.9543708270758306e-05,
+      "loss": 1.0039,
+      "step": 768
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4814022330782992,
+      "learning_rate": 1.9542148314709042e-05,
+      "loss": 0.9401,
+      "step": 769
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7291027375883745,
+      "learning_rate": 1.9540585759115548e-05,
+      "loss": 0.9193,
+      "step": 770
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4139960830591987,
+      "learning_rate": 1.9539020604403502e-05,
+      "loss": 0.9399,
+      "step": 771
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.560926802916431,
+      "learning_rate": 1.95374528509993e-05,
+      "loss": 1.0697,
+      "step": 772
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5753101347232217,
+      "learning_rate": 1.9535882499330044e-05,
+      "loss": 1.0426,
+      "step": 773
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6574101899872662,
+      "learning_rate": 1.9534309549823535e-05,
+      "loss": 0.8504,
+      "step": 774
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8636191659484618,
+      "learning_rate": 1.9532734002908287e-05,
+      "loss": 0.9566,
+      "step": 775
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4122755657238062,
+      "learning_rate": 1.9531155859013527e-05,
+      "loss": 0.9675,
+      "step": 776
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3879651781044995,
+      "learning_rate": 1.9529575118569185e-05,
+      "loss": 0.9536,
+      "step": 777
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.429765313211883,
+      "learning_rate": 1.952799178200589e-05,
+      "loss": 0.9512,
+      "step": 778
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4069032618666792,
+      "learning_rate": 1.9526405849755003e-05,
+      "loss": 0.9536,
+      "step": 779
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.387120905172671,
+      "learning_rate": 1.9524817322248558e-05,
+      "loss": 0.9271,
+      "step": 780
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5871068584995849,
+      "learning_rate": 1.9523226199919326e-05,
+      "loss": 1.0133,
+      "step": 781
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.582381055564859,
+      "learning_rate": 1.9521632483200767e-05,
+      "loss": 0.9479,
+      "step": 782
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4830617390703464,
+      "learning_rate": 1.952003617252705e-05,
+      "loss": 0.9778,
+      "step": 783
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5597433652275832,
+      "learning_rate": 1.9518437268333062e-05,
+      "loss": 0.9918,
+      "step": 784
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4565578636407874,
+      "learning_rate": 1.951683577105438e-05,
+      "loss": 0.9601,
+      "step": 785
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.512021029916176,
+      "learning_rate": 1.9515231681127308e-05,
+      "loss": 1.0207,
+      "step": 786
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9285301152705443,
+      "learning_rate": 1.9513624998988833e-05,
+      "loss": 0.9738,
+      "step": 787
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4349048573688477,
+      "learning_rate": 1.9512015725076666e-05,
+      "loss": 0.9433,
+      "step": 788
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3277804664104127,
+      "learning_rate": 1.9510403859829213e-05,
+      "loss": 1.018,
+      "step": 789
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4533672887046032,
+      "learning_rate": 1.9508789403685594e-05,
+      "loss": 0.9801,
+      "step": 790
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3128964734989959,
+      "learning_rate": 1.950717235708563e-05,
+      "loss": 0.8838,
+      "step": 791
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4273050085952863,
+      "learning_rate": 1.950555272046985e-05,
+      "loss": 1.0446,
+      "step": 792
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4562282311721848,
+      "learning_rate": 1.9503930494279487e-05,
+      "loss": 0.9752,
+      "step": 793
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.434527031003438,
+      "learning_rate": 1.9502305678956476e-05,
+      "loss": 0.973,
+      "step": 794
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.58974759528106,
+      "learning_rate": 1.9500678274943467e-05,
+      "loss": 0.9697,
+      "step": 795
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4101459900385027,
+      "learning_rate": 1.949904828268381e-05,
+      "loss": 0.9928,
+      "step": 796
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4357921883796063,
+      "learning_rate": 1.949741570262156e-05,
+      "loss": 0.9663,
+      "step": 797
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4032743727850312,
+      "learning_rate": 1.949578053520147e-05,
+      "loss": 1.0036,
+      "step": 798
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6426460589990182,
+      "learning_rate": 1.9494142780869013e-05,
+      "loss": 1.0142,
+      "step": 799
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.56254976193343,
+      "learning_rate": 1.9492502440070353e-05,
+      "loss": 1.0101,
+      "step": 800
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5994100190275868,
+      "learning_rate": 1.949085951325237e-05,
+      "loss": 0.9953,
+      "step": 801
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6477139057709178,
+      "learning_rate": 1.9489214000862633e-05,
+      "loss": 0.9475,
+      "step": 802
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8477935190603303,
+      "learning_rate": 1.948756590334943e-05,
+      "loss": 1.0045,
+      "step": 803
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4038817494445788,
+      "learning_rate": 1.9485915221161753e-05,
+      "loss": 0.9542,
+      "step": 804
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4851164779123587,
+      "learning_rate": 1.9484261954749286e-05,
+      "loss": 0.9602,
+      "step": 805
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5985778223552407,
+      "learning_rate": 1.9482606104562428e-05,
+      "loss": 0.9253,
+      "step": 806
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.270414929582952,
+      "learning_rate": 1.9480947671052278e-05,
+      "loss": 0.97,
+      "step": 807
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.334368682506892,
+      "learning_rate": 1.947928665467064e-05,
+      "loss": 1.0086,
+      "step": 808
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6859505358997195,
+      "learning_rate": 1.9477623055870016e-05,
+      "loss": 0.9201,
+      "step": 809
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4899527532372088,
+      "learning_rate": 1.9475956875103623e-05,
+      "loss": 0.938,
+      "step": 810
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.873172251629141,
+      "learning_rate": 1.947428811282537e-05,
+      "loss": 0.9749,
+      "step": 811
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3575691870472932,
+      "learning_rate": 1.947261676948987e-05,
+      "loss": 0.9536,
+      "step": 812
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.412110665361651,
+      "learning_rate": 1.947094284555245e-05,
+      "loss": 0.909,
+      "step": 813
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4137851089996896,
+      "learning_rate": 1.9469266341469136e-05,
+      "loss": 0.9555,
+      "step": 814
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6931908495024213,
+      "learning_rate": 1.9467587257696646e-05,
+      "loss": 0.9518,
+      "step": 815
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.681237618851009,
+      "learning_rate": 1.946590559469241e-05,
+      "loss": 0.9068,
+      "step": 816
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5138215101790222,
+      "learning_rate": 1.946422135291456e-05,
+      "loss": 0.9392,
+      "step": 817
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4173301987688298,
+      "learning_rate": 1.9462534532821934e-05,
+      "loss": 0.9422,
+      "step": 818
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.412644124115625,
+      "learning_rate": 1.946084513487406e-05,
+      "loss": 0.8784,
+      "step": 819
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4431817851421995,
+      "learning_rate": 1.9459153159531188e-05,
+      "loss": 0.9321,
+      "step": 820
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3078977894146726,
+      "learning_rate": 1.9457458607254247e-05,
+      "loss": 0.9175,
+      "step": 821
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.388140340608216,
+      "learning_rate": 1.9455761478504888e-05,
+      "loss": 0.9785,
+      "step": 822
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4435349513776867,
+      "learning_rate": 1.9454061773745448e-05,
+      "loss": 0.9672,
+      "step": 823
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6256350493697957,
+      "learning_rate": 1.9452359493438984e-05,
+      "loss": 1.0384,
+      "step": 824
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4608218298984064,
+      "learning_rate": 1.9450654638049235e-05,
+      "loss": 0.9847,
+      "step": 825
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5046955669192774,
+      "learning_rate": 1.9448947208040655e-05,
+      "loss": 0.9557,
+      "step": 826
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7634757541211445,
+      "learning_rate": 1.9447237203878395e-05,
+      "loss": 0.9673,
+      "step": 827
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7766129783464395,
+      "learning_rate": 1.9445524626028303e-05,
+      "loss": 0.9198,
+      "step": 828
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5245188305010635,
+      "learning_rate": 1.9443809474956937e-05,
+      "loss": 0.9896,
+      "step": 829
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3239636632659209,
+      "learning_rate": 1.944209175113155e-05,
+      "loss": 0.9924,
+      "step": 830
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3373024998275491,
+      "learning_rate": 1.9440371455020094e-05,
+      "loss": 0.9313,
+      "step": 831
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.750510368048217,
+      "learning_rate": 1.943864858709123e-05,
+      "loss": 0.9856,
+      "step": 832
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.518554602865235,
+      "learning_rate": 1.9436923147814317e-05,
+      "loss": 0.9882,
+      "step": 833
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5232333248206844,
+      "learning_rate": 1.9435195137659404e-05,
+      "loss": 0.9433,
+      "step": 834
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.418079630664839,
+      "learning_rate": 1.9433464557097257e-05,
+      "loss": 0.9344,
+      "step": 835
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.357032280868288,
+      "learning_rate": 1.9431731406599325e-05,
+      "loss": 0.9845,
+      "step": 836
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5619016741535259,
+      "learning_rate": 1.9429995686637774e-05,
+      "loss": 1.0308,
+      "step": 837
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.932681854596408,
+      "learning_rate": 1.9428257397685456e-05,
+      "loss": 0.9207,
+      "step": 838
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4217283977659552,
+      "learning_rate": 1.942651654021593e-05,
+      "loss": 0.9916,
+      "step": 839
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4439132663912166,
+      "learning_rate": 1.942477311470346e-05,
+      "loss": 0.8862,
+      "step": 840
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4432247115822097,
+      "learning_rate": 1.9423027121622995e-05,
+      "loss": 1.0597,
+      "step": 841
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4411351480729526,
+      "learning_rate": 1.9421278561450195e-05,
+      "loss": 0.9145,
+      "step": 842
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.721403304238875,
+      "learning_rate": 1.9419527434661418e-05,
+      "loss": 0.9384,
+      "step": 843
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6063558565178955,
+      "learning_rate": 1.9417773741733713e-05,
+      "loss": 0.9649,
+      "step": 844
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3479855763899746,
+      "learning_rate": 1.9416017483144844e-05,
+      "loss": 0.9472,
+      "step": 845
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.400074300593139,
+      "learning_rate": 1.941425865937325e-05,
+      "loss": 0.9639,
+      "step": 846
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5260155309927952,
+      "learning_rate": 1.9412497270898097e-05,
+      "loss": 0.9922,
+      "step": 847
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4907098251675208,
+      "learning_rate": 1.9410733318199233e-05,
+      "loss": 0.9049,
+      "step": 848
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.389156664020996,
+      "learning_rate": 1.9408966801757204e-05,
+      "loss": 0.9888,
+      "step": 849
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4590221786963395,
+      "learning_rate": 1.9407197722053254e-05,
+      "loss": 0.9458,
+      "step": 850
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.483571110934844,
+      "learning_rate": 1.9405426079569337e-05,
+      "loss": 0.9709,
+      "step": 851
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4181297193802171,
+      "learning_rate": 1.940365187478809e-05,
+      "loss": 0.9759,
+      "step": 852
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1491552319143914,
+      "learning_rate": 1.9401875108192863e-05,
+      "loss": 0.3367,
+      "step": 853
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9316460771221197,
+      "learning_rate": 1.940009578026769e-05,
+      "loss": 0.3313,
+      "step": 854
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5256254580771318,
+      "learning_rate": 1.9398313891497314e-05,
+      "loss": 0.9626,
+      "step": 855
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4357723694764464,
+      "learning_rate": 1.9396529442367167e-05,
+      "loss": 0.9401,
+      "step": 856
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9344649260443684,
+      "learning_rate": 1.939474243336338e-05,
+      "loss": 0.3439,
+      "step": 857
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5587911020783471,
+      "learning_rate": 1.939295286497279e-05,
+      "loss": 0.9467,
+      "step": 858
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6964086798975246,
+      "learning_rate": 1.9391160737682924e-05,
+      "loss": 0.9435,
+      "step": 859
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7229682527366523,
+      "learning_rate": 1.9389366051982e-05,
+      "loss": 0.942,
+      "step": 860
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.38940016981017,
+      "learning_rate": 1.9387568808358944e-05,
+      "loss": 1.0144,
+      "step": 861
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4658141574438126,
+      "learning_rate": 1.9385769007303374e-05,
+      "loss": 0.9858,
+      "step": 862
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8948778120606111,
+      "learning_rate": 1.9383966649305608e-05,
+      "loss": 0.8593,
+      "step": 863
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3086717877676333,
+      "learning_rate": 1.9382161734856656e-05,
+      "loss": 0.9415,
+      "step": 864
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.478112034030081,
+      "learning_rate": 1.9380354264448225e-05,
+      "loss": 1.0231,
+      "step": 865
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.759533252891076,
+      "learning_rate": 1.9378544238572723e-05,
+      "loss": 0.9382,
+      "step": 866
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.52135385282774,
+      "learning_rate": 1.937673165772325e-05,
+      "loss": 1.0443,
+      "step": 867
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8633761281823003,
+      "learning_rate": 1.93749165223936e-05,
+      "loss": 0.9817,
+      "step": 868
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4416107043069595,
+      "learning_rate": 1.937309883307827e-05,
+      "loss": 0.9791,
+      "step": 869
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4080848353169404,
+      "learning_rate": 1.9371278590272448e-05,
+      "loss": 0.3452,
+      "step": 870
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6608887481352994,
+      "learning_rate": 1.9369455794472017e-05,
+      "loss": 0.9245,
+      "step": 871
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5834200708782555,
+      "learning_rate": 1.9367630446173554e-05,
+      "loss": 1.0297,
+      "step": 872
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.78904610088039,
+      "learning_rate": 1.936580254587434e-05,
+      "loss": 0.3378,
+      "step": 873
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6835572798600182,
+      "learning_rate": 1.936397209407234e-05,
+      "loss": 0.9627,
+      "step": 874
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8137946136767582,
+      "learning_rate": 1.9362139091266227e-05,
+      "loss": 0.3578,
+      "step": 875
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3846981606506763,
+      "learning_rate": 1.936030353795535e-05,
+      "loss": 0.9149,
+      "step": 876
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5046467292640469,
+      "learning_rate": 1.9358465434639773e-05,
+      "loss": 0.9604,
+      "step": 877
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.444163578471167,
+      "learning_rate": 1.9356624781820246e-05,
+      "loss": 0.9823,
+      "step": 878
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.438098159762969,
+      "learning_rate": 1.9354781579998212e-05,
+      "loss": 1.0007,
+      "step": 879
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4544287499856805,
+      "learning_rate": 1.9352935829675807e-05,
+      "loss": 0.8822,
+      "step": 880
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0810065969506604,
+      "learning_rate": 1.935108753135587e-05,
+      "loss": 0.337,
+      "step": 881
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5879198519748934,
+      "learning_rate": 1.9349236685541924e-05,
+      "loss": 0.9747,
+      "step": 882
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4078565374720287,
+      "learning_rate": 1.934738329273819e-05,
+      "loss": 0.921,
+      "step": 883
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7054212936014457,
+      "learning_rate": 1.9345527353449583e-05,
+      "loss": 1.0073,
+      "step": 884
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4276576001963661,
+      "learning_rate": 1.9343668868181722e-05,
+      "loss": 0.9999,
+      "step": 885
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4928971857318731,
+      "learning_rate": 1.9341807837440893e-05,
+      "loss": 0.9131,
+      "step": 886
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6622502965391146,
+      "learning_rate": 1.9339944261734107e-05,
+      "loss": 0.9812,
+      "step": 887
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4709324898511884,
+      "learning_rate": 1.933807814156904e-05,
+      "loss": 0.8792,
+      "step": 888
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3444750009625994,
+      "learning_rate": 1.933620947745409e-05,
+      "loss": 1.0058,
+      "step": 889
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4711977429638057,
+      "learning_rate": 1.9334338269898317e-05,
+      "loss": 0.9403,
+      "step": 890
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6298787898064873,
+      "learning_rate": 1.9332464519411503e-05,
+      "loss": 0.9068,
+      "step": 891
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3491151887569088,
+      "learning_rate": 1.93305882265041e-05,
+      "loss": 1.0236,
+      "step": 892
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.476311642161689,
+      "learning_rate": 1.9328709391687264e-05,
+      "loss": 0.8896,
+      "step": 893
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5163124967156196,
+      "learning_rate": 1.932682801547284e-05,
+      "loss": 0.8527,
+      "step": 894
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4905473797859936,
+      "learning_rate": 1.9324944098373375e-05,
+      "loss": 0.9881,
+      "step": 895
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3714950006997435,
+      "learning_rate": 1.932305764090209e-05,
+      "loss": 0.9674,
+      "step": 896
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6337608761359723,
+      "learning_rate": 1.9321168643572912e-05,
+      "loss": 0.9402,
+      "step": 897
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7127512545633239,
+      "learning_rate": 1.931927710690046e-05,
+      "loss": 0.9674,
+      "step": 898
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3594732975223536,
+      "learning_rate": 1.9317383031400032e-05,
+      "loss": 0.9658,
+      "step": 899
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7858834527983056,
+      "learning_rate": 1.9315486417587636e-05,
+      "loss": 0.9369,
+      "step": 900
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4115773850347189,
+      "learning_rate": 1.9313587265979952e-05,
+      "loss": 0.988,
+      "step": 901
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3976324350728033,
+      "learning_rate": 1.9311685577094368e-05,
+      "loss": 0.8704,
+      "step": 902
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4979896030892448,
+      "learning_rate": 1.9309781351448956e-05,
+      "loss": 0.9496,
+      "step": 903
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6238756141206292,
+      "learning_rate": 1.9307874589562474e-05,
+      "loss": 0.9588,
+      "step": 904
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6185265965473552,
+      "learning_rate": 1.9305965291954387e-05,
+      "loss": 1.0218,
+      "step": 905
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.474152080416173,
+      "learning_rate": 1.930405345914483e-05,
+      "loss": 0.9648,
+      "step": 906
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4232492041804214,
+      "learning_rate": 1.9302139091654643e-05,
+      "loss": 0.9369,
+      "step": 907
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6567992984183608,
+      "learning_rate": 1.9300222190005352e-05,
+      "loss": 0.9155,
+      "step": 908
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3238791639395158,
+      "learning_rate": 1.9298302754719177e-05,
+      "loss": 0.9655,
+      "step": 909
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4660089223415227,
+      "learning_rate": 1.929638078631902e-05,
+      "loss": 0.9456,
+      "step": 910
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4514647036921116,
+      "learning_rate": 1.9294456285328482e-05,
+      "loss": 0.9856,
+      "step": 911
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6777324221314334,
+      "learning_rate": 1.9292529252271848e-05,
+      "loss": 1.0106,
+      "step": 912
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.658859535255508,
+      "learning_rate": 1.9290599687674095e-05,
+      "loss": 0.9542,
+      "step": 913
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.498134261097887,
+      "learning_rate": 1.928866759206089e-05,
+      "loss": 0.9072,
+      "step": 914
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2699501677904774,
+      "learning_rate": 1.9286732965958593e-05,
+      "loss": 0.9288,
+      "step": 915
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5443765235943152,
+      "learning_rate": 1.928479580989424e-05,
+      "loss": 0.9179,
+      "step": 916
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2770033624895287,
+      "learning_rate": 1.9282856124395578e-05,
+      "loss": 0.9704,
+      "step": 917
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5566634783150148,
+      "learning_rate": 1.9280913909991018e-05,
+      "loss": 0.9557,
+      "step": 918
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6633964988921692,
+      "learning_rate": 1.9278969167209682e-05,
+      "loss": 1.0389,
+      "step": 919
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5818093309050052,
+      "learning_rate": 1.927702189658137e-05,
+      "loss": 0.9543,
+      "step": 920
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2904364719504096,
+      "learning_rate": 1.9275072098636568e-05,
+      "loss": 0.949,
+      "step": 921
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2946359317252538,
+      "learning_rate": 1.927311977390646e-05,
+      "loss": 1.045,
+      "step": 922
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4565977904041179,
+      "learning_rate": 1.927116492292291e-05,
+      "loss": 0.9678,
+      "step": 923
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4590149663044147,
+      "learning_rate": 1.926920754621847e-05,
+      "loss": 0.9891,
+      "step": 924
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3077281024549297,
+      "learning_rate": 1.9267247644326392e-05,
+      "loss": 1.0065,
+      "step": 925
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2745035337413722,
+      "learning_rate": 1.92652852177806e-05,
+      "loss": 0.941,
+      "step": 926
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2482031700459941,
+      "learning_rate": 1.9263320267115716e-05,
+      "loss": 0.907,
+      "step": 927
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4784136639993546,
+      "learning_rate": 1.9261352792867047e-05,
+      "loss": 0.9325,
+      "step": 928
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4441549085819223,
+      "learning_rate": 1.9259382795570588e-05,
+      "loss": 0.9781,
+      "step": 929
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5463356248119329,
+      "learning_rate": 1.9257410275763018e-05,
+      "loss": 0.9718,
+      "step": 930
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.709275710205316,
+      "learning_rate": 1.9255435233981706e-05,
+      "loss": 0.9764,
+      "step": 931
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4669741854822027,
+      "learning_rate": 1.9253457670764707e-05,
+      "loss": 1.0407,
+      "step": 932
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3359093635228503,
+      "learning_rate": 1.9251477586650768e-05,
+      "loss": 0.9561,
+      "step": 933
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.542345499362906,
+      "learning_rate": 1.9249494982179313e-05,
+      "loss": 1.0087,
+      "step": 934
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8106102093891439,
+      "learning_rate": 1.9247509857890466e-05,
+      "loss": 0.9952,
+      "step": 935
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3860013372009408,
+      "learning_rate": 1.9245522214325022e-05,
+      "loss": 0.8171,
+      "step": 936
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9998644483906243,
+      "learning_rate": 1.9243532052024473e-05,
+      "loss": 0.9111,
+      "step": 937
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6482025652518015,
+      "learning_rate": 1.9241539371530994e-05,
+      "loss": 0.9801,
+      "step": 938
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4191058232833251,
+      "learning_rate": 1.9239544173387442e-05,
+      "loss": 0.9511,
+      "step": 939
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.538327250245726,
+      "learning_rate": 1.923754645813737e-05,
+      "loss": 0.9404,
+      "step": 940
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.829724023122164,
+      "learning_rate": 1.923554622632501e-05,
+      "loss": 0.9411,
+      "step": 941
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9249830622789615,
+      "learning_rate": 1.9233543478495277e-05,
+      "loss": 0.3461,
+      "step": 942
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4056394297922208,
+      "learning_rate": 1.9231538215193776e-05,
+      "loss": 0.9774,
+      "step": 943
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4900537377616232,
+      "learning_rate": 1.9229530436966796e-05,
+      "loss": 0.967,
+      "step": 944
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5022591166330743,
+      "learning_rate": 1.9227520144361312e-05,
+      "loss": 1.0246,
+      "step": 945
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3101190616331866,
+      "learning_rate": 1.9225507337924984e-05,
+      "loss": 0.9364,
+      "step": 946
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4243306740225017,
+      "learning_rate": 1.9223492018206155e-05,
+      "loss": 0.9121,
+      "step": 947
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3578198638492456,
+      "learning_rate": 1.922147418575385e-05,
+      "loss": 1.0123,
+      "step": 948
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3629551937544702,
+      "learning_rate": 1.9219453841117787e-05,
+      "loss": 0.8879,
+      "step": 949
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5465134730423564,
+      "learning_rate": 1.921743098484836e-05,
+      "loss": 0.8955,
+      "step": 950
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6151448459099016,
+      "learning_rate": 1.9215405617496657e-05,
+      "loss": 0.9913,
+      "step": 951
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.35675638681422,
+      "learning_rate": 1.921337773961444e-05,
+      "loss": 0.8799,
+      "step": 952
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.546682985265901,
+      "learning_rate": 1.9211347351754154e-05,
+      "loss": 0.9863,
+      "step": 953
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5544977521599055,
+      "learning_rate": 1.9209314454468935e-05,
+      "loss": 0.9633,
+      "step": 954
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7320375822654397,
+      "learning_rate": 1.9207279048312612e-05,
+      "loss": 0.9388,
+      "step": 955
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5608317684472344,
+      "learning_rate": 1.920524113383967e-05,
+      "loss": 0.9125,
+      "step": 956
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3108985978936933,
+      "learning_rate": 1.9203200711605297e-05,
+      "loss": 0.9269,
+      "step": 957
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3625666554468419,
+      "learning_rate": 1.9201157782165366e-05,
+      "loss": 0.9247,
+      "step": 958
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7816393639999777,
+      "learning_rate": 1.9199112346076422e-05,
+      "loss": 0.9125,
+      "step": 959
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6366891972097488,
+      "learning_rate": 1.91970644038957e-05,
+      "loss": 0.894,
+      "step": 960
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3552353751487338,
+      "learning_rate": 1.9195013956181115e-05,
+      "loss": 0.8064,
+      "step": 961
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3206030061198555,
+      "learning_rate": 1.919296100349127e-05,
+      "loss": 0.9903,
+      "step": 962
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8241746844048248,
+      "learning_rate": 1.919090554638544e-05,
+      "loss": 0.9028,
+      "step": 963
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5474018153517208,
+      "learning_rate": 1.9188847585423593e-05,
+      "loss": 1.0327,
+      "step": 964
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6163799059266604,
+      "learning_rate": 1.9186787121166367e-05,
+      "loss": 0.9783,
+      "step": 965
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3994849666886506,
+      "learning_rate": 1.9184724154175096e-05,
+      "loss": 0.9938,
+      "step": 966
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5046244695965882,
+      "learning_rate": 1.9182658685011787e-05,
+      "loss": 0.9087,
+      "step": 967
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5478609211154413,
+      "learning_rate": 1.918059071423913e-05,
+      "loss": 0.8732,
+      "step": 968
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4488964814973997,
+      "learning_rate": 1.9178520242420498e-05,
+      "loss": 1.0123,
+      "step": 969
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6020275813047413,
+      "learning_rate": 1.9176447270119945e-05,
+      "loss": 0.9082,
+      "step": 970
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7740844920167578,
+      "learning_rate": 1.917437179790221e-05,
+      "loss": 0.9443,
+      "step": 971
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4097687692149998,
+      "learning_rate": 1.91722938263327e-05,
+      "loss": 0.9326,
+      "step": 972
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3934202842818284,
+      "learning_rate": 1.9170213355977513e-05,
+      "loss": 0.8556,
+      "step": 973
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.512458460730969,
+      "learning_rate": 1.9168130387403433e-05,
+      "loss": 0.8828,
+      "step": 974
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4849542071255044,
+      "learning_rate": 1.916604492117791e-05,
+      "loss": 1.0031,
+      "step": 975
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4000708016625028,
+      "learning_rate": 1.9163956957869093e-05,
+      "loss": 0.881,
+      "step": 976
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.544050152794303,
+      "learning_rate": 1.9161866498045792e-05,
+      "loss": 0.9371,
+      "step": 977
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4302403444230285,
+      "learning_rate": 1.9159773542277508e-05,
+      "loss": 0.9499,
+      "step": 978
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5398397709811054,
+      "learning_rate": 1.9157678091134424e-05,
+      "loss": 0.9661,
+      "step": 979
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3243422520978139,
+      "learning_rate": 1.9155580145187392e-05,
+      "loss": 0.87,
+      "step": 980
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5311628737769714,
+      "learning_rate": 1.9153479705007953e-05,
+      "loss": 0.952,
+      "step": 981
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.714219641784785,
+      "learning_rate": 1.9151376771168325e-05,
+      "loss": 0.9719,
+      "step": 982
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4822879185420312,
+      "learning_rate": 1.914927134424141e-05,
+      "loss": 0.8965,
+      "step": 983
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4169797535304103,
+      "learning_rate": 1.9147163424800773e-05,
+      "loss": 0.9007,
+      "step": 984
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5664067851039514,
+      "learning_rate": 1.914505301342068e-05,
+      "loss": 0.9321,
+      "step": 985
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7807317343703846,
+      "learning_rate": 1.914294011067606e-05,
+      "loss": 0.9584,
+      "step": 986
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4825237010517485,
+      "learning_rate": 1.914082471714253e-05,
+      "loss": 0.9751,
+      "step": 987
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5146127631427522,
+      "learning_rate": 1.9138706833396374e-05,
+      "loss": 1.0071,
+      "step": 988
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5015119805729173,
+      "learning_rate": 1.913658646001457e-05,
+      "loss": 0.9479,
+      "step": 989
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.262381761804665,
+      "learning_rate": 1.913446359757476e-05,
+      "loss": 0.8898,
+      "step": 990
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3647375669537145,
+      "learning_rate": 1.9132338246655276e-05,
+      "loss": 0.9613,
+      "step": 991
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5864504468508505,
+      "learning_rate": 1.9130210407835118e-05,
+      "loss": 0.9821,
+      "step": 992
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4646656265043707,
+      "learning_rate": 1.912808008169397e-05,
+      "loss": 0.9183,
+      "step": 993
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3796063947437502,
+      "learning_rate": 1.9125947268812192e-05,
+      "loss": 0.8932,
+      "step": 994
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1961649585110579,
+      "learning_rate": 1.9123811969770815e-05,
+      "loss": 0.8055,
+      "step": 995
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5876943102089769,
+      "learning_rate": 1.9121674185151562e-05,
+      "loss": 0.9134,
+      "step": 996
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8498307290511353,
+      "learning_rate": 1.911953391553682e-05,
+      "loss": 0.3122,
+      "step": 997
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6238913915356858,
+      "learning_rate": 1.9117391161509657e-05,
+      "loss": 1.0385,
+      "step": 998
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4142241464608212,
+      "learning_rate": 1.911524592365382e-05,
+      "loss": 0.9436,
+      "step": 999
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.434708617822933,
+      "learning_rate": 1.911309820255373e-05,
+      "loss": 0.931,
+      "step": 1000
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7810939842839393,
+      "learning_rate": 1.9110947998794484e-05,
+      "loss": 0.9253,
+      "step": 1001
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.385218040639362,
+      "learning_rate": 1.910879531296186e-05,
+      "loss": 1.0058,
+      "step": 1002
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.508366568832681,
+      "learning_rate": 1.9106640145642308e-05,
+      "loss": 1.0139,
+      "step": 1003
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6345031087695705,
+      "learning_rate": 1.9104482497422955e-05,
+      "loss": 0.9531,
+      "step": 1004
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.599031381201754,
+      "learning_rate": 1.9102322368891602e-05,
+      "loss": 0.9309,
+      "step": 1005
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4184120177646085,
+      "learning_rate": 1.9100159760636727e-05,
+      "loss": 0.9132,
+      "step": 1006
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9947098543152184,
+      "learning_rate": 1.9097994673247488e-05,
+      "loss": 0.3257,
+      "step": 1007
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4528170629157062,
+      "learning_rate": 1.9095827107313713e-05,
+      "loss": 0.9402,
+      "step": 1008
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7316713373862707,
+      "learning_rate": 1.90936570634259e-05,
+      "loss": 0.9711,
+      "step": 1009
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6357637546313155,
+      "learning_rate": 1.9091484542175244e-05,
+      "loss": 1.0003,
+      "step": 1010
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4904615877707033,
+      "learning_rate": 1.908930954415358e-05,
+      "loss": 0.9172,
+      "step": 1011
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4452231598640641,
+      "learning_rate": 1.9087132069953456e-05,
+      "loss": 0.8875,
+      "step": 1012
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5082600984183923,
+      "learning_rate": 1.908495212016807e-05,
+      "loss": 0.9354,
+      "step": 1013
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.707869940471913,
+      "learning_rate": 1.908276969539129e-05,
+      "loss": 0.968,
+      "step": 1014
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4684966343108745,
+      "learning_rate": 1.9080584796217683e-05,
+      "loss": 0.9227,
+      "step": 1015
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7016849778717056,
+      "learning_rate": 1.9078397423242467e-05,
+      "loss": 0.9422,
+      "step": 1016
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3604224817065067,
+      "learning_rate": 1.9076207577061546e-05,
+      "loss": 0.9189,
+      "step": 1017
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.469361097578541,
+      "learning_rate": 1.907401525827149e-05,
+      "loss": 0.8269,
+      "step": 1018
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6235494558522054,
+      "learning_rate": 1.907182046746956e-05,
+      "loss": 0.9177,
+      "step": 1019
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7633523894576364,
+      "learning_rate": 1.906962320525366e-05,
+      "loss": 0.909,
+      "step": 1020
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5536554711473778,
+      "learning_rate": 1.90674234722224e-05,
+      "loss": 1.0051,
+      "step": 1021
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8125168598147263,
+      "learning_rate": 1.9065221268975037e-05,
+      "loss": 1.0399,
+      "step": 1022
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4833565304871992,
+      "learning_rate": 1.9063016596111516e-05,
+      "loss": 1.0658,
+      "step": 1023
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8551963631314097,
+      "learning_rate": 1.906080945423245e-05,
+      "loss": 1.0431,
+      "step": 1024
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9092081426066301,
+      "learning_rate": 1.9058599843939127e-05,
+      "loss": 0.3583,
+      "step": 1025
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5057059882795931,
+      "learning_rate": 1.9056387765833506e-05,
+      "loss": 0.9491,
+      "step": 1026
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4170223373528612,
+      "learning_rate": 1.905417322051822e-05,
+      "loss": 0.9248,
+      "step": 1027
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8438436951106543,
+      "learning_rate": 1.9051956208596564e-05,
+      "loss": 0.996,
+      "step": 1028
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2880478738362242,
+      "learning_rate": 1.9049736730672518e-05,
+      "loss": 0.9916,
+      "step": 1029
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4935558358995293,
+      "learning_rate": 1.904751478735073e-05,
+      "loss": 0.9485,
+      "step": 1030
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.432688723639302,
+      "learning_rate": 1.9045290379236517e-05,
+      "loss": 0.8824,
+      "step": 1031
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.536095838835383,
+      "learning_rate": 1.904306350693587e-05,
+      "loss": 0.9005,
+      "step": 1032
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4662415218860207,
+      "learning_rate": 1.904083417105545e-05,
+      "loss": 0.9713,
+      "step": 1033
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6070501435355622,
+      "learning_rate": 1.9038602372202594e-05,
+      "loss": 0.9201,
+      "step": 1034
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7527244777157707,
+      "learning_rate": 1.9036368110985296e-05,
+      "loss": 1.0307,
+      "step": 1035
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4700227711415736,
+      "learning_rate": 1.9034131388012237e-05,
+      "loss": 0.9484,
+      "step": 1036
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6424095235107115,
+      "learning_rate": 1.903189220389276e-05,
+      "loss": 0.9261,
+      "step": 1037
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7995022579233025,
+      "learning_rate": 1.9029650559236886e-05,
+      "loss": 0.3572,
+      "step": 1038
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5408550772297291,
+      "learning_rate": 1.9027406454655292e-05,
+      "loss": 0.9673,
+      "step": 1039
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5028551652067113,
+      "learning_rate": 1.9025159890759336e-05,
+      "loss": 0.9257,
+      "step": 1040
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.84265488393321,
+      "learning_rate": 1.902291086816105e-05,
+      "loss": 0.9747,
+      "step": 1041
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.697734685091799,
+      "learning_rate": 1.902065938747312e-05,
+      "loss": 0.9679,
+      "step": 1042
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3202983150764154,
+      "learning_rate": 1.901840544930892e-05,
+      "loss": 0.9037,
+      "step": 1043
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.331114765876868,
+      "learning_rate": 1.9016149054282486e-05,
+      "loss": 0.9499,
+      "step": 1044
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3967544667940832,
+      "learning_rate": 1.9013890203008512e-05,
+      "loss": 0.943,
+      "step": 1045
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2908277576893867,
+      "learning_rate": 1.901162889610238e-05,
+      "loss": 0.9728,
+      "step": 1046
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8109643458551492,
+      "learning_rate": 1.9009365134180128e-05,
+      "loss": 0.327,
+      "step": 1047
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.36900598908847,
+      "learning_rate": 1.9007098917858472e-05,
+      "loss": 0.9419,
+      "step": 1048
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5669325737452564,
+      "learning_rate": 1.9004830247754786e-05,
+      "loss": 0.9227,
+      "step": 1049
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7750426330114113,
+      "learning_rate": 1.9002559124487122e-05,
+      "loss": 1.0072,
+      "step": 1050
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1821655356787864,
+      "learning_rate": 1.9000285548674195e-05,
+      "loss": 0.9281,
+      "step": 1051
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6354043224475265,
+      "learning_rate": 1.8998009520935388e-05,
+      "loss": 0.8816,
+      "step": 1052
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5272528128613672,
+      "learning_rate": 1.899573104189076e-05,
+      "loss": 0.9118,
+      "step": 1053
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3850719391551656,
+      "learning_rate": 1.8993450112161023e-05,
+      "loss": 0.9142,
+      "step": 1054
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4922868585711018,
+      "learning_rate": 1.8991166732367575e-05,
+      "loss": 0.9923,
+      "step": 1055
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8381640173157345,
+      "learning_rate": 1.8988880903132464e-05,
+      "loss": 1.0449,
+      "step": 1056
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3052145866193885,
+      "learning_rate": 1.8986592625078413e-05,
+      "loss": 0.9661,
+      "step": 1057
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3947873200998628,
+      "learning_rate": 1.8984301898828815e-05,
+      "loss": 0.9036,
+      "step": 1058
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4275290221491295,
+      "learning_rate": 1.8982008725007725e-05,
+      "loss": 0.9159,
+      "step": 1059
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5488225845391752,
+      "learning_rate": 1.897971310423987e-05,
+      "loss": 0.9511,
+      "step": 1060
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5371142669658056,
+      "learning_rate": 1.8977415037150638e-05,
+      "loss": 0.9492,
+      "step": 1061
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5189254860002805,
+      "learning_rate": 1.8975114524366086e-05,
+      "loss": 0.9012,
+      "step": 1062
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6247801053067341,
+      "learning_rate": 1.8972811566512934e-05,
+      "loss": 0.9495,
+      "step": 1063
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3726842816594798,
+      "learning_rate": 1.8970506164218578e-05,
+      "loss": 0.9009,
+      "step": 1064
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6496971151751343,
+      "learning_rate": 1.8968198318111067e-05,
+      "loss": 0.9547,
+      "step": 1065
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3988349778562228,
+      "learning_rate": 1.8965888028819125e-05,
+      "loss": 0.9177,
+      "step": 1066
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3479779509105108,
+      "learning_rate": 1.8963575296972137e-05,
+      "loss": 0.9301,
+      "step": 1067
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3679008741276417,
+      "learning_rate": 1.8961260123200158e-05,
+      "loss": 0.9277,
+      "step": 1068
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6981100641923372,
+      "learning_rate": 1.89589425081339e-05,
+      "loss": 1.0307,
+      "step": 1069
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3960011012817106,
+      "learning_rate": 1.895662245240475e-05,
+      "loss": 0.9262,
+      "step": 1070
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4901549104374738,
+      "learning_rate": 1.8954299956644757e-05,
+      "loss": 0.9436,
+      "step": 1071
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6108017616592303,
+      "learning_rate": 1.8951975021486623e-05,
+      "loss": 0.9871,
+      "step": 1072
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4882282932251385,
+      "learning_rate": 1.8949647647563733e-05,
+      "loss": 1.0219,
+      "step": 1073
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3748520648280866,
+      "learning_rate": 1.8947317835510125e-05,
+      "loss": 0.9033,
+      "step": 1074
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3877001909281639,
+      "learning_rate": 1.8944985585960504e-05,
+      "loss": 0.9694,
+      "step": 1075
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5869913628164523,
+      "learning_rate": 1.8942650899550237e-05,
+      "loss": 0.8942,
+      "step": 1076
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8213825145605622,
+      "learning_rate": 1.8940313776915365e-05,
+      "loss": 0.9357,
+      "step": 1077
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3631246519186686,
+      "learning_rate": 1.8937974218692577e-05,
+      "loss": 0.946,
+      "step": 1078
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4662945169032926,
+      "learning_rate": 1.8935632225519232e-05,
+      "loss": 0.8852,
+      "step": 1079
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7868894178492483,
+      "learning_rate": 1.893328779803336e-05,
+      "loss": 0.886,
+      "step": 1080
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6125883006622244,
+      "learning_rate": 1.8930940936873644e-05,
+      "loss": 0.9522,
+      "step": 1081
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4192700614985427,
+      "learning_rate": 1.892859164267943e-05,
+      "loss": 0.962,
+      "step": 1082
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4507946436057837,
+      "learning_rate": 1.892623991609074e-05,
+      "loss": 0.8659,
+      "step": 1083
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.452635378796897,
+      "learning_rate": 1.892388575774824e-05,
+      "loss": 0.9195,
+      "step": 1084
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.490699023182195,
+      "learning_rate": 1.892152916829327e-05,
+      "loss": 0.9209,
+      "step": 1085
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.31694368787653,
+      "learning_rate": 1.8919170148367837e-05,
+      "loss": 0.9767,
+      "step": 1086
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5114321942263294,
+      "learning_rate": 1.891680869861459e-05,
+      "loss": 0.9192,
+      "step": 1087
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8074507190035066,
+      "learning_rate": 1.8914444819676866e-05,
+      "loss": 0.322,
+      "step": 1088
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6575995763331053,
+      "learning_rate": 1.8912078512198642e-05,
+      "loss": 0.9489,
+      "step": 1089
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3927431590611632,
+      "learning_rate": 1.890970977682457e-05,
+      "loss": 0.9797,
+      "step": 1090
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.200757713558766,
+      "learning_rate": 1.8907338614199956e-05,
+      "loss": 0.898,
+      "step": 1091
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.234016828994756,
+      "learning_rate": 1.8904965024970772e-05,
+      "loss": 0.9421,
+      "step": 1092
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6698401157821126,
+      "learning_rate": 1.8902589009783648e-05,
+      "loss": 0.9541,
+      "step": 1093
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3479106849480442,
+      "learning_rate": 1.8900210569285877e-05,
+      "loss": 0.9344,
+      "step": 1094
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7299262482782509,
+      "learning_rate": 1.889782970412541e-05,
+      "loss": 0.3506,
+      "step": 1095
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.562064848148622,
+      "learning_rate": 1.8895446414950864e-05,
+      "loss": 0.9812,
+      "step": 1096
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5229708843396277,
+      "learning_rate": 1.889306070241151e-05,
+      "loss": 0.9483,
+      "step": 1097
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.285892721951612,
+      "learning_rate": 1.8890672567157278e-05,
+      "loss": 0.9576,
+      "step": 1098
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4881002836766306,
+      "learning_rate": 1.888828200983877e-05,
+      "loss": 0.9378,
+      "step": 1099
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6490733820784063,
+      "learning_rate": 1.8885889031107232e-05,
+      "loss": 0.9404,
+      "step": 1100
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5013302620219808,
+      "learning_rate": 1.8883493631614583e-05,
+      "loss": 1.0075,
+      "step": 1101
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3918298825132833,
+      "learning_rate": 1.8881095812013394e-05,
+      "loss": 0.9868,
+      "step": 1102
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5932968584698821,
+      "learning_rate": 1.8878695572956896e-05,
+      "loss": 1.0453,
+      "step": 1103
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.53955576073458,
+      "learning_rate": 1.8876292915098982e-05,
+      "loss": 0.9656,
+      "step": 1104
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3119470894855925,
+      "learning_rate": 1.8873887839094202e-05,
+      "loss": 0.9613,
+      "step": 1105
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6138746611131145,
+      "learning_rate": 1.8871480345597763e-05,
+      "loss": 0.9622,
+      "step": 1106
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4441059876080875,
+      "learning_rate": 1.8869070435265535e-05,
+      "loss": 0.9152,
+      "step": 1107
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5065047857601885,
+      "learning_rate": 1.8866658108754045e-05,
+      "loss": 0.9526,
+      "step": 1108
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5239810994676692,
+      "learning_rate": 1.886424336672047e-05,
+      "loss": 0.9679,
+      "step": 1109
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5806650172817183,
+      "learning_rate": 1.8861826209822662e-05,
+      "loss": 0.9525,
+      "step": 1110
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3122113079226483,
+      "learning_rate": 1.8859406638719117e-05,
+      "loss": 0.9059,
+      "step": 1111
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3703785356113727,
+      "learning_rate": 1.885698465406899e-05,
+      "loss": 0.8664,
+      "step": 1112
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5994870426690446,
+      "learning_rate": 1.8854560256532098e-05,
+      "loss": 0.9265,
+      "step": 1113
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3044443439277875,
+      "learning_rate": 1.885213344676892e-05,
+      "loss": 0.9818,
+      "step": 1114
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.529971318708502,
+      "learning_rate": 1.8849704225440578e-05,
+      "loss": 0.8763,
+      "step": 1115
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4893470818701977,
+      "learning_rate": 1.884727259320886e-05,
+      "loss": 0.9507,
+      "step": 1116
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.42765741786646,
+      "learning_rate": 1.8844838550736215e-05,
+      "loss": 0.9311,
+      "step": 1117
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6135612099403698,
+      "learning_rate": 1.8842402098685735e-05,
+      "loss": 0.9908,
+      "step": 1118
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2903163360701284,
+      "learning_rate": 1.883996323772118e-05,
+      "loss": 0.9129,
+      "step": 1119
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.737358077431292,
+      "learning_rate": 1.883752196850697e-05,
+      "loss": 0.9827,
+      "step": 1120
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4722235091884055,
+      "learning_rate": 1.8835078291708164e-05,
+      "loss": 0.9511,
+      "step": 1121
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.714144716316613,
+      "learning_rate": 1.8832632207990493e-05,
+      "loss": 0.9142,
+      "step": 1122
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.644107492176041,
+      "learning_rate": 1.883018371802033e-05,
+      "loss": 0.9779,
+      "step": 1123
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4983664772258791,
+      "learning_rate": 1.8827732822464723e-05,
+      "loss": 0.9729,
+      "step": 1124
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.554168042305571,
+      "learning_rate": 1.882527952199136e-05,
+      "loss": 0.8862,
+      "step": 1125
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5179550541712927,
+      "learning_rate": 1.8822823817268577e-05,
+      "loss": 0.9066,
+      "step": 1126
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4758783734846341,
+      "learning_rate": 1.8820365708965385e-05,
+      "loss": 0.9422,
+      "step": 1127
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8641786711497633,
+      "learning_rate": 1.881790519775144e-05,
+      "loss": 0.3186,
+      "step": 1128
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5212905696695405,
+      "learning_rate": 1.8815442284297055e-05,
+      "loss": 0.9316,
+      "step": 1129
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2877008471260802,
+      "learning_rate": 1.881297696927319e-05,
+      "loss": 0.9422,
+      "step": 1130
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3505760607630661,
+      "learning_rate": 1.881050925335147e-05,
+      "loss": 1.0274,
+      "step": 1131
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3627921145090331,
+      "learning_rate": 1.8808039137204162e-05,
+      "loss": 0.9084,
+      "step": 1132
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2600794430873121,
+      "learning_rate": 1.8805566621504203e-05,
+      "loss": 0.9465,
+      "step": 1133
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6269075128543034,
+      "learning_rate": 1.8803091706925168e-05,
+      "loss": 0.9375,
+      "step": 1134
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6372073623197263,
+      "learning_rate": 1.8800614394141292e-05,
+      "loss": 0.9401,
+      "step": 1135
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5450912622652375,
+      "learning_rate": 1.8798134683827464e-05,
+      "loss": 0.9609,
+      "step": 1136
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.287283467398738,
+      "learning_rate": 1.879565257665923e-05,
+      "loss": 0.9315,
+      "step": 1137
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7992698432162705,
+      "learning_rate": 1.8793168073312776e-05,
+      "loss": 0.301,
+      "step": 1138
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.585304787051378,
+      "learning_rate": 1.8790681174464956e-05,
+      "loss": 0.933,
+      "step": 1139
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4715045648834797,
+      "learning_rate": 1.8788191880793264e-05,
+      "loss": 0.9508,
+      "step": 1140
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.63923169191026,
+      "learning_rate": 1.878570019297586e-05,
+      "loss": 0.8732,
+      "step": 1141
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5074090349896567,
+      "learning_rate": 1.8783206111691543e-05,
+      "loss": 1.0033,
+      "step": 1142
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4349974021832033,
+      "learning_rate": 1.878070963761977e-05,
+      "loss": 0.9647,
+      "step": 1143
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5126603998596777,
+      "learning_rate": 1.877821077144065e-05,
+      "loss": 0.8579,
+      "step": 1144
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.470026454555716,
+      "learning_rate": 1.8775709513834945e-05,
+      "loss": 0.9294,
+      "step": 1145
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3708458048630978,
+      "learning_rate": 1.8773205865484065e-05,
+      "loss": 0.9371,
+      "step": 1146
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4928609766205274,
+      "learning_rate": 1.8770699827070068e-05,
+      "loss": 0.9347,
+      "step": 1147
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.062066932208199,
+      "learning_rate": 1.8768191399275678e-05,
+      "loss": 1.0131,
+      "step": 1148
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8561238595795774,
+      "learning_rate": 1.876568058278425e-05,
+      "loss": 0.3384,
+      "step": 1149
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.301562715320867,
+      "learning_rate": 1.87631673782798e-05,
+      "loss": 0.955,
+      "step": 1150
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3476308675117146,
+      "learning_rate": 1.8760651786447006e-05,
+      "loss": 0.9473,
+      "step": 1151
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8243190040000618,
+      "learning_rate": 1.8758133807971175e-05,
+      "loss": 0.3618,
+      "step": 1152
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4619404497995054,
+      "learning_rate": 1.8755613443538275e-05,
+      "loss": 0.9,
+      "step": 1153
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5911175782046088,
+      "learning_rate": 1.875309069383492e-05,
+      "loss": 1.0168,
+      "step": 1154
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8607834101988006,
+      "learning_rate": 1.875056555954838e-05,
+      "loss": 0.9482,
+      "step": 1155
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.508935010202807,
+      "learning_rate": 1.8748038041366573e-05,
+      "loss": 0.9337,
+      "step": 1156
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5174446446153425,
+      "learning_rate": 1.8745508139978063e-05,
+      "loss": 0.9258,
+      "step": 1157
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4569585081458498,
+      "learning_rate": 1.8742975856072064e-05,
+      "loss": 0.9465,
+      "step": 1158
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2930779101284477,
+      "learning_rate": 1.874044119033844e-05,
+      "loss": 0.8834,
+      "step": 1159
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3076541901097145,
+      "learning_rate": 1.8737904143467703e-05,
+      "loss": 0.9218,
+      "step": 1160
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4780577759497269,
+      "learning_rate": 1.8735364716151017e-05,
+      "loss": 0.9822,
+      "step": 1161
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3810395348523756,
+      "learning_rate": 1.873282290908019e-05,
+      "loss": 0.9556,
+      "step": 1162
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3310767547670135,
+      "learning_rate": 1.873027872294768e-05,
+      "loss": 0.8971,
+      "step": 1163
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5574735760628649,
+      "learning_rate": 1.8727732158446598e-05,
+      "loss": 0.9192,
+      "step": 1164
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4796583585930962,
+      "learning_rate": 1.872518321627069e-05,
+      "loss": 1.0039,
+      "step": 1165
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4270175746979632,
+      "learning_rate": 1.8722631897114364e-05,
+      "loss": 0.9247,
+      "step": 1166
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5334476959247347,
+      "learning_rate": 1.872007820167267e-05,
+      "loss": 0.9151,
+      "step": 1167
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.355961266972235,
+      "learning_rate": 1.8717522130641305e-05,
+      "loss": 0.8912,
+      "step": 1168
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.415079623023431,
+      "learning_rate": 1.8714963684716605e-05,
+      "loss": 0.9768,
+      "step": 1169
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.346793709686554,
+      "learning_rate": 1.8712402864595575e-05,
+      "loss": 0.9084,
+      "step": 1170
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8871184574100671,
+      "learning_rate": 1.8709839670975843e-05,
+      "loss": 0.318,
+      "step": 1171
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3660321554956938,
+      "learning_rate": 1.8707274104555696e-05,
+      "loss": 0.9352,
+      "step": 1172
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6630718106841917,
+      "learning_rate": 1.8704706166034067e-05,
+      "loss": 0.9703,
+      "step": 1173
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4470234847244023,
+      "learning_rate": 1.870213585611053e-05,
+      "loss": 1.0173,
+      "step": 1174
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.601384974154196,
+      "learning_rate": 1.869956317548531e-05,
+      "loss": 0.8566,
+      "step": 1175
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5269438632914512,
+      "learning_rate": 1.8696988124859278e-05,
+      "loss": 0.9693,
+      "step": 1176
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4813745327098924,
+      "learning_rate": 1.8694410704933944e-05,
+      "loss": 0.9225,
+      "step": 1177
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.103600196574336,
+      "learning_rate": 1.8691830916411473e-05,
+      "loss": 0.9414,
+      "step": 1178
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.414423166825318,
+      "learning_rate": 1.868924875999467e-05,
+      "loss": 0.8825,
+      "step": 1179
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5593897966454986,
+      "learning_rate": 1.8686664236386984e-05,
+      "loss": 0.9553,
+      "step": 1180
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4682804343702456,
+      "learning_rate": 1.8684077346292507e-05,
+      "loss": 0.9155,
+      "step": 1181
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3887671848376817,
+      "learning_rate": 1.8681488090415984e-05,
+      "loss": 0.946,
+      "step": 1182
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3184388157963884,
+      "learning_rate": 1.86788964694628e-05,
+      "loss": 0.853,
+      "step": 1183
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8966923215655658,
+      "learning_rate": 1.8676302484138982e-05,
+      "loss": 0.957,
+      "step": 1184
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2800339679243535,
+      "learning_rate": 1.8673706135151205e-05,
+      "loss": 0.8837,
+      "step": 1185
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.421762006063665,
+      "learning_rate": 1.8671107423206785e-05,
+      "loss": 0.9173,
+      "step": 1186
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5622682411453128,
+      "learning_rate": 1.866850634901368e-05,
+      "loss": 0.9332,
+      "step": 1187
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6950343701530162,
+      "learning_rate": 1.8665902913280498e-05,
+      "loss": 0.9734,
+      "step": 1188
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4402419255614587,
+      "learning_rate": 1.8663297116716488e-05,
+      "loss": 0.9079,
+      "step": 1189
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.837156569521358,
+      "learning_rate": 1.8660688960031533e-05,
+      "loss": 0.8776,
+      "step": 1190
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.614721114037799,
+      "learning_rate": 1.8658078443936175e-05,
+      "loss": 0.9994,
+      "step": 1191
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9409368203318107,
+      "learning_rate": 1.865546556914159e-05,
+      "loss": 0.3557,
+      "step": 1192
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3149849926580364,
+      "learning_rate": 1.865285033635959e-05,
+      "loss": 0.8661,
+      "step": 1193
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3275920780445138,
+      "learning_rate": 1.8650232746302645e-05,
+      "loss": 0.9852,
+      "step": 1194
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.503481201695558,
+      "learning_rate": 1.8647612799683853e-05,
+      "loss": 0.9394,
+      "step": 1195
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4441586334678203,
+      "learning_rate": 1.864499049721696e-05,
+      "loss": 0.9452,
+      "step": 1196
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6884608270762453,
+      "learning_rate": 1.8642365839616358e-05,
+      "loss": 0.917,
+      "step": 1197
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.524430090986756,
+      "learning_rate": 1.863973882759707e-05,
+      "loss": 0.9422,
+      "step": 1198
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.23203692617419,
+      "learning_rate": 1.863710946187477e-05,
+      "loss": 0.9752,
+      "step": 1199
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9073838691946678,
+      "learning_rate": 1.8634477743165772e-05,
+      "loss": 0.3052,
+      "step": 1200
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3363113751570177,
+      "learning_rate": 1.863184367218702e-05,
+      "loss": 0.8275,
+      "step": 1201
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.667548573947656,
+      "learning_rate": 1.862920724965612e-05,
+      "loss": 0.9256,
+      "step": 1202
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4030477472137224,
+      "learning_rate": 1.8626568476291296e-05,
+      "loss": 0.9314,
+      "step": 1203
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.549424248544921,
+      "learning_rate": 1.8623927352811432e-05,
+      "loss": 0.974,
+      "step": 1204
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3990736678480602,
+      "learning_rate": 1.862128387993603e-05,
+      "loss": 0.9349,
+      "step": 1205
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2637281899086543,
+      "learning_rate": 1.8618638058385255e-05,
+      "loss": 0.8247,
+      "step": 1206
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4868923305675628,
+      "learning_rate": 1.8615989888879898e-05,
+      "loss": 0.9991,
+      "step": 1207
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5020154901610467,
+      "learning_rate": 1.8613339372141395e-05,
+      "loss": 0.9165,
+      "step": 1208
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3037904800837983,
+      "learning_rate": 1.861068650889182e-05,
+      "loss": 0.9479,
+      "step": 1209
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8578471956940509,
+      "learning_rate": 1.8608031299853882e-05,
+      "loss": 0.3369,
+      "step": 1210
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4467898306177405,
+      "learning_rate": 1.860537374575094e-05,
+      "loss": 0.959,
+      "step": 1211
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5528269788130764,
+      "learning_rate": 1.860271384730698e-05,
+      "loss": 0.9586,
+      "step": 1212
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.62331017333585,
+      "learning_rate": 1.860005160524663e-05,
+      "loss": 0.9701,
+      "step": 1213
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4502500889769396,
+      "learning_rate": 1.859738702029516e-05,
+      "loss": 0.8916,
+      "step": 1214
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8434507277450702,
+      "learning_rate": 1.859472009317848e-05,
+      "loss": 0.3309,
+      "step": 1215
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4457425366966303,
+      "learning_rate": 1.8592050824623133e-05,
+      "loss": 0.936,
+      "step": 1216
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5009382195871088,
+      "learning_rate": 1.8589379215356302e-05,
+      "loss": 0.9586,
+      "step": 1217
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.788095185668963,
+      "learning_rate": 1.8586705266105802e-05,
+      "loss": 0.9034,
+      "step": 1218
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4663373345382535,
+      "learning_rate": 1.858402897760009e-05,
+      "loss": 0.9832,
+      "step": 1219
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0309724199353967,
+      "learning_rate": 1.858135035056827e-05,
+      "loss": 0.3465,
+      "step": 1220
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4003839865065622,
+      "learning_rate": 1.8578669385740065e-05,
+      "loss": 0.9917,
+      "step": 1221
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.666548967241235,
+      "learning_rate": 1.8575986083845848e-05,
+      "loss": 0.9907,
+      "step": 1222
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3283679585661043,
+      "learning_rate": 1.8573300445616628e-05,
+      "loss": 0.9881,
+      "step": 1223
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6300443702930745,
+      "learning_rate": 1.8570612471784034e-05,
+      "loss": 0.945,
+      "step": 1224
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5459008424683574,
+      "learning_rate": 1.8567922163080357e-05,
+      "loss": 0.9585,
+      "step": 1225
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5981840343820413,
+      "learning_rate": 1.8565229520238504e-05,
+      "loss": 0.9809,
+      "step": 1226
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4133003246228455,
+      "learning_rate": 1.856253454399203e-05,
+      "loss": 1.0399,
+      "step": 1227
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.439196440793315,
+      "learning_rate": 1.855983723507512e-05,
+      "loss": 0.9188,
+      "step": 1228
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7505249285619295,
+      "learning_rate": 1.8557137594222588e-05,
+      "loss": 0.9124,
+      "step": 1229
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5080343821931554,
+      "learning_rate": 1.8554435622169903e-05,
+      "loss": 0.8637,
+      "step": 1230
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5053126787500748,
+      "learning_rate": 1.8551731319653147e-05,
+      "loss": 0.9365,
+      "step": 1231
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3074386455559,
+      "learning_rate": 1.854902468740905e-05,
+      "loss": 0.8649,
+      "step": 1232
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8538496522426393,
+      "learning_rate": 1.8546315726174973e-05,
+      "loss": 0.9378,
+      "step": 1233
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.422759533552292,
+      "learning_rate": 1.854360443668891e-05,
+      "loss": 0.9003,
+      "step": 1234
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.505211989699172,
+      "learning_rate": 1.8540890819689497e-05,
+      "loss": 0.9358,
+      "step": 1235
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5046103513652982,
+      "learning_rate": 1.8538174875915992e-05,
+      "loss": 0.8674,
+      "step": 1236
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4898623205786867,
+      "learning_rate": 1.853545660610829e-05,
+      "loss": 0.9721,
+      "step": 1237
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5865120340360355,
+      "learning_rate": 1.853273601100693e-05,
+      "loss": 0.9269,
+      "step": 1238
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2998279337305358,
+      "learning_rate": 1.8530013091353075e-05,
+      "loss": 0.8891,
+      "step": 1239
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4829570924869244,
+      "learning_rate": 1.8527287847888524e-05,
+      "loss": 0.9925,
+      "step": 1240
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3359230267984405,
+      "learning_rate": 1.8524560281355705e-05,
+      "loss": 0.9579,
+      "step": 1241
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5641327362029283,
+      "learning_rate": 1.8521830392497685e-05,
+      "loss": 0.8456,
+      "step": 1242
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8720659777030559,
+      "learning_rate": 1.8519098182058163e-05,
+      "loss": 0.342,
+      "step": 1243
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3945502943052863,
+      "learning_rate": 1.8516363650781464e-05,
+      "loss": 0.9569,
+      "step": 1244
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5314563644398622,
+      "learning_rate": 1.851362679941255e-05,
+      "loss": 0.9513,
+      "step": 1245
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5244693914070613,
+      "learning_rate": 1.851088762869702e-05,
+      "loss": 0.9881,
+      "step": 1246
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4379823098439193,
+      "learning_rate": 1.850814613938109e-05,
+      "loss": 0.9068,
+      "step": 1247
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7768918065409423,
+      "learning_rate": 1.850540233221163e-05,
+      "loss": 0.9094,
+      "step": 1248
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.486696633712154,
+      "learning_rate": 1.8502656207936117e-05,
+      "loss": 0.9771,
+      "step": 1249
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6675580912852332,
+      "learning_rate": 1.849990776730268e-05,
+      "loss": 0.8733,
+      "step": 1250
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4415213298603884,
+      "learning_rate": 1.849715701106006e-05,
+      "loss": 0.9142,
+      "step": 1251
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5622978049821856,
+      "learning_rate": 1.849440393995765e-05,
+      "loss": 0.9651,
+      "step": 1252
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6258274582781227,
+      "learning_rate": 1.8491648554745457e-05,
+      "loss": 0.9355,
+      "step": 1253
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4049389890520296,
+      "learning_rate": 1.848889085617412e-05,
+      "loss": 0.9626,
+      "step": 1254
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.399927875510835,
+      "learning_rate": 1.848613084499492e-05,
+      "loss": 0.9147,
+      "step": 1255
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3914621889022578,
+      "learning_rate": 1.8483368521959753e-05,
+      "loss": 0.9508,
+      "step": 1256
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3611343501056727,
+      "learning_rate": 1.8480603887821157e-05,
+      "loss": 0.9264,
+      "step": 1257
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4065073539432384,
+      "learning_rate": 1.8477836943332295e-05,
+      "loss": 0.905,
+      "step": 1258
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8270361502400705,
+      "learning_rate": 1.8475067689246954e-05,
+      "loss": 0.9203,
+      "step": 1259
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2900585054698248,
+      "learning_rate": 1.8472296126319557e-05,
+      "loss": 0.9391,
+      "step": 1260
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7535506293630616,
+      "learning_rate": 1.8469522255305156e-05,
+      "loss": 0.858,
+      "step": 1261
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4980231308702827,
+      "learning_rate": 1.846674607695943e-05,
+      "loss": 0.9307,
+      "step": 1262
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5906151925582785,
+      "learning_rate": 1.8463967592038684e-05,
+      "loss": 0.9537,
+      "step": 1263
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4208189938808407,
+      "learning_rate": 1.8461186801299857e-05,
+      "loss": 0.9038,
+      "step": 1264
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2556586972880175,
+      "learning_rate": 1.845840370550051e-05,
+      "loss": 0.9429,
+      "step": 1265
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.328714450291548,
+      "learning_rate": 1.8455618305398836e-05,
+      "loss": 0.9135,
+      "step": 1266
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.462962714126474,
+      "learning_rate": 1.8452830601753655e-05,
+      "loss": 1.0109,
+      "step": 1267
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.341869331469154,
+      "learning_rate": 1.8450040595324416e-05,
+      "loss": 0.992,
+      "step": 1268
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9963080811381916,
+      "learning_rate": 1.844724828687119e-05,
+      "loss": 0.3271,
+      "step": 1269
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6746499326513777,
+      "learning_rate": 1.8444453677154683e-05,
+      "loss": 0.8863,
+      "step": 1270
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5142722834084215,
+      "learning_rate": 1.844165676693622e-05,
+      "loss": 1.0185,
+      "step": 1271
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5738978614988803,
+      "learning_rate": 1.843885755697776e-05,
+      "loss": 0.9318,
+      "step": 1272
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3873979859341738,
+      "learning_rate": 1.843605604804188e-05,
+      "loss": 0.9643,
+      "step": 1273
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.313944898302993,
+      "learning_rate": 1.8433252240891793e-05,
+      "loss": 0.8947,
+      "step": 1274
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4738669964597948,
+      "learning_rate": 1.8430446136291333e-05,
+      "loss": 0.9366,
+      "step": 1275
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6835590107823177,
+      "learning_rate": 1.8427637735004957e-05,
+      "loss": 0.9547,
+      "step": 1276
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.295469361069176,
+      "learning_rate": 1.8424827037797755e-05,
+      "loss": 0.8369,
+      "step": 1277
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5508988789836728,
+      "learning_rate": 1.8422014045435433e-05,
+      "loss": 0.8412,
+      "step": 1278
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7754959084670252,
+      "learning_rate": 1.841919875868433e-05,
+      "loss": 0.9368,
+      "step": 1279
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5582288898203667,
+      "learning_rate": 1.841638117831141e-05,
+      "loss": 1.0028,
+      "step": 1280
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5337265354324514,
+      "learning_rate": 1.8413561305084262e-05,
+      "loss": 0.9136,
+      "step": 1281
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6309186878954023,
+      "learning_rate": 1.841073913977109e-05,
+      "loss": 0.9371,
+      "step": 1282
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4181033272892356,
+      "learning_rate": 1.8407914683140734e-05,
+      "loss": 0.884,
+      "step": 1283
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9395665092408731,
+      "learning_rate": 1.840508793596265e-05,
+      "loss": 0.3704,
+      "step": 1284
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5545111321326446,
+      "learning_rate": 1.8402258899006926e-05,
+      "loss": 1.0153,
+      "step": 1285
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3833091471668526,
+      "learning_rate": 1.839942757304427e-05,
+      "loss": 0.8784,
+      "step": 1286
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5223349775433335,
+      "learning_rate": 1.8396593958846013e-05,
+      "loss": 0.9926,
+      "step": 1287
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7004279027411002,
+      "learning_rate": 1.8393758057184104e-05,
+      "loss": 0.8789,
+      "step": 1288
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.670071427302,
+      "learning_rate": 1.8390919868831126e-05,
+      "loss": 0.9313,
+      "step": 1289
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.407897248691519,
+      "learning_rate": 1.838807939456028e-05,
+      "loss": 0.9167,
+      "step": 1290
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6662236125223822,
+      "learning_rate": 1.8385236635145385e-05,
+      "loss": 0.8904,
+      "step": 1291
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.808553923874879,
+      "learning_rate": 1.8382391591360895e-05,
+      "loss": 0.8968,
+      "step": 1292
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6574027968636342,
+      "learning_rate": 1.837954426398187e-05,
+      "loss": 0.936,
+      "step": 1293
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4557466142668847,
+      "learning_rate": 1.8376694653784003e-05,
+      "loss": 0.9361,
+      "step": 1294
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.287681993450042,
+      "learning_rate": 1.837384276154361e-05,
+      "loss": 0.9523,
+      "step": 1295
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8446314445396148,
+      "learning_rate": 1.8370988588037622e-05,
+      "loss": 0.9189,
+      "step": 1296
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4881626817826161,
+      "learning_rate": 1.8368132134043596e-05,
+      "loss": 0.9017,
+      "step": 1297
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.357751210830098,
+      "learning_rate": 1.836527340033971e-05,
+      "loss": 0.9336,
+      "step": 1298
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8515179790880286,
+      "learning_rate": 1.8362412387704752e-05,
+      "loss": 0.3158,
+      "step": 1299
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.238947413885612,
+      "learning_rate": 1.8359549096918154e-05,
+      "loss": 0.8944,
+      "step": 1300
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4450494488192385,
+      "learning_rate": 1.8356683528759948e-05,
+      "loss": 0.9333,
+      "step": 1301
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.568313467914447,
+      "learning_rate": 1.8353815684010796e-05,
+      "loss": 0.9468,
+      "step": 1302
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4015008304418182,
+      "learning_rate": 1.835094556345198e-05,
+      "loss": 0.8447,
+      "step": 1303
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.450318792332983,
+      "learning_rate": 1.834807316786539e-05,
+      "loss": 0.9758,
+      "step": 1304
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8496855208376316,
+      "learning_rate": 1.834519849803356e-05,
+      "loss": 0.3448,
+      "step": 1305
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.266695364697053,
+      "learning_rate": 1.834232155473962e-05,
+      "loss": 0.963,
+      "step": 1306
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.494139182644815,
+      "learning_rate": 1.833944233876733e-05,
+      "loss": 0.9205,
+      "step": 1307
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4495517673908238,
+      "learning_rate": 1.833656085090107e-05,
+      "loss": 0.9574,
+      "step": 1308
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.601748799909306,
+      "learning_rate": 1.8333677091925834e-05,
+      "loss": 0.9027,
+      "step": 1309
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.394548727241224,
+      "learning_rate": 1.8330791062627244e-05,
+      "loss": 0.9471,
+      "step": 1310
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3164122819054844,
+      "learning_rate": 1.8327902763791523e-05,
+      "loss": 0.9645,
+      "step": 1311
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3643497266778957,
+      "learning_rate": 1.832501219620553e-05,
+      "loss": 0.9542,
+      "step": 1312
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6410030389251806,
+      "learning_rate": 1.8322119360656736e-05,
+      "loss": 0.9107,
+      "step": 1313
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5315100858914268,
+      "learning_rate": 1.831922425793323e-05,
+      "loss": 0.8735,
+      "step": 1314
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.404916554956228,
+      "learning_rate": 1.8316326888823714e-05,
+      "loss": 0.9312,
+      "step": 1315
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2819264250225257,
+      "learning_rate": 1.8313427254117508e-05,
+      "loss": 0.8504,
+      "step": 1316
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4039782439841957,
+      "learning_rate": 1.8310525354604562e-05,
+      "loss": 0.9725,
+      "step": 1317
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5163038220050726,
+      "learning_rate": 1.8307621191075425e-05,
+      "loss": 0.9238,
+      "step": 1318
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5340439562182835,
+      "learning_rate": 1.8304714764321277e-05,
+      "loss": 0.9368,
+      "step": 1319
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7682117268905015,
+      "learning_rate": 1.8301806075133907e-05,
+      "loss": 0.313,
+      "step": 1320
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7105857610583735,
+      "learning_rate": 1.829889512430572e-05,
+      "loss": 0.9423,
+      "step": 1321
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3772880714963722,
+      "learning_rate": 1.829598191262974e-05,
+      "loss": 0.933,
+      "step": 1322
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4981942135446906,
+      "learning_rate": 1.829306644089961e-05,
+      "loss": 0.8442,
+      "step": 1323
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.1867719818237843,
+      "learning_rate": 1.829014870990958e-05,
+      "loss": 1.0171,
+      "step": 1324
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2114262763456343,
+      "learning_rate": 1.8287228720454522e-05,
+      "loss": 0.8926,
+      "step": 1325
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6088282219957097,
+      "learning_rate": 1.8284306473329922e-05,
+      "loss": 0.9661,
+      "step": 1326
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.426200734444402,
+      "learning_rate": 1.828138196933188e-05,
+      "loss": 0.8712,
+      "step": 1327
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9511094050778167,
+      "learning_rate": 1.8278455209257113e-05,
+      "loss": 0.3699,
+      "step": 1328
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.384390392094115,
+      "learning_rate": 1.8275526193902948e-05,
+      "loss": 0.9003,
+      "step": 1329
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3687000048610198,
+      "learning_rate": 1.8272594924067333e-05,
+      "loss": 0.9678,
+      "step": 1330
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6448160341735587,
+      "learning_rate": 1.8269661400548825e-05,
+      "loss": 0.9637,
+      "step": 1331
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.374549829283314,
+      "learning_rate": 1.8266725624146596e-05,
+      "loss": 0.8922,
+      "step": 1332
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3330331359805405,
+      "learning_rate": 1.8263787595660437e-05,
+      "loss": 0.9667,
+      "step": 1333
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.582131879455361,
+      "learning_rate": 1.8260847315890738e-05,
+      "loss": 0.8867,
+      "step": 1334
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5382919600425355,
+      "learning_rate": 1.8257904785638523e-05,
+      "loss": 0.8802,
+      "step": 1335
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4028260207321301,
+      "learning_rate": 1.8254960005705412e-05,
+      "loss": 0.9602,
+      "step": 1336
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5333344542997693,
+      "learning_rate": 1.825201297689365e-05,
+      "loss": 0.9527,
+      "step": 1337
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3300307632337933,
+      "learning_rate": 1.824906370000608e-05,
+      "loss": 0.9405,
+      "step": 1338
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5890640126327098,
+      "learning_rate": 1.8246112175846175e-05,
+      "loss": 0.9841,
+      "step": 1339
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4043446958461396,
+      "learning_rate": 1.8243158405218008e-05,
+      "loss": 0.9513,
+      "step": 1340
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7984676981980211,
+      "learning_rate": 1.824020238892626e-05,
+      "loss": 0.3612,
+      "step": 1341
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9291316161668783,
+      "learning_rate": 1.8237244127776247e-05,
+      "loss": 0.3711,
+      "step": 1342
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4498200775608228,
+      "learning_rate": 1.823428362257387e-05,
+      "loss": 0.9034,
+      "step": 1343
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5454830914803754,
+      "learning_rate": 1.8231320874125656e-05,
+      "loss": 0.9306,
+      "step": 1344
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5392866957701152,
+      "learning_rate": 1.8228355883238737e-05,
+      "loss": 0.9445,
+      "step": 1345
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4680372091249896,
+      "learning_rate": 1.822538865072086e-05,
+      "loss": 0.9346,
+      "step": 1346
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.445781957574018,
+      "learning_rate": 1.822241917738038e-05,
+      "loss": 0.9471,
+      "step": 1347
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9040202862023211,
+      "learning_rate": 1.8219447464026264e-05,
+      "loss": 0.3414,
+      "step": 1348
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.402399291131391,
+      "learning_rate": 1.821647351146809e-05,
+      "loss": 0.9158,
+      "step": 1349
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.452664364463535,
+      "learning_rate": 1.821349732051604e-05,
+      "loss": 0.9418,
+      "step": 1350
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4286752449943367,
+      "learning_rate": 1.8210518891980914e-05,
+      "loss": 0.9238,
+      "step": 1351
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4395874276519312,
+      "learning_rate": 1.820753822667412e-05,
+      "loss": 0.8863,
+      "step": 1352
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4336554191796156,
+      "learning_rate": 1.8204555325407668e-05,
+      "loss": 0.9091,
+      "step": 1353
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4948375622308352,
+      "learning_rate": 1.8201570188994186e-05,
+      "loss": 0.8974,
+      "step": 1354
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.429506926910762,
+      "learning_rate": 1.8198582818246908e-05,
+      "loss": 1.0089,
+      "step": 1355
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3747967290722465,
+      "learning_rate": 1.8195593213979676e-05,
+      "loss": 0.9066,
+      "step": 1356
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4261110348968973,
+      "learning_rate": 1.819260137700694e-05,
+      "loss": 0.937,
+      "step": 1357
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5329262586740509,
+      "learning_rate": 1.8189607308143754e-05,
+      "loss": 0.9663,
+      "step": 1358
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.929523001543622,
+      "learning_rate": 1.8186611008205794e-05,
+      "loss": 0.8981,
+      "step": 1359
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5368894237284176,
+      "learning_rate": 1.8183612478009328e-05,
+      "loss": 0.9369,
+      "step": 1360
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9326727181524269,
+      "learning_rate": 1.8180611718371238e-05,
+      "loss": 0.3465,
+      "step": 1361
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4762773774636195,
+      "learning_rate": 1.8177608730109023e-05,
+      "loss": 0.9596,
+      "step": 1362
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.371631321983861,
+      "learning_rate": 1.8174603514040767e-05,
+      "loss": 0.9181,
+      "step": 1363
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3013884012347359,
+      "learning_rate": 1.817159607098518e-05,
+      "loss": 0.9082,
+      "step": 1364
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6731856603192323,
+      "learning_rate": 1.8168586401761573e-05,
+      "loss": 0.9699,
+      "step": 1365
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.391615561030067,
+      "learning_rate": 1.8165574507189864e-05,
+      "loss": 0.8855,
+      "step": 1366
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3874372256395486,
+      "learning_rate": 1.8162560388090573e-05,
+      "loss": 0.8623,
+      "step": 1367
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5773904585193066,
+      "learning_rate": 1.8159544045284826e-05,
+      "loss": 1.0286,
+      "step": 1368
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2803037197567553,
+      "learning_rate": 1.8156525479594368e-05,
+      "loss": 0.8752,
+      "step": 1369
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3160383657747867,
+      "learning_rate": 1.8153504691841528e-05,
+      "loss": 0.9005,
+      "step": 1370
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2903427164433185,
+      "learning_rate": 1.8150481682849258e-05,
+      "loss": 0.8687,
+      "step": 1371
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.673302462193393,
+      "learning_rate": 1.814745645344111e-05,
+      "loss": 0.9552,
+      "step": 1372
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4645751369878004,
+      "learning_rate": 1.814442900444124e-05,
+      "loss": 0.9533,
+      "step": 1373
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.305852153899607,
+      "learning_rate": 1.8141399336674403e-05,
+      "loss": 0.8746,
+      "step": 1374
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6297915154835378,
+      "learning_rate": 1.8138367450965968e-05,
+      "loss": 0.9031,
+      "step": 1375
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.333331367853201,
+      "learning_rate": 1.8135333348141904e-05,
+      "loss": 0.8936,
+      "step": 1376
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3548416039044349,
+      "learning_rate": 1.813229702902879e-05,
+      "loss": 0.9389,
+      "step": 1377
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4322877165711003,
+      "learning_rate": 1.812925849445379e-05,
+      "loss": 0.9287,
+      "step": 1378
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5219319441841863,
+      "learning_rate": 1.8126217745244696e-05,
+      "loss": 0.9363,
+      "step": 1379
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2925971666289797,
+      "learning_rate": 1.8123174782229884e-05,
+      "loss": 0.992,
+      "step": 1380
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6749370789541775,
+      "learning_rate": 1.8120129606238345e-05,
+      "loss": 0.9655,
+      "step": 1381
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3194344241684788,
+      "learning_rate": 1.811708221809967e-05,
+      "loss": 0.9033,
+      "step": 1382
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1825220465879462,
+      "learning_rate": 1.8114032618644053e-05,
+      "loss": 1.007,
+      "step": 1383
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5392617091277325,
+      "learning_rate": 1.8110980808702282e-05,
+      "loss": 0.923,
+      "step": 1384
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8354536494358633,
+      "learning_rate": 1.810792678910576e-05,
+      "loss": 0.9855,
+      "step": 1385
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5192886406475323,
+      "learning_rate": 1.810487056068648e-05,
+      "loss": 0.9802,
+      "step": 1386
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2407133105067525,
+      "learning_rate": 1.8101812124277053e-05,
+      "loss": 0.9633,
+      "step": 1387
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2591648872515129,
+      "learning_rate": 1.8098751480710675e-05,
+      "loss": 0.9289,
+      "step": 1388
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2967450187360643,
+      "learning_rate": 1.8095688630821147e-05,
+      "loss": 0.9089,
+      "step": 1389
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7139866105338972,
+      "learning_rate": 1.809262357544288e-05,
+      "loss": 0.9372,
+      "step": 1390
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.578663188925344,
+      "learning_rate": 1.8089556315410875e-05,
+      "loss": 1.001,
+      "step": 1391
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5414224895548923,
+      "learning_rate": 1.808648685156074e-05,
+      "loss": 0.8921,
+      "step": 1392
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2827503437579575,
+      "learning_rate": 1.808341518472868e-05,
+      "loss": 0.8932,
+      "step": 1393
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.406253161691714,
+      "learning_rate": 1.8080341315751507e-05,
+      "loss": 0.9096,
+      "step": 1394
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4180396393324481,
+      "learning_rate": 1.8077265245466623e-05,
+      "loss": 0.8994,
+      "step": 1395
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5591949137685248,
+      "learning_rate": 1.8074186974712033e-05,
+      "loss": 1.0144,
+      "step": 1396
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5073974608262897,
+      "learning_rate": 1.8071106504326344e-05,
+      "loss": 0.9537,
+      "step": 1397
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.320117331503544,
+      "learning_rate": 1.8068023835148763e-05,
+      "loss": 0.9297,
+      "step": 1398
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4934937816007674,
+      "learning_rate": 1.8064938968019096e-05,
+      "loss": 0.9413,
+      "step": 1399
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3676624808163538,
+      "learning_rate": 1.8061851903777738e-05,
+      "loss": 0.9321,
+      "step": 1400
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3483786664595354,
+      "learning_rate": 1.8058762643265697e-05,
+      "loss": 0.9168,
+      "step": 1401
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3761018353542291,
+      "learning_rate": 1.805567118732457e-05,
+      "loss": 0.8949,
+      "step": 1402
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4594460539500764,
+      "learning_rate": 1.8052577536796557e-05,
+      "loss": 0.9182,
+      "step": 1403
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2984743489764452,
+      "learning_rate": 1.804948169252445e-05,
+      "loss": 0.844,
+      "step": 1404
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5494533536000397,
+      "learning_rate": 1.804638365535165e-05,
+      "loss": 0.9278,
+      "step": 1405
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4022473724527305,
+      "learning_rate": 1.8043283426122137e-05,
+      "loss": 0.9191,
+      "step": 1406
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3148604664502062,
+      "learning_rate": 1.8040181005680506e-05,
+      "loss": 0.9259,
+      "step": 1407
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4626493958565367,
+      "learning_rate": 1.803707639487194e-05,
+      "loss": 0.9294,
+      "step": 1408
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.406032283141865,
+      "learning_rate": 1.8033969594542223e-05,
+      "loss": 0.9313,
+      "step": 1409
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.477756492755267,
+      "learning_rate": 1.803086060553773e-05,
+      "loss": 0.8774,
+      "step": 1410
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4543686743755668,
+      "learning_rate": 1.8027749428705432e-05,
+      "loss": 1.002,
+      "step": 1411
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2492416700457738,
+      "learning_rate": 1.8024636064892907e-05,
+      "loss": 0.8866,
+      "step": 1412
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2880484233538472,
+      "learning_rate": 1.8021520514948315e-05,
+      "loss": 0.8685,
+      "step": 1413
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.506585854441499,
+      "learning_rate": 1.8018402779720424e-05,
+      "loss": 0.9914,
+      "step": 1414
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6478668966571188,
+      "learning_rate": 1.8015282860058584e-05,
+      "loss": 0.9536,
+      "step": 1415
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4633846760159714,
+      "learning_rate": 1.8012160756812755e-05,
+      "loss": 0.8998,
+      "step": 1416
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4804786427652983,
+      "learning_rate": 1.8009036470833474e-05,
+      "loss": 0.9834,
+      "step": 1417
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.38792622743611,
+      "learning_rate": 1.800591000297189e-05,
+      "loss": 0.8847,
+      "step": 1418
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3723432371266042,
+      "learning_rate": 1.8002781354079736e-05,
+      "loss": 0.9696,
+      "step": 1419
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5180262756319356,
+      "learning_rate": 1.7999650525009343e-05,
+      "loss": 0.9637,
+      "step": 1420
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3724392798698994,
+      "learning_rate": 1.7996517516613637e-05,
+      "loss": 0.8766,
+      "step": 1421
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3903772148104654,
+      "learning_rate": 1.7993382329746135e-05,
+      "loss": 0.9759,
+      "step": 1422
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.520317534911786,
+      "learning_rate": 1.7990244965260945e-05,
+      "loss": 0.9438,
+      "step": 1423
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4104427432317215,
+      "learning_rate": 1.7987105424012777e-05,
+      "loss": 0.9377,
+      "step": 1424
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4114118946859544,
+      "learning_rate": 1.798396370685692e-05,
+      "loss": 0.8924,
+      "step": 1425
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6551549976026878,
+      "learning_rate": 1.7980819814649276e-05,
+      "loss": 0.8985,
+      "step": 1426
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3457685338416856,
+      "learning_rate": 1.797767374824632e-05,
+      "loss": 0.9271,
+      "step": 1427
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4759906886496073,
+      "learning_rate": 1.797452550850513e-05,
+      "loss": 0.9423,
+      "step": 1428
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6250092054528142,
+      "learning_rate": 1.797137509628337e-05,
+      "loss": 0.964,
+      "step": 1429
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4859354525571047,
+      "learning_rate": 1.7968222512439302e-05,
+      "loss": 0.9189,
+      "step": 1430
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.377310656400429,
+      "learning_rate": 1.7965067757831777e-05,
+      "loss": 0.8937,
+      "step": 1431
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3970063281142926,
+      "learning_rate": 1.796191083332024e-05,
+      "loss": 0.9112,
+      "step": 1432
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3131897864979747,
+      "learning_rate": 1.7958751739764716e-05,
+      "loss": 0.8635,
+      "step": 1433
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5551914256220765,
+      "learning_rate": 1.7955590478025833e-05,
+      "loss": 0.9073,
+      "step": 1434
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3823252940333748,
+      "learning_rate": 1.7952427048964808e-05,
+      "loss": 0.9434,
+      "step": 1435
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5950864372965423,
+      "learning_rate": 1.7949261453443446e-05,
+      "loss": 1.0352,
+      "step": 1436
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3557532414566502,
+      "learning_rate": 1.794609369232414e-05,
+      "loss": 0.9306,
+      "step": 1437
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2719863048853584,
+      "learning_rate": 1.794292376646988e-05,
+      "loss": 0.9049,
+      "step": 1438
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5215359331388483,
+      "learning_rate": 1.7939751676744234e-05,
+      "loss": 0.9462,
+      "step": 1439
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4729257154951143,
+      "learning_rate": 1.7936577424011375e-05,
+      "loss": 0.902,
+      "step": 1440
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3584938619130646,
+      "learning_rate": 1.793340100913605e-05,
+      "loss": 0.9406,
+      "step": 1441
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5936746941801514,
+      "learning_rate": 1.7930222432983605e-05,
+      "loss": 0.919,
+      "step": 1442
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5405521711232046,
+      "learning_rate": 1.792704169641997e-05,
+      "loss": 0.859,
+      "step": 1443
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4349608230642699,
+      "learning_rate": 1.7923858800311665e-05,
+      "loss": 0.9123,
+      "step": 1444
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6949011467323114,
+      "learning_rate": 1.7920673745525805e-05,
+      "loss": 0.9873,
+      "step": 1445
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3507833943314131,
+      "learning_rate": 1.7917486532930082e-05,
+      "loss": 0.9045,
+      "step": 1446
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4297559349169837,
+      "learning_rate": 1.7914297163392778e-05,
+      "loss": 0.8762,
+      "step": 1447
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3909769205628935,
+      "learning_rate": 1.7911105637782767e-05,
+      "loss": 0.9199,
+      "step": 1448
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4665736747352947,
+      "learning_rate": 1.790791195696951e-05,
+      "loss": 0.9659,
+      "step": 1449
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8178568375054327,
+      "learning_rate": 1.790471612182306e-05,
+      "loss": 0.8868,
+      "step": 1450
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3791811401591183,
+      "learning_rate": 1.7901518133214034e-05,
+      "loss": 0.9306,
+      "step": 1451
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2551781384772345,
+      "learning_rate": 1.789831799201367e-05,
+      "loss": 0.956,
+      "step": 1452
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5692239067382618,
+      "learning_rate": 1.789511569909376e-05,
+      "loss": 0.9627,
+      "step": 1453
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4792992135526286,
+      "learning_rate": 1.7891911255326705e-05,
+      "loss": 0.844,
+      "step": 1454
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.609924411278348,
+      "learning_rate": 1.7888704661585483e-05,
+      "loss": 0.8855,
+      "step": 1455
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4508613431306197,
+      "learning_rate": 1.7885495918743658e-05,
+      "loss": 0.9068,
+      "step": 1456
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4527909988731622,
+      "learning_rate": 1.7882285027675377e-05,
+      "loss": 0.8861,
+      "step": 1457
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.516929346022876,
+      "learning_rate": 1.7879071989255383e-05,
+      "loss": 0.9063,
+      "step": 1458
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0928756601677976,
+      "learning_rate": 1.787585680435899e-05,
+      "loss": 1.0475,
+      "step": 1459
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7781670277544546,
+      "learning_rate": 1.7872639473862102e-05,
+      "loss": 0.8542,
+      "step": 1460
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3886237598888043,
+      "learning_rate": 1.786941999864121e-05,
+      "loss": 0.9471,
+      "step": 1461
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2231879327501474,
+      "learning_rate": 1.7866198379573393e-05,
+      "loss": 0.9136,
+      "step": 1462
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.524376787491173,
+      "learning_rate": 1.78629746175363e-05,
+      "loss": 0.9295,
+      "step": 1463
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.664827037970853,
+      "learning_rate": 1.785974871340818e-05,
+      "loss": 0.914,
+      "step": 1464
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3099370111612605,
+      "learning_rate": 1.7856520668067854e-05,
+      "loss": 0.92,
+      "step": 1465
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2264179456625537,
+      "learning_rate": 1.785329048239473e-05,
+      "loss": 0.9225,
+      "step": 1466
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.457765702204839,
+      "learning_rate": 1.7850058157268803e-05,
+      "loss": 0.9343,
+      "step": 1467
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7026262898367779,
+      "learning_rate": 1.7846823693570645e-05,
+      "loss": 0.9294,
+      "step": 1468
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5711821904416883,
+      "learning_rate": 1.7843587092181417e-05,
+      "loss": 0.9419,
+      "step": 1469
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7589592954528017,
+      "learning_rate": 1.7840348353982852e-05,
+      "loss": 0.9577,
+      "step": 1470
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5320146437216193,
+      "learning_rate": 1.7837107479857275e-05,
+      "loss": 0.9488,
+      "step": 1471
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7398896491356037,
+      "learning_rate": 1.7833864470687593e-05,
+      "loss": 0.8914,
+      "step": 1472
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8224640300758423,
+      "learning_rate": 1.7830619327357282e-05,
+      "loss": 0.9223,
+      "step": 1473
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8512694112557286,
+      "learning_rate": 1.7827372050750414e-05,
+      "loss": 0.3502,
+      "step": 1474
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3786712015918483,
+      "learning_rate": 1.782412264175164e-05,
+      "loss": 0.9233,
+      "step": 1475
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6252409817637843,
+      "learning_rate": 1.7820871101246185e-05,
+      "loss": 0.9699,
+      "step": 1476
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3838162513732415,
+      "learning_rate": 1.7817617430119858e-05,
+      "loss": 0.9232,
+      "step": 1477
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3881408927702015,
+      "learning_rate": 1.7814361629259052e-05,
+      "loss": 0.8212,
+      "step": 1478
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.696564368734246,
+      "learning_rate": 1.781110369955073e-05,
+      "loss": 0.8954,
+      "step": 1479
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.588319268970133,
+      "learning_rate": 1.780784364188245e-05,
+      "loss": 0.9834,
+      "step": 1480
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.776178921479375,
+      "learning_rate": 1.780458145714233e-05,
+      "loss": 0.8975,
+      "step": 1481
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4175030019600514,
+      "learning_rate": 1.780131714621909e-05,
+      "loss": 0.884,
+      "step": 1482
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3667280586609718,
+      "learning_rate": 1.779805071000202e-05,
+      "loss": 0.8617,
+      "step": 1483
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4043583107085242,
+      "learning_rate": 1.7794782149380977e-05,
+      "loss": 0.9746,
+      "step": 1484
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2685824560777859,
+      "learning_rate": 1.7791511465246413e-05,
+      "loss": 0.9546,
+      "step": 1485
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.243090851419248,
+      "learning_rate": 1.778823865848935e-05,
+      "loss": 0.9619,
+      "step": 1486
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3205551038843497,
+      "learning_rate": 1.7784963730001395e-05,
+      "loss": 0.8851,
+      "step": 1487
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4629639230300375,
+      "learning_rate": 1.7781686680674726e-05,
+      "loss": 0.9435,
+      "step": 1488
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5142142437793134,
+      "learning_rate": 1.77784075114021e-05,
+      "loss": 0.8821,
+      "step": 1489
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5504916747098654,
+      "learning_rate": 1.7775126223076857e-05,
+      "loss": 0.925,
+      "step": 1490
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7386160142418696,
+      "learning_rate": 1.7771842816592907e-05,
+      "loss": 0.952,
+      "step": 1491
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5010075854469,
+      "learning_rate": 1.776855729284474e-05,
+      "loss": 0.9483,
+      "step": 1492
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.322186607129727,
+      "learning_rate": 1.7765269652727427e-05,
+      "loss": 0.878,
+      "step": 1493
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3781340249516472,
+      "learning_rate": 1.7761979897136606e-05,
+      "loss": 0.9551,
+      "step": 1494
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3560533939197716,
+      "learning_rate": 1.7758688026968506e-05,
+      "loss": 0.8735,
+      "step": 1495
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3132814478689676,
+      "learning_rate": 1.7755394043119916e-05,
+      "loss": 0.9825,
+      "step": 1496
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3836426559886001,
+      "learning_rate": 1.775209794648821e-05,
+      "loss": 0.9209,
+      "step": 1497
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6041134441255716,
+      "learning_rate": 1.7748799737971335e-05,
+      "loss": 0.831,
+      "step": 1498
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3286040869005302,
+      "learning_rate": 1.7745499418467814e-05,
+      "loss": 0.9287,
+      "step": 1499
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6941438648508664,
+      "learning_rate": 1.774219698887675e-05,
+      "loss": 0.9639,
+      "step": 1500
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3842512965905442,
+      "learning_rate": 1.7738892450097806e-05,
+      "loss": 0.9151,
+      "step": 1501
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5450894256477892,
+      "learning_rate": 1.7735585803031235e-05,
+      "loss": 0.9361,
+      "step": 1502
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4557121249343312,
+      "learning_rate": 1.7732277048577858e-05,
+      "loss": 0.9788,
+      "step": 1503
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4652782279042018,
+      "learning_rate": 1.7728966187639075e-05,
+      "loss": 0.9639,
+      "step": 1504
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5597329428143736,
+      "learning_rate": 1.7725653221116853e-05,
+      "loss": 0.9454,
+      "step": 1505
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5876108459151903,
+      "learning_rate": 1.772233814991373e-05,
+      "loss": 0.8667,
+      "step": 1506
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.403979079799209,
+      "learning_rate": 1.771902097493283e-05,
+      "loss": 1.0105,
+      "step": 1507
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.34899227080926,
+      "learning_rate": 1.771570169707784e-05,
+      "loss": 0.9293,
+      "step": 1508
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3286380211749882,
+      "learning_rate": 1.7712380317253025e-05,
+      "loss": 0.9212,
+      "step": 1509
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5016283073453969,
+      "learning_rate": 1.7709056836363218e-05,
+      "loss": 0.9574,
+      "step": 1510
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3886239999923242,
+      "learning_rate": 1.7705731255313823e-05,
+      "loss": 0.8998,
+      "step": 1511
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4753339200059699,
+      "learning_rate": 1.770240357501083e-05,
+      "loss": 0.9191,
+      "step": 1512
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3819384385229503,
+      "learning_rate": 1.769907379636078e-05,
+      "loss": 0.9306,
+      "step": 1513
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2355884445510876,
+      "learning_rate": 1.7695741920270807e-05,
+      "loss": 0.8063,
+      "step": 1514
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2513299743575326,
+      "learning_rate": 1.76924079476486e-05,
+      "loss": 0.9152,
+      "step": 1515
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3067684952074006,
+      "learning_rate": 1.768907187940242e-05,
+      "loss": 0.9547,
+      "step": 1516
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.303801981286533,
+      "learning_rate": 1.7685733716441117e-05,
+      "loss": 0.9201,
+      "step": 1517
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6604463368708118,
+      "learning_rate": 1.7682393459674087e-05,
+      "loss": 0.8578,
+      "step": 1518
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3768999718491584,
+      "learning_rate": 1.7679051110011312e-05,
+      "loss": 0.9712,
+      "step": 1519
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5061142318623844,
+      "learning_rate": 1.7675706668363338e-05,
+      "loss": 1.0097,
+      "step": 1520
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3707112868406561,
+      "learning_rate": 1.7672360135641292e-05,
+      "loss": 0.8881,
+      "step": 1521
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6107150275469724,
+      "learning_rate": 1.766901151275685e-05,
+      "loss": 0.9494,
+      "step": 1522
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6488736460415707,
+      "learning_rate": 1.7665660800622274e-05,
+      "loss": 0.9439,
+      "step": 1523
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3736304139249622,
+      "learning_rate": 1.7662308000150393e-05,
+      "loss": 0.9085,
+      "step": 1524
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.998250365499198,
+      "learning_rate": 1.7658953112254603e-05,
+      "loss": 0.9154,
+      "step": 1525
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3879826360340926,
+      "learning_rate": 1.765559613784886e-05,
+      "loss": 0.8934,
+      "step": 1526
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4338861006249786,
+      "learning_rate": 1.7652237077847703e-05,
+      "loss": 0.874,
+      "step": 1527
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.265560115369577,
+      "learning_rate": 1.7648875933166228e-05,
+      "loss": 0.9352,
+      "step": 1528
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3934951587732518,
+      "learning_rate": 1.764551270472011e-05,
+      "loss": 0.9258,
+      "step": 1529
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4487740189011102,
+      "learning_rate": 1.764214739342558e-05,
+      "loss": 0.9284,
+      "step": 1530
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3826910598011186,
+      "learning_rate": 1.7638780000199446e-05,
+      "loss": 0.9548,
+      "step": 1531
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5046801446079305,
+      "learning_rate": 1.7635410525959072e-05,
+      "loss": 0.8838,
+      "step": 1532
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3354668527167677,
+      "learning_rate": 1.76320389716224e-05,
+      "loss": 0.8897,
+      "step": 1533
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5934494801093575,
+      "learning_rate": 1.7628665338107936e-05,
+      "loss": 0.8925,
+      "step": 1534
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.626466152725449,
+      "learning_rate": 1.7625289626334744e-05,
+      "loss": 0.913,
+      "step": 1535
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3699017271698222,
+      "learning_rate": 1.762191183722247e-05,
+      "loss": 0.8034,
+      "step": 1536
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.566841406321332,
+      "learning_rate": 1.761853197169131e-05,
+      "loss": 0.8913,
+      "step": 1537
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.822387670398342,
+      "learning_rate": 1.7615150030662037e-05,
+      "loss": 0.3522,
+      "step": 1538
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5330349052378704,
+      "learning_rate": 1.7611766015055984e-05,
+      "loss": 0.9569,
+      "step": 1539
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6057219439079193,
+      "learning_rate": 1.7608379925795043e-05,
+      "loss": 0.9539,
+      "step": 1540
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.51380736575235,
+      "learning_rate": 1.7604991763801688e-05,
+      "loss": 0.9078,
+      "step": 1541
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2881288145331344,
+      "learning_rate": 1.760160152999894e-05,
+      "loss": 0.8744,
+      "step": 1542
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4834039843708575,
+      "learning_rate": 1.75982092253104e-05,
+      "loss": 0.8857,
+      "step": 1543
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4069587395081684,
+      "learning_rate": 1.759481485066022e-05,
+      "loss": 0.9222,
+      "step": 1544
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4591545008345639,
+      "learning_rate": 1.7591418406973126e-05,
+      "loss": 0.9346,
+      "step": 1545
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4680304717831996,
+      "learning_rate": 1.7588019895174394e-05,
+      "loss": 0.8929,
+      "step": 1546
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5366527807449342,
+      "learning_rate": 1.758461931618988e-05,
+      "loss": 0.9313,
+      "step": 1547
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.537533306872229,
+      "learning_rate": 1.758121667094599e-05,
+      "loss": 0.9283,
+      "step": 1548
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6314236367191264,
+      "learning_rate": 1.7577811960369703e-05,
+      "loss": 0.9175,
+      "step": 1549
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2702649948843467,
+      "learning_rate": 1.7574405185388556e-05,
+      "loss": 0.8699,
+      "step": 1550
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7877521658956605,
+      "learning_rate": 1.7570996346930645e-05,
+      "loss": 0.334,
+      "step": 1551
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6333969863361506,
+      "learning_rate": 1.7567585445924632e-05,
+      "loss": 0.9768,
+      "step": 1552
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6370386810166844,
+      "learning_rate": 1.7564172483299748e-05,
+      "loss": 0.8915,
+      "step": 1553
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3711045983420596,
+      "learning_rate": 1.7560757459985767e-05,
+      "loss": 0.9273,
+      "step": 1554
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.485050782919194,
+      "learning_rate": 1.7557340376913038e-05,
+      "loss": 0.8857,
+      "step": 1555
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8168810212271668,
+      "learning_rate": 1.7553921235012475e-05,
+      "loss": 0.3767,
+      "step": 1556
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.616051608411855,
+      "learning_rate": 1.755050003521554e-05,
+      "loss": 0.8764,
+      "step": 1557
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3313738771877215,
+      "learning_rate": 1.7547076778454268e-05,
+      "loss": 0.9767,
+      "step": 1558
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3897301433064186,
+      "learning_rate": 1.7543651465661244e-05,
+      "loss": 0.9189,
+      "step": 1559
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4656128619022848,
+      "learning_rate": 1.754022409776962e-05,
+      "loss": 0.8615,
+      "step": 1560
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2790980766983986,
+      "learning_rate": 1.75367946757131e-05,
+      "loss": 0.8541,
+      "step": 1561
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6955185866016322,
+      "learning_rate": 1.7533363200425962e-05,
+      "loss": 0.925,
+      "step": 1562
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.476534364447397,
+      "learning_rate": 1.752992967284303e-05,
+      "loss": 0.9457,
+      "step": 1563
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.235884520191518,
+      "learning_rate": 1.7526494093899687e-05,
+      "loss": 0.8745,
+      "step": 1564
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4026909693198406,
+      "learning_rate": 1.7523056464531887e-05,
+      "loss": 0.8942,
+      "step": 1565
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3107086400169938,
+      "learning_rate": 1.751961678567613e-05,
+      "loss": 0.8997,
+      "step": 1566
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3116070336565369,
+      "learning_rate": 1.751617505826948e-05,
+      "loss": 0.8633,
+      "step": 1567
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2509419085030418,
+      "learning_rate": 1.7512731283249563e-05,
+      "loss": 0.8663,
+      "step": 1568
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3406504796458962,
+      "learning_rate": 1.7509285461554552e-05,
+      "loss": 0.9598,
+      "step": 1569
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4142874629654651,
+      "learning_rate": 1.7505837594123186e-05,
+      "loss": 0.8582,
+      "step": 1570
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6180474890259768,
+      "learning_rate": 1.7502387681894765e-05,
+      "loss": 0.9012,
+      "step": 1571
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2981787219717944,
+      "learning_rate": 1.749893572580913e-05,
+      "loss": 0.901,
+      "step": 1572
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6972041695095221,
+      "learning_rate": 1.7495481726806697e-05,
+      "loss": 0.9211,
+      "step": 1573
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4548929762334823,
+      "learning_rate": 1.7492025685828428e-05,
+      "loss": 0.8066,
+      "step": 1574
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2501479299750213,
+      "learning_rate": 1.748856760381584e-05,
+      "loss": 0.9318,
+      "step": 1575
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8617111814603586,
+      "learning_rate": 1.7485107481711014e-05,
+      "loss": 0.9229,
+      "step": 1576
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2702661954527332,
+      "learning_rate": 1.748164532045658e-05,
+      "loss": 0.9123,
+      "step": 1577
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3505731528250098,
+      "learning_rate": 1.747818112099573e-05,
+      "loss": 0.9757,
+      "step": 1578
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.433443666418462,
+      "learning_rate": 1.7474714884272208e-05,
+      "loss": 0.8932,
+      "step": 1579
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.351289736092245,
+      "learning_rate": 1.7471246611230307e-05,
+      "loss": 0.9298,
+      "step": 1580
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.331558394963455,
+      "learning_rate": 1.7467776302814882e-05,
+      "loss": 0.9135,
+      "step": 1581
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.271593672585788,
+      "learning_rate": 1.7464303959971343e-05,
+      "loss": 0.9564,
+      "step": 1582
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.35125491881556,
+      "learning_rate": 1.746082958364565e-05,
+      "loss": 0.8665,
+      "step": 1583
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6404041289200506,
+      "learning_rate": 1.745735317478432e-05,
+      "loss": 0.86,
+      "step": 1584
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3675904075400056,
+      "learning_rate": 1.7453874734334422e-05,
+      "loss": 0.928,
+      "step": 1585
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3558293036200917,
+      "learning_rate": 1.745039426324358e-05,
+      "loss": 0.8523,
+      "step": 1586
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4274944868394677,
+      "learning_rate": 1.7446911762459968e-05,
+      "loss": 0.9072,
+      "step": 1587
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6804319883126988,
+      "learning_rate": 1.7443427232932316e-05,
+      "loss": 0.913,
+      "step": 1588
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.467127302441046,
+      "learning_rate": 1.7439940675609906e-05,
+      "loss": 0.8291,
+      "step": 1589
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3284174232563115,
+      "learning_rate": 1.7436452091442574e-05,
+      "loss": 0.9267,
+      "step": 1590
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.360130199282049,
+      "learning_rate": 1.7432961481380707e-05,
+      "loss": 0.9188,
+      "step": 1591
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5272952655013883,
+      "learning_rate": 1.742946884637524e-05,
+      "loss": 0.9278,
+      "step": 1592
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5232543040435773,
+      "learning_rate": 1.7425974187377665e-05,
+      "loss": 0.9895,
+      "step": 1593
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6917296798562969,
+      "learning_rate": 1.7422477505340024e-05,
+      "loss": 0.8411,
+      "step": 1594
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.68951994592901,
+      "learning_rate": 1.7418978801214906e-05,
+      "loss": 0.9481,
+      "step": 1595
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3740161056431388,
+      "learning_rate": 1.741547807595546e-05,
+      "loss": 0.9881,
+      "step": 1596
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.537902430531926,
+      "learning_rate": 1.7411975330515377e-05,
+      "loss": 0.9062,
+      "step": 1597
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2796198641016825,
+      "learning_rate": 1.7408470565848897e-05,
+      "loss": 0.8651,
+      "step": 1598
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4728954940453995,
+      "learning_rate": 1.7404963782910828e-05,
+      "loss": 0.9563,
+      "step": 1599
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.42905022926618,
+      "learning_rate": 1.74014549826565e-05,
+      "loss": 0.8388,
+      "step": 1600
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.528822518367215,
+      "learning_rate": 1.739794416604181e-05,
+      "loss": 0.9046,
+      "step": 1601
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9858454618629265,
+      "learning_rate": 1.739443133402321e-05,
+      "loss": 0.3389,
+      "step": 1602
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3971105378849025,
+      "learning_rate": 1.739091648755768e-05,
+      "loss": 0.897,
+      "step": 1603
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.687847479723696,
+      "learning_rate": 1.7387399627602773e-05,
+      "loss": 0.9807,
+      "step": 1604
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2858939003245111,
+      "learning_rate": 1.7383880755116567e-05,
+      "loss": 0.9248,
+      "step": 1605
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4310881685870116,
+      "learning_rate": 1.7380359871057707e-05,
+      "loss": 0.9417,
+      "step": 1606
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3008869968709962,
+      "learning_rate": 1.737683697638538e-05,
+      "loss": 0.8745,
+      "step": 1607
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.399665773944484,
+      "learning_rate": 1.7373312072059313e-05,
+      "loss": 0.9181,
+      "step": 1608
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4466663489248912,
+      "learning_rate": 1.7369785159039794e-05,
+      "loss": 0.8917,
+      "step": 1609
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3981887760023952,
+      "learning_rate": 1.7366256238287647e-05,
+      "loss": 0.8774,
+      "step": 1610
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2575035011838098,
+      "learning_rate": 1.736272531076425e-05,
+      "loss": 0.9025,
+      "step": 1611
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6762845343994273,
+      "learning_rate": 1.735919237743152e-05,
+      "loss": 1.0239,
+      "step": 1612
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3580818355800202,
+      "learning_rate": 1.7355657439251933e-05,
+      "loss": 0.9134,
+      "step": 1613
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5195426033400008,
+      "learning_rate": 1.7352120497188497e-05,
+      "loss": 0.8867,
+      "step": 1614
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6766742514404924,
+      "learning_rate": 1.7348581552204776e-05,
+      "loss": 0.9361,
+      "step": 1615
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.597199248728031,
+      "learning_rate": 1.734504060526488e-05,
+      "loss": 0.897,
+      "step": 1616
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.453109387985574,
+      "learning_rate": 1.734149765733345e-05,
+      "loss": 0.9055,
+      "step": 1617
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3422999320495346,
+      "learning_rate": 1.7337952709375688e-05,
+      "loss": 0.8093,
+      "step": 1618
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4037057344651518,
+      "learning_rate": 1.733440576235734e-05,
+      "loss": 1.0066,
+      "step": 1619
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2437857068479816,
+      "learning_rate": 1.7330856817244686e-05,
+      "loss": 0.9456,
+      "step": 1620
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.763271014075135,
+      "learning_rate": 1.732730587500456e-05,
+      "loss": 0.3637,
+      "step": 1621
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2821856490011891,
+      "learning_rate": 1.7323752936604334e-05,
+      "loss": 0.827,
+      "step": 1622
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.456608199721232,
+      "learning_rate": 1.732019800301193e-05,
+      "loss": 0.9262,
+      "step": 1623
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.379231485503792,
+      "learning_rate": 1.7316641075195802e-05,
+      "loss": 0.8474,
+      "step": 1624
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4459960873865894,
+      "learning_rate": 1.7313082154124966e-05,
+      "loss": 0.8392,
+      "step": 1625
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3425809548875087,
+      "learning_rate": 1.7309521240768964e-05,
+      "loss": 0.8642,
+      "step": 1626
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6765583699540803,
+      "learning_rate": 1.7305958336097888e-05,
+      "loss": 0.9333,
+      "step": 1627
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.560699147516626,
+      "learning_rate": 1.730239344108237e-05,
+      "loss": 1.0107,
+      "step": 1628
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.348318388580259,
+      "learning_rate": 1.729882655669359e-05,
+      "loss": 0.9269,
+      "step": 1629
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5533882833493289,
+      "learning_rate": 1.729525768390326e-05,
+      "loss": 0.9577,
+      "step": 1630
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5295036938521527,
+      "learning_rate": 1.729168682368364e-05,
+      "loss": 0.9564,
+      "step": 1631
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5305043425469185,
+      "learning_rate": 1.7288113977007538e-05,
+      "loss": 0.9332,
+      "step": 1632
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5423458888365909,
+      "learning_rate": 1.7284539144848286e-05,
+      "loss": 0.9306,
+      "step": 1633
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.717440114691662,
+      "learning_rate": 1.7280962328179773e-05,
+      "loss": 0.9688,
+      "step": 1634
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.409033535775763,
+      "learning_rate": 1.727738352797642e-05,
+      "loss": 0.9159,
+      "step": 1635
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5232554736574733,
+      "learning_rate": 1.7273802745213193e-05,
+      "loss": 0.9008,
+      "step": 1636
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.731236077805256,
+      "learning_rate": 1.7270219980865594e-05,
+      "loss": 0.9496,
+      "step": 1637
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.671215276759986,
+      "learning_rate": 1.7266635235909664e-05,
+      "loss": 0.8994,
+      "step": 1638
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4765817468135582,
+      "learning_rate": 1.7263048511321995e-05,
+      "loss": 0.9116,
+      "step": 1639
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7645839010152913,
+      "learning_rate": 1.7259459808079705e-05,
+      "loss": 0.9373,
+      "step": 1640
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.947990431203616,
+      "learning_rate": 1.7255869127160452e-05,
+      "loss": 0.3403,
+      "step": 1641
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3994209365249972,
+      "learning_rate": 1.725227646954244e-05,
+      "loss": 0.8872,
+      "step": 1642
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5450412966921077,
+      "learning_rate": 1.724868183620441e-05,
+      "loss": 0.8651,
+      "step": 1643
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.527420349956401,
+      "learning_rate": 1.724508522812563e-05,
+      "loss": 0.8958,
+      "step": 1644
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3164689133300786,
+      "learning_rate": 1.724148664628593e-05,
+      "loss": 0.9359,
+      "step": 1645
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9120700529554698,
+      "learning_rate": 1.7237886091665653e-05,
+      "loss": 0.3342,
+      "step": 1646
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4503632528504864,
+      "learning_rate": 1.7234283565245688e-05,
+      "loss": 1.0021,
+      "step": 1647
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7308955925501437,
+      "learning_rate": 1.723067906800747e-05,
+      "loss": 0.9014,
+      "step": 1648
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2958425856213454,
+      "learning_rate": 1.7227072600932952e-05,
+      "loss": 0.9942,
+      "step": 1649
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4800440259555905,
+      "learning_rate": 1.7223464165004648e-05,
+      "loss": 0.9659,
+      "step": 1650
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5539254964093996,
+      "learning_rate": 1.7219853761205588e-05,
+      "loss": 0.8367,
+      "step": 1651
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.289624835671746,
+      "learning_rate": 1.7216241390519348e-05,
+      "loss": 0.8484,
+      "step": 1652
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.421199623775637,
+      "learning_rate": 1.7212627053930034e-05,
+      "loss": 0.9456,
+      "step": 1653
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2468958742383034,
+      "learning_rate": 1.7209010752422296e-05,
+      "loss": 0.8961,
+      "step": 1654
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4575552758669268,
+      "learning_rate": 1.7205392486981306e-05,
+      "loss": 0.873,
+      "step": 1655
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6644534627810683,
+      "learning_rate": 1.720177225859279e-05,
+      "loss": 0.8779,
+      "step": 1656
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4490642028207916,
+      "learning_rate": 1.7198150068242992e-05,
+      "loss": 0.8806,
+      "step": 1657
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2586906074984874,
+      "learning_rate": 1.71945259169187e-05,
+      "loss": 0.9563,
+      "step": 1658
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5917832239419862,
+      "learning_rate": 1.7190899805607224e-05,
+      "loss": 0.875,
+      "step": 1659
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5328061385610867,
+      "learning_rate": 1.7187271735296428e-05,
+      "loss": 0.878,
+      "step": 1660
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3410265014545164,
+      "learning_rate": 1.718364170697469e-05,
+      "loss": 0.9706,
+      "step": 1661
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.509223176791907,
+      "learning_rate": 1.7180009721630932e-05,
+      "loss": 0.8493,
+      "step": 1662
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2803901623517842,
+      "learning_rate": 1.717637578025461e-05,
+      "loss": 0.8431,
+      "step": 1663
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4815940241647154,
+      "learning_rate": 1.7172739883835705e-05,
+      "loss": 0.8787,
+      "step": 1664
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.456881659271399,
+      "learning_rate": 1.716910203336474e-05,
+      "loss": 0.9107,
+      "step": 1665
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5197804057946243,
+      "learning_rate": 1.716546222983276e-05,
+      "loss": 0.9026,
+      "step": 1666
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1046689574632567,
+      "learning_rate": 1.7161820474231355e-05,
+      "loss": 0.8771,
+      "step": 1667
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4683710941746164,
+      "learning_rate": 1.7158176767552636e-05,
+      "loss": 0.9889,
+      "step": 1668
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2468308201649616,
+      "learning_rate": 1.715453111078925e-05,
+      "loss": 0.8368,
+      "step": 1669
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2475229433643198,
+      "learning_rate": 1.715088350493437e-05,
+      "loss": 0.9047,
+      "step": 1670
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.378033428626109,
+      "learning_rate": 1.714723395098171e-05,
+      "loss": 0.9149,
+      "step": 1671
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3639837916870199,
+      "learning_rate": 1.7143582449925508e-05,
+      "loss": 0.8805,
+      "step": 1672
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3696970441600163,
+      "learning_rate": 1.7139929002760532e-05,
+      "loss": 0.9298,
+      "step": 1673
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6390529912303684,
+      "learning_rate": 1.713627361048208e-05,
+      "loss": 0.8627,
+      "step": 1674
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5105147754660921,
+      "learning_rate": 1.713261627408599e-05,
+      "loss": 0.9467,
+      "step": 1675
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5657000376656307,
+      "learning_rate": 1.7128956994568612e-05,
+      "loss": 0.9782,
+      "step": 1676
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4419605008802805,
+      "learning_rate": 1.7125295772926834e-05,
+      "loss": 0.8904,
+      "step": 1677
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4045057207775635,
+      "learning_rate": 1.7121632610158083e-05,
+      "loss": 0.8312,
+      "step": 1678
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6328565691973866,
+      "learning_rate": 1.7117967507260298e-05,
+      "loss": 0.9271,
+      "step": 1679
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8940965067668736,
+      "learning_rate": 1.7114300465231955e-05,
+      "loss": 0.37,
+      "step": 1680
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6241324922025069,
+      "learning_rate": 1.711063148507206e-05,
+      "loss": 1.0052,
+      "step": 1681
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2804788669084108,
+      "learning_rate": 1.710696056778014e-05,
+      "loss": 0.891,
+      "step": 1682
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4692550105482822,
+      "learning_rate": 1.710328771435626e-05,
+      "loss": 0.9188,
+      "step": 1683
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.317863019117824,
+      "learning_rate": 1.7099612925801005e-05,
+      "loss": 0.976,
+      "step": 1684
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3359038985917746,
+      "learning_rate": 1.7095936203115484e-05,
+      "loss": 0.8763,
+      "step": 1685
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3203848256102904,
+      "learning_rate": 1.709225754730134e-05,
+      "loss": 0.9167,
+      "step": 1686
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4316388214094764,
+      "learning_rate": 1.7088576959360743e-05,
+      "loss": 0.8574,
+      "step": 1687
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2602223115586864,
+      "learning_rate": 1.7084894440296383e-05,
+      "loss": 0.9142,
+      "step": 1688
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6806280530123008,
+      "learning_rate": 1.7081209991111483e-05,
+      "loss": 0.9115,
+      "step": 1689
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4235209282766383,
+      "learning_rate": 1.7077523612809784e-05,
+      "loss": 0.8984,
+      "step": 1690
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8117235008262265,
+      "learning_rate": 1.707383530639556e-05,
+      "loss": 0.3523,
+      "step": 1691
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.493211719258882,
+      "learning_rate": 1.707014507287361e-05,
+      "loss": 0.921,
+      "step": 1692
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.369873990521049,
+      "learning_rate": 1.706645291324925e-05,
+      "loss": 0.8266,
+      "step": 1693
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8139715171475574,
+      "learning_rate": 1.706275882852833e-05,
+      "loss": 0.8617,
+      "step": 1694
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.479826783595611,
+      "learning_rate": 1.705906281971722e-05,
+      "loss": 1.0706,
+      "step": 1695
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.366313854782972,
+      "learning_rate": 1.705536488782281e-05,
+      "loss": 0.8836,
+      "step": 1696
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.338671068202973,
+      "learning_rate": 1.7051665033852525e-05,
+      "loss": 1.0078,
+      "step": 1697
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7228660693594156,
+      "learning_rate": 1.70479632588143e-05,
+      "loss": 0.8365,
+      "step": 1698
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.375946305851306,
+      "learning_rate": 1.704425956371661e-05,
+      "loss": 0.9553,
+      "step": 1699
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6316573948782636,
+      "learning_rate": 1.704055394956844e-05,
+      "loss": 0.8432,
+      "step": 1700
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3826319400179754,
+      "learning_rate": 1.7036846417379295e-05,
+      "loss": 0.861,
+      "step": 1701
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3587403439616919,
+      "learning_rate": 1.7033136968159218e-05,
+      "loss": 0.9736,
+      "step": 1702
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4372309797131653,
+      "learning_rate": 1.7029425602918758e-05,
+      "loss": 0.933,
+      "step": 1703
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5215591013214864,
+      "learning_rate": 1.7025712322669e-05,
+      "loss": 0.9169,
+      "step": 1704
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5997183032553017,
+      "learning_rate": 1.7021997128421537e-05,
+      "loss": 0.8882,
+      "step": 1705
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7303226411417902,
+      "learning_rate": 1.7018280021188497e-05,
+      "loss": 0.8889,
+      "step": 1706
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6925961084332686,
+      "learning_rate": 1.701456100198252e-05,
+      "loss": 0.9085,
+      "step": 1707
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4651990612170933,
+      "learning_rate": 1.7010840071816765e-05,
+      "loss": 0.9634,
+      "step": 1708
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3180207093019016,
+      "learning_rate": 1.7007117231704923e-05,
+      "loss": 0.9305,
+      "step": 1709
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4493006910566164,
+      "learning_rate": 1.7003392482661188e-05,
+      "loss": 0.9591,
+      "step": 1710
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.273046617838578,
+      "learning_rate": 1.69996658257003e-05,
+      "loss": 0.8742,
+      "step": 1711
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3709116051260666,
+      "learning_rate": 1.6995937261837487e-05,
+      "loss": 0.9056,
+      "step": 1712
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5000023810794925,
+      "learning_rate": 1.6992206792088525e-05,
+      "loss": 0.9522,
+      "step": 1713
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4590540419903482,
+      "learning_rate": 1.6988474417469686e-05,
+      "loss": 0.8774,
+      "step": 1714
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3638290203817247,
+      "learning_rate": 1.6984740138997784e-05,
+      "loss": 0.8818,
+      "step": 1715
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.440169521694651,
+      "learning_rate": 1.6981003957690128e-05,
+      "loss": 0.8865,
+      "step": 1716
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5270008015551286,
+      "learning_rate": 1.6977265874564562e-05,
+      "loss": 0.9415,
+      "step": 1717
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4467885309534196,
+      "learning_rate": 1.6973525890639445e-05,
+      "loss": 0.9579,
+      "step": 1718
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4793824705981617,
+      "learning_rate": 1.6969784006933647e-05,
+      "loss": 0.8845,
+      "step": 1719
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6671657851770463,
+      "learning_rate": 1.696604022446656e-05,
+      "loss": 0.8528,
+      "step": 1720
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6090800729202437,
+      "learning_rate": 1.69622945442581e-05,
+      "loss": 0.9804,
+      "step": 1721
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4576746009878208,
+      "learning_rate": 1.6958546967328688e-05,
+      "loss": 0.936,
+      "step": 1722
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6419589455276276,
+      "learning_rate": 1.6954797494699263e-05,
+      "loss": 0.9106,
+      "step": 1723
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4258216995876707,
+      "learning_rate": 1.69510461273913e-05,
+      "loss": 0.9607,
+      "step": 1724
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3383984722612516,
+      "learning_rate": 1.6947292866426757e-05,
+      "loss": 0.8822,
+      "step": 1725
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3589109568707913,
+      "learning_rate": 1.6943537712828138e-05,
+      "loss": 0.8985,
+      "step": 1726
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5050136860770933,
+      "learning_rate": 1.6939780667618445e-05,
+      "loss": 0.9289,
+      "step": 1727
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2657478943957379,
+      "learning_rate": 1.6936021731821202e-05,
+      "loss": 0.9021,
+      "step": 1728
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4600311020031127,
+      "learning_rate": 1.6932260906460448e-05,
+      "loss": 0.9809,
+      "step": 1729
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3254060669012642,
+      "learning_rate": 1.6928498192560737e-05,
+      "loss": 0.8399,
+      "step": 1730
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4588964019562303,
+      "learning_rate": 1.6924733591147127e-05,
+      "loss": 0.8732,
+      "step": 1731
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4744808115714747,
+      "learning_rate": 1.692096710324521e-05,
+      "loss": 0.9501,
+      "step": 1732
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2299454007482096,
+      "learning_rate": 1.6917198729881073e-05,
+      "loss": 0.9078,
+      "step": 1733
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.536771252995333,
+      "learning_rate": 1.6913428472081328e-05,
+      "loss": 0.9818,
+      "step": 1734
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2879916489907295,
+      "learning_rate": 1.69096563308731e-05,
+      "loss": 0.9526,
+      "step": 1735
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2548301381436988,
+      "learning_rate": 1.690588230728402e-05,
+      "loss": 0.858,
+      "step": 1736
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.339787731584422,
+      "learning_rate": 1.690210640234224e-05,
+      "loss": 0.9282,
+      "step": 1737
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2911566774123973,
+      "learning_rate": 1.6898328617076418e-05,
+      "loss": 0.9327,
+      "step": 1738
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2147070686689085,
+      "learning_rate": 1.6894548952515726e-05,
+      "loss": 0.8746,
+      "step": 1739
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0206226631039683,
+      "learning_rate": 1.689076740968985e-05,
+      "loss": 0.3319,
+      "step": 1740
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.392396382674935,
+      "learning_rate": 1.6886983989628985e-05,
+      "loss": 0.9386,
+      "step": 1741
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.450032399078668,
+      "learning_rate": 1.6883198693363843e-05,
+      "loss": 0.9075,
+      "step": 1742
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2890749237155443,
+      "learning_rate": 1.687941152192564e-05,
+      "loss": 0.851,
+      "step": 1743
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3037717708581953,
+      "learning_rate": 1.6875622476346107e-05,
+      "loss": 0.9362,
+      "step": 1744
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7497504761034086,
+      "learning_rate": 1.687183155765748e-05,
+      "loss": 0.8896,
+      "step": 1745
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5210663407270555,
+      "learning_rate": 1.686803876689252e-05,
+      "loss": 0.8484,
+      "step": 1746
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6886943072210736,
+      "learning_rate": 1.6864244105084477e-05,
+      "loss": 0.8443,
+      "step": 1747
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.545226204577288,
+      "learning_rate": 1.6860447573267125e-05,
+      "loss": 0.882,
+      "step": 1748
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3567707695719973,
+      "learning_rate": 1.6856649172474747e-05,
+      "loss": 0.9174,
+      "step": 1749
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2927242405128132,
+      "learning_rate": 1.6852848903742125e-05,
+      "loss": 0.9099,
+      "step": 1750
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.975215208612929,
+      "learning_rate": 1.6849046768104567e-05,
+      "loss": 0.9642,
+      "step": 1751
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.326490441208533,
+      "learning_rate": 1.6845242766597873e-05,
+      "loss": 0.8566,
+      "step": 1752
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.439425723101024,
+      "learning_rate": 1.684143690025836e-05,
+      "loss": 0.9197,
+      "step": 1753
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5118173430325141,
+      "learning_rate": 1.6837629170122846e-05,
+      "loss": 0.8348,
+      "step": 1754
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.564935230117965,
+      "learning_rate": 1.683381957722867e-05,
+      "loss": 0.8856,
+      "step": 1755
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.349188491954696,
+      "learning_rate": 1.6830008122613665e-05,
+      "loss": 0.9681,
+      "step": 1756
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5095425239835352,
+      "learning_rate": 1.6826194807316177e-05,
+      "loss": 0.9068,
+      "step": 1757
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.242826540332664,
+      "learning_rate": 1.6822379632375055e-05,
+      "loss": 0.9144,
+      "step": 1758
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5488815647156526,
+      "learning_rate": 1.6818562598829666e-05,
+      "loss": 0.8395,
+      "step": 1759
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.563083006453182,
+      "learning_rate": 1.681474370771987e-05,
+      "loss": 0.9642,
+      "step": 1760
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4054693068311859,
+      "learning_rate": 1.6810922960086037e-05,
+      "loss": 0.8713,
+      "step": 1761
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5935769881277606,
+      "learning_rate": 1.6807100356969043e-05,
+      "loss": 0.8644,
+      "step": 1762
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9061703708622056,
+      "learning_rate": 1.6803275899410277e-05,
+      "loss": 0.379,
+      "step": 1763
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7638195315045784,
+      "learning_rate": 1.6799449588451618e-05,
+      "loss": 0.9026,
+      "step": 1764
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.172764518794516,
+      "learning_rate": 1.6795621425135465e-05,
+      "loss": 0.9395,
+      "step": 1765
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.251709454980728,
+      "learning_rate": 1.679179141050472e-05,
+      "loss": 0.9014,
+      "step": 1766
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3237932585005006,
+      "learning_rate": 1.678795954560277e-05,
+      "loss": 0.8873,
+      "step": 1767
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5650318780179766,
+      "learning_rate": 1.678412583147353e-05,
+      "loss": 0.8584,
+      "step": 1768
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4289068594858778,
+      "learning_rate": 1.678029026916141e-05,
+      "loss": 0.8444,
+      "step": 1769
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.343947804940275,
+      "learning_rate": 1.677645285971132e-05,
+      "loss": 0.917,
+      "step": 1770
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3209994408204184,
+      "learning_rate": 1.6772613604168677e-05,
+      "loss": 0.9182,
+      "step": 1771
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4723952565519836,
+      "learning_rate": 1.67687725035794e-05,
+      "loss": 0.8935,
+      "step": 1772
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.46843043678263,
+      "learning_rate": 1.676492955898991e-05,
+      "loss": 0.8259,
+      "step": 1773
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2072424121359966,
+      "learning_rate": 1.6761084771447133e-05,
+      "loss": 0.8722,
+      "step": 1774
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3663046738573317,
+      "learning_rate": 1.675723814199849e-05,
+      "loss": 0.9134,
+      "step": 1775
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9270701188670847,
+      "learning_rate": 1.675338967169192e-05,
+      "loss": 0.3224,
+      "step": 1776
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3612592869926525,
+      "learning_rate": 1.674953936157584e-05,
+      "loss": 0.9547,
+      "step": 1777
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3710154622433415,
+      "learning_rate": 1.6745687212699188e-05,
+      "loss": 0.8766,
+      "step": 1778
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5105537637336963,
+      "learning_rate": 1.674183322611139e-05,
+      "loss": 0.9358,
+      "step": 1779
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5378724994613515,
+      "learning_rate": 1.6737977402862383e-05,
+      "loss": 0.9574,
+      "step": 1780
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4201469741444492,
+      "learning_rate": 1.67341197440026e-05,
+      "loss": 0.9326,
+      "step": 1781
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3079506233156866,
+      "learning_rate": 1.673026025058297e-05,
+      "loss": 0.9078,
+      "step": 1782
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3163909932449473,
+      "learning_rate": 1.672639892365493e-05,
+      "loss": 0.9849,
+      "step": 1783
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4954216115828285,
+      "learning_rate": 1.6722535764270405e-05,
+      "loss": 0.8567,
+      "step": 1784
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.440579610231591,
+      "learning_rate": 1.671867077348183e-05,
+      "loss": 0.864,
+      "step": 1785
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.592551097168017,
+      "learning_rate": 1.6714803952342137e-05,
+      "loss": 0.8657,
+      "step": 1786
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5722606410071949,
+      "learning_rate": 1.671093530190475e-05,
+      "loss": 0.8906,
+      "step": 1787
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2369225219673405,
+      "learning_rate": 1.6707064823223603e-05,
+      "loss": 0.9543,
+      "step": 1788
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.366212061594568,
+      "learning_rate": 1.6703192517353116e-05,
+      "loss": 0.856,
+      "step": 1789
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.556054955203911,
+      "learning_rate": 1.6699318385348207e-05,
+      "loss": 0.9443,
+      "step": 1790
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7602980408446298,
+      "learning_rate": 1.6695442428264308e-05,
+      "loss": 0.9147,
+      "step": 1791
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.625324947508096,
+      "learning_rate": 1.669156464715733e-05,
+      "loss": 0.9261,
+      "step": 1792
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2799900914019162,
+      "learning_rate": 1.668768504308369e-05,
+      "loss": 0.9036,
+      "step": 1793
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3627750302606503,
+      "learning_rate": 1.6683803617100293e-05,
+      "loss": 0.9227,
+      "step": 1794
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3568214625493475,
+      "learning_rate": 1.667992037026455e-05,
+      "loss": 0.9658,
+      "step": 1795
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3787841634763032,
+      "learning_rate": 1.6676035303634366e-05,
+      "loss": 0.9085,
+      "step": 1796
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0827242408634352,
+      "learning_rate": 1.6672148418268143e-05,
+      "loss": 0.3393,
+      "step": 1797
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.457818621390555,
+      "learning_rate": 1.6668259715224767e-05,
+      "loss": 0.9083,
+      "step": 1798
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6112687007465818,
+      "learning_rate": 1.6664369195563635e-05,
+      "loss": 0.8646,
+      "step": 1799
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3988407342167224,
+      "learning_rate": 1.666047686034463e-05,
+      "loss": 0.8479,
+      "step": 1800
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3395026123014104,
+      "learning_rate": 1.6656582710628133e-05,
+      "loss": 0.8968,
+      "step": 1801
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7639819492570241,
+      "learning_rate": 1.665268674747501e-05,
+      "loss": 0.92,
+      "step": 1802
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6330661154637676,
+      "learning_rate": 1.664878897194664e-05,
+      "loss": 0.924,
+      "step": 1803
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.532684934724698,
+      "learning_rate": 1.6644889385104875e-05,
+      "loss": 0.9683,
+      "step": 1804
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4137379889089718,
+      "learning_rate": 1.6640987988012077e-05,
+      "loss": 0.9706,
+      "step": 1805
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5235202864625528,
+      "learning_rate": 1.663708478173109e-05,
+      "loss": 0.9588,
+      "step": 1806
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2330784779080506,
+      "learning_rate": 1.6633179767325258e-05,
+      "loss": 0.9011,
+      "step": 1807
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6098673221397422,
+      "learning_rate": 1.6629272945858413e-05,
+      "loss": 0.8826,
+      "step": 1808
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3560816714758392,
+      "learning_rate": 1.662536431839488e-05,
+      "loss": 0.9486,
+      "step": 1809
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.443297722961885,
+      "learning_rate": 1.6621453885999473e-05,
+      "loss": 0.9282,
+      "step": 1810
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.812590067369242,
+      "learning_rate": 1.6617541649737514e-05,
+      "loss": 0.3494,
+      "step": 1811
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.257451338776845,
+      "learning_rate": 1.661362761067479e-05,
+      "loss": 0.9186,
+      "step": 1812
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3689316292121216,
+      "learning_rate": 1.66097117698776e-05,
+      "loss": 0.9328,
+      "step": 1813
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2479203138370454,
+      "learning_rate": 1.6605794128412725e-05,
+      "loss": 0.8719,
+      "step": 1814
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.450862103151999,
+      "learning_rate": 1.6601874687347443e-05,
+      "loss": 0.8834,
+      "step": 1815
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7379808434210793,
+      "learning_rate": 1.6597953447749514e-05,
+      "loss": 0.8571,
+      "step": 1816
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.484289441654545,
+      "learning_rate": 1.659403041068719e-05,
+      "loss": 0.8145,
+      "step": 1817
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3237566141538748,
+      "learning_rate": 1.6590105577229216e-05,
+      "loss": 0.8666,
+      "step": 1818
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3234371492121928,
+      "learning_rate": 1.6586178948444828e-05,
+      "loss": 0.8594,
+      "step": 1819
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6171035312884443,
+      "learning_rate": 1.6582250525403748e-05,
+      "loss": 1.0314,
+      "step": 1820
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3116288007876589,
+      "learning_rate": 1.6578320309176175e-05,
+      "loss": 0.9505,
+      "step": 1821
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4404083353819983,
+      "learning_rate": 1.6574388300832827e-05,
+      "loss": 0.9358,
+      "step": 1822
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4140924574381648,
+      "learning_rate": 1.6570454501444878e-05,
+      "loss": 0.9739,
+      "step": 1823
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3663781024951993,
+      "learning_rate": 1.6566518912084e-05,
+      "loss": 0.8941,
+      "step": 1824
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.586553682341651,
+      "learning_rate": 1.6562581533822374e-05,
+      "loss": 0.8944,
+      "step": 1825
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2136709794298102,
+      "learning_rate": 1.6558642367732635e-05,
+      "loss": 0.8994,
+      "step": 1826
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3946488842382352,
+      "learning_rate": 1.6554701414887927e-05,
+      "loss": 0.9143,
+      "step": 1827
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7669819326937006,
+      "learning_rate": 1.6550758676361872e-05,
+      "loss": 0.9683,
+      "step": 1828
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5094245979500365,
+      "learning_rate": 1.6546814153228576e-05,
+      "loss": 0.9029,
+      "step": 1829
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.464662752413458,
+      "learning_rate": 1.6542867846562644e-05,
+      "loss": 0.9318,
+      "step": 1830
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4866145556006098,
+      "learning_rate": 1.653891975743916e-05,
+      "loss": 0.9326,
+      "step": 1831
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2524207026088447,
+      "learning_rate": 1.653496988693368e-05,
+      "loss": 0.8729,
+      "step": 1832
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4601656154701645,
+      "learning_rate": 1.653101823612227e-05,
+      "loss": 0.9251,
+      "step": 1833
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5345831443120987,
+      "learning_rate": 1.652706480608146e-05,
+      "loss": 0.9465,
+      "step": 1834
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3847654292882612,
+      "learning_rate": 1.652310959788828e-05,
+      "loss": 0.9148,
+      "step": 1835
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5969414738454522,
+      "learning_rate": 1.6519152612620236e-05,
+      "loss": 0.9995,
+      "step": 1836
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3906857096889536,
+      "learning_rate": 1.6515193851355314e-05,
+      "loss": 0.8599,
+      "step": 1837
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2850762316249045,
+      "learning_rate": 1.6511233315172e-05,
+      "loss": 0.8718,
+      "step": 1838
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.278397945351855,
+      "learning_rate": 1.6507271005149246e-05,
+      "loss": 0.8751,
+      "step": 1839
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5388183433366884,
+      "learning_rate": 1.6503306922366497e-05,
+      "loss": 0.8374,
+      "step": 1840
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6401831122625574,
+      "learning_rate": 1.6499341067903675e-05,
+      "loss": 0.9508,
+      "step": 1841
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2463090928773701,
+      "learning_rate": 1.6495373442841192e-05,
+      "loss": 0.9153,
+      "step": 1842
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5078476620067778,
+      "learning_rate": 1.649140404825994e-05,
+      "loss": 0.9205,
+      "step": 1843
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0778705756279008,
+      "learning_rate": 1.6487432885241287e-05,
+      "loss": 0.3591,
+      "step": 1844
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.422654447863389,
+      "learning_rate": 1.6483459954867086e-05,
+      "loss": 0.9037,
+      "step": 1845
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4647929971167872,
+      "learning_rate": 1.6479485258219677e-05,
+      "loss": 0.9241,
+      "step": 1846
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5676452815064608,
+      "learning_rate": 1.6475508796381875e-05,
+      "loss": 0.9066,
+      "step": 1847
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3706296769828599,
+      "learning_rate": 1.6471530570436983e-05,
+      "loss": 0.9679,
+      "step": 1848
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3047505723247015,
+      "learning_rate": 1.646755058146877e-05,
+      "loss": 0.846,
+      "step": 1849
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1552555747114357,
+      "learning_rate": 1.64635688305615e-05,
+      "loss": 0.3565,
+      "step": 1850
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5758281791556659,
+      "learning_rate": 1.6459585318799914e-05,
+      "loss": 0.8125,
+      "step": 1851
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4557076404517675,
+      "learning_rate": 1.6455600047269228e-05,
+      "loss": 0.888,
+      "step": 1852
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8531464666593211,
+      "learning_rate": 1.6451613017055136e-05,
+      "loss": 0.366,
+      "step": 1853
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.236892314965465,
+      "learning_rate": 1.644762422924382e-05,
+      "loss": 0.9813,
+      "step": 1854
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8437537352161382,
+      "learning_rate": 1.6443633684921934e-05,
+      "loss": 0.3252,
+      "step": 1855
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5837266561820837,
+      "learning_rate": 1.643964138517661e-05,
+      "loss": 0.9894,
+      "step": 1856
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4148796484051305,
+      "learning_rate": 1.6435647331095466e-05,
+      "loss": 0.9045,
+      "step": 1857
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2883450361044577,
+      "learning_rate": 1.643165152376659e-05,
+      "loss": 0.9638,
+      "step": 1858
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4272516180682604,
+      "learning_rate": 1.6427653964278547e-05,
+      "loss": 0.9388,
+      "step": 1859
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2096146687684552,
+      "learning_rate": 1.6423654653720385e-05,
+      "loss": 0.9594,
+      "step": 1860
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.489937051433043,
+      "learning_rate": 1.6419653593181628e-05,
+      "loss": 0.7749,
+      "step": 1861
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.355225912322674,
+      "learning_rate": 1.6415650783752274e-05,
+      "loss": 0.9044,
+      "step": 1862
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4914100891211741,
+      "learning_rate": 1.6411646226522793e-05,
+      "loss": 0.9411,
+      "step": 1863
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4348033731294667,
+      "learning_rate": 1.6407639922584148e-05,
+      "loss": 0.9198,
+      "step": 1864
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6303689002119117,
+      "learning_rate": 1.640363187302776e-05,
+      "loss": 0.9531,
+      "step": 1865
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4725151215731733,
+      "learning_rate": 1.639962207894553e-05,
+      "loss": 0.9768,
+      "step": 1866
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.468409529243215,
+      "learning_rate": 1.6395610541429836e-05,
+      "loss": 0.8686,
+      "step": 1867
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3630561403657215,
+      "learning_rate": 1.639159726157354e-05,
+      "loss": 0.8506,
+      "step": 1868
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5142797676875221,
+      "learning_rate": 1.6387582240469963e-05,
+      "loss": 0.9296,
+      "step": 1869
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.298882509111199,
+      "learning_rate": 1.638356547921291e-05,
+      "loss": 0.914,
+      "step": 1870
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8269400359416621,
+      "learning_rate": 1.6379546978896655e-05,
+      "loss": 0.3392,
+      "step": 1871
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5039602492905442,
+      "learning_rate": 1.637552674061595e-05,
+      "loss": 0.8839,
+      "step": 1872
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8194078554819518,
+      "learning_rate": 1.6371504765466014e-05,
+      "loss": 0.2919,
+      "step": 1873
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.499692720862973,
+      "learning_rate": 1.6367481054542556e-05,
+      "loss": 0.9457,
+      "step": 1874
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4548253787151557,
+      "learning_rate": 1.6363455608941734e-05,
+      "loss": 0.9164,
+      "step": 1875
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.277285653653831,
+      "learning_rate": 1.6359428429760192e-05,
+      "loss": 0.9211,
+      "step": 1876
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4171023572477432,
+      "learning_rate": 1.635539951809505e-05,
+      "loss": 0.9129,
+      "step": 1877
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3930811201001108,
+      "learning_rate": 1.6351368875043893e-05,
+      "loss": 0.8563,
+      "step": 1878
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4159797104852054,
+      "learning_rate": 1.6347336501704777e-05,
+      "loss": 0.9508,
+      "step": 1879
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4981195296681808,
+      "learning_rate": 1.6343302399176235e-05,
+      "loss": 0.9181,
+      "step": 1880
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.842047480068682,
+      "learning_rate": 1.633926656855726e-05,
+      "loss": 0.3538,
+      "step": 1881
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4269311709507366,
+      "learning_rate": 1.6335229010947333e-05,
+      "loss": 0.9183,
+      "step": 1882
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4151447097429506,
+      "learning_rate": 1.6331189727446393e-05,
+      "loss": 0.9394,
+      "step": 1883
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4266576503345458,
+      "learning_rate": 1.6327148719154845e-05,
+      "loss": 0.9376,
+      "step": 1884
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3298951960062875,
+      "learning_rate": 1.632310598717358e-05,
+      "loss": 0.9058,
+      "step": 1885
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5969633252954751,
+      "learning_rate": 1.6319061532603947e-05,
+      "loss": 0.8349,
+      "step": 1886
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.517028094613631,
+      "learning_rate": 1.631501535654777e-05,
+      "loss": 0.8948,
+      "step": 1887
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2892759789358808,
+      "learning_rate": 1.6310967460107328e-05,
+      "loss": 0.9436,
+      "step": 1888
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.637203354209914,
+      "learning_rate": 1.630691784438539e-05,
+      "loss": 0.8824,
+      "step": 1889
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0759367610257575,
+      "learning_rate": 1.630286651048518e-05,
+      "loss": 0.3086,
+      "step": 1890
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5267048253037414,
+      "learning_rate": 1.6298813459510396e-05,
+      "loss": 1.0233,
+      "step": 1891
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2437257157224155,
+      "learning_rate": 1.6294758692565197e-05,
+      "loss": 0.959,
+      "step": 1892
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6027426872158737,
+      "learning_rate": 1.6290702210754213e-05,
+      "loss": 0.9224,
+      "step": 1893
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4511738580764633,
+      "learning_rate": 1.628664401518254e-05,
+      "loss": 0.8946,
+      "step": 1894
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.245691753286502,
+      "learning_rate": 1.6282584106955747e-05,
+      "loss": 0.8767,
+      "step": 1895
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.442912440344975,
+      "learning_rate": 1.6278522487179867e-05,
+      "loss": 0.9146,
+      "step": 1896
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5031029190555176,
+      "learning_rate": 1.6274459156961388e-05,
+      "loss": 0.8824,
+      "step": 1897
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3614520676014332,
+      "learning_rate": 1.6270394117407277e-05,
+      "loss": 0.8791,
+      "step": 1898
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.279253632054858,
+      "learning_rate": 1.626632736962497e-05,
+      "loss": 0.9005,
+      "step": 1899
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4344592431206322,
+      "learning_rate": 1.6262258914722352e-05,
+      "loss": 0.9292,
+      "step": 1900
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6104961160353812,
+      "learning_rate": 1.6258188753807783e-05,
+      "loss": 0.916,
+      "step": 1901
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3992711005489513,
+      "learning_rate": 1.625411688799009e-05,
+      "loss": 0.873,
+      "step": 1902
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5572392235306647,
+      "learning_rate": 1.6250043318378563e-05,
+      "loss": 0.8436,
+      "step": 1903
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.313176324273467,
+      "learning_rate": 1.6245968046082952e-05,
+      "loss": 0.8471,
+      "step": 1904
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1381759790585577,
+      "learning_rate": 1.6241891072213475e-05,
+      "loss": 0.3717,
+      "step": 1905
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3214802402811656,
+      "learning_rate": 1.6237812397880806e-05,
+      "loss": 0.9464,
+      "step": 1906
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.763875373979312,
+      "learning_rate": 1.62337320241961e-05,
+      "loss": 0.8829,
+      "step": 1907
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8674387589304035,
+      "learning_rate": 1.6229649952270955e-05,
+      "loss": 0.3582,
+      "step": 1908
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3696017114723968,
+      "learning_rate": 1.622556618321744e-05,
+      "loss": 0.8835,
+      "step": 1909
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.456235674170605,
+      "learning_rate": 1.622148071814809e-05,
+      "loss": 1.0319,
+      "step": 1910
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2798709935201869,
+      "learning_rate": 1.6217393558175897e-05,
+      "loss": 0.8944,
+      "step": 1911
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.393940428120887,
+      "learning_rate": 1.6213304704414314e-05,
+      "loss": 0.9134,
+      "step": 1912
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4409895238945154,
+      "learning_rate": 1.620921415797726e-05,
+      "loss": 0.766,
+      "step": 1913
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6585663289863886,
+      "learning_rate": 1.620512191997911e-05,
+      "loss": 0.8966,
+      "step": 1914
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2650912662057499,
+      "learning_rate": 1.6201027991534705e-05,
+      "loss": 0.8632,
+      "step": 1915
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4393435136270214,
+      "learning_rate": 1.6196932373759337e-05,
+      "loss": 0.9465,
+      "step": 1916
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.334636792244241,
+      "learning_rate": 1.6192835067768776e-05,
+      "loss": 0.8987,
+      "step": 1917
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2041960116127077,
+      "learning_rate": 1.6188736074679237e-05,
+      "loss": 0.8552,
+      "step": 1918
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8990943559996785,
+      "learning_rate": 1.6184635395607395e-05,
+      "loss": 0.3548,
+      "step": 1919
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.1427925275564728,
+      "learning_rate": 1.6180533031670395e-05,
+      "loss": 0.9072,
+      "step": 1920
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4672609933396465,
+      "learning_rate": 1.6176428983985825e-05,
+      "loss": 0.9374,
+      "step": 1921
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4538621301839003,
+      "learning_rate": 1.6172323253671745e-05,
+      "loss": 0.8838,
+      "step": 1922
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5457096936498687,
+      "learning_rate": 1.6168215841846673e-05,
+      "loss": 0.9056,
+      "step": 1923
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3066342115601206,
+      "learning_rate": 1.616410674962958e-05,
+      "loss": 0.923,
+      "step": 1924
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3231451878766702,
+      "learning_rate": 1.615999597813989e-05,
+      "loss": 0.9614,
+      "step": 1925
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4278456217976934,
+      "learning_rate": 1.6155883528497492e-05,
+      "loss": 0.8681,
+      "step": 1926
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4635257055446242,
+      "learning_rate": 1.6151769401822735e-05,
+      "loss": 0.9694,
+      "step": 1927
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4383715502548073,
+      "learning_rate": 1.614765359923642e-05,
+      "loss": 0.8516,
+      "step": 1928
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4921943030277178,
+      "learning_rate": 1.6143536121859805e-05,
+      "loss": 0.8864,
+      "step": 1929
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3852743255540962,
+      "learning_rate": 1.61394169708146e-05,
+      "loss": 0.8516,
+      "step": 1930
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.460610669980455,
+      "learning_rate": 1.613529614722298e-05,
+      "loss": 0.8586,
+      "step": 1931
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2469697667531376,
+      "learning_rate": 1.6131173652207565e-05,
+      "loss": 0.8835,
+      "step": 1932
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.551797209732303,
+      "learning_rate": 1.6127049486891442e-05,
+      "loss": 0.8796,
+      "step": 1933
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4913047406285018,
+      "learning_rate": 1.612292365239815e-05,
+      "loss": 0.9068,
+      "step": 1934
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4759843441186518,
+      "learning_rate": 1.611879614985167e-05,
+      "loss": 0.9129,
+      "step": 1935
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7478676805340232,
+      "learning_rate": 1.6114666980376455e-05,
+      "loss": 0.9263,
+      "step": 1936
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3796207127763283,
+      "learning_rate": 1.6110536145097407e-05,
+      "loss": 0.8591,
+      "step": 1937
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4062992464820359,
+      "learning_rate": 1.6106403645139866e-05,
+      "loss": 0.8777,
+      "step": 1938
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3156829192342445,
+      "learning_rate": 1.6102269481629654e-05,
+      "loss": 0.8738,
+      "step": 1939
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.524495203912162,
+      "learning_rate": 1.6098133655693027e-05,
+      "loss": 0.9024,
+      "step": 1940
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.371689653367193,
+      "learning_rate": 1.6093996168456694e-05,
+      "loss": 0.9143,
+      "step": 1941
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2592867894907427,
+      "learning_rate": 1.6089857021047822e-05,
+      "loss": 0.8772,
+      "step": 1942
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5089332826443895,
+      "learning_rate": 1.608571621459403e-05,
+      "loss": 0.914,
+      "step": 1943
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6946082399306537,
+      "learning_rate": 1.6081573750223388e-05,
+      "loss": 0.9012,
+      "step": 1944
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4515750916543633,
+      "learning_rate": 1.607742962906442e-05,
+      "loss": 0.915,
+      "step": 1945
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4306076097254299,
+      "learning_rate": 1.6073283852246087e-05,
+      "loss": 0.9183,
+      "step": 1946
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2846640967756937,
+      "learning_rate": 1.6069136420897827e-05,
+      "loss": 0.8739,
+      "step": 1947
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3388333354596353,
+      "learning_rate": 1.606498733614951e-05,
+      "loss": 0.93,
+      "step": 1948
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4542791790978111,
+      "learning_rate": 1.606083659913146e-05,
+      "loss": 0.9383,
+      "step": 1949
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.257295988562237,
+      "learning_rate": 1.605668421097445e-05,
+      "loss": 0.9528,
+      "step": 1950
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6888426714181268,
+      "learning_rate": 1.605253017280971e-05,
+      "loss": 0.9161,
+      "step": 1951
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3883795970516037,
+      "learning_rate": 1.6048374485768912e-05,
+      "loss": 0.9688,
+      "step": 1952
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5040407064946615,
+      "learning_rate": 1.604421715098418e-05,
+      "loss": 0.9332,
+      "step": 1953
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.346406664153787,
+      "learning_rate": 1.6040058169588086e-05,
+      "loss": 0.9112,
+      "step": 1954
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6776534994702716,
+      "learning_rate": 1.603589754271365e-05,
+      "loss": 0.9081,
+      "step": 1955
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.330353867940767,
+      "learning_rate": 1.603173527149434e-05,
+      "loss": 0.8344,
+      "step": 1956
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2980454377577682,
+      "learning_rate": 1.602757135706408e-05,
+      "loss": 0.8686,
+      "step": 1957
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4192877721796435,
+      "learning_rate": 1.602340580055723e-05,
+      "loss": 0.9823,
+      "step": 1958
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3824046780890138,
+      "learning_rate": 1.6019238603108605e-05,
+      "loss": 0.8438,
+      "step": 1959
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5052048253862187,
+      "learning_rate": 1.6015069765853462e-05,
+      "loss": 0.8439,
+      "step": 1960
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7751854084455048,
+      "learning_rate": 1.6010899289927513e-05,
+      "loss": 0.9124,
+      "step": 1961
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6650943998610765,
+      "learning_rate": 1.60067271764669e-05,
+      "loss": 0.8793,
+      "step": 1962
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.224744920058577,
+      "learning_rate": 1.600255342660823e-05,
+      "loss": 0.8953,
+      "step": 1963
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7133515574027438,
+      "learning_rate": 1.5998378041488547e-05,
+      "loss": 0.9532,
+      "step": 1964
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2283299082852681,
+      "learning_rate": 1.5994201022245338e-05,
+      "loss": 0.8681,
+      "step": 1965
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1344981198319442,
+      "learning_rate": 1.599002237001654e-05,
+      "loss": 0.9084,
+      "step": 1966
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3177238794130328,
+      "learning_rate": 1.598584208594053e-05,
+      "loss": 0.8337,
+      "step": 1967
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.397189333184374,
+      "learning_rate": 1.5981660171156136e-05,
+      "loss": 0.8724,
+      "step": 1968
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3013798400330248,
+      "learning_rate": 1.5977476626802624e-05,
+      "loss": 0.9118,
+      "step": 1969
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2816872980786425,
+      "learning_rate": 1.5973291454019713e-05,
+      "loss": 0.9026,
+      "step": 1970
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.388597238254469,
+      "learning_rate": 1.5969104653947552e-05,
+      "loss": 0.9196,
+      "step": 1971
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4010803068574738,
+      "learning_rate": 1.5964916227726747e-05,
+      "loss": 0.8973,
+      "step": 1972
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.320947376654923,
+      "learning_rate": 1.5960726176498334e-05,
+      "loss": 0.8945,
+      "step": 1973
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3197637685973724,
+      "learning_rate": 1.5956534501403808e-05,
+      "loss": 0.8162,
+      "step": 1974
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1748864370586247,
+      "learning_rate": 1.5952341203585086e-05,
+      "loss": 0.9614,
+      "step": 1975
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.427350448593724,
+      "learning_rate": 1.594814628418454e-05,
+      "loss": 0.9338,
+      "step": 1976
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4361101187369865,
+      "learning_rate": 1.5943949744344994e-05,
+      "loss": 0.8075,
+      "step": 1977
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.135290864732535,
+      "learning_rate": 1.5939751585209685e-05,
+      "loss": 0.8992,
+      "step": 1978
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4497168782491758,
+      "learning_rate": 1.5935551807922315e-05,
+      "loss": 0.8677,
+      "step": 1979
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3081699111973126,
+      "learning_rate": 1.593135041362702e-05,
+      "loss": 0.8933,
+      "step": 1980
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6592582405629155,
+      "learning_rate": 1.592714740346837e-05,
+      "loss": 0.8719,
+      "step": 1981
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4928880914272047,
+      "learning_rate": 1.592294277859139e-05,
+      "loss": 0.8737,
+      "step": 1982
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6162576501555364,
+      "learning_rate": 1.5918736540141525e-05,
+      "loss": 0.917,
+      "step": 1983
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5205918670432066,
+      "learning_rate": 1.5914528689264677e-05,
+      "loss": 0.8819,
+      "step": 1984
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.209985270959786,
+      "learning_rate": 1.591031922710718e-05,
+      "loss": 0.8604,
+      "step": 1985
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3971386917149347,
+      "learning_rate": 1.5906108154815805e-05,
+      "loss": 0.9026,
+      "step": 1986
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4232462202789518,
+      "learning_rate": 1.5901895473537768e-05,
+      "loss": 0.8754,
+      "step": 1987
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3081648116482805,
+      "learning_rate": 1.5897681184420716e-05,
+      "loss": 0.9134,
+      "step": 1988
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4026718921257686,
+      "learning_rate": 1.589346528861274e-05,
+      "loss": 0.8925,
+      "step": 1989
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4225940104700663,
+      "learning_rate": 1.588924778726236e-05,
+      "loss": 0.904,
+      "step": 1990
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3410152465859781,
+      "learning_rate": 1.588502868151855e-05,
+      "loss": 0.9033,
+      "step": 1991
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8958031573961167,
+      "learning_rate": 1.5880807972530705e-05,
+      "loss": 0.3577,
+      "step": 1992
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.531633992486013,
+      "learning_rate": 1.587658566144866e-05,
+      "loss": 0.8827,
+      "step": 1993
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.375815286653979,
+      "learning_rate": 1.5872361749422694e-05,
+      "loss": 0.9018,
+      "step": 1994
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8890854995027901,
+      "learning_rate": 1.5868136237603516e-05,
+      "loss": 0.9174,
+      "step": 1995
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3732840450533814,
+      "learning_rate": 1.5863909127142268e-05,
+      "loss": 0.9104,
+      "step": 1996
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4159055114551702,
+      "learning_rate": 1.585968041919054e-05,
+      "loss": 0.9233,
+      "step": 1997
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4254468532238582,
+      "learning_rate": 1.585545011490034e-05,
+      "loss": 0.8491,
+      "step": 1998
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3744619905952045,
+      "learning_rate": 1.5851218215424115e-05,
+      "loss": 0.7947,
+      "step": 1999
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2951859119407894,
+      "learning_rate": 1.5846984721914765e-05,
+      "loss": 0.9306,
+      "step": 2000
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2681900657433085,
+      "learning_rate": 1.5842749635525602e-05,
+      "loss": 0.9174,
+      "step": 2001
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4291576663572392,
+      "learning_rate": 1.5838512957410384e-05,
+      "loss": 0.9311,
+      "step": 2002
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.618749993314614,
+      "learning_rate": 1.5834274688723293e-05,
+      "loss": 0.8007,
+      "step": 2003
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.349696689271039,
+      "learning_rate": 1.583003483061896e-05,
+      "loss": 0.9781,
+      "step": 2004
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5430955086319913,
+      "learning_rate": 1.5825793384252432e-05,
+      "loss": 0.93,
+      "step": 2005
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3214809647712604,
+      "learning_rate": 1.582155035077919e-05,
+      "loss": 0.8707,
+      "step": 2006
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3105344398664698,
+      "learning_rate": 1.5817305731355168e-05,
+      "loss": 0.96,
+      "step": 2007
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.526164995153077,
+      "learning_rate": 1.5813059527136708e-05,
+      "loss": 0.8891,
+      "step": 2008
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4207678017094645,
+      "learning_rate": 1.5808811739280592e-05,
+      "loss": 0.8665,
+      "step": 2009
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6024967056664947,
+      "learning_rate": 1.5804562368944042e-05,
+      "loss": 0.9544,
+      "step": 2010
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.300661571971281,
+      "learning_rate": 1.5800311417284695e-05,
+      "loss": 0.9215,
+      "step": 2011
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5023343404831306,
+      "learning_rate": 1.579605888546063e-05,
+      "loss": 0.976,
+      "step": 2012
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.366843390756666,
+      "learning_rate": 1.579180477463036e-05,
+      "loss": 0.9197,
+      "step": 2013
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.637873498302596,
+      "learning_rate": 1.5787549085952816e-05,
+      "loss": 0.9639,
+      "step": 2014
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2562467710254221,
+      "learning_rate": 1.5783291820587366e-05,
+      "loss": 0.889,
+      "step": 2015
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4917293164702217,
+      "learning_rate": 1.5779032979693808e-05,
+      "loss": 0.9411,
+      "step": 2016
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3295242648530172,
+      "learning_rate": 1.5774772564432365e-05,
+      "loss": 0.8543,
+      "step": 2017
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3157751964905036,
+      "learning_rate": 1.577051057596369e-05,
+      "loss": 0.792,
+      "step": 2018
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1483011026868462,
+      "learning_rate": 1.5766247015448875e-05,
+      "loss": 0.9328,
+      "step": 2019
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9090602234923024,
+      "learning_rate": 1.576198188404942e-05,
+      "loss": 0.3851,
+      "step": 2020
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.417335480499871,
+      "learning_rate": 1.5757715182927273e-05,
+      "loss": 0.8626,
+      "step": 2021
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3146688523759742,
+      "learning_rate": 1.5753446913244798e-05,
+      "loss": 0.8823,
+      "step": 2022
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4287771501666913,
+      "learning_rate": 1.5749177076164788e-05,
+      "loss": 0.8827,
+      "step": 2023
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4310137266084093,
+      "learning_rate": 1.5744905672850467e-05,
+      "loss": 0.875,
+      "step": 2024
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.315271893080312,
+      "learning_rate": 1.5740632704465478e-05,
+      "loss": 0.853,
+      "step": 2025
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9589015709291007,
+      "learning_rate": 1.5736358172173903e-05,
+      "loss": 0.9599,
+      "step": 2026
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.461253692329211,
+      "learning_rate": 1.5732082077140235e-05,
+      "loss": 0.9402,
+      "step": 2027
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8454234299989747,
+      "learning_rate": 1.5727804420529405e-05,
+      "loss": 0.3236,
+      "step": 2028
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5518179378300554,
+      "learning_rate": 1.5723525203506758e-05,
+      "loss": 0.8517,
+      "step": 2029
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4448462481686997,
+      "learning_rate": 1.5719244427238086e-05,
+      "loss": 0.8893,
+      "step": 2030
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4342007410924213,
+      "learning_rate": 1.571496209288957e-05,
+      "loss": 0.895,
+      "step": 2031
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3268758677759624,
+      "learning_rate": 1.571067820162785e-05,
+      "loss": 0.868,
+      "step": 2032
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3988853478704937,
+      "learning_rate": 1.5706392754619973e-05,
+      "loss": 0.9087,
+      "step": 2033
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.379079422194281,
+      "learning_rate": 1.5702105753033415e-05,
+      "loss": 0.8396,
+      "step": 2034
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2855426445157951,
+      "learning_rate": 1.5697817198036066e-05,
+      "loss": 0.8341,
+      "step": 2035
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3832603203605502,
+      "learning_rate": 1.569352709079625e-05,
+      "loss": 0.8867,
+      "step": 2036
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3716137618580329,
+      "learning_rate": 1.5689235432482715e-05,
+      "loss": 0.8929,
+      "step": 2037
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.83297503466964,
+      "learning_rate": 1.5684942224264622e-05,
+      "loss": 0.8686,
+      "step": 2038
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5586894365686241,
+      "learning_rate": 1.568064746731156e-05,
+      "loss": 0.9573,
+      "step": 2039
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5354239397855813,
+      "learning_rate": 1.567635116279354e-05,
+      "loss": 0.8918,
+      "step": 2040
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.64151972204166,
+      "learning_rate": 1.5672053311880994e-05,
+      "loss": 0.9438,
+      "step": 2041
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3953362713846793,
+      "learning_rate": 1.5667753915744776e-05,
+      "loss": 0.938,
+      "step": 2042
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3405617411122808,
+      "learning_rate": 1.566345297555616e-05,
+      "loss": 0.811,
+      "step": 2043
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.359147699579635,
+      "learning_rate": 1.5659150492486833e-05,
+      "loss": 0.8493,
+      "step": 2044
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.620516542717152,
+      "learning_rate": 1.565484646770892e-05,
+      "loss": 0.795,
+      "step": 2045
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4572061792913213,
+      "learning_rate": 1.5650540902394954e-05,
+      "loss": 0.8851,
+      "step": 2046
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8338082003450566,
+      "learning_rate": 1.564623379771789e-05,
+      "loss": 0.9402,
+      "step": 2047
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5077881514604323,
+      "learning_rate": 1.5641925154851096e-05,
+      "loss": 0.8017,
+      "step": 2048
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5976108419825408,
+      "learning_rate": 1.5637614974968372e-05,
+      "loss": 0.9932,
+      "step": 2049
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3321729798580437,
+      "learning_rate": 1.5633303259243927e-05,
+      "loss": 0.9171,
+      "step": 2050
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.123374686289018,
+      "learning_rate": 1.562899000885239e-05,
+      "loss": 0.8792,
+      "step": 2051
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6622765319182398,
+      "learning_rate": 1.5624675224968808e-05,
+      "loss": 0.8477,
+      "step": 2052
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.386391314278079,
+      "learning_rate": 1.562035890876865e-05,
+      "loss": 0.867,
+      "step": 2053
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3491757996520053,
+      "learning_rate": 1.5616041061427805e-05,
+      "loss": 0.8796,
+      "step": 2054
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.496533532780325,
+      "learning_rate": 1.5611721684122564e-05,
+      "loss": 0.8806,
+      "step": 2055
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5834088320571502,
+      "learning_rate": 1.5607400778029648e-05,
+      "loss": 0.8817,
+      "step": 2056
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3630594247955332,
+      "learning_rate": 1.5603078344326194e-05,
+      "loss": 0.8974,
+      "step": 2057
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3372602362896633,
+      "learning_rate": 1.5598754384189745e-05,
+      "loss": 0.9163,
+      "step": 2058
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4580252199864865,
+      "learning_rate": 1.5594428898798272e-05,
+      "loss": 0.9119,
+      "step": 2059
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4017634001141213,
+      "learning_rate": 1.559010188933016e-05,
+      "loss": 0.8673,
+      "step": 2060
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6540001126748292,
+      "learning_rate": 1.5585773356964192e-05,
+      "loss": 0.876,
+      "step": 2061
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3125592754807214,
+      "learning_rate": 1.55814433028796e-05,
+      "loss": 0.866,
+      "step": 2062
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4900815461663042,
+      "learning_rate": 1.557711172825599e-05,
+      "loss": 0.8275,
+      "step": 2063
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3004657746564787,
+      "learning_rate": 1.5572778634273417e-05,
+      "loss": 0.8604,
+      "step": 2064
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3775636131276412,
+      "learning_rate": 1.556844402211233e-05,
+      "loss": 0.9643,
+      "step": 2065
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4253117676737015,
+      "learning_rate": 1.556410789295359e-05,
+      "loss": 0.9308,
+      "step": 2066
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4748849308158234,
+      "learning_rate": 1.5559770247978488e-05,
+      "loss": 0.8501,
+      "step": 2067
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5464628867544787,
+      "learning_rate": 1.5555431088368716e-05,
+      "loss": 0.9305,
+      "step": 2068
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4544607798623925,
+      "learning_rate": 1.5551090415306377e-05,
+      "loss": 0.9296,
+      "step": 2069
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2954031861049777,
+      "learning_rate": 1.554674822997399e-05,
+      "loss": 0.8785,
+      "step": 2070
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4661568230617419,
+      "learning_rate": 1.5542404533554493e-05,
+      "loss": 0.9481,
+      "step": 2071
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3282663683825553,
+      "learning_rate": 1.553805932723122e-05,
+      "loss": 0.8221,
+      "step": 2072
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6164971504445893,
+      "learning_rate": 1.5533712612187933e-05,
+      "loss": 0.892,
+      "step": 2073
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3477981043530414,
+      "learning_rate": 1.5529364389608788e-05,
+      "loss": 0.8857,
+      "step": 2074
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2017522890499357,
+      "learning_rate": 1.552501466067837e-05,
+      "loss": 0.8632,
+      "step": 2075
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3419983933224382,
+      "learning_rate": 1.552066342658166e-05,
+      "loss": 0.871,
+      "step": 2076
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3984443235497255,
+      "learning_rate": 1.551631068850405e-05,
+      "loss": 0.8985,
+      "step": 2077
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3442374526304328,
+      "learning_rate": 1.5511956447631355e-05,
+      "loss": 0.899,
+      "step": 2078
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4140398118533117,
+      "learning_rate": 1.5507600705149782e-05,
+      "loss": 0.838,
+      "step": 2079
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.40675016819898,
+      "learning_rate": 1.5503243462245963e-05,
+      "loss": 0.7899,
+      "step": 2080
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4594952960282812,
+      "learning_rate": 1.5498884720106925e-05,
+      "loss": 0.8555,
+      "step": 2081
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3959283942411733,
+      "learning_rate": 1.549452447992011e-05,
+      "loss": 0.9066,
+      "step": 2082
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3857328335296446,
+      "learning_rate": 1.5490162742873372e-05,
+      "loss": 0.8954,
+      "step": 2083
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.498538007015675,
+      "learning_rate": 1.5485799510154965e-05,
+      "loss": 0.9087,
+      "step": 2084
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4580167360026324,
+      "learning_rate": 1.5481434782953557e-05,
+      "loss": 0.8489,
+      "step": 2085
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8358803422050336,
+      "learning_rate": 1.5477068562458212e-05,
+      "loss": 0.3442,
+      "step": 2086
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.482331657051731,
+      "learning_rate": 1.5472700849858417e-05,
+      "loss": 0.8924,
+      "step": 2087
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3434217899117815,
+      "learning_rate": 1.5468331646344056e-05,
+      "loss": 0.9145,
+      "step": 2088
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3673883474505215,
+      "learning_rate": 1.5463960953105416e-05,
+      "loss": 0.8984,
+      "step": 2089
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3328144128814778,
+      "learning_rate": 1.54595887713332e-05,
+      "loss": 0.8554,
+      "step": 2090
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8019379037164139,
+      "learning_rate": 1.5455215102218505e-05,
+      "loss": 0.3203,
+      "step": 2091
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2745206079486524,
+      "learning_rate": 1.5450839946952845e-05,
+      "loss": 0.918,
+      "step": 2092
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7349121136265553,
+      "learning_rate": 1.544646330672813e-05,
+      "loss": 0.8086,
+      "step": 2093
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3974543204107575,
+      "learning_rate": 1.544208518273668e-05,
+      "loss": 0.9076,
+      "step": 2094
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.388588513851327,
+      "learning_rate": 1.5437705576171208e-05,
+      "loss": 0.9572,
+      "step": 2095
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3782447857493447,
+      "learning_rate": 1.543332448822485e-05,
+      "loss": 0.881,
+      "step": 2096
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.206195458602207,
+      "learning_rate": 1.542894192009113e-05,
+      "loss": 0.8796,
+      "step": 2097
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8772987717871875,
+      "learning_rate": 1.5424557872963984e-05,
+      "loss": 0.3326,
+      "step": 2098
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5106064008610238,
+      "learning_rate": 1.5420172348037742e-05,
+      "loss": 0.8839,
+      "step": 2099
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3602666095789804,
+      "learning_rate": 1.5415785346507143e-05,
+      "loss": 0.8538,
+      "step": 2100
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.655156785798686,
+      "learning_rate": 1.5411396869567332e-05,
+      "loss": 0.8745,
+      "step": 2101
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2886400446280795,
+      "learning_rate": 1.5407006918413843e-05,
+      "loss": 0.9232,
+      "step": 2102
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3390862875503675,
+      "learning_rate": 1.540261549424263e-05,
+      "loss": 0.8931,
+      "step": 2103
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.323889844833962,
+      "learning_rate": 1.539822259825003e-05,
+      "loss": 0.9047,
+      "step": 2104
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5010173639424311,
+      "learning_rate": 1.539382823163279e-05,
+      "loss": 0.9984,
+      "step": 2105
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5083787773814015,
+      "learning_rate": 1.538943239558806e-05,
+      "loss": 0.8741,
+      "step": 2106
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3395065711909675,
+      "learning_rate": 1.5385035091313382e-05,
+      "loss": 0.8379,
+      "step": 2107
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4325983485703977,
+      "learning_rate": 1.538063632000671e-05,
+      "loss": 0.9036,
+      "step": 2108
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4259969260769814,
+      "learning_rate": 1.5376236082866384e-05,
+      "loss": 0.8324,
+      "step": 2109
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.424697088188185,
+      "learning_rate": 1.5371834381091152e-05,
+      "loss": 0.8851,
+      "step": 2110
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4600805597194675,
+      "learning_rate": 1.5367431215880156e-05,
+      "loss": 0.9204,
+      "step": 2111
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.582653588464009,
+      "learning_rate": 1.536302658843295e-05,
+      "loss": 0.9067,
+      "step": 2112
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4688398563260845,
+      "learning_rate": 1.5358620499949464e-05,
+      "loss": 0.8702,
+      "step": 2113
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5876682871882286,
+      "learning_rate": 1.5354212951630043e-05,
+      "loss": 0.9259,
+      "step": 2114
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5684207245714736,
+      "learning_rate": 1.5349803944675424e-05,
+      "loss": 0.9044,
+      "step": 2115
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2026772594859434,
+      "learning_rate": 1.5345393480286744e-05,
+      "loss": 0.8738,
+      "step": 2116
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4195604887655733,
+      "learning_rate": 1.5340981559665533e-05,
+      "loss": 0.9376,
+      "step": 2117
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7199205043335877,
+      "learning_rate": 1.5336568184013717e-05,
+      "loss": 0.9292,
+      "step": 2118
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.323290673384635,
+      "learning_rate": 1.5332153354533626e-05,
+      "loss": 0.8559,
+      "step": 2119
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3758325301729055,
+      "learning_rate": 1.532773707242798e-05,
+      "loss": 0.9043,
+      "step": 2120
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3444443266614672,
+      "learning_rate": 1.5323319338899896e-05,
+      "loss": 0.8863,
+      "step": 2121
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5512818974478426,
+      "learning_rate": 1.531890015515289e-05,
+      "loss": 0.8728,
+      "step": 2122
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.323649199554473,
+      "learning_rate": 1.5314479522390856e-05,
+      "loss": 0.9153,
+      "step": 2123
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3178959215559662,
+      "learning_rate": 1.5310057441818115e-05,
+      "loss": 0.858,
+      "step": 2124
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4805177275374284,
+      "learning_rate": 1.5305633914639348e-05,
+      "loss": 0.9368,
+      "step": 2125
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3399720380773832,
+      "learning_rate": 1.530120894205965e-05,
+      "loss": 0.8256,
+      "step": 2126
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.403337365048835,
+      "learning_rate": 1.5296782525284514e-05,
+      "loss": 0.8855,
+      "step": 2127
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4219346361439533,
+      "learning_rate": 1.529235466551981e-05,
+      "loss": 0.8182,
+      "step": 2128
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5538172892263111,
+      "learning_rate": 1.5287925363971807e-05,
+      "loss": 0.958,
+      "step": 2129
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3767689629085758,
+      "learning_rate": 1.5283494621847175e-05,
+      "loss": 0.8653,
+      "step": 2130
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.38441744672304,
+      "learning_rate": 1.5279062440352968e-05,
+      "loss": 0.8887,
+      "step": 2131
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5654703847101017,
+      "learning_rate": 1.527462882069663e-05,
+      "loss": 0.8576,
+      "step": 2132
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3513552714943207,
+      "learning_rate": 1.5270193764086012e-05,
+      "loss": 0.9431,
+      "step": 2133
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2289987316939581,
+      "learning_rate": 1.5265757271729333e-05,
+      "loss": 0.906,
+      "step": 2134
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3670286481398486,
+      "learning_rate": 1.5261319344835225e-05,
+      "loss": 0.8459,
+      "step": 2135
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4874394819727386,
+      "learning_rate": 1.5256879984612698e-05,
+      "loss": 0.9315,
+      "step": 2136
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6014259978018532,
+      "learning_rate": 1.5252439192271156e-05,
+      "loss": 0.8426,
+      "step": 2137
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.438900216382297,
+      "learning_rate": 1.5247996969020394e-05,
+      "loss": 0.8941,
+      "step": 2138
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3001680503945106,
+      "learning_rate": 1.5243553316070596e-05,
+      "loss": 0.9218,
+      "step": 2139
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3216152292556937,
+      "learning_rate": 1.523910823463233e-05,
+      "loss": 0.911,
+      "step": 2140
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.42414910619103,
+      "learning_rate": 1.5234661725916573e-05,
+      "loss": 0.9348,
+      "step": 2141
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3936181311276843,
+      "learning_rate": 1.5230213791134662e-05,
+      "loss": 0.9034,
+      "step": 2142
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3051830574115397,
+      "learning_rate": 1.5225764431498344e-05,
+      "loss": 0.8586,
+      "step": 2143
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.462065254425266,
+      "learning_rate": 1.5221313648219749e-05,
+      "loss": 0.9315,
+      "step": 2144
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.255502092313941,
+      "learning_rate": 1.5216861442511382e-05,
+      "loss": 0.8246,
+      "step": 2145
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3723939520722246,
+      "learning_rate": 1.5212407815586162e-05,
+      "loss": 0.8995,
+      "step": 2146
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2710295646032514,
+      "learning_rate": 1.5207952768657368e-05,
+      "loss": 0.8931,
+      "step": 2147
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9181240153388553,
+      "learning_rate": 1.5203496302938682e-05,
+      "loss": 0.3409,
+      "step": 2148
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5664433875640127,
+      "learning_rate": 1.519903841964417e-05,
+      "loss": 0.885,
+      "step": 2149
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.441685924043669,
+      "learning_rate": 1.519457911998828e-05,
+      "loss": 0.8769,
+      "step": 2150
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.435474668360641,
+      "learning_rate": 1.5190118405185845e-05,
+      "loss": 0.8977,
+      "step": 2151
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1582444251562716,
+      "learning_rate": 1.5185656276452095e-05,
+      "loss": 0.8118,
+      "step": 2152
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6594681368388973,
+      "learning_rate": 1.5181192735002628e-05,
+      "loss": 0.9349,
+      "step": 2153
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3490758688792044,
+      "learning_rate": 1.517672778205344e-05,
+      "loss": 0.848,
+      "step": 2154
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.372940632520386,
+      "learning_rate": 1.5172261418820908e-05,
+      "loss": 0.9409,
+      "step": 2155
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1506409724841307,
+      "learning_rate": 1.5167793646521788e-05,
+      "loss": 0.7848,
+      "step": 2156
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.265659140491366,
+      "learning_rate": 1.5163324466373236e-05,
+      "loss": 0.8734,
+      "step": 2157
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5096840410779218,
+      "learning_rate": 1.5158853879592763e-05,
+      "loss": 0.9241,
+      "step": 2158
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4449940781756598,
+      "learning_rate": 1.515438188739829e-05,
+      "loss": 0.8552,
+      "step": 2159
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.41538473215533,
+      "learning_rate": 1.5149908491008112e-05,
+      "loss": 0.8302,
+      "step": 2160
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2742926506876147,
+      "learning_rate": 1.5145433691640903e-05,
+      "loss": 0.8659,
+      "step": 2161
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2053077593342505,
+      "learning_rate": 1.514095749051572e-05,
+      "loss": 0.8499,
+      "step": 2162
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.407376910059663,
+      "learning_rate": 1.5136479888852006e-05,
+      "loss": 0.8892,
+      "step": 2163
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3727099051272096,
+      "learning_rate": 1.5132000887869583e-05,
+      "loss": 0.9336,
+      "step": 2164
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4988829668066508,
+      "learning_rate": 1.512752048878865e-05,
+      "loss": 0.8863,
+      "step": 2165
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5059552310611017,
+      "learning_rate": 1.5123038692829801e-05,
+      "loss": 0.8937,
+      "step": 2166
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4733453859609438,
+      "learning_rate": 1.5118555501213989e-05,
+      "loss": 0.9127,
+      "step": 2167
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3304119401916974,
+      "learning_rate": 1.5114070915162568e-05,
+      "loss": 0.9368,
+      "step": 2168
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.285562472797957,
+      "learning_rate": 1.5109584935897259e-05,
+      "loss": 0.8714,
+      "step": 2169
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5548034869111933,
+      "learning_rate": 1.5105097564640168e-05,
+      "loss": 0.8717,
+      "step": 2170
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5448864946039587,
+      "learning_rate": 1.5100608802613775e-05,
+      "loss": 0.9049,
+      "step": 2171
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.564177162557625,
+      "learning_rate": 1.5096118651040945e-05,
+      "loss": 0.8878,
+      "step": 2172
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6108031189204868,
+      "learning_rate": 1.5091627111144923e-05,
+      "loss": 0.8345,
+      "step": 2173
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4910460795049896,
+      "learning_rate": 1.508713418414932e-05,
+      "loss": 0.8902,
+      "step": 2174
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.385847744271693,
+      "learning_rate": 1.5082639871278139e-05,
+      "loss": 0.8166,
+      "step": 2175
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6720076702188817,
+      "learning_rate": 1.5078144173755754e-05,
+      "loss": 0.8664,
+      "step": 2176
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4218040804940049,
+      "learning_rate": 1.5073647092806916e-05,
+      "loss": 0.8736,
+      "step": 2177
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4466330153398057,
+      "learning_rate": 1.5069148629656752e-05,
+      "loss": 0.8521,
+      "step": 2178
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.439371853996637,
+      "learning_rate": 1.5064648785530774e-05,
+      "loss": 0.8977,
+      "step": 2179
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2355846765580558,
+      "learning_rate": 1.5060147561654854e-05,
+      "loss": 0.8855,
+      "step": 2180
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.663069990958584,
+      "learning_rate": 1.5055644959255257e-05,
+      "loss": 0.9327,
+      "step": 2181
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4364188024558948,
+      "learning_rate": 1.5051140979558614e-05,
+      "loss": 0.9091,
+      "step": 2182
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3251886880662678,
+      "learning_rate": 1.504663562379193e-05,
+      "loss": 0.8782,
+      "step": 2183
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4595920147654875,
+      "learning_rate": 1.5042128893182595e-05,
+      "loss": 0.8864,
+      "step": 2184
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.387895235482913,
+      "learning_rate": 1.5037620788958359e-05,
+      "loss": 0.9167,
+      "step": 2185
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.237258308477773,
+      "learning_rate": 1.5033111312347357e-05,
+      "loss": 0.8547,
+      "step": 2186
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0173283103746193,
+      "learning_rate": 1.5028600464578099e-05,
+      "loss": 0.3934,
+      "step": 2187
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.695130830690747,
+      "learning_rate": 1.5024088246879456e-05,
+      "loss": 0.8903,
+      "step": 2188
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4066679062034422,
+      "learning_rate": 1.5019574660480685e-05,
+      "loss": 0.8846,
+      "step": 2189
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5432713984385202,
+      "learning_rate": 1.5015059706611413e-05,
+      "loss": 0.8517,
+      "step": 2190
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3007086473686038,
+      "learning_rate": 1.5010543386501634e-05,
+      "loss": 0.8822,
+      "step": 2191
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4091574322799574,
+      "learning_rate": 1.500602570138172e-05,
+      "loss": 0.913,
+      "step": 2192
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5937616848049787,
+      "learning_rate": 1.5001506652482415e-05,
+      "loss": 0.9059,
+      "step": 2193
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6838849755100995,
+      "learning_rate": 1.499698624103483e-05,
+      "loss": 0.9309,
+      "step": 2194
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7340667833071515,
+      "learning_rate": 1.4992464468270451e-05,
+      "loss": 0.9185,
+      "step": 2195
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.401545246123321,
+      "learning_rate": 1.4987941335421132e-05,
+      "loss": 0.901,
+      "step": 2196
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6470834823130729,
+      "learning_rate": 1.4983416843719099e-05,
+      "loss": 0.8706,
+      "step": 2197
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3348927613006705,
+      "learning_rate": 1.497889099439695e-05,
+      "loss": 0.8099,
+      "step": 2198
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3441023905137717,
+      "learning_rate": 1.4974363788687651e-05,
+      "loss": 0.8681,
+      "step": 2199
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4593763048433683,
+      "learning_rate": 1.4969835227824533e-05,
+      "loss": 0.9,
+      "step": 2200
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4980166347017778,
+      "learning_rate": 1.496530531304131e-05,
+      "loss": 0.9216,
+      "step": 2201
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3940221367396826,
+      "learning_rate": 1.4960774045572046e-05,
+      "loss": 0.8919,
+      "step": 2202
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3829439506175594,
+      "learning_rate": 1.495624142665119e-05,
+      "loss": 0.9305,
+      "step": 2203
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4953974947545636,
+      "learning_rate": 1.4951707457513549e-05,
+      "loss": 0.874,
+      "step": 2204
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3430045144049578,
+      "learning_rate": 1.4947172139394301e-05,
+      "loss": 0.8792,
+      "step": 2205
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3363199634631722,
+      "learning_rate": 1.4942635473528994e-05,
+      "loss": 0.9126,
+      "step": 2206
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5989883915243177,
+      "learning_rate": 1.493809746115354e-05,
+      "loss": 0.9173,
+      "step": 2207
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4735093693888386,
+      "learning_rate": 1.4933558103504215e-05,
+      "loss": 0.8741,
+      "step": 2208
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3664428420044636,
+      "learning_rate": 1.4929017401817672e-05,
+      "loss": 0.8533,
+      "step": 2209
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.33577014546894,
+      "learning_rate": 1.4924475357330919e-05,
+      "loss": 0.8291,
+      "step": 2210
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5534383257712345,
+      "learning_rate": 1.491993197128133e-05,
+      "loss": 0.9107,
+      "step": 2211
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5562728766929625,
+      "learning_rate": 1.4915387244906658e-05,
+      "loss": 0.8227,
+      "step": 2212
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3764733689427457,
+      "learning_rate": 1.4910841179445007e-05,
+      "loss": 0.9447,
+      "step": 2213
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9009180427465397,
+      "learning_rate": 1.4906293776134849e-05,
+      "loss": 0.3532,
+      "step": 2214
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3928709491375821,
+      "learning_rate": 1.4901745036215022e-05,
+      "loss": 0.9205,
+      "step": 2215
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4288269772864715,
+      "learning_rate": 1.4897194960924732e-05,
+      "loss": 0.9116,
+      "step": 2216
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5364191948760364,
+      "learning_rate": 1.4892643551503545e-05,
+      "loss": 0.9371,
+      "step": 2217
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2509108428606786,
+      "learning_rate": 1.4888090809191384e-05,
+      "loss": 0.7942,
+      "step": 2218
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2718800392039336,
+      "learning_rate": 1.4883536735228548e-05,
+      "loss": 0.8238,
+      "step": 2219
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3849004137136909,
+      "learning_rate": 1.4878981330855688e-05,
+      "loss": 0.865,
+      "step": 2220
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5411500956725745,
+      "learning_rate": 1.4874424597313828e-05,
+      "loss": 0.9407,
+      "step": 2221
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3353674265555382,
+      "learning_rate": 1.4869866535844337e-05,
+      "loss": 0.8845,
+      "step": 2222
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3698484887338873,
+      "learning_rate": 1.4865307147688967e-05,
+      "loss": 0.9461,
+      "step": 2223
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2663180747978635,
+      "learning_rate": 1.4860746434089817e-05,
+      "loss": 0.8473,
+      "step": 2224
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5572281952642122,
+      "learning_rate": 1.4856184396289348e-05,
+      "loss": 0.8364,
+      "step": 2225
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8211036301561949,
+      "learning_rate": 1.4851621035530392e-05,
+      "loss": 0.3428,
+      "step": 2226
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.203748414150844,
+      "learning_rate": 1.4847056353056126e-05,
+      "loss": 0.8329,
+      "step": 2227
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4945221208212804,
+      "learning_rate": 1.4842490350110103e-05,
+      "loss": 0.9543,
+      "step": 2228
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.19980046983959,
+      "learning_rate": 1.4837923027936223e-05,
+      "loss": 0.9187,
+      "step": 2229
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3635322311079516,
+      "learning_rate": 1.4833354387778753e-05,
+      "loss": 0.8704,
+      "step": 2230
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3559404330255946,
+      "learning_rate": 1.4828784430882315e-05,
+      "loss": 0.8784,
+      "step": 2231
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2887310757991217,
+      "learning_rate": 1.482421315849189e-05,
+      "loss": 0.9186,
+      "step": 2232
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1409952296876684,
+      "learning_rate": 1.4819640571852823e-05,
+      "loss": 0.8771,
+      "step": 2233
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3235816442735124,
+      "learning_rate": 1.4815066672210809e-05,
+      "loss": 0.8059,
+      "step": 2234
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5422495270297572,
+      "learning_rate": 1.4810491460811907e-05,
+      "loss": 0.8857,
+      "step": 2235
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3866930482389916,
+      "learning_rate": 1.4805914938902525e-05,
+      "loss": 0.8656,
+      "step": 2236
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3360330313272575,
+      "learning_rate": 1.4801337107729443e-05,
+      "loss": 0.9133,
+      "step": 2237
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4473589031826841,
+      "learning_rate": 1.4796757968539779e-05,
+      "loss": 0.8265,
+      "step": 2238
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.429804220021035,
+      "learning_rate": 1.4792177522581023e-05,
+      "loss": 0.9466,
+      "step": 2239
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4044861407329503,
+      "learning_rate": 1.4787595771101013e-05,
+      "loss": 0.8658,
+      "step": 2240
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4843499934449191,
+      "learning_rate": 1.4783012715347944e-05,
+      "loss": 0.9163,
+      "step": 2241
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4298231986793588,
+      "learning_rate": 1.4778428356570365e-05,
+      "loss": 0.8582,
+      "step": 2242
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6599928702575104,
+      "learning_rate": 1.4773842696017184e-05,
+      "loss": 0.8379,
+      "step": 2243
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4407991885818552,
+      "learning_rate": 1.4769255734937662e-05,
+      "loss": 0.8623,
+      "step": 2244
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5203572362848454,
+      "learning_rate": 1.4764667474581416e-05,
+      "loss": 0.8996,
+      "step": 2245
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5506097689160783,
+      "learning_rate": 1.4760077916198405e-05,
+      "loss": 0.9055,
+      "step": 2246
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2026901202285056,
+      "learning_rate": 1.475548706103896e-05,
+      "loss": 0.8403,
+      "step": 2247
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.466505080049432,
+      "learning_rate": 1.4750894910353754e-05,
+      "loss": 0.8992,
+      "step": 2248
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.377555539535488,
+      "learning_rate": 1.4746301465393814e-05,
+      "loss": 0.9135,
+      "step": 2249
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4315536753647504,
+      "learning_rate": 1.4741706727410522e-05,
+      "loss": 0.8818,
+      "step": 2250
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.595508009048657,
+      "learning_rate": 1.4737110697655613e-05,
+      "loss": 0.9461,
+      "step": 2251
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4742230168145904,
+      "learning_rate": 1.473251337738117e-05,
+      "loss": 0.812,
+      "step": 2252
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5138875149358872,
+      "learning_rate": 1.472791476783963e-05,
+      "loss": 0.8835,
+      "step": 2253
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5286564667514386,
+      "learning_rate": 1.4723314870283783e-05,
+      "loss": 0.9118,
+      "step": 2254
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3020136164041431,
+      "learning_rate": 1.4718713685966765e-05,
+      "loss": 0.9112,
+      "step": 2255
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3832504311772398,
+      "learning_rate": 1.4714111216142068e-05,
+      "loss": 0.8596,
+      "step": 2256
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.409251480230001,
+      "learning_rate": 1.470950746206353e-05,
+      "loss": 0.917,
+      "step": 2257
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3172843194993251,
+      "learning_rate": 1.4704902424985341e-05,
+      "loss": 0.9592,
+      "step": 2258
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3252559611442432,
+      "learning_rate": 1.4700296106162042e-05,
+      "loss": 0.8878,
+      "step": 2259
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8388986964673923,
+      "learning_rate": 1.4695688506848513e-05,
+      "loss": 0.8958,
+      "step": 2260
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2903790062134557,
+      "learning_rate": 1.4691079628300004e-05,
+      "loss": 0.9224,
+      "step": 2261
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2773115207969163,
+      "learning_rate": 1.4686469471772089e-05,
+      "loss": 0.9451,
+      "step": 2262
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4205982795733245,
+      "learning_rate": 1.4681858038520711e-05,
+      "loss": 0.8737,
+      "step": 2263
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5601666910035128,
+      "learning_rate": 1.4677245329802146e-05,
+      "loss": 0.8408,
+      "step": 2264
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4217344426803438,
+      "learning_rate": 1.4672631346873023e-05,
+      "loss": 0.8615,
+      "step": 2265
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.350937697527269,
+      "learning_rate": 1.466801609099032e-05,
+      "loss": 0.8693,
+      "step": 2266
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3570689840401682,
+      "learning_rate": 1.4663399563411358e-05,
+      "loss": 0.8811,
+      "step": 2267
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3496948672624813,
+      "learning_rate": 1.4658781765393808e-05,
+      "loss": 0.8953,
+      "step": 2268
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.327053535442286,
+      "learning_rate": 1.4654162698195684e-05,
+      "loss": 0.8904,
+      "step": 2269
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5174920065021593,
+      "learning_rate": 1.4649542363075353e-05,
+      "loss": 0.8481,
+      "step": 2270
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4116288497959797,
+      "learning_rate": 1.464492076129151e-05,
+      "loss": 0.9165,
+      "step": 2271
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8629959785761128,
+      "learning_rate": 1.464029789410322e-05,
+      "loss": 0.9483,
+      "step": 2272
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.93831840397108,
+      "learning_rate": 1.4635673762769868e-05,
+      "loss": 0.345,
+      "step": 2273
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1893920115413457,
+      "learning_rate": 1.4631048368551204e-05,
+      "loss": 0.8606,
+      "step": 2274
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3871992796453383,
+      "learning_rate": 1.4626421712707307e-05,
+      "loss": 0.8842,
+      "step": 2275
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6366211405716014,
+      "learning_rate": 1.4621793796498606e-05,
+      "loss": 0.8938,
+      "step": 2276
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.801098204003515,
+      "learning_rate": 1.4617164621185877e-05,
+      "loss": 0.9431,
+      "step": 2277
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.644681898723982,
+      "learning_rate": 1.4612534188030233e-05,
+      "loss": 0.8997,
+      "step": 2278
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3780727954251806,
+      "learning_rate": 1.4607902498293127e-05,
+      "loss": 0.8016,
+      "step": 2279
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3223759282589622,
+      "learning_rate": 1.4603269553236366e-05,
+      "loss": 0.9506,
+      "step": 2280
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3953561376735486,
+      "learning_rate": 1.4598635354122087e-05,
+      "loss": 0.864,
+      "step": 2281
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7313769424374013,
+      "learning_rate": 1.4593999902212775e-05,
+      "loss": 0.869,
+      "step": 2282
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.274980748635463,
+      "learning_rate": 1.4589363198771258e-05,
+      "loss": 0.8288,
+      "step": 2283
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4108575136592878,
+      "learning_rate": 1.4584725245060694e-05,
+      "loss": 0.8715,
+      "step": 2284
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5195966822331568,
+      "learning_rate": 1.45800860423446e-05,
+      "loss": 0.9155,
+      "step": 2285
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3887607298563331,
+      "learning_rate": 1.4575445591886814e-05,
+      "loss": 0.9505,
+      "step": 2286
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3024717728254744,
+      "learning_rate": 1.4570803894951528e-05,
+      "loss": 0.863,
+      "step": 2287
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4590078840953333,
+      "learning_rate": 1.4566160952803268e-05,
+      "loss": 0.8935,
+      "step": 2288
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3611168327986414,
+      "learning_rate": 1.4561516766706893e-05,
+      "loss": 0.9488,
+      "step": 2289
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4840563394419044,
+      "learning_rate": 1.4556871337927615e-05,
+      "loss": 0.8351,
+      "step": 2290
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4736371182479293,
+      "learning_rate": 1.455222466773097e-05,
+      "loss": 0.8934,
+      "step": 2291
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3172184737020978,
+      "learning_rate": 1.4547576757382843e-05,
+      "loss": 0.8857,
+      "step": 2292
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0850948715462614,
+      "learning_rate": 1.4542927608149456e-05,
+      "loss": 0.3364,
+      "step": 2293
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3167077909124218,
+      "learning_rate": 1.453827722129736e-05,
+      "loss": 0.8587,
+      "step": 2294
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4148094617542475,
+      "learning_rate": 1.4533625598093453e-05,
+      "loss": 0.9536,
+      "step": 2295
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3690713852694407,
+      "learning_rate": 1.452897273980496e-05,
+      "loss": 0.9188,
+      "step": 2296
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.815478259601864,
+      "learning_rate": 1.452431864769945e-05,
+      "loss": 0.9335,
+      "step": 2297
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1742590840678162,
+      "learning_rate": 1.451966332304483e-05,
+      "loss": 0.9484,
+      "step": 2298
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8761479848988268,
+      "learning_rate": 1.4515006767109336e-05,
+      "loss": 0.353,
+      "step": 2299
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4187932664831986,
+      "learning_rate": 1.451034898116154e-05,
+      "loss": 0.9211,
+      "step": 2300
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4367702150280137,
+      "learning_rate": 1.4505689966470353e-05,
+      "loss": 0.8344,
+      "step": 2301
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4237007246332254,
+      "learning_rate": 1.4501029724305019e-05,
+      "loss": 0.8346,
+      "step": 2302
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3004173440643338,
+      "learning_rate": 1.4496368255935115e-05,
+      "loss": 0.8892,
+      "step": 2303
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3140514441636295,
+      "learning_rate": 1.4491705562630555e-05,
+      "loss": 0.8881,
+      "step": 2304
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7634660784521374,
+      "learning_rate": 1.4487041645661588e-05,
+      "loss": 0.9144,
+      "step": 2305
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.428872698037481,
+      "learning_rate": 1.448237650629879e-05,
+      "loss": 0.866,
+      "step": 2306
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2516594124180394,
+      "learning_rate": 1.4477710145813074e-05,
+      "loss": 0.945,
+      "step": 2307
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8378396013147641,
+      "learning_rate": 1.4473042565475684e-05,
+      "loss": 0.3709,
+      "step": 2308
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2823692124034947,
+      "learning_rate": 1.44683737665582e-05,
+      "loss": 0.885,
+      "step": 2309
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.335757176459738,
+      "learning_rate": 1.4463703750332532e-05,
+      "loss": 0.8193,
+      "step": 2310
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3921597575946034,
+      "learning_rate": 1.4459032518070917e-05,
+      "loss": 0.8593,
+      "step": 2311
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8924746913222323,
+      "learning_rate": 1.4454360071045933e-05,
+      "loss": 0.9466,
+      "step": 2312
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3528466628879237,
+      "learning_rate": 1.4449686410530478e-05,
+      "loss": 0.8469,
+      "step": 2313
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3656002572194696,
+      "learning_rate": 1.4445011537797788e-05,
+      "loss": 0.9335,
+      "step": 2314
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0229486770446363,
+      "learning_rate": 1.4440335454121428e-05,
+      "loss": 0.3387,
+      "step": 2315
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2594263647563648,
+      "learning_rate": 1.4435658160775296e-05,
+      "loss": 0.8384,
+      "step": 2316
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.349953835192534,
+      "learning_rate": 1.4430979659033609e-05,
+      "loss": 0.8855,
+      "step": 2317
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2380791925715058,
+      "learning_rate": 1.442629995017092e-05,
+      "loss": 0.8738,
+      "step": 2318
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3517152046749763,
+      "learning_rate": 1.4421619035462115e-05,
+      "loss": 0.8757,
+      "step": 2319
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4407721445358312,
+      "learning_rate": 1.44169369161824e-05,
+      "loss": 0.7921,
+      "step": 2320
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4268828757590253,
+      "learning_rate": 1.4412253593607317e-05,
+      "loss": 0.9313,
+      "step": 2321
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3541678774045764,
+      "learning_rate": 1.4407569069012729e-05,
+      "loss": 0.8978,
+      "step": 2322
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4713251981724855,
+      "learning_rate": 1.4402883343674834e-05,
+      "loss": 0.9266,
+      "step": 2323
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3447001931812363,
+      "learning_rate": 1.4398196418870147e-05,
+      "loss": 0.8623,
+      "step": 2324
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8588017581060287,
+      "learning_rate": 1.439350829587552e-05,
+      "loss": 0.3442,
+      "step": 2325
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4449285448265363,
+      "learning_rate": 1.4388818975968126e-05,
+      "loss": 0.8218,
+      "step": 2326
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.66227848501601,
+      "learning_rate": 1.4384128460425467e-05,
+      "loss": 0.8696,
+      "step": 2327
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6301332124844432,
+      "learning_rate": 1.4379436750525362e-05,
+      "loss": 0.9313,
+      "step": 2328
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4705068538346993,
+      "learning_rate": 1.4374743847545967e-05,
+      "loss": 0.9439,
+      "step": 2329
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.513450974373518,
+      "learning_rate": 1.437004975276576e-05,
+      "loss": 0.9266,
+      "step": 2330
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.308554240954074,
+      "learning_rate": 1.4365354467463535e-05,
+      "loss": 0.913,
+      "step": 2331
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5428335746270734,
+      "learning_rate": 1.4360657992918423e-05,
+      "loss": 0.9383,
+      "step": 2332
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5027962142639386,
+      "learning_rate": 1.435596033040987e-05,
+      "loss": 0.8901,
+      "step": 2333
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.335328444994548,
+      "learning_rate": 1.4351261481217655e-05,
+      "loss": 0.9365,
+      "step": 2334
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3755082876140545,
+      "learning_rate": 1.4346561446621865e-05,
+      "loss": 0.9054,
+      "step": 2335
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5176193140888927,
+      "learning_rate": 1.4341860227902923e-05,
+      "loss": 0.9517,
+      "step": 2336
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3000791712698776,
+      "learning_rate": 1.4337157826341575e-05,
+      "loss": 0.9027,
+      "step": 2337
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3134990640577662,
+      "learning_rate": 1.4332454243218878e-05,
+      "loss": 0.8481,
+      "step": 2338
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3828734664624016,
+      "learning_rate": 1.432774947981622e-05,
+      "loss": 0.8451,
+      "step": 2339
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1983359660795525,
+      "learning_rate": 1.4323043537415311e-05,
+      "loss": 0.8485,
+      "step": 2340
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2515336478047323,
+      "learning_rate": 1.4318336417298173e-05,
+      "loss": 0.8891,
+      "step": 2341
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4104676180271158,
+      "learning_rate": 1.431362812074716e-05,
+      "loss": 0.8303,
+      "step": 2342
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3026118771381907,
+      "learning_rate": 1.4308918649044947e-05,
+      "loss": 0.912,
+      "step": 2343
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.298827873316002,
+      "learning_rate": 1.4304208003474508e-05,
+      "loss": 0.8445,
+      "step": 2344
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3275492065140901,
+      "learning_rate": 1.429949618531917e-05,
+      "loss": 0.945,
+      "step": 2345
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7085651428451163,
+      "learning_rate": 1.4294783195862553e-05,
+      "loss": 0.8325,
+      "step": 2346
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8666770819402482,
+      "learning_rate": 1.4290069036388607e-05,
+      "loss": 0.3645,
+      "step": 2347
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9054608973959942,
+      "learning_rate": 1.42853537081816e-05,
+      "loss": 0.3374,
+      "step": 2348
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6988285263369804,
+      "learning_rate": 1.4280637212526116e-05,
+      "loss": 0.9316,
+      "step": 2349
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5271065947375175,
+      "learning_rate": 1.4275919550707058e-05,
+      "loss": 0.8365,
+      "step": 2350
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3399314205247148,
+      "learning_rate": 1.4271200724009648e-05,
+      "loss": 0.9861,
+      "step": 2351
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8278897604834503,
+      "learning_rate": 1.4266480733719426e-05,
+      "loss": 0.9477,
+      "step": 2352
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4272361713264188,
+      "learning_rate": 1.4261759581122243e-05,
+      "loss": 0.9376,
+      "step": 2353
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.904939267805562,
+      "learning_rate": 1.4257037267504277e-05,
+      "loss": 0.341,
+      "step": 2354
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2982544040578603,
+      "learning_rate": 1.425231379415201e-05,
+      "loss": 0.8394,
+      "step": 2355
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.29295131723459,
+      "learning_rate": 1.4247589162352254e-05,
+      "loss": 0.8835,
+      "step": 2356
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8530824722049526,
+      "learning_rate": 1.4242863373392123e-05,
+      "loss": 0.3404,
+      "step": 2357
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3336248101738803,
+      "learning_rate": 1.423813642855905e-05,
+      "loss": 0.8602,
+      "step": 2358
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4707176947410265,
+      "learning_rate": 1.4233408329140796e-05,
+      "loss": 0.9193,
+      "step": 2359
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2490026980804068,
+      "learning_rate": 1.4228679076425414e-05,
+      "loss": 0.8729,
+      "step": 2360
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3143610475663956,
+      "learning_rate": 1.4223948671701289e-05,
+      "loss": 0.7777,
+      "step": 2361
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7865738916708855,
+      "learning_rate": 1.4219217116257111e-05,
+      "loss": 0.3279,
+      "step": 2362
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5862865392761671,
+      "learning_rate": 1.4214484411381885e-05,
+      "loss": 0.8773,
+      "step": 2363
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.248686163702446,
+      "learning_rate": 1.4209750558364936e-05,
+      "loss": 0.9352,
+      "step": 2364
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.45230269430469,
+      "learning_rate": 1.4205015558495893e-05,
+      "loss": 0.8383,
+      "step": 2365
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5173336486325004,
+      "learning_rate": 1.4200279413064695e-05,
+      "loss": 0.9778,
+      "step": 2366
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5354128355012306,
+      "learning_rate": 1.419554212336161e-05,
+      "loss": 0.9282,
+      "step": 2367
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.644502211380055,
+      "learning_rate": 1.4190803690677195e-05,
+      "loss": 0.9084,
+      "step": 2368
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3611401088758242,
+      "learning_rate": 1.4186064116302336e-05,
+      "loss": 0.8821,
+      "step": 2369
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9827002423409152,
+      "learning_rate": 1.4181323401528224e-05,
+      "loss": 0.3581,
+      "step": 2370
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3561769471292597,
+      "learning_rate": 1.4176581547646354e-05,
+      "loss": 0.8966,
+      "step": 2371
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4878033254855738,
+      "learning_rate": 1.4171838555948548e-05,
+      "loss": 0.8066,
+      "step": 2372
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1602231999387538,
+      "learning_rate": 1.4167094427726916e-05,
+      "loss": 0.896,
+      "step": 2373
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5066576986406188,
+      "learning_rate": 1.4162349164273899e-05,
+      "loss": 0.9404,
+      "step": 2374
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5156690852891446,
+      "learning_rate": 1.4157602766882233e-05,
+      "loss": 0.8483,
+      "step": 2375
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8639810755389616,
+      "learning_rate": 1.4152855236844969e-05,
+      "loss": 0.8858,
+      "step": 2376
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.357957809638119,
+      "learning_rate": 1.414810657545546e-05,
+      "loss": 0.8565,
+      "step": 2377
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3921808322111036,
+      "learning_rate": 1.4143356784007383e-05,
+      "loss": 0.9132,
+      "step": 2378
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.438989282458107,
+      "learning_rate": 1.4138605863794703e-05,
+      "loss": 0.9174,
+      "step": 2379
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5333557052984321,
+      "learning_rate": 1.4133853816111703e-05,
+      "loss": 0.8386,
+      "step": 2380
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4108096957277927,
+      "learning_rate": 1.4129100642252978e-05,
+      "loss": 0.8925,
+      "step": 2381
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8295845735459987,
+      "learning_rate": 1.4124346343513411e-05,
+      "loss": 0.3456,
+      "step": 2382
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4893170667023041,
+      "learning_rate": 1.4119590921188217e-05,
+      "loss": 0.9038,
+      "step": 2383
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1956011756265903,
+      "learning_rate": 1.4114834376572898e-05,
+      "loss": 0.8457,
+      "step": 2384
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3722441597081454,
+      "learning_rate": 1.4110076710963269e-05,
+      "loss": 0.937,
+      "step": 2385
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3862584189526093,
+      "learning_rate": 1.4105317925655448e-05,
+      "loss": 0.8846,
+      "step": 2386
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3173203082346663,
+      "learning_rate": 1.4100558021945863e-05,
+      "loss": 0.8444,
+      "step": 2387
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4414543877810566,
+      "learning_rate": 1.4095797001131238e-05,
+      "loss": 0.8533,
+      "step": 2388
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0625404006625905,
+      "learning_rate": 1.4091034864508608e-05,
+      "loss": 0.3078,
+      "step": 2389
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2559781670933106,
+      "learning_rate": 1.408627161337531e-05,
+      "loss": 0.8449,
+      "step": 2390
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.301360216970839,
+      "learning_rate": 1.4081507249028987e-05,
+      "loss": 0.9401,
+      "step": 2391
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4230749835856906,
+      "learning_rate": 1.4076741772767586e-05,
+      "loss": 0.8865,
+      "step": 2392
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3386574935029263,
+      "learning_rate": 1.4071975185889344e-05,
+      "loss": 0.8756,
+      "step": 2393
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.223114574469079,
+      "learning_rate": 1.406720748969282e-05,
+      "loss": 0.863,
+      "step": 2394
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4618617980472217,
+      "learning_rate": 1.4062438685476862e-05,
+      "loss": 0.9081,
+      "step": 2395
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7333851021267135,
+      "learning_rate": 1.4057668774540622e-05,
+      "loss": 0.8528,
+      "step": 2396
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3268877774134125,
+      "learning_rate": 1.405289775818356e-05,
+      "loss": 0.8836,
+      "step": 2397
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4578891937130678,
+      "learning_rate": 1.4048125637705433e-05,
+      "loss": 0.8496,
+      "step": 2398
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2941225995663115,
+      "learning_rate": 1.404335241440629e-05,
+      "loss": 0.8377,
+      "step": 2399
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4734029704413851,
+      "learning_rate": 1.4038578089586493e-05,
+      "loss": 0.8771,
+      "step": 2400
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2483468792100927,
+      "learning_rate": 1.4033802664546704e-05,
+      "loss": 0.9133,
+      "step": 2401
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.591754700973777,
+      "learning_rate": 1.4029026140587876e-05,
+      "loss": 0.8767,
+      "step": 2402
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.371542106718214,
+      "learning_rate": 1.4024248519011266e-05,
+      "loss": 0.8427,
+      "step": 2403
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5535073899273129,
+      "learning_rate": 1.4019469801118429e-05,
+      "loss": 0.8505,
+      "step": 2404
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8314805494354032,
+      "learning_rate": 1.4014689988211223e-05,
+      "loss": 0.8411,
+      "step": 2405
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3193028006073544,
+      "learning_rate": 1.4009909081591798e-05,
+      "loss": 0.9397,
+      "step": 2406
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3300142608972412,
+      "learning_rate": 1.4005127082562603e-05,
+      "loss": 0.8583,
+      "step": 2407
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.375453318841157,
+      "learning_rate": 1.4000343992426391e-05,
+      "loss": 0.8603,
+      "step": 2408
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5195993834766162,
+      "learning_rate": 1.3995559812486205e-05,
+      "loss": 0.85,
+      "step": 2409
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7163499569419223,
+      "learning_rate": 1.399077454404539e-05,
+      "loss": 0.896,
+      "step": 2410
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1600600380932284,
+      "learning_rate": 1.398598818840758e-05,
+      "loss": 0.8758,
+      "step": 2411
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2530655527075543,
+      "learning_rate": 1.3981200746876713e-05,
+      "loss": 0.8189,
+      "step": 2412
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1740726570895015,
+      "learning_rate": 1.397641222075702e-05,
+      "loss": 0.9119,
+      "step": 2413
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3950908263470454,
+      "learning_rate": 1.397162261135303e-05,
+      "loss": 0.8324,
+      "step": 2414
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4170086989173616,
+      "learning_rate": 1.396683191996956e-05,
+      "loss": 0.854,
+      "step": 2415
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3969800062896214,
+      "learning_rate": 1.3962040147911731e-05,
+      "loss": 0.7899,
+      "step": 2416
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3521077625202005,
+      "learning_rate": 1.3957247296484948e-05,
+      "loss": 0.8687,
+      "step": 2417
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.445349802360639,
+      "learning_rate": 1.3952453366994921e-05,
+      "loss": 0.9665,
+      "step": 2418
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5941380245820356,
+      "learning_rate": 1.3947658360747646e-05,
+      "loss": 0.898,
+      "step": 2419
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4172281598521912,
+      "learning_rate": 1.3942862279049418e-05,
+      "loss": 0.8759,
+      "step": 2420
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5887946157405524,
+      "learning_rate": 1.393806512320682e-05,
+      "loss": 0.8981,
+      "step": 2421
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2385421441988698,
+      "learning_rate": 1.3933266894526725e-05,
+      "loss": 0.8691,
+      "step": 2422
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4540322654254652,
+      "learning_rate": 1.392846759431631e-05,
+      "loss": 0.8785,
+      "step": 2423
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5417677430309058,
+      "learning_rate": 1.392366722388303e-05,
+      "loss": 0.8963,
+      "step": 2424
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5877710202933286,
+      "learning_rate": 1.3918865784534647e-05,
+      "loss": 0.9224,
+      "step": 2425
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3737331753464996,
+      "learning_rate": 1.3914063277579195e-05,
+      "loss": 0.8323,
+      "step": 2426
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9291795657350221,
+      "learning_rate": 1.3909259704325018e-05,
+      "loss": 0.3265,
+      "step": 2427
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3074598267548723,
+      "learning_rate": 1.3904455066080737e-05,
+      "loss": 0.8468,
+      "step": 2428
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4620030964665989,
+      "learning_rate": 1.389964936415527e-05,
+      "loss": 0.8903,
+      "step": 2429
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3884231087030878,
+      "learning_rate": 1.3894842599857818e-05,
+      "loss": 0.8725,
+      "step": 2430
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4230503618887171,
+      "learning_rate": 1.3890034774497884e-05,
+      "loss": 0.8849,
+      "step": 2431
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2181851640090477,
+      "learning_rate": 1.3885225889385247e-05,
+      "loss": 0.8085,
+      "step": 2432
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5692719800790655,
+      "learning_rate": 1.3880415945829979e-05,
+      "loss": 0.8339,
+      "step": 2433
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4769324004599105,
+      "learning_rate": 1.3875604945142445e-05,
+      "loss": 0.8536,
+      "step": 2434
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5356410646626897,
+      "learning_rate": 1.3870792888633293e-05,
+      "loss": 0.836,
+      "step": 2435
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8496642812986422,
+      "learning_rate": 1.3865979777613459e-05,
+      "loss": 0.3421,
+      "step": 2436
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.902998542697636,
+      "learning_rate": 1.3861165613394163e-05,
+      "loss": 0.3262,
+      "step": 2437
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4805731183004274,
+      "learning_rate": 1.3856350397286926e-05,
+      "loss": 0.8918,
+      "step": 2438
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4858714328182152,
+      "learning_rate": 1.3851534130603535e-05,
+      "loss": 0.898,
+      "step": 2439
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5888001422534668,
+      "learning_rate": 1.384671681465608e-05,
+      "loss": 0.871,
+      "step": 2440
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.431369314077371,
+      "learning_rate": 1.3841898450756933e-05,
+      "loss": 0.899,
+      "step": 2441
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4392971953326603,
+      "learning_rate": 1.383707904021874e-05,
+      "loss": 0.8709,
+      "step": 2442
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2756894288830853,
+      "learning_rate": 1.383225858435445e-05,
+      "loss": 0.9531,
+      "step": 2443
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2519655137780676,
+      "learning_rate": 1.3827437084477285e-05,
+      "loss": 0.8775,
+      "step": 2444
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.57830987510728,
+      "learning_rate": 1.3822614541900751e-05,
+      "loss": 0.8995,
+      "step": 2445
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2262534736298292,
+      "learning_rate": 1.3817790957938648e-05,
+      "loss": 0.8664,
+      "step": 2446
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4233709655814089,
+      "learning_rate": 1.3812966333905052e-05,
+      "loss": 0.9363,
+      "step": 2447
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4578438543380297,
+      "learning_rate": 1.3808140671114316e-05,
+      "loss": 0.8649,
+      "step": 2448
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.649130293725084,
+      "learning_rate": 1.3803313970881093e-05,
+      "loss": 0.8275,
+      "step": 2449
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4220256227843642,
+      "learning_rate": 1.3798486234520306e-05,
+      "loss": 0.8266,
+      "step": 2450
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3347601554698256,
+      "learning_rate": 1.3793657463347158e-05,
+      "loss": 0.8999,
+      "step": 2451
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2935304682078652,
+      "learning_rate": 1.3788827658677151e-05,
+      "loss": 0.8632,
+      "step": 2452
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2701500851569192,
+      "learning_rate": 1.3783996821826043e-05,
+      "loss": 0.8942,
+      "step": 2453
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2948400189633782,
+      "learning_rate": 1.37791649541099e-05,
+      "loss": 0.8906,
+      "step": 2454
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6108081221311763,
+      "learning_rate": 1.3774332056845047e-05,
+      "loss": 0.8447,
+      "step": 2455
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3971923514038544,
+      "learning_rate": 1.3769498131348102e-05,
+      "loss": 0.8839,
+      "step": 2456
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4268500254530283,
+      "learning_rate": 1.376466317893596e-05,
+      "loss": 0.8898,
+      "step": 2457
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3137124363855885,
+      "learning_rate": 1.3759827200925796e-05,
+      "loss": 0.8794,
+      "step": 2458
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6602536184143386,
+      "learning_rate": 1.375499019863506e-05,
+      "loss": 0.913,
+      "step": 2459
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3080705799663142,
+      "learning_rate": 1.3750152173381488e-05,
+      "loss": 0.917,
+      "step": 2460
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5480241173913074,
+      "learning_rate": 1.3745313126483089e-05,
+      "loss": 0.8507,
+      "step": 2461
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.272820031210293,
+      "learning_rate": 1.3740473059258155e-05,
+      "loss": 0.8996,
+      "step": 2462
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5763211859091717,
+      "learning_rate": 1.3735631973025254e-05,
+      "loss": 0.8086,
+      "step": 2463
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.477612631630901,
+      "learning_rate": 1.3730789869103227e-05,
+      "loss": 0.9521,
+      "step": 2464
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.396855974774549,
+      "learning_rate": 1.3725946748811203e-05,
+      "loss": 0.846,
+      "step": 2465
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3262585287197404,
+      "learning_rate": 1.3721102613468578e-05,
+      "loss": 0.8947,
+      "step": 2466
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2660739251142574,
+      "learning_rate": 1.3716257464395026e-05,
+      "loss": 0.9009,
+      "step": 2467
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2530867373363663,
+      "learning_rate": 1.3711411302910504e-05,
+      "loss": 0.8496,
+      "step": 2468
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4175445103279811,
+      "learning_rate": 1.3706564130335236e-05,
+      "loss": 0.8775,
+      "step": 2469
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3125592279410676,
+      "learning_rate": 1.370171594798973e-05,
+      "loss": 0.9216,
+      "step": 2470
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1922923357579691,
+      "learning_rate": 1.3696866757194757e-05,
+      "loss": 0.8413,
+      "step": 2471
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4614569191972286,
+      "learning_rate": 1.3692016559271377e-05,
+      "loss": 0.9114,
+      "step": 2472
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3003547587881557,
+      "learning_rate": 1.3687165355540915e-05,
+      "loss": 0.8758,
+      "step": 2473
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.425584269333321,
+      "learning_rate": 1.3682313147324972e-05,
+      "loss": 0.8967,
+      "step": 2474
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3289298928392963,
+      "learning_rate": 1.3677459935945425e-05,
+      "loss": 0.8635,
+      "step": 2475
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4773843489489369,
+      "learning_rate": 1.3672605722724422e-05,
+      "loss": 0.9136,
+      "step": 2476
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4183727091611855,
+      "learning_rate": 1.3667750508984383e-05,
+      "loss": 0.8227,
+      "step": 2477
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5000880663818352,
+      "learning_rate": 1.3662894296048004e-05,
+      "loss": 0.9126,
+      "step": 2478
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4049797652567548,
+      "learning_rate": 1.365803708523825e-05,
+      "loss": 0.9545,
+      "step": 2479
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6061232681736757,
+      "learning_rate": 1.365317887787836e-05,
+      "loss": 0.8535,
+      "step": 2480
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7455157869149789,
+      "learning_rate": 1.3648319675291842e-05,
+      "loss": 0.9012,
+      "step": 2481
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4901183284407453,
+      "learning_rate": 1.3643459478802479e-05,
+      "loss": 0.9315,
+      "step": 2482
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4792642505927425,
+      "learning_rate": 1.3638598289734321e-05,
+      "loss": 0.8216,
+      "step": 2483
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6928427346986987,
+      "learning_rate": 1.3633736109411691e-05,
+      "loss": 0.8351,
+      "step": 2484
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.864568784231787,
+      "learning_rate": 1.362887293915918e-05,
+      "loss": 0.3327,
+      "step": 2485
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3418242106089469,
+      "learning_rate": 1.362400878030165e-05,
+      "loss": 0.9269,
+      "step": 2486
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7172047756495814,
+      "learning_rate": 1.3619143634164234e-05,
+      "loss": 0.8961,
+      "step": 2487
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3602502741351972,
+      "learning_rate": 1.3614277502072327e-05,
+      "loss": 0.8704,
+      "step": 2488
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6566686438538933,
+      "learning_rate": 1.3609410385351598e-05,
+      "loss": 0.9263,
+      "step": 2489
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4854773688224303,
+      "learning_rate": 1.3604542285327988e-05,
+      "loss": 0.7858,
+      "step": 2490
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3336200041943738,
+      "learning_rate": 1.3599673203327702e-05,
+      "loss": 0.8677,
+      "step": 2491
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4551009893221094,
+      "learning_rate": 1.3594803140677208e-05,
+      "loss": 0.8108,
+      "step": 2492
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.274539173313676,
+      "learning_rate": 1.358993209870325e-05,
+      "loss": 0.8279,
+      "step": 2493
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3199988964748557,
+      "learning_rate": 1.3585060078732827e-05,
+      "loss": 0.9708,
+      "step": 2494
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4590130321849322,
+      "learning_rate": 1.3580187082093217e-05,
+      "loss": 0.8396,
+      "step": 2495
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4208027758300172,
+      "learning_rate": 1.3575313110111958e-05,
+      "loss": 0.849,
+      "step": 2496
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.366423111174138,
+      "learning_rate": 1.3570438164116852e-05,
+      "loss": 0.8966,
+      "step": 2497
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4501265557683753,
+      "learning_rate": 1.3565562245435974e-05,
+      "loss": 0.8714,
+      "step": 2498
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.464461936145771,
+      "learning_rate": 1.3560685355397651e-05,
+      "loss": 0.9147,
+      "step": 2499
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4975417005284506,
+      "learning_rate": 1.355580749533049e-05,
+      "loss": 0.8207,
+      "step": 2500
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2715553988632624,
+      "learning_rate": 1.3550928666563348e-05,
+      "loss": 0.9026,
+      "step": 2501
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4911337778180596,
+      "learning_rate": 1.3546048870425356e-05,
+      "loss": 0.9768,
+      "step": 2502
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5085502435504217,
+      "learning_rate": 1.3541168108245907e-05,
+      "loss": 0.803,
+      "step": 2503
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.233563560100716,
+      "learning_rate": 1.3536286381354651e-05,
+      "loss": 0.8709,
+      "step": 2504
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5068261819302413,
+      "learning_rate": 1.3531403691081505e-05,
+      "loss": 0.869,
+      "step": 2505
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.165559616409158,
+      "learning_rate": 1.352652003875665e-05,
+      "loss": 0.8803,
+      "step": 2506
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4923555756324929,
+      "learning_rate": 1.3521635425710531e-05,
+      "loss": 0.7875,
+      "step": 2507
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2477350539435565,
+      "learning_rate": 1.351674985327384e-05,
+      "loss": 0.8019,
+      "step": 2508
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.446911170607255,
+      "learning_rate": 1.3511863322777557e-05,
+      "loss": 0.7207,
+      "step": 2509
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5000339245933587,
+      "learning_rate": 1.3506975835552894e-05,
+      "loss": 0.8816,
+      "step": 2510
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.214076603374283,
+      "learning_rate": 1.3502087392931347e-05,
+      "loss": 0.877,
+      "step": 2511
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2646272252487343,
+      "learning_rate": 1.3497197996244655e-05,
+      "loss": 0.8897,
+      "step": 2512
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7950867053199877,
+      "learning_rate": 1.3492307646824832e-05,
+      "loss": 0.3627,
+      "step": 2513
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3758026138988586,
+      "learning_rate": 1.3487416346004139e-05,
+      "loss": 0.8413,
+      "step": 2514
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3099062003135604,
+      "learning_rate": 1.3482524095115099e-05,
+      "loss": 0.9307,
+      "step": 2515
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5371718000171308,
+      "learning_rate": 1.3477630895490501e-05,
+      "loss": 0.8679,
+      "step": 2516
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3586745279810586,
+      "learning_rate": 1.3472736748463387e-05,
+      "loss": 0.897,
+      "step": 2517
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.484723316283319,
+      "learning_rate": 1.3467841655367058e-05,
+      "loss": 0.8904,
+      "step": 2518
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4734172145098396,
+      "learning_rate": 1.3462945617535063e-05,
+      "loss": 0.9074,
+      "step": 2519
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1588953805588347,
+      "learning_rate": 1.3458048636301233e-05,
+      "loss": 0.9002,
+      "step": 2520
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.34901673158023,
+      "learning_rate": 1.3453150712999628e-05,
+      "loss": 0.8588,
+      "step": 2521
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6542838111866047,
+      "learning_rate": 1.3448251848964584e-05,
+      "loss": 0.812,
+      "step": 2522
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.505706654435004,
+      "learning_rate": 1.3443352045530685e-05,
+      "loss": 0.8031,
+      "step": 2523
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6319209658227931,
+      "learning_rate": 1.3438451304032769e-05,
+      "loss": 0.8741,
+      "step": 2524
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.177064904628516,
+      "learning_rate": 1.3433549625805941e-05,
+      "loss": 0.7971,
+      "step": 2525
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2198736714450558,
+      "learning_rate": 1.3428647012185545e-05,
+      "loss": 0.9194,
+      "step": 2526
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5395112085504292,
+      "learning_rate": 1.3423743464507192e-05,
+      "loss": 0.8614,
+      "step": 2527
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3943315998848314,
+      "learning_rate": 1.3418838984106746e-05,
+      "loss": 0.8335,
+      "step": 2528
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2866153910544904,
+      "learning_rate": 1.3413933572320317e-05,
+      "loss": 0.9011,
+      "step": 2529
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3563332109700361,
+      "learning_rate": 1.3409027230484279e-05,
+      "loss": 0.8919,
+      "step": 2530
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3613020663182294,
+      "learning_rate": 1.3404119959935254e-05,
+      "loss": 0.7869,
+      "step": 2531
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.417774532615658,
+      "learning_rate": 1.3399211762010117e-05,
+      "loss": 0.8858,
+      "step": 2532
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4025287240532758,
+      "learning_rate": 1.3394302638045992e-05,
+      "loss": 0.8782,
+      "step": 2533
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3302872168883468,
+      "learning_rate": 1.3389392589380265e-05,
+      "loss": 0.8702,
+      "step": 2534
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3075191782999955,
+      "learning_rate": 1.3384481617350572e-05,
+      "loss": 0.9001,
+      "step": 2535
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3280330170306085,
+      "learning_rate": 1.337956972329479e-05,
+      "loss": 0.8165,
+      "step": 2536
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9618068286745403,
+      "learning_rate": 1.3374656908551055e-05,
+      "loss": 0.3383,
+      "step": 2537
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9820078736957742,
+      "learning_rate": 1.3369743174457756e-05,
+      "loss": 0.8589,
+      "step": 2538
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4487974127617518,
+      "learning_rate": 1.336482852235353e-05,
+      "loss": 0.8904,
+      "step": 2539
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4180925580559416,
+      "learning_rate": 1.3359912953577261e-05,
+      "loss": 0.8851,
+      "step": 2540
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4103470947909094,
+      "learning_rate": 1.335499646946809e-05,
+      "loss": 0.9073,
+      "step": 2541
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1586714010203598,
+      "learning_rate": 1.3350079071365398e-05,
+      "loss": 0.806,
+      "step": 2542
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2868201490887554,
+      "learning_rate": 1.334516076060882e-05,
+      "loss": 0.8684,
+      "step": 2543
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3294892863206522,
+      "learning_rate": 1.334024153853824e-05,
+      "loss": 0.8603,
+      "step": 2544
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4313667566894113,
+      "learning_rate": 1.3335321406493795e-05,
+      "loss": 0.8601,
+      "step": 2545
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3237249911795446,
+      "learning_rate": 1.3330400365815856e-05,
+      "loss": 0.8671,
+      "step": 2546
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4722250369067422,
+      "learning_rate": 1.3325478417845057e-05,
+      "loss": 0.8923,
+      "step": 2547
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3990681850392614,
+      "learning_rate": 1.332055556392227e-05,
+      "loss": 0.9051,
+      "step": 2548
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3288438955049955,
+      "learning_rate": 1.3315631805388613e-05,
+      "loss": 0.8843,
+      "step": 2549
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4187240664384753,
+      "learning_rate": 1.3310707143585458e-05,
+      "loss": 0.8413,
+      "step": 2550
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3800066280227417,
+      "learning_rate": 1.3305781579854417e-05,
+      "loss": 0.9358,
+      "step": 2551
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3792045334128331,
+      "learning_rate": 1.3300855115537347e-05,
+      "loss": 0.9177,
+      "step": 2552
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.491787464884296,
+      "learning_rate": 1.3295927751976358e-05,
+      "loss": 0.923,
+      "step": 2553
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3572395593347804,
+      "learning_rate": 1.3290999490513796e-05,
+      "loss": 0.8991,
+      "step": 2554
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2575532908232072,
+      "learning_rate": 1.3286070332492256e-05,
+      "loss": 0.9602,
+      "step": 2555
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2735467283233441,
+      "learning_rate": 1.3281140279254576e-05,
+      "loss": 0.8785,
+      "step": 2556
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2458656675547382,
+      "learning_rate": 1.3276209332143836e-05,
+      "loss": 0.8074,
+      "step": 2557
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2705728300202341,
+      "learning_rate": 1.327127749250337e-05,
+      "loss": 0.874,
+      "step": 2558
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.416496249220626,
+      "learning_rate": 1.3266344761676735e-05,
+      "loss": 0.8076,
+      "step": 2559
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5030210800590615,
+      "learning_rate": 1.3261411141007757e-05,
+      "loss": 0.8775,
+      "step": 2560
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5277822418731677,
+      "learning_rate": 1.3256476631840478e-05,
+      "loss": 0.8721,
+      "step": 2561
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4173015888312606,
+      "learning_rate": 1.3251541235519199e-05,
+      "loss": 0.9598,
+      "step": 2562
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8482713780177396,
+      "learning_rate": 1.324660495338846e-05,
+      "loss": 0.3593,
+      "step": 2563
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.455733430756094,
+      "learning_rate": 1.324166778679304e-05,
+      "loss": 0.8761,
+      "step": 2564
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3243371592669817,
+      "learning_rate": 1.3236729737077956e-05,
+      "loss": 0.8653,
+      "step": 2565
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3623856523853173,
+      "learning_rate": 1.3231790805588469e-05,
+      "loss": 0.9093,
+      "step": 2566
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3138929309820877,
+      "learning_rate": 1.3226850993670087e-05,
+      "loss": 0.842,
+      "step": 2567
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3399232077601675,
+      "learning_rate": 1.3221910302668538e-05,
+      "loss": 0.8972,
+      "step": 2568
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4822591389479125,
+      "learning_rate": 1.3216968733929817e-05,
+      "loss": 0.7926,
+      "step": 2569
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5428916936054882,
+      "learning_rate": 1.321202628880013e-05,
+      "loss": 0.8091,
+      "step": 2570
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.722716451531505,
+      "learning_rate": 1.3207082968625948e-05,
+      "loss": 0.8542,
+      "step": 2571
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3941513146353774,
+      "learning_rate": 1.3202138774753958e-05,
+      "loss": 0.8951,
+      "step": 2572
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4681943694672512,
+      "learning_rate": 1.3197193708531099e-05,
+      "loss": 0.8389,
+      "step": 2573
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1803899575802237,
+      "learning_rate": 1.3192247771304543e-05,
+      "loss": 0.8591,
+      "step": 2574
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1760188984152213,
+      "learning_rate": 1.3187300964421702e-05,
+      "loss": 0.853,
+      "step": 2575
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6497706524127025,
+      "learning_rate": 1.3182353289230216e-05,
+      "loss": 0.9069,
+      "step": 2576
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4897179977583297,
+      "learning_rate": 1.3177404747077973e-05,
+      "loss": 0.843,
+      "step": 2577
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2520766554365843,
+      "learning_rate": 1.3172455339313091e-05,
+      "loss": 0.8505,
+      "step": 2578
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3593661832797572,
+      "learning_rate": 1.3167505067283926e-05,
+      "loss": 0.897,
+      "step": 2579
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4025352083011648,
+      "learning_rate": 1.3162553932339068e-05,
+      "loss": 0.9526,
+      "step": 2580
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4569447360036825,
+      "learning_rate": 1.315760193582734e-05,
+      "loss": 0.7865,
+      "step": 2581
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4870978178352274,
+      "learning_rate": 1.3152649079097808e-05,
+      "loss": 0.9068,
+      "step": 2582
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4880688420605621,
+      "learning_rate": 1.3147695363499762e-05,
+      "loss": 0.7927,
+      "step": 2583
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2447239161220391,
+      "learning_rate": 1.3142740790382733e-05,
+      "loss": 0.8211,
+      "step": 2584
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3306044742843068,
+      "learning_rate": 1.3137785361096485e-05,
+      "loss": 0.8269,
+      "step": 2585
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3099201931183027,
+      "learning_rate": 1.3132829076991007e-05,
+      "loss": 0.8618,
+      "step": 2586
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9106767858361681,
+      "learning_rate": 1.3127871939416533e-05,
+      "loss": 0.31,
+      "step": 2587
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2283520978232152,
+      "learning_rate": 1.3122913949723522e-05,
+      "loss": 0.9179,
+      "step": 2588
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2005677580368528,
+      "learning_rate": 1.3117955109262668e-05,
+      "loss": 0.9024,
+      "step": 2589
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.263292792296421,
+      "learning_rate": 1.3112995419384894e-05,
+      "loss": 0.89,
+      "step": 2590
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2336260932079128,
+      "learning_rate": 1.3108034881441359e-05,
+      "loss": 0.9273,
+      "step": 2591
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4880816877789842,
+      "learning_rate": 1.3103073496783447e-05,
+      "loss": 0.8324,
+      "step": 2592
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6044059133523552,
+      "learning_rate": 1.309811126676278e-05,
+      "loss": 0.8335,
+      "step": 2593
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4301598164770153,
+      "learning_rate": 1.3093148192731202e-05,
+      "loss": 0.9194,
+      "step": 2594
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.252018050535171,
+      "learning_rate": 1.3088184276040794e-05,
+      "loss": 0.8713,
+      "step": 2595
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5921118664489209,
+      "learning_rate": 1.3083219518043866e-05,
+      "loss": 0.9194,
+      "step": 2596
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3405381890124524,
+      "learning_rate": 1.3078253920092949e-05,
+      "loss": 0.84,
+      "step": 2597
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4458761276096601,
+      "learning_rate": 1.3073287483540811e-05,
+      "loss": 0.9318,
+      "step": 2598
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.338476422419945,
+      "learning_rate": 1.3068320209740448e-05,
+      "loss": 0.9278,
+      "step": 2599
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4647763974282297,
+      "learning_rate": 1.3063352100045079e-05,
+      "loss": 0.9142,
+      "step": 2600
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5151063680605492,
+      "learning_rate": 1.3058383155808159e-05,
+      "loss": 0.923,
+      "step": 2601
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.512645373372797,
+      "learning_rate": 1.3053413378383361e-05,
+      "loss": 0.9433,
+      "step": 2602
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2796770914001458,
+      "learning_rate": 1.304844276912459e-05,
+      "loss": 0.8676,
+      "step": 2603
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8416558820272357,
+      "learning_rate": 1.3043471329385979e-05,
+      "loss": 0.3275,
+      "step": 2604
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.256285641055514,
+      "learning_rate": 1.3038499060521886e-05,
+      "loss": 0.834,
+      "step": 2605
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6829836455198537,
+      "learning_rate": 1.3033525963886888e-05,
+      "loss": 0.9554,
+      "step": 2606
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2218507346624357,
+      "learning_rate": 1.3028552040835802e-05,
+      "loss": 0.8236,
+      "step": 2607
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3481418108057788,
+      "learning_rate": 1.3023577292723655e-05,
+      "loss": 0.8983,
+      "step": 2608
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3004050423185876,
+      "learning_rate": 1.3018601720905708e-05,
+      "loss": 0.9056,
+      "step": 2609
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.318733892226603,
+      "learning_rate": 1.3013625326737444e-05,
+      "loss": 0.8386,
+      "step": 2610
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.452227004798366,
+      "learning_rate": 1.300864811157457e-05,
+      "loss": 0.9305,
+      "step": 2611
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4335034092494865,
+      "learning_rate": 1.3003670076773018e-05,
+      "loss": 0.8872,
+      "step": 2612
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4721167149011813,
+      "learning_rate": 1.2998691223688942e-05,
+      "loss": 0.8891,
+      "step": 2613
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1829638425891258,
+      "learning_rate": 1.2993711553678714e-05,
+      "loss": 0.7439,
+      "step": 2614
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7452711641535947,
+      "learning_rate": 1.2988731068098938e-05,
+      "loss": 0.8013,
+      "step": 2615
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6827409845951422,
+      "learning_rate": 1.2983749768306434e-05,
+      "loss": 0.8366,
+      "step": 2616
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1555665656819616,
+      "learning_rate": 1.2978767655658245e-05,
+      "loss": 0.9188,
+      "step": 2617
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.41296337416045,
+      "learning_rate": 1.2973784731511638e-05,
+      "loss": 0.8922,
+      "step": 2618
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.300809667322695,
+      "learning_rate": 1.2968800997224093e-05,
+      "loss": 0.8401,
+      "step": 2619
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4572320945418977,
+      "learning_rate": 1.2963816454153327e-05,
+      "loss": 0.884,
+      "step": 2620
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.366138714019059,
+      "learning_rate": 1.2958831103657255e-05,
+      "loss": 0.8169,
+      "step": 2621
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.074652625537922,
+      "learning_rate": 1.2953844947094032e-05,
+      "loss": 0.9005,
+      "step": 2622
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5331221097449668,
+      "learning_rate": 1.2948857985822023e-05,
+      "loss": 0.8465,
+      "step": 2623
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9361767961179173,
+      "learning_rate": 1.2943870221199814e-05,
+      "loss": 0.8314,
+      "step": 2624
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4216558384457334,
+      "learning_rate": 1.2938881654586207e-05,
+      "loss": 0.884,
+      "step": 2625
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.300658750064086,
+      "learning_rate": 1.2933892287340226e-05,
+      "loss": 0.8771,
+      "step": 2626
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5011052614431577,
+      "learning_rate": 1.2928902120821111e-05,
+      "loss": 0.8303,
+      "step": 2627
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4151939566629255,
+      "learning_rate": 1.2923911156388327e-05,
+      "loss": 0.8905,
+      "step": 2628
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.286310763637221,
+      "learning_rate": 1.2918919395401544e-05,
+      "loss": 0.8189,
+      "step": 2629
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.473585937303035,
+      "learning_rate": 1.2913926839220654e-05,
+      "loss": 0.8695,
+      "step": 2630
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.56640376533159,
+      "learning_rate": 1.2908933489205776e-05,
+      "loss": 0.8711,
+      "step": 2631
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7078866446708545,
+      "learning_rate": 1.2903939346717226e-05,
+      "loss": 0.8863,
+      "step": 2632
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.529668886969601,
+      "learning_rate": 1.2898944413115553e-05,
+      "loss": 0.9429,
+      "step": 2633
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4737030123806616,
+      "learning_rate": 1.2893948689761509e-05,
+      "loss": 0.8652,
+      "step": 2634
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4191085290993761,
+      "learning_rate": 1.2888952178016075e-05,
+      "loss": 0.8948,
+      "step": 2635
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5087838614167377,
+      "learning_rate": 1.2883954879240428e-05,
+      "loss": 0.8405,
+      "step": 2636
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5296946467445982,
+      "learning_rate": 1.2878956794795978e-05,
+      "loss": 0.8923,
+      "step": 2637
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6408508661878922,
+      "learning_rate": 1.2873957926044336e-05,
+      "loss": 0.9277,
+      "step": 2638
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8587017266729103,
+      "learning_rate": 1.2868958274347334e-05,
+      "loss": 0.3524,
+      "step": 2639
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.504428968991861,
+      "learning_rate": 1.2863957841067018e-05,
+      "loss": 0.8191,
+      "step": 2640
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2794169966525712,
+      "learning_rate": 1.2858956627565635e-05,
+      "loss": 0.8926,
+      "step": 2641
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.469706023189959,
+      "learning_rate": 1.2853954635205665e-05,
+      "loss": 0.8518,
+      "step": 2642
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3393238326733468,
+      "learning_rate": 1.2848951865349778e-05,
+      "loss": 0.8402,
+      "step": 2643
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3353973824938152,
+      "learning_rate": 1.2843948319360875e-05,
+      "loss": 0.9042,
+      "step": 2644
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3159054424701762,
+      "learning_rate": 1.2838943998602054e-05,
+      "loss": 0.8658,
+      "step": 2645
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2605262869922957,
+      "learning_rate": 1.2833938904436634e-05,
+      "loss": 0.8761,
+      "step": 2646
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4916405761479787,
+      "learning_rate": 1.2828933038228136e-05,
+      "loss": 0.8372,
+      "step": 2647
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3571887701209984,
+      "learning_rate": 1.2823926401340296e-05,
+      "loss": 0.8497,
+      "step": 2648
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5049692453395742,
+      "learning_rate": 1.2818918995137066e-05,
+      "loss": 0.9612,
+      "step": 2649
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2500086750534154,
+      "learning_rate": 1.2813910820982596e-05,
+      "loss": 0.844,
+      "step": 2650
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6582192924529107,
+      "learning_rate": 1.2808901880241257e-05,
+      "loss": 0.7362,
+      "step": 2651
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5266517482514566,
+      "learning_rate": 1.2803892174277611e-05,
+      "loss": 0.7828,
+      "step": 2652
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4871359483325157,
+      "learning_rate": 1.2798881704456452e-05,
+      "loss": 0.8531,
+      "step": 2653
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1410052652780025,
+      "learning_rate": 1.2793870472142762e-05,
+      "loss": 0.8159,
+      "step": 2654
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3196910744423755,
+      "learning_rate": 1.2788858478701745e-05,
+      "loss": 0.8573,
+      "step": 2655
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2802450488250747,
+      "learning_rate": 1.27838457254988e-05,
+      "loss": 0.8645,
+      "step": 2656
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4046888308471044,
+      "learning_rate": 1.277883221389954e-05,
+      "loss": 0.8484,
+      "step": 2657
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.28613889152579,
+      "learning_rate": 1.2773817945269793e-05,
+      "loss": 0.852,
+      "step": 2658
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4472431504155767,
+      "learning_rate": 1.2768802920975574e-05,
+      "loss": 0.8969,
+      "step": 2659
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2695417653161836,
+      "learning_rate": 1.2763787142383115e-05,
+      "loss": 0.8396,
+      "step": 2660
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5010635596473825,
+      "learning_rate": 1.2758770610858855e-05,
+      "loss": 0.8211,
+      "step": 2661
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3933435749974046,
+      "learning_rate": 1.2753753327769437e-05,
+      "loss": 0.8946,
+      "step": 2662
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2241554065242581,
+      "learning_rate": 1.2748735294481702e-05,
+      "loss": 0.868,
+      "step": 2663
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3454995838810169,
+      "learning_rate": 1.2743716512362705e-05,
+      "loss": 0.905,
+      "step": 2664
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7739553656448708,
+      "learning_rate": 1.2738696982779699e-05,
+      "loss": 0.9268,
+      "step": 2665
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.474368299268508,
+      "learning_rate": 1.273367670710014e-05,
+      "loss": 0.8748,
+      "step": 2666
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7043444807384016,
+      "learning_rate": 1.2728655686691693e-05,
+      "loss": 0.9116,
+      "step": 2667
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.467698117504435,
+      "learning_rate": 1.2723633922922222e-05,
+      "loss": 0.8484,
+      "step": 2668
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4596530449307805,
+      "learning_rate": 1.2718611417159793e-05,
+      "loss": 0.7909,
+      "step": 2669
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2975092684782483,
+      "learning_rate": 1.2713588170772674e-05,
+      "loss": 0.92,
+      "step": 2670
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.467569241180824,
+      "learning_rate": 1.2708564185129339e-05,
+      "loss": 0.7674,
+      "step": 2671
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5221612175260488,
+      "learning_rate": 1.2703539461598455e-05,
+      "loss": 0.9167,
+      "step": 2672
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.254653860424454,
+      "learning_rate": 1.2698514001548904e-05,
+      "loss": 0.8457,
+      "step": 2673
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2960087155045183,
+      "learning_rate": 1.2693487806349744e-05,
+      "loss": 0.8458,
+      "step": 2674
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3460064173888369,
+      "learning_rate": 1.2688460877370269e-05,
+      "loss": 0.8943,
+      "step": 2675
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4337478179414087,
+      "learning_rate": 1.2683433215979937e-05,
+      "loss": 0.9368,
+      "step": 2676
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.47673147486063,
+      "learning_rate": 1.267840482354843e-05,
+      "loss": 0.8816,
+      "step": 2677
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5435719412580633,
+      "learning_rate": 1.2673375701445618e-05,
+      "loss": 0.8417,
+      "step": 2678
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2359780603931836,
+      "learning_rate": 1.266834585104157e-05,
+      "loss": 0.8785,
+      "step": 2679
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.876629276004696,
+      "learning_rate": 1.2663315273706563e-05,
+      "loss": 0.8788,
+      "step": 2680
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5086605628393848,
+      "learning_rate": 1.2658283970811058e-05,
+      "loss": 0.8967,
+      "step": 2681
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.403250572420167,
+      "learning_rate": 1.2653251943725726e-05,
+      "loss": 0.8932,
+      "step": 2682
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4400827170982606,
+      "learning_rate": 1.2648219193821425e-05,
+      "loss": 0.949,
+      "step": 2683
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2243750905546347,
+      "learning_rate": 1.264318572246922e-05,
+      "loss": 0.8674,
+      "step": 2684
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3373017818195951,
+      "learning_rate": 1.2638151531040359e-05,
+      "loss": 0.9306,
+      "step": 2685
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.416649525260536,
+      "learning_rate": 1.2633116620906306e-05,
+      "loss": 0.91,
+      "step": 2686
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2881412103809202,
+      "learning_rate": 1.2628080993438698e-05,
+      "loss": 0.8594,
+      "step": 2687
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4876763968249898,
+      "learning_rate": 1.2623044650009387e-05,
+      "loss": 0.8607,
+      "step": 2688
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5588882258928356,
+      "learning_rate": 1.261800759199041e-05,
+      "loss": 0.8418,
+      "step": 2689
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3516225980616319,
+      "learning_rate": 1.2612969820753995e-05,
+      "loss": 0.8497,
+      "step": 2690
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8270304329310074,
+      "learning_rate": 1.2607931337672576e-05,
+      "loss": 0.8566,
+      "step": 2691
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5207189984133747,
+      "learning_rate": 1.2602892144118772e-05,
+      "loss": 0.8888,
+      "step": 2692
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3797761311635153,
+      "learning_rate": 1.25978522414654e-05,
+      "loss": 0.8888,
+      "step": 2693
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.240589306573735,
+      "learning_rate": 1.2592811631085464e-05,
+      "loss": 0.8878,
+      "step": 2694
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3866993012877769,
+      "learning_rate": 1.2587770314352175e-05,
+      "loss": 0.8413,
+      "step": 2695
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3706973554348998,
+      "learning_rate": 1.2582728292638913e-05,
+      "loss": 0.894,
+      "step": 2696
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1299601180585,
+      "learning_rate": 1.2577685567319275e-05,
+      "loss": 0.9091,
+      "step": 2697
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.218787378416545,
+      "learning_rate": 1.2572642139767033e-05,
+      "loss": 0.9018,
+      "step": 2698
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4909468981019491,
+      "learning_rate": 1.2567598011356155e-05,
+      "loss": 0.8873,
+      "step": 2699
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2850144349352668,
+      "learning_rate": 1.2562553183460806e-05,
+      "loss": 0.9555,
+      "step": 2700
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7531809528452016,
+      "learning_rate": 1.2557507657455327e-05,
+      "loss": 0.8853,
+      "step": 2701
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.453046608163963,
+      "learning_rate": 1.2552461434714272e-05,
+      "loss": 0.8997,
+      "step": 2702
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3922361506418268,
+      "learning_rate": 1.2547414516612357e-05,
+      "loss": 0.9205,
+      "step": 2703
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2445651435008374,
+      "learning_rate": 1.254236690452451e-05,
+      "loss": 0.947,
+      "step": 2704
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.815467976170979,
+      "learning_rate": 1.2537318599825836e-05,
+      "loss": 0.916,
+      "step": 2705
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3661757724372827,
+      "learning_rate": 1.2532269603891639e-05,
+      "loss": 0.9325,
+      "step": 2706
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2308492163139353,
+      "learning_rate": 1.2527219918097392e-05,
+      "loss": 0.8579,
+      "step": 2707
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4671141113770734,
+      "learning_rate": 1.2522169543818781e-05,
+      "loss": 0.9104,
+      "step": 2708
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4326169950749035,
+      "learning_rate": 1.251711848243166e-05,
+      "loss": 0.8886,
+      "step": 2709
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4063801135692475,
+      "learning_rate": 1.251206673531208e-05,
+      "loss": 0.9006,
+      "step": 2710
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.566053466120827,
+      "learning_rate": 1.2507014303836277e-05,
+      "loss": 0.8439,
+      "step": 2711
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3609889194834883,
+      "learning_rate": 1.250196118938067e-05,
+      "loss": 0.8834,
+      "step": 2712
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4800275343934521,
+      "learning_rate": 1.2496907393321868e-05,
+      "loss": 0.8553,
+      "step": 2713
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5141479770941888,
+      "learning_rate": 1.2491852917036666e-05,
+      "loss": 0.8835,
+      "step": 2714
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.483565736694982,
+      "learning_rate": 1.2486797761902039e-05,
+      "loss": 0.7845,
+      "step": 2715
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.467990221231459,
+      "learning_rate": 1.2481741929295154e-05,
+      "loss": 0.9628,
+      "step": 2716
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.268176929238814,
+      "learning_rate": 1.2476685420593357e-05,
+      "loss": 0.9058,
+      "step": 2717
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.420534497611871,
+      "learning_rate": 1.2471628237174183e-05,
+      "loss": 0.8494,
+      "step": 2718
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4818996343853743,
+      "learning_rate": 1.2466570380415346e-05,
+      "loss": 0.845,
+      "step": 2719
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3036555217723786,
+      "learning_rate": 1.2461511851694743e-05,
+      "loss": 0.8744,
+      "step": 2720
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.34724370612865,
+      "learning_rate": 1.2456452652390463e-05,
+      "loss": 0.949,
+      "step": 2721
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.326086301388255,
+      "learning_rate": 1.2451392783880767e-05,
+      "loss": 0.8408,
+      "step": 2722
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3848436942913445,
+      "learning_rate": 1.2446332247544102e-05,
+      "loss": 0.866,
+      "step": 2723
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3696698310164106,
+      "learning_rate": 1.2441271044759103e-05,
+      "loss": 0.9182,
+      "step": 2724
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8835694625073143,
+      "learning_rate": 1.2436209176904575e-05,
+      "loss": 0.3511,
+      "step": 2725
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4675972909617614,
+      "learning_rate": 1.2431146645359511e-05,
+      "loss": 0.8711,
+      "step": 2726
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.514163643108094,
+      "learning_rate": 1.2426083451503086e-05,
+      "loss": 0.8261,
+      "step": 2727
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.546920932824742,
+      "learning_rate": 1.2421019596714656e-05,
+      "loss": 0.8742,
+      "step": 2728
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3920859996109693,
+      "learning_rate": 1.2415955082373752e-05,
+      "loss": 0.8765,
+      "step": 2729
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5382782046930634,
+      "learning_rate": 1.2410889909860086e-05,
+      "loss": 0.8548,
+      "step": 2730
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2657100721012842,
+      "learning_rate": 1.2405824080553553e-05,
+      "loss": 0.7863,
+      "step": 2731
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.474252277992671,
+      "learning_rate": 1.2400757595834221e-05,
+      "loss": 0.842,
+      "step": 2732
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3477498049673966,
+      "learning_rate": 1.2395690457082348e-05,
+      "loss": 0.8355,
+      "step": 2733
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5706843177942593,
+      "learning_rate": 1.239062266567835e-05,
+      "loss": 0.7975,
+      "step": 2734
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.48539112785759,
+      "learning_rate": 1.2385554223002844e-05,
+      "loss": 0.9006,
+      "step": 2735
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7228948985765387,
+      "learning_rate": 1.238048513043661e-05,
+      "loss": 0.8607,
+      "step": 2736
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.257356729550077,
+      "learning_rate": 1.2375415389360605e-05,
+      "loss": 0.8735,
+      "step": 2737
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.207074651431655,
+      "learning_rate": 1.2370345001155972e-05,
+      "loss": 0.877,
+      "step": 2738
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3810815457071217,
+      "learning_rate": 1.2365273967204018e-05,
+      "loss": 0.8835,
+      "step": 2739
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4066106876767055,
+      "learning_rate": 1.2360202288886243e-05,
+      "loss": 0.8312,
+      "step": 2740
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4209303540974425,
+      "learning_rate": 1.23551299675843e-05,
+      "loss": 0.8382,
+      "step": 2741
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.408524447485463,
+      "learning_rate": 1.2350057004680036e-05,
+      "loss": 0.9103,
+      "step": 2742
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6038285773618228,
+      "learning_rate": 1.2344983401555464e-05,
+      "loss": 0.9332,
+      "step": 2743
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6167742067468192,
+      "learning_rate": 1.2339909159592774e-05,
+      "loss": 0.8909,
+      "step": 2744
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4133871135843694,
+      "learning_rate": 1.2334834280174325e-05,
+      "loss": 0.8257,
+      "step": 2745
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.499130896900979,
+      "learning_rate": 1.2329758764682663e-05,
+      "loss": 0.8813,
+      "step": 2746
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2326498160881572,
+      "learning_rate": 1.2324682614500492e-05,
+      "loss": 0.8458,
+      "step": 2747
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4242312451974273,
+      "learning_rate": 1.2319605831010694e-05,
+      "loss": 0.8885,
+      "step": 2748
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.408793049059259,
+      "learning_rate": 1.231452841559633e-05,
+      "loss": 0.8716,
+      "step": 2749
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.455639623306039,
+      "learning_rate": 1.2309450369640622e-05,
+      "loss": 0.8793,
+      "step": 2750
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4140109090983228,
+      "learning_rate": 1.230437169452698e-05,
+      "loss": 0.9245,
+      "step": 2751
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4044579193397113,
+      "learning_rate": 1.2299292391638962e-05,
+      "loss": 0.8144,
+      "step": 2752
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.444665904007464,
+      "learning_rate": 1.2294212462360318e-05,
+      "loss": 0.8437,
+      "step": 2753
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3885806036247759,
+      "learning_rate": 1.2289131908074958e-05,
+      "loss": 0.8473,
+      "step": 2754
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1797751668856538,
+      "learning_rate": 1.2284050730166968e-05,
+      "loss": 0.8338,
+      "step": 2755
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4853253296410984,
+      "learning_rate": 1.2278968930020597e-05,
+      "loss": 0.8942,
+      "step": 2756
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5549913617412394,
+      "learning_rate": 1.227388650902027e-05,
+      "loss": 0.8534,
+      "step": 2757
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4667515157195492,
+      "learning_rate": 1.2268803468550576e-05,
+      "loss": 0.9162,
+      "step": 2758
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.307285162699302,
+      "learning_rate": 1.226371980999628e-05,
+      "loss": 0.8589,
+      "step": 2759
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2235047744533885,
+      "learning_rate": 1.2258635534742307e-05,
+      "loss": 0.8867,
+      "step": 2760
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4302356823990086,
+      "learning_rate": 1.2253550644173753e-05,
+      "loss": 0.9036,
+      "step": 2761
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3079626262402047,
+      "learning_rate": 1.2248465139675886e-05,
+      "loss": 0.8023,
+      "step": 2762
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5588515332610038,
+      "learning_rate": 1.2243379022634136e-05,
+      "loss": 0.8995,
+      "step": 2763
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.40071126527148,
+      "learning_rate": 1.22382922944341e-05,
+      "loss": 0.8275,
+      "step": 2764
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.883921169199151,
+      "learning_rate": 1.2233204956461545e-05,
+      "loss": 0.8628,
+      "step": 2765
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.366785348602125,
+      "learning_rate": 1.22281170101024e-05,
+      "loss": 0.8359,
+      "step": 2766
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4794186690291042,
+      "learning_rate": 1.2223028456742762e-05,
+      "loss": 0.8399,
+      "step": 2767
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2888013828125635,
+      "learning_rate": 1.2217939297768898e-05,
+      "loss": 0.3306,
+      "step": 2768
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.584539053808563,
+      "learning_rate": 1.221284953456723e-05,
+      "loss": 0.7839,
+      "step": 2769
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8881681733473118,
+      "learning_rate": 1.2207759168524352e-05,
+      "loss": 0.3687,
+      "step": 2770
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.354175990367544,
+      "learning_rate": 1.2202668201027016e-05,
+      "loss": 0.8596,
+      "step": 2771
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3194218486878357,
+      "learning_rate": 1.2197576633462147e-05,
+      "loss": 0.8138,
+      "step": 2772
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3180311779953826,
+      "learning_rate": 1.2192484467216828e-05,
+      "loss": 0.9242,
+      "step": 2773
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4526055706114724,
+      "learning_rate": 1.2187391703678301e-05,
+      "loss": 0.8586,
+      "step": 2774
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7478634320520393,
+      "learning_rate": 1.2182298344233981e-05,
+      "loss": 0.3684,
+      "step": 2775
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5846749574800176,
+      "learning_rate": 1.2177204390271434e-05,
+      "loss": 0.8321,
+      "step": 2776
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3337479796613283,
+      "learning_rate": 1.2172109843178396e-05,
+      "loss": 0.8608,
+      "step": 2777
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4888490935247887,
+      "learning_rate": 1.2167014704342763e-05,
+      "loss": 0.861,
+      "step": 2778
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3817411599158573,
+      "learning_rate": 1.2161918975152593e-05,
+      "loss": 0.8004,
+      "step": 2779
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3615329573280277,
+      "learning_rate": 1.2156822656996098e-05,
+      "loss": 0.8266,
+      "step": 2780
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3344758620295176,
+      "learning_rate": 1.2151725751261659e-05,
+      "loss": 0.8528,
+      "step": 2781
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4694329294799107,
+      "learning_rate": 1.2146628259337814e-05,
+      "loss": 0.8931,
+      "step": 2782
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4841911848345541,
+      "learning_rate": 1.2141530182613255e-05,
+      "loss": 0.8026,
+      "step": 2783
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2637370164993806,
+      "learning_rate": 1.2136431522476847e-05,
+      "loss": 0.8731,
+      "step": 2784
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.368588599531654,
+      "learning_rate": 1.2131332280317598e-05,
+      "loss": 0.8911,
+      "step": 2785
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.359812150332946,
+      "learning_rate": 1.2126232457524686e-05,
+      "loss": 0.9365,
+      "step": 2786
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7891008998006404,
+      "learning_rate": 1.2121132055487442e-05,
+      "loss": 0.3736,
+      "step": 2787
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3053082435655867,
+      "learning_rate": 1.2116031075595356e-05,
+      "loss": 0.8867,
+      "step": 2788
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3282828366933468,
+      "learning_rate": 1.2110929519238077e-05,
+      "loss": 0.8157,
+      "step": 2789
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2593385208966215,
+      "learning_rate": 1.2105827387805413e-05,
+      "loss": 0.8379,
+      "step": 2790
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7885035167213441,
+      "learning_rate": 1.2100724682687317e-05,
+      "loss": 0.3385,
+      "step": 2791
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3351968701184085,
+      "learning_rate": 1.2095621405273912e-05,
+      "loss": 0.7924,
+      "step": 2792
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4256215528645497,
+      "learning_rate": 1.209051755695547e-05,
+      "loss": 0.8995,
+      "step": 2793
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3211608988087316,
+      "learning_rate": 1.2085413139122417e-05,
+      "loss": 0.8325,
+      "step": 2794
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.890842614378777,
+      "learning_rate": 1.2080308153165345e-05,
+      "loss": 0.9104,
+      "step": 2795
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4438560525587463,
+      "learning_rate": 1.2075202600474985e-05,
+      "loss": 0.833,
+      "step": 2796
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4223889251345907,
+      "learning_rate": 1.2070096482442235e-05,
+      "loss": 0.9033,
+      "step": 2797
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2045743103943218,
+      "learning_rate": 1.2064989800458138e-05,
+      "loss": 0.8629,
+      "step": 2798
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2210693271663275,
+      "learning_rate": 1.20598825559139e-05,
+      "loss": 0.8976,
+      "step": 2799
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.988542389387899,
+      "learning_rate": 1.205477475020087e-05,
+      "loss": 0.3307,
+      "step": 2800
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3625822133231007,
+      "learning_rate": 1.2049666384710563e-05,
+      "loss": 0.8764,
+      "step": 2801
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.297299460670785,
+      "learning_rate": 1.204455746083463e-05,
+      "loss": 0.855,
+      "step": 2802
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7781711622658598,
+      "learning_rate": 1.2039447979964887e-05,
+      "loss": 0.8376,
+      "step": 2803
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5087258501606984,
+      "learning_rate": 1.2034337943493297e-05,
+      "loss": 0.8322,
+      "step": 2804
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3860452436860717,
+      "learning_rate": 1.202922735281197e-05,
+      "loss": 0.8896,
+      "step": 2805
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3409551752701763,
+      "learning_rate": 1.202411620931318e-05,
+      "loss": 0.8719,
+      "step": 2806
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5286285808560125,
+      "learning_rate": 1.2019004514389338e-05,
+      "loss": 0.9403,
+      "step": 2807
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3945856784969757,
+      "learning_rate": 1.2013892269433017e-05,
+      "loss": 0.843,
+      "step": 2808
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.261472924616925,
+      "learning_rate": 1.2008779475836923e-05,
+      "loss": 0.8935,
+      "step": 2809
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8824725498730404,
+      "learning_rate": 1.2003666134993928e-05,
+      "loss": 0.3443,
+      "step": 2810
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4178017427283562,
+      "learning_rate": 1.199855224829705e-05,
+      "loss": 0.8828,
+      "step": 2811
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3129133658495373,
+      "learning_rate": 1.1993437817139446e-05,
+      "loss": 0.8583,
+      "step": 2812
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7714852622787589,
+      "learning_rate": 1.1988322842914431e-05,
+      "loss": 0.3175,
+      "step": 2813
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4551300257188933,
+      "learning_rate": 1.1983207327015465e-05,
+      "loss": 0.8628,
+      "step": 2814
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8596732356313029,
+      "learning_rate": 1.197809127083616e-05,
+      "loss": 0.3527,
+      "step": 2815
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3962911552192607,
+      "learning_rate": 1.1972974675770259e-05,
+      "loss": 0.845,
+      "step": 2816
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.355512136677426,
+      "learning_rate": 1.1967857543211679e-05,
+      "loss": 0.8325,
+      "step": 2817
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3582930191457832,
+      "learning_rate": 1.1962739874554452e-05,
+      "loss": 0.8088,
+      "step": 2818
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3228360555774272,
+      "learning_rate": 1.1957621671192785e-05,
+      "loss": 0.7442,
+      "step": 2819
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4636040082620687,
+      "learning_rate": 1.195250293452101e-05,
+      "loss": 0.8396,
+      "step": 2820
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3374091590743906,
+      "learning_rate": 1.1947383665933619e-05,
+      "loss": 0.8866,
+      "step": 2821
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.368567384824542,
+      "learning_rate": 1.1942263866825234e-05,
+      "loss": 0.8447,
+      "step": 2822
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4597333832809862,
+      "learning_rate": 1.1937143538590635e-05,
+      "loss": 0.9666,
+      "step": 2823
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5084731126107642,
+      "learning_rate": 1.1932022682624735e-05,
+      "loss": 0.88,
+      "step": 2824
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2923528531161748,
+      "learning_rate": 1.1926901300322601e-05,
+      "loss": 0.9191,
+      "step": 2825
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.366419643305925,
+      "learning_rate": 1.1921779393079438e-05,
+      "loss": 0.8654,
+      "step": 2826
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2065926290347666,
+      "learning_rate": 1.1916656962290594e-05,
+      "loss": 0.8036,
+      "step": 2827
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.450801029286185,
+      "learning_rate": 1.1911534009351561e-05,
+      "loss": 0.8347,
+      "step": 2828
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3182389218440302,
+      "learning_rate": 1.190641053565797e-05,
+      "loss": 0.8147,
+      "step": 2829
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4333985076189057,
+      "learning_rate": 1.1901286542605604e-05,
+      "loss": 0.9358,
+      "step": 2830
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.41093501648415,
+      "learning_rate": 1.1896162031590367e-05,
+      "loss": 0.8542,
+      "step": 2831
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.408105363375565,
+      "learning_rate": 1.189103700400833e-05,
+      "loss": 0.9133,
+      "step": 2832
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.505468807705461,
+      "learning_rate": 1.1885911461255685e-05,
+      "loss": 0.8975,
+      "step": 2833
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8165674962965024,
+      "learning_rate": 1.1880785404728773e-05,
+      "loss": 0.3573,
+      "step": 2834
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3166676524944276,
+      "learning_rate": 1.1875658835824071e-05,
+      "loss": 0.8793,
+      "step": 2835
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3976809040627696,
+      "learning_rate": 1.1870531755938202e-05,
+      "loss": 0.9186,
+      "step": 2836
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7698839741982167,
+      "learning_rate": 1.1865404166467918e-05,
+      "loss": 0.3131,
+      "step": 2837
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3597761677588358,
+      "learning_rate": 1.1860276068810119e-05,
+      "loss": 0.8338,
+      "step": 2838
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2666371332437028,
+      "learning_rate": 1.1855147464361845e-05,
+      "loss": 0.8526,
+      "step": 2839
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3296651941438953,
+      "learning_rate": 1.1850018354520256e-05,
+      "loss": 0.8487,
+      "step": 2840
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2792481340339643,
+      "learning_rate": 1.1844888740682678e-05,
+      "loss": 0.871,
+      "step": 2841
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.310151728484075,
+      "learning_rate": 1.1839758624246551e-05,
+      "loss": 0.9137,
+      "step": 2842
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3247685553482373,
+      "learning_rate": 1.183462800660946e-05,
+      "loss": 0.8951,
+      "step": 2843
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.997448782121293,
+      "learning_rate": 1.182949688916913e-05,
+      "loss": 0.3656,
+      "step": 2844
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2133007811142893,
+      "learning_rate": 1.1824365273323414e-05,
+      "loss": 0.8791,
+      "step": 2845
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.260885700668015,
+      "learning_rate": 1.1819233160470311e-05,
+      "loss": 0.9284,
+      "step": 2846
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3161530895062066,
+      "learning_rate": 1.1814100552007947e-05,
+      "loss": 0.8815,
+      "step": 2847
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5194529645606776,
+      "learning_rate": 1.1808967449334588e-05,
+      "loss": 0.8993,
+      "step": 2848
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4777492410263449,
+      "learning_rate": 1.1803833853848627e-05,
+      "loss": 0.8779,
+      "step": 2849
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4856803186622674,
+      "learning_rate": 1.1798699766948606e-05,
+      "loss": 0.9052,
+      "step": 2850
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6142362736247533,
+      "learning_rate": 1.1793565190033188e-05,
+      "loss": 0.8999,
+      "step": 2851
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2031496141442488,
+      "learning_rate": 1.1788430124501167e-05,
+      "loss": 0.8131,
+      "step": 2852
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4405244382309563,
+      "learning_rate": 1.1783294571751484e-05,
+      "loss": 0.8967,
+      "step": 2853
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2509649917936982,
+      "learning_rate": 1.1778158533183203e-05,
+      "loss": 0.8103,
+      "step": 2854
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9366071601971184,
+      "learning_rate": 1.1773022010195525e-05,
+      "loss": 0.3502,
+      "step": 2855
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4953320471563223,
+      "learning_rate": 1.1767885004187773e-05,
+      "loss": 0.972,
+      "step": 2856
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2945888171521325,
+      "learning_rate": 1.1762747516559418e-05,
+      "loss": 0.8736,
+      "step": 2857
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5723759488085398,
+      "learning_rate": 1.1757609548710048e-05,
+      "loss": 0.8158,
+      "step": 2858
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.371234215499491,
+      "learning_rate": 1.1752471102039385e-05,
+      "loss": 0.8133,
+      "step": 2859
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2716717003390021,
+      "learning_rate": 1.1747332177947288e-05,
+      "loss": 0.8464,
+      "step": 2860
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4914683723700202,
+      "learning_rate": 1.1742192777833746e-05,
+      "loss": 0.8277,
+      "step": 2861
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6590621473272844,
+      "learning_rate": 1.173705290309886e-05,
+      "loss": 0.9288,
+      "step": 2862
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2368147097401632,
+      "learning_rate": 1.1731912555142883e-05,
+      "loss": 0.8414,
+      "step": 2863
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7567428039344742,
+      "learning_rate": 1.1726771735366186e-05,
+      "loss": 0.3492,
+      "step": 2864
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4984309774600924,
+      "learning_rate": 1.172163044516927e-05,
+      "loss": 0.8697,
+      "step": 2865
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7209098206817808,
+      "learning_rate": 1.1716488685952765e-05,
+      "loss": 0.801,
+      "step": 2866
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2571964336013137,
+      "learning_rate": 1.1711346459117423e-05,
+      "loss": 0.8538,
+      "step": 2867
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3619064877090303,
+      "learning_rate": 1.1706203766064137e-05,
+      "loss": 0.8653,
+      "step": 2868
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3813489568063029,
+      "learning_rate": 1.1701060608193911e-05,
+      "loss": 0.8036,
+      "step": 2869
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3611597609702055,
+      "learning_rate": 1.1695916986907882e-05,
+      "loss": 0.8683,
+      "step": 2870
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3006412427892529,
+      "learning_rate": 1.1690772903607321e-05,
+      "loss": 0.9069,
+      "step": 2871
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4821328935671958,
+      "learning_rate": 1.1685628359693617e-05,
+      "loss": 0.8532,
+      "step": 2872
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3597263010366636,
+      "learning_rate": 1.168048335656828e-05,
+      "loss": 0.8981,
+      "step": 2873
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5102332009229082,
+      "learning_rate": 1.1675337895632953e-05,
+      "loss": 0.8841,
+      "step": 2874
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4767862117676795,
+      "learning_rate": 1.1670191978289402e-05,
+      "loss": 0.7675,
+      "step": 2875
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3012494652866793,
+      "learning_rate": 1.166504560593952e-05,
+      "loss": 0.8775,
+      "step": 2876
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2737512871081325,
+      "learning_rate": 1.1659898779985317e-05,
+      "loss": 0.8733,
+      "step": 2877
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7411447706075702,
+      "learning_rate": 1.1654751501828927e-05,
+      "loss": 0.8581,
+      "step": 2878
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2367130907312387,
+      "learning_rate": 1.164960377287262e-05,
+      "loss": 0.7849,
+      "step": 2879
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0886292963218578,
+      "learning_rate": 1.164445559451877e-05,
+      "loss": 0.8253,
+      "step": 2880
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.420603886123083,
+      "learning_rate": 1.1639306968169887e-05,
+      "loss": 0.9012,
+      "step": 2881
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.323879612056229,
+      "learning_rate": 1.1634157895228599e-05,
+      "loss": 0.9256,
+      "step": 2882
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4298347211039335,
+      "learning_rate": 1.1629008377097655e-05,
+      "loss": 0.8795,
+      "step": 2883
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.524826035680562,
+      "learning_rate": 1.1623858415179925e-05,
+      "loss": 0.9387,
+      "step": 2884
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5203140211244757,
+      "learning_rate": 1.16187080108784e-05,
+      "loss": 0.8467,
+      "step": 2885
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.440477101466983,
+      "learning_rate": 1.1613557165596192e-05,
+      "loss": 0.8511,
+      "step": 2886
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1502314657429593,
+      "learning_rate": 1.1608405880736535e-05,
+      "loss": 0.847,
+      "step": 2887
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5765633492803035,
+      "learning_rate": 1.1603254157702782e-05,
+      "loss": 0.751,
+      "step": 2888
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8804045822620761,
+      "learning_rate": 1.1598101997898398e-05,
+      "loss": 0.3381,
+      "step": 2889
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3894535364982075,
+      "learning_rate": 1.1592949402726983e-05,
+      "loss": 0.9192,
+      "step": 2890
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2866985029933335,
+      "learning_rate": 1.1587796373592237e-05,
+      "loss": 0.7986,
+      "step": 2891
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.276677159318935,
+      "learning_rate": 1.1582642911897991e-05,
+      "loss": 0.7888,
+      "step": 2892
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2928498871123901,
+      "learning_rate": 1.1577489019048191e-05,
+      "loss": 0.931,
+      "step": 2893
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.583216413982874,
+      "learning_rate": 1.15723346964469e-05,
+      "loss": 0.8236,
+      "step": 2894
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.347128758995366,
+      "learning_rate": 1.1567179945498297e-05,
+      "loss": 0.8581,
+      "step": 2895
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9166941366586149,
+      "learning_rate": 1.1562024767606674e-05,
+      "loss": 0.3567,
+      "step": 2896
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3206274084508376,
+      "learning_rate": 1.1556869164176447e-05,
+      "loss": 0.904,
+      "step": 2897
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5217995589546103,
+      "learning_rate": 1.1551713136612146e-05,
+      "loss": 0.9046,
+      "step": 2898
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1762235738597384,
+      "learning_rate": 1.1546556686318414e-05,
+      "loss": 0.8637,
+      "step": 2899
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.452796972715515,
+      "learning_rate": 1.1541399814700006e-05,
+      "loss": 0.9107,
+      "step": 2900
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2723111930384068,
+      "learning_rate": 1.1536242523161802e-05,
+      "loss": 0.8783,
+      "step": 2901
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6607513741072362,
+      "learning_rate": 1.1531084813108788e-05,
+      "loss": 0.8718,
+      "step": 2902
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7620576453091051,
+      "learning_rate": 1.1525926685946067e-05,
+      "loss": 0.3074,
+      "step": 2903
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2063257040316104,
+      "learning_rate": 1.1520768143078853e-05,
+      "loss": 0.8267,
+      "step": 2904
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.465755465024223,
+      "learning_rate": 1.1515609185912475e-05,
+      "loss": 0.839,
+      "step": 2905
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2918454337012537,
+      "learning_rate": 1.151044981585238e-05,
+      "loss": 0.8004,
+      "step": 2906
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.305521007082058,
+      "learning_rate": 1.1505290034304117e-05,
+      "loss": 0.8529,
+      "step": 2907
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6960730395904198,
+      "learning_rate": 1.1500129842673358e-05,
+      "loss": 0.8841,
+      "step": 2908
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4101361044768252,
+      "learning_rate": 1.1494969242365875e-05,
+      "loss": 0.8296,
+      "step": 2909
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4833911822746206,
+      "learning_rate": 1.1489808234787565e-05,
+      "loss": 0.8285,
+      "step": 2910
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2565317749191307,
+      "learning_rate": 1.1484646821344421e-05,
+      "loss": 0.8008,
+      "step": 2911
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2667514264572968,
+      "learning_rate": 1.1479485003442564e-05,
+      "loss": 0.8586,
+      "step": 2912
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2456004069011246,
+      "learning_rate": 1.1474322782488203e-05,
+      "loss": 0.8173,
+      "step": 2913
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2875052640631384,
+      "learning_rate": 1.146916015988768e-05,
+      "loss": 0.7914,
+      "step": 2914
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4086299677893321,
+      "learning_rate": 1.1463997137047431e-05,
+      "loss": 0.8541,
+      "step": 2915
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5384610829695098,
+      "learning_rate": 1.1458833715374005e-05,
+      "loss": 0.8569,
+      "step": 2916
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3547685433381567,
+      "learning_rate": 1.1453669896274066e-05,
+      "loss": 0.8702,
+      "step": 2917
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8553072622370183,
+      "learning_rate": 1.1448505681154373e-05,
+      "loss": 0.3674,
+      "step": 2918
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3995818901317785,
+      "learning_rate": 1.1443341071421804e-05,
+      "loss": 0.8661,
+      "step": 2919
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7528879357256242,
+      "learning_rate": 1.1438176068483345e-05,
+      "loss": 0.8364,
+      "step": 2920
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4409132323644518,
+      "learning_rate": 1.1433010673746079e-05,
+      "loss": 0.8208,
+      "step": 2921
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2489985142145699,
+      "learning_rate": 1.1427844888617203e-05,
+      "loss": 0.8914,
+      "step": 2922
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.445665078717719,
+      "learning_rate": 1.1422678714504022e-05,
+      "loss": 0.8446,
+      "step": 2923
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2702705746766858,
+      "learning_rate": 1.1417512152813944e-05,
+      "loss": 0.9024,
+      "step": 2924
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6022261279739576,
+      "learning_rate": 1.1412345204954477e-05,
+      "loss": 0.8765,
+      "step": 2925
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2438470854978412,
+      "learning_rate": 1.140717787233325e-05,
+      "loss": 0.8555,
+      "step": 2926
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5090866838503327,
+      "learning_rate": 1.1402010156357978e-05,
+      "loss": 0.9202,
+      "step": 2927
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2470917696141193,
+      "learning_rate": 1.1396842058436497e-05,
+      "loss": 0.9295,
+      "step": 2928
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8692983746689249,
+      "learning_rate": 1.139167357997673e-05,
+      "loss": 0.3746,
+      "step": 2929
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4676548941045635,
+      "learning_rate": 1.1386504722386719e-05,
+      "loss": 0.8075,
+      "step": 2930
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3342653470134176,
+      "learning_rate": 1.1381335487074605e-05,
+      "loss": 0.9185,
+      "step": 2931
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.408590592325525,
+      "learning_rate": 1.1376165875448628e-05,
+      "loss": 0.8616,
+      "step": 2932
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3362973144132835,
+      "learning_rate": 1.1370995888917129e-05,
+      "loss": 0.8452,
+      "step": 2933
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3735837687091863,
+      "learning_rate": 1.136582552888856e-05,
+      "loss": 0.9264,
+      "step": 2934
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4534493015251444,
+      "learning_rate": 1.136065479677147e-05,
+      "loss": 0.9038,
+      "step": 2935
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4388074740284413,
+      "learning_rate": 1.1355483693974502e-05,
+      "loss": 0.8554,
+      "step": 2936
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.438160769054234,
+      "learning_rate": 1.1350312221906416e-05,
+      "loss": 0.8509,
+      "step": 2937
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4130747747870107,
+      "learning_rate": 1.1345140381976054e-05,
+      "loss": 0.851,
+      "step": 2938
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3303468400572487,
+      "learning_rate": 1.1339968175592377e-05,
+      "loss": 0.8818,
+      "step": 2939
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2382521904204054,
+      "learning_rate": 1.1334795604164432e-05,
+      "loss": 0.8361,
+      "step": 2940
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8692931512876478,
+      "learning_rate": 1.132962266910137e-05,
+      "loss": 0.3447,
+      "step": 2941
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3881043563485997,
+      "learning_rate": 1.1324449371812441e-05,
+      "loss": 0.8167,
+      "step": 2942
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5907361310975605,
+      "learning_rate": 1.1319275713706997e-05,
+      "loss": 0.8553,
+      "step": 2943
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3510955861245566,
+      "learning_rate": 1.1314101696194479e-05,
+      "loss": 0.8525,
+      "step": 2944
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6029627081359954,
+      "learning_rate": 1.1308927320684438e-05,
+      "loss": 0.8692,
+      "step": 2945
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4410630563192721,
+      "learning_rate": 1.1303752588586512e-05,
+      "loss": 0.813,
+      "step": 2946
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.619421401947289,
+      "learning_rate": 1.1298577501310444e-05,
+      "loss": 0.8304,
+      "step": 2947
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.22101609719287,
+      "learning_rate": 1.1293402060266072e-05,
+      "loss": 0.8797,
+      "step": 2948
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3283268859233963,
+      "learning_rate": 1.128822626686332e-05,
+      "loss": 0.8684,
+      "step": 2949
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3459190436272004,
+      "learning_rate": 1.1283050122512227e-05,
+      "loss": 0.8993,
+      "step": 2950
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5610107572699294,
+      "learning_rate": 1.127787362862291e-05,
+      "loss": 0.8363,
+      "step": 2951
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5798116684447117,
+      "learning_rate": 1.1272696786605595e-05,
+      "loss": 0.9311,
+      "step": 2952
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2809505868761126,
+      "learning_rate": 1.1267519597870591e-05,
+      "loss": 0.8554,
+      "step": 2953
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3269061119256682,
+      "learning_rate": 1.126234206382831e-05,
+      "loss": 0.9096,
+      "step": 2954
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4759068111794678,
+      "learning_rate": 1.1257164185889255e-05,
+      "loss": 0.8255,
+      "step": 2955
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2955341089289438,
+      "learning_rate": 1.1251985965464022e-05,
+      "loss": 0.8678,
+      "step": 2956
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5321555367340742,
+      "learning_rate": 1.12468074039633e-05,
+      "loss": 0.8853,
+      "step": 2957
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.468687954058805,
+      "learning_rate": 1.1241628502797872e-05,
+      "loss": 0.8495,
+      "step": 2958
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2278376172975625,
+      "learning_rate": 1.1236449263378617e-05,
+      "loss": 0.8643,
+      "step": 2959
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4976612381760017,
+      "learning_rate": 1.1231269687116495e-05,
+      "loss": 0.8688,
+      "step": 2960
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3245385024466307,
+      "learning_rate": 1.1226089775422575e-05,
+      "loss": 0.8985,
+      "step": 2961
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.343736999886859,
+      "learning_rate": 1.1220909529707999e-05,
+      "loss": 0.952,
+      "step": 2962
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3539183404231865,
+      "learning_rate": 1.1215728951384013e-05,
+      "loss": 0.8531,
+      "step": 2963
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.297294036975661,
+      "learning_rate": 1.121054804186195e-05,
+      "loss": 0.8817,
+      "step": 2964
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3092166144211002,
+      "learning_rate": 1.1205366802553231e-05,
+      "loss": 0.8348,
+      "step": 2965
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.630493325138277,
+      "learning_rate": 1.1200185234869372e-05,
+      "loss": 0.8637,
+      "step": 2966
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3499751368248618,
+      "learning_rate": 1.1195003340221968e-05,
+      "loss": 0.8746,
+      "step": 2967
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4052956925963385,
+      "learning_rate": 1.1189821120022712e-05,
+      "loss": 0.8543,
+      "step": 2968
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4099089380534122,
+      "learning_rate": 1.1184638575683388e-05,
+      "loss": 0.8621,
+      "step": 2969
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4528705851044637,
+      "learning_rate": 1.1179455708615863e-05,
+      "loss": 0.8594,
+      "step": 2970
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.496986509989778,
+      "learning_rate": 1.1174272520232087e-05,
+      "loss": 0.8853,
+      "step": 2971
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3505102772183077,
+      "learning_rate": 1.1169089011944113e-05,
+      "loss": 0.8235,
+      "step": 2972
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4212502276750465,
+      "learning_rate": 1.116390518516406e-05,
+      "loss": 0.8558,
+      "step": 2973
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7726388038106262,
+      "learning_rate": 1.1158721041304155e-05,
+      "loss": 0.8839,
+      "step": 2974
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4986555163641022,
+      "learning_rate": 1.1153536581776697e-05,
+      "loss": 0.8677,
+      "step": 2975
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3681541104877117,
+      "learning_rate": 1.1148351807994075e-05,
+      "loss": 0.8653,
+      "step": 2976
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2996499989130925,
+      "learning_rate": 1.114316672136877e-05,
+      "loss": 0.8279,
+      "step": 2977
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3546121801295268,
+      "learning_rate": 1.1137981323313335e-05,
+      "loss": 0.8361,
+      "step": 2978
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4983547583141477,
+      "learning_rate": 1.1132795615240419e-05,
+      "loss": 0.8882,
+      "step": 2979
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.303788976441012,
+      "learning_rate": 1.1127609598562754e-05,
+      "loss": 0.8003,
+      "step": 2980
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7365332612079538,
+      "learning_rate": 1.1122423274693152e-05,
+      "loss": 0.8311,
+      "step": 2981
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4166608405928396,
+      "learning_rate": 1.1117236645044506e-05,
+      "loss": 0.8556,
+      "step": 2982
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6934813265242583,
+      "learning_rate": 1.1112049711029807e-05,
+      "loss": 0.9303,
+      "step": 2983
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5709518855600877,
+      "learning_rate": 1.1106862474062107e-05,
+      "loss": 0.9316,
+      "step": 2984
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.857608599957514,
+      "learning_rate": 1.1101674935554561e-05,
+      "loss": 0.3559,
+      "step": 2985
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4075655975510553,
+      "learning_rate": 1.1096487096920394e-05,
+      "loss": 0.8147,
+      "step": 2986
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2997814902937594,
+      "learning_rate": 1.1091298959572919e-05,
+      "loss": 0.7365,
+      "step": 2987
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.40202356909792,
+      "learning_rate": 1.1086110524925524e-05,
+      "loss": 0.8973,
+      "step": 2988
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5691812103090126,
+      "learning_rate": 1.1080921794391684e-05,
+      "loss": 0.8815,
+      "step": 2989
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2316265260256627,
+      "learning_rate": 1.1075732769384948e-05,
+      "loss": 0.8537,
+      "step": 2990
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4167117295449394,
+      "learning_rate": 1.1070543451318956e-05,
+      "loss": 0.8762,
+      "step": 2991
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5047278896943974,
+      "learning_rate": 1.106535384160742e-05,
+      "loss": 0.8648,
+      "step": 2992
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3259441288324516,
+      "learning_rate": 1.1060163941664125e-05,
+      "loss": 0.8455,
+      "step": 2993
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8233852058613227,
+      "learning_rate": 1.1054973752902956e-05,
+      "loss": 0.3473,
+      "step": 2994
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.607580340513171,
+      "learning_rate": 1.1049783276737852e-05,
+      "loss": 0.8417,
+      "step": 2995
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4068987224253178,
+      "learning_rate": 1.1044592514582843e-05,
+      "loss": 0.859,
+      "step": 2996
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7232248174610592,
+      "learning_rate": 1.1039401467852044e-05,
+      "loss": 0.8699,
+      "step": 2997
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8293732241875293,
+      "learning_rate": 1.1034210137959632e-05,
+      "loss": 0.344,
+      "step": 2998
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3692938185626327,
+      "learning_rate": 1.102901852631987e-05,
+      "loss": 0.8968,
+      "step": 2999
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4805334050018335,
+      "learning_rate": 1.1023826634347096e-05,
+      "loss": 0.8902,
+      "step": 3000
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5018954613210944,
+      "learning_rate": 1.1018634463455727e-05,
+      "loss": 0.7854,
+      "step": 3001
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8211642920579347,
+      "learning_rate": 1.101344201506025e-05,
+      "loss": 0.8621,
+      "step": 3002
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.312300602349726,
+      "learning_rate": 1.1008249290575232e-05,
+      "loss": 0.8464,
+      "step": 3003
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5266064494775065,
+      "learning_rate": 1.1003056291415313e-05,
+      "loss": 0.8481,
+      "step": 3004
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.424324660723558,
+      "learning_rate": 1.0997863018995214e-05,
+      "loss": 0.9076,
+      "step": 3005
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3320157303283566,
+      "learning_rate": 1.0992669474729723e-05,
+      "loss": 0.7934,
+      "step": 3006
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3626655538184695,
+      "learning_rate": 1.09874756600337e-05,
+      "loss": 0.8851,
+      "step": 3007
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2323640459914589,
+      "learning_rate": 1.0982281576322091e-05,
+      "loss": 0.8774,
+      "step": 3008
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2549033753309693,
+      "learning_rate": 1.09770872250099e-05,
+      "loss": 0.839,
+      "step": 3009
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3773540758523743,
+      "learning_rate": 1.097189260751222e-05,
+      "loss": 0.8842,
+      "step": 3010
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4174115272284709,
+      "learning_rate": 1.09666977252442e-05,
+      "loss": 0.7974,
+      "step": 3011
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7896154745205465,
+      "learning_rate": 1.0961502579621073e-05,
+      "loss": 0.3597,
+      "step": 3012
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5416169286745272,
+      "learning_rate": 1.0956307172058138e-05,
+      "loss": 0.917,
+      "step": 3013
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2237705760529862,
+      "learning_rate": 1.0951111503970771e-05,
+      "loss": 0.8209,
+      "step": 3014
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2416210967527441,
+      "learning_rate": 1.0945915576774411e-05,
+      "loss": 0.8677,
+      "step": 3015
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.411229555978256,
+      "learning_rate": 1.0940719391884578e-05,
+      "loss": 0.8517,
+      "step": 3016
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2529346528404601,
+      "learning_rate": 1.0935522950716847e-05,
+      "loss": 0.9088,
+      "step": 3017
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3785350836290926,
+      "learning_rate": 1.0930326254686877e-05,
+      "loss": 0.8228,
+      "step": 3018
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4542644593947307,
+      "learning_rate": 1.0925129305210393e-05,
+      "loss": 0.8685,
+      "step": 3019
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6380241157234026,
+      "learning_rate": 1.0919932103703178e-05,
+      "loss": 0.8182,
+      "step": 3020
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2486095574133431,
+      "learning_rate": 1.091473465158111e-05,
+      "loss": 0.8835,
+      "step": 3021
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.381331825300475,
+      "learning_rate": 1.0909536950260102e-05,
+      "loss": 0.9239,
+      "step": 3022
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4046452213582474,
+      "learning_rate": 1.0904339001156158e-05,
+      "loss": 0.9335,
+      "step": 3023
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2256137441999007,
+      "learning_rate": 1.0899140805685342e-05,
+      "loss": 0.8673,
+      "step": 3024
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2910098722823145,
+      "learning_rate": 1.0893942365263786e-05,
+      "loss": 0.8909,
+      "step": 3025
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7420881813974156,
+      "learning_rate": 1.0888743681307686e-05,
+      "loss": 0.9179,
+      "step": 3026
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.333959480098031,
+      "learning_rate": 1.0883544755233313e-05,
+      "loss": 0.8248,
+      "step": 3027
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.586648918889774,
+      "learning_rate": 1.087834558845699e-05,
+      "loss": 0.8658,
+      "step": 3028
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.284939605633235,
+      "learning_rate": 1.0873146182395118e-05,
+      "loss": 0.8498,
+      "step": 3029
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5124325674912475,
+      "learning_rate": 1.086794653846416e-05,
+      "loss": 0.8567,
+      "step": 3030
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4019648420583464,
+      "learning_rate": 1.0862746658080635e-05,
+      "loss": 0.8788,
+      "step": 3031
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.580635005674901,
+      "learning_rate": 1.085754654266114e-05,
+      "loss": 0.7656,
+      "step": 3032
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.166470229364124,
+      "learning_rate": 1.0852346193622332e-05,
+      "loss": 0.8374,
+      "step": 3033
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5701394851701387,
+      "learning_rate": 1.0847145612380922e-05,
+      "loss": 0.9083,
+      "step": 3034
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3089514385850622,
+      "learning_rate": 1.0841944800353696e-05,
+      "loss": 0.8856,
+      "step": 3035
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.444316081077573,
+      "learning_rate": 1.08367437589575e-05,
+      "loss": 0.8077,
+      "step": 3036
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9166723969354483,
+      "learning_rate": 1.083154248960924e-05,
+      "loss": 0.3726,
+      "step": 3037
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5668749227036867,
+      "learning_rate": 1.0826340993725882e-05,
+      "loss": 0.8868,
+      "step": 3038
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.461915218120926,
+      "learning_rate": 1.0821139272724457e-05,
+      "loss": 0.8095,
+      "step": 3039
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3578788523270395,
+      "learning_rate": 1.0815937328022061e-05,
+      "loss": 0.9192,
+      "step": 3040
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1894804610006675,
+      "learning_rate": 1.0810735161035848e-05,
+      "loss": 0.8428,
+      "step": 3041
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7331824111646426,
+      "learning_rate": 1.0805532773183022e-05,
+      "loss": 0.8325,
+      "step": 3042
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4538021086733113,
+      "learning_rate": 1.0800330165880872e-05,
+      "loss": 0.8545,
+      "step": 3043
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3360087497015196,
+      "learning_rate": 1.0795127340546718e-05,
+      "loss": 0.8988,
+      "step": 3044
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5260827250349465,
+      "learning_rate": 1.0789924298597963e-05,
+      "loss": 0.8645,
+      "step": 3045
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.235726106354246,
+      "learning_rate": 1.0784721041452054e-05,
+      "loss": 0.8812,
+      "step": 3046
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4751178029164163,
+      "learning_rate": 1.07795175705265e-05,
+      "loss": 0.8403,
+      "step": 3047
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.553100274117524,
+      "learning_rate": 1.0774313887238874e-05,
+      "loss": 0.832,
+      "step": 3048
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7974727771364543,
+      "learning_rate": 1.0769109993006802e-05,
+      "loss": 0.7906,
+      "step": 3049
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3385744331266738,
+      "learning_rate": 1.0763905889247964e-05,
+      "loss": 0.8528,
+      "step": 3050
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2120850666797633,
+      "learning_rate": 1.0758701577380107e-05,
+      "loss": 0.8312,
+      "step": 3051
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2644943600030492,
+      "learning_rate": 1.0753497058821027e-05,
+      "loss": 0.8489,
+      "step": 3052
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.52135022910106,
+      "learning_rate": 1.0748292334988573e-05,
+      "loss": 0.86,
+      "step": 3053
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2692066026951876,
+      "learning_rate": 1.0743087407300665e-05,
+      "loss": 0.8497,
+      "step": 3054
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2573178233752513,
+      "learning_rate": 1.0737882277175262e-05,
+      "loss": 0.7725,
+      "step": 3055
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3798735119505534,
+      "learning_rate": 1.073267694603039e-05,
+      "loss": 0.9252,
+      "step": 3056
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.603363245926989,
+      "learning_rate": 1.0727471415284119e-05,
+      "loss": 0.927,
+      "step": 3057
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4765345751175483,
+      "learning_rate": 1.072226568635458e-05,
+      "loss": 0.8639,
+      "step": 3058
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6139136049533478,
+      "learning_rate": 1.0717059760659963e-05,
+      "loss": 0.8781,
+      "step": 3059
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5698912907298164,
+      "learning_rate": 1.0711853639618497e-05,
+      "loss": 0.8643,
+      "step": 3060
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3785351653753097,
+      "learning_rate": 1.0706647324648481e-05,
+      "loss": 0.796,
+      "step": 3061
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4642440445685065,
+      "learning_rate": 1.0701440817168251e-05,
+      "loss": 0.8572,
+      "step": 3062
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.219279837177049,
+      "learning_rate": 1.069623411859621e-05,
+      "loss": 0.8526,
+      "step": 3063
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.185054528347692,
+      "learning_rate": 1.0691027230350802e-05,
+      "loss": 0.8926,
+      "step": 3064
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2715295089234997,
+      "learning_rate": 1.0685820153850528e-05,
+      "loss": 0.8697,
+      "step": 3065
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.523485082714657,
+      "learning_rate": 1.0680612890513937e-05,
+      "loss": 0.9524,
+      "step": 3066
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4532688374753664,
+      "learning_rate": 1.0675405441759639e-05,
+      "loss": 0.858,
+      "step": 3067
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8122004899145338,
+      "learning_rate": 1.0670197809006277e-05,
+      "loss": 0.3511,
+      "step": 3068
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3301169183930275,
+      "learning_rate": 1.0664989993672559e-05,
+      "loss": 0.8247,
+      "step": 3069
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5584257986647938,
+      "learning_rate": 1.0659781997177239e-05,
+      "loss": 0.8551,
+      "step": 3070
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.309450681713823,
+      "learning_rate": 1.0654573820939112e-05,
+      "loss": 0.8529,
+      "step": 3071
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.272414766529403,
+      "learning_rate": 1.0649365466377033e-05,
+      "loss": 0.8431,
+      "step": 3072
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5019277346018816,
+      "learning_rate": 1.0644156934909902e-05,
+      "loss": 0.8873,
+      "step": 3073
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3930023571737786,
+      "learning_rate": 1.0638948227956665e-05,
+      "loss": 0.892,
+      "step": 3074
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.420408759341007,
+      "learning_rate": 1.0633739346936318e-05,
+      "loss": 0.8424,
+      "step": 3075
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5070115411704652,
+      "learning_rate": 1.0628530293267908e-05,
+      "loss": 0.8681,
+      "step": 3076
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5695336531245214,
+      "learning_rate": 1.0623321068370515e-05,
+      "loss": 0.8613,
+      "step": 3077
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3391037550427791,
+      "learning_rate": 1.0618111673663283e-05,
+      "loss": 0.9225,
+      "step": 3078
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2236615160097064,
+      "learning_rate": 1.0612902110565393e-05,
+      "loss": 0.9126,
+      "step": 3079
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.469619887780148,
+      "learning_rate": 1.060769238049607e-05,
+      "loss": 0.9166,
+      "step": 3080
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3387055323788486,
+      "learning_rate": 1.0602482484874598e-05,
+      "loss": 0.9164,
+      "step": 3081
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2345736610072682,
+      "learning_rate": 1.0597272425120286e-05,
+      "loss": 0.9115,
+      "step": 3082
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2540792441688027,
+      "learning_rate": 1.05920622026525e-05,
+      "loss": 0.8786,
+      "step": 3083
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4995320379929469,
+      "learning_rate": 1.0586851818890651e-05,
+      "loss": 0.8833,
+      "step": 3084
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3356992198137094,
+      "learning_rate": 1.058164127525419e-05,
+      "loss": 0.8919,
+      "step": 3085
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2781278831127638,
+      "learning_rate": 1.0576430573162612e-05,
+      "loss": 0.8647,
+      "step": 3086
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4111650351043021,
+      "learning_rate": 1.057121971403546e-05,
+      "loss": 0.8607,
+      "step": 3087
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7572311404275531,
+      "learning_rate": 1.0566008699292307e-05,
+      "loss": 0.8936,
+      "step": 3088
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1955921291687412,
+      "learning_rate": 1.0560797530352784e-05,
+      "loss": 0.8447,
+      "step": 3089
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2750046395703591,
+      "learning_rate": 1.0555586208636557e-05,
+      "loss": 0.8708,
+      "step": 3090
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3163310122850227,
+      "learning_rate": 1.0550374735563329e-05,
+      "loss": 0.8144,
+      "step": 3091
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3811984043927703,
+      "learning_rate": 1.0545163112552856e-05,
+      "loss": 0.8669,
+      "step": 3092
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5005440675225892,
+      "learning_rate": 1.053995134102492e-05,
+      "loss": 0.8513,
+      "step": 3093
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1969314959951984,
+      "learning_rate": 1.0534739422399357e-05,
+      "loss": 0.9069,
+      "step": 3094
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3382550662063022,
+      "learning_rate": 1.0529527358096035e-05,
+      "loss": 0.8344,
+      "step": 3095
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2396128741786128,
+      "learning_rate": 1.052431514953486e-05,
+      "loss": 0.893,
+      "step": 3096
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4395120911444843,
+      "learning_rate": 1.051910279813579e-05,
+      "loss": 0.8691,
+      "step": 3097
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5309220462779471,
+      "learning_rate": 1.0513890305318808e-05,
+      "loss": 0.825,
+      "step": 3098
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.477627941894886,
+      "learning_rate": 1.0508677672503942e-05,
+      "loss": 0.8173,
+      "step": 3099
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3164822474426208,
+      "learning_rate": 1.0503464901111253e-05,
+      "loss": 0.9002,
+      "step": 3100
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.558687143348059,
+      "learning_rate": 1.0498251992560851e-05,
+      "loss": 0.8533,
+      "step": 3101
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3918851898004079,
+      "learning_rate": 1.0493038948272866e-05,
+      "loss": 0.899,
+      "step": 3102
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4231218704380626,
+      "learning_rate": 1.0487825769667489e-05,
+      "loss": 0.8851,
+      "step": 3103
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.21505158185986,
+      "learning_rate": 1.0482612458164918e-05,
+      "loss": 0.8337,
+      "step": 3104
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2429226860475722,
+      "learning_rate": 1.0477399015185415e-05,
+      "loss": 0.8279,
+      "step": 3105
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9001127551768887,
+      "learning_rate": 1.0472185442149257e-05,
+      "loss": 0.3617,
+      "step": 3106
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3572454488804095,
+      "learning_rate": 1.046697174047677e-05,
+      "loss": 0.8517,
+      "step": 3107
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3877297510575701,
+      "learning_rate": 1.0461757911588309e-05,
+      "loss": 0.8241,
+      "step": 3108
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4030912606573163,
+      "learning_rate": 1.0456543956904266e-05,
+      "loss": 0.8886,
+      "step": 3109
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.497197921500087,
+      "learning_rate": 1.0451329877845062e-05,
+      "loss": 0.8048,
+      "step": 3110
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.489961032871929,
+      "learning_rate": 1.0446115675831159e-05,
+      "loss": 0.7885,
+      "step": 3111
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4814037120387153,
+      "learning_rate": 1.044090135228305e-05,
+      "loss": 0.829,
+      "step": 3112
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2720968906994399,
+      "learning_rate": 1.0435686908621256e-05,
+      "loss": 0.8484,
+      "step": 3113
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3388453620227982,
+      "learning_rate": 1.0430472346266343e-05,
+      "loss": 0.831,
+      "step": 3114
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.36215736864177,
+      "learning_rate": 1.0425257666638891e-05,
+      "loss": 0.8744,
+      "step": 3115
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3694964305707824,
+      "learning_rate": 1.0420042871159532e-05,
+      "loss": 0.8057,
+      "step": 3116
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6054553500658117,
+      "learning_rate": 1.0414827961248917e-05,
+      "loss": 0.9106,
+      "step": 3117
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1922218942899219,
+      "learning_rate": 1.040961293832773e-05,
+      "loss": 0.8907,
+      "step": 3118
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2974136717474913,
+      "learning_rate": 1.0404397803816686e-05,
+      "loss": 0.8941,
+      "step": 3119
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.55441616256837,
+      "learning_rate": 1.0399182559136536e-05,
+      "loss": 0.8465,
+      "step": 3120
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2929329906802793,
+      "learning_rate": 1.0393967205708049e-05,
+      "loss": 0.7521,
+      "step": 3121
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.470199296020881,
+      "learning_rate": 1.0388751744952037e-05,
+      "loss": 0.8495,
+      "step": 3122
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6260963166730618,
+      "learning_rate": 1.0383536178289336e-05,
+      "loss": 0.876,
+      "step": 3123
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4356338325425162,
+      "learning_rate": 1.0378320507140804e-05,
+      "loss": 0.7491,
+      "step": 3124
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0090143629780062,
+      "learning_rate": 1.0373104732927341e-05,
+      "loss": 0.3394,
+      "step": 3125
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8160227891444077,
+      "learning_rate": 1.036788885706986e-05,
+      "loss": 0.3514,
+      "step": 3126
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.996273928458851,
+      "learning_rate": 1.0362672880989317e-05,
+      "loss": 0.9005,
+      "step": 3127
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2449916088130744,
+      "learning_rate": 1.0357456806106684e-05,
+      "loss": 0.9654,
+      "step": 3128
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3419406548694788,
+      "learning_rate": 1.0352240633842962e-05,
+      "loss": 0.8573,
+      "step": 3129
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2650574674411519,
+      "learning_rate": 1.0347024365619183e-05,
+      "loss": 0.8852,
+      "step": 3130
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6728382321621054,
+      "learning_rate": 1.0341808002856402e-05,
+      "loss": 0.8419,
+      "step": 3131
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.399009813244773,
+      "learning_rate": 1.0336591546975695e-05,
+      "loss": 0.8571,
+      "step": 3132
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3006703809938753,
+      "learning_rate": 1.0331374999398175e-05,
+      "loss": 0.8225,
+      "step": 3133
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3331187596803233,
+      "learning_rate": 1.0326158361544971e-05,
+      "loss": 0.8225,
+      "step": 3134
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3008551275907787,
+      "learning_rate": 1.0320941634837238e-05,
+      "loss": 0.7968,
+      "step": 3135
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.18041499865668,
+      "learning_rate": 1.031572482069616e-05,
+      "loss": 0.9243,
+      "step": 3136
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3027629863451,
+      "learning_rate": 1.0310507920542932e-05,
+      "loss": 0.886,
+      "step": 3137
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.23192960668184,
+      "learning_rate": 1.0305290935798794e-05,
+      "loss": 0.7809,
+      "step": 3138
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2739663324156287,
+      "learning_rate": 1.0300073867884983e-05,
+      "loss": 0.978,
+      "step": 3139
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.291261864464104,
+      "learning_rate": 1.0294856718222782e-05,
+      "loss": 0.8875,
+      "step": 3140
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2803623989107642,
+      "learning_rate": 1.0289639488233484e-05,
+      "loss": 0.8812,
+      "step": 3141
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2952654326714508,
+      "learning_rate": 1.0284422179338404e-05,
+      "loss": 0.798,
+      "step": 3142
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.460529664750074,
+      "learning_rate": 1.0279204792958883e-05,
+      "loss": 0.8852,
+      "step": 3143
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4179800016046116,
+      "learning_rate": 1.0273987330516279e-05,
+      "loss": 0.7886,
+      "step": 3144
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1754818904342017,
+      "learning_rate": 1.0268769793431975e-05,
+      "loss": 0.3502,
+      "step": 3145
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3339787136812287,
+      "learning_rate": 1.026355218312737e-05,
+      "loss": 0.815,
+      "step": 3146
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2073814148820599,
+      "learning_rate": 1.0258334501023886e-05,
+      "loss": 0.822,
+      "step": 3147
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2465137839546458,
+      "learning_rate": 1.025311674854296e-05,
+      "loss": 0.8075,
+      "step": 3148
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2636057749445282,
+      "learning_rate": 1.0247898927106058e-05,
+      "loss": 0.802,
+      "step": 3149
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4012301015759687,
+      "learning_rate": 1.0242681038134654e-05,
+      "loss": 0.8936,
+      "step": 3150
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3593550686087572,
+      "learning_rate": 1.0237463083050246e-05,
+      "loss": 0.8381,
+      "step": 3151
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2268692771142444,
+      "learning_rate": 1.0232245063274349e-05,
+      "loss": 0.8783,
+      "step": 3152
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5626147728728654,
+      "learning_rate": 1.0227026980228492e-05,
+      "loss": 0.9065,
+      "step": 3153
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.149314831118315,
+      "learning_rate": 1.022180883533423e-05,
+      "loss": 0.8502,
+      "step": 3154
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3129715856368063,
+      "learning_rate": 1.0216590630013125e-05,
+      "loss": 0.8691,
+      "step": 3155
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2882633282297973,
+      "learning_rate": 1.0211372365686763e-05,
+      "loss": 0.8097,
+      "step": 3156
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2599390612837662,
+      "learning_rate": 1.0206154043776741e-05,
+      "loss": 0.8594,
+      "step": 3157
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3250497275593773,
+      "learning_rate": 1.0200935665704679e-05,
+      "loss": 0.8597,
+      "step": 3158
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3176112630834818,
+      "learning_rate": 1.0195717232892198e-05,
+      "loss": 0.8682,
+      "step": 3159
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4073732519366353,
+      "learning_rate": 1.0190498746760951e-05,
+      "loss": 0.8688,
+      "step": 3160
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6235617189457237,
+      "learning_rate": 1.0185280208732594e-05,
+      "loss": 0.8239,
+      "step": 3161
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4806289296824688,
+      "learning_rate": 1.0180061620228799e-05,
+      "loss": 0.8933,
+      "step": 3162
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.359189705305406,
+      "learning_rate": 1.0174842982671258e-05,
+      "loss": 0.8886,
+      "step": 3163
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6146281729645078,
+      "learning_rate": 1.0169624297481663e-05,
+      "loss": 0.8727,
+      "step": 3164
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3418625794000443,
+      "learning_rate": 1.016440556608174e-05,
+      "loss": 0.8618,
+      "step": 3165
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.533684622729243,
+      "learning_rate": 1.0159186789893207e-05,
+      "loss": 0.824,
+      "step": 3166
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4036490732862157,
+      "learning_rate": 1.0153967970337803e-05,
+      "loss": 0.875,
+      "step": 3167
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2541933057919359,
+      "learning_rate": 1.0148749108837282e-05,
+      "loss": 0.8879,
+      "step": 3168
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4148761880622192,
+      "learning_rate": 1.0143530206813403e-05,
+      "loss": 0.8635,
+      "step": 3169
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4013081823945985,
+      "learning_rate": 1.013831126568794e-05,
+      "loss": 0.8664,
+      "step": 3170
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.460869006601472,
+      "learning_rate": 1.0133092286882672e-05,
+      "loss": 0.8606,
+      "step": 3171
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4625063813001031,
+      "learning_rate": 1.01278732718194e-05,
+      "loss": 0.8937,
+      "step": 3172
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.409921717755494,
+      "learning_rate": 1.0122654221919921e-05,
+      "loss": 0.8269,
+      "step": 3173
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3715806114322515,
+      "learning_rate": 1.0117435138606054e-05,
+      "loss": 0.861,
+      "step": 3174
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4017339106958044,
+      "learning_rate": 1.0112216023299615e-05,
+      "loss": 0.8726,
+      "step": 3175
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2811030857536037,
+      "learning_rate": 1.0106996877422442e-05,
+      "loss": 0.8857,
+      "step": 3176
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3252947516299698,
+      "learning_rate": 1.0101777702396367e-05,
+      "loss": 0.8071,
+      "step": 3177
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4938043702178325,
+      "learning_rate": 1.0096558499643239e-05,
+      "loss": 0.8945,
+      "step": 3178
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3124407559034774,
+      "learning_rate": 1.0091339270584917e-05,
+      "loss": 0.9057,
+      "step": 3179
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3544737306477355,
+      "learning_rate": 1.008612001664326e-05,
+      "loss": 0.8952,
+      "step": 3180
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6120500993195108,
+      "learning_rate": 1.0080900739240136e-05,
+      "loss": 0.8759,
+      "step": 3181
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6284267917148112,
+      "learning_rate": 1.007568143979742e-05,
+      "loss": 0.8224,
+      "step": 3182
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5481018417614492,
+      "learning_rate": 1.0070462119736993e-05,
+      "loss": 0.7742,
+      "step": 3183
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3091158252467596,
+      "learning_rate": 1.0065242780480742e-05,
+      "loss": 0.8715,
+      "step": 3184
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3840233958393622,
+      "learning_rate": 1.0060023423450562e-05,
+      "loss": 0.8413,
+      "step": 3185
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2993453401413937,
+      "learning_rate": 1.0054804050068343e-05,
+      "loss": 0.8938,
+      "step": 3186
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4014582454857538,
+      "learning_rate": 1.0049584661755993e-05,
+      "loss": 0.9122,
+      "step": 3187
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3691916755540423,
+      "learning_rate": 1.0044365259935413e-05,
+      "loss": 0.8346,
+      "step": 3188
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5906937041227802,
+      "learning_rate": 1.0039145846028515e-05,
+      "loss": 0.8847,
+      "step": 3189
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4258281974379026,
+      "learning_rate": 1.0033926421457208e-05,
+      "loss": 0.8351,
+      "step": 3190
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5022129440543395,
+      "learning_rate": 1.002870698764341e-05,
+      "loss": 0.7858,
+      "step": 3191
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2733292248395236,
+      "learning_rate": 1.0023487546009036e-05,
+      "loss": 0.8423,
+      "step": 3192
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3213549433296046,
+      "learning_rate": 1.0018268097976007e-05,
+      "loss": 0.8121,
+      "step": 3193
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4159834154346203,
+      "learning_rate": 1.0013048644966246e-05,
+      "loss": 0.8386,
+      "step": 3194
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4920112720501866,
+      "learning_rate": 1.0007829188401673e-05,
+      "loss": 0.8227,
+      "step": 3195
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4170278371195022,
+      "learning_rate": 1.0002609729704213e-05,
+      "loss": 0.9145,
+      "step": 3196
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6737666419231418,
+      "learning_rate": 9.997390270295792e-06,
+      "loss": 0.8728,
+      "step": 3197
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5829280813189386,
+      "learning_rate": 9.992170811598332e-06,
+      "loss": 0.8802,
+      "step": 3198
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.93074717171541,
+      "learning_rate": 9.986951355033755e-06,
+      "loss": 0.9011,
+      "step": 3199
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.414408144937566,
+      "learning_rate": 9.981731902023998e-06,
+      "loss": 0.8614,
+      "step": 3200
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3214095022706858,
+      "learning_rate": 9.976512453990967e-06,
+      "loss": 0.8053,
+      "step": 3201
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.279115182772325,
+      "learning_rate": 9.971293012356593e-06,
+      "loss": 0.8094,
+      "step": 3202
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3226841759986698,
+      "learning_rate": 9.966073578542795e-06,
+      "loss": 0.8266,
+      "step": 3203
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2711641476789992,
+      "learning_rate": 9.960854153971488e-06,
+      "loss": 0.8602,
+      "step": 3204
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.26318410505499,
+      "learning_rate": 9.955634740064588e-06,
+      "loss": 0.8106,
+      "step": 3205
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3253027282787253,
+      "learning_rate": 9.95041533824401e-06,
+      "loss": 0.8908,
+      "step": 3206
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4176140941351614,
+      "learning_rate": 9.94519594993166e-06,
+      "loss": 0.8178,
+      "step": 3207
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3476219552970083,
+      "learning_rate": 9.939976576549441e-06,
+      "loss": 0.8324,
+      "step": 3208
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2259315580913956,
+      "learning_rate": 9.934757219519258e-06,
+      "loss": 0.8626,
+      "step": 3209
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.451424786514718,
+      "learning_rate": 9.92953788026301e-06,
+      "loss": 0.874,
+      "step": 3210
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.129632753390078,
+      "learning_rate": 9.924318560202584e-06,
+      "loss": 0.3413,
+      "step": 3211
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7329740719416427,
+      "learning_rate": 9.919099260759867e-06,
+      "loss": 0.8547,
+      "step": 3212
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.39038526653231,
+      "learning_rate": 9.913879983356745e-06,
+      "loss": 0.8298,
+      "step": 3213
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8740425097126813,
+      "learning_rate": 9.908660729415087e-06,
+      "loss": 0.3335,
+      "step": 3214
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2656311716411959,
+      "learning_rate": 9.903441500356761e-06,
+      "loss": 0.8184,
+      "step": 3215
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2894256860447826,
+      "learning_rate": 9.898222297603638e-06,
+      "loss": 0.8612,
+      "step": 3216
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.436470852517631,
+      "learning_rate": 9.893003122577563e-06,
+      "loss": 0.8561,
+      "step": 3217
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2168111019619272,
+      "learning_rate": 9.887783976700387e-06,
+      "loss": 0.9145,
+      "step": 3218
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.415597640203667,
+      "learning_rate": 9.882564861393951e-06,
+      "loss": 0.9051,
+      "step": 3219
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2835811257104337,
+      "learning_rate": 9.877345778080082e-06,
+      "loss": 0.802,
+      "step": 3220
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3501525481448895,
+      "learning_rate": 9.872126728180604e-06,
+      "loss": 0.8079,
+      "step": 3221
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2296740796569845,
+      "learning_rate": 9.866907713117333e-06,
+      "loss": 0.9071,
+      "step": 3222
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.387880318018517,
+      "learning_rate": 9.861688734312064e-06,
+      "loss": 0.8981,
+      "step": 3223
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8682626382927793,
+      "learning_rate": 9.8564697931866e-06,
+      "loss": 0.3573,
+      "step": 3224
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8786518491679626,
+      "learning_rate": 9.851250891162722e-06,
+      "loss": 0.3454,
+      "step": 3225
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.57749481590159,
+      "learning_rate": 9.846032029662199e-06,
+      "loss": 0.8204,
+      "step": 3226
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4232237112521977,
+      "learning_rate": 9.840813210106795e-06,
+      "loss": 0.7425,
+      "step": 3227
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6960350026446973,
+      "learning_rate": 9.83559443391826e-06,
+      "loss": 0.7671,
+      "step": 3228
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5915257082530343,
+      "learning_rate": 9.830375702518339e-06,
+      "loss": 0.7468,
+      "step": 3229
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.485632043549182,
+      "learning_rate": 9.825157017328745e-06,
+      "loss": 0.8364,
+      "step": 3230
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3955019829950641,
+      "learning_rate": 9.8199383797712e-06,
+      "loss": 0.9204,
+      "step": 3231
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8736183677006223,
+      "learning_rate": 9.81471979126741e-06,
+      "loss": 0.3491,
+      "step": 3232
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5361389092132198,
+      "learning_rate": 9.80950125323905e-06,
+      "loss": 0.8685,
+      "step": 3233
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.248868662749267,
+      "learning_rate": 9.804282767107802e-06,
+      "loss": 0.8397,
+      "step": 3234
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.497041755498462,
+      "learning_rate": 9.799064334295324e-06,
+      "loss": 0.84,
+      "step": 3235
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5284794014709355,
+      "learning_rate": 9.79384595622326e-06,
+      "loss": 0.8886,
+      "step": 3236
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5436049519912511,
+      "learning_rate": 9.788627634313237e-06,
+      "loss": 0.8126,
+      "step": 3237
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4701211013224909,
+      "learning_rate": 9.78340936998688e-06,
+      "loss": 0.8344,
+      "step": 3238
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3373669675011004,
+      "learning_rate": 9.778191164665774e-06,
+      "loss": 0.8491,
+      "step": 3239
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3913500381169384,
+      "learning_rate": 9.77297301977151e-06,
+      "loss": 0.7915,
+      "step": 3240
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2765407987784612,
+      "learning_rate": 9.767754936725656e-06,
+      "loss": 0.908,
+      "step": 3241
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4764627405304673,
+      "learning_rate": 9.76253691694976e-06,
+      "loss": 0.8667,
+      "step": 3242
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2246526356102798,
+      "learning_rate": 9.757318961865348e-06,
+      "loss": 0.7937,
+      "step": 3243
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3768426282659891,
+      "learning_rate": 9.752101072893947e-06,
+      "loss": 0.8324,
+      "step": 3244
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.441890983305899,
+      "learning_rate": 9.746883251457043e-06,
+      "loss": 0.911,
+      "step": 3245
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4113821159232731,
+      "learning_rate": 9.741665498976116e-06,
+      "loss": 0.8244,
+      "step": 3246
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7333876751847688,
+      "learning_rate": 9.736447816872632e-06,
+      "loss": 0.905,
+      "step": 3247
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4742538440390462,
+      "learning_rate": 9.731230206568029e-06,
+      "loss": 0.9185,
+      "step": 3248
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3126379831761994,
+      "learning_rate": 9.726012669483723e-06,
+      "loss": 0.818,
+      "step": 3249
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6227053094655446,
+      "learning_rate": 9.720795207041119e-06,
+      "loss": 0.8947,
+      "step": 3250
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.397636898326866,
+      "learning_rate": 9.7155778206616e-06,
+      "loss": 0.7957,
+      "step": 3251
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2708570482693884,
+      "learning_rate": 9.710360511766517e-06,
+      "loss": 0.8147,
+      "step": 3252
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3150074407837038,
+      "learning_rate": 9.705143281777218e-06,
+      "loss": 0.8171,
+      "step": 3253
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2447325295569098,
+      "learning_rate": 9.69992613211502e-06,
+      "loss": 0.8669,
+      "step": 3254
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3215230627426713,
+      "learning_rate": 9.694709064201211e-06,
+      "loss": 0.9005,
+      "step": 3255
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.231424330730008,
+      "learning_rate": 9.68949207945707e-06,
+      "loss": 0.869,
+      "step": 3256
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3507630472738346,
+      "learning_rate": 9.684275179303846e-06,
+      "loss": 0.8506,
+      "step": 3257
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3667070327922466,
+      "learning_rate": 9.679058365162765e-06,
+      "loss": 0.8005,
+      "step": 3258
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3234353216630736,
+      "learning_rate": 9.673841638455029e-06,
+      "loss": 0.8896,
+      "step": 3259
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.267091078564698,
+      "learning_rate": 9.66862500060183e-06,
+      "loss": 0.7921,
+      "step": 3260
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4947600691905483,
+      "learning_rate": 9.663408453024306e-06,
+      "loss": 0.8229,
+      "step": 3261
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4072061255279407,
+      "learning_rate": 9.658191997143602e-06,
+      "loss": 0.8935,
+      "step": 3262
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2321752537233202,
+      "learning_rate": 9.652975634380822e-06,
+      "loss": 0.8934,
+      "step": 3263
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5781972362254921,
+      "learning_rate": 9.647759366157041e-06,
+      "loss": 0.8501,
+      "step": 3264
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4062687031508807,
+      "learning_rate": 9.642543193893318e-06,
+      "loss": 0.8765,
+      "step": 3265
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4509175130300187,
+      "learning_rate": 9.637327119010683e-06,
+      "loss": 0.8286,
+      "step": 3266
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3186831599456152,
+      "learning_rate": 9.632111142930143e-06,
+      "loss": 0.8702,
+      "step": 3267
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3842937185943751,
+      "learning_rate": 9.626895267072662e-06,
+      "loss": 0.845,
+      "step": 3268
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7024665259244136,
+      "learning_rate": 9.621679492859196e-06,
+      "loss": 0.3666,
+      "step": 3269
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4368949917907894,
+      "learning_rate": 9.61646382171067e-06,
+      "loss": 0.8622,
+      "step": 3270
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.345945094989235,
+      "learning_rate": 9.611248255047965e-06,
+      "loss": 0.8835,
+      "step": 3271
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3956786545648319,
+      "learning_rate": 9.606032794291953e-06,
+      "loss": 0.8034,
+      "step": 3272
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3842278321980457,
+      "learning_rate": 9.60081744086347e-06,
+      "loss": 0.7945,
+      "step": 3273
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1970881444119787,
+      "learning_rate": 9.595602196183317e-06,
+      "loss": 0.8815,
+      "step": 3274
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2455472054008807,
+      "learning_rate": 9.590387061672272e-06,
+      "loss": 0.8893,
+      "step": 3275
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.246811117660622,
+      "learning_rate": 9.585172038751086e-06,
+      "loss": 0.8052,
+      "step": 3276
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3471413596257102,
+      "learning_rate": 9.57995712884047e-06,
+      "loss": 0.7828,
+      "step": 3277
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7177921839564563,
+      "learning_rate": 9.57474233336111e-06,
+      "loss": 0.8415,
+      "step": 3278
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3380361617071423,
+      "learning_rate": 9.569527653733662e-06,
+      "loss": 0.8182,
+      "step": 3279
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.302063352352058,
+      "learning_rate": 9.564313091378746e-06,
+      "loss": 0.842,
+      "step": 3280
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.405020572426469,
+      "learning_rate": 9.559098647716953e-06,
+      "loss": 0.8852,
+      "step": 3281
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.259735057519102,
+      "learning_rate": 9.553884324168846e-06,
+      "loss": 0.7663,
+      "step": 3282
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2974090539382952,
+      "learning_rate": 9.54867012215494e-06,
+      "loss": 0.8928,
+      "step": 3283
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.24188124661922,
+      "learning_rate": 9.543456043095736e-06,
+      "loss": 0.845,
+      "step": 3284
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4228821792397963,
+      "learning_rate": 9.538242088411694e-06,
+      "loss": 0.8464,
+      "step": 3285
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3606431413064488,
+      "learning_rate": 9.533028259523233e-06,
+      "loss": 0.8039,
+      "step": 3286
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3872950417549692,
+      "learning_rate": 9.527814557850744e-06,
+      "loss": 0.8638,
+      "step": 3287
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3396394673976084,
+      "learning_rate": 9.522600984814587e-06,
+      "loss": 0.8705,
+      "step": 3288
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1874340624917628,
+      "learning_rate": 9.517387541835087e-06,
+      "loss": 0.8346,
+      "step": 3289
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3985226041855046,
+      "learning_rate": 9.512174230332515e-06,
+      "loss": 0.8497,
+      "step": 3290
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3690590874993591,
+      "learning_rate": 9.506961051727132e-06,
+      "loss": 0.8275,
+      "step": 3291
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3371064616792028,
+      "learning_rate": 9.501748007439154e-06,
+      "loss": 0.8681,
+      "step": 3292
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4895092273553918,
+      "learning_rate": 9.496535098888749e-06,
+      "loss": 0.8545,
+      "step": 3293
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.286697821527991,
+      "learning_rate": 9.491322327496062e-06,
+      "loss": 0.8229,
+      "step": 3294
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7047885919870527,
+      "learning_rate": 9.486109694681195e-06,
+      "loss": 0.8583,
+      "step": 3295
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4133797757255386,
+      "learning_rate": 9.480897201864214e-06,
+      "loss": 0.7621,
+      "step": 3296
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.302556461127079,
+      "learning_rate": 9.47568485046514e-06,
+      "loss": 0.887,
+      "step": 3297
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2363819973855787,
+      "learning_rate": 9.470472641903972e-06,
+      "loss": 0.836,
+      "step": 3298
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2457326430085816,
+      "learning_rate": 9.465260577600645e-06,
+      "loss": 0.8685,
+      "step": 3299
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3753553430532974,
+      "learning_rate": 9.460048658975082e-06,
+      "loss": 0.8869,
+      "step": 3300
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5568293585092103,
+      "learning_rate": 9.454836887447149e-06,
+      "loss": 0.9146,
+      "step": 3301
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9679666599415533,
+      "learning_rate": 9.449625264436673e-06,
+      "loss": 0.8247,
+      "step": 3302
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7636427787163398,
+      "learning_rate": 9.444413791363446e-06,
+      "loss": 0.8434,
+      "step": 3303
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2428006077634597,
+      "learning_rate": 9.43920246964722e-06,
+      "loss": 0.8759,
+      "step": 3304
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.366094269314453,
+      "learning_rate": 9.433991300707694e-06,
+      "loss": 0.908,
+      "step": 3305
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1758315702930813,
+      "learning_rate": 9.428780285964544e-06,
+      "loss": 0.8394,
+      "step": 3306
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.33915486017233,
+      "learning_rate": 9.423569426837387e-06,
+      "loss": 0.9365,
+      "step": 3307
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4125424402958784,
+      "learning_rate": 9.418358724745813e-06,
+      "loss": 0.8071,
+      "step": 3308
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2092331035734156,
+      "learning_rate": 9.413148181109352e-06,
+      "loss": 0.8836,
+      "step": 3309
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.458978251427767,
+      "learning_rate": 9.4079377973475e-06,
+      "loss": 0.8039,
+      "step": 3310
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2949103913006559,
+      "learning_rate": 9.40272757487972e-06,
+      "loss": 0.9294,
+      "step": 3311
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6459365410726812,
+      "learning_rate": 9.397517515125405e-06,
+      "loss": 0.8069,
+      "step": 3312
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1752430513381584,
+      "learning_rate": 9.392307619503928e-06,
+      "loss": 0.781,
+      "step": 3313
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9929648249208682,
+      "learning_rate": 9.38709788943461e-06,
+      "loss": 0.3537,
+      "step": 3314
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1559951024426096,
+      "learning_rate": 9.38188832633672e-06,
+      "loss": 0.8594,
+      "step": 3315
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4160601483941857,
+      "learning_rate": 9.376678931629488e-06,
+      "loss": 0.8624,
+      "step": 3316
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4955144363182897,
+      "learning_rate": 9.371469706732097e-06,
+      "loss": 0.8767,
+      "step": 3317
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2999167934941769,
+      "learning_rate": 9.366260653063684e-06,
+      "loss": 0.8151,
+      "step": 3318
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7342737137155968,
+      "learning_rate": 9.361051772043335e-06,
+      "loss": 0.8603,
+      "step": 3319
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2798142107853252,
+      "learning_rate": 9.355843065090103e-06,
+      "loss": 0.8476,
+      "step": 3320
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2164646593732578,
+      "learning_rate": 9.35063453362297e-06,
+      "loss": 0.8029,
+      "step": 3321
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.593603336023194,
+      "learning_rate": 9.34542617906089e-06,
+      "loss": 0.8905,
+      "step": 3322
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3817291548754649,
+      "learning_rate": 9.340218002822766e-06,
+      "loss": 0.8649,
+      "step": 3323
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2888652484426582,
+      "learning_rate": 9.335010006327443e-06,
+      "loss": 0.887,
+      "step": 3324
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.221618575344612,
+      "learning_rate": 9.329802190993726e-06,
+      "loss": 0.8655,
+      "step": 3325
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.447332046698887,
+      "learning_rate": 9.324594558240361e-06,
+      "loss": 0.8638,
+      "step": 3326
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1928868560124382,
+      "learning_rate": 9.319387109486065e-06,
+      "loss": 0.8639,
+      "step": 3327
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4015089554802516,
+      "learning_rate": 9.314179846149474e-06,
+      "loss": 0.8247,
+      "step": 3328
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4683099065801253,
+      "learning_rate": 9.3089727696492e-06,
+      "loss": 0.7507,
+      "step": 3329
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3350791048660984,
+      "learning_rate": 9.303765881403794e-06,
+      "loss": 0.7957,
+      "step": 3330
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3386527055915662,
+      "learning_rate": 9.298559182831752e-06,
+      "loss": 0.8944,
+      "step": 3331
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2582893112009188,
+      "learning_rate": 9.293352675351524e-06,
+      "loss": 0.8958,
+      "step": 3332
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0103044464980298,
+      "learning_rate": 9.288146360381507e-06,
+      "loss": 0.3316,
+      "step": 3333
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4801365446923147,
+      "learning_rate": 9.28294023934004e-06,
+      "loss": 0.8254,
+      "step": 3334
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6335760982185497,
+      "learning_rate": 9.277734313645422e-06,
+      "loss": 0.7876,
+      "step": 3335
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4513177105310808,
+      "learning_rate": 9.272528584715886e-06,
+      "loss": 0.8784,
+      "step": 3336
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4233294602762923,
+      "learning_rate": 9.267323053969615e-06,
+      "loss": 0.9514,
+      "step": 3337
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.438712034744723,
+      "learning_rate": 9.26211772282474e-06,
+      "loss": 0.8255,
+      "step": 3338
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1731702949554745,
+      "learning_rate": 9.256912592699339e-06,
+      "loss": 0.8511,
+      "step": 3339
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3856830570866687,
+      "learning_rate": 9.251707665011429e-06,
+      "loss": 0.844,
+      "step": 3340
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3257599350250424,
+      "learning_rate": 9.246502941178976e-06,
+      "loss": 0.8203,
+      "step": 3341
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3296645350495884,
+      "learning_rate": 9.2412984226199e-06,
+      "loss": 0.8484,
+      "step": 3342
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3817864978694165,
+      "learning_rate": 9.236094110752038e-06,
+      "loss": 0.8188,
+      "step": 3343
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1829318203702317,
+      "learning_rate": 9.230890006993203e-06,
+      "loss": 0.7452,
+      "step": 3344
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3846450478768437,
+      "learning_rate": 9.225686112761128e-06,
+      "loss": 0.8478,
+      "step": 3345
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6448235104840225,
+      "learning_rate": 9.220482429473502e-06,
+      "loss": 0.7519,
+      "step": 3346
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8019647195171369,
+      "learning_rate": 9.215278958547951e-06,
+      "loss": 0.8446,
+      "step": 3347
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.400620540063252,
+      "learning_rate": 9.210075701402037e-06,
+      "loss": 0.8517,
+      "step": 3348
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3389476630901973,
+      "learning_rate": 9.204872659453285e-06,
+      "loss": 0.8865,
+      "step": 3349
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.177703387294026,
+      "learning_rate": 9.19966983411913e-06,
+      "loss": 0.8552,
+      "step": 3350
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4277524627343439,
+      "learning_rate": 9.194467226816976e-06,
+      "loss": 0.8006,
+      "step": 3351
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2796502211401775,
+      "learning_rate": 9.189264838964159e-06,
+      "loss": 0.8755,
+      "step": 3352
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.285876972064867,
+      "learning_rate": 9.184062671977942e-06,
+      "loss": 0.8977,
+      "step": 3353
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3819258050710548,
+      "learning_rate": 9.178860727275546e-06,
+      "loss": 0.8324,
+      "step": 3354
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3038578165017511,
+      "learning_rate": 9.173659006274125e-06,
+      "loss": 0.8412,
+      "step": 3355
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4912254186650333,
+      "learning_rate": 9.168457510390764e-06,
+      "loss": 0.8058,
+      "step": 3356
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.298473348421068,
+      "learning_rate": 9.163256241042502e-06,
+      "loss": 0.8442,
+      "step": 3357
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4440505382225683,
+      "learning_rate": 9.158055199646306e-06,
+      "loss": 0.8951,
+      "step": 3358
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.403539087713613,
+      "learning_rate": 9.152854387619081e-06,
+      "loss": 0.8092,
+      "step": 3359
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2930896088853676,
+      "learning_rate": 9.147653806377672e-06,
+      "loss": 0.8721,
+      "step": 3360
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7208774789943176,
+      "learning_rate": 9.142453457338864e-06,
+      "loss": 0.3398,
+      "step": 3361
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4569638680276722,
+      "learning_rate": 9.137253341919369e-06,
+      "loss": 0.8183,
+      "step": 3362
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.341705779352439,
+      "learning_rate": 9.132053461535844e-06,
+      "loss": 0.8964,
+      "step": 3363
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3918776147995133,
+      "learning_rate": 9.126853817604887e-06,
+      "loss": 0.9255,
+      "step": 3364
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.362497436396348,
+      "learning_rate": 9.121654411543013e-06,
+      "loss": 0.8283,
+      "step": 3365
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.669726853387932,
+      "learning_rate": 9.11645524476669e-06,
+      "loss": 0.8362,
+      "step": 3366
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.317623392170619,
+      "learning_rate": 9.111256318692316e-06,
+      "loss": 0.8533,
+      "step": 3367
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4399511816211037,
+      "learning_rate": 9.10605763473622e-06,
+      "loss": 0.8598,
+      "step": 3368
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3776198389559628,
+      "learning_rate": 9.100859194314661e-06,
+      "loss": 0.859,
+      "step": 3369
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3374845870134853,
+      "learning_rate": 9.095660998843842e-06,
+      "loss": 0.7726,
+      "step": 3370
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4050668938426265,
+      "learning_rate": 9.090463049739903e-06,
+      "loss": 0.8178,
+      "step": 3371
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2354048619438749,
+      "learning_rate": 9.085265348418894e-06,
+      "loss": 0.803,
+      "step": 3372
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4202285196741902,
+      "learning_rate": 9.08006789629682e-06,
+      "loss": 0.7299,
+      "step": 3373
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.291248405869618,
+      "learning_rate": 9.074870694789613e-06,
+      "loss": 0.8047,
+      "step": 3374
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.309811337370424,
+      "learning_rate": 9.069673745313127e-06,
+      "loss": 0.817,
+      "step": 3375
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2550397976969294,
+      "learning_rate": 9.064477049283157e-06,
+      "loss": 0.7978,
+      "step": 3376
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8764698714408057,
+      "learning_rate": 9.059280608115427e-06,
+      "loss": 0.8226,
+      "step": 3377
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.278061927344398,
+      "learning_rate": 9.054084423225592e-06,
+      "loss": 0.8699,
+      "step": 3378
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5728104985929459,
+      "learning_rate": 9.04888849602923e-06,
+      "loss": 0.8941,
+      "step": 3379
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4322838587259112,
+      "learning_rate": 9.043692827941864e-06,
+      "loss": 0.8641,
+      "step": 3380
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4555162701422832,
+      "learning_rate": 9.03849742037893e-06,
+      "loss": 0.9123,
+      "step": 3381
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.977047434567082,
+      "learning_rate": 9.033302274755803e-06,
+      "loss": 0.3407,
+      "step": 3382
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.174017810728722,
+      "learning_rate": 9.028107392487787e-06,
+      "loss": 0.8131,
+      "step": 3383
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.280065650576498,
+      "learning_rate": 9.022912774990104e-06,
+      "loss": 0.8552,
+      "step": 3384
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5807153998649524,
+      "learning_rate": 9.01771842367791e-06,
+      "loss": 0.8253,
+      "step": 3385
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.67972839106089,
+      "learning_rate": 9.0125243399663e-06,
+      "loss": 0.8966,
+      "step": 3386
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3247030836722151,
+      "learning_rate": 9.007330525270282e-06,
+      "loss": 0.7648,
+      "step": 3387
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.410290600724773,
+      "learning_rate": 9.002136981004787e-06,
+      "loss": 0.8495,
+      "step": 3388
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2745015983919232,
+      "learning_rate": 8.996943708584688e-06,
+      "loss": 0.7973,
+      "step": 3389
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2123064920434417,
+      "learning_rate": 8.991750709424772e-06,
+      "loss": 0.851,
+      "step": 3390
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1945824333649506,
+      "learning_rate": 8.986557984939754e-06,
+      "loss": 0.9181,
+      "step": 3391
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.343760625525537,
+      "learning_rate": 8.981365536544275e-06,
+      "loss": 0.8441,
+      "step": 3392
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4729414867280886,
+      "learning_rate": 8.976173365652909e-06,
+      "loss": 0.9232,
+      "step": 3393
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.307193615565311,
+      "learning_rate": 8.970981473680132e-06,
+      "loss": 0.841,
+      "step": 3394
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8529591022657905,
+      "learning_rate": 8.96578986204037e-06,
+      "loss": 0.3418,
+      "step": 3395
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.315675205634184,
+      "learning_rate": 8.96059853214796e-06,
+      "loss": 0.919,
+      "step": 3396
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.343064447946701,
+      "learning_rate": 8.955407485417158e-06,
+      "loss": 0.8906,
+      "step": 3397
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6882801410498414,
+      "learning_rate": 8.950216723262152e-06,
+      "loss": 0.8458,
+      "step": 3398
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1444624540008776,
+      "learning_rate": 8.94502624709705e-06,
+      "loss": 0.7972,
+      "step": 3399
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3552666556065922,
+      "learning_rate": 8.939836058335878e-06,
+      "loss": 0.8917,
+      "step": 3400
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2170402232237352,
+      "learning_rate": 8.934646158392584e-06,
+      "loss": 0.8539,
+      "step": 3401
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3320480105109125,
+      "learning_rate": 8.92945654868105e-06,
+      "loss": 0.7902,
+      "step": 3402
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4974063389174643,
+      "learning_rate": 8.924267230615054e-06,
+      "loss": 0.8757,
+      "step": 3403
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3847964556426955,
+      "learning_rate": 8.91907820560832e-06,
+      "loss": 0.8244,
+      "step": 3404
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6634492785240853,
+      "learning_rate": 8.913889475074479e-06,
+      "loss": 0.855,
+      "step": 3405
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2321780094987667,
+      "learning_rate": 8.908701040427086e-06,
+      "loss": 0.8548,
+      "step": 3406
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8889174803166736,
+      "learning_rate": 8.903512903079608e-06,
+      "loss": 0.3522,
+      "step": 3407
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3483419339438767,
+      "learning_rate": 8.89832506444544e-06,
+      "loss": 0.7946,
+      "step": 3408
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.492575287984684,
+      "learning_rate": 8.893137525937894e-06,
+      "loss": 0.8886,
+      "step": 3409
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4546793452038402,
+      "learning_rate": 8.887950288970198e-06,
+      "loss": 0.8573,
+      "step": 3410
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6755418484195712,
+      "learning_rate": 8.882763354955495e-06,
+      "loss": 0.8698,
+      "step": 3411
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.404483284734382,
+      "learning_rate": 8.877576725306853e-06,
+      "loss": 0.8878,
+      "step": 3412
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3739944867064433,
+      "learning_rate": 8.87239040143725e-06,
+      "loss": 0.8147,
+      "step": 3413
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4173808389697002,
+      "learning_rate": 8.86720438475958e-06,
+      "loss": 0.8596,
+      "step": 3414
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.409818225063227,
+      "learning_rate": 8.86201867668667e-06,
+      "loss": 0.8626,
+      "step": 3415
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3166725401568788,
+      "learning_rate": 8.856833278631232e-06,
+      "loss": 0.8806,
+      "step": 3416
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3113517627568532,
+      "learning_rate": 8.851648192005925e-06,
+      "loss": 0.8049,
+      "step": 3417
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2117255915024996,
+      "learning_rate": 8.846463418223307e-06,
+      "loss": 0.8432,
+      "step": 3418
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3200814680546045,
+      "learning_rate": 8.841278958695848e-06,
+      "loss": 0.8171,
+      "step": 3419
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3658590691897836,
+      "learning_rate": 8.836094814835941e-06,
+      "loss": 0.8924,
+      "step": 3420
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.446069199066689,
+      "learning_rate": 8.830910988055892e-06,
+      "loss": 0.8487,
+      "step": 3421
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5483252453363934,
+      "learning_rate": 8.825727479767916e-06,
+      "loss": 0.9023,
+      "step": 3422
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3674131084671775,
+      "learning_rate": 8.820544291384138e-06,
+      "loss": 0.8237,
+      "step": 3423
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4079866745582212,
+      "learning_rate": 8.815361424316617e-06,
+      "loss": 0.9318,
+      "step": 3424
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8796023489180789,
+      "learning_rate": 8.81017887997729e-06,
+      "loss": 0.3483,
+      "step": 3425
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3621521539646435,
+      "learning_rate": 8.804996659778036e-06,
+      "loss": 0.8007,
+      "step": 3426
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.361617328513014,
+      "learning_rate": 8.799814765130631e-06,
+      "loss": 0.8484,
+      "step": 3427
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2074680078283484,
+      "learning_rate": 8.79463319744677e-06,
+      "loss": 0.8332,
+      "step": 3428
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6675587637128741,
+      "learning_rate": 8.789451958138053e-06,
+      "loss": 0.8813,
+      "step": 3429
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.524230130943539,
+      "learning_rate": 8.784271048615987e-06,
+      "loss": 0.8003,
+      "step": 3430
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4790775334899682,
+      "learning_rate": 8.779090470292006e-06,
+      "loss": 0.8543,
+      "step": 3431
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4189167971545105,
+      "learning_rate": 8.773910224577428e-06,
+      "loss": 0.8354,
+      "step": 3432
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4887925668887927,
+      "learning_rate": 8.768730312883505e-06,
+      "loss": 0.8881,
+      "step": 3433
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.475849120567027,
+      "learning_rate": 8.763550736621388e-06,
+      "loss": 0.8337,
+      "step": 3434
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6214498519438747,
+      "learning_rate": 8.758371497202131e-06,
+      "loss": 0.8568,
+      "step": 3435
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4446566595112706,
+      "learning_rate": 8.753192596036703e-06,
+      "loss": 0.9014,
+      "step": 3436
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3820789318959346,
+      "learning_rate": 8.748014034535983e-06,
+      "loss": 0.8282,
+      "step": 3437
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.464002998037563,
+      "learning_rate": 8.742835814110746e-06,
+      "loss": 0.8003,
+      "step": 3438
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2255985445536814,
+      "learning_rate": 8.737657936171691e-06,
+      "loss": 0.8896,
+      "step": 3439
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5406771887737536,
+      "learning_rate": 8.73248040212941e-06,
+      "loss": 0.8551,
+      "step": 3440
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3434126896329739,
+      "learning_rate": 8.727303213394408e-06,
+      "loss": 0.8004,
+      "step": 3441
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4663454896525807,
+      "learning_rate": 8.722126371377091e-06,
+      "loss": 0.8203,
+      "step": 3442
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3687337898084877,
+      "learning_rate": 8.716949877487778e-06,
+      "loss": 0.8606,
+      "step": 3443
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2787642891846192,
+      "learning_rate": 8.711773733136684e-06,
+      "loss": 0.9052,
+      "step": 3444
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.301175002767422,
+      "learning_rate": 8.706597939733931e-06,
+      "loss": 0.7972,
+      "step": 3445
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3148539698583814,
+      "learning_rate": 8.701422498689556e-06,
+      "loss": 0.8045,
+      "step": 3446
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4242140520999744,
+      "learning_rate": 8.69624741141349e-06,
+      "loss": 0.8506,
+      "step": 3447
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2213792469591873,
+      "learning_rate": 8.691072679315566e-06,
+      "loss": 0.8421,
+      "step": 3448
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3107817765505518,
+      "learning_rate": 8.685898303805523e-06,
+      "loss": 0.8007,
+      "step": 3449
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5206204643977579,
+      "learning_rate": 8.680724286293008e-06,
+      "loss": 0.8206,
+      "step": 3450
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3542190852626406,
+      "learning_rate": 8.675550628187562e-06,
+      "loss": 0.8753,
+      "step": 3451
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4947709050686147,
+      "learning_rate": 8.670377330898631e-06,
+      "loss": 0.8927,
+      "step": 3452
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3518749479535428,
+      "learning_rate": 8.665204395835573e-06,
+      "loss": 0.8229,
+      "step": 3453
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3676865102993336,
+      "learning_rate": 8.660031824407625e-06,
+      "loss": 0.7903,
+      "step": 3454
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.544069456481627,
+      "learning_rate": 8.654859618023946e-06,
+      "loss": 0.9192,
+      "step": 3455
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2918752392455823,
+      "learning_rate": 8.64968777809359e-06,
+      "loss": 0.8777,
+      "step": 3456
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6232854791696851,
+      "learning_rate": 8.644516306025501e-06,
+      "loss": 0.8583,
+      "step": 3457
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3632774762863789,
+      "learning_rate": 8.639345203228536e-06,
+      "loss": 0.8936,
+      "step": 3458
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3708328409095023,
+      "learning_rate": 8.634174471111445e-06,
+      "loss": 0.8736,
+      "step": 3459
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4050686483564696,
+      "learning_rate": 8.629004111082875e-06,
+      "loss": 0.797,
+      "step": 3460
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3173011813877678,
+      "learning_rate": 8.623834124551375e-06,
+      "loss": 0.8666,
+      "step": 3461
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4803703169698292,
+      "learning_rate": 8.618664512925398e-06,
+      "loss": 0.8583,
+      "step": 3462
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4517891390610331,
+      "learning_rate": 8.613495277613283e-06,
+      "loss": 0.8668,
+      "step": 3463
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2888484589287883,
+      "learning_rate": 8.608326420023272e-06,
+      "loss": 0.9131,
+      "step": 3464
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5421020379296073,
+      "learning_rate": 8.603157941563506e-06,
+      "loss": 0.8667,
+      "step": 3465
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6044311604703163,
+      "learning_rate": 8.597989843642025e-06,
+      "loss": 0.8045,
+      "step": 3466
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3475238617540806,
+      "learning_rate": 8.592822127666751e-06,
+      "loss": 0.819,
+      "step": 3467
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3603085992386053,
+      "learning_rate": 8.58765479504552e-06,
+      "loss": 0.8988,
+      "step": 3468
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4890242177649533,
+      "learning_rate": 8.582487847186061e-06,
+      "loss": 0.8094,
+      "step": 3469
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.502264165481318,
+      "learning_rate": 8.577321285495981e-06,
+      "loss": 0.9264,
+      "step": 3470
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3146033877802417,
+      "learning_rate": 8.572155111382799e-06,
+      "loss": 0.8494,
+      "step": 3471
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.419184554974385,
+      "learning_rate": 8.566989326253924e-06,
+      "loss": 0.8773,
+      "step": 3472
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7184860406104847,
+      "learning_rate": 8.56182393151666e-06,
+      "loss": 0.845,
+      "step": 3473
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3167459116216893,
+      "learning_rate": 8.556658928578196e-06,
+      "loss": 0.856,
+      "step": 3474
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3286731241047276,
+      "learning_rate": 8.55149431884563e-06,
+      "loss": 0.7825,
+      "step": 3475
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5192447149487844,
+      "learning_rate": 8.546330103725937e-06,
+      "loss": 0.847,
+      "step": 3476
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3691414671172901,
+      "learning_rate": 8.541166284625995e-06,
+      "loss": 0.9192,
+      "step": 3477
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.29155082833734,
+      "learning_rate": 8.536002862952572e-06,
+      "loss": 0.8558,
+      "step": 3478
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2134733631509982,
+      "learning_rate": 8.530839840112324e-06,
+      "loss": 0.8644,
+      "step": 3479
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3060670001232697,
+      "learning_rate": 8.525677217511799e-06,
+      "loss": 0.8198,
+      "step": 3480
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4957118686162425,
+      "learning_rate": 8.520514996557443e-06,
+      "loss": 0.8626,
+      "step": 3481
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4513779708200936,
+      "learning_rate": 8.515353178655582e-06,
+      "loss": 0.7608,
+      "step": 3482
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.369259684905322,
+      "learning_rate": 8.510191765212438e-06,
+      "loss": 0.8428,
+      "step": 3483
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3502303387664847,
+      "learning_rate": 8.505030757634125e-06,
+      "loss": 0.846,
+      "step": 3484
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6171504652233177,
+      "learning_rate": 8.499870157326647e-06,
+      "loss": 0.8608,
+      "step": 3485
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4247616518880055,
+      "learning_rate": 8.494709965695885e-06,
+      "loss": 0.866,
+      "step": 3486
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6919014846636853,
+      "learning_rate": 8.489550184147621e-06,
+      "loss": 0.7106,
+      "step": 3487
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1573675685396114,
+      "learning_rate": 8.484390814087527e-06,
+      "loss": 0.8998,
+      "step": 3488
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3370441414586165,
+      "learning_rate": 8.479231856921149e-06,
+      "loss": 0.7731,
+      "step": 3489
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7838595308023129,
+      "learning_rate": 8.474073314053935e-06,
+      "loss": 0.379,
+      "step": 3490
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.878643165676291,
+      "learning_rate": 8.468915186891215e-06,
+      "loss": 0.8244,
+      "step": 3491
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.466068582584559,
+      "learning_rate": 8.4637574768382e-06,
+      "loss": 0.8915,
+      "step": 3492
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.245145688198243,
+      "learning_rate": 8.458600185299994e-06,
+      "loss": 0.8539,
+      "step": 3493
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2236216747054087,
+      "learning_rate": 8.453443313681591e-06,
+      "loss": 0.8388,
+      "step": 3494
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7984914379088176,
+      "learning_rate": 8.448286863387858e-06,
+      "loss": 0.3502,
+      "step": 3495
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2871219535988674,
+      "learning_rate": 8.443130835823553e-06,
+      "loss": 0.8831,
+      "step": 3496
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4578350473236452,
+      "learning_rate": 8.437975232393331e-06,
+      "loss": 0.8541,
+      "step": 3497
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.630661494990603,
+      "learning_rate": 8.432820054501706e-06,
+      "loss": 0.8386,
+      "step": 3498
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.19345090861553,
+      "learning_rate": 8.427665303553101e-06,
+      "loss": 0.9186,
+      "step": 3499
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3315620739233816,
+      "learning_rate": 8.42251098095181e-06,
+      "loss": 0.8752,
+      "step": 3500
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5573633794369377,
+      "learning_rate": 8.41735708810201e-06,
+      "loss": 0.926,
+      "step": 3501
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4227882769144127,
+      "learning_rate": 8.412203626407765e-06,
+      "loss": 0.8436,
+      "step": 3502
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.358640902124898,
+      "learning_rate": 8.407050597273024e-06,
+      "loss": 0.8363,
+      "step": 3503
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4649772487843364,
+      "learning_rate": 8.401898002101605e-06,
+      "loss": 0.8284,
+      "step": 3504
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5623891767300726,
+      "learning_rate": 8.39674584229722e-06,
+      "loss": 0.8587,
+      "step": 3505
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4618227152293575,
+      "learning_rate": 8.391594119263467e-06,
+      "loss": 0.8058,
+      "step": 3506
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3704500303880773,
+      "learning_rate": 8.386442834403811e-06,
+      "loss": 0.8871,
+      "step": 3507
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3659837818879494,
+      "learning_rate": 8.381291989121604e-06,
+      "loss": 0.914,
+      "step": 3508
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4738154680470426,
+      "learning_rate": 8.376141584820078e-06,
+      "loss": 0.8292,
+      "step": 3509
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4942203251145167,
+      "learning_rate": 8.37099162290235e-06,
+      "loss": 0.9181,
+      "step": 3510
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5220244413835853,
+      "learning_rate": 8.365842104771405e-06,
+      "loss": 0.7563,
+      "step": 3511
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3785797095867351,
+      "learning_rate": 8.360693031830114e-06,
+      "loss": 0.7912,
+      "step": 3512
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.387485150658345,
+      "learning_rate": 8.355544405481233e-06,
+      "loss": 0.8284,
+      "step": 3513
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.308382793759505,
+      "learning_rate": 8.350396227127383e-06,
+      "loss": 0.8941,
+      "step": 3514
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7194145651799102,
+      "learning_rate": 8.345248498171073e-06,
+      "loss": 0.8431,
+      "step": 3515
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4640116820681126,
+      "learning_rate": 8.340101220014688e-06,
+      "loss": 0.8107,
+      "step": 3516
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.468349522666423,
+      "learning_rate": 8.334954394060484e-06,
+      "loss": 0.8209,
+      "step": 3517
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.41981655829562,
+      "learning_rate": 8.329808021710598e-06,
+      "loss": 0.8364,
+      "step": 3518
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5125132507195005,
+      "learning_rate": 8.324662104367052e-06,
+      "loss": 0.8311,
+      "step": 3519
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5000362149724618,
+      "learning_rate": 8.319516643431723e-06,
+      "loss": 0.8127,
+      "step": 3520
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4911642523192152,
+      "learning_rate": 8.314371640306386e-06,
+      "loss": 0.8827,
+      "step": 3521
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2776137534712464,
+      "learning_rate": 8.309227096392682e-06,
+      "loss": 0.8137,
+      "step": 3522
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2842261001409296,
+      "learning_rate": 8.30408301309212e-06,
+      "loss": 0.8031,
+      "step": 3523
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.3034707187119876,
+      "learning_rate": 8.298939391806094e-06,
+      "loss": 0.8687,
+      "step": 3524
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5299966016422737,
+      "learning_rate": 8.293796233935864e-06,
+      "loss": 0.8133,
+      "step": 3525
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.597913794173523,
+      "learning_rate": 8.288653540882579e-06,
+      "loss": 0.8824,
+      "step": 3526
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3503685477694625,
+      "learning_rate": 8.283511314047236e-06,
+      "loss": 0.794,
+      "step": 3527
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4766486870624465,
+      "learning_rate": 8.27836955483073e-06,
+      "loss": 0.7188,
+      "step": 3528
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.217303663435195,
+      "learning_rate": 8.273228264633815e-06,
+      "loss": 0.8129,
+      "step": 3529
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4726487926021135,
+      "learning_rate": 8.268087444857119e-06,
+      "loss": 0.7939,
+      "step": 3530
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.327405187544465,
+      "learning_rate": 8.262947096901142e-06,
+      "loss": 0.834,
+      "step": 3531
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.420776379642049,
+      "learning_rate": 8.25780722216626e-06,
+      "loss": 0.8278,
+      "step": 3532
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4610885513979301,
+      "learning_rate": 8.252667822052714e-06,
+      "loss": 0.8712,
+      "step": 3533
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.440062648756464,
+      "learning_rate": 8.247528897960615e-06,
+      "loss": 0.8614,
+      "step": 3534
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7676202136766617,
+      "learning_rate": 8.242390451289959e-06,
+      "loss": 0.8857,
+      "step": 3535
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8676428199801571,
+      "learning_rate": 8.237252483440585e-06,
+      "loss": 0.8328,
+      "step": 3536
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4970313115025857,
+      "learning_rate": 8.232114995812228e-06,
+      "loss": 0.87,
+      "step": 3537
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2127092063754226,
+      "learning_rate": 8.22697798980448e-06,
+      "loss": 0.8517,
+      "step": 3538
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.488339550648583,
+      "learning_rate": 8.2218414668168e-06,
+      "loss": 0.8198,
+      "step": 3539
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4571245210928145,
+      "learning_rate": 8.216705428248519e-06,
+      "loss": 0.7956,
+      "step": 3540
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.303849842229644,
+      "learning_rate": 8.211569875498838e-06,
+      "loss": 0.7641,
+      "step": 3541
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9496303370479445,
+      "learning_rate": 8.206434809966817e-06,
+      "loss": 0.3606,
+      "step": 3542
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7477419033087054,
+      "learning_rate": 8.201300233051395e-06,
+      "loss": 0.8629,
+      "step": 3543
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6319190893897733,
+      "learning_rate": 8.196166146151373e-06,
+      "loss": 0.8791,
+      "step": 3544
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6224880966589676,
+      "learning_rate": 8.191032550665417e-06,
+      "loss": 0.7813,
+      "step": 3545
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3197628632177116,
+      "learning_rate": 8.185899447992056e-06,
+      "loss": 0.8836,
+      "step": 3546
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3899934518654704,
+      "learning_rate": 8.180766839529689e-06,
+      "loss": 0.8408,
+      "step": 3547
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3495105018502578,
+      "learning_rate": 8.175634726676589e-06,
+      "loss": 0.8034,
+      "step": 3548
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3971576308208387,
+      "learning_rate": 8.170503110830874e-06,
+      "loss": 0.814,
+      "step": 3549
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5229597449053567,
+      "learning_rate": 8.16537199339054e-06,
+      "loss": 0.9228,
+      "step": 3550
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.375893798813591,
+      "learning_rate": 8.160241375753452e-06,
+      "loss": 0.8735,
+      "step": 3551
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5051167251845792,
+      "learning_rate": 8.155111259317323e-06,
+      "loss": 0.8423,
+      "step": 3552
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.383524382805488,
+      "learning_rate": 8.149981645479743e-06,
+      "loss": 0.8562,
+      "step": 3553
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3398200181049318,
+      "learning_rate": 8.144852535638161e-06,
+      "loss": 0.789,
+      "step": 3554
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.341858163486133,
+      "learning_rate": 8.139723931189883e-06,
+      "loss": 0.8419,
+      "step": 3555
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.38262743304159,
+      "learning_rate": 8.134595833532084e-06,
+      "loss": 0.8639,
+      "step": 3556
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4483461349131526,
+      "learning_rate": 8.129468244061805e-06,
+      "loss": 0.8975,
+      "step": 3557
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3566356879621455,
+      "learning_rate": 8.12434116417593e-06,
+      "loss": 0.8168,
+      "step": 3558
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4431632473915308,
+      "learning_rate": 8.11921459527123e-06,
+      "loss": 0.8434,
+      "step": 3559
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4308074462698728,
+      "learning_rate": 8.114088538744318e-06,
+      "loss": 0.8442,
+      "step": 3560
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4400251884185313,
+      "learning_rate": 8.108962995991673e-06,
+      "loss": 0.8847,
+      "step": 3561
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3285095590735039,
+      "learning_rate": 8.103837968409634e-06,
+      "loss": 0.8069,
+      "step": 3562
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.208650873849463,
+      "learning_rate": 8.098713457394398e-06,
+      "loss": 0.8725,
+      "step": 3563
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8345193319732066,
+      "learning_rate": 8.093589464342032e-06,
+      "loss": 0.3239,
+      "step": 3564
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.355158976575549,
+      "learning_rate": 8.08846599064844e-06,
+      "loss": 0.8725,
+      "step": 3565
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1012514377760785,
+      "learning_rate": 8.083343037709407e-06,
+      "loss": 0.8556,
+      "step": 3566
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.594654231115963,
+      "learning_rate": 8.078220606920565e-06,
+      "loss": 0.8012,
+      "step": 3567
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2731365304554907,
+      "learning_rate": 8.073098699677402e-06,
+      "loss": 0.8872,
+      "step": 3568
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4688180371261643,
+      "learning_rate": 8.067977317375268e-06,
+      "loss": 0.7722,
+      "step": 3569
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4432357837120806,
+      "learning_rate": 8.062856461409372e-06,
+      "loss": 0.8392,
+      "step": 3570
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2710278294954116,
+      "learning_rate": 8.057736133174768e-06,
+      "loss": 0.8285,
+      "step": 3571
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.430426906028895,
+      "learning_rate": 8.052616334066383e-06,
+      "loss": 0.8659,
+      "step": 3572
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7644353730504525,
+      "learning_rate": 8.047497065478991e-06,
+      "loss": 0.3111,
+      "step": 3573
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3325014165567486,
+      "learning_rate": 8.042378328807217e-06,
+      "loss": 0.937,
+      "step": 3574
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.862486585200898,
+      "learning_rate": 8.037260125445548e-06,
+      "loss": 0.3652,
+      "step": 3575
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.56205962174118,
+      "learning_rate": 8.032142456788328e-06,
+      "loss": 0.8393,
+      "step": 3576
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8122837530815264,
+      "learning_rate": 8.027025324229743e-06,
+      "loss": 0.3588,
+      "step": 3577
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2216100756772725,
+      "learning_rate": 8.021908729163842e-06,
+      "loss": 0.8717,
+      "step": 3578
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2831629169092844,
+      "learning_rate": 8.016792672984538e-06,
+      "loss": 0.9045,
+      "step": 3579
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3243025855867936,
+      "learning_rate": 8.01167715708557e-06,
+      "loss": 0.7933,
+      "step": 3580
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6597515678842947,
+      "learning_rate": 8.006562182860557e-06,
+      "loss": 0.9037,
+      "step": 3581
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2053052844233856,
+      "learning_rate": 8.001447751702955e-06,
+      "loss": 0.8249,
+      "step": 3582
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4288555106012812,
+      "learning_rate": 7.996333865006074e-06,
+      "loss": 0.8947,
+      "step": 3583
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2804017500952973,
+      "learning_rate": 7.99122052416308e-06,
+      "loss": 0.8257,
+      "step": 3584
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4580941241778993,
+      "learning_rate": 7.986107730566985e-06,
+      "loss": 0.9537,
+      "step": 3585
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8079095235166265,
+      "learning_rate": 7.980995485610665e-06,
+      "loss": 0.7771,
+      "step": 3586
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5024538458060308,
+      "learning_rate": 7.975883790686821e-06,
+      "loss": 0.8335,
+      "step": 3587
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.326399495699327,
+      "learning_rate": 7.970772647188029e-06,
+      "loss": 0.862,
+      "step": 3588
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4773395451526725,
+      "learning_rate": 7.965662056506708e-06,
+      "loss": 0.8297,
+      "step": 3589
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3607165366632843,
+      "learning_rate": 7.960552020035118e-06,
+      "loss": 0.7825,
+      "step": 3590
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4867744361578705,
+      "learning_rate": 7.955442539165372e-06,
+      "loss": 0.756,
+      "step": 3591
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5535544862747301,
+      "learning_rate": 7.950333615289442e-06,
+      "loss": 0.9193,
+      "step": 3592
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3525258239208373,
+      "learning_rate": 7.945225249799132e-06,
+      "loss": 0.8587,
+      "step": 3593
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.843060194938453,
+      "learning_rate": 7.9401174440861e-06,
+      "loss": 0.3293,
+      "step": 3594
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2834715816406845,
+      "learning_rate": 7.935010199541864e-06,
+      "loss": 0.8056,
+      "step": 3595
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2828446690292952,
+      "learning_rate": 7.92990351755777e-06,
+      "loss": 0.8823,
+      "step": 3596
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.195200621910457,
+      "learning_rate": 7.924797399525017e-06,
+      "loss": 0.894,
+      "step": 3597
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8263470709706975,
+      "learning_rate": 7.91969184683466e-06,
+      "loss": 0.33,
+      "step": 3598
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.448683042042446,
+      "learning_rate": 7.914586860877584e-06,
+      "loss": 0.7512,
+      "step": 3599
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2014069451228253,
+      "learning_rate": 7.909482443044532e-06,
+      "loss": 0.8261,
+      "step": 3600
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5463862483870354,
+      "learning_rate": 7.904378594726095e-06,
+      "loss": 0.9134,
+      "step": 3601
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3698324144096066,
+      "learning_rate": 7.899275317312686e-06,
+      "loss": 0.9085,
+      "step": 3602
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3879407481755344,
+      "learning_rate": 7.89417261219459e-06,
+      "loss": 0.9213,
+      "step": 3603
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5156555897781723,
+      "learning_rate": 7.889070480761921e-06,
+      "loss": 0.8249,
+      "step": 3604
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2609308982106695,
+      "learning_rate": 7.883968924404645e-06,
+      "loss": 0.8533,
+      "step": 3605
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.245514164639684,
+      "learning_rate": 7.878867944512561e-06,
+      "loss": 0.8531,
+      "step": 3606
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4669113158981286,
+      "learning_rate": 7.873767542475316e-06,
+      "loss": 0.7977,
+      "step": 3607
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2839468008717088,
+      "learning_rate": 7.868667719682409e-06,
+      "loss": 0.8916,
+      "step": 3608
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2653930515538772,
+      "learning_rate": 7.863568477523158e-06,
+      "loss": 0.8507,
+      "step": 3609
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4853824095623467,
+      "learning_rate": 7.858469817386746e-06,
+      "loss": 0.8579,
+      "step": 3610
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2878997511275698,
+      "learning_rate": 7.853371740662193e-06,
+      "loss": 0.7886,
+      "step": 3611
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4375878286482502,
+      "learning_rate": 7.848274248738345e-06,
+      "loss": 0.8143,
+      "step": 3612
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3883573629308774,
+      "learning_rate": 7.843177343003905e-06,
+      "loss": 0.8321,
+      "step": 3613
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8584630196235041,
+      "learning_rate": 7.838081024847412e-06,
+      "loss": 0.9148,
+      "step": 3614
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5934805548791853,
+      "learning_rate": 7.83298529565724e-06,
+      "loss": 0.8588,
+      "step": 3615
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4429527517793959,
+      "learning_rate": 7.827890156821604e-06,
+      "loss": 0.8875,
+      "step": 3616
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.486885162855852,
+      "learning_rate": 7.82279560972857e-06,
+      "loss": 0.888,
+      "step": 3617
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4873931473983737,
+      "learning_rate": 7.817701655766024e-06,
+      "loss": 0.8558,
+      "step": 3618
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4075738847533912,
+      "learning_rate": 7.8126082963217e-06,
+      "loss": 0.8897,
+      "step": 3619
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9865273708747688,
+      "learning_rate": 7.807515532783177e-06,
+      "loss": 0.3497,
+      "step": 3620
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4310453499215643,
+      "learning_rate": 7.802423366537856e-06,
+      "loss": 0.9124,
+      "step": 3621
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.763220100267875,
+      "learning_rate": 7.797331798972986e-06,
+      "loss": 0.7677,
+      "step": 3622
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2520234695227572,
+      "learning_rate": 7.79224083147565e-06,
+      "loss": 0.8828,
+      "step": 3623
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.278512043121873,
+      "learning_rate": 7.787150465432774e-06,
+      "loss": 0.8728,
+      "step": 3624
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2522640838913428,
+      "learning_rate": 7.782060702231103e-06,
+      "loss": 0.8438,
+      "step": 3625
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4182873156551168,
+      "learning_rate": 7.776971543257236e-06,
+      "loss": 0.8658,
+      "step": 3626
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4030868147537157,
+      "learning_rate": 7.771882989897603e-06,
+      "loss": 0.8811,
+      "step": 3627
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3024417600039881,
+      "learning_rate": 7.766795043538457e-06,
+      "loss": 0.8329,
+      "step": 3628
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3326810055412666,
+      "learning_rate": 7.7617077055659e-06,
+      "loss": 0.8909,
+      "step": 3629
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.317713482489219,
+      "learning_rate": 7.756620977365869e-06,
+      "loss": 0.9012,
+      "step": 3630
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7085090234969864,
+      "learning_rate": 7.751534860324116e-06,
+      "loss": 0.8379,
+      "step": 3631
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2171472845485551,
+      "learning_rate": 7.746449355826247e-06,
+      "loss": 0.8305,
+      "step": 3632
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5023959862430087,
+      "learning_rate": 7.741364465257697e-06,
+      "loss": 0.8494,
+      "step": 3633
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5182640140933463,
+      "learning_rate": 7.736280190003723e-06,
+      "loss": 0.8885,
+      "step": 3634
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4046572732113116,
+      "learning_rate": 7.731196531449426e-06,
+      "loss": 0.8659,
+      "step": 3635
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4395281249374392,
+      "learning_rate": 7.726113490979735e-06,
+      "loss": 0.82,
+      "step": 3636
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4571951653779784,
+      "learning_rate": 7.721031069979408e-06,
+      "loss": 0.7843,
+      "step": 3637
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3917649872505053,
+      "learning_rate": 7.715949269833034e-06,
+      "loss": 0.8384,
+      "step": 3638
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3893693682607504,
+      "learning_rate": 7.710868091925047e-06,
+      "loss": 0.8921,
+      "step": 3639
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4430888907167914,
+      "learning_rate": 7.705787537639685e-06,
+      "loss": 0.8301,
+      "step": 3640
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5921294315463197,
+      "learning_rate": 7.70070760836104e-06,
+      "loss": 0.8365,
+      "step": 3641
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4446459186733032,
+      "learning_rate": 7.695628305473025e-06,
+      "loss": 0.7485,
+      "step": 3642
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4433217192139707,
+      "learning_rate": 7.69054963035938e-06,
+      "loss": 0.8406,
+      "step": 3643
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2885077297812098,
+      "learning_rate": 7.685471584403674e-06,
+      "loss": 0.8194,
+      "step": 3644
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4000970364447052,
+      "learning_rate": 7.680394168989306e-06,
+      "loss": 0.8436,
+      "step": 3645
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4945662328620992,
+      "learning_rate": 7.675317385499513e-06,
+      "loss": 0.8877,
+      "step": 3646
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5682556574102344,
+      "learning_rate": 7.670241235317339e-06,
+      "loss": 0.8562,
+      "step": 3647
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.841921139991204,
+      "learning_rate": 7.665165719825676e-06,
+      "loss": 0.3711,
+      "step": 3648
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6112753923052603,
+      "learning_rate": 7.660090840407231e-06,
+      "loss": 0.8236,
+      "step": 3649
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.546960588697949,
+      "learning_rate": 7.65501659844454e-06,
+      "loss": 0.8742,
+      "step": 3650
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4390734031949686,
+      "learning_rate": 7.649942995319965e-06,
+      "loss": 0.892,
+      "step": 3651
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.276461270534326,
+      "learning_rate": 7.644870032415705e-06,
+      "loss": 0.8085,
+      "step": 3652
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3511085423175608,
+      "learning_rate": 7.63979771111376e-06,
+      "loss": 0.8323,
+      "step": 3653
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5268784153174397,
+      "learning_rate": 7.63472603279598e-06,
+      "loss": 0.8848,
+      "step": 3654
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6635770766914166,
+      "learning_rate": 7.629654998844031e-06,
+      "loss": 0.8249,
+      "step": 3655
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3344818618099579,
+      "learning_rate": 7.624584610639397e-06,
+      "loss": 0.8231,
+      "step": 3656
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.308149194289193,
+      "learning_rate": 7.619514869563394e-06,
+      "loss": 0.9221,
+      "step": 3657
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4663931383675721,
+      "learning_rate": 7.6144457769971606e-06,
+      "loss": 0.8183,
+      "step": 3658
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4681913119841432,
+      "learning_rate": 7.609377334321653e-06,
+      "loss": 0.8503,
+      "step": 3659
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2563910463848558,
+      "learning_rate": 7.604309542917656e-06,
+      "loss": 0.8484,
+      "step": 3660
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4152445101026292,
+      "learning_rate": 7.599242404165783e-06,
+      "loss": 0.9022,
+      "step": 3661
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4840868224688974,
+      "learning_rate": 7.59417591944645e-06,
+      "loss": 0.8138,
+      "step": 3662
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4477249515719366,
+      "learning_rate": 7.589110090139917e-06,
+      "loss": 0.8842,
+      "step": 3663
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7909386467859402,
+      "learning_rate": 7.584044917626251e-06,
+      "loss": 0.3137,
+      "step": 3664
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.212369966017015,
+      "learning_rate": 7.5789804032853476e-06,
+      "loss": 0.8048,
+      "step": 3665
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4096588779284505,
+      "learning_rate": 7.573916548496916e-06,
+      "loss": 0.8737,
+      "step": 3666
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4305288869382986,
+      "learning_rate": 7.5688533546404895e-06,
+      "loss": 0.7828,
+      "step": 3667
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5185095041305432,
+      "learning_rate": 7.5637908230954316e-06,
+      "loss": 0.8334,
+      "step": 3668
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1437209378635254,
+      "learning_rate": 7.558728955240901e-06,
+      "loss": 0.781,
+      "step": 3669
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3076771839186683,
+      "learning_rate": 7.553667752455899e-06,
+      "loss": 0.8256,
+      "step": 3670
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1032935318851078,
+      "learning_rate": 7.548607216119237e-06,
+      "loss": 0.8056,
+      "step": 3671
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2407737955590747,
+      "learning_rate": 7.54354734760954e-06,
+      "loss": 0.887,
+      "step": 3672
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3579257105512947,
+      "learning_rate": 7.5384881483052585e-06,
+      "loss": 0.9177,
+      "step": 3673
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4128993609144886,
+      "learning_rate": 7.53342961958466e-06,
+      "loss": 0.8702,
+      "step": 3674
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7175463428207238,
+      "learning_rate": 7.528371762825819e-06,
+      "loss": 0.8744,
+      "step": 3675
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.408134601634781,
+      "learning_rate": 7.5233145794066445e-06,
+      "loss": 0.8289,
+      "step": 3676
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.59831714193309,
+      "learning_rate": 7.518258070704849e-06,
+      "loss": 0.7477,
+      "step": 3677
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3482466130223394,
+      "learning_rate": 7.513202238097963e-06,
+      "loss": 0.7927,
+      "step": 3678
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3275480728326299,
+      "learning_rate": 7.508147082963337e-06,
+      "loss": 0.8057,
+      "step": 3679
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3447884661042737,
+      "learning_rate": 7.503092606678135e-06,
+      "loss": 0.7839,
+      "step": 3680
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5359912256563613,
+      "learning_rate": 7.4980388106193336e-06,
+      "loss": 0.7742,
+      "step": 3681
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3523769505296812,
+      "learning_rate": 7.4929856961637246e-06,
+      "loss": 0.7711,
+      "step": 3682
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.169135366651676,
+      "learning_rate": 7.487933264687921e-06,
+      "loss": 0.8801,
+      "step": 3683
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3607897525132546,
+      "learning_rate": 7.482881517568344e-06,
+      "loss": 0.8532,
+      "step": 3684
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2490016608766912,
+      "learning_rate": 7.477830456181222e-06,
+      "loss": 0.8673,
+      "step": 3685
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2592078201194958,
+      "learning_rate": 7.472780081902608e-06,
+      "loss": 0.83,
+      "step": 3686
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5756763668503362,
+      "learning_rate": 7.467730396108368e-06,
+      "loss": 0.9209,
+      "step": 3687
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.620636344631634,
+      "learning_rate": 7.462681400174165e-06,
+      "loss": 0.8684,
+      "step": 3688
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3609203373673189,
+      "learning_rate": 7.45763309547549e-06,
+      "loss": 0.8377,
+      "step": 3689
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3793255814055534,
+      "learning_rate": 7.452585483387647e-06,
+      "loss": 0.8438,
+      "step": 3690
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3528625032692254,
+      "learning_rate": 7.4475385652857325e-06,
+      "loss": 0.8305,
+      "step": 3691
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.290223113928149,
+      "learning_rate": 7.442492342544672e-06,
+      "loss": 0.8887,
+      "step": 3692
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.238010273805996,
+      "learning_rate": 7.437446816539198e-06,
+      "loss": 0.8474,
+      "step": 3693
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3766904553341708,
+      "learning_rate": 7.432401988643847e-06,
+      "loss": 0.8001,
+      "step": 3694
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3583595981982763,
+      "learning_rate": 7.427357860232971e-06,
+      "loss": 0.7952,
+      "step": 3695
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.437833247870376,
+      "learning_rate": 7.422314432680731e-06,
+      "loss": 0.8208,
+      "step": 3696
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2220461240194171,
+      "learning_rate": 7.417271707361091e-06,
+      "loss": 0.8127,
+      "step": 3697
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3941381966421966,
+      "learning_rate": 7.412229685647829e-06,
+      "loss": 0.8268,
+      "step": 3698
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2628702159097358,
+      "learning_rate": 7.407188368914537e-06,
+      "loss": 0.7658,
+      "step": 3699
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.364032088743096,
+      "learning_rate": 7.402147758534604e-06,
+      "loss": 0.8644,
+      "step": 3700
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4716586267778038,
+      "learning_rate": 7.39710785588123e-06,
+      "loss": 0.8216,
+      "step": 3701
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3870052216362971,
+      "learning_rate": 7.3920686623274265e-06,
+      "loss": 0.8654,
+      "step": 3702
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5895279622266585,
+      "learning_rate": 7.387030179246009e-06,
+      "loss": 0.8984,
+      "step": 3703
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3119306351167366,
+      "learning_rate": 7.381992408009593e-06,
+      "loss": 0.9005,
+      "step": 3704
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4103566526630942,
+      "learning_rate": 7.376955349990613e-06,
+      "loss": 0.8906,
+      "step": 3705
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.208304368040526,
+      "learning_rate": 7.3719190065613035e-06,
+      "loss": 0.8328,
+      "step": 3706
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3522814116222333,
+      "learning_rate": 7.366883379093698e-06,
+      "loss": 0.8812,
+      "step": 3707
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.350054844044675,
+      "learning_rate": 7.361848468959641e-06,
+      "loss": 0.7613,
+      "step": 3708
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3140572026743578,
+      "learning_rate": 7.356814277530785e-06,
+      "loss": 0.8696,
+      "step": 3709
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4975234047132615,
+      "learning_rate": 7.351780806178578e-06,
+      "loss": 0.7983,
+      "step": 3710
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.865545402678677,
+      "learning_rate": 7.346748056274275e-06,
+      "loss": 0.8764,
+      "step": 3711
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.203837494152397,
+      "learning_rate": 7.341716029188946e-06,
+      "loss": 0.852,
+      "step": 3712
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.627818836106272,
+      "learning_rate": 7.33668472629344e-06,
+      "loss": 0.7911,
+      "step": 3713
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.261271191056609,
+      "learning_rate": 7.33165414895843e-06,
+      "loss": 0.8796,
+      "step": 3714
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3952529319423257,
+      "learning_rate": 7.326624298554387e-06,
+      "loss": 0.8235,
+      "step": 3715
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.979079099511998,
+      "learning_rate": 7.321595176451575e-06,
+      "loss": 0.3373,
+      "step": 3716
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.343273951794725,
+      "learning_rate": 7.316566784020067e-06,
+      "loss": 0.8397,
+      "step": 3717
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.353298017010148,
+      "learning_rate": 7.311539122629738e-06,
+      "loss": 0.8213,
+      "step": 3718
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2965598563691165,
+      "learning_rate": 7.306512193650258e-06,
+      "loss": 0.7882,
+      "step": 3719
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4715317017003005,
+      "learning_rate": 7.301485998451101e-06,
+      "loss": 0.8659,
+      "step": 3720
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3947613303265236,
+      "learning_rate": 7.29646053840155e-06,
+      "loss": 0.8331,
+      "step": 3721
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4954970429784855,
+      "learning_rate": 7.291435814870664e-06,
+      "loss": 0.8322,
+      "step": 3722
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5425309948496622,
+      "learning_rate": 7.2864118292273265e-06,
+      "loss": 0.8248,
+      "step": 3723
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4963361955142336,
+      "learning_rate": 7.281388582840209e-06,
+      "loss": 0.8453,
+      "step": 3724
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.564836836012575,
+      "learning_rate": 7.276366077077781e-06,
+      "loss": 0.8985,
+      "step": 3725
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6190949043564615,
+      "learning_rate": 7.271344313308308e-06,
+      "loss": 0.8961,
+      "step": 3726
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3215569959662545,
+      "learning_rate": 7.2663232928998594e-06,
+      "loss": 0.8095,
+      "step": 3727
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5356228497889537,
+      "learning_rate": 7.2613030172203045e-06,
+      "loss": 0.8943,
+      "step": 3728
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7928921749658995,
+      "learning_rate": 7.256283487637297e-06,
+      "loss": 0.7233,
+      "step": 3729
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3531569876865392,
+      "learning_rate": 7.251264705518299e-06,
+      "loss": 0.7773,
+      "step": 3730
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2383020050282971,
+      "learning_rate": 7.246246672230568e-06,
+      "loss": 0.8637,
+      "step": 3731
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4276006084091848,
+      "learning_rate": 7.241229389141147e-06,
+      "loss": 0.8728,
+      "step": 3732
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3888262201532078,
+      "learning_rate": 7.236212857616885e-06,
+      "loss": 0.8606,
+      "step": 3733
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2961132026831341,
+      "learning_rate": 7.231197079024431e-06,
+      "loss": 0.7709,
+      "step": 3734
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5803558567625329,
+      "learning_rate": 7.226182054730208e-06,
+      "loss": 0.8432,
+      "step": 3735
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1878801873635094,
+      "learning_rate": 7.221167786100458e-06,
+      "loss": 0.8552,
+      "step": 3736
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8382884891155093,
+      "learning_rate": 7.216154274501203e-06,
+      "loss": 0.8576,
+      "step": 3737
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2841162966903286,
+      "learning_rate": 7.211141521298259e-06,
+      "loss": 0.8328,
+      "step": 3738
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6146652805902946,
+      "learning_rate": 7.206129527857239e-06,
+      "loss": 0.8061,
+      "step": 3739
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6124781683342146,
+      "learning_rate": 7.201118295543553e-06,
+      "loss": 0.8315,
+      "step": 3740
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.281805271396234,
+      "learning_rate": 7.196107825722392e-06,
+      "loss": 0.8275,
+      "step": 3741
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5154533805229058,
+      "learning_rate": 7.191098119758747e-06,
+      "loss": 0.8083,
+      "step": 3742
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3711137562973996,
+      "learning_rate": 7.1860891790174035e-06,
+      "loss": 0.8973,
+      "step": 3743
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9492173057513767,
+      "learning_rate": 7.181081004862937e-06,
+      "loss": 0.3528,
+      "step": 3744
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5164724081279757,
+      "learning_rate": 7.176073598659706e-06,
+      "loss": 0.8187,
+      "step": 3745
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.509988591902235,
+      "learning_rate": 7.171066961771868e-06,
+      "loss": 0.8555,
+      "step": 3746
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2078658412163263,
+      "learning_rate": 7.1660610955633725e-06,
+      "loss": 0.8119,
+      "step": 3747
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2320890559119315,
+      "learning_rate": 7.16105600139795e-06,
+      "loss": 0.8192,
+      "step": 3748
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9052482208476379,
+      "learning_rate": 7.156051680639127e-06,
+      "loss": 0.3438,
+      "step": 3749
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2174561874258307,
+      "learning_rate": 7.151048134650225e-06,
+      "loss": 0.8238,
+      "step": 3750
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4608942525245714,
+      "learning_rate": 7.146045364794339e-06,
+      "loss": 0.8616,
+      "step": 3751
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.446425289200998,
+      "learning_rate": 7.1410433724343645e-06,
+      "loss": 0.8532,
+      "step": 3752
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4059920375705637,
+      "learning_rate": 7.136042158932988e-06,
+      "loss": 0.8724,
+      "step": 3753
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4354468434054175,
+      "learning_rate": 7.131041725652669e-06,
+      "loss": 0.8495,
+      "step": 3754
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4425730208578538,
+      "learning_rate": 7.126042073955668e-06,
+      "loss": 0.7817,
+      "step": 3755
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.1004703393267685,
+      "learning_rate": 7.121043205204028e-06,
+      "loss": 0.8513,
+      "step": 3756
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4016663062631076,
+      "learning_rate": 7.116045120759575e-06,
+      "loss": 0.8705,
+      "step": 3757
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5700019937475267,
+      "learning_rate": 7.1110478219839295e-06,
+      "loss": 0.8473,
+      "step": 3758
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3322647259274711,
+      "learning_rate": 7.106051310238493e-06,
+      "loss": 0.8125,
+      "step": 3759
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3092225784385094,
+      "learning_rate": 7.101055586884451e-06,
+      "loss": 0.7825,
+      "step": 3760
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6500487803072337,
+      "learning_rate": 7.096060653282777e-06,
+      "loss": 0.7645,
+      "step": 3761
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3108016873552462,
+      "learning_rate": 7.091066510794224e-06,
+      "loss": 0.7239,
+      "step": 3762
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.539918395293901,
+      "learning_rate": 7.086073160779347e-06,
+      "loss": 0.8589,
+      "step": 3763
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6536426308868717,
+      "learning_rate": 7.081080604598458e-06,
+      "loss": 0.889,
+      "step": 3764
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3652522422920201,
+      "learning_rate": 7.076088843611675e-06,
+      "loss": 0.8722,
+      "step": 3765
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7119311230120018,
+      "learning_rate": 7.0710978791788895e-06,
+      "loss": 0.899,
+      "step": 3766
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5380290132918233,
+      "learning_rate": 7.066107712659778e-06,
+      "loss": 0.8224,
+      "step": 3767
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3341886872178037,
+      "learning_rate": 7.061118345413797e-06,
+      "loss": 0.9101,
+      "step": 3768
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5923699897690586,
+      "learning_rate": 7.0561297788001915e-06,
+      "loss": 0.835,
+      "step": 3769
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7803896443895229,
+      "learning_rate": 7.0511420141779805e-06,
+      "loss": 0.8906,
+      "step": 3770
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8546563285124402,
+      "learning_rate": 7.046155052905967e-06,
+      "loss": 0.3626,
+      "step": 3771
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4560847416826381,
+      "learning_rate": 7.04116889634275e-06,
+      "loss": 0.7829,
+      "step": 3772
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.052536155966014,
+      "learning_rate": 7.036183545846678e-06,
+      "loss": 0.3147,
+      "step": 3773
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.286655588429566,
+      "learning_rate": 7.031199002775908e-06,
+      "loss": 0.8641,
+      "step": 3774
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2671804303843404,
+      "learning_rate": 7.0262152684883675e-06,
+      "loss": 0.8589,
+      "step": 3775
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8664162219447997,
+      "learning_rate": 7.021232344341759e-06,
+      "loss": 0.3383,
+      "step": 3776
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8998993016117565,
+      "learning_rate": 7.01625023169357e-06,
+      "loss": 0.3481,
+      "step": 3777
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.507431383589791,
+      "learning_rate": 7.011268931901067e-06,
+      "loss": 0.8369,
+      "step": 3778
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4175251892267902,
+      "learning_rate": 7.006288446321288e-06,
+      "loss": 0.8372,
+      "step": 3779
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3982933032692515,
+      "learning_rate": 7.001308776311061e-06,
+      "loss": 0.7768,
+      "step": 3780
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3049032556142626,
+      "learning_rate": 6.9963299232269824e-06,
+      "loss": 0.9424,
+      "step": 3781
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5430626623329047,
+      "learning_rate": 6.99135188842543e-06,
+      "loss": 0.8321,
+      "step": 3782
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4836573723222872,
+      "learning_rate": 6.986374673262557e-06,
+      "loss": 0.8695,
+      "step": 3783
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2659571191274577,
+      "learning_rate": 6.981398279094292e-06,
+      "loss": 0.7818,
+      "step": 3784
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3177329242212288,
+      "learning_rate": 6.976422707276349e-06,
+      "loss": 0.8151,
+      "step": 3785
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3344675141454254,
+      "learning_rate": 6.971447959164201e-06,
+      "loss": 0.8755,
+      "step": 3786
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4076364874492573,
+      "learning_rate": 6.966474036113112e-06,
+      "loss": 0.7421,
+      "step": 3787
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2840531014831673,
+      "learning_rate": 6.961500939478118e-06,
+      "loss": 0.8807,
+      "step": 3788
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5947739698624976,
+      "learning_rate": 6.956528670614022e-06,
+      "loss": 0.9341,
+      "step": 3789
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4517645588223715,
+      "learning_rate": 6.951557230875412e-06,
+      "loss": 0.8489,
+      "step": 3790
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.334367908480157,
+      "learning_rate": 6.9465866216166436e-06,
+      "loss": 0.8091,
+      "step": 3791
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2659517472773645,
+      "learning_rate": 6.941616844191846e-06,
+      "loss": 0.9292,
+      "step": 3792
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.587963199951103,
+      "learning_rate": 6.936647899954921e-06,
+      "loss": 0.8121,
+      "step": 3793
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3562906354209097,
+      "learning_rate": 6.931679790259558e-06,
+      "loss": 0.7597,
+      "step": 3794
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0654614327558893,
+      "learning_rate": 6.926712516459192e-06,
+      "loss": 0.8222,
+      "step": 3795
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7444124362424673,
+      "learning_rate": 6.921746079907055e-06,
+      "loss": 0.8627,
+      "step": 3796
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5263978097178368,
+      "learning_rate": 6.916780481956139e-06,
+      "loss": 0.8418,
+      "step": 3797
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2133475708321293,
+      "learning_rate": 6.911815723959207e-06,
+      "loss": 0.8363,
+      "step": 3798
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8835223338140338,
+      "learning_rate": 6.906851807268799e-06,
+      "loss": 0.3636,
+      "step": 3799
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5157267914833537,
+      "learning_rate": 6.901888733237226e-06,
+      "loss": 0.8091,
+      "step": 3800
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1870753774516376,
+      "learning_rate": 6.896926503216557e-06,
+      "loss": 0.8568,
+      "step": 3801
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4105007868198842,
+      "learning_rate": 6.891965118558644e-06,
+      "loss": 0.8421,
+      "step": 3802
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5210852221831315,
+      "learning_rate": 6.887004580615107e-06,
+      "loss": 0.8451,
+      "step": 3803
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3933183497037784,
+      "learning_rate": 6.882044890737336e-06,
+      "loss": 0.8844,
+      "step": 3804
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.269943716095466,
+      "learning_rate": 6.8770860502764815e-06,
+      "loss": 0.8371,
+      "step": 3805
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4708111667589325,
+      "learning_rate": 6.8721280605834706e-06,
+      "loss": 0.8056,
+      "step": 3806
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.739870615658819,
+      "learning_rate": 6.867170923008998e-06,
+      "loss": 0.8203,
+      "step": 3807
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.455825154027949,
+      "learning_rate": 6.862214638903519e-06,
+      "loss": 0.8448,
+      "step": 3808
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.348078657376387,
+      "learning_rate": 6.857259209617269e-06,
+      "loss": 0.8221,
+      "step": 3809
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4379066975220443,
+      "learning_rate": 6.85230463650024e-06,
+      "loss": 0.9047,
+      "step": 3810
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2474798450139029,
+      "learning_rate": 6.847350920902194e-06,
+      "loss": 0.8018,
+      "step": 3811
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3708274034449286,
+      "learning_rate": 6.842398064172661e-06,
+      "loss": 0.8674,
+      "step": 3812
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.503655790553231,
+      "learning_rate": 6.837446067660937e-06,
+      "loss": 0.8215,
+      "step": 3813
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1957143126908043,
+      "learning_rate": 6.832494932716078e-06,
+      "loss": 0.8852,
+      "step": 3814
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3507048629675866,
+      "learning_rate": 6.82754466068691e-06,
+      "loss": 0.7933,
+      "step": 3815
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5629797401659629,
+      "learning_rate": 6.822595252922033e-06,
+      "loss": 0.7818,
+      "step": 3816
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2167692097110474,
+      "learning_rate": 6.817646710769788e-06,
+      "loss": 0.9114,
+      "step": 3817
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3911668828061785,
+      "learning_rate": 6.812699035578302e-06,
+      "loss": 0.8691,
+      "step": 3818
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.50195036101614,
+      "learning_rate": 6.807752228695459e-06,
+      "loss": 0.9178,
+      "step": 3819
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3963145929777998,
+      "learning_rate": 6.802806291468903e-06,
+      "loss": 0.8161,
+      "step": 3820
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5922067781428673,
+      "learning_rate": 6.797861225246045e-06,
+      "loss": 0.882,
+      "step": 3821
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2555064110249468,
+      "learning_rate": 6.792917031374053e-06,
+      "loss": 0.8769,
+      "step": 3822
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5183302532560468,
+      "learning_rate": 6.787973711199872e-06,
+      "loss": 0.8616,
+      "step": 3823
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3071454118007868,
+      "learning_rate": 6.783031266070186e-06,
+      "loss": 0.8477,
+      "step": 3824
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.364322909970862,
+      "learning_rate": 6.778089697331462e-06,
+      "loss": 0.8059,
+      "step": 3825
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3547880534425194,
+      "learning_rate": 6.773149006329919e-06,
+      "loss": 0.8548,
+      "step": 3826
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2762865702876471,
+      "learning_rate": 6.768209194411533e-06,
+      "loss": 0.8439,
+      "step": 3827
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9905165024925404,
+      "learning_rate": 6.763270262922047e-06,
+      "loss": 0.3443,
+      "step": 3828
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2404020406374674,
+      "learning_rate": 6.758332213206964e-06,
+      "loss": 0.899,
+      "step": 3829
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2130910789160914,
+      "learning_rate": 6.753395046611543e-06,
+      "loss": 0.8132,
+      "step": 3830
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2329448891459043,
+      "learning_rate": 6.748458764480801e-06,
+      "loss": 0.8692,
+      "step": 3831
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2778135636525336,
+      "learning_rate": 6.743523368159526e-06,
+      "loss": 0.7959,
+      "step": 3832
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2908200131654213,
+      "learning_rate": 6.738588858992248e-06,
+      "loss": 0.8436,
+      "step": 3833
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6573904852195145,
+      "learning_rate": 6.733655238323264e-06,
+      "loss": 0.7732,
+      "step": 3834
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3587608717757185,
+      "learning_rate": 6.728722507496636e-06,
+      "loss": 0.8557,
+      "step": 3835
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3749184577507532,
+      "learning_rate": 6.723790667856166e-06,
+      "loss": 0.8109,
+      "step": 3836
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2553640719276358,
+      "learning_rate": 6.718859720745425e-06,
+      "loss": 0.8166,
+      "step": 3837
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4501132363877314,
+      "learning_rate": 6.71392966750775e-06,
+      "loss": 0.8743,
+      "step": 3838
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4810316813464302,
+      "learning_rate": 6.709000509486207e-06,
+      "loss": 0.8756,
+      "step": 3839
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4574098120818715,
+      "learning_rate": 6.704072248023643e-06,
+      "loss": 0.8671,
+      "step": 3840
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.366405744481634,
+      "learning_rate": 6.6991448844626525e-06,
+      "loss": 0.7989,
+      "step": 3841
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3041910391193647,
+      "learning_rate": 6.694218420145586e-06,
+      "loss": 0.8102,
+      "step": 3842
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3346756083941864,
+      "learning_rate": 6.689292856414545e-06,
+      "loss": 0.8424,
+      "step": 3843
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.286790194971462,
+      "learning_rate": 6.684368194611387e-06,
+      "loss": 0.8577,
+      "step": 3844
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4012391141524787,
+      "learning_rate": 6.679444436077735e-06,
+      "loss": 0.8411,
+      "step": 3845
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3709297292637812,
+      "learning_rate": 6.674521582154945e-06,
+      "loss": 0.846,
+      "step": 3846
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3359358059721798,
+      "learning_rate": 6.669599634184145e-06,
+      "loss": 0.8867,
+      "step": 3847
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.351903414177435,
+      "learning_rate": 6.6646785935062086e-06,
+      "loss": 0.8423,
+      "step": 3848
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5836209975158235,
+      "learning_rate": 6.65975846146176e-06,
+      "loss": 0.8498,
+      "step": 3849
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5405361012060257,
+      "learning_rate": 6.654839239391182e-06,
+      "loss": 0.8913,
+      "step": 3850
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.292856445650205,
+      "learning_rate": 6.649920928634607e-06,
+      "loss": 0.8889,
+      "step": 3851
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.510325321293648,
+      "learning_rate": 6.645003530531915e-06,
+      "loss": 0.753,
+      "step": 3852
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.455399105332562,
+      "learning_rate": 6.640087046422739e-06,
+      "loss": 0.8586,
+      "step": 3853
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5319229369754304,
+      "learning_rate": 6.635171477646475e-06,
+      "loss": 0.8262,
+      "step": 3854
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3414757558103527,
+      "learning_rate": 6.630256825542245e-06,
+      "loss": 0.8045,
+      "step": 3855
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.457036105293591,
+      "learning_rate": 6.625343091448948e-06,
+      "loss": 0.8217,
+      "step": 3856
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8820600391053,
+      "learning_rate": 6.620430276705215e-06,
+      "loss": 0.891,
+      "step": 3857
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3313039782711484,
+      "learning_rate": 6.615518382649433e-06,
+      "loss": 0.8003,
+      "step": 3858
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3311225820621155,
+      "learning_rate": 6.610607410619736e-06,
+      "loss": 0.8075,
+      "step": 3859
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2785665087039633,
+      "learning_rate": 6.605697361954009e-06,
+      "loss": 0.8465,
+      "step": 3860
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3221310916172697,
+      "learning_rate": 6.600788237989888e-06,
+      "loss": 0.8124,
+      "step": 3861
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5172563295142003,
+      "learning_rate": 6.595880040064748e-06,
+      "loss": 0.8953,
+      "step": 3862
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8194082430556378,
+      "learning_rate": 6.590972769515722e-06,
+      "loss": 0.3225,
+      "step": 3863
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4939210063369914,
+      "learning_rate": 6.5860664276796844e-06,
+      "loss": 0.8443,
+      "step": 3864
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3179444486619654,
+      "learning_rate": 6.581161015893257e-06,
+      "loss": 0.8405,
+      "step": 3865
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2136823585580507,
+      "learning_rate": 6.576256535492807e-06,
+      "loss": 0.8095,
+      "step": 3866
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3083798710561638,
+      "learning_rate": 6.571352987814459e-06,
+      "loss": 0.8156,
+      "step": 3867
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8883186901333191,
+      "learning_rate": 6.566450374194062e-06,
+      "loss": 0.3456,
+      "step": 3868
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.639756726795095,
+      "learning_rate": 6.561548695967231e-06,
+      "loss": 0.8333,
+      "step": 3869
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3956511205683717,
+      "learning_rate": 6.55664795446932e-06,
+      "loss": 0.7575,
+      "step": 3870
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5620315189185812,
+      "learning_rate": 6.55174815103542e-06,
+      "loss": 0.8712,
+      "step": 3871
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3727895046453484,
+      "learning_rate": 6.546849287000374e-06,
+      "loss": 0.7996,
+      "step": 3872
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3549539133854809,
+      "learning_rate": 6.5419513636987735e-06,
+      "loss": 0.8423,
+      "step": 3873
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.56441936027866,
+      "learning_rate": 6.53705438246494e-06,
+      "loss": 0.8715,
+      "step": 3874
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4146641938289164,
+      "learning_rate": 6.532158344632946e-06,
+      "loss": 0.7786,
+      "step": 3875
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.539858001292727,
+      "learning_rate": 6.527263251536618e-06,
+      "loss": 0.8717,
+      "step": 3876
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2531133748999301,
+      "learning_rate": 6.522369104509501e-06,
+      "loss": 0.748,
+      "step": 3877
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3791909342711943,
+      "learning_rate": 6.517475904884903e-06,
+      "loss": 0.8501,
+      "step": 3878
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2987783329246365,
+      "learning_rate": 6.512583653995867e-06,
+      "loss": 0.8373,
+      "step": 3879
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1898795238177797,
+      "learning_rate": 6.507692353175172e-06,
+      "loss": 0.8104,
+      "step": 3880
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2514483978762605,
+      "learning_rate": 6.502802003755346e-06,
+      "loss": 0.8873,
+      "step": 3881
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.023076055488548,
+      "learning_rate": 6.497912607068655e-06,
+      "loss": 0.3298,
+      "step": 3882
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3154740687149613,
+      "learning_rate": 6.49302416444711e-06,
+      "loss": 0.8374,
+      "step": 3883
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5074828681952899,
+      "learning_rate": 6.4881366772224475e-06,
+      "loss": 0.8782,
+      "step": 3884
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4126928220273647,
+      "learning_rate": 6.48325014672616e-06,
+      "loss": 0.8281,
+      "step": 3885
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4133387736838636,
+      "learning_rate": 6.478364574289475e-06,
+      "loss": 0.8034,
+      "step": 3886
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.490236299846001,
+      "learning_rate": 6.473479961243353e-06,
+      "loss": 0.8885,
+      "step": 3887
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7767728045855768,
+      "learning_rate": 6.468596308918498e-06,
+      "loss": 0.3206,
+      "step": 3888
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4011839264988704,
+      "learning_rate": 6.463713618645354e-06,
+      "loss": 0.7966,
+      "step": 3889
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6773287629211742,
+      "learning_rate": 6.458831891754096e-06,
+      "loss": 0.7919,
+      "step": 3890
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4943671784022743,
+      "learning_rate": 6.453951129574644e-06,
+      "loss": 0.8574,
+      "step": 3891
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.541524229015405,
+      "learning_rate": 6.449071333436654e-06,
+      "loss": 0.879,
+      "step": 3892
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4072891310124973,
+      "learning_rate": 6.4441925046695135e-06,
+      "loss": 0.7966,
+      "step": 3893
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4710360311103452,
+      "learning_rate": 6.43931464460235e-06,
+      "loss": 0.8769,
+      "step": 3894
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5640893952661052,
+      "learning_rate": 6.434437754564031e-06,
+      "loss": 0.8641,
+      "step": 3895
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.460537627518122,
+      "learning_rate": 6.42956183588315e-06,
+      "loss": 0.7523,
+      "step": 3896
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6129954271919813,
+      "learning_rate": 6.424686889888044e-06,
+      "loss": 0.8699,
+      "step": 3897
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.358066823454367,
+      "learning_rate": 6.419812917906788e-06,
+      "loss": 0.9098,
+      "step": 3898
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4520916308776415,
+      "learning_rate": 6.414939921267176e-06,
+      "loss": 0.8435,
+      "step": 3899
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5296102753201162,
+      "learning_rate": 6.410067901296755e-06,
+      "loss": 0.7561,
+      "step": 3900
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0983438001473078,
+      "learning_rate": 6.405196859322793e-06,
+      "loss": 0.7667,
+      "step": 3901
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4200359407545446,
+      "learning_rate": 6.400326796672302e-06,
+      "loss": 0.8492,
+      "step": 3902
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4481202966194264,
+      "learning_rate": 6.395457714672013e-06,
+      "loss": 0.7654,
+      "step": 3903
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3408678508980298,
+      "learning_rate": 6.390589614648401e-06,
+      "loss": 0.7966,
+      "step": 3904
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1945157426497406,
+      "learning_rate": 6.38572249792768e-06,
+      "loss": 0.8496,
+      "step": 3905
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3099271969298356,
+      "learning_rate": 6.38085636583577e-06,
+      "loss": 0.7965,
+      "step": 3906
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.378734434732506,
+      "learning_rate": 6.375991219698352e-06,
+      "loss": 0.8123,
+      "step": 3907
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.267391995693451,
+      "learning_rate": 6.371127060840823e-06,
+      "loss": 0.7871,
+      "step": 3908
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3450140451845496,
+      "learning_rate": 6.366263890588311e-06,
+      "loss": 0.8619,
+      "step": 3909
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3117907195656504,
+      "learning_rate": 6.3614017102656815e-06,
+      "loss": 0.8353,
+      "step": 3910
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6675054411463346,
+      "learning_rate": 6.356540521197525e-06,
+      "loss": 0.8298,
+      "step": 3911
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.198745127664038,
+      "learning_rate": 6.351680324708162e-06,
+      "loss": 0.8509,
+      "step": 3912
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5782425991233708,
+      "learning_rate": 6.3468211221216425e-06,
+      "loss": 0.8921,
+      "step": 3913
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.584570323596617,
+      "learning_rate": 6.341962914761753e-06,
+      "loss": 0.8397,
+      "step": 3914
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4637526065856552,
+      "learning_rate": 6.337105703952e-06,
+      "loss": 0.8771,
+      "step": 3915
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.366055254760888,
+      "learning_rate": 6.3322494910156195e-06,
+      "loss": 0.8808,
+      "step": 3916
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.373528994815268,
+      "learning_rate": 6.327394277275582e-06,
+      "loss": 0.8317,
+      "step": 3917
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2917610669601896,
+      "learning_rate": 6.322540064054578e-06,
+      "loss": 0.8568,
+      "step": 3918
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.494186823884887,
+      "learning_rate": 6.317686852675029e-06,
+      "loss": 0.9097,
+      "step": 3919
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8268577324928732,
+      "learning_rate": 6.312834644459086e-06,
+      "loss": 0.3386,
+      "step": 3920
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6323789552576708,
+      "learning_rate": 6.307983440728626e-06,
+      "loss": 0.7836,
+      "step": 3921
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5594724936195266,
+      "learning_rate": 6.303133242805244e-06,
+      "loss": 0.8737,
+      "step": 3922
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.094742007491149,
+      "learning_rate": 6.2982840520102725e-06,
+      "loss": 0.7931,
+      "step": 3923
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3402914607399476,
+      "learning_rate": 6.293435869664766e-06,
+      "loss": 0.7525,
+      "step": 3924
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4621459366784988,
+      "learning_rate": 6.2885886970894986e-06,
+      "loss": 0.8691,
+      "step": 3925
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3047944409834007,
+      "learning_rate": 6.283742535604973e-06,
+      "loss": 0.8786,
+      "step": 3926
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.465103721437772,
+      "learning_rate": 6.2788973865314265e-06,
+      "loss": 0.8092,
+      "step": 3927
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7942771827985942,
+      "learning_rate": 6.274053251188798e-06,
+      "loss": 0.3504,
+      "step": 3928
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.597226951149117,
+      "learning_rate": 6.269210130896773e-06,
+      "loss": 0.7935,
+      "step": 3929
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4703019859691049,
+      "learning_rate": 6.26436802697475e-06,
+      "loss": 0.8158,
+      "step": 3930
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5427191665246727,
+      "learning_rate": 6.259526940741848e-06,
+      "loss": 0.7648,
+      "step": 3931
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2875201629071407,
+      "learning_rate": 6.2546868735169134e-06,
+      "loss": 0.8319,
+      "step": 3932
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3976696369725188,
+      "learning_rate": 6.249847826618518e-06,
+      "loss": 0.881,
+      "step": 3933
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3154673383347013,
+      "learning_rate": 6.245009801364945e-06,
+      "loss": 0.8783,
+      "step": 3934
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5589811326473262,
+      "learning_rate": 6.240172799074207e-06,
+      "loss": 0.8259,
+      "step": 3935
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3403328129500314,
+      "learning_rate": 6.2353368210640415e-06,
+      "loss": 0.7911,
+      "step": 3936
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4445005552611252,
+      "learning_rate": 6.230501868651899e-06,
+      "loss": 0.7871,
+      "step": 3937
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3519677742509113,
+      "learning_rate": 6.225667943154956e-06,
+      "loss": 0.8888,
+      "step": 3938
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3556500692872318,
+      "learning_rate": 6.2208350458901035e-06,
+      "loss": 0.8431,
+      "step": 3939
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3622933561612653,
+      "learning_rate": 6.216003178173959e-06,
+      "loss": 0.8037,
+      "step": 3940
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6732516247414286,
+      "learning_rate": 6.211172341322853e-06,
+      "loss": 0.8262,
+      "step": 3941
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5471596227198163,
+      "learning_rate": 6.206342536652841e-06,
+      "loss": 0.8002,
+      "step": 3942
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3813001362247834,
+      "learning_rate": 6.201513765479699e-06,
+      "loss": 0.7342,
+      "step": 3943
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2193124669871471,
+      "learning_rate": 6.196686029118909e-06,
+      "loss": 0.7827,
+      "step": 3944
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3880225036788776,
+      "learning_rate": 6.191859328885685e-06,
+      "loss": 0.8141,
+      "step": 3945
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.409709568420695,
+      "learning_rate": 6.1870336660949526e-06,
+      "loss": 0.8803,
+      "step": 3946
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2217121973751148,
+      "learning_rate": 6.182209042061353e-06,
+      "loss": 0.7302,
+      "step": 3947
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1732205224295202,
+      "learning_rate": 6.177385458099248e-06,
+      "loss": 0.8009,
+      "step": 3948
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7438549960378901,
+      "learning_rate": 6.1725629155227195e-06,
+      "loss": 0.3495,
+      "step": 3949
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8350387263731213,
+      "learning_rate": 6.167741415645552e-06,
+      "loss": 0.3581,
+      "step": 3950
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.39328924804151,
+      "learning_rate": 6.16292095978126e-06,
+      "loss": 0.7978,
+      "step": 3951
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3524587106420576,
+      "learning_rate": 6.158101549243072e-06,
+      "loss": 0.9403,
+      "step": 3952
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.29334385046575,
+      "learning_rate": 6.153283185343921e-06,
+      "loss": 0.8599,
+      "step": 3953
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.310218875874768,
+      "learning_rate": 6.148465869396468e-06,
+      "loss": 0.8455,
+      "step": 3954
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.342882829545087,
+      "learning_rate": 6.14364960271308e-06,
+      "loss": 0.8817,
+      "step": 3955
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3753952730236263,
+      "learning_rate": 6.13883438660584e-06,
+      "loss": 0.8258,
+      "step": 3956
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2785663105760143,
+      "learning_rate": 6.134020222386544e-06,
+      "loss": 0.9149,
+      "step": 3957
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4914870552984836,
+      "learning_rate": 6.1292071113667125e-06,
+      "loss": 0.8149,
+      "step": 3958
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5993073717486268,
+      "learning_rate": 6.124395054857557e-06,
+      "loss": 0.84,
+      "step": 3959
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9809127005303395,
+      "learning_rate": 6.119584054170022e-06,
+      "loss": 0.8733,
+      "step": 3960
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5134555243648176,
+      "learning_rate": 6.114774110614755e-06,
+      "loss": 0.8816,
+      "step": 3961
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3341747472038046,
+      "learning_rate": 6.109965225502119e-06,
+      "loss": 0.8165,
+      "step": 3962
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2948750412643426,
+      "learning_rate": 6.105157400142183e-06,
+      "loss": 0.737,
+      "step": 3963
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6163825528863194,
+      "learning_rate": 6.100350635844731e-06,
+      "loss": 0.8599,
+      "step": 3964
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.530450086494435,
+      "learning_rate": 6.095544933919265e-06,
+      "loss": 0.9085,
+      "step": 3965
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1907440205530424,
+      "learning_rate": 6.090740295674984e-06,
+      "loss": 0.8493,
+      "step": 3966
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.329265067296999,
+      "learning_rate": 6.085936722420806e-06,
+      "loss": 0.7813,
+      "step": 3967
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4042984182537657,
+      "learning_rate": 6.081134215465358e-06,
+      "loss": 0.9183,
+      "step": 3968
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3320858488331586,
+      "learning_rate": 6.076332776116971e-06,
+      "loss": 0.7692,
+      "step": 3969
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3569285439681797,
+      "learning_rate": 6.071532405683691e-06,
+      "loss": 0.8311,
+      "step": 3970
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6704253146128583,
+      "learning_rate": 6.066733105473279e-06,
+      "loss": 0.8487,
+      "step": 3971
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4449896039349552,
+      "learning_rate": 6.061934876793183e-06,
+      "loss": 0.8463,
+      "step": 3972
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6424342794403093,
+      "learning_rate": 6.057137720950583e-06,
+      "loss": 0.8732,
+      "step": 3973
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.878657248510857,
+      "learning_rate": 6.052341639252356e-06,
+      "loss": 0.317,
+      "step": 3974
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.270700163065299,
+      "learning_rate": 6.0475466330050815e-06,
+      "loss": 0.8394,
+      "step": 3975
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2124278507168977,
+      "learning_rate": 6.042752703515054e-06,
+      "loss": 0.8106,
+      "step": 3976
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4378982470423594,
+      "learning_rate": 6.037959852088275e-06,
+      "loss": 0.8534,
+      "step": 3977
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7260439037806536,
+      "learning_rate": 6.0331680800304436e-06,
+      "loss": 0.9067,
+      "step": 3978
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2420710315624883,
+      "learning_rate": 6.028377388646972e-06,
+      "loss": 0.8034,
+      "step": 3979
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.262116983473999,
+      "learning_rate": 6.023587779242981e-06,
+      "loss": 0.8314,
+      "step": 3980
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6167055312494005,
+      "learning_rate": 6.0187992531232895e-06,
+      "loss": 0.8078,
+      "step": 3981
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4401756062530737,
+      "learning_rate": 6.014011811592424e-06,
+      "loss": 0.8451,
+      "step": 3982
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1415996845187493,
+      "learning_rate": 6.0092254559546134e-06,
+      "loss": 0.8364,
+      "step": 3983
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2672171012464313,
+      "learning_rate": 6.004440187513798e-06,
+      "loss": 0.8644,
+      "step": 3984
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3558756309695346,
+      "learning_rate": 5.999656007573612e-06,
+      "loss": 0.8497,
+      "step": 3985
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0870451959470988,
+      "learning_rate": 5.9948729174373975e-06,
+      "loss": 0.8393,
+      "step": 3986
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3579787635137754,
+      "learning_rate": 5.990090918408208e-06,
+      "loss": 0.8521,
+      "step": 3987
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4161645140145918,
+      "learning_rate": 5.985310011788781e-06,
+      "loss": 0.8378,
+      "step": 3988
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1578293920261569,
+      "learning_rate": 5.9805301988815735e-06,
+      "loss": 0.8489,
+      "step": 3989
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4376929182196696,
+      "learning_rate": 5.975751480988739e-06,
+      "loss": 0.8539,
+      "step": 3990
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3919597358876348,
+      "learning_rate": 5.970973859412128e-06,
+      "loss": 0.8758,
+      "step": 3991
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5679502648364925,
+      "learning_rate": 5.966197335453299e-06,
+      "loss": 0.7203,
+      "step": 3992
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5038221224254285,
+      "learning_rate": 5.96142191041351e-06,
+      "loss": 0.8686,
+      "step": 3993
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.259228608163467,
+      "learning_rate": 5.956647585593713e-06,
+      "loss": 0.7934,
+      "step": 3994
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4696726586351705,
+      "learning_rate": 5.9518743622945715e-06,
+      "loss": 0.9047,
+      "step": 3995
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6539421607638625,
+      "learning_rate": 5.9471022418164425e-06,
+      "loss": 0.9049,
+      "step": 3996
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.798190730860522,
+      "learning_rate": 5.942331225459379e-06,
+      "loss": 0.8001,
+      "step": 3997
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4203691466788804,
+      "learning_rate": 5.937561314523142e-06,
+      "loss": 0.8261,
+      "step": 3998
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3277820200394561,
+      "learning_rate": 5.93279251030718e-06,
+      "loss": 0.8414,
+      "step": 3999
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2097297398310745,
+      "learning_rate": 5.928024814110659e-06,
+      "loss": 0.8371,
+      "step": 4000
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4452443989612773,
+      "learning_rate": 5.923258227232418e-06,
+      "loss": 0.8248,
+      "step": 4001
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6710395480287648,
+      "learning_rate": 5.918492750971012e-06,
+      "loss": 0.7991,
+      "step": 4002
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2142557982160682,
+      "learning_rate": 5.913728386624691e-06,
+      "loss": 0.8768,
+      "step": 4003
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4379956360193322,
+      "learning_rate": 5.908965135491394e-06,
+      "loss": 0.8802,
+      "step": 4004
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.462085253898475,
+      "learning_rate": 5.904202998868765e-06,
+      "loss": 0.8475,
+      "step": 4005
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5163560774254725,
+      "learning_rate": 5.899441978054141e-06,
+      "loss": 0.8627,
+      "step": 4006
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3948609192177073,
+      "learning_rate": 5.894682074344554e-06,
+      "loss": 0.8076,
+      "step": 4007
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3766523736177794,
+      "learning_rate": 5.889923289036732e-06,
+      "loss": 0.862,
+      "step": 4008
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3971974033074275,
+      "learning_rate": 5.885165623427106e-06,
+      "loss": 0.8892,
+      "step": 4009
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2614007769122986,
+      "learning_rate": 5.880409078811784e-06,
+      "loss": 0.8506,
+      "step": 4010
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4371098515000917,
+      "learning_rate": 5.875653656486588e-06,
+      "loss": 0.8635,
+      "step": 4011
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5648057066453844,
+      "learning_rate": 5.870899357747029e-06,
+      "loss": 0.842,
+      "step": 4012
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5073605580300786,
+      "learning_rate": 5.8661461838882995e-06,
+      "loss": 0.795,
+      "step": 4013
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2816969417989925,
+      "learning_rate": 5.861394136205301e-06,
+      "loss": 0.8809,
+      "step": 4014
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4620882018500523,
+      "learning_rate": 5.8566432159926215e-06,
+      "loss": 0.817,
+      "step": 4015
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7464170625974282,
+      "learning_rate": 5.851893424544541e-06,
+      "loss": 0.8155,
+      "step": 4016
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5602237458784107,
+      "learning_rate": 5.8471447631550335e-06,
+      "loss": 0.9038,
+      "step": 4017
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4702517664139347,
+      "learning_rate": 5.842397233117771e-06,
+      "loss": 0.9115,
+      "step": 4018
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3053117054077699,
+      "learning_rate": 5.837650835726105e-06,
+      "loss": 0.8173,
+      "step": 4019
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.247199273666786,
+      "learning_rate": 5.832905572273084e-06,
+      "loss": 0.8227,
+      "step": 4020
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5306365270694602,
+      "learning_rate": 5.828161444051456e-06,
+      "loss": 0.9002,
+      "step": 4021
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.471143086801859,
+      "learning_rate": 5.823418452353649e-06,
+      "loss": 0.8279,
+      "step": 4022
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6208905104119133,
+      "learning_rate": 5.8186765984717795e-06,
+      "loss": 0.8232,
+      "step": 4023
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.456329880005253,
+      "learning_rate": 5.813935883697668e-06,
+      "loss": 0.8014,
+      "step": 4024
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2398354316725633,
+      "learning_rate": 5.809196309322811e-06,
+      "loss": 0.8212,
+      "step": 4025
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2797954235807563,
+      "learning_rate": 5.8044578766383945e-06,
+      "loss": 0.8271,
+      "step": 4026
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5748992152078602,
+      "learning_rate": 5.799720586935304e-06,
+      "loss": 0.6803,
+      "step": 4027
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2863578740672688,
+      "learning_rate": 5.794984441504114e-06,
+      "loss": 0.8957,
+      "step": 4028
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8837200664754753,
+      "learning_rate": 5.790249441635067e-06,
+      "loss": 0.3432,
+      "step": 4029
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4133217541661536,
+      "learning_rate": 5.785515588618115e-06,
+      "loss": 0.8378,
+      "step": 4030
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3577327838758657,
+      "learning_rate": 5.780782883742892e-06,
+      "loss": 0.7748,
+      "step": 4031
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4261684936968158,
+      "learning_rate": 5.7760513282987164e-06,
+      "loss": 0.7836,
+      "step": 4032
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3547482066373593,
+      "learning_rate": 5.771320923574588e-06,
+      "loss": 0.8069,
+      "step": 4033
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4813321648114053,
+      "learning_rate": 5.7665916708592095e-06,
+      "loss": 0.8125,
+      "step": 4034
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4367805856328193,
+      "learning_rate": 5.761863571440948e-06,
+      "loss": 0.8009,
+      "step": 4035
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2375233595472444,
+      "learning_rate": 5.757136626607881e-06,
+      "loss": 0.7962,
+      "step": 4036
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.458239181209716,
+      "learning_rate": 5.75241083764775e-06,
+      "loss": 0.8251,
+      "step": 4037
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.382116358714955,
+      "learning_rate": 5.747686205847991e-06,
+      "loss": 0.8765,
+      "step": 4038
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3325681946599095,
+      "learning_rate": 5.742962732495727e-06,
+      "loss": 0.8136,
+      "step": 4039
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4997090663499903,
+      "learning_rate": 5.738240418877758e-06,
+      "loss": 0.8496,
+      "step": 4040
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.217586496867139,
+      "learning_rate": 5.7335192662805784e-06,
+      "loss": 0.8843,
+      "step": 4041
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4073555680070253,
+      "learning_rate": 5.728799275990352e-06,
+      "loss": 0.7038,
+      "step": 4042
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8884718559821404,
+      "learning_rate": 5.7240804492929435e-06,
+      "loss": 0.3639,
+      "step": 4043
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5819973259102533,
+      "learning_rate": 5.7193627874738874e-06,
+      "loss": 0.8256,
+      "step": 4044
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.870378071687995,
+      "learning_rate": 5.714646291818401e-06,
+      "loss": 0.8135,
+      "step": 4045
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9817251565621882,
+      "learning_rate": 5.709930963611394e-06,
+      "loss": 0.345,
+      "step": 4046
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3398380418439002,
+      "learning_rate": 5.70521680413745e-06,
+      "loss": 0.8668,
+      "step": 4047
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6440664391106998,
+      "learning_rate": 5.700503814680831e-06,
+      "loss": 0.7886,
+      "step": 4048
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3966817399587208,
+      "learning_rate": 5.695791996525488e-06,
+      "loss": 0.8683,
+      "step": 4049
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3726707841476058,
+      "learning_rate": 5.691081350955061e-06,
+      "loss": 0.8752,
+      "step": 4050
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.361613889444106,
+      "learning_rate": 5.686371879252841e-06,
+      "loss": 0.7818,
+      "step": 4051
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7504538730873453,
+      "learning_rate": 5.681663582701827e-06,
+      "loss": 0.7762,
+      "step": 4052
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3889290347522594,
+      "learning_rate": 5.676956462584693e-06,
+      "loss": 0.8072,
+      "step": 4053
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3255283767148927,
+      "learning_rate": 5.672250520183784e-06,
+      "loss": 0.8584,
+      "step": 4054
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3036688094172493,
+      "learning_rate": 5.6675457567811235e-06,
+      "loss": 0.8133,
+      "step": 4055
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.361300981954292,
+      "learning_rate": 5.662842173658429e-06,
+      "loss": 0.8778,
+      "step": 4056
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4547585962194465,
+      "learning_rate": 5.65813977209708e-06,
+      "loss": 0.7704,
+      "step": 4057
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.448547585736928,
+      "learning_rate": 5.653438553378137e-06,
+      "loss": 0.8397,
+      "step": 4058
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4918102416633083,
+      "learning_rate": 5.648738518782346e-06,
+      "loss": 0.8288,
+      "step": 4059
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8084724301854964,
+      "learning_rate": 5.64403966959013e-06,
+      "loss": 0.3354,
+      "step": 4060
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5034145269698331,
+      "learning_rate": 5.639342007081581e-06,
+      "loss": 0.8158,
+      "step": 4061
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.493635524194898,
+      "learning_rate": 5.6346455325364664e-06,
+      "loss": 0.8768,
+      "step": 4062
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4314812939754407,
+      "learning_rate": 5.629950247234246e-06,
+      "loss": 0.8369,
+      "step": 4063
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2395893619862246,
+      "learning_rate": 5.625256152454035e-06,
+      "loss": 0.857,
+      "step": 4064
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5027821643797372,
+      "learning_rate": 5.620563249474642e-06,
+      "loss": 0.8198,
+      "step": 4065
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3521031176355667,
+      "learning_rate": 5.615871539574539e-06,
+      "loss": 0.8384,
+      "step": 4066
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3222561811957922,
+      "learning_rate": 5.611181024031875e-06,
+      "loss": 0.8643,
+      "step": 4067
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.359295431636533,
+      "learning_rate": 5.606491704124482e-06,
+      "loss": 0.8021,
+      "step": 4068
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3018932214868242,
+      "learning_rate": 5.6018035811298566e-06,
+      "loss": 0.7823,
+      "step": 4069
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.309949127057004,
+      "learning_rate": 5.597116656325169e-06,
+      "loss": 0.8076,
+      "step": 4070
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.735967448999417,
+      "learning_rate": 5.5924309309872705e-06,
+      "loss": 0.8342,
+      "step": 4071
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3373301623399914,
+      "learning_rate": 5.587746406392689e-06,
+      "loss": 0.8414,
+      "step": 4072
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3232984126205791,
+      "learning_rate": 5.583063083817604e-06,
+      "loss": 0.8589,
+      "step": 4073
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3086289593244262,
+      "learning_rate": 5.578380964537889e-06,
+      "loss": 0.8777,
+      "step": 4074
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6483532694227965,
+      "learning_rate": 5.573700049829083e-06,
+      "loss": 0.7961,
+      "step": 4075
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4501485643654826,
+      "learning_rate": 5.569020340966398e-06,
+      "loss": 0.8308,
+      "step": 4076
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4840859501892305,
+      "learning_rate": 5.564341839224707e-06,
+      "loss": 0.8698,
+      "step": 4077
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3117336579066952,
+      "learning_rate": 5.55966454587857e-06,
+      "loss": 0.8803,
+      "step": 4078
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3251758410963688,
+      "learning_rate": 5.5549884622022165e-06,
+      "loss": 0.7994,
+      "step": 4079
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4255004727189236,
+      "learning_rate": 5.550313589469525e-06,
+      "loss": 0.8774,
+      "step": 4080
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5354747247739804,
+      "learning_rate": 5.54563992895407e-06,
+      "loss": 0.8786,
+      "step": 4081
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4491303743422195,
+      "learning_rate": 5.540967481929085e-06,
+      "loss": 0.8136,
+      "step": 4082
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3099310955432473,
+      "learning_rate": 5.536296249667472e-06,
+      "loss": 0.8844,
+      "step": 4083
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.368648701206005,
+      "learning_rate": 5.531626233441802e-06,
+      "loss": 0.8422,
+      "step": 4084
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1280260924222627,
+      "learning_rate": 5.526957434524319e-06,
+      "loss": 0.7933,
+      "step": 4085
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3166774369964485,
+      "learning_rate": 5.522289854186928e-06,
+      "loss": 0.8385,
+      "step": 4086
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3558668121006436,
+      "learning_rate": 5.517623493701213e-06,
+      "loss": 0.8111,
+      "step": 4087
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.524344565489292,
+      "learning_rate": 5.512958354338416e-06,
+      "loss": 0.8015,
+      "step": 4088
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2776905516453796,
+      "learning_rate": 5.508294437369445e-06,
+      "loss": 0.8844,
+      "step": 4089
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.465207931388753,
+      "learning_rate": 5.503631744064888e-06,
+      "loss": 0.8045,
+      "step": 4090
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3721645713141295,
+      "learning_rate": 5.498970275694987e-06,
+      "loss": 0.9016,
+      "step": 4091
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5834572369998123,
+      "learning_rate": 5.494310033529651e-06,
+      "loss": 0.8416,
+      "step": 4092
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2643175833799456,
+      "learning_rate": 5.489651018838462e-06,
+      "loss": 0.8088,
+      "step": 4093
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2772354625449334,
+      "learning_rate": 5.48499323289067e-06,
+      "loss": 0.8256,
+      "step": 4094
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5651003288236798,
+      "learning_rate": 5.480336676955174e-06,
+      "loss": 0.8689,
+      "step": 4095
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4412919686663415,
+      "learning_rate": 5.475681352300549e-06,
+      "loss": 0.8293,
+      "step": 4096
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9230907750547812,
+      "learning_rate": 5.471027260195043e-06,
+      "loss": 0.8567,
+      "step": 4097
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2018629020863378,
+      "learning_rate": 5.466374401906552e-06,
+      "loss": 0.9062,
+      "step": 4098
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3295145899573564,
+      "learning_rate": 5.461722778702641e-06,
+      "loss": 0.8589,
+      "step": 4099
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.855081309368814,
+      "learning_rate": 5.457072391850543e-06,
+      "loss": 0.3295,
+      "step": 4100
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.405920672266167,
+      "learning_rate": 5.45242324261716e-06,
+      "loss": 0.8583,
+      "step": 4101
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4551644114057882,
+      "learning_rate": 5.447775332269032e-06,
+      "loss": 0.796,
+      "step": 4102
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8446735729888617,
+      "learning_rate": 5.443128662072388e-06,
+      "loss": 0.8019,
+      "step": 4103
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4738961187421096,
+      "learning_rate": 5.43848323329311e-06,
+      "loss": 0.8213,
+      "step": 4104
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.565910716407285,
+      "learning_rate": 5.433839047196738e-06,
+      "loss": 0.7797,
+      "step": 4105
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6286963314989293,
+      "learning_rate": 5.429196105048473e-06,
+      "loss": 0.8268,
+      "step": 4106
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2586299811996409,
+      "learning_rate": 5.424554408113188e-06,
+      "loss": 0.8165,
+      "step": 4107
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3218611626176995,
+      "learning_rate": 5.4199139576554046e-06,
+      "loss": 0.8007,
+      "step": 4108
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8196061249231192,
+      "learning_rate": 5.4152747549393055e-06,
+      "loss": 0.3346,
+      "step": 4109
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2826753355403062,
+      "learning_rate": 5.410636801228748e-06,
+      "loss": 0.9353,
+      "step": 4110
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3633835161211914,
+      "learning_rate": 5.406000097787226e-06,
+      "loss": 0.8477,
+      "step": 4111
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1942060595688722,
+      "learning_rate": 5.4013646458779175e-06,
+      "loss": 0.7308,
+      "step": 4112
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.504643966514461,
+      "learning_rate": 5.396730446763641e-06,
+      "loss": 0.8846,
+      "step": 4113
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3896823833337928,
+      "learning_rate": 5.392097501706877e-06,
+      "loss": 0.8687,
+      "step": 4114
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3155873132696965,
+      "learning_rate": 5.38746581196977e-06,
+      "loss": 0.8437,
+      "step": 4115
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4463019772918246,
+      "learning_rate": 5.382835378814129e-06,
+      "loss": 0.8511,
+      "step": 4116
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.704193711726875,
+      "learning_rate": 5.378206203501397e-06,
+      "loss": 0.8727,
+      "step": 4117
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3219415945216757,
+      "learning_rate": 5.373578287292694e-06,
+      "loss": 0.8379,
+      "step": 4118
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3201682146648561,
+      "learning_rate": 5.3689516314488e-06,
+      "loss": 0.7444,
+      "step": 4119
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4252850701223332,
+      "learning_rate": 5.364326237230135e-06,
+      "loss": 0.8247,
+      "step": 4120
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.536274852478914,
+      "learning_rate": 5.359702105896783e-06,
+      "loss": 0.7844,
+      "step": 4121
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4154694233808633,
+      "learning_rate": 5.355079238708487e-06,
+      "loss": 0.8705,
+      "step": 4122
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3561580606470416,
+      "learning_rate": 5.350457636924654e-06,
+      "loss": 0.8456,
+      "step": 4123
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3336741022385572,
+      "learning_rate": 5.345837301804317e-06,
+      "loss": 0.8787,
+      "step": 4124
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3689592869889338,
+      "learning_rate": 5.341218234606192e-06,
+      "loss": 0.8937,
+      "step": 4125
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7517368444692838,
+      "learning_rate": 5.336600436588644e-06,
+      "loss": 0.8158,
+      "step": 4126
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1664041462136399,
+      "learning_rate": 5.331983909009685e-06,
+      "loss": 0.8189,
+      "step": 4127
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4077388134204758,
+      "learning_rate": 5.327368653126978e-06,
+      "loss": 0.8371,
+      "step": 4128
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3650620502512707,
+      "learning_rate": 5.322754670197859e-06,
+      "loss": 0.894,
+      "step": 4129
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.716730906376721,
+      "learning_rate": 5.318141961479293e-06,
+      "loss": 0.866,
+      "step": 4130
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2337296110005247,
+      "learning_rate": 5.31353052822791e-06,
+      "loss": 0.828,
+      "step": 4131
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3861016246877065,
+      "learning_rate": 5.3089203717e-06,
+      "loss": 0.8222,
+      "step": 4132
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4938032816617672,
+      "learning_rate": 5.304311493151486e-06,
+      "loss": 0.888,
+      "step": 4133
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.71518135983593,
+      "learning_rate": 5.299703893837963e-06,
+      "loss": 0.8306,
+      "step": 4134
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2921420447993095,
+      "learning_rate": 5.2950975750146635e-06,
+      "loss": 0.8228,
+      "step": 4135
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4195008302976238,
+      "learning_rate": 5.290492537936473e-06,
+      "loss": 0.835,
+      "step": 4136
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3045007352867344,
+      "learning_rate": 5.285888783857935e-06,
+      "loss": 0.8481,
+      "step": 4137
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4921538888431292,
+      "learning_rate": 5.281286314033236e-06,
+      "loss": 0.7515,
+      "step": 4138
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3416922138380263,
+      "learning_rate": 5.27668512971622e-06,
+      "loss": 0.9012,
+      "step": 4139
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5248734889505318,
+      "learning_rate": 5.27208523216037e-06,
+      "loss": 0.8502,
+      "step": 4140
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4403566128573408,
+      "learning_rate": 5.267486622618833e-06,
+      "loss": 0.8313,
+      "step": 4141
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.188131326529266,
+      "learning_rate": 5.262889302344391e-06,
+      "loss": 0.8126,
+      "step": 4142
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6992531400247863,
+      "learning_rate": 5.2582932725894785e-06,
+      "loss": 0.7718,
+      "step": 4143
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5632700810564857,
+      "learning_rate": 5.253698534606186e-06,
+      "loss": 0.8667,
+      "step": 4144
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3871500425841599,
+      "learning_rate": 5.249105089646252e-06,
+      "loss": 0.8133,
+      "step": 4145
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3471871057017366,
+      "learning_rate": 5.244512938961044e-06,
+      "loss": 0.8264,
+      "step": 4146
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4019634779215433,
+      "learning_rate": 5.239922083801597e-06,
+      "loss": 0.8421,
+      "step": 4147
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8824573214978029,
+      "learning_rate": 5.235332525418588e-06,
+      "loss": 0.8294,
+      "step": 4148
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3021465681139015,
+      "learning_rate": 5.23074426506234e-06,
+      "loss": 0.8288,
+      "step": 4149
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5319928172697292,
+      "learning_rate": 5.226157303982815e-06,
+      "loss": 0.8887,
+      "step": 4150
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5424103245449776,
+      "learning_rate": 5.221571643429637e-06,
+      "loss": 0.8237,
+      "step": 4151
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5012499275533069,
+      "learning_rate": 5.216987284652061e-06,
+      "loss": 0.8309,
+      "step": 4152
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4417295603772156,
+      "learning_rate": 5.21240422889899e-06,
+      "loss": 0.8205,
+      "step": 4153
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4890165621355274,
+      "learning_rate": 5.20782247741898e-06,
+      "loss": 0.8327,
+      "step": 4154
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.428936557855059,
+      "learning_rate": 5.203242031460222e-06,
+      "loss": 0.7972,
+      "step": 4155
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4052990042221027,
+      "learning_rate": 5.1986628922705605e-06,
+      "loss": 0.8333,
+      "step": 4156
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.444153480792433,
+      "learning_rate": 5.194085061097474e-06,
+      "loss": 0.863,
+      "step": 4157
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5130336474227828,
+      "learning_rate": 5.189508539188097e-06,
+      "loss": 0.8071,
+      "step": 4158
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3586003892488454,
+      "learning_rate": 5.1849333277891946e-06,
+      "loss": 0.8332,
+      "step": 4159
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3918935703624078,
+      "learning_rate": 5.180359428147179e-06,
+      "loss": 0.8001,
+      "step": 4160
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4827910555383252,
+      "learning_rate": 5.175786841508113e-06,
+      "loss": 0.8029,
+      "step": 4161
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.430665932091022,
+      "learning_rate": 5.1712155691176865e-06,
+      "loss": 0.8622,
+      "step": 4162
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.931053049112487,
+      "learning_rate": 5.166645612221251e-06,
+      "loss": 0.8721,
+      "step": 4163
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5829165783618364,
+      "learning_rate": 5.162076972063781e-06,
+      "loss": 0.7307,
+      "step": 4164
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6689632168161572,
+      "learning_rate": 5.1575096498899e-06,
+      "loss": 0.8096,
+      "step": 4165
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4894216003403309,
+      "learning_rate": 5.152943646943876e-06,
+      "loss": 0.8247,
+      "step": 4166
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2528460319343926,
+      "learning_rate": 5.148378964469615e-06,
+      "loss": 0.812,
+      "step": 4167
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1767225782572486,
+      "learning_rate": 5.143815603710654e-06,
+      "loss": 0.801,
+      "step": 4168
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5673012875659664,
+      "learning_rate": 5.139253565910185e-06,
+      "loss": 0.7431,
+      "step": 4169
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.370067746445637,
+      "learning_rate": 5.134692852311035e-06,
+      "loss": 0.7779,
+      "step": 4170
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4659902834969942,
+      "learning_rate": 5.1301334641556665e-06,
+      "loss": 0.8194,
+      "step": 4171
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3976580402838306,
+      "learning_rate": 5.125575402686176e-06,
+      "loss": 0.9117,
+      "step": 4172
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2469693862235234,
+      "learning_rate": 5.121018669144313e-06,
+      "loss": 0.8586,
+      "step": 4173
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.691663335992051,
+      "learning_rate": 5.116463264771456e-06,
+      "loss": 0.8383,
+      "step": 4174
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3874890649344223,
+      "learning_rate": 5.111909190808617e-06,
+      "loss": 0.7871,
+      "step": 4175
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7953108033055187,
+      "learning_rate": 5.107356448496459e-06,
+      "loss": 0.7561,
+      "step": 4176
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6476175565051514,
+      "learning_rate": 5.102805039075267e-06,
+      "loss": 0.806,
+      "step": 4177
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2533106065041095,
+      "learning_rate": 5.098254963784979e-06,
+      "loss": 0.8747,
+      "step": 4178
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.530157907089299,
+      "learning_rate": 5.093706223865151e-06,
+      "loss": 0.8378,
+      "step": 4179
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4607250740016877,
+      "learning_rate": 5.089158820554996e-06,
+      "loss": 0.8273,
+      "step": 4180
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.767049993576241,
+      "learning_rate": 5.084612755093346e-06,
+      "loss": 0.3893,
+      "step": 4181
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9875126919626056,
+      "learning_rate": 5.08006802871867e-06,
+      "loss": 0.8459,
+      "step": 4182
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.404159340204978,
+      "learning_rate": 5.075524642669086e-06,
+      "loss": 0.8032,
+      "step": 4183
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9055791336657648,
+      "learning_rate": 5.07098259818233e-06,
+      "loss": 0.3658,
+      "step": 4184
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3789108751775963,
+      "learning_rate": 5.066441896495786e-06,
+      "loss": 0.8222,
+      "step": 4185
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7762124794798146,
+      "learning_rate": 5.061902538846466e-06,
+      "loss": 0.7915,
+      "step": 4186
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.215253719383097,
+      "learning_rate": 5.057364526471008e-06,
+      "loss": 0.8712,
+      "step": 4187
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2735209839901582,
+      "learning_rate": 5.052827860605702e-06,
+      "loss": 0.8714,
+      "step": 4188
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3704509736368349,
+      "learning_rate": 5.048292542486457e-06,
+      "loss": 0.8347,
+      "step": 4189
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2014146102093959,
+      "learning_rate": 5.0437585733488135e-06,
+      "loss": 0.7817,
+      "step": 4190
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1074998254606334,
+      "learning_rate": 5.039225954427953e-06,
+      "loss": 0.8352,
+      "step": 4191
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6573681869190582,
+      "learning_rate": 5.034694686958692e-06,
+      "loss": 0.8487,
+      "step": 4192
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3367422835041438,
+      "learning_rate": 5.030164772175469e-06,
+      "loss": 0.7788,
+      "step": 4193
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6694863247751166,
+      "learning_rate": 5.025636211312351e-06,
+      "loss": 0.8542,
+      "step": 4194
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5172708831212016,
+      "learning_rate": 5.021109005603053e-06,
+      "loss": 0.8969,
+      "step": 4195
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5531020047176003,
+      "learning_rate": 5.016583156280906e-06,
+      "loss": 0.8551,
+      "step": 4196
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.597165706869662,
+      "learning_rate": 5.012058664578871e-06,
+      "loss": 0.8361,
+      "step": 4197
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.565971783366627,
+      "learning_rate": 5.007535531729548e-06,
+      "loss": 0.8746,
+      "step": 4198
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2939333594867741,
+      "learning_rate": 5.003013758965171e-06,
+      "loss": 0.8803,
+      "step": 4199
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.735253318920913,
+      "learning_rate": 4.9984933475175865e-06,
+      "loss": 0.7513,
+      "step": 4200
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3502436733096745,
+      "learning_rate": 4.9939742986182795e-06,
+      "loss": 0.8239,
+      "step": 4201
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.423550296751868,
+      "learning_rate": 4.989456613498368e-06,
+      "loss": 0.7908,
+      "step": 4202
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.344387202080036,
+      "learning_rate": 4.9849402933885915e-06,
+      "loss": 0.868,
+      "step": 4203
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3923112873098311,
+      "learning_rate": 4.980425339519316e-06,
+      "loss": 0.8074,
+      "step": 4204
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4965230360740287,
+      "learning_rate": 4.975911753120548e-06,
+      "loss": 0.8724,
+      "step": 4205
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4713633702983342,
+      "learning_rate": 4.971399535421904e-06,
+      "loss": 0.8581,
+      "step": 4206
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7559516270412578,
+      "learning_rate": 4.966888687652645e-06,
+      "loss": 0.8259,
+      "step": 4207
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5932226488967949,
+      "learning_rate": 4.9623792110416454e-06,
+      "loss": 0.8088,
+      "step": 4208
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3742652573856473,
+      "learning_rate": 4.9578711068174076e-06,
+      "loss": 0.8385,
+      "step": 4209
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.450995553909466,
+      "learning_rate": 4.953364376208072e-06,
+      "loss": 0.805,
+      "step": 4210
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3650553644614798,
+      "learning_rate": 4.948859020441391e-06,
+      "loss": 0.7591,
+      "step": 4211
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3585927709068117,
+      "learning_rate": 4.944355040744745e-06,
+      "loss": 0.8074,
+      "step": 4212
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.566671540443173,
+      "learning_rate": 4.939852438345145e-06,
+      "loss": 0.8423,
+      "step": 4213
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.947401078361531,
+      "learning_rate": 4.93535121446923e-06,
+      "loss": 0.3548,
+      "step": 4214
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2614752070132242,
+      "learning_rate": 4.93085137034325e-06,
+      "loss": 0.8239,
+      "step": 4215
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.377570802384327,
+      "learning_rate": 4.926352907193086e-06,
+      "loss": 0.8483,
+      "step": 4216
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.579806181819675,
+      "learning_rate": 4.921855826244249e-06,
+      "loss": 0.9109,
+      "step": 4217
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2732063435434646,
+      "learning_rate": 4.917360128721865e-06,
+      "loss": 0.8378,
+      "step": 4218
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4535796117038635,
+      "learning_rate": 4.912865815850682e-06,
+      "loss": 0.8732,
+      "step": 4219
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4451757224547972,
+      "learning_rate": 4.908372888855078e-06,
+      "loss": 0.7574,
+      "step": 4220
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1897169937793317,
+      "learning_rate": 4.903881348959055e-06,
+      "loss": 0.8207,
+      "step": 4221
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5176649737698933,
+      "learning_rate": 4.899391197386229e-06,
+      "loss": 0.8761,
+      "step": 4222
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8269583901484135,
+      "learning_rate": 4.894902435359834e-06,
+      "loss": 0.3339,
+      "step": 4223
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9487676254904482,
+      "learning_rate": 4.890415064102744e-06,
+      "loss": 0.3254,
+      "step": 4224
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.433907412820224,
+      "learning_rate": 4.885929084837436e-06,
+      "loss": 0.8873,
+      "step": 4225
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4160452477875352,
+      "learning_rate": 4.8814444987860125e-06,
+      "loss": 0.8446,
+      "step": 4226
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3603272562532365,
+      "learning_rate": 4.876961307170204e-06,
+      "loss": 0.8355,
+      "step": 4227
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2530137748926036,
+      "learning_rate": 4.87247951121135e-06,
+      "loss": 0.8811,
+      "step": 4228
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5037231705437166,
+      "learning_rate": 4.867999112130422e-06,
+      "loss": 0.7822,
+      "step": 4229
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.310163035230178,
+      "learning_rate": 4.863520111147999e-06,
+      "loss": 0.8287,
+      "step": 4230
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.315104518707782,
+      "learning_rate": 4.859042509484283e-06,
+      "loss": 0.8107,
+      "step": 4231
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6038638490701458,
+      "learning_rate": 4.854566308359102e-06,
+      "loss": 0.8739,
+      "step": 4232
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3270457756469651,
+      "learning_rate": 4.850091508991893e-06,
+      "loss": 0.7879,
+      "step": 4233
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6159304107428625,
+      "learning_rate": 4.845618112601712e-06,
+      "loss": 0.7815,
+      "step": 4234
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5436037631042647,
+      "learning_rate": 4.841146120407239e-06,
+      "loss": 0.7474,
+      "step": 4235
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.353050811299745,
+      "learning_rate": 4.836675533626769e-06,
+      "loss": 0.8678,
+      "step": 4236
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5235886462132675,
+      "learning_rate": 4.832206353478213e-06,
+      "loss": 0.8524,
+      "step": 4237
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.357256117655531,
+      "learning_rate": 4.8277385811790946e-06,
+      "loss": 0.7329,
+      "step": 4238
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5672266047468137,
+      "learning_rate": 4.823272217946563e-06,
+      "loss": 0.8705,
+      "step": 4239
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5066919035439938,
+      "learning_rate": 4.8188072649973775e-06,
+      "loss": 0.8568,
+      "step": 4240
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.414692065142792,
+      "learning_rate": 4.8143437235479085e-06,
+      "loss": 0.7798,
+      "step": 4241
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3251039428528075,
+      "learning_rate": 4.809881594814154e-06,
+      "loss": 0.7959,
+      "step": 4242
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3910431541261903,
+      "learning_rate": 4.805420880011723e-06,
+      "loss": 0.8846,
+      "step": 4243
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.383795968377024,
+      "learning_rate": 4.800961580355833e-06,
+      "loss": 0.8464,
+      "step": 4244
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2411846282655783,
+      "learning_rate": 4.7965036970613175e-06,
+      "loss": 0.8007,
+      "step": 4245
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2800983124410865,
+      "learning_rate": 4.792047231342635e-06,
+      "loss": 0.8502,
+      "step": 4246
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5499003115829288,
+      "learning_rate": 4.787592184413843e-06,
+      "loss": 0.8434,
+      "step": 4247
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.471188831042224,
+      "learning_rate": 4.783138557488618e-06,
+      "loss": 0.8329,
+      "step": 4248
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3411880469545743,
+      "learning_rate": 4.778686351780257e-06,
+      "loss": 0.8156,
+      "step": 4249
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4559445157192166,
+      "learning_rate": 4.774235568501656e-06,
+      "loss": 0.8704,
+      "step": 4250
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4099827842008033,
+      "learning_rate": 4.76978620886534e-06,
+      "loss": 0.8314,
+      "step": 4251
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.271745015969375,
+      "learning_rate": 4.765338274083432e-06,
+      "loss": 0.8672,
+      "step": 4252
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3600439886427995,
+      "learning_rate": 4.7608917653676675e-06,
+      "loss": 0.7909,
+      "step": 4253
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3058487355685737,
+      "learning_rate": 4.7564466839294085e-06,
+      "loss": 0.8075,
+      "step": 4254
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5523066327390431,
+      "learning_rate": 4.752003030979612e-06,
+      "loss": 0.8041,
+      "step": 4255
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6513959068900588,
+      "learning_rate": 4.747560807728847e-06,
+      "loss": 0.7848,
+      "step": 4256
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7969387609787683,
+      "learning_rate": 4.743120015387302e-06,
+      "loss": 0.3539,
+      "step": 4257
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6098973638266458,
+      "learning_rate": 4.7386806551647766e-06,
+      "loss": 0.8377,
+      "step": 4258
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5441044008255504,
+      "learning_rate": 4.73424272827067e-06,
+      "loss": 0.905,
+      "step": 4259
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3169301070293704,
+      "learning_rate": 4.729806235913991e-06,
+      "loss": 0.8537,
+      "step": 4260
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4602267913147935,
+      "learning_rate": 4.725371179303371e-06,
+      "loss": 0.822,
+      "step": 4261
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3661101469718007,
+      "learning_rate": 4.720937559647038e-06,
+      "loss": 0.8142,
+      "step": 4262
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4251445341428535,
+      "learning_rate": 4.716505378152827e-06,
+      "loss": 0.8638,
+      "step": 4263
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.402143744388967,
+      "learning_rate": 4.712074636028192e-06,
+      "loss": 0.7855,
+      "step": 4264
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3368081067566315,
+      "learning_rate": 4.7076453344801965e-06,
+      "loss": 0.7672,
+      "step": 4265
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3909078696083208,
+      "learning_rate": 4.703217474715489e-06,
+      "loss": 0.8124,
+      "step": 4266
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.503184016312308,
+      "learning_rate": 4.698791057940349e-06,
+      "loss": 0.8185,
+      "step": 4267
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6566107701150459,
+      "learning_rate": 4.694366085360656e-06,
+      "loss": 0.8099,
+      "step": 4268
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4594392188776675,
+      "learning_rate": 4.689942558181893e-06,
+      "loss": 0.7954,
+      "step": 4269
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1994972154212105,
+      "learning_rate": 4.6855204776091445e-06,
+      "loss": 0.8343,
+      "step": 4270
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3837925817354688,
+      "learning_rate": 4.681099844847117e-06,
+      "loss": 0.8414,
+      "step": 4271
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9752629913080488,
+      "learning_rate": 4.6766806611001046e-06,
+      "loss": 0.3456,
+      "step": 4272
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.320982936348882,
+      "learning_rate": 4.672262927572021e-06,
+      "loss": 0.8191,
+      "step": 4273
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3972569309398184,
+      "learning_rate": 4.667846645466377e-06,
+      "loss": 0.8556,
+      "step": 4274
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2531189656617716,
+      "learning_rate": 4.663431815986284e-06,
+      "loss": 0.7872,
+      "step": 4275
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4050375927399317,
+      "learning_rate": 4.659018440334472e-06,
+      "loss": 0.8213,
+      "step": 4276
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3031584779981,
+      "learning_rate": 4.654606519713258e-06,
+      "loss": 0.7036,
+      "step": 4277
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3272594759133631,
+      "learning_rate": 4.6501960553245785e-06,
+      "loss": 0.8378,
+      "step": 4278
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5452150899706918,
+      "learning_rate": 4.645787048369958e-06,
+      "loss": 0.8572,
+      "step": 4279
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4658813285962644,
+      "learning_rate": 4.641379500050538e-06,
+      "loss": 0.777,
+      "step": 4280
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5286045913044735,
+      "learning_rate": 4.636973411567055e-06,
+      "loss": 0.8519,
+      "step": 4281
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4282003945465973,
+      "learning_rate": 4.632568784119842e-06,
+      "loss": 0.8141,
+      "step": 4282
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2310525568841226,
+      "learning_rate": 4.628165618908851e-06,
+      "loss": 0.8543,
+      "step": 4283
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.454495744840437,
+      "learning_rate": 4.623763917133621e-06,
+      "loss": 0.8202,
+      "step": 4284
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3034967071451053,
+      "learning_rate": 4.619363679993293e-06,
+      "loss": 0.8683,
+      "step": 4285
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4223198103228434,
+      "learning_rate": 4.614964908686617e-06,
+      "loss": 0.848,
+      "step": 4286
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4667652167346432,
+      "learning_rate": 4.610567604411946e-06,
+      "loss": 0.8771,
+      "step": 4287
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4082472188503694,
+      "learning_rate": 4.606171768367213e-06,
+      "loss": 0.8316,
+      "step": 4288
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.164634589612392,
+      "learning_rate": 4.601777401749972e-06,
+      "loss": 0.7953,
+      "step": 4289
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4609849234906007,
+      "learning_rate": 4.597384505757373e-06,
+      "loss": 0.8111,
+      "step": 4290
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2823083056899898,
+      "learning_rate": 4.592993081586159e-06,
+      "loss": 0.8902,
+      "step": 4291
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4016509244023847,
+      "learning_rate": 4.588603130432671e-06,
+      "loss": 0.8317,
+      "step": 4292
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3713189249898754,
+      "learning_rate": 4.58421465349286e-06,
+      "loss": 0.8631,
+      "step": 4293
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.42265759764696,
+      "learning_rate": 4.579827651962264e-06,
+      "loss": 0.8371,
+      "step": 4294
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.40742206406731,
+      "learning_rate": 4.5754421270360195e-06,
+      "loss": 0.8653,
+      "step": 4295
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3160320282092444,
+      "learning_rate": 4.571058079908869e-06,
+      "loss": 0.8299,
+      "step": 4296
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3970158439288505,
+      "learning_rate": 4.566675511775151e-06,
+      "loss": 0.8876,
+      "step": 4297
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4599217412783911,
+      "learning_rate": 4.562294423828794e-06,
+      "loss": 0.8124,
+      "step": 4298
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4014970157708602,
+      "learning_rate": 4.557914817263324e-06,
+      "loss": 0.8323,
+      "step": 4299
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.416185129477895,
+      "learning_rate": 4.553536693271872e-06,
+      "loss": 0.8175,
+      "step": 4300
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3003775146192469,
+      "learning_rate": 4.5491600530471546e-06,
+      "loss": 0.855,
+      "step": 4301
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3429186142555603,
+      "learning_rate": 4.544784897781495e-06,
+      "loss": 0.7553,
+      "step": 4302
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4239382393830073,
+      "learning_rate": 4.540411228666805e-06,
+      "loss": 0.8836,
+      "step": 4303
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5405966780909657,
+      "learning_rate": 4.536039046894584e-06,
+      "loss": 0.9026,
+      "step": 4304
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4510673546919153,
+      "learning_rate": 4.531668353655948e-06,
+      "loss": 0.8642,
+      "step": 4305
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2502657257701797,
+      "learning_rate": 4.527299150141588e-06,
+      "loss": 0.7621,
+      "step": 4306
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.271358715973476,
+      "learning_rate": 4.52293143754179e-06,
+      "loss": 0.8568,
+      "step": 4307
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.299998817735602,
+      "learning_rate": 4.518565217046446e-06,
+      "loss": 0.7947,
+      "step": 4308
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4167577482552336,
+      "learning_rate": 4.51420048984504e-06,
+      "loss": 0.8466,
+      "step": 4309
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4398217945474072,
+      "learning_rate": 4.50983725712663e-06,
+      "loss": 0.8264,
+      "step": 4310
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7687833904995048,
+      "learning_rate": 4.505475520079889e-06,
+      "loss": 0.3481,
+      "step": 4311
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3900735758358231,
+      "learning_rate": 4.501115279893077e-06,
+      "loss": 0.8601,
+      "step": 4312
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.401890721371135,
+      "learning_rate": 4.4967565377540415e-06,
+      "loss": 0.8584,
+      "step": 4313
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3355339504608454,
+      "learning_rate": 4.492399294850218e-06,
+      "loss": 0.8274,
+      "step": 4314
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2292345351202962,
+      "learning_rate": 4.488043552368649e-06,
+      "loss": 0.7696,
+      "step": 4315
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2980806594846481,
+      "learning_rate": 4.483689311495954e-06,
+      "loss": 0.792,
+      "step": 4316
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4058962227941507,
+      "learning_rate": 4.4793365734183445e-06,
+      "loss": 0.9071,
+      "step": 4317
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.306442520137477,
+      "learning_rate": 4.47498533932163e-06,
+      "loss": 0.7713,
+      "step": 4318
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.282194024578892,
+      "learning_rate": 4.470635610391212e-06,
+      "loss": 0.7357,
+      "step": 4319
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4685406814721074,
+      "learning_rate": 4.466287387812071e-06,
+      "loss": 0.8034,
+      "step": 4320
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5319636023835896,
+      "learning_rate": 4.461940672768779e-06,
+      "loss": 0.8333,
+      "step": 4321
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.379211559144559,
+      "learning_rate": 4.457595466445509e-06,
+      "loss": 0.8597,
+      "step": 4322
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4002588261946511,
+      "learning_rate": 4.453251770026011e-06,
+      "loss": 0.8322,
+      "step": 4323
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3322261544654668,
+      "learning_rate": 4.448909584693626e-06,
+      "loss": 0.8257,
+      "step": 4324
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3120961391034642,
+      "learning_rate": 4.444568911631289e-06,
+      "loss": 0.8341,
+      "step": 4325
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5513946709665167,
+      "learning_rate": 4.4402297520215136e-06,
+      "loss": 0.8406,
+      "step": 4326
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4967581516204458,
+      "learning_rate": 4.435892107046414e-06,
+      "loss": 0.7546,
+      "step": 4327
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5152208290763223,
+      "learning_rate": 4.431555977887679e-06,
+      "loss": 0.9203,
+      "step": 4328
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3350884984260938,
+      "learning_rate": 4.427221365726586e-06,
+      "loss": 0.8077,
+      "step": 4329
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.331257995267486,
+      "learning_rate": 4.422888271744009e-06,
+      "loss": 0.7852,
+      "step": 4330
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8421405159419195,
+      "learning_rate": 4.418556697120408e-06,
+      "loss": 0.3409,
+      "step": 4331
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3465174403666682,
+      "learning_rate": 4.4142266430358085e-06,
+      "loss": 0.8922,
+      "step": 4332
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.342762922766005,
+      "learning_rate": 4.409898110669844e-06,
+      "loss": 0.8451,
+      "step": 4333
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4216569717632133,
+      "learning_rate": 4.40557110120173e-06,
+      "loss": 0.8271,
+      "step": 4334
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2983829872018666,
+      "learning_rate": 4.40124561581026e-06,
+      "loss": 0.8793,
+      "step": 4335
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2717146459867814,
+      "learning_rate": 4.39692165567381e-06,
+      "loss": 0.8251,
+      "step": 4336
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4643920220992683,
+      "learning_rate": 4.392599221970351e-06,
+      "loss": 0.8542,
+      "step": 4337
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2623009940023568,
+      "learning_rate": 4.388278315877441e-06,
+      "loss": 0.807,
+      "step": 4338
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5646845536854703,
+      "learning_rate": 4.3839589385721985e-06,
+      "loss": 0.8487,
+      "step": 4339
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7826344208873324,
+      "learning_rate": 4.379641091231348e-06,
+      "loss": 0.8593,
+      "step": 4340
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.359692297412814,
+      "learning_rate": 4.375324775031194e-06,
+      "loss": 0.77,
+      "step": 4341
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4848570950550797,
+      "learning_rate": 4.3710099911476155e-06,
+      "loss": 0.7718,
+      "step": 4342
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3314098701499462,
+      "learning_rate": 4.3666967407560765e-06,
+      "loss": 0.8079,
+      "step": 4343
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6240926562644307,
+      "learning_rate": 4.362385025031631e-06,
+      "loss": 0.7615,
+      "step": 4344
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2855299700480645,
+      "learning_rate": 4.3580748451489075e-06,
+      "loss": 0.8104,
+      "step": 4345
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.341040577199372,
+      "learning_rate": 4.353766202282113e-06,
+      "loss": 0.7761,
+      "step": 4346
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5248870848845937,
+      "learning_rate": 4.349459097605048e-06,
+      "loss": 0.802,
+      "step": 4347
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8543146268381816,
+      "learning_rate": 4.3451535322910786e-06,
+      "loss": 0.8255,
+      "step": 4348
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8233799083853918,
+      "learning_rate": 4.340849507513168e-06,
+      "loss": 0.3278,
+      "step": 4349
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2563264176359663,
+      "learning_rate": 4.336547024443847e-06,
+      "loss": 0.8228,
+      "step": 4350
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4376301551028559,
+      "learning_rate": 4.332246084255227e-06,
+      "loss": 0.7803,
+      "step": 4351
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7174340065100457,
+      "learning_rate": 4.327946688119006e-06,
+      "loss": 0.7753,
+      "step": 4352
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3727686180883552,
+      "learning_rate": 4.3236488372064656e-06,
+      "loss": 0.8563,
+      "step": 4353
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3912845585130535,
+      "learning_rate": 4.319352532688444e-06,
+      "loss": 0.8571,
+      "step": 4354
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.420853639663344,
+      "learning_rate": 4.31505777573538e-06,
+      "loss": 0.8428,
+      "step": 4355
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3260112354056672,
+      "learning_rate": 4.310764567517288e-06,
+      "loss": 0.7866,
+      "step": 4356
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8249413590469924,
+      "learning_rate": 4.306472909203754e-06,
+      "loss": 0.3208,
+      "step": 4357
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.239801401870534,
+      "learning_rate": 4.302182801963937e-06,
+      "loss": 0.8216,
+      "step": 4358
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6120524165751369,
+      "learning_rate": 4.297894246966586e-06,
+      "loss": 0.759,
+      "step": 4359
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2358976122294756,
+      "learning_rate": 4.29360724538003e-06,
+      "loss": 0.811,
+      "step": 4360
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.360433988134727,
+      "learning_rate": 4.28932179837215e-06,
+      "loss": 0.7975,
+      "step": 4361
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5452096100864234,
+      "learning_rate": 4.2850379071104286e-06,
+      "loss": 0.8033,
+      "step": 4362
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1448321317220125,
+      "learning_rate": 4.280755572761919e-06,
+      "loss": 0.8292,
+      "step": 4363
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2363796351460845,
+      "learning_rate": 4.276474796493243e-06,
+      "loss": 0.8019,
+      "step": 4364
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4440542195220376,
+      "learning_rate": 4.2721955794705985e-06,
+      "loss": 0.9016,
+      "step": 4365
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.584210504907714,
+      "learning_rate": 4.267917922859769e-06,
+      "loss": 0.7839,
+      "step": 4366
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4641191470567296,
+      "learning_rate": 4.263641827826104e-06,
+      "loss": 0.8933,
+      "step": 4367
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3534956107648393,
+      "learning_rate": 4.259367295534524e-06,
+      "loss": 0.8195,
+      "step": 4368
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3247653049878716,
+      "learning_rate": 4.255094327149539e-06,
+      "loss": 0.8095,
+      "step": 4369
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4035554315924368,
+      "learning_rate": 4.250822923835214e-06,
+      "loss": 0.8025,
+      "step": 4370
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7216002650990987,
+      "learning_rate": 4.2465530867552065e-06,
+      "loss": 0.8382,
+      "step": 4371
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5798997943430961,
+      "learning_rate": 4.242284817072732e-06,
+      "loss": 0.8089,
+      "step": 4372
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5969878946250402,
+      "learning_rate": 4.2380181159505815e-06,
+      "loss": 0.8721,
+      "step": 4373
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2793591593365967,
+      "learning_rate": 4.23375298455113e-06,
+      "loss": 0.8012,
+      "step": 4374
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4689644710213003,
+      "learning_rate": 4.22948942403631e-06,
+      "loss": 0.8791,
+      "step": 4375
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3866841902926106,
+      "learning_rate": 4.2252274355676395e-06,
+      "loss": 0.8581,
+      "step": 4376
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4802035009812873,
+      "learning_rate": 4.220967020306194e-06,
+      "loss": 0.8403,
+      "step": 4377
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4783590295574813,
+      "learning_rate": 4.216708179412636e-06,
+      "loss": 0.8812,
+      "step": 4378
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2870513200329659,
+      "learning_rate": 4.212450914047187e-06,
+      "loss": 0.8337,
+      "step": 4379
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.270826924922164,
+      "learning_rate": 4.2081952253696415e-06,
+      "loss": 0.7893,
+      "step": 4380
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5507624634630555,
+      "learning_rate": 4.203941114539367e-06,
+      "loss": 0.8319,
+      "step": 4381
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2441525361405672,
+      "learning_rate": 4.19968858271531e-06,
+      "loss": 0.8358,
+      "step": 4382
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1619526464898216,
+      "learning_rate": 4.195437631055963e-06,
+      "loss": 0.8047,
+      "step": 4383
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.434095952164828,
+      "learning_rate": 4.191188260719408e-06,
+      "loss": 0.782,
+      "step": 4384
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5227068267562682,
+      "learning_rate": 4.186940472863296e-06,
+      "loss": 0.8373,
+      "step": 4385
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2907865150749716,
+      "learning_rate": 4.182694268644837e-06,
+      "loss": 0.8352,
+      "step": 4386
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4436838471854327,
+      "learning_rate": 4.178449649220809e-06,
+      "loss": 0.7934,
+      "step": 4387
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.415141850905036,
+      "learning_rate": 4.174206615747575e-06,
+      "loss": 0.7853,
+      "step": 4388
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5171684604504954,
+      "learning_rate": 4.169965169381045e-06,
+      "loss": 0.8476,
+      "step": 4389
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.393244901387025,
+      "learning_rate": 4.165725311276707e-06,
+      "loss": 0.8279,
+      "step": 4390
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5484051571835435,
+      "learning_rate": 4.161487042589619e-06,
+      "loss": 0.8265,
+      "step": 4391
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5048206819040228,
+      "learning_rate": 4.157250364474398e-06,
+      "loss": 0.7923,
+      "step": 4392
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3434417746758274,
+      "learning_rate": 4.153015278085237e-06,
+      "loss": 0.7748,
+      "step": 4393
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6166106291670272,
+      "learning_rate": 4.148781784575888e-06,
+      "loss": 0.8116,
+      "step": 4394
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3729593152741917,
+      "learning_rate": 4.1445498850996664e-06,
+      "loss": 0.8035,
+      "step": 4395
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.364431645262251,
+      "learning_rate": 4.1403195808094665e-06,
+      "loss": 0.8576,
+      "step": 4396
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3686102739197386,
+      "learning_rate": 4.136090872857732e-06,
+      "loss": 0.8869,
+      "step": 4397
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.231157779023482,
+      "learning_rate": 4.131863762396487e-06,
+      "loss": 0.7824,
+      "step": 4398
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.277919924784873,
+      "learning_rate": 4.127638250577305e-06,
+      "loss": 0.8451,
+      "step": 4399
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.458665641888625,
+      "learning_rate": 4.12341433855134e-06,
+      "loss": 0.7839,
+      "step": 4400
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.277600499547798,
+      "learning_rate": 4.119192027469299e-06,
+      "loss": 0.7722,
+      "step": 4401
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.608756051090134,
+      "learning_rate": 4.114971318481451e-06,
+      "loss": 0.7772,
+      "step": 4402
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5647757563052802,
+      "learning_rate": 4.110752212737641e-06,
+      "loss": 0.7976,
+      "step": 4403
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4142641557962934,
+      "learning_rate": 4.106534711387267e-06,
+      "loss": 0.8794,
+      "step": 4404
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4552195141091302,
+      "learning_rate": 4.102318815579288e-06,
+      "loss": 0.8199,
+      "step": 4405
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4651162520302996,
+      "learning_rate": 4.098104526462235e-06,
+      "loss": 0.7774,
+      "step": 4406
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4399875544218832,
+      "learning_rate": 4.093891845184197e-06,
+      "loss": 0.7903,
+      "step": 4407
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4104297780809647,
+      "learning_rate": 4.0896807728928245e-06,
+      "loss": 0.8198,
+      "step": 4408
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6437664575532518,
+      "learning_rate": 4.085471310735325e-06,
+      "loss": 0.8021,
+      "step": 4409
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3001600527756463,
+      "learning_rate": 4.081263459858479e-06,
+      "loss": 0.7827,
+      "step": 4410
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4810665491336683,
+      "learning_rate": 4.077057221408617e-06,
+      "loss": 0.8264,
+      "step": 4411
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4193027010819241,
+      "learning_rate": 4.072852596531631e-06,
+      "loss": 0.756,
+      "step": 4412
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4906851329260473,
+      "learning_rate": 4.0686495863729854e-06,
+      "loss": 0.8811,
+      "step": 4413
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5196047436609912,
+      "learning_rate": 4.064448192077687e-06,
+      "loss": 0.8586,
+      "step": 4414
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3460065959029994,
+      "learning_rate": 4.060248414790318e-06,
+      "loss": 0.8466,
+      "step": 4415
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5134711741628315,
+      "learning_rate": 4.0560502556550085e-06,
+      "loss": 0.909,
+      "step": 4416
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3049091333601845,
+      "learning_rate": 4.051853715815459e-06,
+      "loss": 0.7882,
+      "step": 4417
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6003473219205104,
+      "learning_rate": 4.0476587964149185e-06,
+      "loss": 0.7748,
+      "step": 4418
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.425819195281636,
+      "learning_rate": 4.0434654985961955e-06,
+      "loss": 0.8293,
+      "step": 4419
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4124346493007718,
+      "learning_rate": 4.039273823501667e-06,
+      "loss": 0.8452,
+      "step": 4420
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3024318707018312,
+      "learning_rate": 4.035083772273254e-06,
+      "loss": 0.8461,
+      "step": 4421
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4441841102431878,
+      "learning_rate": 4.030895346052449e-06,
+      "loss": 0.8181,
+      "step": 4422
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5049591688703727,
+      "learning_rate": 4.02670854598029e-06,
+      "loss": 0.8744,
+      "step": 4423
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6443665194292727,
+      "learning_rate": 4.022523373197376e-06,
+      "loss": 0.7989,
+      "step": 4424
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4515254647557412,
+      "learning_rate": 4.018339828843868e-06,
+      "loss": 0.8733,
+      "step": 4425
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3408149078715867,
+      "learning_rate": 4.014157914059475e-06,
+      "loss": 0.8726,
+      "step": 4426
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1709100827679175,
+      "learning_rate": 4.009977629983464e-06,
+      "loss": 0.9114,
+      "step": 4427
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2973978786715272,
+      "learning_rate": 4.005798977754664e-06,
+      "loss": 0.8347,
+      "step": 4428
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5730852555536845,
+      "learning_rate": 4.001621958511456e-06,
+      "loss": 0.8051,
+      "step": 4429
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.377665937176999,
+      "learning_rate": 3.9974465733917725e-06,
+      "loss": 0.8662,
+      "step": 4430
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4935435142912283,
+      "learning_rate": 3.993272823533101e-06,
+      "loss": 0.8439,
+      "step": 4431
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4429110332712358,
+      "learning_rate": 3.989100710072491e-06,
+      "loss": 0.8767,
+      "step": 4432
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3359158040816939,
+      "learning_rate": 3.98493023414654e-06,
+      "loss": 0.8794,
+      "step": 4433
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4192622894648228,
+      "learning_rate": 3.980761396891396e-06,
+      "loss": 0.8591,
+      "step": 4434
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2474980363873371,
+      "learning_rate": 3.976594199442768e-06,
+      "loss": 0.8068,
+      "step": 4435
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4981168378733734,
+      "learning_rate": 3.972428642935921e-06,
+      "loss": 0.8296,
+      "step": 4436
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.521495261416584,
+      "learning_rate": 3.968264728505662e-06,
+      "loss": 0.7668,
+      "step": 4437
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8689031447958926,
+      "learning_rate": 3.964102457286353e-06,
+      "loss": 0.353,
+      "step": 4438
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4503065130839268,
+      "learning_rate": 3.959941830411918e-06,
+      "loss": 0.7483,
+      "step": 4439
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.717268342539693,
+      "learning_rate": 3.955782849015825e-06,
+      "loss": 0.8872,
+      "step": 4440
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.294991918901895,
+      "learning_rate": 3.95162551423109e-06,
+      "loss": 0.9289,
+      "step": 4441
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3215312489104538,
+      "learning_rate": 3.9474698271902925e-06,
+      "loss": 0.8549,
+      "step": 4442
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4882718133502066,
+      "learning_rate": 3.94331578902555e-06,
+      "loss": 0.8303,
+      "step": 4443
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3237914638052315,
+      "learning_rate": 3.939163400868543e-06,
+      "loss": 0.7986,
+      "step": 4444
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1931634920124807,
+      "learning_rate": 3.935012663850493e-06,
+      "loss": 0.8229,
+      "step": 4445
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4098476345628528,
+      "learning_rate": 3.930863579102173e-06,
+      "loss": 0.7713,
+      "step": 4446
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8840831706959522,
+      "learning_rate": 3.9267161477539155e-06,
+      "loss": 0.8077,
+      "step": 4447
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.398224938343925,
+      "learning_rate": 3.922570370935588e-06,
+      "loss": 0.8577,
+      "step": 4448
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5316889125321147,
+      "learning_rate": 3.918426249776614e-06,
+      "loss": 0.7675,
+      "step": 4449
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3503752169614192,
+      "learning_rate": 3.91428378540597e-06,
+      "loss": 0.8022,
+      "step": 4450
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4521086257012774,
+      "learning_rate": 3.910142978952183e-06,
+      "loss": 0.9199,
+      "step": 4451
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4825339830926825,
+      "learning_rate": 3.906003831543309e-06,
+      "loss": 0.8346,
+      "step": 4452
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7926786554502268,
+      "learning_rate": 3.901866344306975e-06,
+      "loss": 0.3066,
+      "step": 4453
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4031918187469365,
+      "learning_rate": 3.8977305183703464e-06,
+      "loss": 0.8465,
+      "step": 4454
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.303496729651231,
+      "learning_rate": 3.893596354860135e-06,
+      "loss": 0.7074,
+      "step": 4455
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3881690218788287,
+      "learning_rate": 3.889463854902598e-06,
+      "loss": 0.8552,
+      "step": 4456
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5077475931596371,
+      "learning_rate": 3.885333019623544e-06,
+      "loss": 0.8626,
+      "step": 4457
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2328556490642726,
+      "learning_rate": 3.88120385014833e-06,
+      "loss": 0.851,
+      "step": 4458
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4566852373496852,
+      "learning_rate": 3.8770763476018546e-06,
+      "loss": 0.9238,
+      "step": 4459
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1874441428599796,
+      "learning_rate": 3.872950513108558e-06,
+      "loss": 0.7968,
+      "step": 4460
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.499331230846118,
+      "learning_rate": 3.868826347792437e-06,
+      "loss": 0.7806,
+      "step": 4461
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2641836542131464,
+      "learning_rate": 3.864703852777026e-06,
+      "loss": 0.8337,
+      "step": 4462
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4957867001018097,
+      "learning_rate": 3.860583029185403e-06,
+      "loss": 0.7914,
+      "step": 4463
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4038091885109347,
+      "learning_rate": 3.8564638781402e-06,
+      "loss": 0.8702,
+      "step": 4464
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2688078506557041,
+      "learning_rate": 3.85234640076358e-06,
+      "loss": 0.8401,
+      "step": 4465
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3662071405212124,
+      "learning_rate": 3.848230598177266e-06,
+      "loss": 0.8727,
+      "step": 4466
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3091243479518637,
+      "learning_rate": 3.844116471502511e-06,
+      "loss": 0.8693,
+      "step": 4467
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3986915075986504,
+      "learning_rate": 3.840004021860113e-06,
+      "loss": 0.8571,
+      "step": 4468
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6132602141858485,
+      "learning_rate": 3.835893250370426e-06,
+      "loss": 0.8687,
+      "step": 4469
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3911932630700203,
+      "learning_rate": 3.831784158153331e-06,
+      "loss": 0.8043,
+      "step": 4470
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.588859496540571,
+      "learning_rate": 3.827676746328256e-06,
+      "loss": 0.7905,
+      "step": 4471
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8065331349743698,
+      "learning_rate": 3.823571016014176e-06,
+      "loss": 0.7978,
+      "step": 4472
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3069954555703998,
+      "learning_rate": 3.819466968329613e-06,
+      "loss": 0.8102,
+      "step": 4473
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4073311775442594,
+      "learning_rate": 3.815364604392607e-06,
+      "loss": 0.8363,
+      "step": 4474
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4183997492295515,
+      "learning_rate": 3.811263925320765e-06,
+      "loss": 0.7498,
+      "step": 4475
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.381584869194829,
+      "learning_rate": 3.8071649322312256e-06,
+      "loss": 0.8316,
+      "step": 4476
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3462547575136914,
+      "learning_rate": 3.803067626240665e-06,
+      "loss": 0.7858,
+      "step": 4477
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4051069216657956,
+      "learning_rate": 3.7989720084653003e-06,
+      "loss": 0.8219,
+      "step": 4478
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.427378322741663,
+      "learning_rate": 3.7948780800208916e-06,
+      "loss": 0.7915,
+      "step": 4479
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.494173768243142,
+      "learning_rate": 3.790785842022746e-06,
+      "loss": 0.7498,
+      "step": 4480
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3285230152403515,
+      "learning_rate": 3.7866952955856895e-06,
+      "loss": 0.8455,
+      "step": 4481
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3673930752694237,
+      "learning_rate": 3.7826064418241037e-06,
+      "loss": 0.9068,
+      "step": 4482
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2136249684188323,
+      "learning_rate": 3.7785192818519113e-06,
+      "loss": 0.8506,
+      "step": 4483
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5244707577198364,
+      "learning_rate": 3.774433816782561e-06,
+      "loss": 0.8534,
+      "step": 4484
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5745068705303786,
+      "learning_rate": 3.7703500477290456e-06,
+      "loss": 0.7814,
+      "step": 4485
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.592223918711669,
+      "learning_rate": 3.7662679758039023e-06,
+      "loss": 0.9493,
+      "step": 4486
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5047488657350458,
+      "learning_rate": 3.762187602119192e-06,
+      "loss": 0.9013,
+      "step": 4487
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.422928141075951,
+      "learning_rate": 3.758108927786528e-06,
+      "loss": 0.8844,
+      "step": 4488
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4594127608280405,
+      "learning_rate": 3.7540319539170522e-06,
+      "loss": 0.8185,
+      "step": 4489
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.283019489636463,
+      "learning_rate": 3.7499566816214384e-06,
+      "loss": 0.8206,
+      "step": 4490
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4077586644524112,
+      "learning_rate": 3.7458831120099126e-06,
+      "loss": 0.8837,
+      "step": 4491
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4688753948936477,
+      "learning_rate": 3.741811246192223e-06,
+      "loss": 0.8671,
+      "step": 4492
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3062604438755507,
+      "learning_rate": 3.737741085277653e-06,
+      "loss": 0.832,
+      "step": 4493
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3191198677494795,
+      "learning_rate": 3.7336726303750327e-06,
+      "loss": 0.8666,
+      "step": 4494
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3160726229886406,
+      "learning_rate": 3.729605882592724e-06,
+      "loss": 0.8172,
+      "step": 4495
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3893386799854455,
+      "learning_rate": 3.7255408430386164e-06,
+      "loss": 0.9093,
+      "step": 4496
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6660702026747376,
+      "learning_rate": 3.7214775128201363e-06,
+      "loss": 0.7978,
+      "step": 4497
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8523426572511208,
+      "learning_rate": 3.717415893044254e-06,
+      "loss": 0.3175,
+      "step": 4498
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.478682894674482,
+      "learning_rate": 3.713355984817463e-06,
+      "loss": 0.756,
+      "step": 4499
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3604547005555878,
+      "learning_rate": 3.7092977892457905e-06,
+      "loss": 0.7553,
+      "step": 4500
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3451571055792968,
+      "learning_rate": 3.7052413074348038e-06,
+      "loss": 0.8328,
+      "step": 4501
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3022451064624914,
+      "learning_rate": 3.7011865404896085e-06,
+      "loss": 0.8438,
+      "step": 4502
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8326903282664518,
+      "learning_rate": 3.6971334895148202e-06,
+      "loss": 0.323,
+      "step": 4503
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.192988267359808,
+      "learning_rate": 3.6930821556146092e-06,
+      "loss": 0.8123,
+      "step": 4504
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3951098419854824,
+      "learning_rate": 3.689032539892673e-06,
+      "loss": 0.8104,
+      "step": 4505
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4705445329487081,
+      "learning_rate": 3.684984643452236e-06,
+      "loss": 0.8273,
+      "step": 4506
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8865627669422081,
+      "learning_rate": 3.6809384673960545e-06,
+      "loss": 0.3551,
+      "step": 4507
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.314786424509152,
+      "learning_rate": 3.676894012826422e-06,
+      "loss": 0.8539,
+      "step": 4508
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4987192037801202,
+      "learning_rate": 3.6728512808451554e-06,
+      "loss": 0.8732,
+      "step": 4509
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4703414023818968,
+      "learning_rate": 3.668810272553612e-06,
+      "loss": 0.8551,
+      "step": 4510
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4844069358251264,
+      "learning_rate": 3.6647709890526708e-06,
+      "loss": 0.7939,
+      "step": 4511
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.470149797985317,
+      "learning_rate": 3.66073343144274e-06,
+      "loss": 0.8611,
+      "step": 4512
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6220625981119294,
+      "learning_rate": 3.65669760082377e-06,
+      "loss": 0.8424,
+      "step": 4513
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4752904695871116,
+      "learning_rate": 3.6526634982952225e-06,
+      "loss": 0.774,
+      "step": 4514
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4321083603226317,
+      "learning_rate": 3.648631124956108e-06,
+      "loss": 0.8124,
+      "step": 4515
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8698157084910847,
+      "learning_rate": 3.6446004819049473e-06,
+      "loss": 0.9217,
+      "step": 4516
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5719660360999446,
+      "learning_rate": 3.640571570239807e-06,
+      "loss": 0.7643,
+      "step": 4517
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4693490484540859,
+      "learning_rate": 3.63654439105827e-06,
+      "loss": 0.7788,
+      "step": 4518
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3287259362184631,
+      "learning_rate": 3.632518945457446e-06,
+      "loss": 0.8596,
+      "step": 4519
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0685690862662183,
+      "learning_rate": 3.6284952345339864e-06,
+      "loss": 0.8181,
+      "step": 4520
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3796398572590949,
+      "learning_rate": 3.624473259384056e-06,
+      "loss": 0.7646,
+      "step": 4521
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5895217515535458,
+      "learning_rate": 3.6204530211033482e-06,
+      "loss": 0.8225,
+      "step": 4522
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2936802992151082,
+      "learning_rate": 3.616434520787091e-06,
+      "loss": 0.7523,
+      "step": 4523
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3218001471850482,
+      "learning_rate": 3.6124177595300415e-06,
+      "loss": 0.8659,
+      "step": 4524
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.205757299044214,
+      "learning_rate": 3.608402738426462e-06,
+      "loss": 0.7237,
+      "step": 4525
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.708876816225869,
+      "learning_rate": 3.6043894585701623e-06,
+      "loss": 0.836,
+      "step": 4526
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5839417936337774,
+      "learning_rate": 3.6003779210544733e-06,
+      "loss": 0.8183,
+      "step": 4527
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8846497180840347,
+      "learning_rate": 3.596368126972245e-06,
+      "loss": 0.3329,
+      "step": 4528
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4422604083409127,
+      "learning_rate": 3.592360077415853e-06,
+      "loss": 0.7988,
+      "step": 4529
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.472374270583407,
+      "learning_rate": 3.588353773477208e-06,
+      "loss": 0.8148,
+      "step": 4530
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4220890575714356,
+      "learning_rate": 3.5843492162477312e-06,
+      "loss": 0.8144,
+      "step": 4531
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3744379426931526,
+      "learning_rate": 3.5803464068183734e-06,
+      "loss": 0.8854,
+      "step": 4532
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.161171529106772,
+      "learning_rate": 3.576345346279614e-06,
+      "loss": 0.8274,
+      "step": 4533
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1475818034696188,
+      "learning_rate": 3.5723460357214547e-06,
+      "loss": 0.8043,
+      "step": 4534
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3847066230471472,
+      "learning_rate": 3.568348476233414e-06,
+      "loss": 0.8396,
+      "step": 4535
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8590872022588721,
+      "learning_rate": 3.564352668904535e-06,
+      "loss": 0.3347,
+      "step": 4536
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.335976897043588,
+      "learning_rate": 3.5603586148233917e-06,
+      "loss": 0.8158,
+      "step": 4537
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1876977925211192,
+      "learning_rate": 3.556366315078068e-06,
+      "loss": 0.8106,
+      "step": 4538
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7558165132342765,
+      "learning_rate": 3.5523757707561836e-06,
+      "loss": 0.8079,
+      "step": 4539
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.294637040993705,
+      "learning_rate": 3.548386982944868e-06,
+      "loss": 0.7753,
+      "step": 4540
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3278888502965889,
+      "learning_rate": 3.544399952730776e-06,
+      "loss": 0.8824,
+      "step": 4541
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3314996033151159,
+      "learning_rate": 3.5404146812000893e-06,
+      "loss": 0.8609,
+      "step": 4542
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2715993813099422,
+      "learning_rate": 3.536431169438502e-06,
+      "loss": 0.7481,
+      "step": 4543
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3580886671360501,
+      "learning_rate": 3.5324494185312317e-06,
+      "loss": 0.8293,
+      "step": 4544
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7015666525503723,
+      "learning_rate": 3.5284694295630183e-06,
+      "loss": 0.8196,
+      "step": 4545
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3454876887831788,
+      "learning_rate": 3.5244912036181276e-06,
+      "loss": 0.7904,
+      "step": 4546
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5431661256265312,
+      "learning_rate": 3.5205147417803253e-06,
+      "loss": 0.7576,
+      "step": 4547
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4191286988496183,
+      "learning_rate": 3.5165400451329147e-06,
+      "loss": 0.7237,
+      "step": 4548
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2165767235546667,
+      "learning_rate": 3.512567114758717e-06,
+      "loss": 0.8291,
+      "step": 4549
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3471825644852433,
+      "learning_rate": 3.5085959517400645e-06,
+      "loss": 0.8174,
+      "step": 4550
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6173246205749865,
+      "learning_rate": 3.504626557158808e-06,
+      "loss": 0.9089,
+      "step": 4551
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.414567089414755,
+      "learning_rate": 3.500658932096327e-06,
+      "loss": 0.8711,
+      "step": 4552
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3942766464714509,
+      "learning_rate": 3.4966930776335083e-06,
+      "loss": 0.7604,
+      "step": 4553
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5862658787876505,
+      "learning_rate": 3.492728994850756e-06,
+      "loss": 0.8437,
+      "step": 4554
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2819101272914573,
+      "learning_rate": 3.488766684828e-06,
+      "loss": 0.6886,
+      "step": 4555
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3757993091208154,
+      "learning_rate": 3.4848061486446848e-06,
+      "loss": 0.8845,
+      "step": 4556
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3169207925692148,
+      "learning_rate": 3.4808473873797675e-06,
+      "loss": 0.8157,
+      "step": 4557
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4490506770052904,
+      "learning_rate": 3.476890402111721e-06,
+      "loss": 0.8317,
+      "step": 4558
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4116023229659287,
+      "learning_rate": 3.472935193918542e-06,
+      "loss": 0.8157,
+      "step": 4559
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3167486368112151,
+      "learning_rate": 3.4689817638777355e-06,
+      "loss": 0.8077,
+      "step": 4560
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4326459463102414,
+      "learning_rate": 3.4650301130663224e-06,
+      "loss": 0.8216,
+      "step": 4561
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3305761221434707,
+      "learning_rate": 3.461080242560847e-06,
+      "loss": 0.7662,
+      "step": 4562
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3511201907912784,
+      "learning_rate": 3.4571321534373557e-06,
+      "loss": 0.8214,
+      "step": 4563
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1735433298064246,
+      "learning_rate": 3.453185846771425e-06,
+      "loss": 0.8022,
+      "step": 4564
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3616895332747592,
+      "learning_rate": 3.449241323638135e-06,
+      "loss": 0.8511,
+      "step": 4565
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9705964966280717,
+      "learning_rate": 3.445298585112077e-06,
+      "loss": 0.3074,
+      "step": 4566
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3728402129417934,
+      "learning_rate": 3.441357632267365e-06,
+      "loss": 0.7986,
+      "step": 4567
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5934863065207023,
+      "learning_rate": 3.437418466177631e-06,
+      "loss": 0.8624,
+      "step": 4568
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0172151807894863,
+      "learning_rate": 3.4334810879159987e-06,
+      "loss": 0.3119,
+      "step": 4569
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2824732818374256,
+      "learning_rate": 3.429545498555126e-06,
+      "loss": 0.8599,
+      "step": 4570
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8087818000797187,
+      "learning_rate": 3.4256116991671773e-06,
+      "loss": 0.3331,
+      "step": 4571
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4284111016003391,
+      "learning_rate": 3.421679690823827e-06,
+      "loss": 0.9121,
+      "step": 4572
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.394947740458682,
+      "learning_rate": 3.417749474596257e-06,
+      "loss": 0.8109,
+      "step": 4573
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4927535044631843,
+      "learning_rate": 3.4138210515551717e-06,
+      "loss": 0.8471,
+      "step": 4574
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4700875374449653,
+      "learning_rate": 3.4098944227707873e-06,
+      "loss": 0.8439,
+      "step": 4575
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3124245539620338,
+      "learning_rate": 3.4059695893128133e-06,
+      "loss": 0.7606,
+      "step": 4576
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.046113454040001,
+      "learning_rate": 3.4020465522504876e-06,
+      "loss": 0.8573,
+      "step": 4577
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4425325593587695,
+      "learning_rate": 3.3981253126525593e-06,
+      "loss": 0.8124,
+      "step": 4578
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3061538938709223,
+      "learning_rate": 3.394205871587277e-06,
+      "loss": 0.9021,
+      "step": 4579
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3968656672948951,
+      "learning_rate": 3.3902882301224016e-06,
+      "loss": 0.7762,
+      "step": 4580
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4402304327269095,
+      "learning_rate": 3.386372389325213e-06,
+      "loss": 0.7976,
+      "step": 4581
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2529998998971583,
+      "learning_rate": 3.382458350262493e-06,
+      "loss": 0.8118,
+      "step": 4582
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.267057599225208,
+      "learning_rate": 3.378546114000527e-06,
+      "loss": 0.7101,
+      "step": 4583
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.283348846725599,
+      "learning_rate": 3.374635681605125e-06,
+      "loss": 0.7633,
+      "step": 4584
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3763833026881758,
+      "learning_rate": 3.3707270541415895e-06,
+      "loss": 0.8306,
+      "step": 4585
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3870988351802296,
+      "learning_rate": 3.3668202326747433e-06,
+      "loss": 0.7298,
+      "step": 4586
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4636531429043995,
+      "learning_rate": 3.3629152182689117e-06,
+      "loss": 0.8473,
+      "step": 4587
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2662382020093548,
+      "learning_rate": 3.3590120119879233e-06,
+      "loss": 0.8124,
+      "step": 4588
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3072294567765188,
+      "learning_rate": 3.3551106148951262e-06,
+      "loss": 0.8475,
+      "step": 4589
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.442508742375798,
+      "learning_rate": 3.351211028053365e-06,
+      "loss": 0.9055,
+      "step": 4590
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2960453014907136,
+      "learning_rate": 3.3473132525249918e-06,
+      "loss": 0.899,
+      "step": 4591
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3906475300383587,
+      "learning_rate": 3.3434172893718707e-06,
+      "loss": 0.8027,
+      "step": 4592
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5760171429057548,
+      "learning_rate": 3.339523139655373e-06,
+      "loss": 0.9177,
+      "step": 4593
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.269209711731106,
+      "learning_rate": 3.3356308044363683e-06,
+      "loss": 0.7466,
+      "step": 4594
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2789226031816263,
+      "learning_rate": 3.3317402847752344e-06,
+      "loss": 0.7593,
+      "step": 4595
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3426214513594168,
+      "learning_rate": 3.327851581731859e-06,
+      "loss": 0.8698,
+      "step": 4596
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4033046215828673,
+      "learning_rate": 3.323964696365638e-06,
+      "loss": 0.835,
+      "step": 4597
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7531404692601436,
+      "learning_rate": 3.320079629735452e-06,
+      "loss": 0.8819,
+      "step": 4598
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2343122264135091,
+      "learning_rate": 3.316196382899709e-06,
+      "loss": 0.8291,
+      "step": 4599
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6588369653195205,
+      "learning_rate": 3.312314956916315e-06,
+      "loss": 0.8245,
+      "step": 4600
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3280495776470922,
+      "learning_rate": 3.3084353528426727e-06,
+      "loss": 0.8668,
+      "step": 4601
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1922568419490234,
+      "learning_rate": 3.3045575717356926e-06,
+      "loss": 0.8691,
+      "step": 4602
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3327769509244611,
+      "learning_rate": 3.3006816146517927e-06,
+      "loss": 0.8104,
+      "step": 4603
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5039261123039187,
+      "learning_rate": 3.29680748264689e-06,
+      "loss": 0.8354,
+      "step": 4604
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4390163328280725,
+      "learning_rate": 3.2929351767763997e-06,
+      "loss": 0.7603,
+      "step": 4605
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3093446171542662,
+      "learning_rate": 3.289064698095251e-06,
+      "loss": 0.8822,
+      "step": 4606
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2121243759306402,
+      "learning_rate": 3.2851960476578647e-06,
+      "loss": 0.8336,
+      "step": 4607
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3479061060948208,
+      "learning_rate": 3.281329226518173e-06,
+      "loss": 0.8563,
+      "step": 4608
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.319931036535498,
+      "learning_rate": 3.2774642357296006e-06,
+      "loss": 0.9039,
+      "step": 4609
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5990038192494218,
+      "learning_rate": 3.2736010763450744e-06,
+      "loss": 0.8383,
+      "step": 4610
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2487795250856075,
+      "learning_rate": 3.2697397494170336e-06,
+      "loss": 0.8858,
+      "step": 4611
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3727514626108848,
+      "learning_rate": 3.2658802559974046e-06,
+      "loss": 0.7983,
+      "step": 4612
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6093886812332248,
+      "learning_rate": 3.2620225971376187e-06,
+      "loss": 0.7997,
+      "step": 4613
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4311652526221224,
+      "learning_rate": 3.25816677388861e-06,
+      "loss": 0.8374,
+      "step": 4614
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.480377833611945,
+      "learning_rate": 3.2543127873008164e-06,
+      "loss": 0.8615,
+      "step": 4615
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7145031751574502,
+      "learning_rate": 3.2504606384241642e-06,
+      "loss": 0.8197,
+      "step": 4616
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4812887752815658,
+      "learning_rate": 3.246610328308083e-06,
+      "loss": 0.8922,
+      "step": 4617
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3992676335543859,
+      "learning_rate": 3.2427618580015107e-06,
+      "loss": 0.8049,
+      "step": 4618
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3787676502468011,
+      "learning_rate": 3.2389152285528726e-06,
+      "loss": 0.8205,
+      "step": 4619
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.251255421970912,
+      "learning_rate": 3.235070441010092e-06,
+      "loss": 0.7366,
+      "step": 4620
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4918743929733893,
+      "learning_rate": 3.2312274964206013e-06,
+      "loss": 0.8301,
+      "step": 4621
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4498453366587167,
+      "learning_rate": 3.2273863958313257e-06,
+      "loss": 0.8271,
+      "step": 4622
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4591424695500785,
+      "learning_rate": 3.2235471402886833e-06,
+      "loss": 0.778,
+      "step": 4623
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5772348191311392,
+      "learning_rate": 3.2197097308385916e-06,
+      "loss": 0.7478,
+      "step": 4624
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2070137646868462,
+      "learning_rate": 3.2158741685264715e-06,
+      "loss": 0.8594,
+      "step": 4625
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3933171559042201,
+      "learning_rate": 3.212040454397234e-06,
+      "loss": 0.7904,
+      "step": 4626
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1952114779349141,
+      "learning_rate": 3.208208589495284e-06,
+      "loss": 0.8301,
+      "step": 4627
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9752023943417202,
+      "learning_rate": 3.204378574864535e-06,
+      "loss": 0.3427,
+      "step": 4628
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.505884330508093,
+      "learning_rate": 3.200550411548381e-06,
+      "loss": 0.8588,
+      "step": 4629
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4492764493607064,
+      "learning_rate": 3.1967241005897264e-06,
+      "loss": 0.7862,
+      "step": 4630
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3536520320859997,
+      "learning_rate": 3.19289964303096e-06,
+      "loss": 0.87,
+      "step": 4631
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.709930752978526,
+      "learning_rate": 3.189077039913967e-06,
+      "loss": 0.8146,
+      "step": 4632
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2768763524056255,
+      "learning_rate": 3.1852562922801346e-06,
+      "loss": 0.8036,
+      "step": 4633
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.555682858690599,
+      "learning_rate": 3.181437401170335e-06,
+      "loss": 0.8123,
+      "step": 4634
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7700029491121926,
+      "learning_rate": 3.177620367624946e-06,
+      "loss": 0.3377,
+      "step": 4635
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6109098874231431,
+      "learning_rate": 3.1738051926838243e-06,
+      "loss": 0.7974,
+      "step": 4636
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2557265543669884,
+      "learning_rate": 3.169991877386338e-06,
+      "loss": 0.8346,
+      "step": 4637
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4542267529644586,
+      "learning_rate": 3.1661804227713334e-06,
+      "loss": 0.7901,
+      "step": 4638
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6090037064572964,
+      "learning_rate": 3.162370829877154e-06,
+      "loss": 0.91,
+      "step": 4639
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.274934944311655,
+      "learning_rate": 3.158563099741644e-06,
+      "loss": 0.8863,
+      "step": 4640
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3647634460765492,
+      "learning_rate": 3.1547572334021315e-06,
+      "loss": 0.865,
+      "step": 4641
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2058996602858776,
+      "learning_rate": 3.150953231895435e-06,
+      "loss": 0.785,
+      "step": 4642
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4880652940567194,
+      "learning_rate": 3.1471510962578743e-06,
+      "loss": 0.7924,
+      "step": 4643
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8404726191235011,
+      "learning_rate": 3.143350827525257e-06,
+      "loss": 0.3487,
+      "step": 4644
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5028895663383457,
+      "learning_rate": 3.139552426732879e-06,
+      "loss": 0.8436,
+      "step": 4645
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.326259575024106,
+      "learning_rate": 3.1357558949155266e-06,
+      "loss": 0.8326,
+      "step": 4646
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6337339029389253,
+      "learning_rate": 3.1319612331074856e-06,
+      "loss": 0.841,
+      "step": 4647
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8376977060434554,
+      "learning_rate": 3.128168442342523e-06,
+      "loss": 0.3272,
+      "step": 4648
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2234325713177738,
+      "learning_rate": 3.1243775236538963e-06,
+      "loss": 0.8344,
+      "step": 4649
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4849761993966566,
+      "learning_rate": 3.1205884780743633e-06,
+      "loss": 0.7679,
+      "step": 4650
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.409717975148358,
+      "learning_rate": 3.116801306636158e-06,
+      "loss": 0.8178,
+      "step": 4651
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6829071780253515,
+      "learning_rate": 3.113016010371016e-06,
+      "loss": 0.8338,
+      "step": 4652
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3574278330500635,
+      "learning_rate": 3.1092325903101518e-06,
+      "loss": 0.8474,
+      "step": 4653
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7394067196124647,
+      "learning_rate": 3.105451047484277e-06,
+      "loss": 0.8034,
+      "step": 4654
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.550439496662501,
+      "learning_rate": 3.1016713829235866e-06,
+      "loss": 0.8021,
+      "step": 4655
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8544490547193098,
+      "learning_rate": 3.0978935976577617e-06,
+      "loss": 0.3281,
+      "step": 4656
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4015045524023768,
+      "learning_rate": 3.0941176927159812e-06,
+      "loss": 0.8779,
+      "step": 4657
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.248961740182372,
+      "learning_rate": 3.0903436691269006e-06,
+      "loss": 0.7945,
+      "step": 4658
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6171210132513463,
+      "learning_rate": 3.0865715279186724e-06,
+      "loss": 0.7699,
+      "step": 4659
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3981970844782021,
+      "learning_rate": 3.0828012701189316e-06,
+      "loss": 0.8269,
+      "step": 4660
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6148611015657328,
+      "learning_rate": 3.079032896754793e-06,
+      "loss": 0.8701,
+      "step": 4661
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.585843536806766,
+      "learning_rate": 3.075266408852876e-06,
+      "loss": 0.8013,
+      "step": 4662
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3893948971771075,
+      "learning_rate": 3.07150180743927e-06,
+      "loss": 0.7804,
+      "step": 4663
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6279533993207522,
+      "learning_rate": 3.0677390935395533e-06,
+      "loss": 0.8204,
+      "step": 4664
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6195709964541622,
+      "learning_rate": 3.063978268178798e-06,
+      "loss": 0.8633,
+      "step": 4665
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3495268361733943,
+      "learning_rate": 3.0602193323815563e-06,
+      "loss": 0.7675,
+      "step": 4666
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3265433402188649,
+      "learning_rate": 3.056462287171865e-06,
+      "loss": 0.7991,
+      "step": 4667
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.538274529414804,
+      "learning_rate": 3.052707133573244e-06,
+      "loss": 0.8395,
+      "step": 4668
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.503590284014218,
+      "learning_rate": 3.0489538726087053e-06,
+      "loss": 0.8402,
+      "step": 4669
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.337043396494602,
+      "learning_rate": 3.0452025053007396e-06,
+      "loss": 0.8318,
+      "step": 4670
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3908823420129273,
+      "learning_rate": 3.0414530326713176e-06,
+      "loss": 0.816,
+      "step": 4671
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5278305451144205,
+      "learning_rate": 3.0377054557419028e-06,
+      "loss": 0.8631,
+      "step": 4672
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8858852060718336,
+      "learning_rate": 3.0339597755334414e-06,
+      "loss": 0.3367,
+      "step": 4673
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5073235835848535,
+      "learning_rate": 3.0302159930663575e-06,
+      "loss": 0.8395,
+      "step": 4674
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2255792625215078,
+      "learning_rate": 3.026474109360559e-06,
+      "loss": 0.8122,
+      "step": 4675
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.395127827997497,
+      "learning_rate": 3.0227341254354405e-06,
+      "loss": 0.8483,
+      "step": 4676
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3967492405927933,
+      "learning_rate": 3.0189960423098765e-06,
+      "loss": 0.8411,
+      "step": 4677
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8058305842031075,
+      "learning_rate": 3.015259861002219e-06,
+      "loss": 0.8427,
+      "step": 4678
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5646683107989443,
+      "learning_rate": 3.0115255825303148e-06,
+      "loss": 0.8319,
+      "step": 4679
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2708654072709185,
+      "learning_rate": 3.0077932079114768e-06,
+      "loss": 0.8519,
+      "step": 4680
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.579503433481849,
+      "learning_rate": 3.004062738162514e-06,
+      "loss": 0.7811,
+      "step": 4681
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6116411714193701,
+      "learning_rate": 3.000334174299705e-06,
+      "loss": 0.8498,
+      "step": 4682
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.252487935769062,
+      "learning_rate": 2.996607517338811e-06,
+      "loss": 0.7534,
+      "step": 4683
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.284323911219279,
+      "learning_rate": 2.9928827682950825e-06,
+      "loss": 0.8281,
+      "step": 4684
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5534638017082685,
+      "learning_rate": 2.9891599281832395e-06,
+      "loss": 0.85,
+      "step": 4685
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6245453157152632,
+      "learning_rate": 2.9854389980174837e-06,
+      "loss": 0.8166,
+      "step": 4686
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3335684069117697,
+      "learning_rate": 2.9817199788115025e-06,
+      "loss": 0.8264,
+      "step": 4687
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3994176475920879,
+      "learning_rate": 2.978002871578466e-06,
+      "loss": 0.8092,
+      "step": 4688
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5112244909411698,
+      "learning_rate": 2.9742876773310037e-06,
+      "loss": 0.8667,
+      "step": 4689
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.343841591197542,
+      "learning_rate": 2.9705743970812416e-06,
+      "loss": 0.8188,
+      "step": 4690
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7326316432330489,
+      "learning_rate": 2.9668630318407854e-06,
+      "loss": 0.8499,
+      "step": 4691
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5315243319565888,
+      "learning_rate": 2.963153582620708e-06,
+      "loss": 0.7917,
+      "step": 4692
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2618766492473392,
+      "learning_rate": 2.9594460504315637e-06,
+      "loss": 0.8141,
+      "step": 4693
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6655788340978797,
+      "learning_rate": 2.9557404362833898e-06,
+      "loss": 0.8314,
+      "step": 4694
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1972884025087276,
+      "learning_rate": 2.9520367411856997e-06,
+      "loss": 0.7529,
+      "step": 4695
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6391126667287583,
+      "learning_rate": 2.9483349661474792e-06,
+      "loss": 0.8067,
+      "step": 4696
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2406837178785362,
+      "learning_rate": 2.944635112177192e-06,
+      "loss": 0.8418,
+      "step": 4697
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.192417237615125,
+      "learning_rate": 2.9409371802827856e-06,
+      "loss": 0.8405,
+      "step": 4698
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4562592851973102,
+      "learning_rate": 2.937241171471674e-06,
+      "loss": 0.8118,
+      "step": 4699
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3906665270744865,
+      "learning_rate": 2.933547086750752e-06,
+      "loss": 0.7273,
+      "step": 4700
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.313457905803475,
+      "learning_rate": 2.9298549271263944e-06,
+      "loss": 0.8816,
+      "step": 4701
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9222697651164837,
+      "learning_rate": 2.9261646936044408e-06,
+      "loss": 0.8614,
+      "step": 4702
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.419822423094064,
+      "learning_rate": 2.9224763871902186e-06,
+      "loss": 0.7413,
+      "step": 4703
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.24121525941542,
+      "learning_rate": 2.9187900088885224e-06,
+      "loss": 0.8137,
+      "step": 4704
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.326930666804502,
+      "learning_rate": 2.9151055597036195e-06,
+      "loss": 0.7895,
+      "step": 4705
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5322445177906812,
+      "learning_rate": 2.9114230406392608e-06,
+      "loss": 0.8607,
+      "step": 4706
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6097622215755245,
+      "learning_rate": 2.907742452698664e-06,
+      "loss": 0.8665,
+      "step": 4707
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8712730757703152,
+      "learning_rate": 2.90406379688452e-06,
+      "loss": 0.3446,
+      "step": 4708
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5671107629470997,
+      "learning_rate": 2.900387074198997e-06,
+      "loss": 0.8077,
+      "step": 4709
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.39163987424883,
+      "learning_rate": 2.8967122856437435e-06,
+      "loss": 0.826,
+      "step": 4710
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2525924643573303,
+      "learning_rate": 2.893039432219861e-06,
+      "loss": 0.829,
+      "step": 4711
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.695902427309275,
+      "learning_rate": 2.8893685149279417e-06,
+      "loss": 0.8593,
+      "step": 4712
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4347093500493817,
+      "learning_rate": 2.885699534768047e-06,
+      "loss": 0.8287,
+      "step": 4713
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1842828765108258,
+      "learning_rate": 2.882032492739706e-06,
+      "loss": 0.8064,
+      "step": 4714
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2926222705767059,
+      "learning_rate": 2.8783673898419194e-06,
+      "loss": 0.802,
+      "step": 4715
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5344490395161348,
+      "learning_rate": 2.874704227073164e-06,
+      "loss": 0.8158,
+      "step": 4716
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2923744714475556,
+      "learning_rate": 2.871043005431394e-06,
+      "loss": 0.8168,
+      "step": 4717
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3479601994120898,
+      "learning_rate": 2.8673837259140138e-06,
+      "loss": 0.8182,
+      "step": 4718
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.573127027851483,
+      "learning_rate": 2.863726389517918e-06,
+      "loss": 0.8421,
+      "step": 4719
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8362987299905271,
+      "learning_rate": 2.860070997239469e-06,
+      "loss": 0.778,
+      "step": 4720
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8000931830824157,
+      "learning_rate": 2.856417550074495e-06,
+      "loss": 0.3055,
+      "step": 4721
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.362140206692828,
+      "learning_rate": 2.852766049018291e-06,
+      "loss": 0.8111,
+      "step": 4722
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2076487516405343,
+      "learning_rate": 2.8491164950656313e-06,
+      "loss": 0.9246,
+      "step": 4723
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3447822124349311,
+      "learning_rate": 2.8454688892107518e-06,
+      "loss": 0.8586,
+      "step": 4724
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3955087041084533,
+      "learning_rate": 2.841823232447366e-06,
+      "loss": 0.82,
+      "step": 4725
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3295635657878024,
+      "learning_rate": 2.8381795257686485e-06,
+      "loss": 0.848,
+      "step": 4726
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.220902574267021,
+      "learning_rate": 2.8345377701672404e-06,
+      "loss": 0.8345,
+      "step": 4727
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7915873389831936,
+      "learning_rate": 2.8308979666352644e-06,
+      "loss": 0.7619,
+      "step": 4728
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4689146489195513,
+      "learning_rate": 2.8272601161643e-06,
+      "loss": 0.863,
+      "step": 4729
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7847020878284674,
+      "learning_rate": 2.8236242197453943e-06,
+      "loss": 0.3314,
+      "step": 4730
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.416161603309013,
+      "learning_rate": 2.819990278369069e-06,
+      "loss": 0.781,
+      "step": 4731
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5261454088558803,
+      "learning_rate": 2.816358293025314e-06,
+      "loss": 0.755,
+      "step": 4732
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2681060949319873,
+      "learning_rate": 2.812728264703577e-06,
+      "loss": 0.8548,
+      "step": 4733
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5987856907762272,
+      "learning_rate": 2.8091001943927764e-06,
+      "loss": 0.8853,
+      "step": 4734
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2115621240193155,
+      "learning_rate": 2.8054740830813056e-06,
+      "loss": 0.7916,
+      "step": 4735
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.294371233962966,
+      "learning_rate": 2.8018499317570115e-06,
+      "loss": 0.7853,
+      "step": 4736
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2931185953172708,
+      "learning_rate": 2.79822774140721e-06,
+      "loss": 0.8079,
+      "step": 4737
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9602101266890735,
+      "learning_rate": 2.794607513018691e-06,
+      "loss": 0.9138,
+      "step": 4738
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2451759040301955,
+      "learning_rate": 2.79098924757771e-06,
+      "loss": 0.8024,
+      "step": 4739
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1412515935946763,
+      "learning_rate": 2.7873729460699684e-06,
+      "loss": 0.7975,
+      "step": 4740
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.439696610974413,
+      "learning_rate": 2.7837586094806535e-06,
+      "loss": 0.8942,
+      "step": 4741
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3840311150998554,
+      "learning_rate": 2.7801462387944145e-06,
+      "loss": 0.8432,
+      "step": 4742
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3094201083595784,
+      "learning_rate": 2.7765358349953554e-06,
+      "loss": 0.7871,
+      "step": 4743
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4368429651957844,
+      "learning_rate": 2.772927399067048e-06,
+      "loss": 0.8224,
+      "step": 4744
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4615771160797788,
+      "learning_rate": 2.769320931992535e-06,
+      "loss": 0.7728,
+      "step": 4745
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.411504566498782,
+      "learning_rate": 2.765716434754315e-06,
+      "loss": 0.8077,
+      "step": 4746
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3903705385543468,
+      "learning_rate": 2.762113908334351e-06,
+      "loss": 0.7877,
+      "step": 4747
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.401003650205048,
+      "learning_rate": 2.758513353714073e-06,
+      "loss": 0.7891,
+      "step": 4748
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4304251526689573,
+      "learning_rate": 2.7549147718743684e-06,
+      "loss": 0.7866,
+      "step": 4749
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.43539533658667,
+      "learning_rate": 2.7513181637955945e-06,
+      "loss": 0.7968,
+      "step": 4750
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8902246449580287,
+      "learning_rate": 2.74772353045756e-06,
+      "loss": 0.3248,
+      "step": 4751
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3298765406444581,
+      "learning_rate": 2.74413087283955e-06,
+      "loss": 0.8225,
+      "step": 4752
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3946123640528336,
+      "learning_rate": 2.7405401919202967e-06,
+      "loss": 0.7738,
+      "step": 4753
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.347772011415199,
+      "learning_rate": 2.736951488678006e-06,
+      "loss": 0.8393,
+      "step": 4754
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3137958145716673,
+      "learning_rate": 2.733364764090336e-06,
+      "loss": 0.8634,
+      "step": 4755
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1784630872795685,
+      "learning_rate": 2.729780019134408e-06,
+      "loss": 0.827,
+      "step": 4756
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3337123175933627,
+      "learning_rate": 2.726197254786811e-06,
+      "loss": 0.8242,
+      "step": 4757
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5238580966770316,
+      "learning_rate": 2.722616472023585e-06,
+      "loss": 0.8548,
+      "step": 4758
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2206352970352585,
+      "learning_rate": 2.7190376718202304e-06,
+      "loss": 0.8462,
+      "step": 4759
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6015753745515238,
+      "learning_rate": 2.715460855151716e-06,
+      "loss": 0.7925,
+      "step": 4760
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3470440517563254,
+      "learning_rate": 2.711886022992469e-06,
+      "loss": 0.8337,
+      "step": 4761
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5101020753934624,
+      "learning_rate": 2.708313176316363e-06,
+      "loss": 0.8181,
+      "step": 4762
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0197095820056687,
+      "learning_rate": 2.7047423160967433e-06,
+      "loss": 0.3047,
+      "step": 4763
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3298754453630075,
+      "learning_rate": 2.701173443306414e-06,
+      "loss": 0.7523,
+      "step": 4764
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.296041938494606,
+      "learning_rate": 2.6976065589176337e-06,
+      "loss": 0.8572,
+      "step": 4765
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3128053176640733,
+      "learning_rate": 2.694041663902114e-06,
+      "loss": 0.8599,
+      "step": 4766
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.762393339079883,
+      "learning_rate": 2.690478759231039e-06,
+      "loss": 0.8208,
+      "step": 4767
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8830039129364213,
+      "learning_rate": 2.686917845875038e-06,
+      "loss": 0.8371,
+      "step": 4768
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.345728826900248,
+      "learning_rate": 2.683358924804198e-06,
+      "loss": 0.8638,
+      "step": 4769
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6009295468874667,
+      "learning_rate": 2.679801996988075e-06,
+      "loss": 0.9421,
+      "step": 4770
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.309622731891372,
+      "learning_rate": 2.676247063395668e-06,
+      "loss": 0.8002,
+      "step": 4771
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4069561645583646,
+      "learning_rate": 2.6726941249954443e-06,
+      "loss": 0.8256,
+      "step": 4772
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7235899090492846,
+      "learning_rate": 2.669143182755315e-06,
+      "loss": 0.8037,
+      "step": 4773
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3571092828848632,
+      "learning_rate": 2.6655942376426635e-06,
+      "loss": 0.7565,
+      "step": 4774
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9341718825705039,
+      "learning_rate": 2.6620472906243123e-06,
+      "loss": 0.8373,
+      "step": 4775
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5728898313872097,
+      "learning_rate": 2.6585023426665534e-06,
+      "loss": 0.8731,
+      "step": 4776
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4449965260473816,
+      "learning_rate": 2.6549593947351258e-06,
+      "loss": 0.8004,
+      "step": 4777
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3828991570502336,
+      "learning_rate": 2.6514184477952244e-06,
+      "loss": 0.7977,
+      "step": 4778
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3006525559527446,
+      "learning_rate": 2.6478795028115046e-06,
+      "loss": 0.8631,
+      "step": 4779
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3091878232855372,
+      "learning_rate": 2.644342560748071e-06,
+      "loss": 0.7841,
+      "step": 4780
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3827949202222882,
+      "learning_rate": 2.6408076225684808e-06,
+      "loss": 0.8129,
+      "step": 4781
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4711646043708109,
+      "learning_rate": 2.6372746892357514e-06,
+      "loss": 0.7798,
+      "step": 4782
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4743790598241788,
+      "learning_rate": 2.6337437617123586e-06,
+      "loss": 0.825,
+      "step": 4783
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4533201043090753,
+      "learning_rate": 2.630214840960209e-06,
+      "loss": 0.8003,
+      "step": 4784
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8438111336521879,
+      "learning_rate": 2.626687927940688e-06,
+      "loss": 0.7874,
+      "step": 4785
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4532204437700442,
+      "learning_rate": 2.623163023614623e-06,
+      "loss": 0.7926,
+      "step": 4786
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2711403768875933,
+      "learning_rate": 2.6196401289422955e-06,
+      "loss": 0.8532,
+      "step": 4787
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3025418386631398,
+      "learning_rate": 2.6161192448834348e-06,
+      "loss": 0.8037,
+      "step": 4788
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3535184750107214,
+      "learning_rate": 2.6126003723972325e-06,
+      "loss": 0.8376,
+      "step": 4789
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3210481316875209,
+      "learning_rate": 2.609083512442323e-06,
+      "loss": 0.8505,
+      "step": 4790
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.419981395754241,
+      "learning_rate": 2.6055686659767944e-06,
+      "loss": 0.8112,
+      "step": 4791
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4878739089908137,
+      "learning_rate": 2.6020558339581893e-06,
+      "loss": 0.8181,
+      "step": 4792
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4601986426827283,
+      "learning_rate": 2.5985450173435035e-06,
+      "loss": 0.8827,
+      "step": 4793
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4281533105431454,
+      "learning_rate": 2.5950362170891774e-06,
+      "loss": 0.8276,
+      "step": 4794
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4420822730530534,
+      "learning_rate": 2.591529434151101e-06,
+      "loss": 0.886,
+      "step": 4795
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3207776696938347,
+      "learning_rate": 2.5880246694846266e-06,
+      "loss": 0.7761,
+      "step": 4796
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3662813726105645,
+      "learning_rate": 2.584521924044544e-06,
+      "loss": 0.8614,
+      "step": 4797
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4237891023099956,
+      "learning_rate": 2.5810211987850953e-06,
+      "loss": 0.7869,
+      "step": 4798
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.351238349675533,
+      "learning_rate": 2.577522494659981e-06,
+      "loss": 0.8733,
+      "step": 4799
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.499371123609028,
+      "learning_rate": 2.574025812622337e-06,
+      "loss": 0.8017,
+      "step": 4800
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4614065648888892,
+      "learning_rate": 2.5705311536247622e-06,
+      "loss": 0.8006,
+      "step": 4801
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6781911389185815,
+      "learning_rate": 2.567038518619297e-06,
+      "loss": 0.8639,
+      "step": 4802
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.38417838784059,
+      "learning_rate": 2.5635479085574267e-06,
+      "loss": 0.8496,
+      "step": 4803
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.379796672030245,
+      "learning_rate": 2.5600593243900927e-06,
+      "loss": 0.8425,
+      "step": 4804
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5074689083269697,
+      "learning_rate": 2.5565727670676887e-06,
+      "loss": 0.8421,
+      "step": 4805
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1869306218867617,
+      "learning_rate": 2.5530882375400358e-06,
+      "loss": 0.8157,
+      "step": 4806
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1947729113732086,
+      "learning_rate": 2.549605736756423e-06,
+      "loss": 0.7717,
+      "step": 4807
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.632941741710884,
+      "learning_rate": 2.5461252656655813e-06,
+      "loss": 0.81,
+      "step": 4808
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2887486762177378,
+      "learning_rate": 2.5426468252156846e-06,
+      "loss": 0.8519,
+      "step": 4809
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5638530999744125,
+      "learning_rate": 2.5391704163543516e-06,
+      "loss": 0.8048,
+      "step": 4810
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2852429651293653,
+      "learning_rate": 2.535696040028658e-06,
+      "loss": 0.7383,
+      "step": 4811
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4417582130002655,
+      "learning_rate": 2.532223697185122e-06,
+      "loss": 0.852,
+      "step": 4812
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3789184896666142,
+      "learning_rate": 2.528753388769697e-06,
+      "loss": 0.8387,
+      "step": 4813
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6858790082569208,
+      "learning_rate": 2.5252851157277945e-06,
+      "loss": 0.8815,
+      "step": 4814
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5436280749219542,
+      "learning_rate": 2.521818879004271e-06,
+      "loss": 0.7918,
+      "step": 4815
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.384937980482031,
+      "learning_rate": 2.518354679543422e-06,
+      "loss": 0.8232,
+      "step": 4816
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7151801927878798,
+      "learning_rate": 2.514892518288988e-06,
+      "loss": 0.3207,
+      "step": 4817
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.469952707223259,
+      "learning_rate": 2.511432396184165e-06,
+      "loss": 0.8778,
+      "step": 4818
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3119519720184325,
+      "learning_rate": 2.5079743141715786e-06,
+      "loss": 0.8678,
+      "step": 4819
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4311901218994671,
+      "learning_rate": 2.504518273193306e-06,
+      "loss": 0.7825,
+      "step": 4820
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2659290869708009,
+      "learning_rate": 2.501064274190872e-06,
+      "loss": 0.7607,
+      "step": 4821
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5034695090230736,
+      "learning_rate": 2.497612318105237e-06,
+      "loss": 0.8115,
+      "step": 4822
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5035590080569217,
+      "learning_rate": 2.4941624058768143e-06,
+      "loss": 0.8338,
+      "step": 4823
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5047508871338022,
+      "learning_rate": 2.4907145384454514e-06,
+      "loss": 0.8547,
+      "step": 4824
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.468443482517806,
+      "learning_rate": 2.4872687167504393e-06,
+      "loss": 0.7849,
+      "step": 4825
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6360440024163683,
+      "learning_rate": 2.4838249417305214e-06,
+      "loss": 0.8144,
+      "step": 4826
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3644007756030407,
+      "learning_rate": 2.480383214323875e-06,
+      "loss": 0.8602,
+      "step": 4827
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8615700366887754,
+      "learning_rate": 2.476943535468117e-06,
+      "loss": 0.3462,
+      "step": 4828
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3804834628441032,
+      "learning_rate": 2.4735059061003143e-06,
+      "loss": 0.7757,
+      "step": 4829
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.446131978464802,
+      "learning_rate": 2.470070327156975e-06,
+      "loss": 0.8712,
+      "step": 4830
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3351877796965814,
+      "learning_rate": 2.4666367995740414e-06,
+      "loss": 0.8211,
+      "step": 4831
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5299335779708552,
+      "learning_rate": 2.4632053242869005e-06,
+      "loss": 0.7839,
+      "step": 4832
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.411685526488035,
+      "learning_rate": 2.4597759022303813e-06,
+      "loss": 0.8079,
+      "step": 4833
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6286817259844446,
+      "learning_rate": 2.4563485343387596e-06,
+      "loss": 0.8306,
+      "step": 4834
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4052947097769446,
+      "learning_rate": 2.4529232215457333e-06,
+      "loss": 0.7712,
+      "step": 4835
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2947740452447218,
+      "learning_rate": 2.4494999647844574e-06,
+      "loss": 0.8235,
+      "step": 4836
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.459633565323603,
+      "learning_rate": 2.4460787649875263e-06,
+      "loss": 0.8135,
+      "step": 4837
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7226874463290647,
+      "learning_rate": 2.4426596230869624e-06,
+      "loss": 0.8139,
+      "step": 4838
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3856042822890058,
+      "learning_rate": 2.439242540014236e-06,
+      "loss": 0.8395,
+      "step": 4839
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4054211193631352,
+      "learning_rate": 2.4358275167002564e-06,
+      "loss": 0.8215,
+      "step": 4840
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4133799114657297,
+      "learning_rate": 2.4324145540753697e-06,
+      "loss": 0.7281,
+      "step": 4841
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5179157716082834,
+      "learning_rate": 2.429003653069357e-06,
+      "loss": 0.8626,
+      "step": 4842
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5471082388716655,
+      "learning_rate": 2.4255948146114483e-06,
+      "loss": 0.8076,
+      "step": 4843
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2998987395997688,
+      "learning_rate": 2.4221880396302977e-06,
+      "loss": 0.8017,
+      "step": 4844
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.400121432864329,
+      "learning_rate": 2.418783329054013e-06,
+      "loss": 0.7756,
+      "step": 4845
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.267401055427947,
+      "learning_rate": 2.4153806838101257e-06,
+      "loss": 0.9318,
+      "step": 4846
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5193188115508298,
+      "learning_rate": 2.4119801048256096e-06,
+      "loss": 0.7923,
+      "step": 4847
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6039467045122593,
+      "learning_rate": 2.4085815930268807e-06,
+      "loss": 0.843,
+      "step": 4848
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4440248114301522,
+      "learning_rate": 2.4051851493397835e-06,
+      "loss": 0.8449,
+      "step": 4849
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4906822059158764,
+      "learning_rate": 2.401790774689602e-06,
+      "loss": 0.7083,
+      "step": 4850
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4548501133684855,
+      "learning_rate": 2.3983984700010587e-06,
+      "loss": 0.8982,
+      "step": 4851
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.371044080600788,
+      "learning_rate": 2.395008236198315e-06,
+      "loss": 0.8117,
+      "step": 4852
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3581950825970681,
+      "learning_rate": 2.39162007420496e-06,
+      "loss": 0.8198,
+      "step": 4853
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4548097151697144,
+      "learning_rate": 2.3882339849440206e-06,
+      "loss": 0.7942,
+      "step": 4854
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3890936347842693,
+      "learning_rate": 2.384849969337967e-06,
+      "loss": 0.8049,
+      "step": 4855
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3257029986849345,
+      "learning_rate": 2.381468028308693e-06,
+      "loss": 0.7812,
+      "step": 4856
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3278843233877267,
+      "learning_rate": 2.378088162777532e-06,
+      "loss": 0.8543,
+      "step": 4857
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8270242264498211,
+      "learning_rate": 2.3747103736652543e-06,
+      "loss": 0.3562,
+      "step": 4858
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.531645704249799,
+      "learning_rate": 2.371334661892066e-06,
+      "loss": 0.8645,
+      "step": 4859
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2539261492519806,
+      "learning_rate": 2.3679610283776023e-06,
+      "loss": 0.7827,
+      "step": 4860
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.498486350573901,
+      "learning_rate": 2.3645894740409294e-06,
+      "loss": 0.6973,
+      "step": 4861
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3829209674439278,
+      "learning_rate": 2.361219999800558e-06,
+      "loss": 0.7672,
+      "step": 4862
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7305582746679888,
+      "learning_rate": 2.3578526065744223e-06,
+      "loss": 0.8456,
+      "step": 4863
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.475110020817717,
+      "learning_rate": 2.3544872952798913e-06,
+      "loss": 0.7948,
+      "step": 4864
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2527863256211342,
+      "learning_rate": 2.351124066833773e-06,
+      "loss": 0.8552,
+      "step": 4865
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6116378382492695,
+      "learning_rate": 2.3477629221522992e-06,
+      "loss": 0.8105,
+      "step": 4866
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5100930193290318,
+      "learning_rate": 2.3444038621511435e-06,
+      "loss": 0.8675,
+      "step": 4867
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4794449361173752,
+      "learning_rate": 2.341046887745403e-06,
+      "loss": 0.799,
+      "step": 4868
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3365418004334388,
+      "learning_rate": 2.3376919998496083e-06,
+      "loss": 0.8863,
+      "step": 4869
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2790553167018173,
+      "learning_rate": 2.3343391993777274e-06,
+      "loss": 0.7871,
+      "step": 4870
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4909981686352511,
+      "learning_rate": 2.330988487243152e-06,
+      "loss": 0.7723,
+      "step": 4871
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5190453671884316,
+      "learning_rate": 2.3276398643587127e-06,
+      "loss": 0.7908,
+      "step": 4872
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.891691762679748,
+      "learning_rate": 2.3242933316366613e-06,
+      "loss": 0.3192,
+      "step": 4873
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2974824933178404,
+      "learning_rate": 2.320948889988691e-06,
+      "loss": 0.8117,
+      "step": 4874
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4645519601759032,
+      "learning_rate": 2.3176065403259184e-06,
+      "loss": 0.7029,
+      "step": 4875
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6171250187280124,
+      "learning_rate": 2.314266283558887e-06,
+      "loss": 0.8657,
+      "step": 4876
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5250751355932701,
+      "learning_rate": 2.310928120597581e-06,
+      "loss": 0.8258,
+      "step": 4877
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2718792633944718,
+      "learning_rate": 2.307592052351406e-06,
+      "loss": 0.8567,
+      "step": 4878
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3808272262240875,
+      "learning_rate": 2.3042580797291956e-06,
+      "loss": 0.7549,
+      "step": 4879
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3979685119512733,
+      "learning_rate": 2.300926203639218e-06,
+      "loss": 0.8039,
+      "step": 4880
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.283493816858713,
+      "learning_rate": 2.2975964249891723e-06,
+      "loss": 0.7298,
+      "step": 4881
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.437431135549487,
+      "learning_rate": 2.294268744686178e-06,
+      "loss": 0.8407,
+      "step": 4882
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4352529977511503,
+      "learning_rate": 2.2909431636367853e-06,
+      "loss": 0.8531,
+      "step": 4883
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2454554161989482,
+      "learning_rate": 2.2876196827469776e-06,
+      "loss": 0.7123,
+      "step": 4884
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7406299700143835,
+      "learning_rate": 2.284298302922162e-06,
+      "loss": 0.8469,
+      "step": 4885
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4686450819323873,
+      "learning_rate": 2.2809790250671717e-06,
+      "loss": 0.7989,
+      "step": 4886
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3334465335420826,
+      "learning_rate": 2.2776618500862725e-06,
+      "loss": 0.8732,
+      "step": 4887
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4262962802414627,
+      "learning_rate": 2.2743467788831496e-06,
+      "loss": 0.7672,
+      "step": 4888
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6001630996112,
+      "learning_rate": 2.2710338123609264e-06,
+      "loss": 0.8628,
+      "step": 4889
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3171500823310665,
+      "learning_rate": 2.267722951422141e-06,
+      "loss": 0.8298,
+      "step": 4890
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3425310172683729,
+      "learning_rate": 2.2644141969687674e-06,
+      "loss": 0.7635,
+      "step": 4891
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4090316969685044,
+      "learning_rate": 2.2611075499021985e-06,
+      "loss": 0.8712,
+      "step": 4892
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2114792288844773,
+      "learning_rate": 2.257803011123254e-06,
+      "loss": 0.7515,
+      "step": 4893
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5790682207030717,
+      "learning_rate": 2.2545005815321875e-06,
+      "loss": 0.821,
+      "step": 4894
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5206421610230947,
+      "learning_rate": 2.2512002620286653e-06,
+      "loss": 0.8152,
+      "step": 4895
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9888513366390224,
+      "learning_rate": 2.2479020535117924e-06,
+      "loss": 0.321,
+      "step": 4896
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2941426662874576,
+      "learning_rate": 2.2446059568800872e-06,
+      "loss": 0.8273,
+      "step": 4897
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3714165837312662,
+      "learning_rate": 2.241311973031496e-06,
+      "loss": 0.7805,
+      "step": 4898
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2380605968849425,
+      "learning_rate": 2.2380201028633954e-06,
+      "loss": 0.8354,
+      "step": 4899
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2413457569161983,
+      "learning_rate": 2.2347303472725778e-06,
+      "loss": 0.7661,
+      "step": 4900
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5217746633983296,
+      "learning_rate": 2.2314427071552624e-06,
+      "loss": 0.7944,
+      "step": 4901
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.416094124689887,
+      "learning_rate": 2.228157183407096e-06,
+      "loss": 0.7765,
+      "step": 4902
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3228778795096796,
+      "learning_rate": 2.2248737769231497e-06,
+      "loss": 0.8357,
+      "step": 4903
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.427865929680922,
+      "learning_rate": 2.2215924885979035e-06,
+      "loss": 0.7986,
+      "step": 4904
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.32154536809842,
+      "learning_rate": 2.218313319325277e-06,
+      "loss": 0.785,
+      "step": 4905
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3219310486524143,
+      "learning_rate": 2.2150362699986084e-06,
+      "loss": 0.8328,
+      "step": 4906
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1973314757230091,
+      "learning_rate": 2.2117613415106533e-06,
+      "loss": 0.7628,
+      "step": 4907
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3054948681445049,
+      "learning_rate": 2.20848853475359e-06,
+      "loss": 0.8044,
+      "step": 4908
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4631344104437332,
+      "learning_rate": 2.2052178506190267e-06,
+      "loss": 0.821,
+      "step": 4909
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3585536985981372,
+      "learning_rate": 2.201949289997983e-06,
+      "loss": 0.8458,
+      "step": 4910
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4709928522920759,
+      "learning_rate": 2.1986828537809103e-06,
+      "loss": 0.8193,
+      "step": 4911
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.274851194222755,
+      "learning_rate": 2.195418542857669e-06,
+      "loss": 0.8186,
+      "step": 4912
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2363040252795432,
+      "learning_rate": 2.1921563581175553e-06,
+      "loss": 0.8189,
+      "step": 4913
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.455114285448573,
+      "learning_rate": 2.1888963004492735e-06,
+      "loss": 0.7653,
+      "step": 4914
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3326152922264465,
+      "learning_rate": 2.1856383707409513e-06,
+      "loss": 0.8291,
+      "step": 4915
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4497609910594378,
+      "learning_rate": 2.1823825698801424e-06,
+      "loss": 0.8672,
+      "step": 4916
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2494401601952263,
+      "learning_rate": 2.179128898753814e-06,
+      "loss": 0.8037,
+      "step": 4917
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.364618215557938,
+      "learning_rate": 2.1758773582483594e-06,
+      "loss": 0.8156,
+      "step": 4918
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.70733421984757,
+      "learning_rate": 2.172627949249586e-06,
+      "loss": 0.7832,
+      "step": 4919
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.33817857650755,
+      "learning_rate": 2.169380672642719e-06,
+      "loss": 0.8458,
+      "step": 4920
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4138624408050462,
+      "learning_rate": 2.166135529312412e-06,
+      "loss": 0.8032,
+      "step": 4921
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.558824437261852,
+      "learning_rate": 2.1628925201427285e-06,
+      "loss": 0.3233,
+      "step": 4922
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3523664770635242,
+      "learning_rate": 2.159651646017151e-06,
+      "loss": 0.8588,
+      "step": 4923
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.280496848768631,
+      "learning_rate": 2.156412907818586e-06,
+      "loss": 0.7931,
+      "step": 4924
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5848470060493143,
+      "learning_rate": 2.1531763064293587e-06,
+      "loss": 0.8278,
+      "step": 4925
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3847246345565354,
+      "learning_rate": 2.1499418427312002e-06,
+      "loss": 0.8234,
+      "step": 4926
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4651939431258614,
+      "learning_rate": 2.146709517605271e-06,
+      "loss": 0.857,
+      "step": 4927
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4799943764088812,
+      "learning_rate": 2.14347933193215e-06,
+      "loss": 0.8846,
+      "step": 4928
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3409975307065036,
+      "learning_rate": 2.1402512865918246e-06,
+      "loss": 0.7431,
+      "step": 4929
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6089024807819252,
+      "learning_rate": 2.1370253824637023e-06,
+      "loss": 0.8386,
+      "step": 4930
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.371510306547513,
+      "learning_rate": 2.1338016204266087e-06,
+      "loss": 0.8359,
+      "step": 4931
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3809211425779309,
+      "learning_rate": 2.1305800013587906e-06,
+      "loss": 0.801,
+      "step": 4932
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.477976659202525,
+      "learning_rate": 2.1273605261379015e-06,
+      "loss": 0.8911,
+      "step": 4933
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3540908511805727,
+      "learning_rate": 2.124143195641013e-06,
+      "loss": 0.7525,
+      "step": 4934
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.23683413249256,
+      "learning_rate": 2.1209280107446195e-06,
+      "loss": 0.8613,
+      "step": 4935
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3717767666163534,
+      "learning_rate": 2.117714972324624e-06,
+      "loss": 0.7137,
+      "step": 4936
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6411127512101575,
+      "learning_rate": 2.1145040812563432e-06,
+      "loss": 0.8101,
+      "step": 4937
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3018105332386591,
+      "learning_rate": 2.1112953384145195e-06,
+      "loss": 0.7273,
+      "step": 4938
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5321370184354948,
+      "learning_rate": 2.1080887446732955e-06,
+      "loss": 0.8928,
+      "step": 4939
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4569233337916576,
+      "learning_rate": 2.1048843009062424e-06,
+      "loss": 0.7999,
+      "step": 4940
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4689522208340593,
+      "learning_rate": 2.1016820079863366e-06,
+      "loss": 0.8395,
+      "step": 4941
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2633881729492629,
+      "learning_rate": 2.0984818667859665e-06,
+      "loss": 0.8091,
+      "step": 4942
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.429791619986001,
+      "learning_rate": 2.0952838781769446e-06,
+      "loss": 0.8175,
+      "step": 4943
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4025088986031191,
+      "learning_rate": 2.09208804303049e-06,
+      "loss": 0.7822,
+      "step": 4944
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4240692389770133,
+      "learning_rate": 2.088894362217233e-06,
+      "loss": 0.745,
+      "step": 4945
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4629725228317716,
+      "learning_rate": 2.0857028366072217e-06,
+      "loss": 0.8252,
+      "step": 4946
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.828667828838698,
+      "learning_rate": 2.0825134670699232e-06,
+      "loss": 0.3224,
+      "step": 4947
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9034732524242843,
+      "learning_rate": 2.0793262544741965e-06,
+      "loss": 0.766,
+      "step": 4948
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5508099602556762,
+      "learning_rate": 2.076141199688333e-06,
+      "loss": 0.7234,
+      "step": 4949
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2106260760612215,
+      "learning_rate": 2.072958303580033e-06,
+      "loss": 0.8143,
+      "step": 4950
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.305842215539109,
+      "learning_rate": 2.0697775670164e-06,
+      "loss": 0.7832,
+      "step": 4951
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3152337498290907,
+      "learning_rate": 2.0665989908639526e-06,
+      "loss": 0.9129,
+      "step": 4952
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5543578004364778,
+      "learning_rate": 2.063422575988626e-06,
+      "loss": 0.8357,
+      "step": 4953
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3797778192897858,
+      "learning_rate": 2.060248323255769e-06,
+      "loss": 0.7668,
+      "step": 4954
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6743870511952674,
+      "learning_rate": 2.057076233530123e-06,
+      "loss": 0.8455,
+      "step": 4955
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3367015399415796,
+      "learning_rate": 2.053906307675859e-06,
+      "loss": 0.8379,
+      "step": 4956
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4031000342728666,
+      "learning_rate": 2.0507385465565557e-06,
+      "loss": 0.8457,
+      "step": 4957
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4750816732006027,
+      "learning_rate": 2.0475729510351937e-06,
+      "loss": 0.875,
+      "step": 4958
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2764264248423196,
+      "learning_rate": 2.044409521974168e-06,
+      "loss": 0.7769,
+      "step": 4959
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8383751048744504,
+      "learning_rate": 2.0412482602352877e-06,
+      "loss": 0.3404,
+      "step": 4960
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.326247915354842,
+      "learning_rate": 2.038089166679763e-06,
+      "loss": 0.8359,
+      "step": 4961
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4383470239990104,
+      "learning_rate": 2.0349322421682238e-06,
+      "loss": 0.8909,
+      "step": 4962
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.244927983400676,
+      "learning_rate": 2.0317774875606997e-06,
+      "loss": 0.8782,
+      "step": 4963
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4768380476919143,
+      "learning_rate": 2.0286249037166316e-06,
+      "loss": 0.8641,
+      "step": 4964
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.56417525546673,
+      "learning_rate": 2.025474491494874e-06,
+      "loss": 0.833,
+      "step": 4965
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4872032852321055,
+      "learning_rate": 2.0223262517536833e-06,
+      "loss": 0.882,
+      "step": 4966
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7139346492564387,
+      "learning_rate": 2.0191801853507264e-06,
+      "loss": 0.8147,
+      "step": 4967
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5696320945675561,
+      "learning_rate": 2.0160362931430788e-06,
+      "loss": 0.8434,
+      "step": 4968
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2844456676431177,
+      "learning_rate": 2.0128945759872264e-06,
+      "loss": 0.7763,
+      "step": 4969
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5926982188283063,
+      "learning_rate": 2.009755034739057e-06,
+      "loss": 0.7946,
+      "step": 4970
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3866952652352946,
+      "learning_rate": 2.0066176702538675e-06,
+      "loss": 0.853,
+      "step": 4971
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.281355354510402,
+      "learning_rate": 2.0034824833863652e-06,
+      "loss": 0.7551,
+      "step": 4972
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4413213343018103,
+      "learning_rate": 2.0003494749906595e-06,
+      "loss": 0.7746,
+      "step": 4973
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3317028127368096,
+      "learning_rate": 1.9972186459202656e-06,
+      "loss": 0.7318,
+      "step": 4974
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.578013515901678,
+      "learning_rate": 1.9940899970281115e-06,
+      "loss": 0.8293,
+      "step": 4975
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1725247862507482,
+      "learning_rate": 1.9909635291665307e-06,
+      "loss": 0.8366,
+      "step": 4976
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4371700128390135,
+      "learning_rate": 1.9878392431872506e-06,
+      "loss": 0.8911,
+      "step": 4977
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3495542491524504,
+      "learning_rate": 1.984717139941417e-06,
+      "loss": 0.765,
+      "step": 4978
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.382734345662889,
+      "learning_rate": 1.9815972202795796e-06,
+      "loss": 0.9208,
+      "step": 4979
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.542160043162817,
+      "learning_rate": 1.9784794850516874e-06,
+      "loss": 0.9225,
+      "step": 4980
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4904646948596165,
+      "learning_rate": 1.9753639351070954e-06,
+      "loss": 0.8672,
+      "step": 4981
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2483181367610416,
+      "learning_rate": 1.9722505712945715e-06,
+      "loss": 0.7887,
+      "step": 4982
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8341131122132301,
+      "learning_rate": 1.969139394462277e-06,
+      "loss": 0.3182,
+      "step": 4983
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3884108280596756,
+      "learning_rate": 1.9660304054577815e-06,
+      "loss": 0.7783,
+      "step": 4984
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4760301515876666,
+      "learning_rate": 1.962923605128064e-06,
+      "loss": 0.7501,
+      "step": 4985
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3547770414492288,
+      "learning_rate": 1.9598189943194965e-06,
+      "loss": 0.8309,
+      "step": 4986
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2618513214805087,
+      "learning_rate": 1.956716573877867e-06,
+      "loss": 0.7802,
+      "step": 4987
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3440339937459915,
+      "learning_rate": 1.953616344648357e-06,
+      "loss": 0.7934,
+      "step": 4988
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.555677460578652,
+      "learning_rate": 1.9505183074755516e-06,
+      "loss": 0.8707,
+      "step": 4989
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4685647396165218,
+      "learning_rate": 1.9474224632034442e-06,
+      "loss": 0.8026,
+      "step": 4990
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3080752904300719,
+      "learning_rate": 1.944328812675432e-06,
+      "loss": 0.776,
+      "step": 4991
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1745329560056474,
+      "learning_rate": 1.941237356734307e-06,
+      "loss": 0.8533,
+      "step": 4992
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4447081765168766,
+      "learning_rate": 1.938148096222264e-06,
+      "loss": 0.778,
+      "step": 4993
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.398319234056097,
+      "learning_rate": 1.935061031980909e-06,
+      "loss": 0.8075,
+      "step": 4994
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3836104009133627,
+      "learning_rate": 1.9319761648512404e-06,
+      "loss": 0.8145,
+      "step": 4995
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4189440523103918,
+      "learning_rate": 1.9288934956736572e-06,
+      "loss": 0.833,
+      "step": 4996
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.381502206556557,
+      "learning_rate": 1.9258130252879683e-06,
+      "loss": 0.8189,
+      "step": 4997
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.337234391122214,
+      "learning_rate": 1.922734754533382e-06,
+      "loss": 0.8829,
+      "step": 4998
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6450991646663586,
+      "learning_rate": 1.919658684248494e-06,
+      "loss": 0.8132,
+      "step": 4999
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3288605896889056,
+      "learning_rate": 1.9165848152713174e-06,
+      "loss": 0.8382,
+      "step": 5000
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5920815616961541,
+      "learning_rate": 1.91351314843926e-06,
+      "loss": 0.8538,
+      "step": 5001
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3735396530560529,
+      "learning_rate": 1.910443684589127e-06,
+      "loss": 0.8138,
+      "step": 5002
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.373939426866001,
+      "learning_rate": 1.9073764245571204e-06,
+      "loss": 0.7805,
+      "step": 5003
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2317954445971735,
+      "learning_rate": 1.904311369178854e-06,
+      "loss": 0.7774,
+      "step": 5004
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5220337733999811,
+      "learning_rate": 1.9012485192893283e-06,
+      "loss": 0.8089,
+      "step": 5005
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.310509976096398,
+      "learning_rate": 1.8981878757229478e-06,
+      "loss": 0.8564,
+      "step": 5006
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4455000890961793,
+      "learning_rate": 1.89512943931352e-06,
+      "loss": 0.858,
+      "step": 5007
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7857974338097528,
+      "learning_rate": 1.892073210894242e-06,
+      "loss": 0.3182,
+      "step": 5008
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3308503413194692,
+      "learning_rate": 1.8890191912977207e-06,
+      "loss": 0.8175,
+      "step": 5009
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3915294695467166,
+      "learning_rate": 1.88596738135595e-06,
+      "loss": 0.8437,
+      "step": 5010
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.240944446585345,
+      "learning_rate": 1.8829177819003308e-06,
+      "loss": 0.7826,
+      "step": 5011
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2770327924497142,
+      "learning_rate": 1.879870393761657e-06,
+      "loss": 0.8177,
+      "step": 5012
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2381078308343385,
+      "learning_rate": 1.8768252177701184e-06,
+      "loss": 0.9168,
+      "step": 5013
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4272174806235682,
+      "learning_rate": 1.8737822547553086e-06,
+      "loss": 0.7523,
+      "step": 5014
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6680535645552008,
+      "learning_rate": 1.870741505546212e-06,
+      "loss": 0.8354,
+      "step": 5015
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4755194337855941,
+      "learning_rate": 1.8677029709712147e-06,
+      "loss": 0.8325,
+      "step": 5016
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2488777755732807,
+      "learning_rate": 1.8646666518580968e-06,
+      "loss": 0.7864,
+      "step": 5017
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3643468344444352,
+      "learning_rate": 1.8616325490340326e-06,
+      "loss": 0.7791,
+      "step": 5018
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4526373419239407,
+      "learning_rate": 1.858600663325597e-06,
+      "loss": 0.7599,
+      "step": 5019
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6143843941003542,
+      "learning_rate": 1.8555709955587654e-06,
+      "loss": 0.8687,
+      "step": 5020
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4101868161193605,
+      "learning_rate": 1.8525435465588914e-06,
+      "loss": 0.8205,
+      "step": 5021
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5821432756861946,
+      "learning_rate": 1.8495183171507415e-06,
+      "loss": 0.8963,
+      "step": 5022
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3947321943354474,
+      "learning_rate": 1.8464953081584735e-06,
+      "loss": 0.8352,
+      "step": 5023
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1690344660963368,
+      "learning_rate": 1.843474520405637e-06,
+      "loss": 0.7914,
+      "step": 5024
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5407844933383912,
+      "learning_rate": 1.8404559547151746e-06,
+      "loss": 0.8012,
+      "step": 5025
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.306542936373972,
+      "learning_rate": 1.8374396119094307e-06,
+      "loss": 0.781,
+      "step": 5026
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5559624315763207,
+      "learning_rate": 1.8344254928101401e-06,
+      "loss": 0.8492,
+      "step": 5027
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3933385107918228,
+      "learning_rate": 1.8314135982384274e-06,
+      "loss": 0.8823,
+      "step": 5028
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.614022823429547,
+      "learning_rate": 1.828403929014818e-06,
+      "loss": 0.811,
+      "step": 5029
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7507100288045265,
+      "learning_rate": 1.8253964859592343e-06,
+      "loss": 0.8707,
+      "step": 5030
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4960798867980243,
+      "learning_rate": 1.8223912698909807e-06,
+      "loss": 0.782,
+      "step": 5031
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.299790092116909,
+      "learning_rate": 1.81938828162876e-06,
+      "loss": 0.81,
+      "step": 5032
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.475638757867349,
+      "learning_rate": 1.8163875219906735e-06,
+      "loss": 0.8497,
+      "step": 5033
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4561037161521124,
+      "learning_rate": 1.8133889917942093e-06,
+      "loss": 0.8392,
+      "step": 5034
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2019121982238032,
+      "learning_rate": 1.8103926918562465e-06,
+      "loss": 0.7549,
+      "step": 5035
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.387506539929717,
+      "learning_rate": 1.807398622993064e-06,
+      "loss": 0.8627,
+      "step": 5036
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2750852611407033,
+      "learning_rate": 1.804406786020325e-06,
+      "loss": 0.8511,
+      "step": 5037
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2466001064784635,
+      "learning_rate": 1.8014171817530934e-06,
+      "loss": 0.806,
+      "step": 5038
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6682169107628473,
+      "learning_rate": 1.7984298110058162e-06,
+      "loss": 0.8147,
+      "step": 5039
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4502595546891326,
+      "learning_rate": 1.7954446745923325e-06,
+      "loss": 0.7062,
+      "step": 5040
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3866398244882665,
+      "learning_rate": 1.7924617733258831e-06,
+      "loss": 0.8324,
+      "step": 5041
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6159636316932655,
+      "learning_rate": 1.7894811080190888e-06,
+      "loss": 0.8245,
+      "step": 5042
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4197749897517802,
+      "learning_rate": 1.7865026794839625e-06,
+      "loss": 0.7807,
+      "step": 5043
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3108587270617214,
+      "learning_rate": 1.7835264885319127e-06,
+      "loss": 0.7052,
+      "step": 5044
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3088003671565214,
+      "learning_rate": 1.7805525359737375e-06,
+      "loss": 0.7817,
+      "step": 5045
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8993869325934439,
+      "learning_rate": 1.7775808226196222e-06,
+      "loss": 0.3394,
+      "step": 5046
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.457712258802836,
+      "learning_rate": 1.7746113492791407e-06,
+      "loss": 0.8551,
+      "step": 5047
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.543952776597069,
+      "learning_rate": 1.7716441167612618e-06,
+      "loss": 0.9814,
+      "step": 5048
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2144080335884664,
+      "learning_rate": 1.7686791258743475e-06,
+      "loss": 0.8539,
+      "step": 5049
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2271072338277673,
+      "learning_rate": 1.765716377426131e-06,
+      "loss": 0.7837,
+      "step": 5050
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4025052988354734,
+      "learning_rate": 1.7627558722237525e-06,
+      "loss": 0.8025,
+      "step": 5051
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6193514129916942,
+      "learning_rate": 1.759797611073738e-06,
+      "loss": 0.821,
+      "step": 5052
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2738069425504421,
+      "learning_rate": 1.7568415947819973e-06,
+      "loss": 0.8177,
+      "step": 5053
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4691338430692886,
+      "learning_rate": 1.753887824153827e-06,
+      "loss": 0.812,
+      "step": 5054
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3654618332988104,
+      "learning_rate": 1.7509362999939217e-06,
+      "loss": 0.8219,
+      "step": 5055
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.467219133208719,
+      "learning_rate": 1.7479870231063544e-06,
+      "loss": 0.8599,
+      "step": 5056
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3472446670074512,
+      "learning_rate": 1.7450399942945884e-06,
+      "loss": 0.7314,
+      "step": 5057
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3045122091667622,
+      "learning_rate": 1.74209521436148e-06,
+      "loss": 0.7127,
+      "step": 5058
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4135689229651887,
+      "learning_rate": 1.739152684109262e-06,
+      "loss": 0.7722,
+      "step": 5059
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.406852187590174,
+      "learning_rate": 1.7362124043395678e-06,
+      "loss": 0.8515,
+      "step": 5060
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4771758425393624,
+      "learning_rate": 1.7332743758534076e-06,
+      "loss": 0.8116,
+      "step": 5061
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2923726165470755,
+      "learning_rate": 1.730338599451178e-06,
+      "loss": 0.8634,
+      "step": 5062
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3782086843537737,
+      "learning_rate": 1.727405075932671e-06,
+      "loss": 0.8849,
+      "step": 5063
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4676736092137324,
+      "learning_rate": 1.7244738060970567e-06,
+      "loss": 0.7694,
+      "step": 5064
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.361870870256793,
+      "learning_rate": 1.7215447907428907e-06,
+      "loss": 0.8263,
+      "step": 5065
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.707800114797023,
+      "learning_rate": 1.7186180306681221e-06,
+      "loss": 0.8249,
+      "step": 5066
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.200244593714298,
+      "learning_rate": 1.7156935266700814e-06,
+      "loss": 0.7765,
+      "step": 5067
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5507454063251245,
+      "learning_rate": 1.712771279545482e-06,
+      "loss": 0.7326,
+      "step": 5068
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3112502186302735,
+      "learning_rate": 1.7098512900904219e-06,
+      "loss": 0.8443,
+      "step": 5069
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5777756149604878,
+      "learning_rate": 1.706933559100391e-06,
+      "loss": 0.8657,
+      "step": 5070
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3952305573733161,
+      "learning_rate": 1.7040180873702629e-06,
+      "loss": 0.8325,
+      "step": 5071
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3971682854548926,
+      "learning_rate": 1.7011048756942817e-06,
+      "loss": 0.8687,
+      "step": 5072
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4110709285696406,
+      "learning_rate": 1.6981939248660938e-06,
+      "loss": 0.7954,
+      "step": 5073
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3690681434932765,
+      "learning_rate": 1.6952852356787241e-06,
+      "loss": 0.7396,
+      "step": 5074
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3798590500790802,
+      "learning_rate": 1.6923788089245764e-06,
+      "loss": 0.818,
+      "step": 5075
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8469252675699344,
+      "learning_rate": 1.6894746453954402e-06,
+      "loss": 0.3169,
+      "step": 5076
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.484077686098373,
+      "learning_rate": 1.6865727458824931e-06,
+      "loss": 0.8365,
+      "step": 5077
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3619481949965073,
+      "learning_rate": 1.6836731111762915e-06,
+      "loss": 0.8204,
+      "step": 5078
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3995813295076152,
+      "learning_rate": 1.6807757420667737e-06,
+      "loss": 0.8433,
+      "step": 5079
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4731730141318553,
+      "learning_rate": 1.6778806393432656e-06,
+      "loss": 0.811,
+      "step": 5080
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3800042268396344,
+      "learning_rate": 1.67498780379447e-06,
+      "loss": 0.8236,
+      "step": 5081
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.279325942808528,
+      "learning_rate": 1.6720972362084798e-06,
+      "loss": 0.7728,
+      "step": 5082
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5436306783039384,
+      "learning_rate": 1.6692089373727616e-06,
+      "loss": 0.8697,
+      "step": 5083
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8720506866655976,
+      "learning_rate": 1.6663229080741672e-06,
+      "loss": 0.3233,
+      "step": 5084
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2081759415999347,
+      "learning_rate": 1.6634391490989333e-06,
+      "loss": 0.8524,
+      "step": 5085
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2787716712005615,
+      "learning_rate": 1.6605576612326745e-06,
+      "loss": 0.8367,
+      "step": 5086
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5530114736359613,
+      "learning_rate": 1.6576784452603833e-06,
+      "loss": 0.8714,
+      "step": 5087
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6558686381911316,
+      "learning_rate": 1.654801501966442e-06,
+      "loss": 0.7839,
+      "step": 5088
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3853548777546105,
+      "learning_rate": 1.6519268321346104e-06,
+      "loss": 0.8155,
+      "step": 5089
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.486518980760601,
+      "learning_rate": 1.6490544365480266e-06,
+      "loss": 0.8509,
+      "step": 5090
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4579222916588561,
+      "learning_rate": 1.6461843159892055e-06,
+      "loss": 0.8492,
+      "step": 5091
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.577792589426919,
+      "learning_rate": 1.643316471240054e-06,
+      "loss": 0.7621,
+      "step": 5092
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6237117734367241,
+      "learning_rate": 1.6404509030818493e-06,
+      "loss": 0.8365,
+      "step": 5093
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1183655103434567,
+      "learning_rate": 1.6375876122952483e-06,
+      "loss": 0.7826,
+      "step": 5094
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7174729913457785,
+      "learning_rate": 1.6347265996602935e-06,
+      "loss": 0.8131,
+      "step": 5095
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3645984385530405,
+      "learning_rate": 1.6318678659564046e-06,
+      "loss": 0.8068,
+      "step": 5096
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3493603353754806,
+      "learning_rate": 1.629011411962379e-06,
+      "loss": 0.8071,
+      "step": 5097
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5601298165219113,
+      "learning_rate": 1.6261572384563895e-06,
+      "loss": 0.827,
+      "step": 5098
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4175912621247566,
+      "learning_rate": 1.6233053462159965e-06,
+      "loss": 0.858,
+      "step": 5099
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4559034541473175,
+      "learning_rate": 1.6204557360181328e-06,
+      "loss": 0.7735,
+      "step": 5100
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9364697875566013,
+      "learning_rate": 1.6176084086391075e-06,
+      "loss": 0.346,
+      "step": 5101
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1845600901364681,
+      "learning_rate": 1.6147633648546157e-06,
+      "loss": 0.7923,
+      "step": 5102
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.450168938188232,
+      "learning_rate": 1.6119206054397218e-06,
+      "loss": 0.7317,
+      "step": 5103
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3708730666751399,
+      "learning_rate": 1.6090801311688764e-06,
+      "loss": 0.7384,
+      "step": 5104
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4215141437146308,
+      "learning_rate": 1.6062419428158993e-06,
+      "loss": 0.7724,
+      "step": 5105
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3292631720923311,
+      "learning_rate": 1.603406041153991e-06,
+      "loss": 0.843,
+      "step": 5106
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4004163008162296,
+      "learning_rate": 1.6005724269557322e-06,
+      "loss": 0.8301,
+      "step": 5107
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2867549262030802,
+      "learning_rate": 1.5977411009930743e-06,
+      "loss": 0.8876,
+      "step": 5108
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.644813344075044,
+      "learning_rate": 1.5949120640373517e-06,
+      "loss": 0.836,
+      "step": 5109
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2354139093641139,
+      "learning_rate": 1.5920853168592676e-06,
+      "loss": 0.7942,
+      "step": 5110
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6468215746420534,
+      "learning_rate": 1.5892608602289129e-06,
+      "loss": 0.7912,
+      "step": 5111
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.295500220522812,
+      "learning_rate": 1.5864386949157419e-06,
+      "loss": 0.8419,
+      "step": 5112
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2374374334246054,
+      "learning_rate": 1.5836188216885895e-06,
+      "loss": 0.8372,
+      "step": 5113
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2816757266810712,
+      "learning_rate": 1.5808012413156715e-06,
+      "loss": 0.7773,
+      "step": 5114
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2494049406769672,
+      "learning_rate": 1.5779859545645714e-06,
+      "loss": 0.7565,
+      "step": 5115
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6918840884151998,
+      "learning_rate": 1.5751729622022494e-06,
+      "loss": 0.8635,
+      "step": 5116
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7806019367469736,
+      "learning_rate": 1.5723622649950442e-06,
+      "loss": 0.879,
+      "step": 5117
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.288220068171052,
+      "learning_rate": 1.5695538637086693e-06,
+      "loss": 0.7994,
+      "step": 5118
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5279081227515365,
+      "learning_rate": 1.5667477591082092e-06,
+      "loss": 0.7591,
+      "step": 5119
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2645970967450206,
+      "learning_rate": 1.5639439519581212e-06,
+      "loss": 0.8308,
+      "step": 5120
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4517170394186936,
+      "learning_rate": 1.5611424430222432e-06,
+      "loss": 0.7999,
+      "step": 5121
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2419227495652463,
+      "learning_rate": 1.5583432330637826e-06,
+      "loss": 0.7769,
+      "step": 5122
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3925831321155473,
+      "learning_rate": 1.5555463228453193e-06,
+      "loss": 0.818,
+      "step": 5123
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1159012716213748,
+      "learning_rate": 1.5527517131288128e-06,
+      "loss": 0.806,
+      "step": 5124
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3369591867649087,
+      "learning_rate": 1.5499594046755862e-06,
+      "loss": 0.7984,
+      "step": 5125
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5889560273695866,
+      "learning_rate": 1.5471693982463476e-06,
+      "loss": 0.8504,
+      "step": 5126
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3704076351712626,
+      "learning_rate": 1.5443816946011658e-06,
+      "loss": 0.7968,
+      "step": 5127
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2752378607388382,
+      "learning_rate": 1.5415962944994933e-06,
+      "loss": 0.795,
+      "step": 5128
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.281199917308717,
+      "learning_rate": 1.5388131987001464e-06,
+      "loss": 0.8541,
+      "step": 5129
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.666358964922515,
+      "learning_rate": 1.5360324079613175e-06,
+      "loss": 0.7873,
+      "step": 5130
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3802410741887337,
+      "learning_rate": 1.533253923040573e-06,
+      "loss": 0.8144,
+      "step": 5131
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.531867712825312,
+      "learning_rate": 1.5304777446948448e-06,
+      "loss": 0.8894,
+      "step": 5132
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4776925475583678,
+      "learning_rate": 1.527703873680445e-06,
+      "loss": 0.6912,
+      "step": 5133
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.349230626892801,
+      "learning_rate": 1.52493231075305e-06,
+      "loss": 0.816,
+      "step": 5134
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5196554622287257,
+      "learning_rate": 1.522163056667708e-06,
+      "loss": 0.8735,
+      "step": 5135
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1917645134622263,
+      "learning_rate": 1.5193961121788448e-06,
+      "loss": 0.813,
+      "step": 5136
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3821789218130558,
+      "learning_rate": 1.5166314780402492e-06,
+      "loss": 0.7479,
+      "step": 5137
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.388687201440847,
+      "learning_rate": 1.513869155005082e-06,
+      "loss": 0.6952,
+      "step": 5138
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6593997196468362,
+      "learning_rate": 1.5111091438258796e-06,
+      "loss": 0.7562,
+      "step": 5139
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3286637013102585,
+      "learning_rate": 1.5083514452545488e-06,
+      "loss": 0.7483,
+      "step": 5140
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.472185802193547,
+      "learning_rate": 1.5055960600423524e-06,
+      "loss": 0.7627,
+      "step": 5141
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4975917477913243,
+      "learning_rate": 1.5028429889399388e-06,
+      "loss": 0.7848,
+      "step": 5142
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3177877331234338,
+      "learning_rate": 1.5000922326973233e-06,
+      "loss": 0.7978,
+      "step": 5143
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3262964942932212,
+      "learning_rate": 1.4973437920638856e-06,
+      "loss": 0.82,
+      "step": 5144
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3916283748787606,
+      "learning_rate": 1.4945976677883723e-06,
+      "loss": 0.8408,
+      "step": 5145
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8167666399544077,
+      "learning_rate": 1.4918538606189104e-06,
+      "loss": 0.3028,
+      "step": 5146
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.466419347263555,
+      "learning_rate": 1.4891123713029832e-06,
+      "loss": 0.8332,
+      "step": 5147
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5608451639345016,
+      "learning_rate": 1.4863732005874509e-06,
+      "loss": 0.8029,
+      "step": 5148
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.483068594634602,
+      "learning_rate": 1.483636349218538e-06,
+      "loss": 0.859,
+      "step": 5149
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3787146139980786,
+      "learning_rate": 1.4809018179418388e-06,
+      "loss": 0.8477,
+      "step": 5150
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5476422742995382,
+      "learning_rate": 1.4781696075023156e-06,
+      "loss": 0.8135,
+      "step": 5151
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3864963484112915,
+      "learning_rate": 1.4754397186442947e-06,
+      "loss": 0.8719,
+      "step": 5152
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.231744563492654,
+      "learning_rate": 1.4727121521114784e-06,
+      "loss": 0.8273,
+      "step": 5153
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4045198380272634,
+      "learning_rate": 1.4699869086469242e-06,
+      "loss": 0.8591,
+      "step": 5154
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2808932456086721,
+      "learning_rate": 1.4672639889930707e-06,
+      "loss": 0.8424,
+      "step": 5155
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3762489188730964,
+      "learning_rate": 1.464543393891712e-06,
+      "loss": 0.8835,
+      "step": 5156
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3941458248266982,
+      "learning_rate": 1.461825124084012e-06,
+      "loss": 0.8291,
+      "step": 5157
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3477287729071803,
+      "learning_rate": 1.4591091803105072e-06,
+      "loss": 0.9062,
+      "step": 5158
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.88062324507363,
+      "learning_rate": 1.4563955633110926e-06,
+      "loss": 0.8585,
+      "step": 5159
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3350133669309923,
+      "learning_rate": 1.453684273825029e-06,
+      "loss": 0.7884,
+      "step": 5160
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4369804795006753,
+      "learning_rate": 1.450975312590951e-06,
+      "loss": 0.8529,
+      "step": 5161
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2923006156700652,
+      "learning_rate": 1.448268680346857e-06,
+      "loss": 0.807,
+      "step": 5162
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9007617646845156,
+      "learning_rate": 1.4455643778301e-06,
+      "loss": 0.344,
+      "step": 5163
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1399077465316996,
+      "learning_rate": 1.442862405777411e-06,
+      "loss": 0.8053,
+      "step": 5164
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0635498720443177,
+      "learning_rate": 1.440162764924884e-06,
+      "loss": 0.343,
+      "step": 5165
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5914569245998573,
+      "learning_rate": 1.4374654560079725e-06,
+      "loss": 0.739,
+      "step": 5166
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5971842443977484,
+      "learning_rate": 1.434770479761497e-06,
+      "loss": 0.7933,
+      "step": 5167
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2015511204815645,
+      "learning_rate": 1.4320778369196443e-06,
+      "loss": 0.8044,
+      "step": 5168
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4362604930112997,
+      "learning_rate": 1.4293875282159698e-06,
+      "loss": 0.8369,
+      "step": 5169
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8827222244134763,
+      "learning_rate": 1.4266995543833772e-06,
+      "loss": 0.3448,
+      "step": 5170
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3814882632983223,
+      "learning_rate": 1.424013916154151e-06,
+      "loss": 0.8432,
+      "step": 5171
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4217982598964616,
+      "learning_rate": 1.421330614259936e-06,
+      "loss": 0.8473,
+      "step": 5172
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3341171357763792,
+      "learning_rate": 1.4186496494317325e-06,
+      "loss": 0.6961,
+      "step": 5173
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.285275248696766,
+      "learning_rate": 1.415971022399909e-06,
+      "loss": 0.7634,
+      "step": 5174
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5157854453678503,
+      "learning_rate": 1.4132947338942016e-06,
+      "loss": 0.8047,
+      "step": 5175
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4247176449015786,
+      "learning_rate": 1.4106207846437003e-06,
+      "loss": 0.8821,
+      "step": 5176
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.653967286890549,
+      "learning_rate": 1.407949175376867e-06,
+      "loss": 0.802,
+      "step": 5177
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4985355987498457,
+      "learning_rate": 1.4052799068215206e-06,
+      "loss": 0.8427,
+      "step": 5178
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3427708324060312,
+      "learning_rate": 1.4026129797048393e-06,
+      "loss": 0.8274,
+      "step": 5179
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.280001738428238,
+      "learning_rate": 1.399948394753372e-06,
+      "loss": 0.7792,
+      "step": 5180
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.374135420193684,
+      "learning_rate": 1.3972861526930249e-06,
+      "loss": 0.8397,
+      "step": 5181
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2226981416356757,
+      "learning_rate": 1.394626254249063e-06,
+      "loss": 0.7853,
+      "step": 5182
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4674074778926702,
+      "learning_rate": 1.391968700146118e-06,
+      "loss": 0.8222,
+      "step": 5183
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5871191672779557,
+      "learning_rate": 1.389313491108184e-06,
+      "loss": 0.8152,
+      "step": 5184
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3496408327745233,
+      "learning_rate": 1.386660627858607e-06,
+      "loss": 0.7802,
+      "step": 5185
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3111024366906372,
+      "learning_rate": 1.384010111120102e-06,
+      "loss": 0.8052,
+      "step": 5186
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1420869260492195,
+      "learning_rate": 1.3813619416147472e-06,
+      "loss": 0.7785,
+      "step": 5187
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.283622370561614,
+      "learning_rate": 1.378716120063973e-06,
+      "loss": 0.8175,
+      "step": 5188
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5492890022227461,
+      "learning_rate": 1.3760726471885722e-06,
+      "loss": 0.8577,
+      "step": 5189
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2629827170830878,
+      "learning_rate": 1.3734315237087027e-06,
+      "loss": 0.8756,
+      "step": 5190
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3352656041201154,
+      "learning_rate": 1.3707927503438833e-06,
+      "loss": 0.8213,
+      "step": 5191
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4744377885865863,
+      "learning_rate": 1.3681563278129794e-06,
+      "loss": 0.8156,
+      "step": 5192
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.261076051561022,
+      "learning_rate": 1.3655222568342308e-06,
+      "loss": 0.8368,
+      "step": 5193
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3144799466393677,
+      "learning_rate": 1.3628905381252322e-06,
+      "loss": 0.7739,
+      "step": 5194
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6395217552344568,
+      "learning_rate": 1.3602611724029335e-06,
+      "loss": 0.8914,
+      "step": 5195
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4516780722571418,
+      "learning_rate": 1.3576341603836462e-06,
+      "loss": 0.8117,
+      "step": 5196
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7639226824773393,
+      "learning_rate": 1.3550095027830435e-06,
+      "loss": 0.327,
+      "step": 5197
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.405972291095598,
+      "learning_rate": 1.352387200316152e-06,
+      "loss": 0.796,
+      "step": 5198
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3665672772124704,
+      "learning_rate": 1.3497672536973594e-06,
+      "loss": 0.8,
+      "step": 5199
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5911074262771294,
+      "learning_rate": 1.3471496636404124e-06,
+      "loss": 0.8445,
+      "step": 5200
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.355154899160837,
+      "learning_rate": 1.344534430858413e-06,
+      "loss": 0.7902,
+      "step": 5201
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8715510862397434,
+      "learning_rate": 1.3419215560638265e-06,
+      "loss": 0.3488,
+      "step": 5202
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.361774718862833,
+      "learning_rate": 1.3393110399684695e-06,
+      "loss": 0.8451,
+      "step": 5203
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.559890431159658,
+      "learning_rate": 1.3367028832835149e-06,
+      "loss": 0.8582,
+      "step": 5204
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2309376073119616,
+      "learning_rate": 1.3340970867195014e-06,
+      "loss": 0.7595,
+      "step": 5205
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.886840777863905,
+      "learning_rate": 1.331493650986323e-06,
+      "loss": 0.8158,
+      "step": 5206
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.292222095543013,
+      "learning_rate": 1.328892576793217e-06,
+      "loss": 0.801,
+      "step": 5207
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3361651773133782,
+      "learning_rate": 1.3262938648487955e-06,
+      "loss": 0.7986,
+      "step": 5208
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3923923153124236,
+      "learning_rate": 1.3236975158610178e-06,
+      "loss": 0.7587,
+      "step": 5209
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4426340641060156,
+      "learning_rate": 1.321103530537201e-06,
+      "loss": 0.8023,
+      "step": 5210
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5681359879012584,
+      "learning_rate": 1.318511909584016e-06,
+      "loss": 0.8019,
+      "step": 5211
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3857614171103256,
+      "learning_rate": 1.3159226537074933e-06,
+      "loss": 0.7778,
+      "step": 5212
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4395662483914837,
+      "learning_rate": 1.3133357636130217e-06,
+      "loss": 0.8223,
+      "step": 5213
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3313365261463554,
+      "learning_rate": 1.3107512400053335e-06,
+      "loss": 0.8549,
+      "step": 5214
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5227750878618302,
+      "learning_rate": 1.3081690835885274e-06,
+      "loss": 0.8435,
+      "step": 5215
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3533725628515563,
+      "learning_rate": 1.3055892950660576e-06,
+      "loss": 0.7815,
+      "step": 5216
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.473533881110601,
+      "learning_rate": 1.303011875140726e-06,
+      "loss": 0.7739,
+      "step": 5217
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3354294731870167,
+      "learning_rate": 1.3004368245146913e-06,
+      "loss": 0.8477,
+      "step": 5218
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3491595773941265,
+      "learning_rate": 1.2978641438894735e-06,
+      "loss": 0.7942,
+      "step": 5219
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9494856539621644,
+      "learning_rate": 1.2952938339659382e-06,
+      "loss": 0.3366,
+      "step": 5220
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7534712854780335,
+      "learning_rate": 1.2927258954443066e-06,
+      "loss": 0.3279,
+      "step": 5221
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4389656767180963,
+      "learning_rate": 1.2901603290241615e-06,
+      "loss": 0.8152,
+      "step": 5222
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1974218423745941,
+      "learning_rate": 1.287597135404428e-06,
+      "loss": 0.8119,
+      "step": 5223
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3273289107250414,
+      "learning_rate": 1.2850363152833955e-06,
+      "loss": 0.8064,
+      "step": 5224
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5597652368428019,
+      "learning_rate": 1.2824778693587014e-06,
+      "loss": 0.8237,
+      "step": 5225
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2511474636321367,
+      "learning_rate": 1.2799217983273326e-06,
+      "loss": 0.7949,
+      "step": 5226
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5563569857503288,
+      "learning_rate": 1.2773681028856366e-06,
+      "loss": 0.834,
+      "step": 5227
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4451860262976652,
+      "learning_rate": 1.2748167837293113e-06,
+      "loss": 0.7915,
+      "step": 5228
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2503151693729173,
+      "learning_rate": 1.2722678415534062e-06,
+      "loss": 0.823,
+      "step": 5229
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4464895649053773,
+      "learning_rate": 1.2697212770523205e-06,
+      "loss": 0.773,
+      "step": 5230
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5416266960657676,
+      "learning_rate": 1.2671770909198122e-06,
+      "loss": 0.836,
+      "step": 5231
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4724177842876036,
+      "learning_rate": 1.264635283848985e-06,
+      "loss": 0.8444,
+      "step": 5232
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.582534792319502,
+      "learning_rate": 1.262095856532297e-06,
+      "loss": 0.8361,
+      "step": 5233
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4605725394792384,
+      "learning_rate": 1.2595588096615596e-06,
+      "loss": 0.8503,
+      "step": 5234
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4133550787009541,
+      "learning_rate": 1.2570241439279386e-06,
+      "loss": 0.8771,
+      "step": 5235
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3808401442325793,
+      "learning_rate": 1.254491860021938e-06,
+      "loss": 0.8312,
+      "step": 5236
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4008230940828748,
+      "learning_rate": 1.2519619586334253e-06,
+      "loss": 0.8276,
+      "step": 5237
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5126666613648918,
+      "learning_rate": 1.249434440451619e-06,
+      "loss": 0.7822,
+      "step": 5238
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.359159874953255,
+      "learning_rate": 1.2469093061650816e-06,
+      "loss": 0.8306,
+      "step": 5239
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.188031279306702,
+      "learning_rate": 1.2443865564617274e-06,
+      "loss": 0.7191,
+      "step": 5240
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2935344036400596,
+      "learning_rate": 1.2418661920288278e-06,
+      "loss": 0.8145,
+      "step": 5241
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.465430177924679,
+      "learning_rate": 1.2393482135529954e-06,
+      "loss": 0.7872,
+      "step": 5242
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6050514319236442,
+      "learning_rate": 1.2368326217201976e-06,
+      "loss": 0.7719,
+      "step": 5243
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3161406362257608,
+      "learning_rate": 1.2343194172157535e-06,
+      "loss": 0.7653,
+      "step": 5244
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4057114056497189,
+      "learning_rate": 1.2318086007243257e-06,
+      "loss": 0.7967,
+      "step": 5245
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.355464617863054,
+      "learning_rate": 1.2293001729299336e-06,
+      "loss": 0.8514,
+      "step": 5246
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3290198530128567,
+      "learning_rate": 1.2267941345159385e-06,
+      "loss": 0.903,
+      "step": 5247
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4605060025587326,
+      "learning_rate": 1.2242904861650574e-06,
+      "loss": 0.8178,
+      "step": 5248
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5257403629731268,
+      "learning_rate": 1.2217892285593513e-06,
+      "loss": 0.8219,
+      "step": 5249
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4309607982488288,
+      "learning_rate": 1.219290362380231e-06,
+      "loss": 0.816,
+      "step": 5250
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.379907772489096,
+      "learning_rate": 1.2167938883084595e-06,
+      "loss": 0.857,
+      "step": 5251
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3503707243987104,
+      "learning_rate": 1.2142998070241407e-06,
+      "loss": 0.7346,
+      "step": 5252
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2630054258278525,
+      "learning_rate": 1.2118081192067365e-06,
+      "loss": 0.8236,
+      "step": 5253
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.412888702304788,
+      "learning_rate": 1.2093188255350485e-06,
+      "loss": 0.8277,
+      "step": 5254
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4827971391816805,
+      "learning_rate": 1.2068319266872264e-06,
+      "loss": 0.7889,
+      "step": 5255
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4803430483756137,
+      "learning_rate": 1.2043474233407737e-06,
+      "loss": 0.8284,
+      "step": 5256
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.269311023464135,
+      "learning_rate": 1.2018653161725392e-06,
+      "loss": 0.7998,
+      "step": 5257
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5161565531822676,
+      "learning_rate": 1.1993856058587117e-06,
+      "loss": 0.8194,
+      "step": 5258
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3442780647806312,
+      "learning_rate": 1.1969082930748343e-06,
+      "loss": 0.8079,
+      "step": 5259
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6027828190404312,
+      "learning_rate": 1.1944333784957996e-06,
+      "loss": 0.8415,
+      "step": 5260
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7932396540712283,
+      "learning_rate": 1.191960862795839e-06,
+      "loss": 0.7442,
+      "step": 5261
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4315670470706994,
+      "learning_rate": 1.1894907466485317e-06,
+      "loss": 0.7775,
+      "step": 5262
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4226688111864627,
+      "learning_rate": 1.1870230307268116e-06,
+      "loss": 0.8064,
+      "step": 5263
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2801781861889887,
+      "learning_rate": 1.1845577157029474e-06,
+      "loss": 0.7717,
+      "step": 5264
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3965523743246373,
+      "learning_rate": 1.1820948022485602e-06,
+      "loss": 0.8608,
+      "step": 5265
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4052247880012798,
+      "learning_rate": 1.1796342910346147e-06,
+      "loss": 0.7928,
+      "step": 5266
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7039930762705902,
+      "learning_rate": 1.1771761827314254e-06,
+      "loss": 0.766,
+      "step": 5267
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2651174905273497,
+      "learning_rate": 1.1747204780086462e-06,
+      "loss": 0.7832,
+      "step": 5268
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4302796924854877,
+      "learning_rate": 1.1722671775352778e-06,
+      "loss": 0.8402,
+      "step": 5269
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.518725516225925,
+      "learning_rate": 1.16981628197967e-06,
+      "loss": 0.7862,
+      "step": 5270
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.296674607346402,
+      "learning_rate": 1.1673677920095116e-06,
+      "loss": 0.7981,
+      "step": 5271
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.531081827815868,
+      "learning_rate": 1.1649217082918385e-06,
+      "loss": 0.8326,
+      "step": 5272
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.511847252768553,
+      "learning_rate": 1.1624780314930339e-06,
+      "loss": 0.8506,
+      "step": 5273
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.550381314621408,
+      "learning_rate": 1.16003676227882e-06,
+      "loss": 0.8099,
+      "step": 5274
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2903987273412068,
+      "learning_rate": 1.157597901314268e-06,
+      "loss": 0.8061,
+      "step": 5275
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4829916038290576,
+      "learning_rate": 1.1551614492637908e-06,
+      "loss": 0.8013,
+      "step": 5276
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3837284352139283,
+      "learning_rate": 1.152727406791142e-06,
+      "loss": 0.7984,
+      "step": 5277
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2642908157715427,
+      "learning_rate": 1.1502957745594256e-06,
+      "loss": 0.8012,
+      "step": 5278
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2597127367347019,
+      "learning_rate": 1.1478665532310839e-06,
+      "loss": 0.8059,
+      "step": 5279
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4216110900259018,
+      "learning_rate": 1.1454397434679022e-06,
+      "loss": 0.7939,
+      "step": 5280
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4507115309062562,
+      "learning_rate": 1.1430153459310112e-06,
+      "loss": 0.8835,
+      "step": 5281
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3149758811423164,
+      "learning_rate": 1.1405933612808862e-06,
+      "loss": 0.7915,
+      "step": 5282
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2809584063375505,
+      "learning_rate": 1.1381737901773405e-06,
+      "loss": 0.7705,
+      "step": 5283
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8853702626503501,
+      "learning_rate": 1.13575663327953e-06,
+      "loss": 0.3126,
+      "step": 5284
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5293814597969675,
+      "learning_rate": 1.1333418912459593e-06,
+      "loss": 0.8855,
+      "step": 5285
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4713699841422747,
+      "learning_rate": 1.1309295647344675e-06,
+      "loss": 0.8459,
+      "step": 5286
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4923331579048758,
+      "learning_rate": 1.1285196544022392e-06,
+      "loss": 0.9115,
+      "step": 5287
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5148807411438387,
+      "learning_rate": 1.126112160905799e-06,
+      "loss": 0.854,
+      "step": 5288
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4175981087772667,
+      "learning_rate": 1.12370708490102e-06,
+      "loss": 0.8184,
+      "step": 5289
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.319453889951722,
+      "learning_rate": 1.1213044270431062e-06,
+      "loss": 0.7745,
+      "step": 5290
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3227461926955983,
+      "learning_rate": 1.1189041879866081e-06,
+      "loss": 0.8795,
+      "step": 5291
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4965099466708551,
+      "learning_rate": 1.1165063683854193e-06,
+      "loss": 0.8549,
+      "step": 5292
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3253127729833423,
+      "learning_rate": 1.1141109688927709e-06,
+      "loss": 0.7515,
+      "step": 5293
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2909171698824828,
+      "learning_rate": 1.1117179901612328e-06,
+      "loss": 0.7636,
+      "step": 5294
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.35830437185751,
+      "learning_rate": 1.109327432842725e-06,
+      "loss": 0.8658,
+      "step": 5295
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4579160007306255,
+      "learning_rate": 1.106939297588494e-06,
+      "loss": 0.8115,
+      "step": 5296
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1547597410538475,
+      "learning_rate": 1.1045535850491396e-06,
+      "loss": 0.726,
+      "step": 5297
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4124817785843204,
+      "learning_rate": 1.1021702958745917e-06,
+      "loss": 0.7804,
+      "step": 5298
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5272846100716195,
+      "learning_rate": 1.0997894307141244e-06,
+      "loss": 0.7293,
+      "step": 5299
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.494225150566031,
+      "learning_rate": 1.0974109902163544e-06,
+      "loss": 0.8458,
+      "step": 5300
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4533799573309003,
+      "learning_rate": 1.0950349750292311e-06,
+      "loss": 0.8197,
+      "step": 5301
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2727239809694235,
+      "learning_rate": 1.0926613858000456e-06,
+      "loss": 0.8111,
+      "step": 5302
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6093514941869236,
+      "learning_rate": 1.0902902231754309e-06,
+      "loss": 0.7813,
+      "step": 5303
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3827534908161128,
+      "learning_rate": 1.087921487801359e-06,
+      "loss": 0.7859,
+      "step": 5304
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6754433217814297,
+      "learning_rate": 1.0855551803231368e-06,
+      "loss": 0.8533,
+      "step": 5305
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6292151625882672,
+      "learning_rate": 1.0831913013854101e-06,
+      "loss": 0.7774,
+      "step": 5306
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5114736136468767,
+      "learning_rate": 1.080829851632167e-06,
+      "loss": 0.9358,
+      "step": 5307
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.34756071125226,
+      "learning_rate": 1.0784708317067316e-06,
+      "loss": 0.7975,
+      "step": 5308
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4740206983406128,
+      "learning_rate": 1.0761142422517623e-06,
+      "loss": 0.8691,
+      "step": 5309
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8178306808735657,
+      "learning_rate": 1.073760083909262e-06,
+      "loss": 0.3266,
+      "step": 5310
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3789162554391414,
+      "learning_rate": 1.0714083573205702e-06,
+      "loss": 0.8508,
+      "step": 5311
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6046089656912332,
+      "learning_rate": 1.069059063126361e-06,
+      "loss": 0.8279,
+      "step": 5312
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3729483486329883,
+      "learning_rate": 1.066712201966642e-06,
+      "loss": 0.7292,
+      "step": 5313
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.554574565164168,
+      "learning_rate": 1.06436777448077e-06,
+      "loss": 0.8377,
+      "step": 5314
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2737446981961735,
+      "learning_rate": 1.0620257813074274e-06,
+      "loss": 0.7761,
+      "step": 5315
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3167963072651943,
+      "learning_rate": 1.0596862230846371e-06,
+      "loss": 0.7619,
+      "step": 5316
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3108927647644004,
+      "learning_rate": 1.0573491004497637e-06,
+      "loss": 0.7994,
+      "step": 5317
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6003325438068274,
+      "learning_rate": 1.055014414039498e-06,
+      "loss": 0.7997,
+      "step": 5318
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5489227270840964,
+      "learning_rate": 1.0526821644898777e-06,
+      "loss": 0.8034,
+      "step": 5319
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5553913907194774,
+      "learning_rate": 1.0503523524362701e-06,
+      "loss": 0.811,
+      "step": 5320
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2811553664992879,
+      "learning_rate": 1.048024978513379e-06,
+      "loss": 0.8333,
+      "step": 5321
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4998375288075945,
+      "learning_rate": 1.0457000433552478e-06,
+      "loss": 0.8023,
+      "step": 5322
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.36212382944794,
+      "learning_rate": 1.0433775475952511e-06,
+      "loss": 0.8567,
+      "step": 5323
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.593792342109697,
+      "learning_rate": 1.0410574918660998e-06,
+      "loss": 0.8977,
+      "step": 5324
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3122936523672994,
+      "learning_rate": 1.0387398767998425e-06,
+      "loss": 0.8681,
+      "step": 5325
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5096601261818865,
+      "learning_rate": 1.036424703027863e-06,
+      "loss": 0.7307,
+      "step": 5326
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.345246935979127,
+      "learning_rate": 1.0341119711808778e-06,
+      "loss": 0.8123,
+      "step": 5327
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2909270153714156,
+      "learning_rate": 1.0318016818889343e-06,
+      "loss": 0.8591,
+      "step": 5328
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4654301830089067,
+      "learning_rate": 1.0294938357814254e-06,
+      "loss": 0.844,
+      "step": 5329
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4175686455664849,
+      "learning_rate": 1.0271884334870685e-06,
+      "loss": 0.8032,
+      "step": 5330
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3158467742401752,
+      "learning_rate": 1.0248854756339176e-06,
+      "loss": 0.8303,
+      "step": 5331
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6588025958390515,
+      "learning_rate": 1.0225849628493634e-06,
+      "loss": 0.8009,
+      "step": 5332
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4908201108952597,
+      "learning_rate": 1.020286895760132e-06,
+      "loss": 0.7647,
+      "step": 5333
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4396252186359273,
+      "learning_rate": 1.0179912749922772e-06,
+      "loss": 0.8503,
+      "step": 5334
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4358349329447966,
+      "learning_rate": 1.0156981011711875e-06,
+      "loss": 0.8255,
+      "step": 5335
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4040875955391896,
+      "learning_rate": 1.01340737492159e-06,
+      "loss": 0.8663,
+      "step": 5336
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.318424795434939,
+      "learning_rate": 1.011119096867541e-06,
+      "loss": 0.8305,
+      "step": 5337
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4382492133089426,
+      "learning_rate": 1.0088332676324285e-06,
+      "loss": 0.7864,
+      "step": 5338
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3338805922075627,
+      "learning_rate": 1.006549887838978e-06,
+      "loss": 0.8641,
+      "step": 5339
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3725193076412905,
+      "learning_rate": 1.0042689581092424e-06,
+      "loss": 0.8306,
+      "step": 5340
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5409036415565949,
+      "learning_rate": 1.001990479064613e-06,
+      "loss": 0.8476,
+      "step": 5341
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2461600480335424,
+      "learning_rate": 9.997144513258095e-07,
+      "loss": 0.8206,
+      "step": 5342
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2934695733205006,
+      "learning_rate": 9.974408755128817e-07,
+      "loss": 0.8108,
+      "step": 5343
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3059599262291892,
+      "learning_rate": 9.951697522452175e-07,
+      "loss": 0.8188,
+      "step": 5344
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5349864871918681,
+      "learning_rate": 9.92901082141531e-07,
+      "loss": 0.7302,
+      "step": 5345
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3265620431573484,
+      "learning_rate": 9.90634865819874e-07,
+      "loss": 0.8108,
+      "step": 5346
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3310033028438562,
+      "learning_rate": 9.883711038976218e-07,
+      "loss": 0.7748,
+      "step": 5347
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3102875927753372,
+      "learning_rate": 9.861097969914901e-07,
+      "loss": 0.8176,
+      "step": 5348
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2864623274013998,
+      "learning_rate": 9.838509457175183e-07,
+      "loss": 0.7972,
+      "step": 5349
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8661303435429006,
+      "learning_rate": 9.815945506910795e-07,
+      "loss": 0.7891,
+      "step": 5350
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2877256765941425,
+      "learning_rate": 9.793406125268801e-07,
+      "loss": 0.8037,
+      "step": 5351
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2962416343179906,
+      "learning_rate": 9.770891318389542e-07,
+      "loss": 0.8638,
+      "step": 5352
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5687350990529993,
+      "learning_rate": 9.748401092406657e-07,
+      "loss": 0.8161,
+      "step": 5353
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4920707477211714,
+      "learning_rate": 9.7259354534471e-07,
+      "loss": 0.9224,
+      "step": 5354
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2192861824876913,
+      "learning_rate": 9.703494407631176e-07,
+      "loss": 0.7941,
+      "step": 5355
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5672857248534733,
+      "learning_rate": 9.681077961072405e-07,
+      "loss": 0.8246,
+      "step": 5356
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5497155407251213,
+      "learning_rate": 9.658686119877636e-07,
+      "loss": 0.8071,
+      "step": 5357
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3386746656980388,
+      "learning_rate": 9.636318890147057e-07,
+      "loss": 0.7863,
+      "step": 5358
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5672472814984217,
+      "learning_rate": 9.613976277974101e-07,
+      "loss": 0.8036,
+      "step": 5359
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5231821647045571,
+      "learning_rate": 9.591658289445504e-07,
+      "loss": 0.7557,
+      "step": 5360
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3801120883445477,
+      "learning_rate": 9.569364930641323e-07,
+      "loss": 0.8045,
+      "step": 5361
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8228643643179944,
+      "learning_rate": 9.547096207634843e-07,
+      "loss": 0.3415,
+      "step": 5362
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3445452212561466,
+      "learning_rate": 9.524852126492734e-07,
+      "loss": 0.8334,
+      "step": 5363
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.235738286638803,
+      "learning_rate": 9.50263269327486e-07,
+      "loss": 0.7663,
+      "step": 5364
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4175873432611288,
+      "learning_rate": 9.480437914034402e-07,
+      "loss": 0.8618,
+      "step": 5365
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3902692460298143,
+      "learning_rate": 9.458267794817866e-07,
+      "loss": 0.8039,
+      "step": 5366
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.578026350639442,
+      "learning_rate": 9.436122341664955e-07,
+      "loss": 0.7876,
+      "step": 5367
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.654528290912296,
+      "learning_rate": 9.414001560608743e-07,
+      "loss": 0.8799,
+      "step": 5368
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4148816579935815,
+      "learning_rate": 9.391905457675499e-07,
+      "loss": 0.8536,
+      "step": 5369
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.496873148969768,
+      "learning_rate": 9.369834038884862e-07,
+      "loss": 0.8801,
+      "step": 5370
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4258834802217042,
+      "learning_rate": 9.347787310249668e-07,
+      "loss": 0.8301,
+      "step": 5371
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4403554793817133,
+      "learning_rate": 9.325765277776033e-07,
+      "loss": 0.7295,
+      "step": 5372
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2911688797344099,
+      "learning_rate": 9.303767947463416e-07,
+      "loss": 0.8392,
+      "step": 5373
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.323787987675487,
+      "learning_rate": 9.281795325304454e-07,
+      "loss": 0.757,
+      "step": 5374
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2607834851321609,
+      "learning_rate": 9.259847417285084e-07,
+      "loss": 0.7858,
+      "step": 5375
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4108271330432955,
+      "learning_rate": 9.237924229384554e-07,
+      "loss": 0.78,
+      "step": 5376
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7111532192887853,
+      "learning_rate": 9.216025767575376e-07,
+      "loss": 0.7721,
+      "step": 5377
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2593002953273762,
+      "learning_rate": 9.194152037823211e-07,
+      "loss": 0.8091,
+      "step": 5378
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4634044404422597,
+      "learning_rate": 9.172303046087105e-07,
+      "loss": 0.768,
+      "step": 5379
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4037826170798435,
+      "learning_rate": 9.150478798319351e-07,
+      "loss": 0.8737,
+      "step": 5380
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3267436961216668,
+      "learning_rate": 9.128679300465459e-07,
+      "loss": 0.8347,
+      "step": 5381
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2077387885650863,
+      "learning_rate": 9.10690455846418e-07,
+      "loss": 0.796,
+      "step": 5382
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3934984824027679,
+      "learning_rate": 9.085154578247613e-07,
+      "loss": 0.8502,
+      "step": 5383
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.62194190989826,
+      "learning_rate": 9.063429365740995e-07,
+      "loss": 0.8264,
+      "step": 5384
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.585135511983675,
+      "learning_rate": 9.041728926862914e-07,
+      "loss": 0.7999,
+      "step": 5385
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8390152753872244,
+      "learning_rate": 9.020053267525142e-07,
+      "loss": 0.3605,
+      "step": 5386
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4046042792933673,
+      "learning_rate": 8.998402393632755e-07,
+      "loss": 0.8167,
+      "step": 5387
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1615341329555229,
+      "learning_rate": 8.976776311084024e-07,
+      "loss": 0.8176,
+      "step": 5388
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4072841530583289,
+      "learning_rate": 8.95517502577048e-07,
+      "loss": 0.8128,
+      "step": 5389
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3823471458201901,
+      "learning_rate": 8.933598543576938e-07,
+      "loss": 0.8239,
+      "step": 5390
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.499590611719277,
+      "learning_rate": 8.912046870381397e-07,
+      "loss": 0.725,
+      "step": 5391
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5252665177385483,
+      "learning_rate": 8.89052001205517e-07,
+      "loss": 0.7988,
+      "step": 5392
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3328610950814526,
+      "learning_rate": 8.869017974462735e-07,
+      "loss": 0.8139,
+      "step": 5393
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3309228437173843,
+      "learning_rate": 8.847540763461815e-07,
+      "loss": 0.8512,
+      "step": 5394
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6036483440232907,
+      "learning_rate": 8.826088384903453e-07,
+      "loss": 0.7421,
+      "step": 5395
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9962180854763197,
+      "learning_rate": 8.804660844631841e-07,
+      "loss": 0.3408,
+      "step": 5396
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.516697891600251,
+      "learning_rate": 8.783258148484397e-07,
+      "loss": 0.8111,
+      "step": 5397
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4205706700086675,
+      "learning_rate": 8.761880302291847e-07,
+      "loss": 0.8041,
+      "step": 5398
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2827767156454926,
+      "learning_rate": 8.740527311878133e-07,
+      "loss": 0.8701,
+      "step": 5399
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1579982019837285,
+      "learning_rate": 8.719199183060323e-07,
+      "loss": 0.7499,
+      "step": 5400
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.351868499723918,
+      "learning_rate": 8.697895921648824e-07,
+      "loss": 0.7565,
+      "step": 5401
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6339733190319774,
+      "learning_rate": 8.676617533447251e-07,
+      "loss": 0.8408,
+      "step": 5402
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.567736801982584,
+      "learning_rate": 8.655364024252411e-07,
+      "loss": 0.7769,
+      "step": 5403
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.50482176372812,
+      "learning_rate": 8.634135399854315e-07,
+      "loss": 0.8301,
+      "step": 5404
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.423011554789644,
+      "learning_rate": 8.612931666036262e-07,
+      "loss": 0.778,
+      "step": 5405
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3759096740562957,
+      "learning_rate": 8.59175282857475e-07,
+      "loss": 0.8251,
+      "step": 5406
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3709497749780362,
+      "learning_rate": 8.570598893239413e-07,
+      "loss": 0.8121,
+      "step": 5407
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.571229915797805,
+      "learning_rate": 8.549469865793214e-07,
+      "loss": 0.7938,
+      "step": 5408
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3566478525853967,
+      "learning_rate": 8.528365751992284e-07,
+      "loss": 0.8598,
+      "step": 5409
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.300280629298944,
+      "learning_rate": 8.507286557585948e-07,
+      "loss": 0.7454,
+      "step": 5410
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5006524631441105,
+      "learning_rate": 8.486232288316754e-07,
+      "loss": 0.7674,
+      "step": 5411
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3890595451619632,
+      "learning_rate": 8.465202949920492e-07,
+      "loss": 0.7835,
+      "step": 5412
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.42601130145009,
+      "learning_rate": 8.444198548126103e-07,
+      "loss": 0.84,
+      "step": 5413
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4124990014108572,
+      "learning_rate": 8.423219088655788e-07,
+      "loss": 0.8597,
+      "step": 5414
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.261855303494956,
+      "learning_rate": 8.402264577224928e-07,
+      "loss": 0.8045,
+      "step": 5415
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5508183587772695,
+      "learning_rate": 8.381335019542091e-07,
+      "loss": 0.8461,
+      "step": 5416
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3596466658976296,
+      "learning_rate": 8.360430421309096e-07,
+      "loss": 0.8641,
+      "step": 5417
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6317987363582516,
+      "learning_rate": 8.339550788220907e-07,
+      "loss": 0.8218,
+      "step": 5418
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4675176959863563,
+      "learning_rate": 8.318696125965698e-07,
+      "loss": 0.7307,
+      "step": 5419
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8528473305496553,
+      "learning_rate": 8.297866440224889e-07,
+      "loss": 0.7819,
+      "step": 5420
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.505922752808056,
+      "learning_rate": 8.277061736673064e-07,
+      "loss": 0.8035,
+      "step": 5421
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3763313161030533,
+      "learning_rate": 8.256282020977957e-07,
+      "loss": 0.7318,
+      "step": 5422
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3138751938883437,
+      "learning_rate": 8.235527298800549e-07,
+      "loss": 0.7871,
+      "step": 5423
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5854933749888387,
+      "learning_rate": 8.21479757579503e-07,
+      "loss": 0.7521,
+      "step": 5424
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4420092451811641,
+      "learning_rate": 8.194092857608726e-07,
+      "loss": 0.7972,
+      "step": 5425
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2656269198567935,
+      "learning_rate": 8.173413149882147e-07,
+      "loss": 0.8772,
+      "step": 5426
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.236147276742048,
+      "learning_rate": 8.152758458249055e-07,
+      "loss": 0.7712,
+      "step": 5427
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4593082021368307,
+      "learning_rate": 8.132128788336368e-07,
+      "loss": 0.7778,
+      "step": 5428
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.0272526794073764,
+      "learning_rate": 8.111524145764116e-07,
+      "loss": 0.7643,
+      "step": 5429
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2789420403807168,
+      "learning_rate": 8.090944536145606e-07,
+      "loss": 0.8856,
+      "step": 5430
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3897070356306853,
+      "learning_rate": 8.070389965087311e-07,
+      "loss": 0.7865,
+      "step": 5431
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4157138614732654,
+      "learning_rate": 8.04986043818885e-07,
+      "loss": 0.8944,
+      "step": 5432
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8019803837538457,
+      "learning_rate": 8.029355961043006e-07,
+      "loss": 0.8657,
+      "step": 5433
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5378706681514454,
+      "learning_rate": 8.008876539235799e-07,
+      "loss": 0.7841,
+      "step": 5434
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8344634631822374,
+      "learning_rate": 7.988422178346378e-07,
+      "loss": 0.7618,
+      "step": 5435
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2374102484211469,
+      "learning_rate": 7.967992883947051e-07,
+      "loss": 0.7747,
+      "step": 5436
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2646487026140678,
+      "learning_rate": 7.947588661603345e-07,
+      "loss": 0.846,
+      "step": 5437
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4172264736204214,
+      "learning_rate": 7.927209516873924e-07,
+      "loss": 0.7937,
+      "step": 5438
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.401876491281377,
+      "learning_rate": 7.906855455310647e-07,
+      "loss": 0.762,
+      "step": 5439
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3406368138386309,
+      "learning_rate": 7.886526482458501e-07,
+      "loss": 0.8578,
+      "step": 5440
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.488750226179341,
+      "learning_rate": 7.866222603855656e-07,
+      "loss": 0.7329,
+      "step": 5441
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.328318248040999,
+      "learning_rate": 7.845943825033442e-07,
+      "loss": 0.8179,
+      "step": 5442
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.640855250777269,
+      "learning_rate": 7.825690151516418e-07,
+      "loss": 0.8538,
+      "step": 5443
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3985598460015132,
+      "learning_rate": 7.805461588822161e-07,
+      "loss": 0.8194,
+      "step": 5444
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2270716252892366,
+      "learning_rate": 7.785258142461516e-07,
+      "loss": 0.7748,
+      "step": 5445
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4379233450872904,
+      "learning_rate": 7.765079817938493e-07,
+      "loss": 0.8291,
+      "step": 5446
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7274313936811894,
+      "learning_rate": 7.744926620750193e-07,
+      "loss": 0.8701,
+      "step": 5447
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.396616671371132,
+      "learning_rate": 7.724798556386892e-07,
+      "loss": 0.8344,
+      "step": 5448
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.336563401528519,
+      "learning_rate": 7.704695630332048e-07,
+      "loss": 0.8124,
+      "step": 5449
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4954696840226143,
+      "learning_rate": 7.684617848062281e-07,
+      "loss": 0.762,
+      "step": 5450
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5133165286948758,
+      "learning_rate": 7.664565215047259e-07,
+      "loss": 0.7706,
+      "step": 5451
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8045847876975954,
+      "learning_rate": 7.644537736749924e-07,
+      "loss": 0.3582,
+      "step": 5452
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6282576349070783,
+      "learning_rate": 7.624535418626323e-07,
+      "loss": 0.8762,
+      "step": 5453
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3720551086072128,
+      "learning_rate": 7.604558266125606e-07,
+      "loss": 0.87,
+      "step": 5454
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3727016225175706,
+      "learning_rate": 7.584606284690099e-07,
+      "loss": 0.8063,
+      "step": 5455
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.710926184524886,
+      "learning_rate": 7.56467947975531e-07,
+      "loss": 0.33,
+      "step": 5456
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5290127486619944,
+      "learning_rate": 7.544777856749818e-07,
+      "loss": 0.764,
+      "step": 5457
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.476218180903004,
+      "learning_rate": 7.524901421095365e-07,
+      "loss": 0.7946,
+      "step": 5458
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4853940508732788,
+      "learning_rate": 7.505050178206874e-07,
+      "loss": 0.7391,
+      "step": 5459
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.302246677325604,
+      "learning_rate": 7.485224133492341e-07,
+      "loss": 0.7194,
+      "step": 5460
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3149103903981434,
+      "learning_rate": 7.465423292352947e-07,
+      "loss": 0.8352,
+      "step": 5461
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8798223283112991,
+      "learning_rate": 7.445647660182987e-07,
+      "loss": 0.3494,
+      "step": 5462
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.522163359278408,
+      "learning_rate": 7.425897242369861e-07,
+      "loss": 0.7849,
+      "step": 5463
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3410667052737066,
+      "learning_rate": 7.406172044294157e-07,
+      "loss": 0.844,
+      "step": 5464
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6111707390880774,
+      "learning_rate": 7.386472071329543e-07,
+      "loss": 0.7971,
+      "step": 5465
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7121131539656658,
+      "learning_rate": 7.366797328842856e-07,
+      "loss": 0.7864,
+      "step": 5466
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4595184395565537,
+      "learning_rate": 7.347147822194012e-07,
+      "loss": 0.8347,
+      "step": 5467
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3449681189057456,
+      "learning_rate": 7.327523556736104e-07,
+      "loss": 0.771,
+      "step": 5468
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.849130815798013,
+      "learning_rate": 7.307924537815314e-07,
+      "loss": 0.3358,
+      "step": 5469
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7104378742622606,
+      "learning_rate": 7.288350770770935e-07,
+      "loss": 0.7534,
+      "step": 5470
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4261393345610522,
+      "learning_rate": 7.268802260935415e-07,
+      "loss": 0.8283,
+      "step": 5471
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5201861801478926,
+      "learning_rate": 7.249279013634348e-07,
+      "loss": 0.7815,
+      "step": 5472
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.343667400324821,
+      "learning_rate": 7.229781034186323e-07,
+      "loss": 0.815,
+      "step": 5473
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5539203289309018,
+      "learning_rate": 7.210308327903182e-07,
+      "loss": 0.866,
+      "step": 5474
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.286695711263135,
+      "learning_rate": 7.190860900089824e-07,
+      "loss": 0.8009,
+      "step": 5475
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6320062465131093,
+      "learning_rate": 7.171438756044258e-07,
+      "loss": 0.8289,
+      "step": 5476
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3890640218820804,
+      "learning_rate": 7.152041901057594e-07,
+      "loss": 0.845,
+      "step": 5477
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.413327070645842,
+      "learning_rate": 7.132670340414106e-07,
+      "loss": 0.7139,
+      "step": 5478
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.468835601189063,
+      "learning_rate": 7.113324079391115e-07,
+      "loss": 0.8212,
+      "step": 5479
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3552067686242206,
+      "learning_rate": 7.094003123259063e-07,
+      "loss": 0.7286,
+      "step": 5480
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7745663179327958,
+      "learning_rate": 7.07470747728155e-07,
+      "loss": 0.3087,
+      "step": 5481
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8653127395843776,
+      "learning_rate": 7.055437146715194e-07,
+      "loss": 0.3353,
+      "step": 5482
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4735506488603232,
+      "learning_rate": 7.036192136809816e-07,
+      "loss": 0.8447,
+      "step": 5483
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3002765280600637,
+      "learning_rate": 7.016972452808246e-07,
+      "loss": 0.8245,
+      "step": 5484
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5306605803009683,
+      "learning_rate": 6.997778099946495e-07,
+      "loss": 0.8325,
+      "step": 5485
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2574716003850852,
+      "learning_rate": 6.978609083453602e-07,
+      "loss": 0.7942,
+      "step": 5486
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4004290720451564,
+      "learning_rate": 6.959465408551724e-07,
+      "loss": 0.8169,
+      "step": 5487
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7767445728051512,
+      "learning_rate": 6.940347080456178e-07,
+      "loss": 0.7818,
+      "step": 5488
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8671196689315561,
+      "learning_rate": 6.921254104375264e-07,
+      "loss": 0.7944,
+      "step": 5489
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8901832388452009,
+      "learning_rate": 6.902186485510476e-07,
+      "loss": 0.3507,
+      "step": 5490
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.555749173658839,
+      "learning_rate": 6.883144229056349e-07,
+      "loss": 0.7814,
+      "step": 5491
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5763145320767764,
+      "learning_rate": 6.864127340200499e-07,
+      "loss": 0.8514,
+      "step": 5492
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5476265403769007,
+      "learning_rate": 6.845135824123672e-07,
+      "loss": 0.7905,
+      "step": 5493
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.523068056181864,
+      "learning_rate": 6.826169685999707e-07,
+      "loss": 0.8697,
+      "step": 5494
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3408861217060435,
+      "learning_rate": 6.807228930995436e-07,
+      "loss": 0.7475,
+      "step": 5495
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6021598010303677,
+      "learning_rate": 6.788313564270877e-07,
+      "loss": 0.7784,
+      "step": 5496
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.255160562599145,
+      "learning_rate": 6.769423590979107e-07,
+      "loss": 0.7566,
+      "step": 5497
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.459850173754575,
+      "learning_rate": 6.750559016266278e-07,
+      "loss": 0.8862,
+      "step": 5498
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2259314483168189,
+      "learning_rate": 6.731719845271589e-07,
+      "loss": 0.8708,
+      "step": 5499
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8215457099087253,
+      "learning_rate": 6.71290608312739e-07,
+      "loss": 0.324,
+      "step": 5500
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3397638645351855,
+      "learning_rate": 6.694117734959038e-07,
+      "loss": 0.8443,
+      "step": 5501
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7981340288996954,
+      "learning_rate": 6.675354805885004e-07,
+      "loss": 0.746,
+      "step": 5502
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2923434415092097,
+      "learning_rate": 6.656617301016833e-07,
+      "loss": 0.8123,
+      "step": 5503
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3998882431368602,
+      "learning_rate": 6.637905225459129e-07,
+      "loss": 0.8918,
+      "step": 5504
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3408067185781782,
+      "learning_rate": 6.619218584309595e-07,
+      "loss": 0.8396,
+      "step": 5505
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5629337984151004,
+      "learning_rate": 6.600557382658956e-07,
+      "loss": 0.8606,
+      "step": 5506
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2827171718458665,
+      "learning_rate": 6.58192162559107e-07,
+      "loss": 0.7833,
+      "step": 5507
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5926487819510016,
+      "learning_rate": 6.563311318182819e-07,
+      "loss": 0.8573,
+      "step": 5508
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3365687306740295,
+      "learning_rate": 6.544726465504159e-07,
+      "loss": 0.785,
+      "step": 5509
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4941721694708086,
+      "learning_rate": 6.526167072618117e-07,
+      "loss": 0.8926,
+      "step": 5510
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6306568945177617,
+      "learning_rate": 6.507633144580783e-07,
+      "loss": 0.8306,
+      "step": 5511
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3309414676196005,
+      "learning_rate": 6.489124686441328e-07,
+      "loss": 0.8119,
+      "step": 5512
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.350611640950385,
+      "learning_rate": 6.470641703241942e-07,
+      "loss": 0.8308,
+      "step": 5513
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5266014982169664,
+      "learning_rate": 6.452184200017897e-07,
+      "loss": 0.802,
+      "step": 5514
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3800896660305717,
+      "learning_rate": 6.433752181797548e-07,
+      "loss": 0.7722,
+      "step": 5515
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4764110144357536,
+      "learning_rate": 6.415345653602278e-07,
+      "loss": 0.9001,
+      "step": 5516
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.47773841803913,
+      "learning_rate": 6.396964620446522e-07,
+      "loss": 0.8194,
+      "step": 5517
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.531522237044744,
+      "learning_rate": 6.378609087337773e-07,
+      "loss": 0.8356,
+      "step": 5518
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6942495642861382,
+      "learning_rate": 6.360279059276619e-07,
+      "loss": 0.8148,
+      "step": 5519
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5477608446558635,
+      "learning_rate": 6.341974541256635e-07,
+      "loss": 0.8666,
+      "step": 5520
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7404343605727923,
+      "learning_rate": 6.323695538264474e-07,
+      "loss": 0.7703,
+      "step": 5521
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6828815094293763,
+      "learning_rate": 6.305442055279864e-07,
+      "loss": 0.8369,
+      "step": 5522
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8073488456206463,
+      "learning_rate": 6.287214097275551e-07,
+      "loss": 0.3474,
+      "step": 5523
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4320248794622075,
+      "learning_rate": 6.269011669217306e-07,
+      "loss": 0.7903,
+      "step": 5524
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3042153254278122,
+      "learning_rate": 6.250834776063996e-07,
+      "loss": 0.8653,
+      "step": 5525
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.51746061230947,
+      "learning_rate": 6.232683422767516e-07,
+      "loss": 0.8365,
+      "step": 5526
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.524819469245884,
+      "learning_rate": 6.214557614272787e-07,
+      "loss": 0.8246,
+      "step": 5527
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.524765059771098,
+      "learning_rate": 6.196457355517749e-07,
+      "loss": 0.8655,
+      "step": 5528
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4067262253779294,
+      "learning_rate": 6.178382651433456e-07,
+      "loss": 0.8146,
+      "step": 5529
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4242190307139813,
+      "learning_rate": 6.160333506943939e-07,
+      "loss": 0.8066,
+      "step": 5530
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3184105411066747,
+      "learning_rate": 6.142309926966273e-07,
+      "loss": 0.7667,
+      "step": 5531
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2522231047447183,
+      "learning_rate": 6.124311916410586e-07,
+      "loss": 0.7682,
+      "step": 5532
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.762020493920924,
+      "learning_rate": 6.106339480180023e-07,
+      "loss": 0.3376,
+      "step": 5533
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.692216606394664,
+      "learning_rate": 6.088392623170802e-07,
+      "loss": 0.323,
+      "step": 5534
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4347446094264773,
+      "learning_rate": 6.070471350272111e-07,
+      "loss": 0.8295,
+      "step": 5535
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.519851797856534,
+      "learning_rate": 6.052575666366189e-07,
+      "loss": 0.799,
+      "step": 5536
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.606896747478483,
+      "learning_rate": 6.034705576328348e-07,
+      "loss": 0.7856,
+      "step": 5537
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3655201144935198,
+      "learning_rate": 6.016861085026881e-07,
+      "loss": 0.8737,
+      "step": 5538
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2841255727759426,
+      "learning_rate": 5.999042197323102e-07,
+      "loss": 0.7866,
+      "step": 5539
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4415901882124487,
+      "learning_rate": 5.98124891807137e-07,
+      "loss": 0.7892,
+      "step": 5540
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5249203336731263,
+      "learning_rate": 5.963481252119096e-07,
+      "loss": 0.8783,
+      "step": 5541
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4693523158618806,
+      "learning_rate": 5.945739204306666e-07,
+      "loss": 0.8708,
+      "step": 5542
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3794433369062153,
+      "learning_rate": 5.928022779467468e-07,
+      "loss": 0.7912,
+      "step": 5543
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4985344943644534,
+      "learning_rate": 5.910331982428008e-07,
+      "loss": 0.7766,
+      "step": 5544
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2925112994083339,
+      "learning_rate": 5.892666818007698e-07,
+      "loss": 0.814,
+      "step": 5545
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5027866760237576,
+      "learning_rate": 5.87502729101902e-07,
+      "loss": 0.8749,
+      "step": 5546
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3905884397715607,
+      "learning_rate": 5.857413406267476e-07,
+      "loss": 0.7729,
+      "step": 5547
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7936364542930816,
+      "learning_rate": 5.839825168551594e-07,
+      "loss": 0.8335,
+      "step": 5548
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8466180819678006,
+      "learning_rate": 5.822262582662874e-07,
+      "loss": 0.3208,
+      "step": 5549
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3251947558970396,
+      "learning_rate": 5.804725653385846e-07,
+      "loss": 0.8499,
+      "step": 5550
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5271264846107917,
+      "learning_rate": 5.787214385498063e-07,
+      "loss": 0.763,
+      "step": 5551
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.152165909153071,
+      "learning_rate": 5.769728783770068e-07,
+      "loss": 0.8192,
+      "step": 5552
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6271298290796294,
+      "learning_rate": 5.752268852965426e-07,
+      "loss": 0.8475,
+      "step": 5553
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3203849134253383,
+      "learning_rate": 5.734834597840699e-07,
+      "loss": 0.797,
+      "step": 5554
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3207920677052762,
+      "learning_rate": 5.717426023145456e-07,
+      "loss": 0.8899,
+      "step": 5555
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2730090226383586,
+      "learning_rate": 5.70004313362229e-07,
+      "loss": 0.7354,
+      "step": 5556
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8767600359349941,
+      "learning_rate": 5.682685934006771e-07,
+      "loss": 0.3161,
+      "step": 5557
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3433818867387002,
+      "learning_rate": 5.66535442902747e-07,
+      "loss": 0.8635,
+      "step": 5558
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3512846779696344,
+      "learning_rate": 5.648048623405977e-07,
+      "loss": 0.7641,
+      "step": 5559
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6683006821796067,
+      "learning_rate": 5.630768521856866e-07,
+      "loss": 0.775,
+      "step": 5560
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4823825426374606,
+      "learning_rate": 5.613514129087693e-07,
+      "loss": 0.8246,
+      "step": 5561
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3919054884048734,
+      "learning_rate": 5.596285449799055e-07,
+      "loss": 0.8031,
+      "step": 5562
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4806001033738834,
+      "learning_rate": 5.579082488684529e-07,
+      "loss": 0.8067,
+      "step": 5563
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3357663379577107,
+      "learning_rate": 5.561905250430665e-07,
+      "loss": 0.7998,
+      "step": 5564
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.531265730679734,
+      "learning_rate": 5.54475373971699e-07,
+      "loss": 0.8269,
+      "step": 5565
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4685833036148803,
+      "learning_rate": 5.527627961216087e-07,
+      "loss": 0.7988,
+      "step": 5566
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2159730003415292,
+      "learning_rate": 5.510527919593478e-07,
+      "loss": 0.8186,
+      "step": 5567
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2849258380078918,
+      "learning_rate": 5.493453619507672e-07,
+      "loss": 0.8817,
+      "step": 5568
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5648903920254402,
+      "learning_rate": 5.47640506561018e-07,
+      "loss": 0.7554,
+      "step": 5569
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6900244694896094,
+      "learning_rate": 5.45938226254552e-07,
+      "loss": 0.6941,
+      "step": 5570
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5218016559289356,
+      "learning_rate": 5.442385214951151e-07,
+      "loss": 0.9481,
+      "step": 5571
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4683341530787384,
+      "learning_rate": 5.425413927457546e-07,
+      "loss": 0.8084,
+      "step": 5572
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3661652503510426,
+      "learning_rate": 5.408468404688161e-07,
+      "loss": 0.7409,
+      "step": 5573
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5785955846368618,
+      "learning_rate": 5.391548651259415e-07,
+      "loss": 0.8118,
+      "step": 5574
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3431355331484052,
+      "learning_rate": 5.374654671780688e-07,
+      "loss": 0.7645,
+      "step": 5575
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7792403348809664,
+      "learning_rate": 5.357786470854421e-07,
+      "loss": 0.3245,
+      "step": 5576
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4823851097081293,
+      "learning_rate": 5.340944053075925e-07,
+      "loss": 0.7747,
+      "step": 5577
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.277916916741586,
+      "learning_rate": 5.324127423033576e-07,
+      "loss": 0.8175,
+      "step": 5578
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3557016683753234,
+      "learning_rate": 5.307336585308676e-07,
+      "loss": 0.76,
+      "step": 5579
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.922831357478769,
+      "learning_rate": 5.290571544475487e-07,
+      "loss": 0.8002,
+      "step": 5580
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3428219498627414,
+      "learning_rate": 5.273832305101312e-07,
+      "loss": 0.7891,
+      "step": 5581
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6923478910405894,
+      "learning_rate": 5.257118871746347e-07,
+      "loss": 0.8549,
+      "step": 5582
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9669896886463294,
+      "learning_rate": 5.240431248963807e-07,
+      "loss": 0.7725,
+      "step": 5583
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4131559718399938,
+      "learning_rate": 5.223769441299842e-07,
+      "loss": 0.8149,
+      "step": 5584
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3580317598835006,
+      "learning_rate": 5.207133453293633e-07,
+      "loss": 0.869,
+      "step": 5585
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5810420275553005,
+      "learning_rate": 5.190523289477234e-07,
+      "loss": 0.8476,
+      "step": 5586
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5946614158111023,
+      "learning_rate": 5.173938954375734e-07,
+      "loss": 0.7343,
+      "step": 5587
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4162044017842432,
+      "learning_rate": 5.157380452507166e-07,
+      "loss": 0.7515,
+      "step": 5588
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4924453436435994,
+      "learning_rate": 5.140847788382508e-07,
+      "loss": 0.7205,
+      "step": 5589
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1336140246980457,
+      "learning_rate": 5.124340966505715e-07,
+      "loss": 0.8262,
+      "step": 5590
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2883493032750521,
+      "learning_rate": 5.107859991373698e-07,
+      "loss": 0.7726,
+      "step": 5591
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4137513986868562,
+      "learning_rate": 5.091404867476368e-07,
+      "loss": 0.8049,
+      "step": 5592
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4944600792253941,
+      "learning_rate": 5.074975599296494e-07,
+      "loss": 0.8724,
+      "step": 5593
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5726363959198584,
+      "learning_rate": 5.058572191309896e-07,
+      "loss": 0.7507,
+      "step": 5594
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3018099444596989,
+      "learning_rate": 5.042194647985311e-07,
+      "loss": 0.8027,
+      "step": 5595
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8361215656168353,
+      "learning_rate": 5.025842973784445e-07,
+      "loss": 0.3687,
+      "step": 5596
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4082944360247134,
+      "learning_rate": 5.009517173161904e-07,
+      "loss": 0.7881,
+      "step": 5597
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4449359525720271,
+      "learning_rate": 4.993217250565341e-07,
+      "loss": 0.8128,
+      "step": 5598
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.320181308243942,
+      "learning_rate": 4.976943210435247e-07,
+      "loss": 0.8845,
+      "step": 5599
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2916138701326245,
+      "learning_rate": 4.960695057205178e-07,
+      "loss": 0.8052,
+      "step": 5600
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6095151545087636,
+      "learning_rate": 4.944472795301535e-07,
+      "loss": 0.8178,
+      "step": 5601
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2586029315015101,
+      "learning_rate": 4.928276429143719e-07,
+      "loss": 0.8256,
+      "step": 5602
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1629534324801705,
+      "learning_rate": 4.912105963144076e-07,
+      "loss": 0.7967,
+      "step": 5603
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5487660399184402,
+      "learning_rate": 4.895961401707882e-07,
+      "loss": 0.9005,
+      "step": 5604
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3227978470588002,
+      "learning_rate": 4.879842749233366e-07,
+      "loss": 0.8163,
+      "step": 5605
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5109991700832777,
+      "learning_rate": 4.863750010111667e-07,
+      "loss": 0.8554,
+      "step": 5606
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5408918321344962,
+      "learning_rate": 4.847683188726938e-07,
+      "loss": 0.834,
+      "step": 5607
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4740615357595384,
+      "learning_rate": 4.831642289456184e-07,
+      "loss": 0.776,
+      "step": 5608
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6157480395234871,
+      "learning_rate": 4.815627316669403e-07,
+      "loss": 0.8456,
+      "step": 5609
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5792749058518485,
+      "learning_rate": 4.799638274729513e-07,
+      "loss": 0.7485,
+      "step": 5610
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3733168227419978,
+      "learning_rate": 4.783675167992385e-07,
+      "loss": 0.8078,
+      "step": 5611
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4780729785826854,
+      "learning_rate": 4.767738000806765e-07,
+      "loss": 0.7943,
+      "step": 5612
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3982431080838194,
+      "learning_rate": 4.7518267775144233e-07,
+      "loss": 0.8297,
+      "step": 5613
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.502008260997234,
+      "learning_rate": 4.7359415024500143e-07,
+      "loss": 0.7837,
+      "step": 5614
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4574214972621047,
+      "learning_rate": 4.7200821799410767e-07,
+      "loss": 0.7926,
+      "step": 5615
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.379308878171987,
+      "learning_rate": 4.7042488143081766e-07,
+      "loss": 0.8188,
+      "step": 5616
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4145438747481005,
+      "learning_rate": 4.6884414098647415e-07,
+      "loss": 0.8721,
+      "step": 5617
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.613216605519247,
+      "learning_rate": 4.6726599709171483e-07,
+      "loss": 0.8165,
+      "step": 5618
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4773727639827547,
+      "learning_rate": 4.6569045017646807e-07,
+      "loss": 0.7462,
+      "step": 5619
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4956098918245366,
+      "learning_rate": 4.641175006699594e-07,
+      "loss": 0.8706,
+      "step": 5620
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5918035167719349,
+      "learning_rate": 4.625471490007005e-07,
+      "loss": 0.7302,
+      "step": 5621
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4508785616254272,
+      "learning_rate": 4.609793955964992e-07,
+      "loss": 0.7664,
+      "step": 5622
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8383526417479003,
+      "learning_rate": 4.5941424088445485e-07,
+      "loss": 0.3414,
+      "step": 5623
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.479634468237048,
+      "learning_rate": 4.578516852909609e-07,
+      "loss": 0.8519,
+      "step": 5624
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2563251978160062,
+      "learning_rate": 4.5629172924169793e-07,
+      "loss": 0.7963,
+      "step": 5625
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8709697894454793,
+      "learning_rate": 4.547343731616405e-07,
+      "loss": 0.3568,
+      "step": 5626
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6784947130135892,
+      "learning_rate": 4.5317961747505803e-07,
+      "loss": 0.7842,
+      "step": 5627
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3702008832897716,
+      "learning_rate": 4.5162746260550614e-07,
+      "loss": 0.8407,
+      "step": 5628
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4197127045172724,
+      "learning_rate": 4.500779089758378e-07,
+      "loss": 0.7327,
+      "step": 5629
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6621593334476215,
+      "learning_rate": 4.4853095700819196e-07,
+      "loss": 0.7664,
+      "step": 5630
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4717464798905782,
+      "learning_rate": 4.469866071239992e-07,
+      "loss": 0.8429,
+      "step": 5631
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.228909530926282,
+      "learning_rate": 4.4544485974398757e-07,
+      "loss": 0.7894,
+      "step": 5632
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3941502144040971,
+      "learning_rate": 4.439057152881676e-07,
+      "loss": 0.8283,
+      "step": 5633
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4131273759198115,
+      "learning_rate": 4.4236917417584513e-07,
+      "loss": 0.8454,
+      "step": 5634
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2921710045579249,
+      "learning_rate": 4.4083523682561747e-07,
+      "loss": 0.7612,
+      "step": 5635
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9042498986254954,
+      "learning_rate": 4.393039036553748e-07,
+      "loss": 0.3424,
+      "step": 5636
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.425707116224639,
+      "learning_rate": 4.377751750822867e-07,
+      "loss": 0.7579,
+      "step": 5637
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.404853764762308,
+      "learning_rate": 4.362490515228257e-07,
+      "loss": 0.7832,
+      "step": 5638
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3824882861362426,
+      "learning_rate": 4.3472553339275136e-07,
+      "loss": 0.8301,
+      "step": 5639
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.403492763650546,
+      "learning_rate": 4.3320462110710946e-07,
+      "loss": 0.7324,
+      "step": 5640
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6395437605924126,
+      "learning_rate": 4.316863150802375e-07,
+      "loss": 0.8387,
+      "step": 5641
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.0039441902629087,
+      "learning_rate": 4.301706157257657e-07,
+      "loss": 0.7473,
+      "step": 5642
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.425808234808583,
+      "learning_rate": 4.286575234566148e-07,
+      "loss": 0.861,
+      "step": 5643
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5896814372415942,
+      "learning_rate": 4.271470386849874e-07,
+      "loss": 0.7355,
+      "step": 5644
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1431877612810804,
+      "learning_rate": 4.256391618223843e-07,
+      "loss": 0.7865,
+      "step": 5645
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4845729222228934,
+      "learning_rate": 4.241338932795935e-07,
+      "loss": 0.8244,
+      "step": 5646
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2599954262434736,
+      "learning_rate": 4.226312334666904e-07,
+      "loss": 0.7877,
+      "step": 5647
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9118153649973397,
+      "learning_rate": 4.211311827930398e-07,
+      "loss": 0.8042,
+      "step": 5648
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.441999835914149,
+      "learning_rate": 4.196337416672991e-07,
+      "loss": 0.7767,
+      "step": 5649
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3878284849497384,
+      "learning_rate": 4.1813891049740986e-07,
+      "loss": 0.8145,
+      "step": 5650
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3252025954033044,
+      "learning_rate": 4.166466896906085e-07,
+      "loss": 0.8314,
+      "step": 5651
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5864556137454788,
+      "learning_rate": 4.1515707965341547e-07,
+      "loss": 0.87,
+      "step": 5652
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2391506371172691,
+      "learning_rate": 4.136700807916405e-07,
+      "loss": 0.833,
+      "step": 5653
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8091769775687769,
+      "learning_rate": 4.121856935103863e-07,
+      "loss": 0.3439,
+      "step": 5654
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2063803846918737,
+      "learning_rate": 4.107039182140382e-07,
+      "loss": 0.7326,
+      "step": 5655
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3166823348787937,
+      "learning_rate": 4.0922475530627224e-07,
+      "loss": 0.766,
+      "step": 5656
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2419849461528696,
+      "learning_rate": 4.0774820519005385e-07,
+      "loss": 0.8341,
+      "step": 5657
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7446180424132262,
+      "learning_rate": 4.0627426826763903e-07,
+      "loss": 0.3214,
+      "step": 5658
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2046703003269057,
+      "learning_rate": 4.048029449405633e-07,
+      "loss": 0.7415,
+      "step": 5659
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.366660248043041,
+      "learning_rate": 4.033342356096592e-07,
+      "loss": 0.8327,
+      "step": 5660
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5712890435332019,
+      "learning_rate": 4.0186814067504356e-07,
+      "loss": 0.7994,
+      "step": 5661
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5218286706238888,
+      "learning_rate": 4.0040466053612006e-07,
+      "loss": 0.749,
+      "step": 5662
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4804011633953513,
+      "learning_rate": 3.9894379559158094e-07,
+      "loss": 0.8082,
+      "step": 5663
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3076894909652519,
+      "learning_rate": 3.974855462394067e-07,
+      "loss": 0.7848,
+      "step": 5664
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3696682382059755,
+      "learning_rate": 3.960299128768663e-07,
+      "loss": 0.8904,
+      "step": 5665
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4192875923835468,
+      "learning_rate": 3.945768959005114e-07,
+      "loss": 0.7434,
+      "step": 5666
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.375736066649245,
+      "learning_rate": 3.931264957061842e-07,
+      "loss": 0.7329,
+      "step": 5667
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1753555300218785,
+      "learning_rate": 3.9167871268901536e-07,
+      "loss": 0.8189,
+      "step": 5668
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4362959198587142,
+      "learning_rate": 3.9023354724342157e-07,
+      "loss": 0.8188,
+      "step": 5669
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2652781064094962,
+      "learning_rate": 3.887909997631012e-07,
+      "loss": 0.8346,
+      "step": 5670
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4904698020936544,
+      "learning_rate": 3.8735107064104994e-07,
+      "loss": 0.8342,
+      "step": 5671
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2775343350310018,
+      "learning_rate": 3.8591376026954063e-07,
+      "loss": 0.7385,
+      "step": 5672
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.37766653477396,
+      "learning_rate": 3.844790690401357e-07,
+      "loss": 0.8394,
+      "step": 5673
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.337868472424067,
+      "learning_rate": 3.830469973436868e-07,
+      "loss": 0.8119,
+      "step": 5674
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4295056858043067,
+      "learning_rate": 3.816175455703264e-07,
+      "loss": 0.8056,
+      "step": 5675
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5694926657806871,
+      "learning_rate": 3.8019071410948183e-07,
+      "loss": 0.8523,
+      "step": 5676
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2646407491678548,
+      "learning_rate": 3.7876650334985776e-07,
+      "loss": 0.8516,
+      "step": 5677
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3175515871969035,
+      "learning_rate": 3.7734491367944716e-07,
+      "loss": 0.774,
+      "step": 5678
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4966343084915807,
+      "learning_rate": 3.7592594548553354e-07,
+      "loss": 0.8444,
+      "step": 5679
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5380559779080085,
+      "learning_rate": 3.745095991546821e-07,
+      "loss": 0.7379,
+      "step": 5680
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5638468957229843,
+      "learning_rate": 3.7309587507274313e-07,
+      "loss": 0.874,
+      "step": 5681
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5280627926449353,
+      "learning_rate": 3.716847736248541e-07,
+      "loss": 0.833,
+      "step": 5682
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.597118067858388,
+      "learning_rate": 3.7027629519544085e-07,
+      "loss": 0.8244,
+      "step": 5683
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2254150894660858,
+      "learning_rate": 3.688704401682086e-07,
+      "loss": 0.7363,
+      "step": 5684
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7515129771875544,
+      "learning_rate": 3.6746720892615216e-07,
+      "loss": 0.7591,
+      "step": 5685
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.290912804426233,
+      "learning_rate": 3.6606660185154906e-07,
+      "loss": 0.7509,
+      "step": 5686
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4624522912311337,
+      "learning_rate": 3.646686193259674e-07,
+      "loss": 0.8895,
+      "step": 5687
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7426132218709403,
+      "learning_rate": 3.632732617302515e-07,
+      "loss": 0.7744,
+      "step": 5688
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2918204617286333,
+      "learning_rate": 3.61880529444536e-07,
+      "loss": 0.8737,
+      "step": 5689
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3938178560445185,
+      "learning_rate": 3.604904228482431e-07,
+      "loss": 0.8338,
+      "step": 5690
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7747638442811526,
+      "learning_rate": 3.5910294232007206e-07,
+      "loss": 0.318,
+      "step": 5691
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4327858108584866,
+      "learning_rate": 3.5771808823801266e-07,
+      "loss": 0.7994,
+      "step": 5692
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.317227857114016,
+      "learning_rate": 3.563358609793377e-07,
+      "loss": 0.863,
+      "step": 5693
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.20140643349516,
+      "learning_rate": 3.5495626092060367e-07,
+      "loss": 0.8001,
+      "step": 5694
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2908751920652093,
+      "learning_rate": 3.535792884376499e-07,
+      "loss": 0.7814,
+      "step": 5695
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3672627462388711,
+      "learning_rate": 3.5220494390560414e-07,
+      "loss": 0.824,
+      "step": 5696
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3559451317325133,
+      "learning_rate": 3.508332276988735e-07,
+      "loss": 0.7527,
+      "step": 5697
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.1095629750254083,
+      "learning_rate": 3.494641401911536e-07,
+      "loss": 0.8565,
+      "step": 5698
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4507735682537302,
+      "learning_rate": 3.4809768175542046e-07,
+      "loss": 0.7472,
+      "step": 5699
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3294757204953986,
+      "learning_rate": 3.467338527639341e-07,
+      "loss": 0.8088,
+      "step": 5700
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2775825720526357,
+      "learning_rate": 3.453726535882418e-07,
+      "loss": 0.7575,
+      "step": 5701
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3716570975219533,
+      "learning_rate": 3.44014084599168e-07,
+      "loss": 0.7697,
+      "step": 5702
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.1948988808928853,
+      "learning_rate": 3.4265814616682766e-07,
+      "loss": 0.8386,
+      "step": 5703
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3841450582602188,
+      "learning_rate": 3.4130483866061327e-07,
+      "loss": 0.8273,
+      "step": 5704
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4319063684211157,
+      "learning_rate": 3.3995416244920643e-07,
+      "loss": 0.8186,
+      "step": 5705
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2233245701878226,
+      "learning_rate": 3.386061179005651e-07,
+      "loss": 0.8159,
+      "step": 5706
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3536741943419865,
+      "learning_rate": 3.372607053819355e-07,
+      "loss": 0.7854,
+      "step": 5707
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4069484512647155,
+      "learning_rate": 3.3591792525984324e-07,
+      "loss": 0.7626,
+      "step": 5708
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4740732576304418,
+      "learning_rate": 3.345777779001036e-07,
+      "loss": 0.8711,
+      "step": 5709
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3409633886595875,
+      "learning_rate": 3.3324026366780224e-07,
+      "loss": 0.8287,
+      "step": 5710
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.609613561772463,
+      "learning_rate": 3.3190538292732e-07,
+      "loss": 0.8199,
+      "step": 5711
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4331344627553741,
+      "learning_rate": 3.305731360423159e-07,
+      "loss": 0.8844,
+      "step": 5712
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5242148885673126,
+      "learning_rate": 3.2924352337572743e-07,
+      "loss": 0.8308,
+      "step": 5713
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2901727381889982,
+      "learning_rate": 3.2791654528977924e-07,
+      "loss": 0.7207,
+      "step": 5714
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.563864787816798,
+      "learning_rate": 3.2659220214597666e-07,
+      "loss": 0.7675,
+      "step": 5715
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4557242411892386,
+      "learning_rate": 3.2527049430510883e-07,
+      "loss": 0.7986,
+      "step": 5716
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4548453199159757,
+      "learning_rate": 3.239514221272411e-07,
+      "loss": 0.8285,
+      "step": 5717
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3284064120991832,
+      "learning_rate": 3.226349859717293e-07,
+      "loss": 0.7902,
+      "step": 5718
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3425087366161716,
+      "learning_rate": 3.2132118619720545e-07,
+      "loss": 0.8894,
+      "step": 5719
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3559409941788485,
+      "learning_rate": 3.2001002316158434e-07,
+      "loss": 0.8341,
+      "step": 5720
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2115435356620574,
+      "learning_rate": 3.1870149722206366e-07,
+      "loss": 0.768,
+      "step": 5721
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.649036182371031,
+      "learning_rate": 3.1739560873512155e-07,
+      "loss": 0.7992,
+      "step": 5722
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.559163989533164,
+      "learning_rate": 3.1609235805651896e-07,
+      "loss": 0.7702,
+      "step": 5723
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7531969286376207,
+      "learning_rate": 3.147917455412952e-07,
+      "loss": 0.3293,
+      "step": 5724
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4354961593907734,
+      "learning_rate": 3.134937715437758e-07,
+      "loss": 0.7518,
+      "step": 5725
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.432405222740573,
+      "learning_rate": 3.121984364175612e-07,
+      "loss": 0.8483,
+      "step": 5726
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4661140452564443,
+      "learning_rate": 3.109057405155402e-07,
+      "loss": 0.851,
+      "step": 5727
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4970222286258723,
+      "learning_rate": 3.0961568418987673e-07,
+      "loss": 0.7869,
+      "step": 5728
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3134968327089118,
+      "learning_rate": 3.0832826779201633e-07,
+      "loss": 0.7834,
+      "step": 5729
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.565020378607771,
+      "learning_rate": 3.070434916726905e-07,
+      "loss": 0.8209,
+      "step": 5730
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.25055365073191,
+      "learning_rate": 3.0576135618190393e-07,
+      "loss": 0.809,
+      "step": 5731
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4163288000333403,
+      "learning_rate": 3.044818616689471e-07,
+      "loss": 0.8356,
+      "step": 5732
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6504098987018232,
+      "learning_rate": 3.032050084823901e-07,
+      "loss": 0.7658,
+      "step": 5733
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5054144548375827,
+      "learning_rate": 3.019307969700824e-07,
+      "loss": 0.7849,
+      "step": 5734
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.257743883926915,
+      "learning_rate": 3.006592274791553e-07,
+      "loss": 0.7839,
+      "step": 5735
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7418143965875021,
+      "learning_rate": 2.993903003560172e-07,
+      "loss": 0.347,
+      "step": 5736
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4109724112032695,
+      "learning_rate": 2.981240159463616e-07,
+      "loss": 0.7923,
+      "step": 5737
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7100957145276017,
+      "learning_rate": 2.9686037459515707e-07,
+      "loss": 0.8123,
+      "step": 5738
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.265140717476901,
+      "learning_rate": 2.9559937664665474e-07,
+      "loss": 0.8439,
+      "step": 5739
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4061792739990566,
+      "learning_rate": 2.9434102244438544e-07,
+      "loss": 0.7573,
+      "step": 5740
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4601666232251374,
+      "learning_rate": 2.9308531233115947e-07,
+      "loss": 0.842,
+      "step": 5741
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4607137620926964,
+      "learning_rate": 2.918322466490686e-07,
+      "loss": 0.7802,
+      "step": 5742
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.240896137033796,
+      "learning_rate": 2.905818257394799e-07,
+      "loss": 0.8078,
+      "step": 5743
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5087117270985708,
+      "learning_rate": 2.8933404994304417e-07,
+      "loss": 0.8278,
+      "step": 5744
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1564547870488182,
+      "learning_rate": 2.8808891959968946e-07,
+      "loss": 0.7826,
+      "step": 5745
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5820855943741345,
+      "learning_rate": 2.868464350486222e-07,
+      "loss": 0.7924,
+      "step": 5746
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2096314649385618,
+      "learning_rate": 2.856065966283317e-07,
+      "loss": 0.8096,
+      "step": 5747
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4196107410686543,
+      "learning_rate": 2.8436940467658213e-07,
+      "loss": 0.8378,
+      "step": 5748
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5321116200715066,
+      "learning_rate": 2.831348595304206e-07,
+      "loss": 0.8525,
+      "step": 5749
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2809909829144148,
+      "learning_rate": 2.8190296152617035e-07,
+      "loss": 0.8575,
+      "step": 5750
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3459590328040532,
+      "learning_rate": 2.8067371099943286e-07,
+      "loss": 0.7978,
+      "step": 5751
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4054637538827472,
+      "learning_rate": 2.794471082850936e-07,
+      "loss": 0.8092,
+      "step": 5752
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5590754149523856,
+      "learning_rate": 2.7822315371730965e-07,
+      "loss": 0.8959,
+      "step": 5753
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6985122247706366,
+      "learning_rate": 2.7700184762952e-07,
+      "loss": 0.8294,
+      "step": 5754
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3573197367546328,
+      "learning_rate": 2.7578319035444277e-07,
+      "loss": 0.8212,
+      "step": 5755
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.392973744666387,
+      "learning_rate": 2.7456718222407584e-07,
+      "loss": 0.7665,
+      "step": 5756
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6180717502938349,
+      "learning_rate": 2.7335382356969196e-07,
+      "loss": 0.7704,
+      "step": 5757
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3204315868258285,
+      "learning_rate": 2.721431147218412e-07,
+      "loss": 0.7897,
+      "step": 5758
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.371477493153698,
+      "learning_rate": 2.709350560103574e-07,
+      "loss": 0.7673,
+      "step": 5759
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3514601663204908,
+      "learning_rate": 2.697296477643474e-07,
+      "loss": 0.8255,
+      "step": 5760
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8155569462638977,
+      "learning_rate": 2.6852689031219626e-07,
+      "loss": 0.3485,
+      "step": 5761
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9354440457369974,
+      "learning_rate": 2.6732678398157077e-07,
+      "loss": 0.3088,
+      "step": 5762
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4171039194161557,
+      "learning_rate": 2.6612932909941267e-07,
+      "loss": 0.8598,
+      "step": 5763
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4026105515365053,
+      "learning_rate": 2.6493452599194115e-07,
+      "loss": 0.8516,
+      "step": 5764
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3673707866425937,
+      "learning_rate": 2.637423749846524e-07,
+      "loss": 0.8645,
+      "step": 5765
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3910387982626506,
+      "learning_rate": 2.625528764023222e-07,
+      "loss": 0.8392,
+      "step": 5766
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6040481956092347,
+      "learning_rate": 2.6136603056900356e-07,
+      "loss": 0.8448,
+      "step": 5767
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3336763333481905,
+      "learning_rate": 2.601818378080245e-07,
+      "loss": 0.7069,
+      "step": 5768
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4230752636876016,
+      "learning_rate": 2.590002984419937e-07,
+      "loss": 0.8964,
+      "step": 5769
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5616355304672134,
+      "learning_rate": 2.578214127927925e-07,
+      "loss": 0.7656,
+      "step": 5770
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2842804742287206,
+      "learning_rate": 2.56645181181584e-07,
+      "loss": 0.7792,
+      "step": 5771
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2785013185814305,
+      "learning_rate": 2.5547160392880523e-07,
+      "loss": 0.8141,
+      "step": 5772
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3467578498724593,
+      "learning_rate": 2.543006813541704e-07,
+      "loss": 0.804,
+      "step": 5773
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5006719055435538,
+      "learning_rate": 2.531324137766722e-07,
+      "loss": 0.782,
+      "step": 5774
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7416609759890418,
+      "learning_rate": 2.5196680151457933e-07,
+      "loss": 0.3341,
+      "step": 5775
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4277388106876046,
+      "learning_rate": 2.508038448854344e-07,
+      "loss": 0.7196,
+      "step": 5776
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3245600243737297,
+      "learning_rate": 2.496435442060607e-07,
+      "loss": 0.8227,
+      "step": 5777
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6937775650983733,
+      "learning_rate": 2.484858997925566e-07,
+      "loss": 0.884,
+      "step": 5778
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5324028976144402,
+      "learning_rate": 2.47330911960294e-07,
+      "loss": 0.8383,
+      "step": 5779
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.499900944985622,
+      "learning_rate": 2.461785810239259e-07,
+      "loss": 0.7424,
+      "step": 5780
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4783272415343476,
+      "learning_rate": 2.4502890729737773e-07,
+      "loss": 0.8746,
+      "step": 5781
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8030600954335674,
+      "learning_rate": 2.4388189109385227e-07,
+      "loss": 0.3103,
+      "step": 5782
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4651735214333335,
+      "learning_rate": 2.427375327258286e-07,
+      "loss": 0.7848,
+      "step": 5783
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.424424955555753,
+      "learning_rate": 2.4159583250506157e-07,
+      "loss": 0.813,
+      "step": 5784
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2942199156536125,
+      "learning_rate": 2.4045679074258253e-07,
+      "loss": 0.7786,
+      "step": 5785
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2761286488554746,
+      "learning_rate": 2.393204077486966e-07,
+      "loss": 0.8546,
+      "step": 5786
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3751141577982744,
+      "learning_rate": 2.3818668383298605e-07,
+      "loss": 0.7699,
+      "step": 5787
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2053829577308015,
+      "learning_rate": 2.3705561930430942e-07,
+      "loss": 0.7559,
+      "step": 5788
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4305610437375362,
+      "learning_rate": 2.3592721447079912e-07,
+      "loss": 0.7992,
+      "step": 5789
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4770834277619957,
+      "learning_rate": 2.348014696398626e-07,
+      "loss": 0.7178,
+      "step": 5790
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3655621390311026,
+      "learning_rate": 2.3367838511818675e-07,
+      "loss": 0.7927,
+      "step": 5791
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3468260025142138,
+      "learning_rate": 2.325579612117279e-07,
+      "loss": 0.8511,
+      "step": 5792
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2992623628646924,
+      "learning_rate": 2.3144019822572194e-07,
+      "loss": 0.8055,
+      "step": 5793
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4099751115734618,
+      "learning_rate": 2.303250964646786e-07,
+      "loss": 0.9016,
+      "step": 5794
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2671931671175782,
+      "learning_rate": 2.2921265623238042e-07,
+      "loss": 0.8611,
+      "step": 5795
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3088528006625328,
+      "learning_rate": 2.2810287783188833e-07,
+      "loss": 0.8901,
+      "step": 5796
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.229162341618068,
+      "learning_rate": 2.2699576156553715e-07,
+      "loss": 0.7881,
+      "step": 5797
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5970655307555903,
+      "learning_rate": 2.258913077349334e-07,
+      "loss": 0.7907,
+      "step": 5798
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3501786523290782,
+      "learning_rate": 2.2478951664096305e-07,
+      "loss": 0.8763,
+      "step": 5799
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5031802536797807,
+      "learning_rate": 2.236903885837849e-07,
+      "loss": 0.8488,
+      "step": 5800
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.40396595678273,
+      "learning_rate": 2.2259392386282829e-07,
+      "loss": 0.8053,
+      "step": 5801
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6262999095678248,
+      "learning_rate": 2.215001227768032e-07,
+      "loss": 0.866,
+      "step": 5802
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.566119752877557,
+      "learning_rate": 2.204089856236913e-07,
+      "loss": 0.8574,
+      "step": 5803
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.0439017214184325,
+      "learning_rate": 2.1932051270074807e-07,
+      "loss": 0.8062,
+      "step": 5804
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2900990306883946,
+      "learning_rate": 2.182347043045019e-07,
+      "loss": 0.7578,
+      "step": 5805
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.255523284046389,
+      "learning_rate": 2.1715156073075838e-07,
+      "loss": 0.8227,
+      "step": 5806
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2764864711141906,
+      "learning_rate": 2.1607108227459594e-07,
+      "loss": 0.8184,
+      "step": 5807
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5518717216853828,
+      "learning_rate": 2.1499326923036688e-07,
+      "loss": 0.7082,
+      "step": 5808
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.441591985490072,
+      "learning_rate": 2.1391812189169526e-07,
+      "loss": 0.7736,
+      "step": 5809
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3558148044040281,
+      "learning_rate": 2.1284564055148337e-07,
+      "loss": 0.8754,
+      "step": 5810
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3069128863328823,
+      "learning_rate": 2.1177582550190313e-07,
+      "loss": 0.8436,
+      "step": 5811
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2052003900463464,
+      "learning_rate": 2.1070867703440135e-07,
+      "loss": 0.8353,
+      "step": 5812
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2648404826156798,
+      "learning_rate": 2.0964419543970104e-07,
+      "loss": 0.8447,
+      "step": 5813
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.374705865676168,
+      "learning_rate": 2.0858238100779248e-07,
+      "loss": 0.7815,
+      "step": 5814
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4942750430884373,
+      "learning_rate": 2.075232340279465e-07,
+      "loss": 0.8381,
+      "step": 5815
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4149789047090398,
+      "learning_rate": 2.0646675478870337e-07,
+      "loss": 0.8018,
+      "step": 5816
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3489783443129002,
+      "learning_rate": 2.0541294357787512e-07,
+      "loss": 0.7473,
+      "step": 5817
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.559304940942349,
+      "learning_rate": 2.0436180068255207e-07,
+      "loss": 0.7405,
+      "step": 5818
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8691919450931498,
+      "learning_rate": 2.0331332638909184e-07,
+      "loss": 0.3197,
+      "step": 5819
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.34528276095489,
+      "learning_rate": 2.022675209831282e-07,
+      "loss": 0.7089,
+      "step": 5820
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5229308679764808,
+      "learning_rate": 2.0122438474956764e-07,
+      "loss": 0.7801,
+      "step": 5821
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4417133934472999,
+      "learning_rate": 2.0018391797259063e-07,
+      "loss": 0.894,
+      "step": 5822
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4295438149692528,
+      "learning_rate": 1.9914612093564822e-07,
+      "loss": 0.7677,
+      "step": 5823
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.424046308705004,
+      "learning_rate": 1.9811099392146427e-07,
+      "loss": 0.8054,
+      "step": 5824
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.586981646932607,
+      "learning_rate": 1.970785372120354e-07,
+      "loss": 0.8708,
+      "step": 5825
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4394830897303752,
+      "learning_rate": 1.960487510886333e-07,
+      "loss": 0.8099,
+      "step": 5826
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.367622377169449,
+      "learning_rate": 1.95021635831798e-07,
+      "loss": 0.8392,
+      "step": 5827
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4465438810187796,
+      "learning_rate": 1.9399719172134458e-07,
+      "loss": 0.7998,
+      "step": 5828
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.291521783283156,
+      "learning_rate": 1.9297541903636196e-07,
+      "loss": 0.8343,
+      "step": 5829
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3826197827338123,
+      "learning_rate": 1.9195631805520642e-07,
+      "loss": 0.7827,
+      "step": 5830
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.444376988759316,
+      "learning_rate": 1.909398890555092e-07,
+      "loss": 0.7899,
+      "step": 5831
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.423264810014063,
+      "learning_rate": 1.8992613231417546e-07,
+      "loss": 0.8071,
+      "step": 5832
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3749603650492794,
+      "learning_rate": 1.889150481073798e-07,
+      "loss": 0.7778,
+      "step": 5833
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1917279946400148,
+      "learning_rate": 1.8790663671056863e-07,
+      "loss": 0.8071,
+      "step": 5834
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6873534881824053,
+      "learning_rate": 1.8690089839846215e-07,
+      "loss": 0.8151,
+      "step": 5835
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3043343232306424,
+      "learning_rate": 1.8589783344504897e-07,
+      "loss": 0.7807,
+      "step": 5836
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.403143849092268,
+      "learning_rate": 1.8489744212359495e-07,
+      "loss": 0.7829,
+      "step": 5837
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4177916412246403,
+      "learning_rate": 1.8389972470663208e-07,
+      "loss": 0.8537,
+      "step": 5838
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4761373501349313,
+      "learning_rate": 1.829046814659663e-07,
+      "loss": 0.8183,
+      "step": 5839
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6106246111815314,
+      "learning_rate": 1.8191231267267629e-07,
+      "loss": 0.7468,
+      "step": 5840
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5383885597551368,
+      "learning_rate": 1.8092261859710802e-07,
+      "loss": 0.8877,
+      "step": 5841
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4866331046979355,
+      "learning_rate": 1.799355995088836e-07,
+      "loss": 0.873,
+      "step": 5842
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2227693663851267,
+      "learning_rate": 1.7895125567689354e-07,
+      "loss": 0.7574,
+      "step": 5843
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5931120663971288,
+      "learning_rate": 1.7796958736929992e-07,
+      "loss": 0.8464,
+      "step": 5844
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4743775431811985,
+      "learning_rate": 1.7699059485353775e-07,
+      "loss": 0.8229,
+      "step": 5845
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3121949136465074,
+      "learning_rate": 1.7601427839630814e-07,
+      "loss": 0.8245,
+      "step": 5846
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3186576484653143,
+      "learning_rate": 1.7504063826359053e-07,
+      "loss": 0.7679,
+      "step": 5847
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.451073230532813,
+      "learning_rate": 1.740696747206294e-07,
+      "loss": 0.8182,
+      "step": 5848
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4891040810206257,
+      "learning_rate": 1.731013880319421e-07,
+      "loss": 0.8541,
+      "step": 5849
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6127505777593139,
+      "learning_rate": 1.7213577846131647e-07,
+      "loss": 0.768,
+      "step": 5850
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4168771254069639,
+      "learning_rate": 1.7117284627181207e-07,
+      "loss": 0.8061,
+      "step": 5851
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7554694624331924,
+      "learning_rate": 1.7021259172575688e-07,
+      "loss": 0.7914,
+      "step": 5852
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4722342547396627,
+      "learning_rate": 1.6925501508475162e-07,
+      "loss": 0.8351,
+      "step": 5853
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.47263088508031,
+      "learning_rate": 1.6830011660966648e-07,
+      "loss": 0.743,
+      "step": 5854
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.458456344113911,
+      "learning_rate": 1.673478965606423e-07,
+      "loss": 0.8334,
+      "step": 5855
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.567902450936999,
+      "learning_rate": 1.6639835519708826e-07,
+      "loss": 0.7675,
+      "step": 5856
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5001548405290912,
+      "learning_rate": 1.6545149277768845e-07,
+      "loss": 0.8957,
+      "step": 5857
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2718217207377867,
+      "learning_rate": 1.6450730956039328e-07,
+      "loss": 0.8219,
+      "step": 5858
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.399894317827479,
+      "learning_rate": 1.6356580580242253e-07,
+      "loss": 0.7816,
+      "step": 5859
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2900403114445638,
+      "learning_rate": 1.626269817602699e-07,
+      "loss": 0.8787,
+      "step": 5860
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7878762616528001,
+      "learning_rate": 1.616908376896964e-07,
+      "loss": 0.8854,
+      "step": 5861
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3663184728286413,
+      "learning_rate": 1.6075737384573354e-07,
+      "loss": 0.7603,
+      "step": 5862
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.349524226173142,
+      "learning_rate": 1.5982659048268124e-07,
+      "loss": 0.8257,
+      "step": 5863
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4152968930642944,
+      "learning_rate": 1.588984878541133e-07,
+      "loss": 0.8594,
+      "step": 5864
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.247527228596684,
+      "learning_rate": 1.5797306621286757e-07,
+      "loss": 0.8295,
+      "step": 5865
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6125036587045254,
+      "learning_rate": 1.5705032581105563e-07,
+      "loss": 0.8281,
+      "step": 5866
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.388090769253597,
+      "learning_rate": 1.561302669000586e-07,
+      "loss": 0.7455,
+      "step": 5867
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6150775709923049,
+      "learning_rate": 1.5521288973052274e-07,
+      "loss": 0.7759,
+      "step": 5868
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4672832343014335,
+      "learning_rate": 1.5429819455237137e-07,
+      "loss": 0.7622,
+      "step": 5869
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3014316808506072,
+      "learning_rate": 1.5338618161478857e-07,
+      "loss": 0.7879,
+      "step": 5870
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3312620941052646,
+      "learning_rate": 1.5247685116623335e-07,
+      "loss": 0.7632,
+      "step": 5871
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3173266735373872,
+      "learning_rate": 1.5157020345443195e-07,
+      "loss": 0.8153,
+      "step": 5872
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4105801473563526,
+      "learning_rate": 1.5066623872638242e-07,
+      "loss": 0.7818,
+      "step": 5873
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4172251212164064,
+      "learning_rate": 1.497649572283466e-07,
+      "loss": 0.8219,
+      "step": 5874
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7075194494363592,
+      "learning_rate": 1.4886635920586036e-07,
+      "loss": 0.7655,
+      "step": 5875
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3608861038672677,
+      "learning_rate": 1.479704449037256e-07,
+      "loss": 0.8056,
+      "step": 5876
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3032055929073807,
+      "learning_rate": 1.4707721456601486e-07,
+      "loss": 0.7612,
+      "step": 5877
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2737091672766117,
+      "learning_rate": 1.46186668436068e-07,
+      "loss": 0.828,
+      "step": 5878
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4108201191953755,
+      "learning_rate": 1.4529880675649534e-07,
+      "loss": 0.8359,
+      "step": 5879
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3237744474967679,
+      "learning_rate": 1.444136297691734e-07,
+      "loss": 0.7642,
+      "step": 5880
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5203590957346371,
+      "learning_rate": 1.435311377152493e-07,
+      "loss": 0.8492,
+      "step": 5881
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.378149566571992,
+      "learning_rate": 1.426513308351385e-07,
+      "loss": 0.8823,
+      "step": 5882
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3302852095499067,
+      "learning_rate": 1.4177420936852482e-07,
+      "loss": 0.8698,
+      "step": 5883
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.354800238092628,
+      "learning_rate": 1.4089977355436045e-07,
+      "loss": 0.7549,
+      "step": 5884
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.428505610941251,
+      "learning_rate": 1.4002802363086486e-07,
+      "loss": 0.8756,
+      "step": 5885
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.329027119162686,
+      "learning_rate": 1.3915895983552806e-07,
+      "loss": 0.8298,
+      "step": 5886
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.721343783945078,
+      "learning_rate": 1.3829258240510624e-07,
+      "loss": 0.8259,
+      "step": 5887
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5230271593741773,
+      "learning_rate": 1.374288915756228e-07,
+      "loss": 0.8706,
+      "step": 5888
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3166890735485766,
+      "learning_rate": 1.3656788758237504e-07,
+      "loss": 0.8202,
+      "step": 5889
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4521925091549335,
+      "learning_rate": 1.3570957065991987e-07,
+      "loss": 0.7959,
+      "step": 5890
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.488225433036324,
+      "learning_rate": 1.3485394104209015e-07,
+      "loss": 0.8386,
+      "step": 5891
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5876528050394962,
+      "learning_rate": 1.340009989619806e-07,
+      "loss": 0.8517,
+      "step": 5892
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6780228672079986,
+      "learning_rate": 1.3315074465195533e-07,
+      "loss": 0.8263,
+      "step": 5893
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3947894678084862,
+      "learning_rate": 1.3230317834365013e-07,
+      "loss": 0.8081,
+      "step": 5894
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3386729772210906,
+      "learning_rate": 1.3145830026796368e-07,
+      "loss": 0.8344,
+      "step": 5895
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3676530992070035,
+      "learning_rate": 1.3061611065506409e-07,
+      "loss": 0.7983,
+      "step": 5896
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.376522024793113,
+      "learning_rate": 1.2977660973438667e-07,
+      "loss": 0.8538,
+      "step": 5897
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4251120228011547,
+      "learning_rate": 1.2893979773463516e-07,
+      "loss": 0.8602,
+      "step": 5898
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3524789476634684,
+      "learning_rate": 1.2810567488378055e-07,
+      "loss": 0.8322,
+      "step": 5899
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.419861557425791,
+      "learning_rate": 1.2727424140905998e-07,
+      "loss": 0.8024,
+      "step": 5900
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6305277412976746,
+      "learning_rate": 1.264454975369789e-07,
+      "loss": 0.792,
+      "step": 5901
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.49304891607622,
+      "learning_rate": 1.2561944349331223e-07,
+      "loss": 0.8088,
+      "step": 5902
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.435138526240153,
+      "learning_rate": 1.247960795030967e-07,
+      "loss": 0.7957,
+      "step": 5903
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.433288698326768,
+      "learning_rate": 1.239754057906406e-07,
+      "loss": 0.7976,
+      "step": 5904
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4804757326912585,
+      "learning_rate": 1.2315742257951847e-07,
+      "loss": 0.7664,
+      "step": 5905
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2560605396744795,
+      "learning_rate": 1.22342130092572e-07,
+      "loss": 0.8395,
+      "step": 5906
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4650800145666492,
+      "learning_rate": 1.215295285519069e-07,
+      "loss": 0.8587,
+      "step": 5907
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2593450191575843,
+      "learning_rate": 1.2071961817890053e-07,
+      "loss": 0.833,
+      "step": 5908
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4699047810677959,
+      "learning_rate": 1.1991239919419529e-07,
+      "loss": 0.8585,
+      "step": 5909
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3490132541749875,
+      "learning_rate": 1.1910787181769745e-07,
+      "loss": 0.7749,
+      "step": 5910
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.567113981633362,
+      "learning_rate": 1.1830603626858394e-07,
+      "loss": 0.8151,
+      "step": 5911
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3243047196127378,
+      "learning_rate": 1.175068927652967e-07,
+      "loss": 0.8241,
+      "step": 5912
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4202187344578385,
+      "learning_rate": 1.1671044152554378e-07,
+      "loss": 0.8026,
+      "step": 5913
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2572403915015684,
+      "learning_rate": 1.1591668276630274e-07,
+      "loss": 0.8172,
+      "step": 5914
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5306349037121818,
+      "learning_rate": 1.1512561670381172e-07,
+      "loss": 0.8662,
+      "step": 5915
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.491862956424143,
+      "learning_rate": 1.1433724355358167e-07,
+      "loss": 0.852,
+      "step": 5916
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.8334847579658969,
+      "learning_rate": 1.1355156353038743e-07,
+      "loss": 0.3109,
+      "step": 5917
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3824879779382424,
+      "learning_rate": 1.127685768482667e-07,
+      "loss": 0.7762,
+      "step": 5918
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1403659251171714,
+      "learning_rate": 1.1198828372052994e-07,
+      "loss": 0.8054,
+      "step": 5919
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5357591343676864,
+      "learning_rate": 1.1121068435974935e-07,
+      "loss": 0.8803,
+      "step": 5920
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3189133462148945,
+      "learning_rate": 1.1043577897776547e-07,
+      "loss": 0.8414,
+      "step": 5921
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.36018412176385,
+      "learning_rate": 1.0966356778568055e-07,
+      "loss": 0.7611,
+      "step": 5922
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4635754988760137,
+      "learning_rate": 1.0889405099386962e-07,
+      "loss": 0.8387,
+      "step": 5923
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4805583134134297,
+      "learning_rate": 1.0812722881197058e-07,
+      "loss": 0.8366,
+      "step": 5924
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.494126185802427,
+      "learning_rate": 1.0736310144888296e-07,
+      "loss": 0.7916,
+      "step": 5925
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4274387053467852,
+      "learning_rate": 1.0660166911277914e-07,
+      "loss": 0.8019,
+      "step": 5926
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3518156182581955,
+      "learning_rate": 1.0584293201109541e-07,
+      "loss": 0.8356,
+      "step": 5927
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5696182032737453,
+      "learning_rate": 1.0508689035052977e-07,
+      "loss": 0.8265,
+      "step": 5928
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2474297443895914,
+      "learning_rate": 1.0433354433705078e-07,
+      "loss": 0.8143,
+      "step": 5929
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.581888202393035,
+      "learning_rate": 1.0358289417588874e-07,
+      "loss": 0.7142,
+      "step": 5930
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4667049054038683,
+      "learning_rate": 1.0283494007154448e-07,
+      "loss": 0.7494,
+      "step": 5931
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3020620999194406,
+      "learning_rate": 1.0208968222777838e-07,
+      "loss": 0.8307,
+      "step": 5932
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4478430464836454,
+      "learning_rate": 1.0134712084762022e-07,
+      "loss": 0.7827,
+      "step": 5933
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3773379106121946,
+      "learning_rate": 1.0060725613336375e-07,
+      "loss": 0.7321,
+      "step": 5934
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.776565208323955,
+      "learning_rate": 9.987008828656997e-08,
+      "loss": 0.8597,
+      "step": 5935
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3488961397075148,
+      "learning_rate": 9.913561750806378e-08,
+      "loss": 0.7394,
+      "step": 5936
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4416859951031693,
+      "learning_rate": 9.84038439979329e-08,
+      "loss": 0.817,
+      "step": 5937
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4380756850306522,
+      "learning_rate": 9.767476795553454e-08,
+      "loss": 0.8757,
+      "step": 5938
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.695429294711731,
+      "learning_rate": 9.69483895794876e-08,
+      "loss": 0.7779,
+      "step": 5939
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2509237966196682,
+      "learning_rate": 9.622470906767933e-08,
+      "loss": 0.7999,
+      "step": 5940
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.350896842295312,
+      "learning_rate": 9.550372661725982e-08,
+      "loss": 0.8087,
+      "step": 5941
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3849225076441671,
+      "learning_rate": 9.478544242464415e-08,
+      "loss": 0.8576,
+      "step": 5942
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2556512594942453,
+      "learning_rate": 9.40698566855125e-08,
+      "loss": 0.7988,
+      "step": 5943
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2394643352962174,
+      "learning_rate": 9.335696959481e-08,
+      "loss": 0.7981,
+      "step": 5944
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.404950753675489,
+      "learning_rate": 9.264678134674687e-08,
+      "loss": 0.8559,
+      "step": 5945
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3717047085463214,
+      "learning_rate": 9.193929213480057e-08,
+      "loss": 0.8441,
+      "step": 5946
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6095642087413555,
+      "learning_rate": 9.123450215170693e-08,
+      "loss": 0.8259,
+      "step": 5947
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2763733290894483,
+      "learning_rate": 9.053241158947123e-08,
+      "loss": 0.8115,
+      "step": 5948
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.247716192007064,
+      "learning_rate": 8.983302063936272e-08,
+      "loss": 0.8302,
+      "step": 5949
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.8542310500982013,
+      "learning_rate": 8.913632949191564e-08,
+      "loss": 0.304,
+      "step": 5950
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4659786888529376,
+      "learning_rate": 8.844233833692595e-08,
+      "loss": 0.7214,
+      "step": 5951
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.623691449780856,
+      "learning_rate": 8.775104736345796e-08,
+      "loss": 0.8332,
+      "step": 5952
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3750660972530822,
+      "learning_rate": 8.70624567598366e-08,
+      "loss": 0.8126,
+      "step": 5953
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.295990380084933,
+      "learning_rate": 8.637656671365402e-08,
+      "loss": 0.7977,
+      "step": 5954
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3852379885039,
+      "learning_rate": 8.56933774117652e-08,
+      "loss": 0.8191,
+      "step": 5955
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2844819938716439,
+      "learning_rate": 8.501288904029014e-08,
+      "loss": 0.8717,
+      "step": 5956
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4356604583076804,
+      "learning_rate": 8.433510178461168e-08,
+      "loss": 0.7886,
+      "step": 5957
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2211143045287962,
+      "learning_rate": 8.366001582937988e-08,
+      "loss": 0.8174,
+      "step": 5958
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4134980591334174,
+      "learning_rate": 8.29876313585043e-08,
+      "loss": 0.8097,
+      "step": 5959
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4097338384573557,
+      "learning_rate": 8.231794855516173e-08,
+      "loss": 0.8465,
+      "step": 5960
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4082632898501515,
+      "learning_rate": 8.165096760179181e-08,
+      "loss": 0.7845,
+      "step": 5961
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4579992211019106,
+      "learning_rate": 8.09866886801014e-08,
+      "loss": 0.7439,
+      "step": 5962
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.40576536384155,
+      "learning_rate": 8.032511197105353e-08,
+      "loss": 0.7708,
+      "step": 5963
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5105139369446332,
+      "learning_rate": 7.966623765488513e-08,
+      "loss": 0.8049,
+      "step": 5964
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6613008205619584,
+      "learning_rate": 7.901006591108817e-08,
+      "loss": 0.8259,
+      "step": 5965
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.8131023040634394,
+      "learning_rate": 7.83565969184219e-08,
+      "loss": 0.3392,
+      "step": 5966
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.465097045615563,
+      "learning_rate": 7.770583085491168e-08,
+      "loss": 0.764,
+      "step": 5967
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2741823051451713,
+      "learning_rate": 7.705776789784237e-08,
+      "loss": 0.8095,
+      "step": 5968
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2719255229981568,
+      "learning_rate": 7.641240822376495e-08,
+      "loss": 0.8299,
+      "step": 5969
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0409545486884109,
+      "learning_rate": 7.576975200849212e-08,
+      "loss": 0.3578,
+      "step": 5970
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.474968810832697,
+      "learning_rate": 7.512979942710163e-08,
+      "loss": 0.8277,
+      "step": 5971
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4946718337515887,
+      "learning_rate": 7.449255065393624e-08,
+      "loss": 0.7364,
+      "step": 5972
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4117878512225581,
+      "learning_rate": 7.385800586259595e-08,
+      "loss": 0.8121,
+      "step": 5973
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5797484587992223,
+      "learning_rate": 7.32261652259525e-08,
+      "loss": 0.7546,
+      "step": 5974
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.387185293616104,
+      "learning_rate": 7.259702891613374e-08,
+      "loss": 0.8616,
+      "step": 5975
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9568487670474821,
+      "learning_rate": 7.19705971045348e-08,
+      "loss": 0.8477,
+      "step": 5976
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4321704585995494,
+      "learning_rate": 7.134686996181361e-08,
+      "loss": 0.8604,
+      "step": 5977
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5060501515851128,
+      "learning_rate": 7.07258476578887e-08,
+      "loss": 0.8725,
+      "step": 5978
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9044262904670698,
+      "learning_rate": 7.010753036194584e-08,
+      "loss": 0.3279,
+      "step": 5979
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6601701113178298,
+      "learning_rate": 6.949191824243028e-08,
+      "loss": 0.7786,
+      "step": 5980
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.8763852330342072,
+      "learning_rate": 6.887901146705344e-08,
+      "loss": 0.3228,
+      "step": 5981
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.1411771368753003,
+      "learning_rate": 6.82688102027862e-08,
+      "loss": 0.7869,
+      "step": 5982
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.650915739499833,
+      "learning_rate": 6.766131461586445e-08,
+      "loss": 0.881,
+      "step": 5983
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4097177148543596,
+      "learning_rate": 6.705652487178693e-08,
+      "loss": 0.8293,
+      "step": 5984
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4450980633097819,
+      "learning_rate": 6.645444113531519e-08,
+      "loss": 0.8028,
+      "step": 5985
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3876390338360636,
+      "learning_rate": 6.585506357047466e-08,
+      "loss": 0.6795,
+      "step": 5986
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.509803181165102,
+      "learning_rate": 6.525839234055032e-08,
+      "loss": 0.7796,
+      "step": 5987
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2895552415925915,
+      "learning_rate": 6.46644276080921e-08,
+      "loss": 0.8759,
+      "step": 5988
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.315683841578733,
+      "learning_rate": 6.407316953491393e-08,
+      "loss": 0.8095,
+      "step": 5989
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6831936006681572,
+      "learning_rate": 6.348461828208919e-08,
+      "loss": 0.9101,
+      "step": 5990
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3062773471803266,
+      "learning_rate": 6.289877400995625e-08,
+      "loss": 0.8375,
+      "step": 5991
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3658429434827144,
+      "learning_rate": 6.231563687811526e-08,
+      "loss": 0.8275,
+      "step": 5992
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4373605345627107,
+      "learning_rate": 6.173520704542802e-08,
+      "loss": 0.7993,
+      "step": 5993
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3827320105085605,
+      "learning_rate": 6.115748467002136e-08,
+      "loss": 0.749,
+      "step": 5994
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.456378912182639,
+      "learning_rate": 6.058246990928051e-08,
+      "loss": 0.8514,
+      "step": 5995
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4169548786911967,
+      "learning_rate": 6.001016291985795e-08,
+      "loss": 0.7819,
+      "step": 5996
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3853744680268891,
+      "learning_rate": 5.944056385766339e-08,
+      "loss": 0.7561,
+      "step": 5997
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5661707615483282,
+      "learning_rate": 5.887367287787271e-08,
+      "loss": 0.8222,
+      "step": 5998
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5322182626541996,
+      "learning_rate": 5.8309490134921265e-08,
+      "loss": 0.8458,
+      "step": 5999
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.275804848899459,
+      "learning_rate": 5.774801578251055e-08,
+      "loss": 0.8712,
+      "step": 6000
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2960530952725058,
+      "learning_rate": 5.718924997359932e-08,
+      "loss": 0.867,
+      "step": 6001
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2131069530999965,
+      "learning_rate": 5.663319286041136e-08,
+      "loss": 0.7631,
+      "step": 6002
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4120027427986936,
+      "learning_rate": 5.6079844594433276e-08,
+      "loss": 0.807,
+      "step": 6003
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3749311007230216,
+      "learning_rate": 5.552920532641004e-08,
+      "loss": 0.8161,
+      "step": 6004
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.39230097143429,
+      "learning_rate": 5.498127520635277e-08,
+      "loss": 0.7324,
+      "step": 6005
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5222429155496597,
+      "learning_rate": 5.4436054383532054e-08,
+      "loss": 0.7997,
+      "step": 6006
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4188826127714713,
+      "learning_rate": 5.389354300648131e-08,
+      "loss": 0.8076,
+      "step": 6007
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.437679005574304,
+      "learning_rate": 5.3353741222995634e-08,
+      "loss": 0.7932,
+      "step": 6008
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.351626526619759,
+      "learning_rate": 5.281664918013185e-08,
+      "loss": 0.8316,
+      "step": 6009
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3894137568362077,
+      "learning_rate": 5.2282267024207355e-08,
+      "loss": 0.7973,
+      "step": 6010
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4942096875005764,
+      "learning_rate": 5.1750594900805697e-08,
+      "loss": 0.8346,
+      "step": 6011
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1696680674096718,
+      "learning_rate": 5.1221632954765455e-08,
+      "loss": 0.7659,
+      "step": 6012
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4419059430015377,
+      "learning_rate": 5.069538133019247e-08,
+      "loss": 0.7898,
+      "step": 6013
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3167764846289038,
+      "learning_rate": 5.017184017045207e-08,
+      "loss": 0.7442,
+      "step": 6014
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.633980105415254,
+      "learning_rate": 4.965100961817126e-08,
+      "loss": 0.8692,
+      "step": 6015
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5289417612126126,
+      "learning_rate": 4.913288981523878e-08,
+      "loss": 0.8913,
+      "step": 6016
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3586197082568885,
+      "learning_rate": 4.8617480902802826e-08,
+      "loss": 0.7637,
+      "step": 6017
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5081706498786276,
+      "learning_rate": 4.8104783021277746e-08,
+      "loss": 0.888,
+      "step": 6018
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4960566682926195,
+      "learning_rate": 4.759479631033514e-08,
+      "loss": 0.8311,
+      "step": 6019
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4447719645743997,
+      "learning_rate": 4.708752090890944e-08,
+      "loss": 0.7546,
+      "step": 6020
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5115737130823927,
+      "learning_rate": 4.6582956955196765e-08,
+      "loss": 0.8294,
+      "step": 6021
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5736588424209623,
+      "learning_rate": 4.608110458665382e-08,
+      "loss": 0.8316,
+      "step": 6022
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5378659203007923,
+      "learning_rate": 4.558196394000014e-08,
+      "loss": 0.8298,
+      "step": 6023
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.616295368893459,
+      "learning_rate": 4.508553515121472e-08,
+      "loss": 0.8117,
+      "step": 6024
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5298792855446002,
+      "learning_rate": 4.4591818355538276e-08,
+      "loss": 0.8507,
+      "step": 6025
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4244093526252097,
+      "learning_rate": 4.41008136874721e-08,
+      "loss": 0.765,
+      "step": 6026
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3393998759296428,
+      "learning_rate": 4.361252128078031e-08,
+      "loss": 0.8079,
+      "step": 6027
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6515032465743607,
+      "learning_rate": 4.31269412684876e-08,
+      "loss": 0.7501,
+      "step": 6028
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5129715172948077,
+      "learning_rate": 4.264407378287927e-08,
+      "loss": 0.8271,
+      "step": 6029
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3527769513029613,
+      "learning_rate": 4.216391895550121e-08,
+      "loss": 0.8913,
+      "step": 6030
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3918578410493412,
+      "learning_rate": 4.168647691716099e-08,
+      "loss": 0.8635,
+      "step": 6031
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5348407843467056,
+      "learning_rate": 4.12117477979257e-08,
+      "loss": 0.8776,
+      "step": 6032
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8687932235725033,
+      "learning_rate": 4.0739731727127416e-08,
+      "loss": 0.8466,
+      "step": 6033
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.532864897772446,
+      "learning_rate": 4.027042883335441e-08,
+      "loss": 0.7994,
+      "step": 6034
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2850539100040388,
+      "learning_rate": 3.980383924445774e-08,
+      "loss": 0.8858,
+      "step": 6035
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4409219786362937,
+      "learning_rate": 3.933996308755017e-08,
+      "loss": 0.8464,
+      "step": 6036
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0407069090233243,
+      "learning_rate": 3.887880048900394e-08,
+      "loss": 0.8182,
+      "step": 6037
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5844661118852745,
+      "learning_rate": 3.8420351574453005e-08,
+      "loss": 0.8188,
+      "step": 6038
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3785471624245913,
+      "learning_rate": 3.796461646878968e-08,
+      "loss": 0.7843,
+      "step": 6039
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4713201419024522,
+      "learning_rate": 3.751159529617021e-08,
+      "loss": 0.8047,
+      "step": 6040
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3056823940316953,
+      "learning_rate": 3.706128818001031e-08,
+      "loss": 0.8285,
+      "step": 6041
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5702077067932025,
+      "learning_rate": 3.6613695242984085e-08,
+      "loss": 0.8139,
+      "step": 6042
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3816812792656656,
+      "learning_rate": 3.616881660703064e-08,
+      "loss": 0.8125,
+      "step": 6043
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4030506954133668,
+      "learning_rate": 3.5726652393346385e-08,
+      "loss": 0.7792,
+      "step": 6044
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4277002614925574,
+      "learning_rate": 3.528720272238828e-08,
+      "loss": 0.8238,
+      "step": 6045
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4315246317765293,
+      "learning_rate": 3.485046771387612e-08,
+      "loss": 0.8171,
+      "step": 6046
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5521084057513086,
+      "learning_rate": 3.441644748678585e-08,
+      "loss": 0.7468,
+      "step": 6047
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4959855862254767,
+      "learning_rate": 3.398514215935955e-08,
+      "loss": 0.7329,
+      "step": 6048
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5231993256745697,
+      "learning_rate": 3.355655184909545e-08,
+      "loss": 0.8365,
+      "step": 6049
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.394047115473417,
+      "learning_rate": 3.313067667275238e-08,
+      "loss": 0.8588,
+      "step": 6050
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.430275989406736,
+      "learning_rate": 3.270751674635197e-08,
+      "loss": 0.8711,
+      "step": 6051
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3594894210300184,
+      "learning_rate": 3.228707218517313e-08,
+      "loss": 0.8275,
+      "step": 6052
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4291910435840587,
+      "learning_rate": 3.186934310375866e-08,
+      "loss": 0.8332,
+      "step": 6053
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2868411359157137,
+      "learning_rate": 3.1454329615907554e-08,
+      "loss": 0.6927,
+      "step": 6054
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3345822074021731,
+      "learning_rate": 3.104203183468157e-08,
+      "loss": 0.7665,
+      "step": 6055
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4961335118390904,
+      "learning_rate": 3.0632449872401994e-08,
+      "loss": 0.8143,
+      "step": 6056
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4461928798625012,
+      "learning_rate": 3.0225583840650665e-08,
+      "loss": 0.8178,
+      "step": 6057
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5197553873852232,
+      "learning_rate": 2.982143385026892e-08,
+      "loss": 0.8036,
+      "step": 6058
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4189539191712903,
+      "learning_rate": 2.9420000011357585e-08,
+      "loss": 0.7886,
+      "step": 6059
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.596172604547849,
+      "learning_rate": 2.9021282433279173e-08,
+      "loss": 0.7904,
+      "step": 6060
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3080037964953877,
+      "learning_rate": 2.8625281224654578e-08,
+      "loss": 0.8707,
+      "step": 6061
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.769712903001942,
+      "learning_rate": 2.8231996493366387e-08,
+      "loss": 0.8191,
+      "step": 6062
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5817813805477723,
+      "learning_rate": 2.7841428346556676e-08,
+      "loss": 0.8729,
+      "step": 6063
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5928000840250816,
+      "learning_rate": 2.745357689062478e-08,
+      "loss": 0.722,
+      "step": 6064
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4393314468011846,
+      "learning_rate": 2.706844223123395e-08,
+      "loss": 0.7586,
+      "step": 6065
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2804768754946843,
+      "learning_rate": 2.6686024473304706e-08,
+      "loss": 0.7986,
+      "step": 6066
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6364664734585934,
+      "learning_rate": 2.6306323721018156e-08,
+      "loss": 0.8213,
+      "step": 6067
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4654030533192521,
+      "learning_rate": 2.5929340077816e-08,
+      "loss": 0.7619,
+      "step": 6068
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5345204495134486,
+      "learning_rate": 2.555507364639831e-08,
+      "loss": 0.7735,
+      "step": 6069
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3764229274509767,
+      "learning_rate": 2.5183524528725744e-08,
+      "loss": 0.7974,
+      "step": 6070
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3556201957395027,
+      "learning_rate": 2.481469282601845e-08,
+      "loss": 0.7894,
+      "step": 6071
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8722287818885825,
+      "learning_rate": 2.444857863875605e-08,
+      "loss": 0.7785,
+      "step": 6072
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3450157101216205,
+      "learning_rate": 2.408518206667876e-08,
+      "loss": 0.831,
+      "step": 6073
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3797760704152304,
+      "learning_rate": 2.3724503208786276e-08,
+      "loss": 0.7917,
+      "step": 6074
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4672671693026282,
+      "learning_rate": 2.3366542163336668e-08,
+      "loss": 0.7926,
+      "step": 6075
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4267259107256909,
+      "learning_rate": 2.3011299027847488e-08,
+      "loss": 0.8005,
+      "step": 6076
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4567062760819132,
+      "learning_rate": 2.2658773899097986e-08,
+      "loss": 0.7699,
+      "step": 6077
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5142317154996165,
+      "learning_rate": 2.230896687312578e-08,
+      "loss": 0.8172,
+      "step": 6078
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5882842378405284,
+      "learning_rate": 2.196187804522798e-08,
+      "loss": 0.8247,
+      "step": 6079
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4677942392546506,
+      "learning_rate": 2.1617507509960058e-08,
+      "loss": 0.7849,
+      "step": 6080
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4774514970971258,
+      "learning_rate": 2.1275855361140297e-08,
+      "loss": 0.7637,
+      "step": 6081
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3527214781113057,
+      "learning_rate": 2.0936921691842028e-08,
+      "loss": 0.7576,
+      "step": 6082
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4375647181734748,
+      "learning_rate": 2.0600706594400278e-08,
+      "loss": 0.7837,
+      "step": 6083
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5067429893984299,
+      "learning_rate": 2.0267210160409557e-08,
+      "loss": 0.8706,
+      "step": 6084
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3868167916415504,
+      "learning_rate": 1.9936432480723854e-08,
+      "loss": 0.8293,
+      "step": 6085
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5452650800589995,
+      "learning_rate": 1.9608373645456648e-08,
+      "loss": 0.8111,
+      "step": 6086
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3868176277262179,
+      "learning_rate": 1.9283033743978663e-08,
+      "loss": 0.7749,
+      "step": 6087
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2735195250510247,
+      "learning_rate": 1.896041286492345e-08,
+      "loss": 0.7893,
+      "step": 6088
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4559457163774372,
+      "learning_rate": 1.86405110961807e-08,
+      "loss": 0.808,
+      "step": 6089
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3190427658426584,
+      "learning_rate": 1.8323328524899597e-08,
+      "loss": 0.8573,
+      "step": 6090
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4582286151402442,
+      "learning_rate": 1.8008865237491014e-08,
+      "loss": 0.7686,
+      "step": 6091
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3004729310777106,
+      "learning_rate": 1.7697121319621978e-08,
+      "loss": 0.8467,
+      "step": 6092
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.405625089237877,
+      "learning_rate": 1.7388096856221227e-08,
+      "loss": 0.8259,
+      "step": 6093
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2260298377192678,
+      "learning_rate": 1.7081791931475855e-08,
+      "loss": 0.7753,
+      "step": 6094
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4916007551046446,
+      "learning_rate": 1.677820662883134e-08,
+      "loss": 0.8167,
+      "step": 6095
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6494424805076153,
+      "learning_rate": 1.647734103099152e-08,
+      "loss": 0.8343,
+      "step": 6096
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2326458864916343,
+      "learning_rate": 1.6179195219921952e-08,
+      "loss": 0.8227,
+      "step": 6097
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.350790440680907,
+      "learning_rate": 1.588376927684432e-08,
+      "loss": 0.7803,
+      "step": 6098
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2189293203724412,
+      "learning_rate": 1.5591063282242026e-08,
+      "loss": 0.8056,
+      "step": 6099
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4244769701920426,
+      "learning_rate": 1.530107731585684e-08,
+      "loss": 0.811,
+      "step": 6100
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4914891727239066,
+      "learning_rate": 1.5013811456687787e-08,
+      "loss": 0.7379,
+      "step": 6101
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4561880896162003,
+      "learning_rate": 1.4729265782993384e-08,
+      "loss": 0.7709,
+      "step": 6102
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.243431717256865,
+      "learning_rate": 1.4447440372292732e-08,
+      "loss": 0.7952,
+      "step": 6103
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.446095102016903,
+      "learning_rate": 1.41683353013633e-08,
+      "loss": 0.7761,
+      "step": 6104
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.397530094170945,
+      "learning_rate": 1.3891950646239827e-08,
+      "loss": 0.8049,
+      "step": 6105
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3857259951042646,
+      "learning_rate": 1.3618286482218745e-08,
+      "loss": 0.797,
+      "step": 6106
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1867196246615432,
+      "learning_rate": 1.3347342883851532e-08,
+      "loss": 0.8376,
+      "step": 6107
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4819705336364355,
+      "learning_rate": 1.3079119924952477e-08,
+      "loss": 0.8393,
+      "step": 6108
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2937477896735987,
+      "learning_rate": 1.2813617678592017e-08,
+      "loss": 0.6799,
+      "step": 6109
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2776305182263576,
+      "learning_rate": 1.2550836217101182e-08,
+      "loss": 0.7636,
+      "step": 6110
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6383007320360798,
+      "learning_rate": 1.2290775612067151e-08,
+      "loss": 0.7967,
+      "step": 6111
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4671573405522675,
+      "learning_rate": 1.2033435934338811e-08,
+      "loss": 0.8732,
+      "step": 6112
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7301246309342323,
+      "learning_rate": 1.1778817254022301e-08,
+      "loss": 0.3319,
+      "step": 6113
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4879643611008029,
+      "learning_rate": 1.1526919640483247e-08,
+      "loss": 0.8405,
+      "step": 6114
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9903820302078314,
+      "learning_rate": 1.1277743162345644e-08,
+      "loss": 0.3122,
+      "step": 6115
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3711695296927295,
+      "learning_rate": 1.103128788749075e-08,
+      "loss": 0.8504,
+      "step": 6116
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5351068195121127,
+      "learning_rate": 1.0787553883061519e-08,
+      "loss": 0.8389,
+      "step": 6117
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4588913881526042,
+      "learning_rate": 1.0546541215455952e-08,
+      "loss": 0.8379,
+      "step": 6118
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8925496498815968,
+      "learning_rate": 1.0308249950333749e-08,
+      "loss": 0.8005,
+      "step": 6119
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5061494545400904,
+      "learning_rate": 1.0072680152611869e-08,
+      "loss": 0.7942,
+      "step": 6120
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.8543378478037552,
+      "learning_rate": 9.839831886465644e-09,
+      "loss": 0.3638,
+      "step": 6121
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.331011318751303,
+      "learning_rate": 9.60970521532878e-09,
+      "loss": 0.8169,
+      "step": 6122
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2828089124487685,
+      "learning_rate": 9.382300201896676e-09,
+      "loss": 0.7759,
+      "step": 6123
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.8686874520860466,
+      "learning_rate": 9.157616908117562e-09,
+      "loss": 0.3333,
+      "step": 6124
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3888901218435323,
+      "learning_rate": 8.935655395203579e-09,
+      "loss": 0.7946,
+      "step": 6125
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4573400741590914,
+      "learning_rate": 8.716415723621918e-09,
+      "loss": 0.7958,
+      "step": 6126
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.483994570719936,
+      "learning_rate": 8.499897953100355e-09,
+      "loss": 0.7756,
+      "step": 6127
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8224020392741154,
+      "learning_rate": 8.28610214262393e-09,
+      "loss": 0.829,
+      "step": 6128
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.262798375189001,
+      "learning_rate": 8.075028350436054e-09,
+      "loss": 0.783,
+      "step": 6129
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1642630087575414,
+      "learning_rate": 7.866676634039617e-09,
+      "loss": 0.824,
+      "step": 6130
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5237177737718797,
+      "learning_rate": 7.66104705019588e-09,
+      "loss": 0.858,
+      "step": 6131
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4149524011921286,
+      "learning_rate": 7.45813965492337e-09,
+      "loss": 0.7364,
+      "step": 6132
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3735161572175374,
+      "learning_rate": 7.257954503498976e-09,
+      "loss": 0.8095,
+      "step": 6133
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.8473921839141729,
+      "learning_rate": 7.060491650459078e-09,
+      "loss": 0.3238,
+      "step": 6134
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4612953103579018,
+      "learning_rate": 6.8657511495984205e-09,
+      "loss": 0.8739,
+      "step": 6135
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3094344299731606,
+      "learning_rate": 6.673733053970122e-09,
+      "loss": 0.7723,
+      "step": 6136
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3434094348142434,
+      "learning_rate": 6.4844374158834485e-09,
+      "loss": 0.8265,
+      "step": 6137
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3199177337354602,
+      "learning_rate": 6.297864286910482e-09,
+      "loss": 0.7932,
+      "step": 6138
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4078135364853601,
+      "learning_rate": 6.114013717876121e-09,
+      "loss": 0.8214,
+      "step": 6139
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4473654201659836,
+      "learning_rate": 5.9328857588680785e-09,
+      "loss": 0.8276,
+      "step": 6140
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.319312787929375,
+      "learning_rate": 5.754480459229106e-09,
+      "loss": 0.7507,
+      "step": 6141
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2457349748029742,
+      "learning_rate": 5.5787978675636566e-09,
+      "loss": 0.7601,
+      "step": 6142
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3780636201892504,
+      "learning_rate": 5.405838031731225e-09,
+      "loss": 0.8638,
+      "step": 6143
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8567096917267634,
+      "learning_rate": 5.235600998850787e-09,
+      "loss": 0.8195,
+      "step": 6144
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2323067013536686,
+      "learning_rate": 5.068086815300799e-09,
+      "loss": 0.8351,
+      "step": 6145
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.827715935755992,
+      "learning_rate": 4.9032955267158675e-09,
+      "loss": 0.3583,
+      "step": 6146
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5427901550944183,
+      "learning_rate": 4.741227177988972e-09,
+      "loss": 0.7252,
+      "step": 6147
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3712506707102023,
+      "learning_rate": 4.581881813272571e-09,
+      "loss": 0.8916,
+      "step": 6148
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4310982562815822,
+      "learning_rate": 4.4252594759774945e-09,
+      "loss": 0.8746,
+      "step": 6149
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5608320727557936,
+      "learning_rate": 4.271360208770725e-09,
+      "loss": 0.794,
+      "step": 6150
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4210714660898054,
+      "learning_rate": 4.120184053579834e-09,
+      "loss": 0.8428,
+      "step": 6151
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4890500233258597,
+      "learning_rate": 3.971731051588545e-09,
+      "loss": 0.7812,
+      "step": 6152
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2871833329232851,
+      "learning_rate": 3.826001243240063e-09,
+      "loss": 0.8156,
+      "step": 6153
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.328261716201147,
+      "learning_rate": 3.682994668234852e-09,
+      "loss": 0.7434,
+      "step": 6154
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4325254198988706,
+      "learning_rate": 3.542711365531748e-09,
+      "loss": 0.8456,
+      "step": 6155
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3260906999433153,
+      "learning_rate": 3.405151373347959e-09,
+      "loss": 0.7743,
+      "step": 6156
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3132202508138637,
+      "learning_rate": 3.270314729159063e-09,
+      "loss": 0.887,
+      "step": 6157
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3073780337540433,
+      "learning_rate": 3.138201469697899e-09,
+      "loss": 0.7788,
+      "step": 6158
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2896284566045197,
+      "learning_rate": 3.008811630955677e-09,
+      "loss": 0.7936,
+      "step": 6159
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5002828842883926,
+      "learning_rate": 2.882145248181978e-09,
+      "loss": 0.787,
+      "step": 6160
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2962109469921965,
+      "learning_rate": 2.7582023558847537e-09,
+      "loss": 0.8232,
+      "step": 6161
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5180957910363062,
+      "learning_rate": 2.6369829878281074e-09,
+      "loss": 0.8467,
+      "step": 6162
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7372282031733024,
+      "learning_rate": 2.5184871770356224e-09,
+      "loss": 0.7418,
+      "step": 6163
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4654194714554771,
+      "learning_rate": 2.4027149557903638e-09,
+      "loss": 0.8389,
+      "step": 6164
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3683123593280433,
+      "learning_rate": 2.2896663556304378e-09,
+      "loss": 0.7355,
+      "step": 6165
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3456378032986156,
+      "learning_rate": 2.1793414073545407e-09,
+      "loss": 0.8504,
+      "step": 6166
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5512644509822757,
+      "learning_rate": 2.0717401410164097e-09,
+      "loss": 0.8361,
+      "step": 6167
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.389581947451672,
+      "learning_rate": 1.9668625859314838e-09,
+      "loss": 0.8439,
+      "step": 6168
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3836868345287536,
+      "learning_rate": 1.8647087706702425e-09,
+      "loss": 0.9227,
+      "step": 6169
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3243755069156071,
+      "learning_rate": 1.7652787230637569e-09,
+      "loss": 0.8093,
+      "step": 6170
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8334673913570718,
+      "learning_rate": 1.6685724701970274e-09,
+      "loss": 0.742,
+      "step": 6171
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3236095326788275,
+      "learning_rate": 1.5745900384167567e-09,
+      "loss": 0.8124,
+      "step": 6172
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6421420406379479,
+      "learning_rate": 1.4833314533269084e-09,
+      "loss": 0.7681,
+      "step": 6173
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4591228363232178,
+      "learning_rate": 1.3947967397887064e-09,
+      "loss": 0.795,
+      "step": 6174
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3895153416197268,
+      "learning_rate": 1.3089859219195256e-09,
+      "loss": 0.8513,
+      "step": 6175
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3891368133246338,
+      "learning_rate": 1.2258990230995527e-09,
+      "loss": 0.8113,
+      "step": 6176
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.291173652580304,
+      "learning_rate": 1.1455360659617942e-09,
+      "loss": 0.8816,
+      "step": 6177
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4979692610587292,
+      "learning_rate": 1.0678970724009585e-09,
+      "loss": 0.7963,
+      "step": 6178
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3732226050397678,
+      "learning_rate": 9.92982063565684e-10,
+      "loss": 0.8813,
+      "step": 6179
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.674897591152258,
+      "learning_rate": 9.207910598674208e-10,
+      "loss": 0.8686,
+      "step": 6180
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.8966911187644984,
+      "learning_rate": 8.513240809715495e-10,
+      "loss": 0.3084,
+      "step": 6181
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4861555870307426,
+      "learning_rate": 7.845811458029317e-10,
+      "loss": 0.8592,
+      "step": 6182
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.852847697792143,
+      "learning_rate": 7.205622725448003e-10,
+      "loss": 0.3285,
+      "step": 6183
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3861945867513388,
+      "learning_rate": 6.592674786376485e-10,
+      "loss": 0.8363,
+      "step": 6184
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5794724339711894,
+      "learning_rate": 6.006967807781206e-10,
+      "loss": 0.8016,
+      "step": 6185
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5225244295364593,
+      "learning_rate": 5.448501949256724e-10,
+      "loss": 0.7794,
+      "step": 6186
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.2166138736397747,
+      "learning_rate": 4.917277362914696e-10,
+      "loss": 0.7963,
+      "step": 6187
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.2835831478607436,
+      "learning_rate": 4.413294193483797e-10,
+      "loss": 0.7081,
+      "step": 6188
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6787691466850847,
+      "learning_rate": 3.936552578276409e-10,
+      "loss": 0.8822,
+      "step": 6189
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.394811230156793,
+      "learning_rate": 3.48705264715532e-10,
+      "loss": 0.8101,
+      "step": 6190
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5505917310816706,
+      "learning_rate": 3.0647945225781294e-10,
+      "loss": 0.8211,
+      "step": 6191
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6743771874917042,
+      "learning_rate": 2.669778319586147e-10,
+      "loss": 0.7956,
+      "step": 6192
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4424851500042852,
+      "learning_rate": 2.3020041457821885e-10,
+      "loss": 0.8436,
+      "step": 6193
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7040445148789318,
+      "learning_rate": 1.9614721013749838e-10,
+      "loss": 0.825,
+      "step": 6194
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6273866673422035,
+      "learning_rate": 1.6481822791125646e-10,
+      "loss": 0.8281,
+      "step": 6195
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3892537432422891,
+      "learning_rate": 1.3621347643710814e-10,
+      "loss": 0.8789,
+      "step": 6196
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5215961491144232,
+      "learning_rate": 1.1033296350548839e-10,
+      "loss": 0.7913,
+      "step": 6197
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6205482070520882,
+      "learning_rate": 8.717669616742364e-11,
+      "loss": 0.8448,
+      "step": 6198
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3315640375465965,
+      "learning_rate": 6.674468073231133e-11,
+      "loss": 0.7669,
+      "step": 6199
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5160521220035885,
+      "learning_rate": 4.903692276569949e-11,
+      "loss": 0.842,
+      "step": 6200
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.376235957430811,
+      "learning_rate": 3.405342709150716e-11,
+      "loss": 0.8409,
+      "step": 6201
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4748114213775856,
+      "learning_rate": 2.1794197790914183e-11,
+      "loss": 0.8638,
+      "step": 6202
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.2362651379526777,
+      "learning_rate": 1.2259238205691858e-11,
+      "loss": 0.8505,
+      "step": 6203
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4117071250211006,
+      "learning_rate": 5.4485509326518415e-12,
+      "loss": 0.8824,
+      "step": 6204
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.2812310641186235,
+      "learning_rate": 1.362137825866583e-12,
+      "loss": 0.7853,
+      "step": 6205
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.8623132640479105,
+      "learning_rate": 0.0,
+      "loss": 0.4962,
+      "step": 6206
+    },
+    {
+      "epoch": 1.0,
+      "step": 6206,
+      "total_flos": 2.654469138789748e+19,
+      "train_loss": 0.8537314998750106,
+      "train_runtime": 50545.8721,
+      "train_samples_per_second": 15.717,
+      "train_steps_per_second": 0.123
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 6206,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 20000,
+  "total_flos": 2.654469138789748e+19,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}