diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,65144 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 9302,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.132667040745654,
+      "learning_rate": 3.571428571428572e-08,
+      "loss": 1.1077,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.5832425816734204,
+      "learning_rate": 7.142857142857144e-08,
+      "loss": 1.0251,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.081331918493752,
+      "learning_rate": 1.0714285714285716e-07,
+      "loss": 1.1024,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.8633328566479195,
+      "learning_rate": 1.4285714285714287e-07,
+      "loss": 1.1216,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.7936032871635522,
+      "learning_rate": 1.7857142857142858e-07,
+      "loss": 1.0761,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.083283544203863,
+      "learning_rate": 2.142857142857143e-07,
+      "loss": 1.1378,
+      "step": 6
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.990553192653829,
+      "learning_rate": 2.5000000000000004e-07,
+      "loss": 1.1512,
+      "step": 7
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.780291733936853,
+      "learning_rate": 2.8571428571428575e-07,
+      "loss": 1.1272,
+      "step": 8
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.798410582076193,
+      "learning_rate": 3.214285714285714e-07,
+      "loss": 1.1643,
+      "step": 9
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.618792809252698,
+      "learning_rate": 3.5714285714285716e-07,
+      "loss": 1.0503,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.489419747018667,
+      "learning_rate": 3.9285714285714286e-07,
+      "loss": 1.1526,
+      "step": 11
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.737918990138968,
+      "learning_rate": 4.285714285714286e-07,
+      "loss": 1.1737,
+      "step": 12
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.9032250615675395,
+      "learning_rate": 4.642857142857143e-07,
+      "loss": 1.0996,
+      "step": 13
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.7614359569379131,
+      "learning_rate": 5.000000000000001e-07,
+      "loss": 1.0415,
+      "step": 14
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.6846012859223602,
+      "learning_rate": 5.357142857142857e-07,
+      "loss": 1.1706,
+      "step": 15
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.363759732901119,
+      "learning_rate": 5.714285714285715e-07,
+      "loss": 1.1177,
+      "step": 16
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.965490335364641,
+      "learning_rate": 6.071428571428572e-07,
+      "loss": 1.1345,
+      "step": 17
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.217408774339712,
+      "learning_rate": 6.428571428571428e-07,
+      "loss": 1.1446,
+      "step": 18
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.288632652626488,
+      "learning_rate": 6.785714285714286e-07,
+      "loss": 1.1303,
+      "step": 19
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.036957671044967,
+      "learning_rate": 7.142857142857143e-07,
+      "loss": 1.1462,
+      "step": 20
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.503082109980841,
+      "learning_rate": 7.5e-07,
+      "loss": 1.1061,
+      "step": 21
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.075934264654508,
+      "learning_rate": 7.857142857142857e-07,
+      "loss": 1.1608,
+      "step": 22
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.8841677561501564,
+      "learning_rate": 8.214285714285715e-07,
+      "loss": 1.1,
+      "step": 23
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.7592592592734663,
+      "learning_rate": 8.571428571428572e-07,
+      "loss": 1.0412,
+      "step": 24
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.2771766359738415,
+      "learning_rate": 8.928571428571429e-07,
+      "loss": 1.1736,
+      "step": 25
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.9519357053838924,
+      "learning_rate": 9.285714285714287e-07,
+      "loss": 1.1072,
+      "step": 26
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.6181542903595982,
+      "learning_rate": 9.642857142857145e-07,
+      "loss": 1.1146,
+      "step": 27
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.1157188676659278,
+      "learning_rate": 1.0000000000000002e-06,
+      "loss": 1.1193,
+      "step": 28
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.85238806546678,
+      "learning_rate": 1.0357142857142859e-06,
+      "loss": 1.1558,
+      "step": 29
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.7945059133569435,
+      "learning_rate": 1.0714285714285714e-06,
+      "loss": 1.0713,
+      "step": 30
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.182934667495162,
+      "learning_rate": 1.1071428571428573e-06,
+      "loss": 1.1067,
+      "step": 31
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.60556636198004,
+      "learning_rate": 1.142857142857143e-06,
+      "loss": 1.0752,
+      "step": 32
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.286711288434339,
+      "learning_rate": 1.1785714285714287e-06,
+      "loss": 1.0178,
+      "step": 33
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.6556224312393986,
+      "learning_rate": 1.2142857142857144e-06,
+      "loss": 1.0089,
+      "step": 34
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.799681523216164,
+      "learning_rate": 1.25e-06,
+      "loss": 1.1192,
+      "step": 35
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.448133385030778,
+      "learning_rate": 1.2857142857142856e-06,
+      "loss": 1.0778,
+      "step": 36
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.1832669597916006,
+      "learning_rate": 1.3214285714285715e-06,
+      "loss": 1.0193,
+      "step": 37
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.3479972378836447,
+      "learning_rate": 1.3571428571428572e-06,
+      "loss": 1.0401,
+      "step": 38
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.1381509060973554,
+      "learning_rate": 1.392857142857143e-06,
+      "loss": 1.0512,
+      "step": 39
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.8779552164019765,
+      "learning_rate": 1.4285714285714286e-06,
+      "loss": 1.063,
+      "step": 40
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.428905843479653,
+      "learning_rate": 1.4642857142857145e-06,
+      "loss": 1.0743,
+      "step": 41
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.124776423877994,
+      "learning_rate": 1.5e-06,
+      "loss": 1.0816,
+      "step": 42
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.0109705235401445,
+      "learning_rate": 1.535714285714286e-06,
+      "loss": 0.9992,
+      "step": 43
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.1162201902784097,
+      "learning_rate": 1.5714285714285714e-06,
+      "loss": 1.0692,
+      "step": 44
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.142106122929943,
+      "learning_rate": 1.6071428571428574e-06,
+      "loss": 1.135,
+      "step": 45
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.065089444816147,
+      "learning_rate": 1.642857142857143e-06,
+      "loss": 1.0484,
+      "step": 46
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8847204976170282,
+      "learning_rate": 1.6785714285714286e-06,
+      "loss": 1.1077,
+      "step": 47
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0327528518531324,
+      "learning_rate": 1.7142857142857145e-06,
+      "loss": 1.1642,
+      "step": 48
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0341463283734225,
+      "learning_rate": 1.75e-06,
+      "loss": 1.1166,
+      "step": 49
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8338487229571316,
+      "learning_rate": 1.7857142857142859e-06,
+      "loss": 0.9445,
+      "step": 50
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.967377778585595,
+      "learning_rate": 1.8214285714285716e-06,
+      "loss": 1.1091,
+      "step": 51
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1284920128828375,
+      "learning_rate": 1.8571428571428573e-06,
+      "loss": 1.1126,
+      "step": 52
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0672831930590867,
+      "learning_rate": 1.892857142857143e-06,
+      "loss": 1.0507,
+      "step": 53
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0333319879417613,
+      "learning_rate": 1.928571428571429e-06,
+      "loss": 1.0415,
+      "step": 54
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8786659219203434,
+      "learning_rate": 1.9642857142857144e-06,
+      "loss": 1.0165,
+      "step": 55
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.918407820837012,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 1.0638,
+      "step": 56
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8516765852067223,
+      "learning_rate": 2.035714285714286e-06,
+      "loss": 1.0447,
+      "step": 57
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.4320660831231042,
+      "learning_rate": 2.0714285714285717e-06,
+      "loss": 1.0313,
+      "step": 58
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8872259842859693,
+      "learning_rate": 2.1071428571428572e-06,
+      "loss": 1.0493,
+      "step": 59
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2014917685697886,
+      "learning_rate": 2.1428571428571427e-06,
+      "loss": 1.0829,
+      "step": 60
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7494056578946249,
+      "learning_rate": 2.1785714285714286e-06,
+      "loss": 1.0281,
+      "step": 61
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8926843897702263,
+      "learning_rate": 2.2142857142857146e-06,
+      "loss": 1.0901,
+      "step": 62
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9982540726924571,
+      "learning_rate": 2.25e-06,
+      "loss": 1.0307,
+      "step": 63
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9553156105718255,
+      "learning_rate": 2.285714285714286e-06,
+      "loss": 1.0379,
+      "step": 64
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7508115743290498,
+      "learning_rate": 2.321428571428572e-06,
+      "loss": 1.0175,
+      "step": 65
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.76567899787695,
+      "learning_rate": 2.3571428571428574e-06,
+      "loss": 0.9992,
+      "step": 66
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9818949212936914,
+      "learning_rate": 2.3928571428571433e-06,
+      "loss": 1.0146,
+      "step": 67
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9235911390313425,
+      "learning_rate": 2.428571428571429e-06,
+      "loss": 1.0556,
+      "step": 68
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9334278051472502,
+      "learning_rate": 2.4642857142857147e-06,
+      "loss": 0.9865,
+      "step": 69
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9175828101791368,
+      "learning_rate": 2.5e-06,
+      "loss": 0.9776,
+      "step": 70
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9516094394347074,
+      "learning_rate": 2.5357142857142857e-06,
+      "loss": 0.933,
+      "step": 71
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.802534782906283,
+      "learning_rate": 2.571428571428571e-06,
+      "loss": 1.016,
+      "step": 72
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9136428941209254,
+      "learning_rate": 2.6071428571428575e-06,
+      "loss": 0.9673,
+      "step": 73
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.94501972139913,
+      "learning_rate": 2.642857142857143e-06,
+      "loss": 1.0887,
+      "step": 74
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.913293449765966,
+      "learning_rate": 2.6785714285714285e-06,
+      "loss": 1.0172,
+      "step": 75
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.817372303588891,
+      "learning_rate": 2.7142857142857144e-06,
+      "loss": 1.1163,
+      "step": 76
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9481917224020602,
+      "learning_rate": 2.7500000000000004e-06,
+      "loss": 1.011,
+      "step": 77
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.788267732281819,
+      "learning_rate": 2.785714285714286e-06,
+      "loss": 0.997,
+      "step": 78
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8784637878723836,
+      "learning_rate": 2.8214285714285718e-06,
+      "loss": 0.9679,
+      "step": 79
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8537341419288107,
+      "learning_rate": 2.8571428571428573e-06,
+      "loss": 1.0761,
+      "step": 80
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8078542130867514,
+      "learning_rate": 2.892857142857143e-06,
+      "loss": 1.082,
+      "step": 81
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0120693392018367,
+      "learning_rate": 2.928571428571429e-06,
+      "loss": 0.9997,
+      "step": 82
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.858289061402852,
+      "learning_rate": 2.9642857142857146e-06,
+      "loss": 0.9857,
+      "step": 83
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.443760850229742,
+      "learning_rate": 3e-06,
+      "loss": 1.0857,
+      "step": 84
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8929874508264222,
+      "learning_rate": 3.0357142857142856e-06,
+      "loss": 1.0229,
+      "step": 85
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6712403104381248,
+      "learning_rate": 3.071428571428572e-06,
+      "loss": 0.9918,
+      "step": 86
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9400055474881428,
+      "learning_rate": 3.1071428571428574e-06,
+      "loss": 0.9699,
+      "step": 87
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8587272521034581,
+      "learning_rate": 3.142857142857143e-06,
+      "loss": 1.0221,
+      "step": 88
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7873447084415033,
+      "learning_rate": 3.178571428571429e-06,
+      "loss": 0.9635,
+      "step": 89
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.840087749643695,
+      "learning_rate": 3.2142857142857147e-06,
+      "loss": 1.0235,
+      "step": 90
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.727243481814505,
+      "learning_rate": 3.2500000000000002e-06,
+      "loss": 0.9754,
+      "step": 91
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8327650161161178,
+      "learning_rate": 3.285714285714286e-06,
+      "loss": 0.9965,
+      "step": 92
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9350250589762261,
+      "learning_rate": 3.3214285714285716e-06,
+      "loss": 1.0383,
+      "step": 93
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8991733726284912,
+      "learning_rate": 3.357142857142857e-06,
+      "loss": 1.0678,
+      "step": 94
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7973465752671096,
+      "learning_rate": 3.3928571428571435e-06,
+      "loss": 1.0978,
+      "step": 95
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9533674203712352,
+      "learning_rate": 3.428571428571429e-06,
+      "loss": 0.9195,
+      "step": 96
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9088003679630476,
+      "learning_rate": 3.4642857142857145e-06,
+      "loss": 0.9903,
+      "step": 97
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8372828926002955,
+      "learning_rate": 3.5e-06,
+      "loss": 1.0251,
+      "step": 98
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9696122216873582,
+      "learning_rate": 3.5357142857142863e-06,
+      "loss": 1.0034,
+      "step": 99
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8380300025142526,
+      "learning_rate": 3.5714285714285718e-06,
+      "loss": 0.9751,
+      "step": 100
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.847954449033685,
+      "learning_rate": 3.6071428571428573e-06,
+      "loss": 0.9459,
+      "step": 101
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8832981594056148,
+      "learning_rate": 3.642857142857143e-06,
+      "loss": 1.0059,
+      "step": 102
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8547492680917304,
+      "learning_rate": 3.678571428571429e-06,
+      "loss": 0.9932,
+      "step": 103
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8657425751623185,
+      "learning_rate": 3.7142857142857146e-06,
+      "loss": 1.002,
+      "step": 104
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6665780702081876,
+      "learning_rate": 3.7500000000000005e-06,
+      "loss": 1.0593,
+      "step": 105
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7964712148810413,
+      "learning_rate": 3.785714285714286e-06,
+      "loss": 0.9619,
+      "step": 106
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7553910708732687,
+      "learning_rate": 3.8214285714285715e-06,
+      "loss": 1.01,
+      "step": 107
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7747469002432539,
+      "learning_rate": 3.857142857142858e-06,
+      "loss": 1.0268,
+      "step": 108
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9173742281169184,
+      "learning_rate": 3.892857142857143e-06,
+      "loss": 1.0891,
+      "step": 109
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7949439015706234,
+      "learning_rate": 3.928571428571429e-06,
+      "loss": 0.983,
+      "step": 110
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.867520643718796,
+      "learning_rate": 3.964285714285714e-06,
+      "loss": 0.9982,
+      "step": 111
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8034726808554176,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 1.0578,
+      "step": 112
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8141924506809295,
+      "learning_rate": 4.035714285714286e-06,
+      "loss": 1.001,
+      "step": 113
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7976283640270174,
+      "learning_rate": 4.071428571428572e-06,
+      "loss": 0.9854,
+      "step": 114
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.797596603852026,
+      "learning_rate": 4.107142857142857e-06,
+      "loss": 0.9949,
+      "step": 115
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6659524586669274,
+      "learning_rate": 4.1428571428571435e-06,
+      "loss": 1.0108,
+      "step": 116
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8330019349258238,
+      "learning_rate": 4.178571428571429e-06,
+      "loss": 1.044,
+      "step": 117
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7917795096457647,
+      "learning_rate": 4.2142857142857145e-06,
+      "loss": 0.9551,
+      "step": 118
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8926908130434266,
+      "learning_rate": 4.25e-06,
+      "loss": 1.0879,
+      "step": 119
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8494192643378685,
+      "learning_rate": 4.2857142857142855e-06,
+      "loss": 1.0516,
+      "step": 120
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7928814318941266,
+      "learning_rate": 4.321428571428572e-06,
+      "loss": 0.9981,
+      "step": 121
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7244277737182516,
+      "learning_rate": 4.357142857142857e-06,
+      "loss": 1.0243,
+      "step": 122
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8920283111381664,
+      "learning_rate": 4.392857142857143e-06,
+      "loss": 1.0406,
+      "step": 123
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8754037128801189,
+      "learning_rate": 4.428571428571429e-06,
+      "loss": 0.9638,
+      "step": 124
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.911665558253273,
+      "learning_rate": 4.464285714285715e-06,
+      "loss": 1.0164,
+      "step": 125
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.772633550590763,
+      "learning_rate": 4.5e-06,
+      "loss": 0.9896,
+      "step": 126
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.669546660563049,
+      "learning_rate": 4.5357142857142865e-06,
+      "loss": 0.9915,
+      "step": 127
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9082535381339016,
+      "learning_rate": 4.571428571428572e-06,
+      "loss": 0.998,
+      "step": 128
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8817733444830658,
+      "learning_rate": 4.6071428571428574e-06,
+      "loss": 0.9828,
+      "step": 129
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9198395496696505,
+      "learning_rate": 4.642857142857144e-06,
+      "loss": 1.0489,
+      "step": 130
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8826157598716975,
+      "learning_rate": 4.678571428571429e-06,
+      "loss": 0.9634,
+      "step": 131
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.735718457310413,
+      "learning_rate": 4.714285714285715e-06,
+      "loss": 0.9558,
+      "step": 132
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8876893880232615,
+      "learning_rate": 4.75e-06,
+      "loss": 0.9628,
+      "step": 133
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8906222378020202,
+      "learning_rate": 4.785714285714287e-06,
+      "loss": 1.0632,
+      "step": 134
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7780584176322876,
+      "learning_rate": 4.821428571428572e-06,
+      "loss": 0.8832,
+      "step": 135
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.949660343989006,
+      "learning_rate": 4.857142857142858e-06,
+      "loss": 0.9787,
+      "step": 136
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8651817922616587,
+      "learning_rate": 4.892857142857143e-06,
+      "loss": 1.0215,
+      "step": 137
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8595636989861604,
+      "learning_rate": 4.928571428571429e-06,
+      "loss": 1.0211,
+      "step": 138
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9155569626611333,
+      "learning_rate": 4.964285714285715e-06,
+      "loss": 1.0025,
+      "step": 139
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.854851167247576,
+      "learning_rate": 5e-06,
+      "loss": 1.0933,
+      "step": 140
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1191389008275445,
+      "learning_rate": 5.035714285714286e-06,
+      "loss": 0.9626,
+      "step": 141
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8318167230835649,
+      "learning_rate": 5.071428571428571e-06,
+      "loss": 1.0578,
+      "step": 142
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9756106842049228,
+      "learning_rate": 5.107142857142857e-06,
+      "loss": 1.0221,
+      "step": 143
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8017593586744073,
+      "learning_rate": 5.142857142857142e-06,
+      "loss": 1.0654,
+      "step": 144
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8686875983722218,
+      "learning_rate": 5.1785714285714296e-06,
+      "loss": 1.0484,
+      "step": 145
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.142993315548542,
+      "learning_rate": 5.214285714285715e-06,
+      "loss": 1.0742,
+      "step": 146
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.758586210268953,
+      "learning_rate": 5.2500000000000006e-06,
+      "loss": 0.9645,
+      "step": 147
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8497120419127009,
+      "learning_rate": 5.285714285714286e-06,
+      "loss": 0.966,
+      "step": 148
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.19940116235944,
+      "learning_rate": 5.3214285714285715e-06,
+      "loss": 1.0068,
+      "step": 149
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9107416277972262,
+      "learning_rate": 5.357142857142857e-06,
+      "loss": 0.9777,
+      "step": 150
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7854396855590269,
+      "learning_rate": 5.392857142857143e-06,
+      "loss": 1.0025,
+      "step": 151
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.849147562026073,
+      "learning_rate": 5.428571428571429e-06,
+      "loss": 1.0689,
+      "step": 152
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9479106386506027,
+      "learning_rate": 5.464285714285714e-06,
+      "loss": 1.0963,
+      "step": 153
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8601881293482652,
+      "learning_rate": 5.500000000000001e-06,
+      "loss": 0.9894,
+      "step": 154
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8068654290580078,
+      "learning_rate": 5.535714285714286e-06,
+      "loss": 1.0484,
+      "step": 155
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9335778532937649,
+      "learning_rate": 5.571428571428572e-06,
+      "loss": 0.9808,
+      "step": 156
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.671782156709297,
+      "learning_rate": 5.607142857142858e-06,
+      "loss": 0.9491,
+      "step": 157
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.865357045098957,
+      "learning_rate": 5.6428571428571435e-06,
+      "loss": 1.0005,
+      "step": 158
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8325287278744955,
+      "learning_rate": 5.678571428571429e-06,
+      "loss": 1.0134,
+      "step": 159
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8969848642682183,
+      "learning_rate": 5.7142857142857145e-06,
+      "loss": 0.9595,
+      "step": 160
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0133514679110736,
+      "learning_rate": 5.75e-06,
+      "loss": 1.0015,
+      "step": 161
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.913586407953343,
+      "learning_rate": 5.785714285714286e-06,
+      "loss": 0.9709,
+      "step": 162
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7144487151715317,
+      "learning_rate": 5.821428571428573e-06,
+      "loss": 0.8927,
+      "step": 163
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.851837740629236,
+      "learning_rate": 5.857142857142858e-06,
+      "loss": 0.9865,
+      "step": 164
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8280151614241462,
+      "learning_rate": 5.892857142857144e-06,
+      "loss": 0.9441,
+      "step": 165
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7242042075904989,
+      "learning_rate": 5.928571428571429e-06,
+      "loss": 0.984,
+      "step": 166
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9161404562957403,
+      "learning_rate": 5.964285714285715e-06,
+      "loss": 1.0448,
+      "step": 167
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7210219289051765,
+      "learning_rate": 6e-06,
+      "loss": 1.0076,
+      "step": 168
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9161873976847124,
+      "learning_rate": 6.035714285714286e-06,
+      "loss": 1.0101,
+      "step": 169
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.994381188481452,
+      "learning_rate": 6.071428571428571e-06,
+      "loss": 1.0199,
+      "step": 170
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8155916608768936,
+      "learning_rate": 6.107142857142858e-06,
+      "loss": 1.0366,
+      "step": 171
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8529987765332334,
+      "learning_rate": 6.142857142857144e-06,
+      "loss": 1.0381,
+      "step": 172
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7646415075275224,
+      "learning_rate": 6.178571428571429e-06,
+      "loss": 0.9645,
+      "step": 173
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9556687103694002,
+      "learning_rate": 6.214285714285715e-06,
+      "loss": 1.0198,
+      "step": 174
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8690329443893716,
+      "learning_rate": 6.25e-06,
+      "loss": 0.9835,
+      "step": 175
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9878692524987729,
+      "learning_rate": 6.285714285714286e-06,
+      "loss": 1.0341,
+      "step": 176
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7552142151229924,
+      "learning_rate": 6.321428571428571e-06,
+      "loss": 1.0388,
+      "step": 177
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9164847564790324,
+      "learning_rate": 6.357142857142858e-06,
+      "loss": 0.9507,
+      "step": 178
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8482424875142105,
+      "learning_rate": 6.392857142857143e-06,
+      "loss": 0.9355,
+      "step": 179
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8769436691779848,
+      "learning_rate": 6.4285714285714295e-06,
+      "loss": 1.1162,
+      "step": 180
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7499514391959903,
+      "learning_rate": 6.464285714285715e-06,
+      "loss": 1.0159,
+      "step": 181
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7872268649535845,
+      "learning_rate": 6.5000000000000004e-06,
+      "loss": 0.9953,
+      "step": 182
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8434107714580807,
+      "learning_rate": 6.535714285714286e-06,
+      "loss": 0.9673,
+      "step": 183
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.932556414731293,
+      "learning_rate": 6.571428571428572e-06,
+      "loss": 1.0159,
+      "step": 184
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7812262305105497,
+      "learning_rate": 6.607142857142858e-06,
+      "loss": 1.0057,
+      "step": 185
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9146929237583254,
+      "learning_rate": 6.642857142857143e-06,
+      "loss": 0.952,
+      "step": 186
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.050237576845038,
+      "learning_rate": 6.678571428571429e-06,
+      "loss": 1.0187,
+      "step": 187
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8055046835895017,
+      "learning_rate": 6.714285714285714e-06,
+      "loss": 0.9374,
+      "step": 188
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7428815471800148,
+      "learning_rate": 6.750000000000001e-06,
+      "loss": 0.9654,
+      "step": 189
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7290508613297726,
+      "learning_rate": 6.785714285714287e-06,
+      "loss": 0.979,
+      "step": 190
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8709292729971096,
+      "learning_rate": 6.8214285714285724e-06,
+      "loss": 1.032,
+      "step": 191
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8539278040187954,
+      "learning_rate": 6.857142857142858e-06,
+      "loss": 0.9922,
+      "step": 192
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8943122990472057,
+      "learning_rate": 6.892857142857143e-06,
+      "loss": 0.9647,
+      "step": 193
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9487611423888436,
+      "learning_rate": 6.928571428571429e-06,
+      "loss": 0.958,
+      "step": 194
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9710707176492916,
+      "learning_rate": 6.964285714285714e-06,
+      "loss": 0.9513,
+      "step": 195
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8810856608114985,
+      "learning_rate": 7e-06,
+      "loss": 1.0234,
+      "step": 196
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9246220265109757,
+      "learning_rate": 7.035714285714287e-06,
+      "loss": 0.9576,
+      "step": 197
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9171919203292742,
+      "learning_rate": 7.0714285714285726e-06,
+      "loss": 1.0097,
+      "step": 198
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8111290026886595,
+      "learning_rate": 7.107142857142858e-06,
+      "loss": 1.0222,
+      "step": 199
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8889835487175084,
+      "learning_rate": 7.1428571428571436e-06,
+      "loss": 0.9164,
+      "step": 200
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9183040195395535,
+      "learning_rate": 7.178571428571429e-06,
+      "loss": 1.0267,
+      "step": 201
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.796736996058039,
+      "learning_rate": 7.2142857142857145e-06,
+      "loss": 0.9211,
+      "step": 202
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.725403341102231,
+      "learning_rate": 7.25e-06,
+      "loss": 0.9405,
+      "step": 203
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.803756098788917,
+      "learning_rate": 7.285714285714286e-06,
+      "loss": 0.9899,
+      "step": 204
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8910097314906544,
+      "learning_rate": 7.321428571428572e-06,
+      "loss": 0.9688,
+      "step": 205
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7738192196269755,
+      "learning_rate": 7.357142857142858e-06,
+      "loss": 1.0411,
+      "step": 206
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8355213449343943,
+      "learning_rate": 7.392857142857144e-06,
+      "loss": 0.9451,
+      "step": 207
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9872759071908677,
+      "learning_rate": 7.428571428571429e-06,
+      "loss": 1.0579,
+      "step": 208
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.036220596836638,
+      "learning_rate": 7.464285714285715e-06,
+      "loss": 1.0807,
+      "step": 209
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.046736507093887,
+      "learning_rate": 7.500000000000001e-06,
+      "loss": 0.9566,
+      "step": 210
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.834145401119039,
+      "learning_rate": 7.5357142857142865e-06,
+      "loss": 0.9485,
+      "step": 211
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8167955776419433,
+      "learning_rate": 7.571428571428572e-06,
+      "loss": 1.0094,
+      "step": 212
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7953286542018367,
+      "learning_rate": 7.6071428571428575e-06,
+      "loss": 0.9822,
+      "step": 213
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.934893444982141,
+      "learning_rate": 7.642857142857143e-06,
+      "loss": 0.9906,
+      "step": 214
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8117607424872288,
+      "learning_rate": 7.67857142857143e-06,
+      "loss": 0.9645,
+      "step": 215
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8050905557334838,
+      "learning_rate": 7.714285714285716e-06,
+      "loss": 0.9649,
+      "step": 216
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0057293784376,
+      "learning_rate": 7.75e-06,
+      "loss": 1.005,
+      "step": 217
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7653909964252372,
+      "learning_rate": 7.785714285714287e-06,
+      "loss": 0.9747,
+      "step": 218
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8667043820377498,
+      "learning_rate": 7.821428571428571e-06,
+      "loss": 1.0398,
+      "step": 219
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9457652331065496,
+      "learning_rate": 7.857142857142858e-06,
+      "loss": 0.9192,
+      "step": 220
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8650452583254298,
+      "learning_rate": 7.892857142857144e-06,
+      "loss": 0.9619,
+      "step": 221
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8425348807571091,
+      "learning_rate": 7.928571428571429e-06,
+      "loss": 0.9311,
+      "step": 222
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.095624587213627,
+      "learning_rate": 7.964285714285715e-06,
+      "loss": 1.0646,
+      "step": 223
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0660069130445873,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 1.0495,
+      "step": 224
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8776626176180948,
+      "learning_rate": 8.035714285714286e-06,
+      "loss": 0.8902,
+      "step": 225
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8569786061484821,
+      "learning_rate": 8.071428571428572e-06,
+      "loss": 1.0566,
+      "step": 226
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.944510348870039,
+      "learning_rate": 8.107142857142859e-06,
+      "loss": 0.9745,
+      "step": 227
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7657772602076074,
+      "learning_rate": 8.142857142857143e-06,
+      "loss": 1.0124,
+      "step": 228
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8081497041727166,
+      "learning_rate": 8.17857142857143e-06,
+      "loss": 0.9777,
+      "step": 229
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7912166529817861,
+      "learning_rate": 8.214285714285714e-06,
+      "loss": 1.0445,
+      "step": 230
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8123940070158067,
+      "learning_rate": 8.25e-06,
+      "loss": 1.0406,
+      "step": 231
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8576805497501068,
+      "learning_rate": 8.285714285714287e-06,
+      "loss": 0.9653,
+      "step": 232
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9225745882521939,
+      "learning_rate": 8.321428571428573e-06,
+      "loss": 1.0605,
+      "step": 233
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7395879984469136,
+      "learning_rate": 8.357142857142858e-06,
+      "loss": 0.9636,
+      "step": 234
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9006397505812607,
+      "learning_rate": 8.392857142857144e-06,
+      "loss": 0.9565,
+      "step": 235
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.291257477520151,
+      "learning_rate": 8.428571428571429e-06,
+      "loss": 1.043,
+      "step": 236
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8203677694879266,
+      "learning_rate": 8.464285714285715e-06,
+      "loss": 0.9956,
+      "step": 237
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7966100172996862,
+      "learning_rate": 8.5e-06,
+      "loss": 1.0042,
+      "step": 238
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7573668767925414,
+      "learning_rate": 8.535714285714286e-06,
+      "loss": 0.9407,
+      "step": 239
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8643526191975748,
+      "learning_rate": 8.571428571428571e-06,
+      "loss": 1.0195,
+      "step": 240
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8545606002743658,
+      "learning_rate": 8.607142857142859e-06,
+      "loss": 1.012,
+      "step": 241
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8015753058981967,
+      "learning_rate": 8.642857142857144e-06,
+      "loss": 0.9649,
+      "step": 242
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1035101534370813,
+      "learning_rate": 8.67857142857143e-06,
+      "loss": 0.9757,
+      "step": 243
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7262336482910128,
+      "learning_rate": 8.714285714285715e-06,
+      "loss": 1.0357,
+      "step": 244
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.781661870931318,
+      "learning_rate": 8.750000000000001e-06,
+      "loss": 0.9269,
+      "step": 245
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7344842248404833,
+      "learning_rate": 8.785714285714286e-06,
+      "loss": 0.978,
+      "step": 246
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2122058143163352,
+      "learning_rate": 8.821428571428572e-06,
+      "loss": 0.9117,
+      "step": 247
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7336709645662929,
+      "learning_rate": 8.857142857142858e-06,
+      "loss": 0.9877,
+      "step": 248
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7301957684985823,
+      "learning_rate": 8.892857142857143e-06,
+      "loss": 0.8862,
+      "step": 249
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9050622733091505,
+      "learning_rate": 8.92857142857143e-06,
+      "loss": 0.912,
+      "step": 250
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8358700381944053,
+      "learning_rate": 8.964285714285716e-06,
+      "loss": 0.9989,
+      "step": 251
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.735802137705787,
+      "learning_rate": 9e-06,
+      "loss": 0.9906,
+      "step": 252
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7854181328182608,
+      "learning_rate": 9.035714285714287e-06,
+      "loss": 1.0675,
+      "step": 253
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9189468675670451,
+      "learning_rate": 9.071428571428573e-06,
+      "loss": 0.9858,
+      "step": 254
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.032427778485921,
+      "learning_rate": 9.107142857142858e-06,
+      "loss": 1.0544,
+      "step": 255
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8066424778786374,
+      "learning_rate": 9.142857142857144e-06,
+      "loss": 1.0026,
+      "step": 256
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9110162535480548,
+      "learning_rate": 9.178571428571429e-06,
+      "loss": 0.9976,
+      "step": 257
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7861930824601886,
+      "learning_rate": 9.214285714285715e-06,
+      "loss": 0.9512,
+      "step": 258
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.770632806920664,
+      "learning_rate": 9.250000000000001e-06,
+      "loss": 0.9302,
+      "step": 259
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9557164474413924,
+      "learning_rate": 9.285714285714288e-06,
+      "loss": 1.0303,
+      "step": 260
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.81306497888572,
+      "learning_rate": 9.321428571428572e-06,
+      "loss": 0.961,
+      "step": 261
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1778120221806583,
+      "learning_rate": 9.357142857142859e-06,
+      "loss": 1.0658,
+      "step": 262
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9161924776149604,
+      "learning_rate": 9.392857142857143e-06,
+      "loss": 0.9645,
+      "step": 263
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8556829068296476,
+      "learning_rate": 9.42857142857143e-06,
+      "loss": 0.8909,
+      "step": 264
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8808322974032767,
+      "learning_rate": 9.464285714285714e-06,
+      "loss": 1.0483,
+      "step": 265
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0288988239443093,
+      "learning_rate": 9.5e-06,
+      "loss": 1.0503,
+      "step": 266
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9520986111176257,
+      "learning_rate": 9.535714285714287e-06,
+      "loss": 1.0207,
+      "step": 267
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.807397773832005,
+      "learning_rate": 9.571428571428573e-06,
+      "loss": 0.9126,
+      "step": 268
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7665364979436666,
+      "learning_rate": 9.607142857142858e-06,
+      "loss": 0.9769,
+      "step": 269
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0276055140102667,
+      "learning_rate": 9.642857142857144e-06,
+      "loss": 0.9747,
+      "step": 270
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9350435118211642,
+      "learning_rate": 9.678571428571429e-06,
+      "loss": 0.9097,
+      "step": 271
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8864951736289273,
+      "learning_rate": 9.714285714285715e-06,
+      "loss": 1.0829,
+      "step": 272
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.714042820673516,
+      "learning_rate": 9.75e-06,
+      "loss": 0.9118,
+      "step": 273
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7691063494705084,
+      "learning_rate": 9.785714285714286e-06,
+      "loss": 0.9189,
+      "step": 274
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9451414181368836,
+      "learning_rate": 9.821428571428573e-06,
+      "loss": 0.9581,
+      "step": 275
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7953846137041707,
+      "learning_rate": 9.857142857142859e-06,
+      "loss": 0.9632,
+      "step": 276
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8061258794044173,
+      "learning_rate": 9.892857142857143e-06,
+      "loss": 0.9539,
+      "step": 277
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.007758057953556,
+      "learning_rate": 9.92857142857143e-06,
+      "loss": 1.1074,
+      "step": 278
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9535907710363158,
+      "learning_rate": 9.964285714285714e-06,
+      "loss": 1.0491,
+      "step": 279
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.838674484696826,
+      "learning_rate": 1e-05,
+      "loss": 1.0559,
+      "step": 280
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9413909246598793,
+      "learning_rate": 9.999999696866382e-06,
+      "loss": 0.9741,
+      "step": 281
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9014065870768257,
+      "learning_rate": 9.999998787465562e-06,
+      "loss": 1.0312,
+      "step": 282
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9284806265774397,
+      "learning_rate": 9.999997271797653e-06,
+      "loss": 1.0811,
+      "step": 283
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7991993311833812,
+      "learning_rate": 9.999995149862836e-06,
+      "loss": 0.938,
+      "step": 284
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.826482791858458,
+      "learning_rate": 9.999992421661369e-06,
+      "loss": 1.0089,
+      "step": 285
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7407124628486075,
+      "learning_rate": 9.999989087193582e-06,
+      "loss": 1.0458,
+      "step": 286
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8169931241354556,
+      "learning_rate": 9.999985146459882e-06,
+      "loss": 0.9556,
+      "step": 287
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.852519306063477,
+      "learning_rate": 9.999980599460747e-06,
+      "loss": 0.9774,
+      "step": 288
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.79173319189818,
+      "learning_rate": 9.999975446196725e-06,
+      "loss": 0.9972,
+      "step": 289
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.981412318188983,
+      "learning_rate": 9.999969686668442e-06,
+      "loss": 0.9291,
+      "step": 290
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8168744903192728,
+      "learning_rate": 9.9999633208766e-06,
+      "loss": 0.9561,
+      "step": 291
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6672658485947125,
+      "learning_rate": 9.999956348821965e-06,
+      "loss": 0.9927,
+      "step": 292
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7859722035567993,
+      "learning_rate": 9.999948770505386e-06,
+      "loss": 0.9362,
+      "step": 293
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7537741457767284,
+      "learning_rate": 9.99994058592778e-06,
+      "loss": 1.0143,
+      "step": 294
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8105585270396467,
+      "learning_rate": 9.999931795090143e-06,
+      "loss": 0.9353,
+      "step": 295
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7608932258417644,
+      "learning_rate": 9.999922397993537e-06,
+      "loss": 0.9088,
+      "step": 296
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0198290778560577,
+      "learning_rate": 9.999912394639105e-06,
+      "loss": 0.8845,
+      "step": 297
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7707371639021934,
+      "learning_rate": 9.999901785028055e-06,
+      "loss": 0.9302,
+      "step": 298
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8824966806420331,
+      "learning_rate": 9.999890569161676e-06,
+      "loss": 0.9837,
+      "step": 299
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8653828960816923,
+      "learning_rate": 9.99987874704133e-06,
+      "loss": 1.0813,
+      "step": 300
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8399969651795043,
+      "learning_rate": 9.99986631866845e-06,
+      "loss": 1.0106,
+      "step": 301
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8724158233403128,
+      "learning_rate": 9.999853284044539e-06,
+      "loss": 0.9073,
+      "step": 302
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8148194528040043,
+      "learning_rate": 9.99983964317118e-06,
+      "loss": 0.9504,
+      "step": 303
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8698145593419844,
+      "learning_rate": 9.99982539605003e-06,
+      "loss": 1.0199,
+      "step": 304
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.948495131611859,
+      "learning_rate": 9.999810542682811e-06,
+      "loss": 1.0659,
+      "step": 305
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0240633345617711,
+      "learning_rate": 9.999795083071328e-06,
+      "loss": 1.0572,
+      "step": 306
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9271320490521096,
+      "learning_rate": 9.999779017217454e-06,
+      "loss": 0.9408,
+      "step": 307
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.799310225281412,
+      "learning_rate": 9.999762345123137e-06,
+      "loss": 1.0931,
+      "step": 308
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8771991773467342,
+      "learning_rate": 9.999745066790398e-06,
+      "loss": 1.0826,
+      "step": 309
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.911607519215812,
+      "learning_rate": 9.999727182221335e-06,
+      "loss": 1.0542,
+      "step": 310
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8276600341947764,
+      "learning_rate": 9.999708691418113e-06,
+      "loss": 0.963,
+      "step": 311
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9260927865590318,
+      "learning_rate": 9.999689594382977e-06,
+      "loss": 0.9583,
+      "step": 312
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.769568358359543,
+      "learning_rate": 9.99966989111824e-06,
+      "loss": 0.9126,
+      "step": 313
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9279884601418389,
+      "learning_rate": 9.999649581626292e-06,
+      "loss": 0.957,
+      "step": 314
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0169806444231524,
+      "learning_rate": 9.999628665909597e-06,
+      "loss": 1.0059,
+      "step": 315
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.769920040310968,
+      "learning_rate": 9.999607143970689e-06,
+      "loss": 0.9079,
+      "step": 316
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3205038104644193,
+      "learning_rate": 9.99958501581218e-06,
+      "loss": 1.0608,
+      "step": 317
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9634857245188009,
+      "learning_rate": 9.999562281436749e-06,
+      "loss": 0.8918,
+      "step": 318
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7511875874435554,
+      "learning_rate": 9.999538940847156e-06,
+      "loss": 0.9796,
+      "step": 319
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7254704409678374,
+      "learning_rate": 9.999514994046232e-06,
+      "loss": 1.0014,
+      "step": 320
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7490919194395176,
+      "learning_rate": 9.999490441036877e-06,
+      "loss": 1.0146,
+      "step": 321
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9439622195078756,
+      "learning_rate": 9.999465281822072e-06,
+      "loss": 0.9571,
+      "step": 322
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.782253938648281,
+      "learning_rate": 9.999439516404864e-06,
+      "loss": 1.0272,
+      "step": 323
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9432956938428907,
+      "learning_rate": 9.99941314478838e-06,
+      "loss": 0.9866,
+      "step": 324
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8152523837360088,
+      "learning_rate": 9.999386166975818e-06,
+      "loss": 0.9852,
+      "step": 325
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7937956475411325,
+      "learning_rate": 9.999358582970444e-06,
+      "loss": 1.0544,
+      "step": 326
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.795258404936494,
+      "learning_rate": 9.99933039277561e-06,
+      "loss": 1.0072,
+      "step": 327
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.960989877077747,
+      "learning_rate": 9.999301596394728e-06,
+      "loss": 1.0839,
+      "step": 328
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.8300759682454655,
+      "learning_rate": 9.999272193831294e-06,
+      "loss": 0.9527,
+      "step": 329
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7395742534137697,
+      "learning_rate": 9.999242185088871e-06,
+      "loss": 0.9977,
+      "step": 330
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.790302817546493,
+      "learning_rate": 9.999211570171098e-06,
+      "loss": 1.0114,
+      "step": 331
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9012583446470444,
+      "learning_rate": 9.999180349081688e-06,
+      "loss": 0.981,
+      "step": 332
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7391960130633908,
+      "learning_rate": 9.999148521824424e-06,
+      "loss": 1.0103,
+      "step": 333
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8080327218213963,
+      "learning_rate": 9.999116088403168e-06,
+      "loss": 1.0077,
+      "step": 334
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.789355024851214,
+      "learning_rate": 9.999083048821852e-06,
+      "loss": 1.0322,
+      "step": 335
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7604125625680604,
+      "learning_rate": 9.999049403084481e-06,
+      "loss": 1.0077,
+      "step": 336
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.751969691466195,
+      "learning_rate": 9.999015151195137e-06,
+      "loss": 0.9632,
+      "step": 337
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9422430315733938,
+      "learning_rate": 9.99898029315797e-06,
+      "loss": 1.045,
+      "step": 338
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8526265905252293,
+      "learning_rate": 9.998944828977209e-06,
+      "loss": 0.9418,
+      "step": 339
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7598020549446747,
+      "learning_rate": 9.998908758657153e-06,
+      "loss": 0.9381,
+      "step": 340
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.947987701677072,
+      "learning_rate": 9.998872082202178e-06,
+      "loss": 0.9885,
+      "step": 341
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8449518459226886,
+      "learning_rate": 9.998834799616729e-06,
+      "loss": 0.9486,
+      "step": 342
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7242455557867136,
+      "learning_rate": 9.998796910905326e-06,
+      "loss": 1.0132,
+      "step": 343
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8378995486721486,
+      "learning_rate": 9.998758416072563e-06,
+      "loss": 0.9658,
+      "step": 344
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7452730650892918,
+      "learning_rate": 9.99871931512311e-06,
+      "loss": 0.9142,
+      "step": 345
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7216184771824707,
+      "learning_rate": 9.998679608061706e-06,
+      "loss": 0.8848,
+      "step": 346
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.771244228117099,
+      "learning_rate": 9.998639294893168e-06,
+      "loss": 0.9513,
+      "step": 347
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8524535352268292,
+      "learning_rate": 9.998598375622382e-06,
+      "loss": 0.878,
+      "step": 348
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7574556626857367,
+      "learning_rate": 9.99855685025431e-06,
+      "loss": 0.9874,
+      "step": 349
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.936301479766512,
+      "learning_rate": 9.998514718793986e-06,
+      "loss": 1.01,
+      "step": 350
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.743238898998547,
+      "learning_rate": 9.998471981246522e-06,
+      "loss": 0.9206,
+      "step": 351
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.908407426072886,
+      "learning_rate": 9.998428637617096e-06,
+      "loss": 1.0315,
+      "step": 352
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.887323151948872,
+      "learning_rate": 9.998384687910968e-06,
+      "loss": 1.0054,
+      "step": 353
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8689358502968432,
+      "learning_rate": 9.998340132133462e-06,
+      "loss": 0.9786,
+      "step": 354
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.766866877247742,
+      "learning_rate": 9.998294970289984e-06,
+      "loss": 0.974,
+      "step": 355
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.668049969560548,
+      "learning_rate": 9.998249202386008e-06,
+      "loss": 0.9332,
+      "step": 356
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7810907957387372,
+      "learning_rate": 9.998202828427087e-06,
+      "loss": 1.072,
+      "step": 357
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7810003122175218,
+      "learning_rate": 9.99815584841884e-06,
+      "loss": 1.0325,
+      "step": 358
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8694430603696126,
+      "learning_rate": 9.998108262366965e-06,
+      "loss": 0.9572,
+      "step": 359
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1378034244315067,
+      "learning_rate": 9.998060070277232e-06,
+      "loss": 0.9627,
+      "step": 360
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8597088740879761,
+      "learning_rate": 9.998011272155486e-06,
+      "loss": 0.9268,
+      "step": 361
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9075515554522582,
+      "learning_rate": 9.997961868007641e-06,
+      "loss": 0.9384,
+      "step": 362
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1093648375171188,
+      "learning_rate": 9.99791185783969e-06,
+      "loss": 1.0893,
+      "step": 363
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6774363515926407,
+      "learning_rate": 9.997861241657695e-06,
+      "loss": 0.9866,
+      "step": 364
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9226192210210127,
+      "learning_rate": 9.997810019467794e-06,
+      "loss": 0.9917,
+      "step": 365
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7369964385383962,
+      "learning_rate": 9.997758191276199e-06,
+      "loss": 0.9783,
+      "step": 366
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.768290530051918,
+      "learning_rate": 9.997705757089193e-06,
+      "loss": 0.9573,
+      "step": 367
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7267396405436175,
+      "learning_rate": 9.997652716913135e-06,
+      "loss": 0.9405,
+      "step": 368
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7974486258138789,
+      "learning_rate": 9.997599070754454e-06,
+      "loss": 0.972,
+      "step": 369
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7718921771971852,
+      "learning_rate": 9.997544818619658e-06,
+      "loss": 0.9679,
+      "step": 370
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7677117622153677,
+      "learning_rate": 9.99748996051532e-06,
+      "loss": 1.0525,
+      "step": 371
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8489389044812485,
+      "learning_rate": 9.997434496448099e-06,
+      "loss": 0.9741,
+      "step": 372
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8615316587238664,
+      "learning_rate": 9.997378426424716e-06,
+      "loss": 0.9277,
+      "step": 373
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7626686084934176,
+      "learning_rate": 9.99732175045197e-06,
+      "loss": 1.0037,
+      "step": 374
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.842797240944744,
+      "learning_rate": 9.997264468536734e-06,
+      "loss": 0.9722,
+      "step": 375
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.740147069318428,
+      "learning_rate": 9.99720658068595e-06,
+      "loss": 0.9897,
+      "step": 376
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4163507091164966,
+      "learning_rate": 9.997148086906643e-06,
+      "loss": 1.0379,
+      "step": 377
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7718206364251659,
+      "learning_rate": 9.997088987205902e-06,
+      "loss": 1.0424,
+      "step": 378
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7341436912414976,
+      "learning_rate": 9.997029281590893e-06,
+      "loss": 0.9789,
+      "step": 379
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8998843933957248,
+      "learning_rate": 9.996968970068857e-06,
+      "loss": 0.9529,
+      "step": 380
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8303203647864688,
+      "learning_rate": 9.996908052647107e-06,
+      "loss": 1.0387,
+      "step": 381
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0217903943934874,
+      "learning_rate": 9.996846529333028e-06,
+      "loss": 1.0489,
+      "step": 382
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7377397366551048,
+      "learning_rate": 9.99678440013408e-06,
+      "loss": 0.8791,
+      "step": 383
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9158974151548998,
+      "learning_rate": 9.996721665057796e-06,
+      "loss": 0.8786,
+      "step": 384
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8046078589216081,
+      "learning_rate": 9.996658324111786e-06,
+      "loss": 0.9756,
+      "step": 385
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.723119485274888,
+      "learning_rate": 9.996594377303727e-06,
+      "loss": 0.9463,
+      "step": 386
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.822813652888169,
+      "learning_rate": 9.996529824641375e-06,
+      "loss": 0.919,
+      "step": 387
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7650058929683152,
+      "learning_rate": 9.996464666132554e-06,
+      "loss": 0.9869,
+      "step": 388
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8839699219901074,
+      "learning_rate": 9.996398901785166e-06,
+      "loss": 1.0376,
+      "step": 389
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.989083906714401,
+      "learning_rate": 9.996332531607189e-06,
+      "loss": 0.9248,
+      "step": 390
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9767177439526261,
+      "learning_rate": 9.996265555606666e-06,
+      "loss": 1.1028,
+      "step": 391
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8001547919049068,
+      "learning_rate": 9.996197973791722e-06,
+      "loss": 0.9378,
+      "step": 392
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8601221209189258,
+      "learning_rate": 9.996129786170545e-06,
+      "loss": 0.9148,
+      "step": 393
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7460548937345546,
+      "learning_rate": 9.996060992751411e-06,
+      "loss": 0.9483,
+      "step": 394
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.948467329702224,
+      "learning_rate": 9.995991593542655e-06,
+      "loss": 1.0333,
+      "step": 395
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6954938973889058,
+      "learning_rate": 9.995921588552694e-06,
+      "loss": 0.9458,
+      "step": 396
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7805111411197816,
+      "learning_rate": 9.99585097779002e-06,
+      "loss": 1.0234,
+      "step": 397
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8173023960655723,
+      "learning_rate": 9.995779761263188e-06,
+      "loss": 0.9241,
+      "step": 398
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.780629413302553,
+      "learning_rate": 9.995707938980841e-06,
+      "loss": 0.9113,
+      "step": 399
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.923387366399058,
+      "learning_rate": 9.99563551095168e-06,
+      "loss": 0.9415,
+      "step": 400
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1846628128580843,
+      "learning_rate": 9.995562477184492e-06,
+      "loss": 1.0783,
+      "step": 401
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.067450718611197,
+      "learning_rate": 9.995488837688132e-06,
+      "loss": 0.9961,
+      "step": 402
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8192408543082925,
+      "learning_rate": 9.995414592471528e-06,
+      "loss": 0.9835,
+      "step": 403
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9682916396457573,
+      "learning_rate": 9.995339741543682e-06,
+      "loss": 0.9839,
+      "step": 404
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8957509477365946,
+      "learning_rate": 9.995264284913671e-06,
+      "loss": 1.0281,
+      "step": 405
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8091428564274485,
+      "learning_rate": 9.995188222590646e-06,
+      "loss": 0.989,
+      "step": 406
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7407792514054405,
+      "learning_rate": 9.995111554583825e-06,
+      "loss": 0.9978,
+      "step": 407
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.826377586139703,
+      "learning_rate": 9.995034280902508e-06,
+      "loss": 1.001,
+      "step": 408
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7977044321030833,
+      "learning_rate": 9.994956401556067e-06,
+      "loss": 0.8442,
+      "step": 409
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7611367649859104,
+      "learning_rate": 9.994877916553937e-06,
+      "loss": 1.0238,
+      "step": 410
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9552743499553855,
+      "learning_rate": 9.994798825905644e-06,
+      "loss": 1.0599,
+      "step": 411
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2054686137262969,
+      "learning_rate": 9.99471912962077e-06,
+      "loss": 1.0891,
+      "step": 412
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6706010293151303,
+      "learning_rate": 9.994638827708986e-06,
+      "loss": 0.9328,
+      "step": 413
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.948189979164849,
+      "learning_rate": 9.994557920180024e-06,
+      "loss": 1.0057,
+      "step": 414
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8265247740545638,
+      "learning_rate": 9.994476407043694e-06,
+      "loss": 0.895,
+      "step": 415
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.5305730347541093,
+      "learning_rate": 9.99439428830988e-06,
+      "loss": 0.953,
+      "step": 416
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9373030008073546,
+      "learning_rate": 9.994311563988543e-06,
+      "loss": 0.916,
+      "step": 417
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8555670580350845,
+      "learning_rate": 9.99422823408971e-06,
+      "loss": 0.9907,
+      "step": 418
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.763068302574431,
+      "learning_rate": 9.994144298623485e-06,
+      "loss": 0.875,
+      "step": 419
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2126877008153958,
+      "learning_rate": 9.994059757600047e-06,
+      "loss": 1.1077,
+      "step": 420
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0319541692222705,
+      "learning_rate": 9.993974611029646e-06,
+      "loss": 1.0642,
+      "step": 421
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7862262757935998,
+      "learning_rate": 9.993888858922606e-06,
+      "loss": 0.9159,
+      "step": 422
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8430679573996356,
+      "learning_rate": 9.993802501289327e-06,
+      "loss": 0.9196,
+      "step": 423
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2034916194647973,
+      "learning_rate": 9.993715538140277e-06,
+      "loss": 1.0918,
+      "step": 424
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1434693347387117,
+      "learning_rate": 9.993627969486003e-06,
+      "loss": 1.0497,
+      "step": 425
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9667992296409438,
+      "learning_rate": 9.993539795337122e-06,
+      "loss": 1.0522,
+      "step": 426
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7356235034275533,
+      "learning_rate": 9.993451015704324e-06,
+      "loss": 0.8628,
+      "step": 427
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.03162997263821,
+      "learning_rate": 9.993361630598377e-06,
+      "loss": 1.0795,
+      "step": 428
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.095996321926192,
+      "learning_rate": 9.993271640030116e-06,
+      "loss": 1.054,
+      "step": 429
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8531287594286143,
+      "learning_rate": 9.993181044010456e-06,
+      "loss": 0.9752,
+      "step": 430
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7220756831092734,
+      "learning_rate": 9.993089842550379e-06,
+      "loss": 0.9634,
+      "step": 431
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7837394850945063,
+      "learning_rate": 9.992998035660946e-06,
+      "loss": 0.9995,
+      "step": 432
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7110706526802253,
+      "learning_rate": 9.992905623353287e-06,
+      "loss": 0.9145,
+      "step": 433
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6534140467040328,
+      "learning_rate": 9.992812605638606e-06,
+      "loss": 0.9078,
+      "step": 434
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.746210291124398,
+      "learning_rate": 9.992718982528187e-06,
+      "loss": 0.8916,
+      "step": 435
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.935310891172882,
+      "learning_rate": 9.992624754033377e-06,
+      "loss": 0.9946,
+      "step": 436
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7873889464337964,
+      "learning_rate": 9.992529920165603e-06,
+      "loss": 0.8361,
+      "step": 437
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8612968636748695,
+      "learning_rate": 9.992434480936365e-06,
+      "loss": 1.0078,
+      "step": 438
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7918394896269425,
+      "learning_rate": 9.992338436357235e-06,
+      "loss": 0.9402,
+      "step": 439
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9078193197942677,
+      "learning_rate": 9.992241786439859e-06,
+      "loss": 0.9495,
+      "step": 440
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8319620670045862,
+      "learning_rate": 9.992144531195955e-06,
+      "loss": 0.935,
+      "step": 441
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8449293510185172,
+      "learning_rate": 9.992046670637316e-06,
+      "loss": 0.9734,
+      "step": 442
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8730803588651936,
+      "learning_rate": 9.991948204775808e-06,
+      "loss": 0.9007,
+      "step": 443
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8714182092309293,
+      "learning_rate": 9.99184913362337e-06,
+      "loss": 1.009,
+      "step": 444
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.811743033637239,
+      "learning_rate": 9.991749457192015e-06,
+      "loss": 1.0229,
+      "step": 445
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.89992424778901,
+      "learning_rate": 9.991649175493829e-06,
+      "loss": 0.9306,
+      "step": 446
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.767758615947386,
+      "learning_rate": 9.991548288540971e-06,
+      "loss": 0.8862,
+      "step": 447
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.717951201708389,
+      "learning_rate": 9.991446796345676e-06,
+      "loss": 0.9847,
+      "step": 448
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8102586640244867,
+      "learning_rate": 9.991344698920248e-06,
+      "loss": 1.0063,
+      "step": 449
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8310093291758218,
+      "learning_rate": 9.991241996277068e-06,
+      "loss": 0.8896,
+      "step": 450
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.770758327814944,
+      "learning_rate": 9.991138688428588e-06,
+      "loss": 1.0795,
+      "step": 451
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8460990532070476,
+      "learning_rate": 9.991034775387336e-06,
+      "loss": 0.9816,
+      "step": 452
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6429231499080639,
+      "learning_rate": 9.99093025716591e-06,
+      "loss": 0.9052,
+      "step": 453
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9432220732897085,
+      "learning_rate": 9.990825133776983e-06,
+      "loss": 1.0062,
+      "step": 454
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7485747471408606,
+      "learning_rate": 9.990719405233302e-06,
+      "loss": 0.9437,
+      "step": 455
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.790664740982909,
+      "learning_rate": 9.99061307154769e-06,
+      "loss": 1.0697,
+      "step": 456
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8680188794611297,
+      "learning_rate": 9.990506132733036e-06,
+      "loss": 1.025,
+      "step": 457
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2723494579556252,
+      "learning_rate": 9.99039858880231e-06,
+      "loss": 1.0975,
+      "step": 458
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7684429078817845,
+      "learning_rate": 9.99029043976855e-06,
+      "loss": 0.9523,
+      "step": 459
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9800672354643611,
+      "learning_rate": 9.990181685644869e-06,
+      "loss": 0.9782,
+      "step": 460
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7497641597424318,
+      "learning_rate": 9.990072326444456e-06,
+      "loss": 0.8998,
+      "step": 461
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7678804701646167,
+      "learning_rate": 9.98996236218057e-06,
+      "loss": 1.0159,
+      "step": 462
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8333399979742488,
+      "learning_rate": 9.989851792866544e-06,
+      "loss": 0.9202,
+      "step": 463
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8342458871717588,
+      "learning_rate": 9.989740618515787e-06,
+      "loss": 0.8786,
+      "step": 464
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8275362399611232,
+      "learning_rate": 9.989628839141777e-06,
+      "loss": 0.9931,
+      "step": 465
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.776468802584567,
+      "learning_rate": 9.989516454758066e-06,
+      "loss": 0.9564,
+      "step": 466
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.779785337605654,
+      "learning_rate": 9.989403465378286e-06,
+      "loss": 0.9898,
+      "step": 467
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2474927320282596,
+      "learning_rate": 9.989289871016134e-06,
+      "loss": 1.057,
+      "step": 468
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8832695871158862,
+      "learning_rate": 9.989175671685383e-06,
+      "loss": 0.9224,
+      "step": 469
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8812481153777465,
+      "learning_rate": 9.989060867399884e-06,
+      "loss": 1.075,
+      "step": 470
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.852361736150292,
+      "learning_rate": 9.988945458173553e-06,
+      "loss": 1.0411,
+      "step": 471
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7585999424822527,
+      "learning_rate": 9.988829444020388e-06,
+      "loss": 0.9183,
+      "step": 472
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9517448301169975,
+      "learning_rate": 9.988712824954452e-06,
+      "loss": 0.9947,
+      "step": 473
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.928191014054902,
+      "learning_rate": 9.988595600989887e-06,
+      "loss": 0.966,
+      "step": 474
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7355021778443462,
+      "learning_rate": 9.988477772140908e-06,
+      "loss": 0.9682,
+      "step": 475
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8370788754365992,
+      "learning_rate": 9.988359338421801e-06,
+      "loss": 1.0217,
+      "step": 476
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8807883614600507,
+      "learning_rate": 9.988240299846927e-06,
+      "loss": 0.9342,
+      "step": 477
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3272969544545015,
+      "learning_rate": 9.988120656430718e-06,
+      "loss": 1.0586,
+      "step": 478
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9627495429852833,
+      "learning_rate": 9.988000408187684e-06,
+      "loss": 0.9488,
+      "step": 479
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8990878635514887,
+      "learning_rate": 9.987879555132404e-06,
+      "loss": 0.9213,
+      "step": 480
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9523811405836686,
+      "learning_rate": 9.987758097279535e-06,
+      "loss": 0.9519,
+      "step": 481
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9348596004746337,
+      "learning_rate": 9.987636034643798e-06,
+      "loss": 0.9765,
+      "step": 482
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8986060031488083,
+      "learning_rate": 9.987513367239995e-06,
+      "loss": 0.8812,
+      "step": 483
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7834458054495186,
+      "learning_rate": 9.987390095083005e-06,
+      "loss": 0.9296,
+      "step": 484
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7732203264226643,
+      "learning_rate": 9.987266218187771e-06,
+      "loss": 0.8164,
+      "step": 485
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9347718463354657,
+      "learning_rate": 9.987141736569314e-06,
+      "loss": 0.9611,
+      "step": 486
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8765773287844287,
+      "learning_rate": 9.98701665024273e-06,
+      "loss": 0.9491,
+      "step": 487
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8150677633824404,
+      "learning_rate": 9.986890959223181e-06,
+      "loss": 0.9206,
+      "step": 488
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8867460056400567,
+      "learning_rate": 9.986764663525913e-06,
+      "loss": 0.9396,
+      "step": 489
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8298813753799374,
+      "learning_rate": 9.986637763166236e-06,
+      "loss": 1.0279,
+      "step": 490
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6398417232609694,
+      "learning_rate": 9.986510258159541e-06,
+      "loss": 0.9621,
+      "step": 491
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7268728251216694,
+      "learning_rate": 9.986382148521283e-06,
+      "loss": 0.9665,
+      "step": 492
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8633207629215438,
+      "learning_rate": 9.986253434267002e-06,
+      "loss": 0.8896,
+      "step": 493
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8284720123352507,
+      "learning_rate": 9.9861241154123e-06,
+      "loss": 0.965,
+      "step": 494
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8156493599253845,
+      "learning_rate": 9.98599419197286e-06,
+      "loss": 1.0073,
+      "step": 495
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9796908925385863,
+      "learning_rate": 9.985863663964434e-06,
+      "loss": 1.0066,
+      "step": 496
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8337812904898183,
+      "learning_rate": 9.98573253140285e-06,
+      "loss": 0.9401,
+      "step": 497
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7559436180098102,
+      "learning_rate": 9.98560079430401e-06,
+      "loss": 1.019,
+      "step": 498
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1679472626168819,
+      "learning_rate": 9.985468452683882e-06,
+      "loss": 1.072,
+      "step": 499
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.922452023770267,
+      "learning_rate": 9.98533550655852e-06,
+      "loss": 1.0815,
+      "step": 500
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7237522608965452,
+      "learning_rate": 9.985201955944039e-06,
+      "loss": 0.891,
+      "step": 501
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8582005561984585,
+      "learning_rate": 9.985067800856636e-06,
+      "loss": 1.0206,
+      "step": 502
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7448129739558724,
+      "learning_rate": 9.984933041312575e-06,
+      "loss": 0.9692,
+      "step": 503
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8282086515360507,
+      "learning_rate": 9.984797677328195e-06,
+      "loss": 0.9114,
+      "step": 504
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7351393366243573,
+      "learning_rate": 9.984661708919913e-06,
+      "loss": 0.9995,
+      "step": 505
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7591151114795303,
+      "learning_rate": 9.984525136104214e-06,
+      "loss": 1.0112,
+      "step": 506
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.868634073105971,
+      "learning_rate": 9.98438795889766e-06,
+      "loss": 0.9551,
+      "step": 507
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7998162548535677,
+      "learning_rate": 9.984250177316881e-06,
+      "loss": 0.9458,
+      "step": 508
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8821961179539208,
+      "learning_rate": 9.984111791378583e-06,
+      "loss": 0.9718,
+      "step": 509
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7585068521771228,
+      "learning_rate": 9.983972801099548e-06,
+      "loss": 0.9697,
+      "step": 510
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7915986259264736,
+      "learning_rate": 9.98383320649663e-06,
+      "loss": 0.9561,
+      "step": 511
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.678142637420618,
+      "learning_rate": 9.983693007586754e-06,
+      "loss": 0.8566,
+      "step": 512
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.749509545081998,
+      "learning_rate": 9.983552204386916e-06,
+      "loss": 0.936,
+      "step": 513
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.834234008138445,
+      "learning_rate": 9.983410796914197e-06,
+      "loss": 0.952,
+      "step": 514
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.033294029345438,
+      "learning_rate": 9.983268785185736e-06,
+      "loss": 1.003,
+      "step": 515
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.888426292024446,
+      "learning_rate": 9.983126169218754e-06,
+      "loss": 1.036,
+      "step": 516
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8609013930333695,
+      "learning_rate": 9.982982949030546e-06,
+      "loss": 0.9874,
+      "step": 517
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0116723727488406,
+      "learning_rate": 9.982839124638474e-06,
+      "loss": 1.0292,
+      "step": 518
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9039604388176836,
+      "learning_rate": 9.982694696059982e-06,
+      "loss": 0.9762,
+      "step": 519
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7936355511215125,
+      "learning_rate": 9.982549663312582e-06,
+      "loss": 0.8692,
+      "step": 520
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7573773134145743,
+      "learning_rate": 9.982404026413854e-06,
+      "loss": 0.9193,
+      "step": 521
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.874692404142349,
+      "learning_rate": 9.982257785381464e-06,
+      "loss": 0.9192,
+      "step": 522
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.761098423029259,
+      "learning_rate": 9.982110940233139e-06,
+      "loss": 1.0955,
+      "step": 523
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6651444864648541,
+      "learning_rate": 9.981963490986687e-06,
+      "loss": 0.9885,
+      "step": 524
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.863108919064453,
+      "learning_rate": 9.981815437659987e-06,
+      "loss": 0.9783,
+      "step": 525
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7341218868163277,
+      "learning_rate": 9.98166678027099e-06,
+      "loss": 0.859,
+      "step": 526
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8583856446853357,
+      "learning_rate": 9.98151751883772e-06,
+      "loss": 0.9862,
+      "step": 527
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7902381185921945,
+      "learning_rate": 9.981367653378278e-06,
+      "loss": 0.9678,
+      "step": 528
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7232482103771993,
+      "learning_rate": 9.981217183910834e-06,
+      "loss": 0.9854,
+      "step": 529
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7245166406569976,
+      "learning_rate": 9.981066110453634e-06,
+      "loss": 1.044,
+      "step": 530
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8755340432400052,
+      "learning_rate": 9.980914433024997e-06,
+      "loss": 0.9439,
+      "step": 531
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8180651767153069,
+      "learning_rate": 9.980762151643313e-06,
+      "loss": 0.9877,
+      "step": 532
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6642117033393633,
+      "learning_rate": 9.980609266327046e-06,
+      "loss": 0.9304,
+      "step": 533
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.770751243847106,
+      "learning_rate": 9.980455777094733e-06,
+      "loss": 0.9437,
+      "step": 534
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6630684700108163,
+      "learning_rate": 9.98030168396499e-06,
+      "loss": 0.9379,
+      "step": 535
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8048619734291211,
+      "learning_rate": 9.980146986956495e-06,
+      "loss": 0.9545,
+      "step": 536
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8298762898602046,
+      "learning_rate": 9.979991686088012e-06,
+      "loss": 0.9512,
+      "step": 537
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9104819626973562,
+      "learning_rate": 9.979835781378366e-06,
+      "loss": 1.0086,
+      "step": 538
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8346423261750149,
+      "learning_rate": 9.979679272846463e-06,
+      "loss": 0.9459,
+      "step": 539
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7542480905298683,
+      "learning_rate": 9.979522160511282e-06,
+      "loss": 1.0184,
+      "step": 540
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7604193676146613,
+      "learning_rate": 9.979364444391873e-06,
+      "loss": 0.9596,
+      "step": 541
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7285763067588011,
+      "learning_rate": 9.979206124507357e-06,
+      "loss": 0.9392,
+      "step": 542
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7086746994545379,
+      "learning_rate": 9.979047200876932e-06,
+      "loss": 0.9275,
+      "step": 543
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8537858338204671,
+      "learning_rate": 9.97888767351987e-06,
+      "loss": 1.0081,
+      "step": 544
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8793186820578498,
+      "learning_rate": 9.978727542455512e-06,
+      "loss": 0.9517,
+      "step": 545
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7469288461512746,
+      "learning_rate": 9.978566807703275e-06,
+      "loss": 1.1121,
+      "step": 546
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8215703390582285,
+      "learning_rate": 9.978405469282648e-06,
+      "loss": 0.9125,
+      "step": 547
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1155966186705373,
+      "learning_rate": 9.978243527213196e-06,
+      "loss": 1.0629,
+      "step": 548
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7365975649787089,
+      "learning_rate": 9.978080981514554e-06,
+      "loss": 0.9202,
+      "step": 549
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7687962806959705,
+      "learning_rate": 9.977917832206431e-06,
+      "loss": 0.9539,
+      "step": 550
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7911541862755394,
+      "learning_rate": 9.977754079308608e-06,
+      "loss": 0.9884,
+      "step": 551
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9017310930732227,
+      "learning_rate": 9.977589722840943e-06,
+      "loss": 0.9943,
+      "step": 552
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.138313033014973,
+      "learning_rate": 9.977424762823364e-06,
+      "loss": 0.845,
+      "step": 553
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8215438597774032,
+      "learning_rate": 9.977259199275871e-06,
+      "loss": 0.9565,
+      "step": 554
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9462909274871398,
+      "learning_rate": 9.977093032218544e-06,
+      "loss": 1.0496,
+      "step": 555
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8921178982747826,
+      "learning_rate": 9.976926261671524e-06,
+      "loss": 0.9938,
+      "step": 556
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8287017846381217,
+      "learning_rate": 9.97675888765504e-06,
+      "loss": 0.9369,
+      "step": 557
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7895611904073543,
+      "learning_rate": 9.976590910189382e-06,
+      "loss": 0.9526,
+      "step": 558
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.846818399416353,
+      "learning_rate": 9.976422329294919e-06,
+      "loss": 0.9289,
+      "step": 559
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8702061937596497,
+      "learning_rate": 9.976253144992094e-06,
+      "loss": 0.9949,
+      "step": 560
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.955839330032564,
+      "learning_rate": 9.976083357301419e-06,
+      "loss": 0.9309,
+      "step": 561
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8023863188200453,
+      "learning_rate": 9.97591296624348e-06,
+      "loss": 0.9942,
+      "step": 562
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7903171343827091,
+      "learning_rate": 9.97574197183894e-06,
+      "loss": 0.9201,
+      "step": 563
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8089959609657595,
+      "learning_rate": 9.975570374108531e-06,
+      "loss": 0.8746,
+      "step": 564
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8127768129223576,
+      "learning_rate": 9.975398173073064e-06,
+      "loss": 0.9709,
+      "step": 565
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9045708550657638,
+      "learning_rate": 9.975225368753412e-06,
+      "loss": 1.0174,
+      "step": 566
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7450261882038183,
+      "learning_rate": 9.975051961170533e-06,
+      "loss": 0.953,
+      "step": 567
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9166097491832965,
+      "learning_rate": 9.974877950345453e-06,
+      "loss": 0.9467,
+      "step": 568
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8483370251536706,
+      "learning_rate": 9.97470333629927e-06,
+      "loss": 0.9745,
+      "step": 569
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9224517206725216,
+      "learning_rate": 9.974528119053158e-06,
+      "loss": 1.0604,
+      "step": 570
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8135361467431292,
+      "learning_rate": 9.97435229862836e-06,
+      "loss": 0.989,
+      "step": 571
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7391744135358165,
+      "learning_rate": 9.974175875046197e-06,
+      "loss": 0.8963,
+      "step": 572
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8259354012034896,
+      "learning_rate": 9.973998848328062e-06,
+      "loss": 0.9903,
+      "step": 573
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.873932729050753,
+      "learning_rate": 9.973821218495417e-06,
+      "loss": 0.8828,
+      "step": 574
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7524517468682828,
+      "learning_rate": 9.973642985569803e-06,
+      "loss": 0.9067,
+      "step": 575
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0015384145181394,
+      "learning_rate": 9.973464149572829e-06,
+      "loss": 1.0994,
+      "step": 576
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.897425966186243,
+      "learning_rate": 9.97328471052618e-06,
+      "loss": 1.0066,
+      "step": 577
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2178348468792712,
+      "learning_rate": 9.973104668451617e-06,
+      "loss": 1.0475,
+      "step": 578
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.912013919624642,
+      "learning_rate": 9.972924023370967e-06,
+      "loss": 1.006,
+      "step": 579
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0686877461036826,
+      "learning_rate": 9.972742775306135e-06,
+      "loss": 1.0357,
+      "step": 580
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.017591536114641,
+      "learning_rate": 9.972560924279098e-06,
+      "loss": 1.0012,
+      "step": 581
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.2379380531776647,
+      "learning_rate": 9.972378470311906e-06,
+      "loss": 1.0291,
+      "step": 582
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9020642874721987,
+      "learning_rate": 9.97219541342668e-06,
+      "loss": 0.9986,
+      "step": 583
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.892875164281699,
+      "learning_rate": 9.97201175364562e-06,
+      "loss": 0.9579,
+      "step": 584
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.830561365390589,
+      "learning_rate": 9.971827490990993e-06,
+      "loss": 0.9995,
+      "step": 585
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9935730223840498,
+      "learning_rate": 9.971642625485144e-06,
+      "loss": 0.9408,
+      "step": 586
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8769784060918686,
+      "learning_rate": 9.971457157150485e-06,
+      "loss": 0.9592,
+      "step": 587
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.04349603128331,
+      "learning_rate": 9.971271086009509e-06,
+      "loss": 0.9998,
+      "step": 588
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9254072544653607,
+      "learning_rate": 9.971084412084773e-06,
+      "loss": 0.9745,
+      "step": 589
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.2928783245331386,
+      "learning_rate": 9.970897135398914e-06,
+      "loss": 1.1144,
+      "step": 590
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.859779802438604,
+      "learning_rate": 9.97070925597464e-06,
+      "loss": 1.0651,
+      "step": 591
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9521111212451354,
+      "learning_rate": 9.970520773834734e-06,
+      "loss": 0.8919,
+      "step": 592
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0769150280121682,
+      "learning_rate": 9.970331689002046e-06,
+      "loss": 1.0606,
+      "step": 593
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9367091767594273,
+      "learning_rate": 9.970142001499505e-06,
+      "loss": 0.932,
+      "step": 594
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9241743588307845,
+      "learning_rate": 9.969951711350114e-06,
+      "loss": 0.9728,
+      "step": 595
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7815244053234291,
+      "learning_rate": 9.969760818576943e-06,
+      "loss": 0.9051,
+      "step": 596
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7541410479484068,
+      "learning_rate": 9.969569323203138e-06,
+      "loss": 0.9613,
+      "step": 597
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9976257598875988,
+      "learning_rate": 9.96937722525192e-06,
+      "loss": 0.9828,
+      "step": 598
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9730089547222298,
+      "learning_rate": 9.969184524746585e-06,
+      "loss": 0.9612,
+      "step": 599
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9912894581740013,
+      "learning_rate": 9.96899122171049e-06,
+      "loss": 0.9905,
+      "step": 600
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8575772946937077,
+      "learning_rate": 9.968797316167083e-06,
+      "loss": 0.958,
+      "step": 601
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8819876995599274,
+      "learning_rate": 9.96860280813987e-06,
+      "loss": 0.9882,
+      "step": 602
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.6457416407968006,
+      "learning_rate": 9.968407697652436e-06,
+      "loss": 1.0823,
+      "step": 603
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.817364641800979,
+      "learning_rate": 9.96821198472844e-06,
+      "loss": 0.9558,
+      "step": 604
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8771531816842049,
+      "learning_rate": 9.968015669391613e-06,
+      "loss": 1.0313,
+      "step": 605
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.837738421712122,
+      "learning_rate": 9.967818751665759e-06,
+      "loss": 1.0452,
+      "step": 606
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9799716024859662,
+      "learning_rate": 9.967621231574754e-06,
+      "loss": 0.9136,
+      "step": 607
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8973139433595638,
+      "learning_rate": 9.96742310914255e-06,
+      "loss": 1.0104,
+      "step": 608
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8616875016391605,
+      "learning_rate": 9.967224384393168e-06,
+      "loss": 1.0116,
+      "step": 609
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9552365160279477,
+      "learning_rate": 9.967025057350705e-06,
+      "loss": 0.9468,
+      "step": 610
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8682790571888328,
+      "learning_rate": 9.96682512803933e-06,
+      "loss": 0.934,
+      "step": 611
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8512425230724565,
+      "learning_rate": 9.966624596483284e-06,
+      "loss": 1.0202,
+      "step": 612
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7897736543785032,
+      "learning_rate": 9.966423462706885e-06,
+      "loss": 0.9423,
+      "step": 613
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8637879693105739,
+      "learning_rate": 9.966221726734518e-06,
+      "loss": 1.016,
+      "step": 614
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8289382477344223,
+      "learning_rate": 9.966019388590645e-06,
+      "loss": 1.0386,
+      "step": 615
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7184673263447736,
+      "learning_rate": 9.965816448299801e-06,
+      "loss": 0.9467,
+      "step": 616
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9270740641333812,
+      "learning_rate": 9.965612905886593e-06,
+      "loss": 0.9502,
+      "step": 617
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8449413187934345,
+      "learning_rate": 9.965408761375702e-06,
+      "loss": 0.9439,
+      "step": 618
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8221184142114624,
+      "learning_rate": 9.965204014791879e-06,
+      "loss": 0.9395,
+      "step": 619
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8938004785244031,
+      "learning_rate": 9.964998666159952e-06,
+      "loss": 0.9725,
+      "step": 620
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0583541343532343,
+      "learning_rate": 9.96479271550482e-06,
+      "loss": 0.854,
+      "step": 621
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.956440267051593,
+      "learning_rate": 9.964586162851455e-06,
+      "loss": 0.9236,
+      "step": 622
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.775693996598594,
+      "learning_rate": 9.964379008224902e-06,
+      "loss": 0.9357,
+      "step": 623
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6530095208317594,
+      "learning_rate": 9.96417125165028e-06,
+      "loss": 0.9268,
+      "step": 624
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9327439029302136,
+      "learning_rate": 9.963962893152779e-06,
+      "loss": 0.9302,
+      "step": 625
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8169518109892377,
+      "learning_rate": 9.963753932757663e-06,
+      "loss": 0.9559,
+      "step": 626
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9687021807597014,
+      "learning_rate": 9.96354437049027e-06,
+      "loss": 0.9264,
+      "step": 627
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7503101529375042,
+      "learning_rate": 9.963334206376011e-06,
+      "loss": 0.9384,
+      "step": 628
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0987366385180621,
+      "learning_rate": 9.96312344044037e-06,
+      "loss": 1.0687,
+      "step": 629
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7913385847332246,
+      "learning_rate": 9.962912072708898e-06,
+      "loss": 0.9251,
+      "step": 630
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7342714295246835,
+      "learning_rate": 9.96270010320723e-06,
+      "loss": 1.038,
+      "step": 631
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7780518676028247,
+      "learning_rate": 9.962487531961064e-06,
+      "loss": 0.9799,
+      "step": 632
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.788720066371652,
+      "learning_rate": 9.962274358996178e-06,
+      "loss": 0.9225,
+      "step": 633
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7654704240372205,
+      "learning_rate": 9.962060584338416e-06,
+      "loss": 0.9222,
+      "step": 634
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7625696357265461,
+      "learning_rate": 9.961846208013703e-06,
+      "loss": 0.9726,
+      "step": 635
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6841560229671029,
+      "learning_rate": 9.961631230048032e-06,
+      "loss": 0.9397,
+      "step": 636
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9103175919775225,
+      "learning_rate": 9.961415650467468e-06,
+      "loss": 1.0677,
+      "step": 637
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.768671834701704,
+      "learning_rate": 9.961199469298151e-06,
+      "loss": 1.0266,
+      "step": 638
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.78625536991518,
+      "learning_rate": 9.960982686566295e-06,
+      "loss": 0.9044,
+      "step": 639
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7628255178663492,
+      "learning_rate": 9.960765302298185e-06,
+      "loss": 0.8897,
+      "step": 640
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7237950091053256,
+      "learning_rate": 9.960547316520182e-06,
+      "loss": 0.9906,
+      "step": 641
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8392334080878924,
+      "learning_rate": 9.960328729258713e-06,
+      "loss": 1.0882,
+      "step": 642
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8005071344569594,
+      "learning_rate": 9.960109540540283e-06,
+      "loss": 1.0159,
+      "step": 643
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7761629381340716,
+      "learning_rate": 9.959889750391474e-06,
+      "loss": 1.0221,
+      "step": 644
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7060716439274553,
+      "learning_rate": 9.959669358838932e-06,
+      "loss": 0.9978,
+      "step": 645
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8953033852689745,
+      "learning_rate": 9.959448365909383e-06,
+      "loss": 0.9311,
+      "step": 646
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7863509286896406,
+      "learning_rate": 9.959226771629623e-06,
+      "loss": 0.8935,
+      "step": 647
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.728604807127596,
+      "learning_rate": 9.959004576026516e-06,
+      "loss": 0.9478,
+      "step": 648
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8419916827797762,
+      "learning_rate": 9.95878177912701e-06,
+      "loss": 0.9079,
+      "step": 649
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2087797719122222,
+      "learning_rate": 9.958558380958118e-06,
+      "loss": 1.0666,
+      "step": 650
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7791755105817448,
+      "learning_rate": 9.958334381546927e-06,
+      "loss": 0.9997,
+      "step": 651
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8707170135919433,
+      "learning_rate": 9.958109780920598e-06,
+      "loss": 0.8551,
+      "step": 652
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9541392269333469,
+      "learning_rate": 9.957884579106365e-06,
+      "loss": 0.9551,
+      "step": 653
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.905994967707482,
+      "learning_rate": 9.957658776131536e-06,
+      "loss": 0.9826,
+      "step": 654
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.899777785878718,
+      "learning_rate": 9.957432372023487e-06,
+      "loss": 0.9648,
+      "step": 655
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0979333253902503,
+      "learning_rate": 9.957205366809673e-06,
+      "loss": 1.1097,
+      "step": 656
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8775999868428007,
+      "learning_rate": 9.956977760517617e-06,
+      "loss": 1.0705,
+      "step": 657
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9362360310686466,
+      "learning_rate": 9.956749553174919e-06,
+      "loss": 0.9588,
+      "step": 658
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9388211625572105,
+      "learning_rate": 9.956520744809249e-06,
+      "loss": 0.9761,
+      "step": 659
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8519198696963795,
+      "learning_rate": 9.956291335448351e-06,
+      "loss": 1.003,
+      "step": 660
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8454020936116626,
+      "learning_rate": 9.95606132512004e-06,
+      "loss": 0.9225,
+      "step": 661
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8060025455197584,
+      "learning_rate": 9.95583071385221e-06,
+      "loss": 1.0429,
+      "step": 662
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7171723636224656,
+      "learning_rate": 9.95559950167282e-06,
+      "loss": 0.9743,
+      "step": 663
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8787139336762622,
+      "learning_rate": 9.955367688609904e-06,
+      "loss": 1.003,
+      "step": 664
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6749247472999584,
+      "learning_rate": 9.955135274691575e-06,
+      "loss": 0.934,
+      "step": 665
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7332859395977531,
+      "learning_rate": 9.954902259946008e-06,
+      "loss": 0.9183,
+      "step": 666
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9021831293417901,
+      "learning_rate": 9.954668644401462e-06,
+      "loss": 1.0768,
+      "step": 667
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.856143973447642,
+      "learning_rate": 9.95443442808626e-06,
+      "loss": 1.015,
+      "step": 668
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8090075540328545,
+      "learning_rate": 9.954199611028803e-06,
+      "loss": 0.9164,
+      "step": 669
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7650216485821708,
+      "learning_rate": 9.953964193257563e-06,
+      "loss": 1.0329,
+      "step": 670
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.708200062633118,
+      "learning_rate": 9.953728174801089e-06,
+      "loss": 0.9691,
+      "step": 671
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.842674374882094,
+      "learning_rate": 9.953491555687994e-06,
+      "loss": 1.066,
+      "step": 672
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7718209773301226,
+      "learning_rate": 9.95325433594697e-06,
+      "loss": 0.917,
+      "step": 673
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4880659865595764,
+      "learning_rate": 9.95301651560678e-06,
+      "loss": 1.0497,
+      "step": 674
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1691065286797593,
+      "learning_rate": 9.952778094696263e-06,
+      "loss": 1.0805,
+      "step": 675
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7628223270057626,
+      "learning_rate": 9.952539073244326e-06,
+      "loss": 0.9391,
+      "step": 676
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9530968609954608,
+      "learning_rate": 9.952299451279956e-06,
+      "loss": 1.0834,
+      "step": 677
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7191467821085589,
+      "learning_rate": 9.952059228832202e-06,
+      "loss": 0.987,
+      "step": 678
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9518416576120086,
+      "learning_rate": 9.951818405930194e-06,
+      "loss": 0.915,
+      "step": 679
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8824786204070405,
+      "learning_rate": 9.951576982603133e-06,
+      "loss": 1.0249,
+      "step": 680
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9540247887414999,
+      "learning_rate": 9.951334958880292e-06,
+      "loss": 0.9899,
+      "step": 681
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8103136107534201,
+      "learning_rate": 9.951092334791018e-06,
+      "loss": 0.9051,
+      "step": 682
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9171593119399666,
+      "learning_rate": 9.950849110364729e-06,
+      "loss": 1.0154,
+      "step": 683
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8212591907813733,
+      "learning_rate": 9.950605285630917e-06,
+      "loss": 0.9529,
+      "step": 684
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7956146149315781,
+      "learning_rate": 9.950360860619147e-06,
+      "loss": 0.9842,
+      "step": 685
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8475635658461347,
+      "learning_rate": 9.950115835359055e-06,
+      "loss": 0.9212,
+      "step": 686
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6732753547058545,
+      "learning_rate": 9.949870209880355e-06,
+      "loss": 0.9607,
+      "step": 687
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7881646180243027,
+      "learning_rate": 9.949623984212825e-06,
+      "loss": 0.9131,
+      "step": 688
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7632314528143287,
+      "learning_rate": 9.949377158386324e-06,
+      "loss": 0.9592,
+      "step": 689
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.72638724645718,
+      "learning_rate": 9.949129732430779e-06,
+      "loss": 0.9474,
+      "step": 690
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.75036474727148,
+      "learning_rate": 9.948881706376193e-06,
+      "loss": 0.9568,
+      "step": 691
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6302994955465049,
+      "learning_rate": 9.948633080252637e-06,
+      "loss": 1.0904,
+      "step": 692
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.720832326694608,
+      "learning_rate": 9.94838385409026e-06,
+      "loss": 0.9587,
+      "step": 693
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0458443644499276,
+      "learning_rate": 9.94813402791928e-06,
+      "loss": 1.0918,
+      "step": 694
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9654229081399957,
+      "learning_rate": 9.947883601769992e-06,
+      "loss": 0.9023,
+      "step": 695
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9860269735905858,
+      "learning_rate": 9.947632575672758e-06,
+      "loss": 0.9692,
+      "step": 696
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7071028298728652,
+      "learning_rate": 9.947380949658018e-06,
+      "loss": 0.9506,
+      "step": 697
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7538962637600446,
+      "learning_rate": 9.947128723756282e-06,
+      "loss": 0.8981,
+      "step": 698
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2947361655354426,
+      "learning_rate": 9.946875897998131e-06,
+      "loss": 1.0773,
+      "step": 699
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8541641733446406,
+      "learning_rate": 9.946622472414224e-06,
+      "loss": 0.9427,
+      "step": 700
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9265828758974637,
+      "learning_rate": 9.946368447035291e-06,
+      "loss": 0.9609,
+      "step": 701
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.81874713320279,
+      "learning_rate": 9.946113821892129e-06,
+      "loss": 0.9073,
+      "step": 702
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0227154371529525,
+      "learning_rate": 9.945858597015614e-06,
+      "loss": 1.0915,
+      "step": 703
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6918596382511752,
+      "learning_rate": 9.945602772436693e-06,
+      "loss": 0.9359,
+      "step": 704
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8193262539606974,
+      "learning_rate": 9.945346348186385e-06,
+      "loss": 0.9487,
+      "step": 705
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7522003758107525,
+      "learning_rate": 9.945089324295785e-06,
+      "loss": 0.9456,
+      "step": 706
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.896463535938412,
+      "learning_rate": 9.944831700796055e-06,
+      "loss": 0.9862,
+      "step": 707
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8321035180909384,
+      "learning_rate": 9.944573477718435e-06,
+      "loss": 1.0236,
+      "step": 708
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6590614298208861,
+      "learning_rate": 9.944314655094233e-06,
+      "loss": 0.9222,
+      "step": 709
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8113615471791165,
+      "learning_rate": 9.944055232954834e-06,
+      "loss": 1.0818,
+      "step": 710
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7238807744908444,
+      "learning_rate": 9.943795211331692e-06,
+      "loss": 0.9948,
+      "step": 711
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6970854361188064,
+      "learning_rate": 9.943534590256338e-06,
+      "loss": 0.993,
+      "step": 712
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.146543441409252,
+      "learning_rate": 9.943273369760373e-06,
+      "loss": 1.0946,
+      "step": 713
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8770392079366378,
+      "learning_rate": 9.943011549875468e-06,
+      "loss": 0.9566,
+      "step": 714
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9521917888426603,
+      "learning_rate": 9.942749130633371e-06,
+      "loss": 1.0605,
+      "step": 715
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7807992782522637,
+      "learning_rate": 9.942486112065903e-06,
+      "loss": 0.9806,
+      "step": 716
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9385575090178524,
+      "learning_rate": 9.942222494204956e-06,
+      "loss": 0.8975,
+      "step": 717
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7069462399149318,
+      "learning_rate": 9.941958277082491e-06,
+      "loss": 0.9507,
+      "step": 718
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8106305629744721,
+      "learning_rate": 9.941693460730549e-06,
+      "loss": 0.9822,
+      "step": 719
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8201826080289825,
+      "learning_rate": 9.941428045181236e-06,
+      "loss": 0.9576,
+      "step": 720
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.763219712486219,
+      "learning_rate": 9.941162030466738e-06,
+      "loss": 0.9385,
+      "step": 721
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8086066736383897,
+      "learning_rate": 9.940895416619308e-06,
+      "loss": 1.0173,
+      "step": 722
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4087884379310707,
+      "learning_rate": 9.940628203671277e-06,
+      "loss": 1.0614,
+      "step": 723
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7889845618807796,
+      "learning_rate": 9.940360391655042e-06,
+      "loss": 0.99,
+      "step": 724
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8224591862233448,
+      "learning_rate": 9.940091980603077e-06,
+      "loss": 1.0077,
+      "step": 725
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7157721857885622,
+      "learning_rate": 9.93982297054793e-06,
+      "loss": 0.9752,
+      "step": 726
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7476637291527775,
+      "learning_rate": 9.939553361522216e-06,
+      "loss": 0.9338,
+      "step": 727
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8725732835583182,
+      "learning_rate": 9.939283153558628e-06,
+      "loss": 0.9515,
+      "step": 728
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8196883956992354,
+      "learning_rate": 9.93901234668993e-06,
+      "loss": 0.9523,
+      "step": 729
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.852767824352539,
+      "learning_rate": 9.938740940948958e-06,
+      "loss": 0.8896,
+      "step": 730
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8279283348121713,
+      "learning_rate": 9.93846893636862e-06,
+      "loss": 0.9793,
+      "step": 731
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7601982567831413,
+      "learning_rate": 9.938196332981897e-06,
+      "loss": 0.9993,
+      "step": 732
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8075529151933076,
+      "learning_rate": 9.937923130821845e-06,
+      "loss": 0.9624,
+      "step": 733
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8088169140742196,
+      "learning_rate": 9.93764932992159e-06,
+      "loss": 0.8712,
+      "step": 734
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7456650635646105,
+      "learning_rate": 9.937374930314329e-06,
+      "loss": 0.8955,
+      "step": 735
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.820080867070887,
+      "learning_rate": 9.937099932033338e-06,
+      "loss": 0.972,
+      "step": 736
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8135064108896235,
+      "learning_rate": 9.93682433511196e-06,
+      "loss": 0.9319,
+      "step": 737
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8341605777113597,
+      "learning_rate": 9.93654813958361e-06,
+      "loss": 1.0143,
+      "step": 738
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.689388119074655,
+      "learning_rate": 9.93627134548178e-06,
+      "loss": 0.9801,
+      "step": 739
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0866878366430799,
+      "learning_rate": 9.935993952840033e-06,
+      "loss": 1.0823,
+      "step": 740
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8593270576815693,
+      "learning_rate": 9.935715961691999e-06,
+      "loss": 0.9909,
+      "step": 741
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8616278854311283,
+      "learning_rate": 9.93543737207139e-06,
+      "loss": 1.0867,
+      "step": 742
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8438166281667455,
+      "learning_rate": 9.935158184011986e-06,
+      "loss": 1.077,
+      "step": 743
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6302606809454292,
+      "learning_rate": 9.934878397547636e-06,
+      "loss": 0.8916,
+      "step": 744
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7674304057868386,
+      "learning_rate": 9.934598012712268e-06,
+      "loss": 0.913,
+      "step": 745
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7740834844671542,
+      "learning_rate": 9.93431702953988e-06,
+      "loss": 1.0073,
+      "step": 746
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.742712634550345,
+      "learning_rate": 9.934035448064538e-06,
+      "loss": 0.9415,
+      "step": 747
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8245864128902234,
+      "learning_rate": 9.933753268320391e-06,
+      "loss": 0.9103,
+      "step": 748
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7466188233865187,
+      "learning_rate": 9.933470490341651e-06,
+      "loss": 0.9975,
+      "step": 749
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8248312212161273,
+      "learning_rate": 9.933187114162604e-06,
+      "loss": 1.0739,
+      "step": 750
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7719863992975844,
+      "learning_rate": 9.932903139817612e-06,
+      "loss": 0.9883,
+      "step": 751
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8415070340387925,
+      "learning_rate": 9.932618567341107e-06,
+      "loss": 0.9719,
+      "step": 752
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.796157909232651,
+      "learning_rate": 9.932333396767597e-06,
+      "loss": 0.9084,
+      "step": 753
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7453697848339382,
+      "learning_rate": 9.932047628131659e-06,
+      "loss": 0.8782,
+      "step": 754
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7117725873989906,
+      "learning_rate": 9.931761261467943e-06,
+      "loss": 0.974,
+      "step": 755
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6937052868547227,
+      "learning_rate": 9.93147429681117e-06,
+      "loss": 0.9048,
+      "step": 756
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7995422616418353,
+      "learning_rate": 9.931186734196137e-06,
+      "loss": 0.9049,
+      "step": 757
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7728240453817152,
+      "learning_rate": 9.930898573657712e-06,
+      "loss": 0.9533,
+      "step": 758
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6739001112850005,
+      "learning_rate": 9.930609815230838e-06,
+      "loss": 0.9506,
+      "step": 759
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4264406399843927,
+      "learning_rate": 9.930320458950525e-06,
+      "loss": 1.0621,
+      "step": 760
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9943067906601395,
+      "learning_rate": 9.930030504851857e-06,
+      "loss": 0.9967,
+      "step": 761
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8847757964544065,
+      "learning_rate": 9.929739952969995e-06,
+      "loss": 0.9251,
+      "step": 762
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9348071031777235,
+      "learning_rate": 9.929448803340167e-06,
+      "loss": 1.0657,
+      "step": 763
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8507551867523058,
+      "learning_rate": 9.929157055997677e-06,
+      "loss": 1.0051,
+      "step": 764
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8662079598995955,
+      "learning_rate": 9.928864710977903e-06,
+      "loss": 0.9403,
+      "step": 765
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6924564742815962,
+      "learning_rate": 9.928571768316289e-06,
+      "loss": 0.8989,
+      "step": 766
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7992176984343593,
+      "learning_rate": 9.928278228048358e-06,
+      "loss": 1.0326,
+      "step": 767
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7175018853464668,
+      "learning_rate": 9.927984090209698e-06,
+      "loss": 1.0425,
+      "step": 768
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3479413320398927,
+      "learning_rate": 9.927689354835983e-06,
+      "loss": 1.0618,
+      "step": 769
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.840870084538823,
+      "learning_rate": 9.927394021962942e-06,
+      "loss": 0.9972,
+      "step": 770
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8291575510899423,
+      "learning_rate": 9.927098091626388e-06,
+      "loss": 1.0217,
+      "step": 771
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.815872423949907,
+      "learning_rate": 9.926801563862205e-06,
+      "loss": 1.0031,
+      "step": 772
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7699865405879576,
+      "learning_rate": 9.926504438706348e-06,
+      "loss": 1.0249,
+      "step": 773
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8638404384370981,
+      "learning_rate": 9.926206716194842e-06,
+      "loss": 0.9735,
+      "step": 774
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8466288413453076,
+      "learning_rate": 9.92590839636379e-06,
+      "loss": 0.8829,
+      "step": 775
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9281617018916826,
+      "learning_rate": 9.925609479249363e-06,
+      "loss": 0.9683,
+      "step": 776
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.863361796006873,
+      "learning_rate": 9.925309964887803e-06,
+      "loss": 0.9628,
+      "step": 777
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9116470414934745,
+      "learning_rate": 9.925009853315433e-06,
+      "loss": 1.036,
+      "step": 778
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7730855246377493,
+      "learning_rate": 9.924709144568635e-06,
+      "loss": 0.9992,
+      "step": 779
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9445079988777663,
+      "learning_rate": 9.924407838683878e-06,
+      "loss": 0.9462,
+      "step": 780
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7966060909369934,
+      "learning_rate": 9.924105935697694e-06,
+      "loss": 0.9882,
+      "step": 781
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7747116960351288,
+      "learning_rate": 9.923803435646688e-06,
+      "loss": 0.9429,
+      "step": 782
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7830219888159082,
+      "learning_rate": 9.923500338567542e-06,
+      "loss": 0.8295,
+      "step": 783
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7177629078829748,
+      "learning_rate": 9.923196644497004e-06,
+      "loss": 0.9394,
+      "step": 784
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8316271252569067,
+      "learning_rate": 9.9228923534719e-06,
+      "loss": 0.968,
+      "step": 785
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9414775391803776,
+      "learning_rate": 9.922587465529128e-06,
+      "loss": 1.0062,
+      "step": 786
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7061566906701089,
+      "learning_rate": 9.922281980705654e-06,
+      "loss": 0.9501,
+      "step": 787
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.029514244238478,
+      "learning_rate": 9.921975899038521e-06,
+      "loss": 0.8408,
+      "step": 788
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.671320723077056,
+      "learning_rate": 9.92166922056484e-06,
+      "loss": 0.9702,
+      "step": 789
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9060640324258404,
+      "learning_rate": 9.921361945321801e-06,
+      "loss": 0.9949,
+      "step": 790
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7566011580452403,
+      "learning_rate": 9.921054073346658e-06,
+      "loss": 0.8964,
+      "step": 791
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7578820136460285,
+      "learning_rate": 9.920745604676744e-06,
+      "loss": 0.9451,
+      "step": 792
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8078783994268042,
+      "learning_rate": 9.92043653934946e-06,
+      "loss": 0.9713,
+      "step": 793
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7780143426814108,
+      "learning_rate": 9.920126877402283e-06,
+      "loss": 0.8989,
+      "step": 794
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7502681274849756,
+      "learning_rate": 9.91981661887276e-06,
+      "loss": 0.9007,
+      "step": 795
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8431526590150353,
+      "learning_rate": 9.919505763798508e-06,
+      "loss": 1.0035,
+      "step": 796
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7281172804129425,
+      "learning_rate": 9.919194312217227e-06,
+      "loss": 0.8723,
+      "step": 797
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.750776978331979,
+      "learning_rate": 9.918882264166672e-06,
+      "loss": 0.9387,
+      "step": 798
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8585927174532484,
+      "learning_rate": 9.918569619684687e-06,
+      "loss": 0.9037,
+      "step": 799
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.794742139687047,
+      "learning_rate": 9.918256378809178e-06,
+      "loss": 0.9911,
+      "step": 800
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.69951349192749,
+      "learning_rate": 9.917942541578127e-06,
+      "loss": 0.9548,
+      "step": 801
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9640403251318759,
+      "learning_rate": 9.917628108029589e-06,
+      "loss": 1.094,
+      "step": 802
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.688444744519535,
+      "learning_rate": 9.91731307820169e-06,
+      "loss": 0.9455,
+      "step": 803
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.759844535189644,
+      "learning_rate": 9.916997452132625e-06,
+      "loss": 0.9644,
+      "step": 804
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8403238606779961,
+      "learning_rate": 9.91668122986067e-06,
+      "loss": 1.0178,
+      "step": 805
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9280707944688548,
+      "learning_rate": 9.916364411424164e-06,
+      "loss": 1.0721,
+      "step": 806
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7325376430503456,
+      "learning_rate": 9.916046996861524e-06,
+      "loss": 0.9298,
+      "step": 807
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7058257406671034,
+      "learning_rate": 9.915728986211237e-06,
+      "loss": 0.9109,
+      "step": 808
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.930390188761472,
+      "learning_rate": 9.915410379511866e-06,
+      "loss": 0.9922,
+      "step": 809
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7360505353013305,
+      "learning_rate": 9.915091176802037e-06,
+      "loss": 0.9615,
+      "step": 810
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9161336619173658,
+      "learning_rate": 9.91477137812046e-06,
+      "loss": 1.1304,
+      "step": 811
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7382963703023766,
+      "learning_rate": 9.914450983505908e-06,
+      "loss": 0.9176,
+      "step": 812
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7092472530553176,
+      "learning_rate": 9.914129992997232e-06,
+      "loss": 0.8817,
+      "step": 813
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8866513087285066,
+      "learning_rate": 9.913808406633354e-06,
+      "loss": 1.0628,
+      "step": 814
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7451650908923817,
+      "learning_rate": 9.913486224453266e-06,
+      "loss": 0.9583,
+      "step": 815
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7807043554573663,
+      "learning_rate": 9.913163446496034e-06,
+      "loss": 0.9031,
+      "step": 816
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8449517807695555,
+      "learning_rate": 9.912840072800795e-06,
+      "loss": 1.0842,
+      "step": 817
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9266464083336783,
+      "learning_rate": 9.912516103406759e-06,
+      "loss": 0.9285,
+      "step": 818
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.811910705681027,
+      "learning_rate": 9.912191538353213e-06,
+      "loss": 0.901,
+      "step": 819
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7882865572956181,
+      "learning_rate": 9.911866377679505e-06,
+      "loss": 0.912,
+      "step": 820
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7969028009982195,
+      "learning_rate": 9.911540621425067e-06,
+      "loss": 0.8945,
+      "step": 821
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0191178327763168,
+      "learning_rate": 9.911214269629394e-06,
+      "loss": 1.089,
+      "step": 822
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.741187876151006,
+      "learning_rate": 9.91088732233206e-06,
+      "loss": 0.8335,
+      "step": 823
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.732133495398089,
+      "learning_rate": 9.910559779572708e-06,
+      "loss": 0.9242,
+      "step": 824
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7343768237601058,
+      "learning_rate": 9.910231641391053e-06,
+      "loss": 0.9482,
+      "step": 825
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7418414048653905,
+      "learning_rate": 9.909902907826884e-06,
+      "loss": 0.9877,
+      "step": 826
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7349789307668178,
+      "learning_rate": 9.90957357892006e-06,
+      "loss": 0.9746,
+      "step": 827
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7533232100316827,
+      "learning_rate": 9.909243654710514e-06,
+      "loss": 0.9751,
+      "step": 828
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8124054867314066,
+      "learning_rate": 9.90891313523825e-06,
+      "loss": 0.8991,
+      "step": 829
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8526191852080973,
+      "learning_rate": 9.908582020543345e-06,
+      "loss": 0.9134,
+      "step": 830
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7946310001884849,
+      "learning_rate": 9.908250310665947e-06,
+      "loss": 0.9749,
+      "step": 831
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6836509484891176,
+      "learning_rate": 9.907918005646277e-06,
+      "loss": 0.9909,
+      "step": 832
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6965789082414533,
+      "learning_rate": 9.90758510552463e-06,
+      "loss": 0.8995,
+      "step": 833
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8365483575889425,
+      "learning_rate": 9.907251610341369e-06,
+      "loss": 0.9616,
+      "step": 834
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7993294714917092,
+      "learning_rate": 9.906917520136932e-06,
+      "loss": 1.0056,
+      "step": 835
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7558556448383609,
+      "learning_rate": 9.90658283495183e-06,
+      "loss": 0.986,
+      "step": 836
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8431305729429366,
+      "learning_rate": 9.906247554826645e-06,
+      "loss": 0.9029,
+      "step": 837
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8425183998237253,
+      "learning_rate": 9.905911679802028e-06,
+      "loss": 0.96,
+      "step": 838
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9681811428674642,
+      "learning_rate": 9.905575209918705e-06,
+      "loss": 1.0814,
+      "step": 839
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7098033686864729,
+      "learning_rate": 9.905238145217477e-06,
+      "loss": 0.9165,
+      "step": 840
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.768674657630725,
+      "learning_rate": 9.904900485739215e-06,
+      "loss": 0.9518,
+      "step": 841
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7481140936861697,
+      "learning_rate": 9.904562231524856e-06,
+      "loss": 0.8626,
+      "step": 842
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7232018000134481,
+      "learning_rate": 9.904223382615418e-06,
+      "loss": 1.032,
+      "step": 843
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.460709642359016,
+      "learning_rate": 9.90388393905199e-06,
+      "loss": 1.0068,
+      "step": 844
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7847390679904496,
+      "learning_rate": 9.903543900875727e-06,
+      "loss": 0.9547,
+      "step": 845
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8188187988716773,
+      "learning_rate": 9.903203268127861e-06,
+      "loss": 0.8809,
+      "step": 846
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8959188841722867,
+      "learning_rate": 9.902862040849694e-06,
+      "loss": 1.045,
+      "step": 847
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9304520747760463,
+      "learning_rate": 9.902520219082604e-06,
+      "loss": 0.9593,
+      "step": 848
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6853604254058532,
+      "learning_rate": 9.902177802868033e-06,
+      "loss": 0.997,
+      "step": 849
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9046856757594037,
+      "learning_rate": 9.901834792247504e-06,
+      "loss": 0.9276,
+      "step": 850
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.856163675430937,
+      "learning_rate": 9.901491187262608e-06,
+      "loss": 0.8387,
+      "step": 851
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6901082580280227,
+      "learning_rate": 9.901146987955008e-06,
+      "loss": 0.9335,
+      "step": 852
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7626665574979539,
+      "learning_rate": 9.900802194366439e-06,
+      "loss": 0.9634,
+      "step": 853
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7189162383293837,
+      "learning_rate": 9.900456806538707e-06,
+      "loss": 0.8783,
+      "step": 854
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.798664545332496,
+      "learning_rate": 9.900110824513693e-06,
+      "loss": 0.9507,
+      "step": 855
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7933570794802838,
+      "learning_rate": 9.899764248333348e-06,
+      "loss": 1.0097,
+      "step": 856
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8045860649415821,
+      "learning_rate": 9.899417078039695e-06,
+      "loss": 0.9356,
+      "step": 857
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.793830812148432,
+      "learning_rate": 9.899069313674832e-06,
+      "loss": 0.9589,
+      "step": 858
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8818990693521418,
+      "learning_rate": 9.898720955280926e-06,
+      "loss": 0.872,
+      "step": 859
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7388290430716535,
+      "learning_rate": 9.898372002900214e-06,
+      "loss": 0.9796,
+      "step": 860
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8184896287341015,
+      "learning_rate": 9.89802245657501e-06,
+      "loss": 0.938,
+      "step": 861
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.751753767791437,
+      "learning_rate": 9.897672316347697e-06,
+      "loss": 0.9233,
+      "step": 862
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.777873794624135,
+      "learning_rate": 9.897321582260731e-06,
+      "loss": 0.9132,
+      "step": 863
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.875554789561168,
+      "learning_rate": 9.896970254356638e-06,
+      "loss": 0.997,
+      "step": 864
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7159141108327025,
+      "learning_rate": 9.896618332678021e-06,
+      "loss": 0.9756,
+      "step": 865
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.802988055170041,
+      "learning_rate": 9.89626581726755e-06,
+      "loss": 0.9552,
+      "step": 866
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.788193234998761,
+      "learning_rate": 9.895912708167968e-06,
+      "loss": 0.939,
+      "step": 867
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.967937412553483,
+      "learning_rate": 9.89555900542209e-06,
+      "loss": 0.9564,
+      "step": 868
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9584590750178426,
+      "learning_rate": 9.895204709072807e-06,
+      "loss": 0.8972,
+      "step": 869
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.776914883093344,
+      "learning_rate": 9.894849819163076e-06,
+      "loss": 0.9836,
+      "step": 870
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.657850519074155,
+      "learning_rate": 9.894494335735929e-06,
+      "loss": 0.9783,
+      "step": 871
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.82980919779469,
+      "learning_rate": 9.894138258834471e-06,
+      "loss": 0.9402,
+      "step": 872
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7292430648274233,
+      "learning_rate": 9.893781588501876e-06,
+      "loss": 1.0016,
+      "step": 873
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.845714277354703,
+      "learning_rate": 9.893424324781392e-06,
+      "loss": 0.919,
+      "step": 874
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6966738622607047,
+      "learning_rate": 9.893066467716337e-06,
+      "loss": 0.9552,
+      "step": 875
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6981289358887794,
+      "learning_rate": 9.892708017350106e-06,
+      "loss": 0.9049,
+      "step": 876
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8229690190196253,
+      "learning_rate": 9.892348973726159e-06,
+      "loss": 0.954,
+      "step": 877
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7485689573753123,
+      "learning_rate": 9.891989336888033e-06,
+      "loss": 0.9521,
+      "step": 878
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6752052159666504,
+      "learning_rate": 9.891629106879333e-06,
+      "loss": 0.9024,
+      "step": 879
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7313966774914158,
+      "learning_rate": 9.891268283743742e-06,
+      "loss": 0.9875,
+      "step": 880
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.830922032405926,
+      "learning_rate": 9.890906867525007e-06,
+      "loss": 0.9455,
+      "step": 881
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8109162507724996,
+      "learning_rate": 9.890544858266955e-06,
+      "loss": 0.9871,
+      "step": 882
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7417205306759183,
+      "learning_rate": 9.890182256013477e-06,
+      "loss": 0.9317,
+      "step": 883
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7853655787558373,
+      "learning_rate": 9.889819060808541e-06,
+      "loss": 0.9226,
+      "step": 884
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9554376378711412,
+      "learning_rate": 9.889455272696187e-06,
+      "loss": 1.0656,
+      "step": 885
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8857809456398171,
+      "learning_rate": 9.889090891720526e-06,
+      "loss": 0.8675,
+      "step": 886
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6424155974287267,
+      "learning_rate": 9.888725917925736e-06,
+      "loss": 0.8298,
+      "step": 887
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7808522499120072,
+      "learning_rate": 9.888360351356077e-06,
+      "loss": 0.8964,
+      "step": 888
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7627066056269047,
+      "learning_rate": 9.887994192055872e-06,
+      "loss": 1.0054,
+      "step": 889
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.979767168464868,
+      "learning_rate": 9.887627440069519e-06,
+      "loss": 1.0152,
+      "step": 890
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.74866393019557,
+      "learning_rate": 9.887260095441489e-06,
+      "loss": 0.9841,
+      "step": 891
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7157595730196746,
+      "learning_rate": 9.886892158216324e-06,
+      "loss": 0.9529,
+      "step": 892
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7763716533564495,
+      "learning_rate": 9.886523628438635e-06,
+      "loss": 0.928,
+      "step": 893
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8212550824086182,
+      "learning_rate": 9.886154506153112e-06,
+      "loss": 0.8341,
+      "step": 894
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8343347038075255,
+      "learning_rate": 9.88578479140451e-06,
+      "loss": 0.9257,
+      "step": 895
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7051211703837204,
+      "learning_rate": 9.885414484237657e-06,
+      "loss": 0.9491,
+      "step": 896
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7976184374351234,
+      "learning_rate": 9.885043584697457e-06,
+      "loss": 0.9456,
+      "step": 897
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8060206001502357,
+      "learning_rate": 9.88467209282888e-06,
+      "loss": 1.0235,
+      "step": 898
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.81298686945845,
+      "learning_rate": 9.88430000867697e-06,
+      "loss": 0.9829,
+      "step": 899
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.970570114340302,
+      "learning_rate": 9.883927332286849e-06,
+      "loss": 1.0613,
+      "step": 900
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7854817827966023,
+      "learning_rate": 9.883554063703697e-06,
+      "loss": 0.9537,
+      "step": 901
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.927996103870089,
+      "learning_rate": 9.883180202972781e-06,
+      "loss": 1.0228,
+      "step": 902
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.856514683923491,
+      "learning_rate": 9.882805750139432e-06,
+      "loss": 0.9382,
+      "step": 903
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8808958480854923,
+      "learning_rate": 9.88243070524905e-06,
+      "loss": 1.0414,
+      "step": 904
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8125918194183057,
+      "learning_rate": 9.882055068347114e-06,
+      "loss": 0.9859,
+      "step": 905
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8565888928975418,
+      "learning_rate": 9.881678839479169e-06,
+      "loss": 0.9809,
+      "step": 906
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8011154265151827,
+      "learning_rate": 9.881302018690834e-06,
+      "loss": 0.9417,
+      "step": 907
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8422978245339494,
+      "learning_rate": 9.880924606027802e-06,
+      "loss": 0.9778,
+      "step": 908
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7574599608364259,
+      "learning_rate": 9.880546601535835e-06,
+      "loss": 0.9358,
+      "step": 909
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.821164318921587,
+      "learning_rate": 9.880168005260766e-06,
+      "loss": 1.0327,
+      "step": 910
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.81373773273064,
+      "learning_rate": 9.879788817248503e-06,
+      "loss": 0.9495,
+      "step": 911
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8325467629904657,
+      "learning_rate": 9.879409037545022e-06,
+      "loss": 0.9436,
+      "step": 912
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7707345048997194,
+      "learning_rate": 9.879028666196373e-06,
+      "loss": 0.9573,
+      "step": 913
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6085397190552688,
+      "learning_rate": 9.878647703248678e-06,
+      "loss": 0.8504,
+      "step": 914
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.662080121129869,
+      "learning_rate": 9.878266148748129e-06,
+      "loss": 0.9391,
+      "step": 915
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2466025279825363,
+      "learning_rate": 9.877884002740992e-06,
+      "loss": 1.0855,
+      "step": 916
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8343435128597174,
+      "learning_rate": 9.877501265273603e-06,
+      "loss": 0.982,
+      "step": 917
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7925314912201602,
+      "learning_rate": 9.87711793639237e-06,
+      "loss": 0.9993,
+      "step": 918
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7650708300779172,
+      "learning_rate": 9.876734016143775e-06,
+      "loss": 0.8859,
+      "step": 919
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8964994349322162,
+      "learning_rate": 9.876349504574365e-06,
+      "loss": 0.9441,
+      "step": 920
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7176745169036531,
+      "learning_rate": 9.875964401730769e-06,
+      "loss": 0.9091,
+      "step": 921
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6933410878690973,
+      "learning_rate": 9.875578707659678e-06,
+      "loss": 0.8961,
+      "step": 922
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7363253413599269,
+      "learning_rate": 9.87519242240786e-06,
+      "loss": 0.987,
+      "step": 923
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.907173957984541,
+      "learning_rate": 9.874805546022152e-06,
+      "loss": 1.0431,
+      "step": 924
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0433577696012308,
+      "learning_rate": 9.874418078549467e-06,
+      "loss": 1.0777,
+      "step": 925
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9675203321833825,
+      "learning_rate": 9.874030020036787e-06,
+      "loss": 0.9205,
+      "step": 926
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8234458795355297,
+      "learning_rate": 9.873641370531161e-06,
+      "loss": 0.9537,
+      "step": 927
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7185461654092753,
+      "learning_rate": 9.873252130079718e-06,
+      "loss": 0.9812,
+      "step": 928
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7871474028242238,
+      "learning_rate": 9.872862298729653e-06,
+      "loss": 0.9619,
+      "step": 929
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.943918554776973,
+      "learning_rate": 9.872471876528235e-06,
+      "loss": 1.0908,
+      "step": 930
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8069191798692539,
+      "learning_rate": 9.872080863522805e-06,
+      "loss": 0.921,
+      "step": 931
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.793440181508253,
+      "learning_rate": 9.871689259760772e-06,
+      "loss": 0.9064,
+      "step": 932
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.789904121655422,
+      "learning_rate": 9.871297065289623e-06,
+      "loss": 0.9087,
+      "step": 933
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8885909772120167,
+      "learning_rate": 9.87090428015691e-06,
+      "loss": 1.0759,
+      "step": 934
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7725739192521233,
+      "learning_rate": 9.870510904410262e-06,
+      "loss": 0.9812,
+      "step": 935
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7880024017925895,
+      "learning_rate": 9.870116938097374e-06,
+      "loss": 1.0041,
+      "step": 936
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7802382088925373,
+      "learning_rate": 9.869722381266016e-06,
+      "loss": 0.9234,
+      "step": 937
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.74797009086092,
+      "learning_rate": 9.869327233964032e-06,
+      "loss": 0.968,
+      "step": 938
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7845851679050015,
+      "learning_rate": 9.868931496239335e-06,
+      "loss": 0.8331,
+      "step": 939
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.786188650736567,
+      "learning_rate": 9.868535168139907e-06,
+      "loss": 0.9683,
+      "step": 940
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7498469482370298,
+      "learning_rate": 9.868138249713805e-06,
+      "loss": 0.9917,
+      "step": 941
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7550107891927942,
+      "learning_rate": 9.86774074100916e-06,
+      "loss": 0.9222,
+      "step": 942
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7734088676702036,
+      "learning_rate": 9.867342642074165e-06,
+      "loss": 0.957,
+      "step": 943
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8130629504027158,
+      "learning_rate": 9.866943952957097e-06,
+      "loss": 0.9783,
+      "step": 944
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.783181479678194,
+      "learning_rate": 9.866544673706294e-06,
+      "loss": 0.952,
+      "step": 945
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8070743492645691,
+      "learning_rate": 9.866144804370172e-06,
+      "loss": 0.957,
+      "step": 946
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7865090166401243,
+      "learning_rate": 9.865744344997217e-06,
+      "loss": 0.9805,
+      "step": 947
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.736712499937743,
+      "learning_rate": 9.865343295635986e-06,
+      "loss": 0.9228,
+      "step": 948
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7343624218746638,
+      "learning_rate": 9.864941656335105e-06,
+      "loss": 0.9255,
+      "step": 949
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6993565969738738,
+      "learning_rate": 9.864539427143277e-06,
+      "loss": 0.8653,
+      "step": 950
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7952386953609754,
+      "learning_rate": 9.864136608109274e-06,
+      "loss": 0.9186,
+      "step": 951
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7070412665095087,
+      "learning_rate": 9.863733199281938e-06,
+      "loss": 0.9977,
+      "step": 952
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7091637326615385,
+      "learning_rate": 9.863329200710183e-06,
+      "loss": 0.9348,
+      "step": 953
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7783376184079491,
+      "learning_rate": 9.862924612442996e-06,
+      "loss": 0.9954,
+      "step": 954
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9280078714553308,
+      "learning_rate": 9.862519434529435e-06,
+      "loss": 0.9597,
+      "step": 955
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8527421676289169,
+      "learning_rate": 9.862113667018628e-06,
+      "loss": 1.004,
+      "step": 956
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8654319695692816,
+      "learning_rate": 9.861707309959777e-06,
+      "loss": 1.0228,
+      "step": 957
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8068180700733758,
+      "learning_rate": 9.861300363402154e-06,
+      "loss": 0.9579,
+      "step": 958
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7765839260998324,
+      "learning_rate": 9.860892827395104e-06,
+      "loss": 0.96,
+      "step": 959
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.821942317171618,
+      "learning_rate": 9.860484701988038e-06,
+      "loss": 0.9458,
+      "step": 960
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6575431038625992,
+      "learning_rate": 9.860075987230446e-06,
+      "loss": 0.9847,
+      "step": 961
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6364947500360234,
+      "learning_rate": 9.859666683171886e-06,
+      "loss": 0.8716,
+      "step": 962
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8407579272755996,
+      "learning_rate": 9.859256789861986e-06,
+      "loss": 0.9524,
+      "step": 963
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8312335699813784,
+      "learning_rate": 9.858846307350449e-06,
+      "loss": 0.9431,
+      "step": 964
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7870624002242408,
+      "learning_rate": 9.858435235687046e-06,
+      "loss": 0.9827,
+      "step": 965
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.778812764134227,
+      "learning_rate": 9.85802357492162e-06,
+      "loss": 0.9986,
+      "step": 966
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8624679934326214,
+      "learning_rate": 9.857611325104088e-06,
+      "loss": 0.9496,
+      "step": 967
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0749965364838965,
+      "learning_rate": 9.857198486284436e-06,
+      "loss": 0.9997,
+      "step": 968
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.962839183190603,
+      "learning_rate": 9.856785058512722e-06,
+      "loss": 1.088,
+      "step": 969
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.826861056262288,
+      "learning_rate": 9.856371041839077e-06,
+      "loss": 0.9124,
+      "step": 970
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7998874824167328,
+      "learning_rate": 9.8559564363137e-06,
+      "loss": 1.0311,
+      "step": 971
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8646915243817053,
+      "learning_rate": 9.855541241986864e-06,
+      "loss": 0.9469,
+      "step": 972
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7261471930948713,
+      "learning_rate": 9.855125458908912e-06,
+      "loss": 0.8794,
+      "step": 973
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8115072836574875,
+      "learning_rate": 9.854709087130261e-06,
+      "loss": 0.9968,
+      "step": 974
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7395941661014613,
+      "learning_rate": 9.854292126701397e-06,
+      "loss": 0.9884,
+      "step": 975
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6195324062942877,
+      "learning_rate": 9.853874577672875e-06,
+      "loss": 0.9168,
+      "step": 976
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9336280077133761,
+      "learning_rate": 9.853456440095329e-06,
+      "loss": 1.0776,
+      "step": 977
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6924047582395587,
+      "learning_rate": 9.853037714019455e-06,
+      "loss": 0.9587,
+      "step": 978
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7302928999595852,
+      "learning_rate": 9.85261839949603e-06,
+      "loss": 0.9836,
+      "step": 979
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8250592146880602,
+      "learning_rate": 9.852198496575893e-06,
+      "loss": 0.969,
+      "step": 980
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5987937940221362,
+      "learning_rate": 9.85177800530996e-06,
+      "loss": 0.93,
+      "step": 981
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8164032112643722,
+      "learning_rate": 9.851356925749218e-06,
+      "loss": 1.0004,
+      "step": 982
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8784685537036336,
+      "learning_rate": 9.850935257944722e-06,
+      "loss": 0.8654,
+      "step": 983
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7868661690290244,
+      "learning_rate": 9.850513001947604e-06,
+      "loss": 0.938,
+      "step": 984
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.756513101119036,
+      "learning_rate": 9.850090157809062e-06,
+      "loss": 1.0043,
+      "step": 985
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8329671738514197,
+      "learning_rate": 9.849666725580366e-06,
+      "loss": 0.935,
+      "step": 986
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7801831076219174,
+      "learning_rate": 9.849242705312863e-06,
+      "loss": 0.9776,
+      "step": 987
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8129292290267487,
+      "learning_rate": 9.848818097057962e-06,
+      "loss": 0.9517,
+      "step": 988
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7482559362234948,
+      "learning_rate": 9.848392900867151e-06,
+      "loss": 1.0212,
+      "step": 989
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9672338540711949,
+      "learning_rate": 9.847967116791985e-06,
+      "loss": 0.9408,
+      "step": 990
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7897730064924482,
+      "learning_rate": 9.847540744884094e-06,
+      "loss": 1.0049,
+      "step": 991
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6817636084648129,
+      "learning_rate": 9.847113785195175e-06,
+      "loss": 0.9242,
+      "step": 992
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8249791233016752,
+      "learning_rate": 9.846686237776999e-06,
+      "loss": 1.006,
+      "step": 993
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9357126804069176,
+      "learning_rate": 9.846258102681407e-06,
+      "loss": 0.881,
+      "step": 994
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8692610867033181,
+      "learning_rate": 9.845829379960312e-06,
+      "loss": 0.8461,
+      "step": 995
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8470217278899053,
+      "learning_rate": 9.845400069665701e-06,
+      "loss": 0.9798,
+      "step": 996
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0139386725154416,
+      "learning_rate": 9.844970171849625e-06,
+      "loss": 1.0896,
+      "step": 997
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1602670435074516,
+      "learning_rate": 9.844539686564213e-06,
+      "loss": 1.0017,
+      "step": 998
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8588536576422883,
+      "learning_rate": 9.844108613861663e-06,
+      "loss": 1.0023,
+      "step": 999
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.700574006958072,
+      "learning_rate": 9.843676953794242e-06,
+      "loss": 0.8836,
+      "step": 1000
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7181997383951344,
+      "learning_rate": 9.843244706414292e-06,
+      "loss": 0.9453,
+      "step": 1001
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8298449700677073,
+      "learning_rate": 9.842811871774224e-06,
+      "loss": 0.9068,
+      "step": 1002
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7618353343782998,
+      "learning_rate": 9.842378449926521e-06,
+      "loss": 0.9387,
+      "step": 1003
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8723621462859346,
+      "learning_rate": 9.841944440923738e-06,
+      "loss": 0.9411,
+      "step": 1004
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7180383125276784,
+      "learning_rate": 9.841509844818498e-06,
+      "loss": 0.9533,
+      "step": 1005
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7886269511756854,
+      "learning_rate": 9.841074661663497e-06,
+      "loss": 0.9312,
+      "step": 1006
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.218821639247775,
+      "learning_rate": 9.840638891511505e-06,
+      "loss": 1.0581,
+      "step": 1007
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7451178320447245,
+      "learning_rate": 9.840202534415358e-06,
+      "loss": 0.9473,
+      "step": 1008
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8057909459822308,
+      "learning_rate": 9.839765590427968e-06,
+      "loss": 0.9667,
+      "step": 1009
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7384824199482714,
+      "learning_rate": 9.839328059602316e-06,
+      "loss": 0.9005,
+      "step": 1010
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6484595911957376,
+      "learning_rate": 9.838889941991451e-06,
+      "loss": 0.9097,
+      "step": 1011
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7415414223270598,
+      "learning_rate": 9.838451237648499e-06,
+      "loss": 0.9625,
+      "step": 1012
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8318880362447547,
+      "learning_rate": 9.838011946626654e-06,
+      "loss": 0.9928,
+      "step": 1013
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0768110764943803,
+      "learning_rate": 9.83757206897918e-06,
+      "loss": 1.051,
+      "step": 1014
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.931089254987397,
+      "learning_rate": 9.837131604759416e-06,
+      "loss": 0.9729,
+      "step": 1015
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7960000185747274,
+      "learning_rate": 9.836690554020769e-06,
+      "loss": 0.9264,
+      "step": 1016
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.762389443245749,
+      "learning_rate": 9.836248916816718e-06,
+      "loss": 0.9156,
+      "step": 1017
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.752816001636378,
+      "learning_rate": 9.835806693200812e-06,
+      "loss": 0.9202,
+      "step": 1018
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.898639377393412,
+      "learning_rate": 9.835363883226675e-06,
+      "loss": 0.9206,
+      "step": 1019
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.669329753830521,
+      "learning_rate": 9.834920486947994e-06,
+      "loss": 0.9295,
+      "step": 1020
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.750657614307251,
+      "learning_rate": 9.834476504418537e-06,
+      "loss": 0.9744,
+      "step": 1021
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.928201607176009,
+      "learning_rate": 9.834031935692136e-06,
+      "loss": 0.8937,
+      "step": 1022
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7317418931000343,
+      "learning_rate": 9.833586780822698e-06,
+      "loss": 0.9318,
+      "step": 1023
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8596109093142383,
+      "learning_rate": 9.8331410398642e-06,
+      "loss": 0.9689,
+      "step": 1024
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7974015634012337,
+      "learning_rate": 9.832694712870687e-06,
+      "loss": 0.9388,
+      "step": 1025
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7786555667025423,
+      "learning_rate": 9.83224779989628e-06,
+      "loss": 0.9819,
+      "step": 1026
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7524301458796527,
+      "learning_rate": 9.831800300995167e-06,
+      "loss": 1.0324,
+      "step": 1027
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6551019641741704,
+      "learning_rate": 9.83135221622161e-06,
+      "loss": 1.0417,
+      "step": 1028
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.694109910291689,
+      "learning_rate": 9.830903545629942e-06,
+      "loss": 0.9071,
+      "step": 1029
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7588396521708929,
+      "learning_rate": 9.830454289274563e-06,
+      "loss": 0.8541,
+      "step": 1030
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7291720210703856,
+      "learning_rate": 9.830004447209949e-06,
+      "loss": 0.968,
+      "step": 1031
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.700967122701236,
+      "learning_rate": 9.829554019490644e-06,
+      "loss": 0.9379,
+      "step": 1032
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.634810475927666,
+      "learning_rate": 9.829103006171264e-06,
+      "loss": 0.9892,
+      "step": 1033
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6760866654998925,
+      "learning_rate": 9.828651407306495e-06,
+      "loss": 0.9784,
+      "step": 1034
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6740977614798291,
+      "learning_rate": 9.828199222951099e-06,
+      "loss": 0.9834,
+      "step": 1035
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6694694417771454,
+      "learning_rate": 9.8277464531599e-06,
+      "loss": 0.9068,
+      "step": 1036
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.161989127970139,
+      "learning_rate": 9.8272930979878e-06,
+      "loss": 1.0748,
+      "step": 1037
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9743562979785556,
+      "learning_rate": 9.826839157489768e-06,
+      "loss": 0.9554,
+      "step": 1038
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8637476009368845,
+      "learning_rate": 9.826384631720849e-06,
+      "loss": 0.947,
+      "step": 1039
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.810381557043378,
+      "learning_rate": 9.825929520736155e-06,
+      "loss": 0.9726,
+      "step": 1040
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8172722380972584,
+      "learning_rate": 9.825473824590867e-06,
+      "loss": 0.9383,
+      "step": 1041
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8940697338345922,
+      "learning_rate": 9.825017543340244e-06,
+      "loss": 0.9698,
+      "step": 1042
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.812893484359614,
+      "learning_rate": 9.82456067703961e-06,
+      "loss": 0.9293,
+      "step": 1043
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.769184440301945,
+      "learning_rate": 9.82410322574436e-06,
+      "loss": 0.9881,
+      "step": 1044
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7476337618368556,
+      "learning_rate": 9.823645189509961e-06,
+      "loss": 0.9206,
+      "step": 1045
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.712289227917583,
+      "learning_rate": 9.823186568391957e-06,
+      "loss": 0.8945,
+      "step": 1046
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8362004300488781,
+      "learning_rate": 9.82272736244595e-06,
+      "loss": 0.9478,
+      "step": 1047
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7044043586888376,
+      "learning_rate": 9.822267571727624e-06,
+      "loss": 0.8737,
+      "step": 1048
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7451132865069126,
+      "learning_rate": 9.821807196292732e-06,
+      "loss": 0.9271,
+      "step": 1049
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1777265584003949,
+      "learning_rate": 9.821346236197092e-06,
+      "loss": 1.057,
+      "step": 1050
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7969767375332188,
+      "learning_rate": 9.820884691496601e-06,
+      "loss": 0.9736,
+      "step": 1051
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7379516077333648,
+      "learning_rate": 9.82042256224722e-06,
+      "loss": 0.9556,
+      "step": 1052
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.643244233541863,
+      "learning_rate": 9.819959848504986e-06,
+      "loss": 0.9117,
+      "step": 1053
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8117506157191972,
+      "learning_rate": 9.819496550326002e-06,
+      "loss": 0.9194,
+      "step": 1054
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8572414590857225,
+      "learning_rate": 9.819032667766447e-06,
+      "loss": 0.8787,
+      "step": 1055
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7268845782212832,
+      "learning_rate": 9.818568200882566e-06,
+      "loss": 0.9105,
+      "step": 1056
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7499943828483917,
+      "learning_rate": 9.81810314973068e-06,
+      "loss": 0.9514,
+      "step": 1057
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1539247255425136,
+      "learning_rate": 9.817637514367175e-06,
+      "loss": 1.0687,
+      "step": 1058
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0071822664097232,
+      "learning_rate": 9.817171294848514e-06,
+      "loss": 1.0907,
+      "step": 1059
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7064070197943426,
+      "learning_rate": 9.816704491231226e-06,
+      "loss": 0.9668,
+      "step": 1060
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0232510412023972,
+      "learning_rate": 9.816237103571913e-06,
+      "loss": 1.1092,
+      "step": 1061
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9001613872713028,
+      "learning_rate": 9.815769131927247e-06,
+      "loss": 0.9514,
+      "step": 1062
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8633149316002269,
+      "learning_rate": 9.81530057635397e-06,
+      "loss": 0.8809,
+      "step": 1063
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.670818616827823,
+      "learning_rate": 9.814831436908897e-06,
+      "loss": 0.9813,
+      "step": 1064
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6838320261347404,
+      "learning_rate": 9.814361713648914e-06,
+      "loss": 0.9507,
+      "step": 1065
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7167523344969506,
+      "learning_rate": 9.813891406630975e-06,
+      "loss": 0.9961,
+      "step": 1066
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8456744000164487,
+      "learning_rate": 9.81342051591211e-06,
+      "loss": 0.9546,
+      "step": 1067
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9280740862459023,
+      "learning_rate": 9.812949041549408e-06,
+      "loss": 0.9787,
+      "step": 1068
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.84611676794399,
+      "learning_rate": 9.812476983600046e-06,
+      "loss": 0.9256,
+      "step": 1069
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.800919827128619,
+      "learning_rate": 9.812004342121257e-06,
+      "loss": 0.9749,
+      "step": 1070
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8455604703422732,
+      "learning_rate": 9.811531117170353e-06,
+      "loss": 1.0197,
+      "step": 1071
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.685741152848005,
+      "learning_rate": 9.811057308804711e-06,
+      "loss": 1.023,
+      "step": 1072
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.710455384377104,
+      "learning_rate": 9.810582917081786e-06,
+      "loss": 1.0491,
+      "step": 1073
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7600125017366672,
+      "learning_rate": 9.810107942059097e-06,
+      "loss": 0.972,
+      "step": 1074
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6024080768493354,
+      "learning_rate": 9.809632383794236e-06,
+      "loss": 0.9531,
+      "step": 1075
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8615663794713686,
+      "learning_rate": 9.809156242344868e-06,
+      "loss": 1.0343,
+      "step": 1076
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7532352719641884,
+      "learning_rate": 9.808679517768727e-06,
+      "loss": 0.9091,
+      "step": 1077
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.77917124210932,
+      "learning_rate": 9.808202210123617e-06,
+      "loss": 0.9827,
+      "step": 1078
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7836323310069275,
+      "learning_rate": 9.80772431946741e-06,
+      "loss": 0.8918,
+      "step": 1079
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7774652821414163,
+      "learning_rate": 9.807245845858056e-06,
+      "loss": 0.9467,
+      "step": 1080
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0184982058455514,
+      "learning_rate": 9.80676678935357e-06,
+      "loss": 0.967,
+      "step": 1081
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7003372981375924,
+      "learning_rate": 9.80628715001204e-06,
+      "loss": 0.9151,
+      "step": 1082
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8329061591491158,
+      "learning_rate": 9.805806927891622e-06,
+      "loss": 0.9438,
+      "step": 1083
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.774727187159285,
+      "learning_rate": 9.805326123050547e-06,
+      "loss": 0.9494,
+      "step": 1084
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8568606668103063,
+      "learning_rate": 9.804844735547113e-06,
+      "loss": 0.9996,
+      "step": 1085
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2791177485487049,
+      "learning_rate": 9.804362765439688e-06,
+      "loss": 1.0832,
+      "step": 1086
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8032529382948508,
+      "learning_rate": 9.803880212786716e-06,
+      "loss": 0.9669,
+      "step": 1087
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7127224239865193,
+      "learning_rate": 9.803397077646706e-06,
+      "loss": 0.9389,
+      "step": 1088
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8149416370023337,
+      "learning_rate": 9.80291336007824e-06,
+      "loss": 0.8991,
+      "step": 1089
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.760311639847556,
+      "learning_rate": 9.802429060139973e-06,
+      "loss": 0.9495,
+      "step": 1090
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8429835200172284,
+      "learning_rate": 9.801944177890626e-06,
+      "loss": 0.9852,
+      "step": 1091
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8350620456825846,
+      "learning_rate": 9.801458713388991e-06,
+      "loss": 0.9803,
+      "step": 1092
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.882108257764037,
+      "learning_rate": 9.800972666693936e-06,
+      "loss": 0.9215,
+      "step": 1093
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9755543710455752,
+      "learning_rate": 9.800486037864393e-06,
+      "loss": 0.9311,
+      "step": 1094
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1432965521731466,
+      "learning_rate": 9.799998826959366e-06,
+      "loss": 1.0493,
+      "step": 1095
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8840183059162294,
+      "learning_rate": 9.799511034037934e-06,
+      "loss": 0.957,
+      "step": 1096
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7858058105779706,
+      "learning_rate": 9.799022659159243e-06,
+      "loss": 0.905,
+      "step": 1097
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9409426172326323,
+      "learning_rate": 9.798533702382512e-06,
+      "loss": 1.0083,
+      "step": 1098
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8394467919465047,
+      "learning_rate": 9.798044163767022e-06,
+      "loss": 0.9388,
+      "step": 1099
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9554723120353845,
+      "learning_rate": 9.797554043372139e-06,
+      "loss": 0.9869,
+      "step": 1100
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.741561855656167,
+      "learning_rate": 9.79706334125729e-06,
+      "loss": 0.9568,
+      "step": 1101
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9362479774256827,
+      "learning_rate": 9.796572057481969e-06,
+      "loss": 1.0453,
+      "step": 1102
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7547957918734698,
+      "learning_rate": 9.79608019210575e-06,
+      "loss": 1.0017,
+      "step": 1103
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9790835346110744,
+      "learning_rate": 9.795587745188277e-06,
+      "loss": 0.9622,
+      "step": 1104
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7746915491668618,
+      "learning_rate": 9.795094716789252e-06,
+      "loss": 0.9815,
+      "step": 1105
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7750417008918324,
+      "learning_rate": 9.794601106968466e-06,
+      "loss": 0.9146,
+      "step": 1106
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.831413815780519,
+      "learning_rate": 9.794106915785764e-06,
+      "loss": 0.9479,
+      "step": 1107
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6643258761149287,
+      "learning_rate": 9.79361214330107e-06,
+      "loss": 0.9926,
+      "step": 1108
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8919240235227575,
+      "learning_rate": 9.793116789574379e-06,
+      "loss": 1.0029,
+      "step": 1109
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7438980242862387,
+      "learning_rate": 9.792620854665752e-06,
+      "loss": 0.883,
+      "step": 1110
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8180112242115445,
+      "learning_rate": 9.792124338635325e-06,
+      "loss": 0.993,
+      "step": 1111
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.816636483697899,
+      "learning_rate": 9.7916272415433e-06,
+      "loss": 1.0401,
+      "step": 1112
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7509090735478468,
+      "learning_rate": 9.791129563449953e-06,
+      "loss": 0.9632,
+      "step": 1113
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8267204320972894,
+      "learning_rate": 9.790631304415627e-06,
+      "loss": 0.9639,
+      "step": 1114
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7711127633528847,
+      "learning_rate": 9.790132464500741e-06,
+      "loss": 1.001,
+      "step": 1115
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1077124772900868,
+      "learning_rate": 9.789633043765781e-06,
+      "loss": 1.0949,
+      "step": 1116
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9305025462701941,
+      "learning_rate": 9.7891330422713e-06,
+      "loss": 1.0779,
+      "step": 1117
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7243184975084127,
+      "learning_rate": 9.788632460077928e-06,
+      "loss": 0.9928,
+      "step": 1118
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7037677988003548,
+      "learning_rate": 9.788131297246362e-06,
+      "loss": 0.972,
+      "step": 1119
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7798806860339067,
+      "learning_rate": 9.787629553837368e-06,
+      "loss": 0.947,
+      "step": 1120
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6582626056215375,
+      "learning_rate": 9.787127229911783e-06,
+      "loss": 0.8571,
+      "step": 1121
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.033400470010361,
+      "learning_rate": 9.786624325530521e-06,
+      "loss": 0.9424,
+      "step": 1122
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.839132253265899,
+      "learning_rate": 9.786120840754557e-06,
+      "loss": 0.8955,
+      "step": 1123
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3442596771492645,
+      "learning_rate": 9.78561677564494e-06,
+      "loss": 1.0709,
+      "step": 1124
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7035337853042138,
+      "learning_rate": 9.78511213026279e-06,
+      "loss": 0.9714,
+      "step": 1125
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.932330612454489,
+      "learning_rate": 9.784606904669298e-06,
+      "loss": 0.9643,
+      "step": 1126
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8038822373487082,
+      "learning_rate": 9.784101098925723e-06,
+      "loss": 1.0149,
+      "step": 1127
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6897023886943041,
+      "learning_rate": 9.783594713093399e-06,
+      "loss": 0.9779,
+      "step": 1128
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8829041784730604,
+      "learning_rate": 9.783087747233722e-06,
+      "loss": 1.0874,
+      "step": 1129
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8730841857749171,
+      "learning_rate": 9.782580201408165e-06,
+      "loss": 1.0855,
+      "step": 1130
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8725353698414466,
+      "learning_rate": 9.782072075678273e-06,
+      "loss": 0.8449,
+      "step": 1131
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0773312155579124,
+      "learning_rate": 9.781563370105654e-06,
+      "loss": 0.9481,
+      "step": 1132
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9102606387047252,
+      "learning_rate": 9.781054084751993e-06,
+      "loss": 0.8468,
+      "step": 1133
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7930604081479018,
+      "learning_rate": 9.78054421967904e-06,
+      "loss": 0.9337,
+      "step": 1134
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8473971319906888,
+      "learning_rate": 9.78003377494862e-06,
+      "loss": 0.9305,
+      "step": 1135
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.85176056028402,
+      "learning_rate": 9.779522750622626e-06,
+      "loss": 0.9797,
+      "step": 1136
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7397960692669328,
+      "learning_rate": 9.77901114676302e-06,
+      "loss": 0.9492,
+      "step": 1137
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.73717477710704,
+      "learning_rate": 9.778498963431838e-06,
+      "loss": 0.8659,
+      "step": 1138
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7682856161597194,
+      "learning_rate": 9.77798620069118e-06,
+      "loss": 0.9341,
+      "step": 1139
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8309610001723984,
+      "learning_rate": 9.777472858603227e-06,
+      "loss": 0.9559,
+      "step": 1140
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7363188810192791,
+      "learning_rate": 9.776958937230216e-06,
+      "loss": 0.8937,
+      "step": 1141
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.871640621744642,
+      "learning_rate": 9.776444436634467e-06,
+      "loss": 0.86,
+      "step": 1142
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7276594801751906,
+      "learning_rate": 9.775929356878362e-06,
+      "loss": 0.9755,
+      "step": 1143
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7628302899777013,
+      "learning_rate": 9.775413698024358e-06,
+      "loss": 0.9392,
+      "step": 1144
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.039096521583918,
+      "learning_rate": 9.77489746013498e-06,
+      "loss": 1.0588,
+      "step": 1145
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0926795984818383,
+      "learning_rate": 9.774380643272824e-06,
+      "loss": 1.0825,
+      "step": 1146
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6992906775863967,
+      "learning_rate": 9.773863247500554e-06,
+      "loss": 0.9219,
+      "step": 1147
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8617765892884641,
+      "learning_rate": 9.773345272880908e-06,
+      "loss": 0.954,
+      "step": 1148
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.829431328990108,
+      "learning_rate": 9.77282671947669e-06,
+      "loss": 0.935,
+      "step": 1149
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8932176065547315,
+      "learning_rate": 9.77230758735078e-06,
+      "loss": 1.1008,
+      "step": 1150
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7197101910280772,
+      "learning_rate": 9.771787876566121e-06,
+      "loss": 0.8337,
+      "step": 1151
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6810711385239918,
+      "learning_rate": 9.771267587185732e-06,
+      "loss": 0.9056,
+      "step": 1152
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7277927675279006,
+      "learning_rate": 9.770746719272696e-06,
+      "loss": 0.9002,
+      "step": 1153
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7259773786406831,
+      "learning_rate": 9.770225272890177e-06,
+      "loss": 0.9414,
+      "step": 1154
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7368740244304055,
+      "learning_rate": 9.769703248101397e-06,
+      "loss": 0.9126,
+      "step": 1155
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8125964343318284,
+      "learning_rate": 9.769180644969653e-06,
+      "loss": 0.9877,
+      "step": 1156
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.129001504975424,
+      "learning_rate": 9.768657463558316e-06,
+      "loss": 1.0956,
+      "step": 1157
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6693471625847693,
+      "learning_rate": 9.76813370393082e-06,
+      "loss": 0.9593,
+      "step": 1158
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.874832684996006,
+      "learning_rate": 9.767609366150675e-06,
+      "loss": 0.8864,
+      "step": 1159
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7556508292751147,
+      "learning_rate": 9.767084450281457e-06,
+      "loss": 1.0714,
+      "step": 1160
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8851490659830241,
+      "learning_rate": 9.766558956386814e-06,
+      "loss": 0.9562,
+      "step": 1161
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.724498562009284,
+      "learning_rate": 9.766032884530466e-06,
+      "loss": 0.9199,
+      "step": 1162
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.779881992187159,
+      "learning_rate": 9.765506234776201e-06,
+      "loss": 0.9725,
+      "step": 1163
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7560831366724476,
+      "learning_rate": 9.764979007187874e-06,
+      "loss": 0.9687,
+      "step": 1164
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8925677576649702,
+      "learning_rate": 9.764451201829415e-06,
+      "loss": 0.9948,
+      "step": 1165
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.735458900323121,
+      "learning_rate": 9.763922818764824e-06,
+      "loss": 0.9654,
+      "step": 1166
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6651433238646258,
+      "learning_rate": 9.763393858058165e-06,
+      "loss": 0.9009,
+      "step": 1167
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8590380913076896,
+      "learning_rate": 9.76286431977358e-06,
+      "loss": 0.882,
+      "step": 1168
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1555532803839814,
+      "learning_rate": 9.762334203975277e-06,
+      "loss": 1.055,
+      "step": 1169
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7482108525015174,
+      "learning_rate": 9.761803510727532e-06,
+      "loss": 0.9774,
+      "step": 1170
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.836801299609463,
+      "learning_rate": 9.761272240094695e-06,
+      "loss": 1.0109,
+      "step": 1171
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7148262171388866,
+      "learning_rate": 9.760740392141186e-06,
+      "loss": 0.8835,
+      "step": 1172
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8583512028818457,
+      "learning_rate": 9.76020796693149e-06,
+      "loss": 1.0661,
+      "step": 1173
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.771380201318604,
+      "learning_rate": 9.759674964530167e-06,
+      "loss": 0.9806,
+      "step": 1174
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7048611706393197,
+      "learning_rate": 9.759141385001846e-06,
+      "loss": 0.9796,
+      "step": 1175
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.666934776115197,
+      "learning_rate": 9.758607228411226e-06,
+      "loss": 0.9507,
+      "step": 1176
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8579198394441183,
+      "learning_rate": 9.758072494823072e-06,
+      "loss": 0.8535,
+      "step": 1177
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.731610721355881,
+      "learning_rate": 9.757537184302225e-06,
+      "loss": 0.8855,
+      "step": 1178
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8567773312664542,
+      "learning_rate": 9.757001296913594e-06,
+      "loss": 0.9325,
+      "step": 1179
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7115101500042478,
+      "learning_rate": 9.756464832722155e-06,
+      "loss": 0.9759,
+      "step": 1180
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7928370572520054,
+      "learning_rate": 9.755927791792956e-06,
+      "loss": 0.9614,
+      "step": 1181
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8802490173635868,
+      "learning_rate": 9.755390174191117e-06,
+      "loss": 0.9717,
+      "step": 1182
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6889013721166843,
+      "learning_rate": 9.754851979981827e-06,
+      "loss": 0.9714,
+      "step": 1183
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7813643126599725,
+      "learning_rate": 9.75431320923034e-06,
+      "loss": 0.9536,
+      "step": 1184
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.76605807977974,
+      "learning_rate": 9.753773862001985e-06,
+      "loss": 0.9965,
+      "step": 1185
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5913301669398983,
+      "learning_rate": 9.753233938362161e-06,
+      "loss": 0.8939,
+      "step": 1186
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6881318965280296,
+      "learning_rate": 9.752693438376337e-06,
+      "loss": 0.9475,
+      "step": 1187
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0947588332008547,
+      "learning_rate": 9.752152362110046e-06,
+      "loss": 1.0438,
+      "step": 1188
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9686766294967777,
+      "learning_rate": 9.751610709628898e-06,
+      "loss": 1.0701,
+      "step": 1189
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8673464143885388,
+      "learning_rate": 9.751068480998572e-06,
+      "loss": 0.987,
+      "step": 1190
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7844467966453497,
+      "learning_rate": 9.750525676284813e-06,
+      "loss": 0.9515,
+      "step": 1191
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7696494984768467,
+      "learning_rate": 9.749982295553437e-06,
+      "loss": 0.9391,
+      "step": 1192
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8158550291199695,
+      "learning_rate": 9.749438338870332e-06,
+      "loss": 1.0003,
+      "step": 1193
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7976566013889366,
+      "learning_rate": 9.748893806301455e-06,
+      "loss": 0.9392,
+      "step": 1194
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9332542237192982,
+      "learning_rate": 9.748348697912833e-06,
+      "loss": 0.9493,
+      "step": 1195
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.604440152543112,
+      "learning_rate": 9.74780301377056e-06,
+      "loss": 1.0833,
+      "step": 1196
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8255608816105517,
+      "learning_rate": 9.747256753940803e-06,
+      "loss": 0.9578,
+      "step": 1197
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7809087848827114,
+      "learning_rate": 9.746709918489801e-06,
+      "loss": 0.8719,
+      "step": 1198
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.65431478005747,
+      "learning_rate": 9.746162507483854e-06,
+      "loss": 0.9051,
+      "step": 1199
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9627018848274116,
+      "learning_rate": 9.745614520989341e-06,
+      "loss": 0.9859,
+      "step": 1200
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.913291515226628,
+      "learning_rate": 9.745065959072708e-06,
+      "loss": 0.9065,
+      "step": 1201
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8686220949555998,
+      "learning_rate": 9.74451682180047e-06,
+      "loss": 0.9349,
+      "step": 1202
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9723400638928938,
+      "learning_rate": 9.743967109239207e-06,
+      "loss": 0.8769,
+      "step": 1203
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8034736380117218,
+      "learning_rate": 9.743416821455577e-06,
+      "loss": 0.9652,
+      "step": 1204
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.81550973541998,
+      "learning_rate": 9.742865958516307e-06,
+      "loss": 0.9602,
+      "step": 1205
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7542400887946263,
+      "learning_rate": 9.742314520488188e-06,
+      "loss": 0.9109,
+      "step": 1206
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7646559733238234,
+      "learning_rate": 9.741762507438084e-06,
+      "loss": 0.8948,
+      "step": 1207
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8779758906380966,
+      "learning_rate": 9.741209919432928e-06,
+      "loss": 0.9726,
+      "step": 1208
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8320607531141995,
+      "learning_rate": 9.740656756539724e-06,
+      "loss": 0.9371,
+      "step": 1209
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8765413675366769,
+      "learning_rate": 9.740103018825544e-06,
+      "loss": 0.9057,
+      "step": 1210
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9908868725287112,
+      "learning_rate": 9.739548706357533e-06,
+      "loss": 0.904,
+      "step": 1211
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6610421176473633,
+      "learning_rate": 9.7389938192029e-06,
+      "loss": 0.9286,
+      "step": 1212
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9231587102047503,
+      "learning_rate": 9.73843835742893e-06,
+      "loss": 0.9101,
+      "step": 1213
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8222129549290134,
+      "learning_rate": 9.737882321102973e-06,
+      "loss": 1.0269,
+      "step": 1214
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8299883463824855,
+      "learning_rate": 9.73732571029245e-06,
+      "loss": 0.9079,
+      "step": 1215
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.712060848562136,
+      "learning_rate": 9.736768525064852e-06,
+      "loss": 0.8776,
+      "step": 1216
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8585795908956828,
+      "learning_rate": 9.736210765487741e-06,
+      "loss": 0.9741,
+      "step": 1217
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.841056951620945,
+      "learning_rate": 9.735652431628748e-06,
+      "loss": 0.9681,
+      "step": 1218
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8485960722266441,
+      "learning_rate": 9.73509352355557e-06,
+      "loss": 0.8837,
+      "step": 1219
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9668234966147329,
+      "learning_rate": 9.734534041335976e-06,
+      "loss": 0.9026,
+      "step": 1220
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7998547441960084,
+      "learning_rate": 9.73397398503781e-06,
+      "loss": 0.9372,
+      "step": 1221
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7652738569703286,
+      "learning_rate": 9.733413354728978e-06,
+      "loss": 0.8862,
+      "step": 1222
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8306394523063005,
+      "learning_rate": 9.732852150477456e-06,
+      "loss": 0.9897,
+      "step": 1223
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7943897000149613,
+      "learning_rate": 9.732290372351295e-06,
+      "loss": 0.8629,
+      "step": 1224
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7186497667282286,
+      "learning_rate": 9.731728020418612e-06,
+      "loss": 0.8658,
+      "step": 1225
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8285182177430324,
+      "learning_rate": 9.731165094747593e-06,
+      "loss": 0.8435,
+      "step": 1226
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.763297876847239,
+      "learning_rate": 9.730601595406497e-06,
+      "loss": 0.9361,
+      "step": 1227
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8446768054265583,
+      "learning_rate": 9.730037522463647e-06,
+      "loss": 1.0085,
+      "step": 1228
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8919520221084818,
+      "learning_rate": 9.72947287598744e-06,
+      "loss": 0.9534,
+      "step": 1229
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.812571374767467,
+      "learning_rate": 9.728907656046344e-06,
+      "loss": 0.9691,
+      "step": 1230
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8595529542956106,
+      "learning_rate": 9.72834186270889e-06,
+      "loss": 1.0088,
+      "step": 1231
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8890378025150192,
+      "learning_rate": 9.727775496043685e-06,
+      "loss": 0.9857,
+      "step": 1232
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8604894870630209,
+      "learning_rate": 9.727208556119401e-06,
+      "loss": 1.0162,
+      "step": 1233
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9827885363499883,
+      "learning_rate": 9.726641043004782e-06,
+      "loss": 1.074,
+      "step": 1234
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7994014319664224,
+      "learning_rate": 9.726072956768644e-06,
+      "loss": 0.9154,
+      "step": 1235
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0611412327825405,
+      "learning_rate": 9.725504297479865e-06,
+      "loss": 0.9611,
+      "step": 1236
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7433485532841393,
+      "learning_rate": 9.7249350652074e-06,
+      "loss": 1.0075,
+      "step": 1237
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7258647403929175,
+      "learning_rate": 9.724365260020267e-06,
+      "loss": 0.9223,
+      "step": 1238
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.777195581948761,
+      "learning_rate": 9.723794881987561e-06,
+      "loss": 0.9695,
+      "step": 1239
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7538343864266914,
+      "learning_rate": 9.723223931178439e-06,
+      "loss": 0.9321,
+      "step": 1240
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.740926222844274,
+      "learning_rate": 9.722652407662129e-06,
+      "loss": 0.9413,
+      "step": 1241
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9250698917754199,
+      "learning_rate": 9.722080311507938e-06,
+      "loss": 0.9888,
+      "step": 1242
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8705263775410779,
+      "learning_rate": 9.721507642785225e-06,
+      "loss": 0.896,
+      "step": 1243
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.71888811367962,
+      "learning_rate": 9.720934401563438e-06,
+      "loss": 1.0403,
+      "step": 1244
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6833230077152903,
+      "learning_rate": 9.720360587912075e-06,
+      "loss": 1.0131,
+      "step": 1245
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6538990010705523,
+      "learning_rate": 9.71978620190072e-06,
+      "loss": 0.9447,
+      "step": 1246
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7347180559872641,
+      "learning_rate": 9.719211243599015e-06,
+      "loss": 0.8859,
+      "step": 1247
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8130683519951711,
+      "learning_rate": 9.718635713076677e-06,
+      "loss": 1.0101,
+      "step": 1248
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6560385235192612,
+      "learning_rate": 9.718059610403493e-06,
+      "loss": 0.9231,
+      "step": 1249
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8177146048601935,
+      "learning_rate": 9.717482935649313e-06,
+      "loss": 0.9771,
+      "step": 1250
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7395339593735508,
+      "learning_rate": 9.716905688884065e-06,
+      "loss": 0.9146,
+      "step": 1251
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7389112591296354,
+      "learning_rate": 9.716327870177739e-06,
+      "loss": 0.929,
+      "step": 1252
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7990127525203634,
+      "learning_rate": 9.7157494796004e-06,
+      "loss": 0.9416,
+      "step": 1253
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8008901225028542,
+      "learning_rate": 9.71517051722218e-06,
+      "loss": 0.9843,
+      "step": 1254
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.695225925972305,
+      "learning_rate": 9.714590983113279e-06,
+      "loss": 0.9486,
+      "step": 1255
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0323666532676747,
+      "learning_rate": 9.714010877343967e-06,
+      "loss": 0.9838,
+      "step": 1256
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7474522306776232,
+      "learning_rate": 9.713430199984583e-06,
+      "loss": 0.9453,
+      "step": 1257
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7329963737080902,
+      "learning_rate": 9.71284895110554e-06,
+      "loss": 0.9651,
+      "step": 1258
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.655091167228909,
+      "learning_rate": 9.712267130777313e-06,
+      "loss": 0.8903,
+      "step": 1259
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6270740211826473,
+      "learning_rate": 9.71168473907045e-06,
+      "loss": 1.0478,
+      "step": 1260
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8726126384154236,
+      "learning_rate": 9.71110177605557e-06,
+      "loss": 0.9691,
+      "step": 1261
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8305500669363823,
+      "learning_rate": 9.710518241803357e-06,
+      "loss": 1.0793,
+      "step": 1262
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7136656451697307,
+      "learning_rate": 9.70993413638457e-06,
+      "loss": 0.9529,
+      "step": 1263
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.813780300934652,
+      "learning_rate": 9.709349459870028e-06,
+      "loss": 0.9345,
+      "step": 1264
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6472900963296162,
+      "learning_rate": 9.70876421233063e-06,
+      "loss": 0.9406,
+      "step": 1265
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.723428613503388,
+      "learning_rate": 9.708178393837338e-06,
+      "loss": 1.0019,
+      "step": 1266
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.858267166961602,
+      "learning_rate": 9.707592004461183e-06,
+      "loss": 0.9419,
+      "step": 1267
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8143513651790806,
+      "learning_rate": 9.707005044273268e-06,
+      "loss": 0.914,
+      "step": 1268
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7384561200174327,
+      "learning_rate": 9.706417513344764e-06,
+      "loss": 0.9817,
+      "step": 1269
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7266648002075469,
+      "learning_rate": 9.705829411746912e-06,
+      "loss": 0.9013,
+      "step": 1270
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7628945304962549,
+      "learning_rate": 9.70524073955102e-06,
+      "loss": 0.978,
+      "step": 1271
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7109629466915632,
+      "learning_rate": 9.704651496828466e-06,
+      "loss": 0.9939,
+      "step": 1272
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6675938200800113,
+      "learning_rate": 9.704061683650701e-06,
+      "loss": 0.8742,
+      "step": 1273
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6415771502746803,
+      "learning_rate": 9.703471300089237e-06,
+      "loss": 0.8555,
+      "step": 1274
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0159705629237386,
+      "learning_rate": 9.702880346215664e-06,
+      "loss": 1.061,
+      "step": 1275
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8782974841500362,
+      "learning_rate": 9.702288822101634e-06,
+      "loss": 0.9643,
+      "step": 1276
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9142163252362512,
+      "learning_rate": 9.701696727818875e-06,
+      "loss": 0.9218,
+      "step": 1277
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7945349869411686,
+      "learning_rate": 9.701104063439177e-06,
+      "loss": 0.9468,
+      "step": 1278
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6823058108530389,
+      "learning_rate": 9.700510829034404e-06,
+      "loss": 0.894,
+      "step": 1279
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6471700079296356,
+      "learning_rate": 9.699917024676489e-06,
+      "loss": 0.9257,
+      "step": 1280
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6269068923806016,
+      "learning_rate": 9.699322650437433e-06,
+      "loss": 0.8255,
+      "step": 1281
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7222566133072286,
+      "learning_rate": 9.698727706389301e-06,
+      "loss": 0.9025,
+      "step": 1282
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.96198972310481,
+      "learning_rate": 9.698132192604239e-06,
+      "loss": 0.8239,
+      "step": 1283
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7052648327891688,
+      "learning_rate": 9.69753610915445e-06,
+      "loss": 0.9713,
+      "step": 1284
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7640973430706142,
+      "learning_rate": 9.696939456112213e-06,
+      "loss": 0.9105,
+      "step": 1285
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7377684409884142,
+      "learning_rate": 9.696342233549874e-06,
+      "loss": 0.8673,
+      "step": 1286
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8058118218246957,
+      "learning_rate": 9.695744441539849e-06,
+      "loss": 1.0093,
+      "step": 1287
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.747296920871417,
+      "learning_rate": 9.695146080154622e-06,
+      "loss": 0.9903,
+      "step": 1288
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.76291060426145,
+      "learning_rate": 9.694547149466745e-06,
+      "loss": 0.9424,
+      "step": 1289
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8177909459653936,
+      "learning_rate": 9.693947649548843e-06,
+      "loss": 0.9511,
+      "step": 1290
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.750848039221567,
+      "learning_rate": 9.693347580473604e-06,
+      "loss": 0.8586,
+      "step": 1291
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8107141727760652,
+      "learning_rate": 9.692746942313792e-06,
+      "loss": 0.9398,
+      "step": 1292
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1208810158622644,
+      "learning_rate": 9.692145735142236e-06,
+      "loss": 1.0863,
+      "step": 1293
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.758835623577077,
+      "learning_rate": 9.691543959031831e-06,
+      "loss": 1.0542,
+      "step": 1294
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8757005104982023,
+      "learning_rate": 9.690941614055547e-06,
+      "loss": 0.9939,
+      "step": 1295
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8712646055539242,
+      "learning_rate": 9.690338700286422e-06,
+      "loss": 0.9999,
+      "step": 1296
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.81666852819241,
+      "learning_rate": 9.689735217797559e-06,
+      "loss": 0.8466,
+      "step": 1297
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6671873570281903,
+      "learning_rate": 9.689131166662131e-06,
+      "loss": 0.9924,
+      "step": 1298
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9037719001113205,
+      "learning_rate": 9.688526546953384e-06,
+      "loss": 0.9326,
+      "step": 1299
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7457763336401118,
+      "learning_rate": 9.68792135874463e-06,
+      "loss": 0.9576,
+      "step": 1300
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.790134778334665,
+      "learning_rate": 9.687315602109248e-06,
+      "loss": 0.9133,
+      "step": 1301
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9387600455288883,
+      "learning_rate": 9.686709277120692e-06,
+      "loss": 0.9792,
+      "step": 1302
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7072807738289963,
+      "learning_rate": 9.686102383852478e-06,
+      "loss": 0.9369,
+      "step": 1303
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.858926821639685,
+      "learning_rate": 9.685494922378194e-06,
+      "loss": 1.0107,
+      "step": 1304
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8107108248500081,
+      "learning_rate": 9.684886892771496e-06,
+      "loss": 0.9232,
+      "step": 1305
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.210888916342686,
+      "learning_rate": 9.684278295106112e-06,
+      "loss": 0.8916,
+      "step": 1306
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9312078253379261,
+      "learning_rate": 9.683669129455837e-06,
+      "loss": 1.0018,
+      "step": 1307
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.693339068322031,
+      "learning_rate": 9.683059395894531e-06,
+      "loss": 0.8573,
+      "step": 1308
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7609878790371245,
+      "learning_rate": 9.682449094496129e-06,
+      "loss": 0.9593,
+      "step": 1309
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7791779636723473,
+      "learning_rate": 9.681838225334632e-06,
+      "loss": 0.8852,
+      "step": 1310
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8435741831611834,
+      "learning_rate": 9.681226788484108e-06,
+      "loss": 0.8646,
+      "step": 1311
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9005094583290882,
+      "learning_rate": 9.680614784018699e-06,
+      "loss": 0.9362,
+      "step": 1312
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6833583214904495,
+      "learning_rate": 9.68000221201261e-06,
+      "loss": 0.9593,
+      "step": 1313
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.805499964168594,
+      "learning_rate": 9.67938907254012e-06,
+      "loss": 0.9615,
+      "step": 1314
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3072561085902317,
+      "learning_rate": 9.67877536567557e-06,
+      "loss": 1.1319,
+      "step": 1315
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7784117500025727,
+      "learning_rate": 9.678161091493378e-06,
+      "loss": 0.9815,
+      "step": 1316
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8026128541578461,
+      "learning_rate": 9.677546250068025e-06,
+      "loss": 0.9723,
+      "step": 1317
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.486980377449916,
+      "learning_rate": 9.676930841474064e-06,
+      "loss": 0.9959,
+      "step": 1318
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8863922286334577,
+      "learning_rate": 9.676314865786114e-06,
+      "loss": 1.0542,
+      "step": 1319
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.75046126424522,
+      "learning_rate": 9.675698323078865e-06,
+      "loss": 0.9753,
+      "step": 1320
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9399716336946635,
+      "learning_rate": 9.675081213427076e-06,
+      "loss": 1.0143,
+      "step": 1321
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7157181648743232,
+      "learning_rate": 9.67446353690557e-06,
+      "loss": 0.9659,
+      "step": 1322
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8282768376815017,
+      "learning_rate": 9.673845293589246e-06,
+      "loss": 0.9337,
+      "step": 1323
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7252122580881186,
+      "learning_rate": 9.673226483553068e-06,
+      "loss": 0.9336,
+      "step": 1324
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8514618276661683,
+      "learning_rate": 9.672607106872067e-06,
+      "loss": 0.9351,
+      "step": 1325
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6597692093469811,
+      "learning_rate": 9.671987163621344e-06,
+      "loss": 0.9571,
+      "step": 1326
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7046655882246773,
+      "learning_rate": 9.671366653876071e-06,
+      "loss": 1.0213,
+      "step": 1327
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.1981672147474396,
+      "learning_rate": 9.670745577711488e-06,
+      "loss": 0.9167,
+      "step": 1328
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0390628311009635,
+      "learning_rate": 9.670123935202903e-06,
+      "loss": 1.0578,
+      "step": 1329
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8354635814158557,
+      "learning_rate": 9.669501726425687e-06,
+      "loss": 0.9165,
+      "step": 1330
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7475007969030938,
+      "learning_rate": 9.668878951455289e-06,
+      "loss": 0.9494,
+      "step": 1331
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.066607607222415,
+      "learning_rate": 9.668255610367223e-06,
+      "loss": 1.0763,
+      "step": 1332
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9332562758103905,
+      "learning_rate": 9.66763170323707e-06,
+      "loss": 1.0519,
+      "step": 1333
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7690239520875373,
+      "learning_rate": 9.66700723014048e-06,
+      "loss": 0.8634,
+      "step": 1334
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8655210938415188,
+      "learning_rate": 9.666382191153175e-06,
+      "loss": 0.8994,
+      "step": 1335
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7214682814997069,
+      "learning_rate": 9.665756586350942e-06,
+      "loss": 0.8422,
+      "step": 1336
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9189520783844412,
+      "learning_rate": 9.665130415809636e-06,
+      "loss": 0.9295,
+      "step": 1337
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7555819689715926,
+      "learning_rate": 9.664503679605186e-06,
+      "loss": 0.9372,
+      "step": 1338
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5120281178561668,
+      "learning_rate": 9.663876377813583e-06,
+      "loss": 1.0979,
+      "step": 1339
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8887715016352102,
+      "learning_rate": 9.663248510510891e-06,
+      "loss": 0.9317,
+      "step": 1340
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.78285620538769,
+      "learning_rate": 9.66262007777324e-06,
+      "loss": 0.9119,
+      "step": 1341
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8099054179515521,
+      "learning_rate": 9.661991079676829e-06,
+      "loss": 0.9395,
+      "step": 1342
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8409298147675015,
+      "learning_rate": 9.661361516297929e-06,
+      "loss": 1.0039,
+      "step": 1343
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7532029746844264,
+      "learning_rate": 9.660731387712874e-06,
+      "loss": 0.8862,
+      "step": 1344
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8677637739314561,
+      "learning_rate": 9.66010069399807e-06,
+      "loss": 1.0623,
+      "step": 1345
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6661236538454791,
+      "learning_rate": 9.659469435229992e-06,
+      "loss": 0.9802,
+      "step": 1346
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7228145119940494,
+      "learning_rate": 9.658837611485181e-06,
+      "loss": 0.8587,
+      "step": 1347
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9827889608803497,
+      "learning_rate": 9.658205222840249e-06,
+      "loss": 0.9886,
+      "step": 1348
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7001093162072172,
+      "learning_rate": 9.657572269371873e-06,
+      "loss": 0.9809,
+      "step": 1349
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7098572325468708,
+      "learning_rate": 9.656938751156803e-06,
+      "loss": 0.9224,
+      "step": 1350
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8442169760629519,
+      "learning_rate": 9.656304668271855e-06,
+      "loss": 0.9046,
+      "step": 1351
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7725951821894057,
+      "learning_rate": 9.655670020793911e-06,
+      "loss": 0.903,
+      "step": 1352
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.81089926782642,
+      "learning_rate": 9.655034808799928e-06,
+      "loss": 0.9439,
+      "step": 1353
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7046825599693822,
+      "learning_rate": 9.654399032366925e-06,
+      "loss": 0.8834,
+      "step": 1354
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0804309527702856,
+      "learning_rate": 9.653762691571995e-06,
+      "loss": 1.0607,
+      "step": 1355
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.695882114293869,
+      "learning_rate": 9.653125786492295e-06,
+      "loss": 1.0056,
+      "step": 1356
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7420609459047227,
+      "learning_rate": 9.652488317205048e-06,
+      "loss": 0.9403,
+      "step": 1357
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.793860685473266,
+      "learning_rate": 9.651850283787557e-06,
+      "loss": 0.9239,
+      "step": 1358
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7042076451463992,
+      "learning_rate": 9.65121168631718e-06,
+      "loss": 0.9065,
+      "step": 1359
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7708722354834414,
+      "learning_rate": 9.65057252487135e-06,
+      "loss": 0.9533,
+      "step": 1360
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6386179550890096,
+      "learning_rate": 9.64993279952757e-06,
+      "loss": 0.9549,
+      "step": 1361
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8536643936212485,
+      "learning_rate": 9.649292510363408e-06,
+      "loss": 0.9241,
+      "step": 1362
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7339524449091333,
+      "learning_rate": 9.648651657456498e-06,
+      "loss": 0.9297,
+      "step": 1363
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7631514901166518,
+      "learning_rate": 9.64801024088455e-06,
+      "loss": 0.9363,
+      "step": 1364
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6965774736226111,
+      "learning_rate": 9.647368260725336e-06,
+      "loss": 0.8785,
+      "step": 1365
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.706803745351107,
+      "learning_rate": 9.646725717056697e-06,
+      "loss": 0.911,
+      "step": 1366
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7600289695431448,
+      "learning_rate": 9.646082609956546e-06,
+      "loss": 0.915,
+      "step": 1367
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7488046788826288,
+      "learning_rate": 9.645438939502863e-06,
+      "loss": 0.9704,
+      "step": 1368
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7130732761905072,
+      "learning_rate": 9.644794705773691e-06,
+      "loss": 0.9354,
+      "step": 1369
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1601707805369146,
+      "learning_rate": 9.644149908847149e-06,
+      "loss": 1.06,
+      "step": 1370
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.69304970988638,
+      "learning_rate": 9.64350454880142e-06,
+      "loss": 0.9424,
+      "step": 1371
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9005509963414922,
+      "learning_rate": 9.642858625714753e-06,
+      "loss": 0.9743,
+      "step": 1372
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7852006723405949,
+      "learning_rate": 9.642212139665474e-06,
+      "loss": 0.9634,
+      "step": 1373
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8685194685507618,
+      "learning_rate": 9.641565090731967e-06,
+      "loss": 1.0291,
+      "step": 1374
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8566019828610887,
+      "learning_rate": 9.640917478992693e-06,
+      "loss": 0.9851,
+      "step": 1375
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7854845875722785,
+      "learning_rate": 9.640269304526175e-06,
+      "loss": 0.8802,
+      "step": 1376
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7813969315515208,
+      "learning_rate": 9.639620567411005e-06,
+      "loss": 0.9305,
+      "step": 1377
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7420129421277366,
+      "learning_rate": 9.638971267725847e-06,
+      "loss": 1.0297,
+      "step": 1378
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8008666170015144,
+      "learning_rate": 9.638321405549429e-06,
+      "loss": 0.9802,
+      "step": 1379
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.958862742974025,
+      "learning_rate": 9.63767098096055e-06,
+      "loss": 0.7991,
+      "step": 1380
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7704898569672072,
+      "learning_rate": 9.637019994038076e-06,
+      "loss": 0.8174,
+      "step": 1381
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9454146184495025,
+      "learning_rate": 9.636368444860942e-06,
+      "loss": 0.97,
+      "step": 1382
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6956410727746987,
+      "learning_rate": 9.63571633350815e-06,
+      "loss": 0.8851,
+      "step": 1383
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7753421821822153,
+      "learning_rate": 9.635063660058772e-06,
+      "loss": 0.979,
+      "step": 1384
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.826476811546799,
+      "learning_rate": 9.634410424591942e-06,
+      "loss": 1.0011,
+      "step": 1385
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.770390285888198,
+      "learning_rate": 9.633756627186874e-06,
+      "loss": 0.9638,
+      "step": 1386
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7045419233686316,
+      "learning_rate": 9.63310226792284e-06,
+      "loss": 0.8683,
+      "step": 1387
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7270966668884549,
+      "learning_rate": 9.632447346879182e-06,
+      "loss": 1.0009,
+      "step": 1388
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.022446249033957,
+      "learning_rate": 9.631791864135313e-06,
+      "loss": 0.9553,
+      "step": 1389
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7648309623355223,
+      "learning_rate": 9.631135819770711e-06,
+      "loss": 0.8933,
+      "step": 1390
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7833654836449024,
+      "learning_rate": 9.630479213864928e-06,
+      "loss": 0.9827,
+      "step": 1391
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.740149213184367,
+      "learning_rate": 9.629822046497575e-06,
+      "loss": 0.9161,
+      "step": 1392
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7662673731343899,
+      "learning_rate": 9.629164317748336e-06,
+      "loss": 0.9325,
+      "step": 1393
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7731996284175147,
+      "learning_rate": 9.628506027696967e-06,
+      "loss": 0.8705,
+      "step": 1394
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7094195388265079,
+      "learning_rate": 9.627847176423283e-06,
+      "loss": 0.8823,
+      "step": 1395
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9022802877599772,
+      "learning_rate": 9.627187764007175e-06,
+      "loss": 0.8905,
+      "step": 1396
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8615816218104457,
+      "learning_rate": 9.6265277905286e-06,
+      "loss": 0.961,
+      "step": 1397
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7684005418903954,
+      "learning_rate": 9.625867256067577e-06,
+      "loss": 0.9207,
+      "step": 1398
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8845015944913246,
+      "learning_rate": 9.625206160704204e-06,
+      "loss": 0.9769,
+      "step": 1399
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.625433004093357,
+      "learning_rate": 9.624544504518637e-06,
+      "loss": 0.9197,
+      "step": 1400
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6422140537518477,
+      "learning_rate": 9.623882287591106e-06,
+      "loss": 0.8393,
+      "step": 1401
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8103291103331431,
+      "learning_rate": 9.623219510001907e-06,
+      "loss": 0.9786,
+      "step": 1402
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7953363283505914,
+      "learning_rate": 9.622556171831404e-06,
+      "loss": 0.9641,
+      "step": 1403
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7144848456947013,
+      "learning_rate": 9.621892273160028e-06,
+      "loss": 0.9311,
+      "step": 1404
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8153083903694607,
+      "learning_rate": 9.62122781406828e-06,
+      "loss": 0.9042,
+      "step": 1405
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7453583393893406,
+      "learning_rate": 9.62056279463673e-06,
+      "loss": 0.9359,
+      "step": 1406
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.659512785089936,
+      "learning_rate": 9.619897214946007e-06,
+      "loss": 0.8652,
+      "step": 1407
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.733328565287686,
+      "learning_rate": 9.619231075076824e-06,
+      "loss": 0.9588,
+      "step": 1408
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6997106195565865,
+      "learning_rate": 9.618564375109946e-06,
+      "loss": 0.9283,
+      "step": 1409
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8751968559908663,
+      "learning_rate": 9.617897115126216e-06,
+      "loss": 1.0544,
+      "step": 1410
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9064539780729668,
+      "learning_rate": 9.617229295206538e-06,
+      "loss": 0.9258,
+      "step": 1411
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6711881808897096,
+      "learning_rate": 9.616560915431892e-06,
+      "loss": 0.9444,
+      "step": 1412
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8847200174043122,
+      "learning_rate": 9.61589197588332e-06,
+      "loss": 0.9827,
+      "step": 1413
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8738157795787422,
+      "learning_rate": 9.61522247664193e-06,
+      "loss": 1.0495,
+      "step": 1414
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7425273221469568,
+      "learning_rate": 9.614552417788906e-06,
+      "loss": 0.9573,
+      "step": 1415
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7575288608399968,
+      "learning_rate": 9.613881799405492e-06,
+      "loss": 0.9183,
+      "step": 1416
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8834375103441199,
+      "learning_rate": 9.613210621573003e-06,
+      "loss": 0.9823,
+      "step": 1417
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7690878548189666,
+      "learning_rate": 9.612538884372821e-06,
+      "loss": 0.9225,
+      "step": 1418
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6542470625678858,
+      "learning_rate": 9.611866587886399e-06,
+      "loss": 0.9525,
+      "step": 1419
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7621813487367117,
+      "learning_rate": 9.611193732195254e-06,
+      "loss": 0.9662,
+      "step": 1420
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1506346680658193,
+      "learning_rate": 9.61052031738097e-06,
+      "loss": 1.0839,
+      "step": 1421
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8449001583598301,
+      "learning_rate": 9.609846343525204e-06,
+      "loss": 0.9651,
+      "step": 1422
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7414555121333162,
+      "learning_rate": 9.609171810709676e-06,
+      "loss": 0.9264,
+      "step": 1423
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9024877412828572,
+      "learning_rate": 9.608496719016176e-06,
+      "loss": 1.0762,
+      "step": 1424
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.752707612053309,
+      "learning_rate": 9.60782106852656e-06,
+      "loss": 0.8828,
+      "step": 1425
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0026795989633763,
+      "learning_rate": 9.607144859322755e-06,
+      "loss": 1.0746,
+      "step": 1426
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8172629353399616,
+      "learning_rate": 9.60646809148675e-06,
+      "loss": 0.9938,
+      "step": 1427
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6832733744100667,
+      "learning_rate": 9.60579076510061e-06,
+      "loss": 0.9274,
+      "step": 1428
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9206543551158937,
+      "learning_rate": 9.605112880246463e-06,
+      "loss": 1.0608,
+      "step": 1429
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7606615110417487,
+      "learning_rate": 9.6044344370065e-06,
+      "loss": 0.8544,
+      "step": 1430
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7427976038830009,
+      "learning_rate": 9.603755435462989e-06,
+      "loss": 0.9507,
+      "step": 1431
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6525341772533055,
+      "learning_rate": 9.60307587569826e-06,
+      "loss": 0.8523,
+      "step": 1432
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8312518195293592,
+      "learning_rate": 9.602395757794712e-06,
+      "loss": 0.9351,
+      "step": 1433
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7165173862702916,
+      "learning_rate": 9.60171508183481e-06,
+      "loss": 0.9171,
+      "step": 1434
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8730417273271636,
+      "learning_rate": 9.601033847901092e-06,
+      "loss": 0.9468,
+      "step": 1435
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8074233371503403,
+      "learning_rate": 9.600352056076154e-06,
+      "loss": 0.9305,
+      "step": 1436
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0297949535982869,
+      "learning_rate": 9.599669706442674e-06,
+      "loss": 1.1048,
+      "step": 1437
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.913167798266604,
+      "learning_rate": 9.59898679908338e-06,
+      "loss": 0.8638,
+      "step": 1438
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8267080809909488,
+      "learning_rate": 9.598303334081084e-06,
+      "loss": 0.916,
+      "step": 1439
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8397781019634896,
+      "learning_rate": 9.597619311518658e-06,
+      "loss": 0.9389,
+      "step": 1440
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7805768704962996,
+      "learning_rate": 9.596934731479038e-06,
+      "loss": 0.8918,
+      "step": 1441
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8042871516598047,
+      "learning_rate": 9.596249594045233e-06,
+      "loss": 0.9516,
+      "step": 1442
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7528771138655306,
+      "learning_rate": 9.59556389930032e-06,
+      "loss": 0.9505,
+      "step": 1443
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0070859145349267,
+      "learning_rate": 9.594877647327441e-06,
+      "loss": 1.0661,
+      "step": 1444
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.700460230704621,
+      "learning_rate": 9.594190838209806e-06,
+      "loss": 0.8955,
+      "step": 1445
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.796119565528654,
+      "learning_rate": 9.593503472030693e-06,
+      "loss": 1.0348,
+      "step": 1446
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7041173326119494,
+      "learning_rate": 9.592815548873448e-06,
+      "loss": 0.8967,
+      "step": 1447
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7426517227021265,
+      "learning_rate": 9.592127068821484e-06,
+      "loss": 1.0777,
+      "step": 1448
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.848973488224146,
+      "learning_rate": 9.591438031958282e-06,
+      "loss": 1.0168,
+      "step": 1449
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8584809255109977,
+      "learning_rate": 9.59074843836739e-06,
+      "loss": 0.9333,
+      "step": 1450
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6375930421210068,
+      "learning_rate": 9.59005828813242e-06,
+      "loss": 0.9812,
+      "step": 1451
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6748034967136696,
+      "learning_rate": 9.589367581337062e-06,
+      "loss": 0.854,
+      "step": 1452
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7422802462084397,
+      "learning_rate": 9.58867631806506e-06,
+      "loss": 0.9287,
+      "step": 1453
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8116740123279091,
+      "learning_rate": 9.587984498400236e-06,
+      "loss": 1.0021,
+      "step": 1454
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7490236950223288,
+      "learning_rate": 9.587292122426475e-06,
+      "loss": 0.9234,
+      "step": 1455
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6769719591071819,
+      "learning_rate": 9.586599190227728e-06,
+      "loss": 0.9893,
+      "step": 1456
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8218259667147807,
+      "learning_rate": 9.585905701888019e-06,
+      "loss": 0.9153,
+      "step": 1457
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7692596437189836,
+      "learning_rate": 9.58521165749143e-06,
+      "loss": 0.9391,
+      "step": 1458
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9319441460901332,
+      "learning_rate": 9.584517057122122e-06,
+      "loss": 0.9197,
+      "step": 1459
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.785876796632594,
+      "learning_rate": 9.583821900864316e-06,
+      "loss": 0.9373,
+      "step": 1460
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7772596216568746,
+      "learning_rate": 9.583126188802301e-06,
+      "loss": 0.9584,
+      "step": 1461
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7079353356490585,
+      "learning_rate": 9.582429921020436e-06,
+      "loss": 0.9423,
+      "step": 1462
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8163617814990844,
+      "learning_rate": 9.581733097603146e-06,
+      "loss": 0.9219,
+      "step": 1463
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8811264923395157,
+      "learning_rate": 9.581035718634919e-06,
+      "loss": 0.9903,
+      "step": 1464
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6829421499951367,
+      "learning_rate": 9.58033778420032e-06,
+      "loss": 0.9204,
+      "step": 1465
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7351549270581252,
+      "learning_rate": 9.579639294383974e-06,
+      "loss": 0.9297,
+      "step": 1466
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.923461698038161,
+      "learning_rate": 9.578940249270573e-06,
+      "loss": 0.9457,
+      "step": 1467
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8909901997231626,
+      "learning_rate": 9.578240648944882e-06,
+      "loss": 0.9275,
+      "step": 1468
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7369715401791137,
+      "learning_rate": 9.577540493491731e-06,
+      "loss": 0.9252,
+      "step": 1469
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7988155171833973,
+      "learning_rate": 9.576839782996013e-06,
+      "loss": 1.0008,
+      "step": 1470
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7887788494593262,
+      "learning_rate": 9.576138517542691e-06,
+      "loss": 1.0239,
+      "step": 1471
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.807498054208189,
+      "learning_rate": 9.575436697216798e-06,
+      "loss": 0.9261,
+      "step": 1472
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9799835770722665,
+      "learning_rate": 9.574734322103431e-06,
+      "loss": 0.9492,
+      "step": 1473
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0846591782420674,
+      "learning_rate": 9.574031392287756e-06,
+      "loss": 1.0683,
+      "step": 1474
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6984755730529613,
+      "learning_rate": 9.573327907855006e-06,
+      "loss": 0.8705,
+      "step": 1475
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9126039045040988,
+      "learning_rate": 9.572623868890482e-06,
+      "loss": 0.9198,
+      "step": 1476
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.776189328225635,
+      "learning_rate": 9.571919275479549e-06,
+      "loss": 0.9325,
+      "step": 1477
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8854589844152346,
+      "learning_rate": 9.57121412770764e-06,
+      "loss": 1.0957,
+      "step": 1478
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6644984567232182,
+      "learning_rate": 9.570508425660261e-06,
+      "loss": 0.879,
+      "step": 1479
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8963164863101774,
+      "learning_rate": 9.569802169422978e-06,
+      "loss": 1.0621,
+      "step": 1480
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6499888132985112,
+      "learning_rate": 9.569095359081426e-06,
+      "loss": 0.9274,
+      "step": 1481
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.775357426320905,
+      "learning_rate": 9.56838799472131e-06,
+      "loss": 0.8879,
+      "step": 1482
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7619305700154273,
+      "learning_rate": 9.567680076428401e-06,
+      "loss": 0.9135,
+      "step": 1483
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8658049953436682,
+      "learning_rate": 9.566971604288535e-06,
+      "loss": 1.0586,
+      "step": 1484
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.883498273910403,
+      "learning_rate": 9.566262578387616e-06,
+      "loss": 1.0952,
+      "step": 1485
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7860736113050653,
+      "learning_rate": 9.56555299881162e-06,
+      "loss": 0.999,
+      "step": 1486
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.788308283411982,
+      "learning_rate": 9.564842865646581e-06,
+      "loss": 1.0242,
+      "step": 1487
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7915569612269753,
+      "learning_rate": 9.564132178978607e-06,
+      "loss": 0.9776,
+      "step": 1488
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9244510751017048,
+      "learning_rate": 9.563420938893872e-06,
+      "loss": 0.9698,
+      "step": 1489
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7809793029634062,
+      "learning_rate": 9.562709145478616e-06,
+      "loss": 0.9303,
+      "step": 1490
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.794132424495073,
+      "learning_rate": 9.561996798819146e-06,
+      "loss": 0.9136,
+      "step": 1491
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7885808940988799,
+      "learning_rate": 9.561283899001837e-06,
+      "loss": 0.9912,
+      "step": 1492
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9952579796826119,
+      "learning_rate": 9.560570446113128e-06,
+      "loss": 1.0934,
+      "step": 1493
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7479208534644994,
+      "learning_rate": 9.559856440239531e-06,
+      "loss": 0.9353,
+      "step": 1494
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6682315968058588,
+      "learning_rate": 9.55914188146762e-06,
+      "loss": 0.8799,
+      "step": 1495
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9186530244838425,
+      "learning_rate": 9.558426769884039e-06,
+      "loss": 0.9481,
+      "step": 1496
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6631815023849306,
+      "learning_rate": 9.557711105575496e-06,
+      "loss": 0.9056,
+      "step": 1497
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7643626574075302,
+      "learning_rate": 9.55699488862877e-06,
+      "loss": 0.8479,
+      "step": 1498
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8944318627905647,
+      "learning_rate": 9.556278119130702e-06,
+      "loss": 0.9493,
+      "step": 1499
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7930574629291882,
+      "learning_rate": 9.555560797168203e-06,
+      "loss": 0.8814,
+      "step": 1500
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7956590508579462,
+      "learning_rate": 9.554842922828254e-06,
+      "loss": 0.8887,
+      "step": 1501
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6233746068238766,
+      "learning_rate": 9.554124496197899e-06,
+      "loss": 0.8853,
+      "step": 1502
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8621655083509916,
+      "learning_rate": 9.553405517364246e-06,
+      "loss": 0.9964,
+      "step": 1503
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9785469971424361,
+      "learning_rate": 9.552685986414477e-06,
+      "loss": 1.0724,
+      "step": 1504
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6539520670444752,
+      "learning_rate": 9.551965903435836e-06,
+      "loss": 0.9225,
+      "step": 1505
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.795626046826408,
+      "learning_rate": 9.551245268515637e-06,
+      "loss": 0.9606,
+      "step": 1506
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.853922286442062,
+      "learning_rate": 9.550524081741256e-06,
+      "loss": 0.9395,
+      "step": 1507
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7391171053569394,
+      "learning_rate": 9.549802343200146e-06,
+      "loss": 0.9164,
+      "step": 1508
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6615961641332677,
+      "learning_rate": 9.549080052979814e-06,
+      "loss": 0.9132,
+      "step": 1509
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9129865662740381,
+      "learning_rate": 9.548357211167842e-06,
+      "loss": 1.0955,
+      "step": 1510
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7780192844165246,
+      "learning_rate": 9.54763381785188e-06,
+      "loss": 1.0247,
+      "step": 1511
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8701512239910684,
+      "learning_rate": 9.546909873119637e-06,
+      "loss": 1.0618,
+      "step": 1512
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.825190625487844,
+      "learning_rate": 9.5461853770589e-06,
+      "loss": 1.049,
+      "step": 1513
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7308811109640356,
+      "learning_rate": 9.545460329757508e-06,
+      "loss": 0.8756,
+      "step": 1514
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8222613359962494,
+      "learning_rate": 9.544734731303383e-06,
+      "loss": 0.8985,
+      "step": 1515
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7045491989571553,
+      "learning_rate": 9.544008581784503e-06,
+      "loss": 0.9322,
+      "step": 1516
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9376396489539779,
+      "learning_rate": 9.543281881288918e-06,
+      "loss": 1.0352,
+      "step": 1517
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8429203598553712,
+      "learning_rate": 9.542554629904742e-06,
+      "loss": 0.9538,
+      "step": 1518
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8521445078904208,
+      "learning_rate": 9.541826827720156e-06,
+      "loss": 0.9153,
+      "step": 1519
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7726482231253609,
+      "learning_rate": 9.541098474823408e-06,
+      "loss": 0.9313,
+      "step": 1520
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.942629681432199,
+      "learning_rate": 9.540369571302816e-06,
+      "loss": 0.9992,
+      "step": 1521
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7332496241713975,
+      "learning_rate": 9.539640117246758e-06,
+      "loss": 0.8966,
+      "step": 1522
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7502675788458923,
+      "learning_rate": 9.538910112743689e-06,
+      "loss": 0.9081,
+      "step": 1523
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8238649915952863,
+      "learning_rate": 9.538179557882119e-06,
+      "loss": 0.8455,
+      "step": 1524
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6886277968199084,
+      "learning_rate": 9.53744845275063e-06,
+      "loss": 0.8818,
+      "step": 1525
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.735663462504771,
+      "learning_rate": 9.536716797437876e-06,
+      "loss": 1.0018,
+      "step": 1526
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7631497545731312,
+      "learning_rate": 9.535984592032568e-06,
+      "loss": 0.9073,
+      "step": 1527
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.686686930994495,
+      "learning_rate": 9.535251836623491e-06,
+      "loss": 0.8776,
+      "step": 1528
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7324681530689945,
+      "learning_rate": 9.534518531299494e-06,
+      "loss": 0.9576,
+      "step": 1529
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.99698055434455,
+      "learning_rate": 9.533784676149493e-06,
+      "loss": 1.0729,
+      "step": 1530
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.932345823975987,
+      "learning_rate": 9.533050271262468e-06,
+      "loss": 1.0633,
+      "step": 1531
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6957282205133997,
+      "learning_rate": 9.53231531672747e-06,
+      "loss": 0.9416,
+      "step": 1532
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9018591607564523,
+      "learning_rate": 9.531579812633616e-06,
+      "loss": 0.9547,
+      "step": 1533
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7956514332877205,
+      "learning_rate": 9.530843759070085e-06,
+      "loss": 0.9372,
+      "step": 1534
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.751794833552981,
+      "learning_rate": 9.530107156126129e-06,
+      "loss": 0.9329,
+      "step": 1535
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8049222621106435,
+      "learning_rate": 9.529370003891063e-06,
+      "loss": 0.9409,
+      "step": 1536
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8368229853757845,
+      "learning_rate": 9.528632302454267e-06,
+      "loss": 0.9251,
+      "step": 1537
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7712436479689058,
+      "learning_rate": 9.527894051905194e-06,
+      "loss": 1.0048,
+      "step": 1538
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6860835177225004,
+      "learning_rate": 9.527155252333357e-06,
+      "loss": 0.9355,
+      "step": 1539
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.818524618931703,
+      "learning_rate": 9.526415903828338e-06,
+      "loss": 0.9591,
+      "step": 1540
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.73315252080244,
+      "learning_rate": 9.525676006479785e-06,
+      "loss": 0.9697,
+      "step": 1541
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7616940147017521,
+      "learning_rate": 9.524935560377414e-06,
+      "loss": 0.864,
+      "step": 1542
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.855426647962338,
+      "learning_rate": 9.524194565611007e-06,
+      "loss": 0.9304,
+      "step": 1543
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8273627965257853,
+      "learning_rate": 9.52345302227041e-06,
+      "loss": 1.0417,
+      "step": 1544
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8129023435010687,
+      "learning_rate": 9.522710930445542e-06,
+      "loss": 0.9342,
+      "step": 1545
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8393727213013886,
+      "learning_rate": 9.52196829022638e-06,
+      "loss": 0.9886,
+      "step": 1546
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8010432444774538,
+      "learning_rate": 9.521225101702973e-06,
+      "loss": 0.9397,
+      "step": 1547
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8917464923998197,
+      "learning_rate": 9.520481364965436e-06,
+      "loss": 0.9216,
+      "step": 1548
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7180294102427878,
+      "learning_rate": 9.519737080103949e-06,
+      "loss": 0.886,
+      "step": 1549
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.716168255277452,
+      "learning_rate": 9.518992247208757e-06,
+      "loss": 1.0659,
+      "step": 1550
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4818956791541413,
+      "learning_rate": 9.51824686637018e-06,
+      "loss": 1.0851,
+      "step": 1551
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7873947719702918,
+      "learning_rate": 9.517500937678591e-06,
+      "loss": 0.8801,
+      "step": 1552
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8554870371227414,
+      "learning_rate": 9.51675446122444e-06,
+      "loss": 0.9329,
+      "step": 1553
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8147428582141027,
+      "learning_rate": 9.516007437098238e-06,
+      "loss": 0.8973,
+      "step": 1554
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7790162313804563,
+      "learning_rate": 9.515259865390565e-06,
+      "loss": 0.8941,
+      "step": 1555
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8090950689132357,
+      "learning_rate": 9.514511746192067e-06,
+      "loss": 0.9054,
+      "step": 1556
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.834340272516845,
+      "learning_rate": 9.513763079593457e-06,
+      "loss": 0.9553,
+      "step": 1557
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.829696085841028,
+      "learning_rate": 9.513013865685511e-06,
+      "loss": 0.9906,
+      "step": 1558
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7354323245986263,
+      "learning_rate": 9.512264104559077e-06,
+      "loss": 0.9353,
+      "step": 1559
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.782968330099768,
+      "learning_rate": 9.511513796305063e-06,
+      "loss": 0.9714,
+      "step": 1560
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7612703633592532,
+      "learning_rate": 9.510762941014446e-06,
+      "loss": 0.8955,
+      "step": 1561
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.5247018338685168,
+      "learning_rate": 9.510011538778275e-06,
+      "loss": 0.9456,
+      "step": 1562
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7817367562990276,
+      "learning_rate": 9.509259589687653e-06,
+      "loss": 0.9536,
+      "step": 1563
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6190247079874307,
+      "learning_rate": 9.50850709383376e-06,
+      "loss": 1.014,
+      "step": 1564
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.278281096942023,
+      "learning_rate": 9.507754051307842e-06,
+      "loss": 1.0577,
+      "step": 1565
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8701996432779897,
+      "learning_rate": 9.507000462201202e-06,
+      "loss": 1.0803,
+      "step": 1566
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9105876298342516,
+      "learning_rate": 9.50624632660522e-06,
+      "loss": 1.0066,
+      "step": 1567
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9314157988698444,
+      "learning_rate": 9.505491644611334e-06,
+      "loss": 0.9443,
+      "step": 1568
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6952717504623727,
+      "learning_rate": 9.504736416311053e-06,
+      "loss": 0.9094,
+      "step": 1569
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7325830326206304,
+      "learning_rate": 9.503980641795952e-06,
+      "loss": 0.8888,
+      "step": 1570
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9293331985245432,
+      "learning_rate": 9.503224321157671e-06,
+      "loss": 0.9527,
+      "step": 1571
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7845539183150374,
+      "learning_rate": 9.502467454487916e-06,
+      "loss": 0.9819,
+      "step": 1572
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8699764226903586,
+      "learning_rate": 9.501710041878457e-06,
+      "loss": 0.9186,
+      "step": 1573
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7201167658587087,
+      "learning_rate": 9.50095208342114e-06,
+      "loss": 0.9955,
+      "step": 1574
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7885324426374476,
+      "learning_rate": 9.500193579207863e-06,
+      "loss": 0.9664,
+      "step": 1575
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7543327371272046,
+      "learning_rate": 9.499434529330601e-06,
+      "loss": 0.9355,
+      "step": 1576
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8054051675052007,
+      "learning_rate": 9.498674933881392e-06,
+      "loss": 0.9407,
+      "step": 1577
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9167670872810607,
+      "learning_rate": 9.497914792952334e-06,
+      "loss": 0.9918,
+      "step": 1578
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6363675984794586,
+      "learning_rate": 9.497154106635604e-06,
+      "loss": 0.9507,
+      "step": 1579
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8280664953722363,
+      "learning_rate": 9.496392875023433e-06,
+      "loss": 0.9117,
+      "step": 1580
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8179916217237777,
+      "learning_rate": 9.495631098208125e-06,
+      "loss": 0.9088,
+      "step": 1581
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8249914220542176,
+      "learning_rate": 9.494868776282047e-06,
+      "loss": 0.9137,
+      "step": 1582
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7461654862068758,
+      "learning_rate": 9.494105909337635e-06,
+      "loss": 0.9636,
+      "step": 1583
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8669692812582928,
+      "learning_rate": 9.493342497467386e-06,
+      "loss": 0.9348,
+      "step": 1584
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6526695338393733,
+      "learning_rate": 9.49257854076387e-06,
+      "loss": 0.8929,
+      "step": 1585
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8097276602422698,
+      "learning_rate": 9.491814039319716e-06,
+      "loss": 0.9531,
+      "step": 1586
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7848455431796935,
+      "learning_rate": 9.491048993227626e-06,
+      "loss": 0.9157,
+      "step": 1587
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8571682511734318,
+      "learning_rate": 9.490283402580362e-06,
+      "loss": 0.9349,
+      "step": 1588
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7167589560386909,
+      "learning_rate": 9.489517267470754e-06,
+      "loss": 0.8696,
+      "step": 1589
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.951280991041683,
+      "learning_rate": 9.488750587991701e-06,
+      "loss": 1.0203,
+      "step": 1590
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7177153904997686,
+      "learning_rate": 9.487983364236163e-06,
+      "loss": 1.0105,
+      "step": 1591
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8735576881426454,
+      "learning_rate": 9.48721559629717e-06,
+      "loss": 0.8333,
+      "step": 1592
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7055730593186267,
+      "learning_rate": 9.486447284267816e-06,
+      "loss": 1.0496,
+      "step": 1593
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6816028860308425,
+      "learning_rate": 9.485678428241264e-06,
+      "loss": 0.9392,
+      "step": 1594
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7563785817696747,
+      "learning_rate": 9.484909028310735e-06,
+      "loss": 0.9545,
+      "step": 1595
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9811548929859668,
+      "learning_rate": 9.484139084569526e-06,
+      "loss": 1.0301,
+      "step": 1596
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0281016942196723,
+      "learning_rate": 9.483368597110993e-06,
+      "loss": 1.0035,
+      "step": 1597
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.067732895678647,
+      "learning_rate": 9.48259756602856e-06,
+      "loss": 1.0671,
+      "step": 1598
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.07628810549085,
+      "learning_rate": 9.48182599141572e-06,
+      "loss": 1.0963,
+      "step": 1599
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6807206024487984,
+      "learning_rate": 9.481053873366027e-06,
+      "loss": 0.9125,
+      "step": 1600
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7711055404415776,
+      "learning_rate": 9.480281211973103e-06,
+      "loss": 0.9392,
+      "step": 1601
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8789025480854984,
+      "learning_rate": 9.479508007330638e-06,
+      "loss": 0.9708,
+      "step": 1602
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8169271454458398,
+      "learning_rate": 9.478734259532382e-06,
+      "loss": 0.9445,
+      "step": 1603
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8440287717479045,
+      "learning_rate": 9.477959968672156e-06,
+      "loss": 0.9278,
+      "step": 1604
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7337011372022615,
+      "learning_rate": 9.477185134843847e-06,
+      "loss": 1.0127,
+      "step": 1605
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.779199996416407,
+      "learning_rate": 9.476409758141404e-06,
+      "loss": 1.0169,
+      "step": 1606
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.626350079092121,
+      "learning_rate": 9.475633838658847e-06,
+      "loss": 0.8917,
+      "step": 1607
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9024358530279832,
+      "learning_rate": 9.474857376490257e-06,
+      "loss": 1.0131,
+      "step": 1608
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.74767419509368,
+      "learning_rate": 9.474080371729782e-06,
+      "loss": 0.8896,
+      "step": 1609
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6394150252177642,
+      "learning_rate": 9.473302824471637e-06,
+      "loss": 0.8858,
+      "step": 1610
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7423654745419432,
+      "learning_rate": 9.472524734810105e-06,
+      "loss": 0.9331,
+      "step": 1611
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7644524200453582,
+      "learning_rate": 9.471746102839527e-06,
+      "loss": 0.9179,
+      "step": 1612
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.220360376672148,
+      "learning_rate": 9.470966928654321e-06,
+      "loss": 1.0978,
+      "step": 1613
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8032750110490727,
+      "learning_rate": 9.470187212348958e-06,
+      "loss": 0.9178,
+      "step": 1614
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8356720451855801,
+      "learning_rate": 9.469406954017985e-06,
+      "loss": 0.9292,
+      "step": 1615
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9477171034498245,
+      "learning_rate": 9.468626153756011e-06,
+      "loss": 1.1025,
+      "step": 1616
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.862931298075745,
+      "learning_rate": 9.46784481165771e-06,
+      "loss": 1.0104,
+      "step": 1617
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.768769683658307,
+      "learning_rate": 9.467062927817822e-06,
+      "loss": 0.9502,
+      "step": 1618
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.88976214116001,
+      "learning_rate": 9.466280502331156e-06,
+      "loss": 1.0323,
+      "step": 1619
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7701421486604245,
+      "learning_rate": 9.465497535292579e-06,
+      "loss": 0.938,
+      "step": 1620
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.765281719228903,
+      "learning_rate": 9.464714026797032e-06,
+      "loss": 0.8698,
+      "step": 1621
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8079039872934388,
+      "learning_rate": 9.463929976939516e-06,
+      "loss": 0.954,
+      "step": 1622
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7234135617783135,
+      "learning_rate": 9.463145385815102e-06,
+      "loss": 0.9305,
+      "step": 1623
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6557501508638002,
+      "learning_rate": 9.462360253518923e-06,
+      "loss": 0.8903,
+      "step": 1624
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8457024241595237,
+      "learning_rate": 9.461574580146179e-06,
+      "loss": 0.9379,
+      "step": 1625
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8071902686092984,
+      "learning_rate": 9.460788365792135e-06,
+      "loss": 0.9648,
+      "step": 1626
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9248373207448017,
+      "learning_rate": 9.460001610552125e-06,
+      "loss": 0.9292,
+      "step": 1627
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.740968994604588,
+      "learning_rate": 9.45921431452154e-06,
+      "loss": 0.9059,
+      "step": 1628
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8340951487338804,
+      "learning_rate": 9.45842647779585e-06,
+      "loss": 0.9302,
+      "step": 1629
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7695954559215235,
+      "learning_rate": 9.457638100470577e-06,
+      "loss": 0.8537,
+      "step": 1630
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.086350342743236,
+      "learning_rate": 9.456849182641318e-06,
+      "loss": 1.0556,
+      "step": 1631
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8687061764855715,
+      "learning_rate": 9.45605972440373e-06,
+      "loss": 0.893,
+      "step": 1632
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9159875056050855,
+      "learning_rate": 9.455269725853538e-06,
+      "loss": 0.9313,
+      "step": 1633
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7296572387104225,
+      "learning_rate": 9.45447918708653e-06,
+      "loss": 0.9379,
+      "step": 1634
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7670363777478117,
+      "learning_rate": 9.453688108198567e-06,
+      "loss": 0.9185,
+      "step": 1635
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.656734510161265,
+      "learning_rate": 9.452896489285563e-06,
+      "loss": 0.9733,
+      "step": 1636
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7392638484626881,
+      "learning_rate": 9.45210433044351e-06,
+      "loss": 0.976,
+      "step": 1637
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0073051601870442,
+      "learning_rate": 9.451311631768459e-06,
+      "loss": 1.0418,
+      "step": 1638
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7278805469677507,
+      "learning_rate": 9.450518393356527e-06,
+      "loss": 0.9823,
+      "step": 1639
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7944010060562974,
+      "learning_rate": 9.449724615303894e-06,
+      "loss": 0.9177,
+      "step": 1640
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7058312883371887,
+      "learning_rate": 9.448930297706813e-06,
+      "loss": 0.9326,
+      "step": 1641
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8196673192989918,
+      "learning_rate": 9.448135440661595e-06,
+      "loss": 0.9506,
+      "step": 1642
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7291427316863979,
+      "learning_rate": 9.44734004426462e-06,
+      "loss": 0.9078,
+      "step": 1643
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8403164886108503,
+      "learning_rate": 9.446544108612332e-06,
+      "loss": 0.9225,
+      "step": 1644
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9127663778519497,
+      "learning_rate": 9.445747633801241e-06,
+      "loss": 0.93,
+      "step": 1645
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7592111347234174,
+      "learning_rate": 9.444950619927924e-06,
+      "loss": 0.8959,
+      "step": 1646
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8022025253155796,
+      "learning_rate": 9.44415306708902e-06,
+      "loss": 0.9831,
+      "step": 1647
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.820025829321247,
+      "learning_rate": 9.443354975381233e-06,
+      "loss": 0.8694,
+      "step": 1648
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.750461400262516,
+      "learning_rate": 9.44255634490134e-06,
+      "loss": 0.8559,
+      "step": 1649
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.830341755116359,
+      "learning_rate": 9.441757175746171e-06,
+      "loss": 0.834,
+      "step": 1650
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7265623475677363,
+      "learning_rate": 9.440957468012633e-06,
+      "loss": 0.8895,
+      "step": 1651
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6538500801793754,
+      "learning_rate": 9.440157221797692e-06,
+      "loss": 0.904,
+      "step": 1652
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8371413480294632,
+      "learning_rate": 9.43935643719838e-06,
+      "loss": 0.8971,
+      "step": 1653
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8165757942635419,
+      "learning_rate": 9.438555114311796e-06,
+      "loss": 0.9567,
+      "step": 1654
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.81669489268732,
+      "learning_rate": 9.4377532532351e-06,
+      "loss": 0.8942,
+      "step": 1655
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8451808848962241,
+      "learning_rate": 9.436950854065526e-06,
+      "loss": 0.9541,
+      "step": 1656
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8458044779466614,
+      "learning_rate": 9.436147916900361e-06,
+      "loss": 0.8467,
+      "step": 1657
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.722324078905247,
+      "learning_rate": 9.43534444183697e-06,
+      "loss": 0.8859,
+      "step": 1658
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.691866988807825,
+      "learning_rate": 9.434540428972772e-06,
+      "loss": 0.9492,
+      "step": 1659
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7502877887173267,
+      "learning_rate": 9.433735878405262e-06,
+      "loss": 0.8941,
+      "step": 1660
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7792898371870884,
+      "learning_rate": 9.43293079023199e-06,
+      "loss": 1.025,
+      "step": 1661
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.762664681017269,
+      "learning_rate": 9.432125164550576e-06,
+      "loss": 0.8761,
+      "step": 1662
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8601389065199223,
+      "learning_rate": 9.431319001458705e-06,
+      "loss": 1.0363,
+      "step": 1663
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7526931412635853,
+      "learning_rate": 9.43051230105413e-06,
+      "loss": 0.9018,
+      "step": 1664
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.114019706039584,
+      "learning_rate": 9.429705063434664e-06,
+      "loss": 1.0547,
+      "step": 1665
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8480240255875005,
+      "learning_rate": 9.428897288698189e-06,
+      "loss": 0.9316,
+      "step": 1666
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.635479378220326,
+      "learning_rate": 9.428088976942647e-06,
+      "loss": 0.9115,
+      "step": 1667
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.790600979480228,
+      "learning_rate": 9.427280128266049e-06,
+      "loss": 0.9439,
+      "step": 1668
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7525994600566974,
+      "learning_rate": 9.426470742766476e-06,
+      "loss": 0.9443,
+      "step": 1669
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7789619974972335,
+      "learning_rate": 9.425660820542064e-06,
+      "loss": 0.8716,
+      "step": 1670
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8117298472593428,
+      "learning_rate": 9.424850361691019e-06,
+      "loss": 1.0137,
+      "step": 1671
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7097333731050373,
+      "learning_rate": 9.424039366311612e-06,
+      "loss": 0.9298,
+      "step": 1672
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2594893555758384,
+      "learning_rate": 9.423227834502183e-06,
+      "loss": 1.0603,
+      "step": 1673
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8179521064441009,
+      "learning_rate": 9.422415766361125e-06,
+      "loss": 0.9371,
+      "step": 1674
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.730241451741859,
+      "learning_rate": 9.421603161986912e-06,
+      "loss": 0.9825,
+      "step": 1675
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8318522423858585,
+      "learning_rate": 9.420790021478073e-06,
+      "loss": 0.9114,
+      "step": 1676
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.659267656269558,
+      "learning_rate": 9.419976344933201e-06,
+      "loss": 0.9546,
+      "step": 1677
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9429111508718713,
+      "learning_rate": 9.419162132450961e-06,
+      "loss": 0.9045,
+      "step": 1678
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7985025022189522,
+      "learning_rate": 9.418347384130076e-06,
+      "loss": 1.0022,
+      "step": 1679
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7071283665607146,
+      "learning_rate": 9.41753210006934e-06,
+      "loss": 0.9458,
+      "step": 1680
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7996364207984064,
+      "learning_rate": 9.416716280367605e-06,
+      "loss": 0.9557,
+      "step": 1681
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.851427660718819,
+      "learning_rate": 9.415899925123796e-06,
+      "loss": 0.9572,
+      "step": 1682
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6162901433304024,
+      "learning_rate": 9.415083034436896e-06,
+      "loss": 0.9134,
+      "step": 1683
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8304412921571063,
+      "learning_rate": 9.414265608405956e-06,
+      "loss": 1.0524,
+      "step": 1684
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7312831483848148,
+      "learning_rate": 9.413447647130096e-06,
+      "loss": 0.9604,
+      "step": 1685
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6867956895027407,
+      "learning_rate": 9.412629150708493e-06,
+      "loss": 0.9964,
+      "step": 1686
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7260900851908219,
+      "learning_rate": 9.41181011924039e-06,
+      "loss": 0.8596,
+      "step": 1687
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8338422456172219,
+      "learning_rate": 9.4109905528251e-06,
+      "loss": 0.9942,
+      "step": 1688
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 6.276983262554703,
+      "learning_rate": 9.410170451562002e-06,
+      "loss": 0.8987,
+      "step": 1689
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7464955134675695,
+      "learning_rate": 9.40934981555053e-06,
+      "loss": 0.9934,
+      "step": 1690
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9891399845125455,
+      "learning_rate": 9.408528644890191e-06,
+      "loss": 1.0958,
+      "step": 1691
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8236482413130668,
+      "learning_rate": 9.407706939680557e-06,
+      "loss": 0.97,
+      "step": 1692
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.956203143244961,
+      "learning_rate": 9.406884700021261e-06,
+      "loss": 0.9836,
+      "step": 1693
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7783697360440844,
+      "learning_rate": 9.406061926012001e-06,
+      "loss": 0.9919,
+      "step": 1694
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7980040165228948,
+      "learning_rate": 9.405238617752543e-06,
+      "loss": 0.9264,
+      "step": 1695
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8977718336965468,
+      "learning_rate": 9.404414775342715e-06,
+      "loss": 0.9411,
+      "step": 1696
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.716048278216333,
+      "learning_rate": 9.403590398882411e-06,
+      "loss": 0.8783,
+      "step": 1697
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8687999347529645,
+      "learning_rate": 9.402765488471592e-06,
+      "loss": 0.9226,
+      "step": 1698
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7921786998645939,
+      "learning_rate": 9.401940044210276e-06,
+      "loss": 1.0053,
+      "step": 1699
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8893247857046835,
+      "learning_rate": 9.401114066198556e-06,
+      "loss": 0.9306,
+      "step": 1700
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.911770251978091,
+      "learning_rate": 9.400287554536582e-06,
+      "loss": 0.9063,
+      "step": 1701
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7639527171051343,
+      "learning_rate": 9.399460509324571e-06,
+      "loss": 0.947,
+      "step": 1702
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.678642982496663,
+      "learning_rate": 9.398632930662805e-06,
+      "loss": 0.9691,
+      "step": 1703
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7640254632149746,
+      "learning_rate": 9.397804818651634e-06,
+      "loss": 0.9074,
+      "step": 1704
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7275336488265538,
+      "learning_rate": 9.396976173391467e-06,
+      "loss": 0.9465,
+      "step": 1705
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6870582043790836,
+      "learning_rate": 9.39614699498278e-06,
+      "loss": 0.9308,
+      "step": 1706
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.875282164695613,
+      "learning_rate": 9.395317283526113e-06,
+      "loss": 0.8677,
+      "step": 1707
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0021100237895224,
+      "learning_rate": 9.394487039122074e-06,
+      "loss": 1.0353,
+      "step": 1708
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.685631021467314,
+      "learning_rate": 9.39365626187133e-06,
+      "loss": 0.9138,
+      "step": 1709
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6326056023592486,
+      "learning_rate": 9.392824951874618e-06,
+      "loss": 0.9474,
+      "step": 1710
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8360342479292169,
+      "learning_rate": 9.391993109232735e-06,
+      "loss": 1.0661,
+      "step": 1711
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7775888298773734,
+      "learning_rate": 9.391160734046548e-06,
+      "loss": 0.8829,
+      "step": 1712
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7382127354905041,
+      "learning_rate": 9.390327826416981e-06,
+      "loss": 0.956,
+      "step": 1713
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8915432667767361,
+      "learning_rate": 9.38949438644503e-06,
+      "loss": 1.0573,
+      "step": 1714
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8271534237491605,
+      "learning_rate": 9.388660414231752e-06,
+      "loss": 0.946,
+      "step": 1715
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8903463557802382,
+      "learning_rate": 9.387825909878269e-06,
+      "loss": 0.9542,
+      "step": 1716
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.756842681333753,
+      "learning_rate": 9.386990873485767e-06,
+      "loss": 0.9901,
+      "step": 1717
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7395299761818248,
+      "learning_rate": 9.386155305155498e-06,
+      "loss": 0.8932,
+      "step": 1718
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.007461065323802,
+      "learning_rate": 9.385319204988777e-06,
+      "loss": 0.8891,
+      "step": 1719
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7450897912687702,
+      "learning_rate": 9.384482573086983e-06,
+      "loss": 0.9365,
+      "step": 1720
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.74741028697665,
+      "learning_rate": 9.383645409551561e-06,
+      "loss": 0.9097,
+      "step": 1721
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9872578539790238,
+      "learning_rate": 9.382807714484021e-06,
+      "loss": 1.0602,
+      "step": 1722
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8811157894796312,
+      "learning_rate": 9.381969487985936e-06,
+      "loss": 1.0649,
+      "step": 1723
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8183150384506244,
+      "learning_rate": 9.381130730158944e-06,
+      "loss": 0.9838,
+      "step": 1724
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6657369674080504,
+      "learning_rate": 9.380291441104748e-06,
+      "loss": 0.9889,
+      "step": 1725
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7400245509389738,
+      "learning_rate": 9.379451620925112e-06,
+      "loss": 0.9416,
+      "step": 1726
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.773389312861087,
+      "learning_rate": 9.378611269721867e-06,
+      "loss": 0.9152,
+      "step": 1727
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7173663822412004,
+      "learning_rate": 9.377770387596911e-06,
+      "loss": 0.954,
+      "step": 1728
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.886290186201423,
+      "learning_rate": 9.376928974652205e-06,
+      "loss": 0.9685,
+      "step": 1729
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.173724170079782,
+      "learning_rate": 9.37608703098977e-06,
+      "loss": 1.0623,
+      "step": 1730
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7162980369423855,
+      "learning_rate": 9.375244556711696e-06,
+      "loss": 0.8327,
+      "step": 1731
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.818891667608514,
+      "learning_rate": 9.374401551920135e-06,
+      "loss": 0.9076,
+      "step": 1732
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.807792908512656,
+      "learning_rate": 9.373558016717306e-06,
+      "loss": 0.8853,
+      "step": 1733
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.770136249884862,
+      "learning_rate": 9.37271395120549e-06,
+      "loss": 0.9191,
+      "step": 1734
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8245802616585581,
+      "learning_rate": 9.37186935548703e-06,
+      "loss": 1.0523,
+      "step": 1735
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.682712444213008,
+      "learning_rate": 9.371024229664342e-06,
+      "loss": 0.8898,
+      "step": 1736
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9603860104010589,
+      "learning_rate": 9.370178573839894e-06,
+      "loss": 0.9586,
+      "step": 1737
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8444404000709858,
+      "learning_rate": 9.36933238811623e-06,
+      "loss": 1.0851,
+      "step": 1738
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7520007827510016,
+      "learning_rate": 9.368485672595949e-06,
+      "loss": 0.9783,
+      "step": 1739
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8687984153549504,
+      "learning_rate": 9.36763842738172e-06,
+      "loss": 0.9161,
+      "step": 1740
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.785100797770794,
+      "learning_rate": 9.366790652576274e-06,
+      "loss": 0.9118,
+      "step": 1741
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8303576721024377,
+      "learning_rate": 9.365942348282407e-06,
+      "loss": 0.9912,
+      "step": 1742
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8976531811747362,
+      "learning_rate": 9.365093514602979e-06,
+      "loss": 1.0925,
+      "step": 1743
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.806749114395195,
+      "learning_rate": 9.364244151640913e-06,
+      "loss": 0.9934,
+      "step": 1744
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7479843534337158,
+      "learning_rate": 9.363394259499197e-06,
+      "loss": 0.8819,
+      "step": 1745
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8609222091437427,
+      "learning_rate": 9.362543838280886e-06,
+      "loss": 1.0513,
+      "step": 1746
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.77354517792749,
+      "learning_rate": 9.361692888089094e-06,
+      "loss": 0.9278,
+      "step": 1747
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.79386575789392,
+      "learning_rate": 9.360841409027002e-06,
+      "loss": 0.9378,
+      "step": 1748
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.744791541652462,
+      "learning_rate": 9.359989401197853e-06,
+      "loss": 0.9286,
+      "step": 1749
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7402018810828206,
+      "learning_rate": 9.359136864704962e-06,
+      "loss": 0.9058,
+      "step": 1750
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7596778410455853,
+      "learning_rate": 9.358283799651694e-06,
+      "loss": 0.9815,
+      "step": 1751
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.754761807858723,
+      "learning_rate": 9.357430206141492e-06,
+      "loss": 0.9455,
+      "step": 1752
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7327405049413305,
+      "learning_rate": 9.356576084277856e-06,
+      "loss": 0.9538,
+      "step": 1753
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8396020467896548,
+      "learning_rate": 9.355721434164349e-06,
+      "loss": 0.92,
+      "step": 1754
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7726794497885092,
+      "learning_rate": 9.354866255904602e-06,
+      "loss": 0.8798,
+      "step": 1755
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9095979274070243,
+      "learning_rate": 9.354010549602308e-06,
+      "loss": 0.916,
+      "step": 1756
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6597055764282813,
+      "learning_rate": 9.353154315361223e-06,
+      "loss": 0.9121,
+      "step": 1757
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8546264695961023,
+      "learning_rate": 9.352297553285172e-06,
+      "loss": 0.9795,
+      "step": 1758
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.686545908487543,
+      "learning_rate": 9.351440263478036e-06,
+      "loss": 0.9389,
+      "step": 1759
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.685972669062883,
+      "learning_rate": 9.350582446043768e-06,
+      "loss": 0.9533,
+      "step": 1760
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6002553818014105,
+      "learning_rate": 9.34972410108638e-06,
+      "loss": 0.9161,
+      "step": 1761
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7821144165092586,
+      "learning_rate": 9.348865228709947e-06,
+      "loss": 0.9715,
+      "step": 1762
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8242500155584662,
+      "learning_rate": 9.348005829018613e-06,
+      "loss": 0.9391,
+      "step": 1763
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8326264201222315,
+      "learning_rate": 9.347145902116583e-06,
+      "loss": 0.9884,
+      "step": 1764
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6548577983005488,
+      "learning_rate": 9.346285448108125e-06,
+      "loss": 0.9427,
+      "step": 1765
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9314200512681357,
+      "learning_rate": 9.345424467097572e-06,
+      "loss": 0.9757,
+      "step": 1766
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.650666830597433,
+      "learning_rate": 9.344562959189321e-06,
+      "loss": 0.9408,
+      "step": 1767
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.766967884709704,
+      "learning_rate": 9.343700924487835e-06,
+      "loss": 0.8825,
+      "step": 1768
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7275370066164668,
+      "learning_rate": 9.342838363097634e-06,
+      "loss": 0.9065,
+      "step": 1769
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.781654308048231,
+      "learning_rate": 9.341975275123314e-06,
+      "loss": 0.8703,
+      "step": 1770
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0388997801494995,
+      "learning_rate": 9.34111166066952e-06,
+      "loss": 0.9073,
+      "step": 1771
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9582044532087621,
+      "learning_rate": 9.34024751984097e-06,
+      "loss": 0.9287,
+      "step": 1772
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8782088926863287,
+      "learning_rate": 9.339382852742447e-06,
+      "loss": 0.9203,
+      "step": 1773
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7621544218526413,
+      "learning_rate": 9.338517659478792e-06,
+      "loss": 0.9262,
+      "step": 1774
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8200918598568752,
+      "learning_rate": 9.337651940154914e-06,
+      "loss": 0.9448,
+      "step": 1775
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7470440485922911,
+      "learning_rate": 9.336785694875785e-06,
+      "loss": 0.9773,
+      "step": 1776
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5642024307005644,
+      "learning_rate": 9.335918923746438e-06,
+      "loss": 0.8281,
+      "step": 1777
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7097256380956931,
+      "learning_rate": 9.335051626871973e-06,
+      "loss": 0.9031,
+      "step": 1778
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.697528024305107,
+      "learning_rate": 9.334183804357555e-06,
+      "loss": 0.9291,
+      "step": 1779
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8407619815860479,
+      "learning_rate": 9.333315456308407e-06,
+      "loss": 0.893,
+      "step": 1780
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8555201941547579,
+      "learning_rate": 9.332446582829822e-06,
+      "loss": 0.8904,
+      "step": 1781
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9246445703760912,
+      "learning_rate": 9.33157718402715e-06,
+      "loss": 1.0409,
+      "step": 1782
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6467760915998826,
+      "learning_rate": 9.330707260005814e-06,
+      "loss": 0.9291,
+      "step": 1783
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.829094922897823,
+      "learning_rate": 9.329836810871291e-06,
+      "loss": 0.9296,
+      "step": 1784
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.695543105295909,
+      "learning_rate": 9.328965836729128e-06,
+      "loss": 0.8795,
+      "step": 1785
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8156998307582315,
+      "learning_rate": 9.328094337684933e-06,
+      "loss": 0.9188,
+      "step": 1786
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6837655551636772,
+      "learning_rate": 9.327222313844377e-06,
+      "loss": 0.872,
+      "step": 1787
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.230197384319136,
+      "learning_rate": 9.326349765313199e-06,
+      "loss": 1.007,
+      "step": 1788
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9701909453835805,
+      "learning_rate": 9.325476692197197e-06,
+      "loss": 1.0777,
+      "step": 1789
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6385497318681554,
+      "learning_rate": 9.324603094602232e-06,
+      "loss": 0.8702,
+      "step": 1790
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.841278967636332,
+      "learning_rate": 9.323728972634234e-06,
+      "loss": 0.9683,
+      "step": 1791
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7037357736735774,
+      "learning_rate": 9.322854326399192e-06,
+      "loss": 0.9455,
+      "step": 1792
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7448140330582516,
+      "learning_rate": 9.32197915600316e-06,
+      "loss": 0.9036,
+      "step": 1793
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7820526916193953,
+      "learning_rate": 9.321103461552254e-06,
+      "loss": 0.919,
+      "step": 1794
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.700915342819063,
+      "learning_rate": 9.320227243152657e-06,
+      "loss": 0.9322,
+      "step": 1795
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7778023382917663,
+      "learning_rate": 9.319350500910613e-06,
+      "loss": 1.0182,
+      "step": 1796
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7556500912438093,
+      "learning_rate": 9.318473234932428e-06,
+      "loss": 0.932,
+      "step": 1797
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7218221285388764,
+      "learning_rate": 9.317595445324477e-06,
+      "loss": 0.8063,
+      "step": 1798
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8407928481093185,
+      "learning_rate": 9.316717132193193e-06,
+      "loss": 0.9927,
+      "step": 1799
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9881181254512363,
+      "learning_rate": 9.315838295645074e-06,
+      "loss": 0.9339,
+      "step": 1800
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7534256507588757,
+      "learning_rate": 9.314958935786684e-06,
+      "loss": 0.9479,
+      "step": 1801
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7442417586911638,
+      "learning_rate": 9.314079052724645e-06,
+      "loss": 0.93,
+      "step": 1802
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7582023410301424,
+      "learning_rate": 9.313198646565648e-06,
+      "loss": 0.9448,
+      "step": 1803
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8034612792690485,
+      "learning_rate": 9.312317717416448e-06,
+      "loss": 0.9446,
+      "step": 1804
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8207857613822092,
+      "learning_rate": 9.311436265383856e-06,
+      "loss": 0.9265,
+      "step": 1805
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.755879362342169,
+      "learning_rate": 9.310554290574754e-06,
+      "loss": 0.9036,
+      "step": 1806
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8196916446957612,
+      "learning_rate": 9.309671793096082e-06,
+      "loss": 0.9843,
+      "step": 1807
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8259028164946325,
+      "learning_rate": 9.308788773054848e-06,
+      "loss": 0.9441,
+      "step": 1808
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7506992265911372,
+      "learning_rate": 9.307905230558123e-06,
+      "loss": 0.9003,
+      "step": 1809
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9348390311276369,
+      "learning_rate": 9.307021165713034e-06,
+      "loss": 1.0728,
+      "step": 1810
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8950757941280718,
+      "learning_rate": 9.30613657862678e-06,
+      "loss": 1.089,
+      "step": 1811
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7776572878461376,
+      "learning_rate": 9.305251469406621e-06,
+      "loss": 0.8855,
+      "step": 1812
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7863774216540826,
+      "learning_rate": 9.304365838159882e-06,
+      "loss": 0.9823,
+      "step": 1813
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.552522256098939,
+      "learning_rate": 9.303479684993943e-06,
+      "loss": 0.9288,
+      "step": 1814
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8666547722967624,
+      "learning_rate": 9.302593010016255e-06,
+      "loss": 0.9639,
+      "step": 1815
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7985689090779444,
+      "learning_rate": 9.301705813334332e-06,
+      "loss": 0.8416,
+      "step": 1816
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7903888698289745,
+      "learning_rate": 9.30081809505575e-06,
+      "loss": 0.9327,
+      "step": 1817
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7064702630194581,
+      "learning_rate": 9.299929855288145e-06,
+      "loss": 0.8514,
+      "step": 1818
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.746260069970438,
+      "learning_rate": 9.299041094139222e-06,
+      "loss": 0.9537,
+      "step": 1819
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7040369892385567,
+      "learning_rate": 9.298151811716745e-06,
+      "loss": 0.9251,
+      "step": 1820
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7503350095215338,
+      "learning_rate": 9.297262008128544e-06,
+      "loss": 0.9434,
+      "step": 1821
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.824164418442658,
+      "learning_rate": 9.296371683482508e-06,
+      "loss": 0.9574,
+      "step": 1822
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.604268613892651,
+      "learning_rate": 9.295480837886595e-06,
+      "loss": 0.9298,
+      "step": 1823
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7468451005464873,
+      "learning_rate": 9.294589471448819e-06,
+      "loss": 0.9784,
+      "step": 1824
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7733493999230137,
+      "learning_rate": 9.293697584277266e-06,
+      "loss": 0.8951,
+      "step": 1825
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6832605816026116,
+      "learning_rate": 9.292805176480077e-06,
+      "loss": 0.987,
+      "step": 1826
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.877231066029553,
+      "learning_rate": 9.291912248165462e-06,
+      "loss": 0.9943,
+      "step": 1827
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7085726284566523,
+      "learning_rate": 9.291018799441692e-06,
+      "loss": 0.9338,
+      "step": 1828
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7256740659041032,
+      "learning_rate": 9.290124830417097e-06,
+      "loss": 0.9365,
+      "step": 1829
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.538575987113729,
+      "learning_rate": 9.289230341200075e-06,
+      "loss": 0.9307,
+      "step": 1830
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3475312689214456,
+      "learning_rate": 9.288335331899089e-06,
+      "loss": 1.0863,
+      "step": 1831
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8768824636934855,
+      "learning_rate": 9.28743980262266e-06,
+      "loss": 1.0205,
+      "step": 1832
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9390586983641143,
+      "learning_rate": 9.286543753479372e-06,
+      "loss": 0.9533,
+      "step": 1833
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7434974476243281,
+      "learning_rate": 9.285647184577878e-06,
+      "loss": 0.9595,
+      "step": 1834
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7208624336344924,
+      "learning_rate": 9.284750096026886e-06,
+      "loss": 0.9126,
+      "step": 1835
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.847226770154086,
+      "learning_rate": 9.283852487935174e-06,
+      "loss": 0.8367,
+      "step": 1836
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0007329963835727,
+      "learning_rate": 9.282954360411579e-06,
+      "loss": 0.9466,
+      "step": 1837
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8307194966465414,
+      "learning_rate": 9.282055713565e-06,
+      "loss": 0.9192,
+      "step": 1838
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6531047767734504,
+      "learning_rate": 9.28115654750441e-06,
+      "loss": 0.9956,
+      "step": 1839
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7991623053045893,
+      "learning_rate": 9.280256862338822e-06,
+      "loss": 0.8799,
+      "step": 1840
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8347461879243223,
+      "learning_rate": 9.279356658177337e-06,
+      "loss": 0.9448,
+      "step": 1841
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0182955352522247,
+      "learning_rate": 9.278455935129103e-06,
+      "loss": 1.0577,
+      "step": 1842
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7375096391564306,
+      "learning_rate": 9.277554693303337e-06,
+      "loss": 0.9574,
+      "step": 1843
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.838341923274779,
+      "learning_rate": 9.276652932809315e-06,
+      "loss": 0.9177,
+      "step": 1844
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8903074648011653,
+      "learning_rate": 9.275750653756385e-06,
+      "loss": 1.0053,
+      "step": 1845
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.738829040351453,
+      "learning_rate": 9.274847856253946e-06,
+      "loss": 0.953,
+      "step": 1846
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8822018117669301,
+      "learning_rate": 9.273944540411465e-06,
+      "loss": 0.9243,
+      "step": 1847
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6463768866952586,
+      "learning_rate": 9.273040706338476e-06,
+      "loss": 0.9678,
+      "step": 1848
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.769238046477046,
+      "learning_rate": 9.272136354144569e-06,
+      "loss": 0.9719,
+      "step": 1849
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7172992350183847,
+      "learning_rate": 9.271231483939403e-06,
+      "loss": 0.9165,
+      "step": 1850
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.021291352187464,
+      "learning_rate": 9.27032609583269e-06,
+      "loss": 0.9904,
+      "step": 1851
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7460976182349062,
+      "learning_rate": 9.269420189934219e-06,
+      "loss": 1.016,
+      "step": 1852
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8008142021154498,
+      "learning_rate": 9.26851376635383e-06,
+      "loss": 0.9852,
+      "step": 1853
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7223077810668341,
+      "learning_rate": 9.267606825201433e-06,
+      "loss": 0.9899,
+      "step": 1854
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7058910764921766,
+      "learning_rate": 9.266699366586992e-06,
+      "loss": 0.9553,
+      "step": 1855
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7452840246724481,
+      "learning_rate": 9.265791390620547e-06,
+      "loss": 0.9465,
+      "step": 1856
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.3819751045892827,
+      "learning_rate": 9.26488289741219e-06,
+      "loss": 0.8812,
+      "step": 1857
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8648699474764825,
+      "learning_rate": 9.263973887072074e-06,
+      "loss": 0.9226,
+      "step": 1858
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7958055821328507,
+      "learning_rate": 9.263064359710428e-06,
+      "loss": 0.872,
+      "step": 1859
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0999266959027558,
+      "learning_rate": 9.262154315437528e-06,
+      "loss": 1.0782,
+      "step": 1860
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8704991863521319,
+      "learning_rate": 9.261243754363726e-06,
+      "loss": 0.9617,
+      "step": 1861
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8873386507933301,
+      "learning_rate": 9.260332676599426e-06,
+      "loss": 1.1003,
+      "step": 1862
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8304825682040373,
+      "learning_rate": 9.259421082255104e-06,
+      "loss": 0.9339,
+      "step": 1863
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8233412860484728,
+      "learning_rate": 9.258508971441289e-06,
+      "loss": 0.9645,
+      "step": 1864
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.801967920479368,
+      "learning_rate": 9.25759634426858e-06,
+      "loss": 0.8628,
+      "step": 1865
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.816464119954529,
+      "learning_rate": 9.256683200847638e-06,
+      "loss": 0.9329,
+      "step": 1866
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7761026302078886,
+      "learning_rate": 9.255769541289182e-06,
+      "loss": 0.9358,
+      "step": 1867
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.009542918545504,
+      "learning_rate": 9.254855365703997e-06,
+      "loss": 0.9197,
+      "step": 1868
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7656313833329986,
+      "learning_rate": 9.253940674202931e-06,
+      "loss": 0.9428,
+      "step": 1869
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8082558098668886,
+      "learning_rate": 9.25302546689689e-06,
+      "loss": 0.9944,
+      "step": 1870
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.761377002250604,
+      "learning_rate": 9.25210974389685e-06,
+      "loss": 0.9599,
+      "step": 1871
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6759222998936374,
+      "learning_rate": 9.251193505313845e-06,
+      "loss": 0.9614,
+      "step": 1872
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6489028487714614,
+      "learning_rate": 9.250276751258972e-06,
+      "loss": 0.9479,
+      "step": 1873
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.815609536416996,
+      "learning_rate": 9.249359481843389e-06,
+      "loss": 0.9816,
+      "step": 1874
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.959187744583045,
+      "learning_rate": 9.24844169717832e-06,
+      "loss": 0.9655,
+      "step": 1875
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0772192460809054,
+      "learning_rate": 9.247523397375047e-06,
+      "loss": 1.0347,
+      "step": 1876
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.802144048816886,
+      "learning_rate": 9.24660458254492e-06,
+      "loss": 0.9378,
+      "step": 1877
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4350032285232037,
+      "learning_rate": 9.245685252799346e-06,
+      "loss": 1.0747,
+      "step": 1878
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.1314340863301897,
+      "learning_rate": 9.244765408249798e-06,
+      "loss": 0.9392,
+      "step": 1879
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.754997241278803,
+      "learning_rate": 9.243845049007811e-06,
+      "loss": 1.0273,
+      "step": 1880
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7184841740951815,
+      "learning_rate": 9.242924175184982e-06,
+      "loss": 0.9322,
+      "step": 1881
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8224043722200962,
+      "learning_rate": 9.242002786892967e-06,
+      "loss": 0.9573,
+      "step": 1882
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.727080341621473,
+      "learning_rate": 9.241080884243492e-06,
+      "loss": 0.9657,
+      "step": 1883
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.739393047694169,
+      "learning_rate": 9.240158467348337e-06,
+      "loss": 0.9188,
+      "step": 1884
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9086289176822075,
+      "learning_rate": 9.239235536319351e-06,
+      "loss": 0.9648,
+      "step": 1885
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1104797586184179,
+      "learning_rate": 9.23831209126844e-06,
+      "loss": 1.0312,
+      "step": 1886
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7186667571263123,
+      "learning_rate": 9.237388132307576e-06,
+      "loss": 0.9911,
+      "step": 1887
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9059308660291685,
+      "learning_rate": 9.236463659548793e-06,
+      "loss": 1.0682,
+      "step": 1888
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8484201095349662,
+      "learning_rate": 9.235538673104187e-06,
+      "loss": 0.9949,
+      "step": 1889
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.712884427017372,
+      "learning_rate": 9.234613173085913e-06,
+      "loss": 0.9316,
+      "step": 1890
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7710726546352922,
+      "learning_rate": 9.233687159606195e-06,
+      "loss": 0.9493,
+      "step": 1891
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7899813051033093,
+      "learning_rate": 9.232760632777311e-06,
+      "loss": 0.9703,
+      "step": 1892
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7220805068269962,
+      "learning_rate": 9.23183359271161e-06,
+      "loss": 0.9483,
+      "step": 1893
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2041934668183552,
+      "learning_rate": 9.230906039521495e-06,
+      "loss": 1.0683,
+      "step": 1894
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.800817264327702,
+      "learning_rate": 9.229977973319436e-06,
+      "loss": 0.9755,
+      "step": 1895
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7557702104161894,
+      "learning_rate": 9.229049394217966e-06,
+      "loss": 0.935,
+      "step": 1896
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7523909232200132,
+      "learning_rate": 9.228120302329678e-06,
+      "loss": 0.995,
+      "step": 1897
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7341058691265612,
+      "learning_rate": 9.227190697767224e-06,
+      "loss": 0.9823,
+      "step": 1898
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8093974763259668,
+      "learning_rate": 9.226260580643327e-06,
+      "loss": 1.0165,
+      "step": 1899
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.701566868782907,
+      "learning_rate": 9.225329951070762e-06,
+      "loss": 0.9336,
+      "step": 1900
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7643830016985842,
+      "learning_rate": 9.224398809162376e-06,
+      "loss": 0.9272,
+      "step": 1901
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7193677240663048,
+      "learning_rate": 9.22346715503107e-06,
+      "loss": 0.9115,
+      "step": 1902
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6523688576396802,
+      "learning_rate": 9.222534988789811e-06,
+      "loss": 1.0473,
+      "step": 1903
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8254339957045995,
+      "learning_rate": 9.221602310551627e-06,
+      "loss": 0.9849,
+      "step": 1904
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8164326724473965,
+      "learning_rate": 9.220669120429608e-06,
+      "loss": 1.0118,
+      "step": 1905
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6952945644758017,
+      "learning_rate": 9.219735418536907e-06,
+      "loss": 0.9142,
+      "step": 1906
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8127470274743847,
+      "learning_rate": 9.21880120498674e-06,
+      "loss": 0.9711,
+      "step": 1907
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7777075350040212,
+      "learning_rate": 9.217866479892383e-06,
+      "loss": 0.952,
+      "step": 1908
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8675985061156481,
+      "learning_rate": 9.216931243367173e-06,
+      "loss": 1.0273,
+      "step": 1909
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7844054403162997,
+      "learning_rate": 9.215995495524512e-06,
+      "loss": 0.8971,
+      "step": 1910
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9795868481997869,
+      "learning_rate": 9.215059236477864e-06,
+      "loss": 1.0587,
+      "step": 1911
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7062858728779509,
+      "learning_rate": 9.21412246634075e-06,
+      "loss": 0.8878,
+      "step": 1912
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.253490183547763,
+      "learning_rate": 9.21318518522676e-06,
+      "loss": 0.9562,
+      "step": 1913
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6131464294757167,
+      "learning_rate": 9.21224739324954e-06,
+      "loss": 0.9003,
+      "step": 1914
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8289091336103582,
+      "learning_rate": 9.211309090522803e-06,
+      "loss": 0.9115,
+      "step": 1915
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7645742491245373,
+      "learning_rate": 9.21037027716032e-06,
+      "loss": 0.8872,
+      "step": 1916
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6715217605712385,
+      "learning_rate": 9.209430953275925e-06,
+      "loss": 0.9059,
+      "step": 1917
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7572363930732027,
+      "learning_rate": 9.208491118983515e-06,
+      "loss": 0.9447,
+      "step": 1918
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7309616359720637,
+      "learning_rate": 9.207550774397048e-06,
+      "loss": 0.9098,
+      "step": 1919
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6613283057277448,
+      "learning_rate": 9.206609919630543e-06,
+      "loss": 0.8194,
+      "step": 1920
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.820017003936623,
+      "learning_rate": 9.205668554798084e-06,
+      "loss": 0.9482,
+      "step": 1921
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7141743574117203,
+      "learning_rate": 9.204726680013813e-06,
+      "loss": 0.9112,
+      "step": 1922
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7079502598769296,
+      "learning_rate": 9.203784295391938e-06,
+      "loss": 0.955,
+      "step": 1923
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7130405763889902,
+      "learning_rate": 9.202841401046722e-06,
+      "loss": 0.92,
+      "step": 1924
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8284579852362663,
+      "learning_rate": 9.201897997092497e-06,
+      "loss": 1.0011,
+      "step": 1925
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9640964629548258,
+      "learning_rate": 9.200954083643654e-06,
+      "loss": 1.0599,
+      "step": 1926
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8076322487427539,
+      "learning_rate": 9.200009660814646e-06,
+      "loss": 0.9872,
+      "step": 1927
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7460418256181771,
+      "learning_rate": 9.199064728719988e-06,
+      "loss": 0.9384,
+      "step": 1928
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6510572174913778,
+      "learning_rate": 9.198119287474254e-06,
+      "loss": 0.9243,
+      "step": 1929
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7380002672061294,
+      "learning_rate": 9.197173337192082e-06,
+      "loss": 1.0062,
+      "step": 1930
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.065750187065165,
+      "learning_rate": 9.196226877988174e-06,
+      "loss": 0.9388,
+      "step": 1931
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.799031228706279,
+      "learning_rate": 9.195279909977293e-06,
+      "loss": 0.8353,
+      "step": 1932
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7871456475913279,
+      "learning_rate": 9.194332433274256e-06,
+      "loss": 0.8984,
+      "step": 1933
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6330175217381413,
+      "learning_rate": 9.193384447993954e-06,
+      "loss": 0.8531,
+      "step": 1934
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7079918150996587,
+      "learning_rate": 9.192435954251329e-06,
+      "loss": 0.9251,
+      "step": 1935
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8067540213974942,
+      "learning_rate": 9.191486952161393e-06,
+      "loss": 0.8999,
+      "step": 1936
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7074488265497858,
+      "learning_rate": 9.190537441839212e-06,
+      "loss": 0.8949,
+      "step": 1937
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6080487497123068,
+      "learning_rate": 9.189587423399919e-06,
+      "loss": 0.9929,
+      "step": 1938
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6855637728507937,
+      "learning_rate": 9.188636896958708e-06,
+      "loss": 0.9412,
+      "step": 1939
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6811233933981744,
+      "learning_rate": 9.187685862630833e-06,
+      "loss": 0.9483,
+      "step": 1940
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.83258543017693,
+      "learning_rate": 9.18673432053161e-06,
+      "loss": 0.9655,
+      "step": 1941
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6355337096686275,
+      "learning_rate": 9.185782270776416e-06,
+      "loss": 0.964,
+      "step": 1942
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7439263912013652,
+      "learning_rate": 9.184829713480691e-06,
+      "loss": 1.0006,
+      "step": 1943
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.753830070938035,
+      "learning_rate": 9.183876648759937e-06,
+      "loss": 0.9466,
+      "step": 1944
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7433409829098994,
+      "learning_rate": 9.182923076729716e-06,
+      "loss": 0.9783,
+      "step": 1945
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8154363994905607,
+      "learning_rate": 9.181968997505649e-06,
+      "loss": 0.9937,
+      "step": 1946
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7301117743495247,
+      "learning_rate": 9.181014411203426e-06,
+      "loss": 0.8778,
+      "step": 1947
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7462066214556098,
+      "learning_rate": 9.18005931793879e-06,
+      "loss": 0.8916,
+      "step": 1948
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7346272739587119,
+      "learning_rate": 9.179103717827551e-06,
+      "loss": 0.886,
+      "step": 1949
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8157250308424318,
+      "learning_rate": 9.178147610985578e-06,
+      "loss": 0.8509,
+      "step": 1950
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.907719918967169,
+      "learning_rate": 9.177190997528805e-06,
+      "loss": 0.8829,
+      "step": 1951
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.070732855239176,
+      "learning_rate": 9.176233877573219e-06,
+      "loss": 0.9263,
+      "step": 1952
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.881909811871556,
+      "learning_rate": 9.175276251234881e-06,
+      "loss": 0.8981,
+      "step": 1953
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.2023823781281506,
+      "learning_rate": 9.1743181186299e-06,
+      "loss": 0.9203,
+      "step": 1954
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8743617936782138,
+      "learning_rate": 9.17335947987446e-06,
+      "loss": 0.874,
+      "step": 1955
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6858633318633545,
+      "learning_rate": 9.172400335084793e-06,
+      "loss": 0.9295,
+      "step": 1956
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8299749008338706,
+      "learning_rate": 9.171440684377201e-06,
+      "loss": 0.906,
+      "step": 1957
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8708404177649354,
+      "learning_rate": 9.170480527868046e-06,
+      "loss": 0.9131,
+      "step": 1958
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7831789643063127,
+      "learning_rate": 9.169519865673748e-06,
+      "loss": 0.9699,
+      "step": 1959
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6263784491331625,
+      "learning_rate": 9.168558697910792e-06,
+      "loss": 0.9357,
+      "step": 1960
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8531462836832402,
+      "learning_rate": 9.167597024695722e-06,
+      "loss": 0.973,
+      "step": 1961
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7458217831272926,
+      "learning_rate": 9.166634846145146e-06,
+      "loss": 0.9441,
+      "step": 1962
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8338101350552134,
+      "learning_rate": 9.165672162375731e-06,
+      "loss": 0.8176,
+      "step": 1963
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7833876917242943,
+      "learning_rate": 9.164708973504203e-06,
+      "loss": 0.9604,
+      "step": 1964
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.744128167259554,
+      "learning_rate": 9.163745279647356e-06,
+      "loss": 0.8866,
+      "step": 1965
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0425432478350012,
+      "learning_rate": 9.162781080922039e-06,
+      "loss": 1.0636,
+      "step": 1966
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.741259031907156,
+      "learning_rate": 9.161816377445164e-06,
+      "loss": 0.9246,
+      "step": 1967
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8325636565956467,
+      "learning_rate": 9.160851169333704e-06,
+      "loss": 0.9681,
+      "step": 1968
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8282740156943096,
+      "learning_rate": 9.159885456704696e-06,
+      "loss": 0.9145,
+      "step": 1969
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8454847036119734,
+      "learning_rate": 9.158919239675237e-06,
+      "loss": 1.0265,
+      "step": 1970
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8524124044119996,
+      "learning_rate": 9.157952518362478e-06,
+      "loss": 0.9422,
+      "step": 1971
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7076001566816923,
+      "learning_rate": 9.156985292883645e-06,
+      "loss": 0.9501,
+      "step": 1972
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8150945945841483,
+      "learning_rate": 9.156017563356013e-06,
+      "loss": 0.9645,
+      "step": 1973
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8068897688800742,
+      "learning_rate": 9.155049329896924e-06,
+      "loss": 0.8709,
+      "step": 1974
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7240076575381553,
+      "learning_rate": 9.154080592623777e-06,
+      "loss": 0.8868,
+      "step": 1975
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7820911229507752,
+      "learning_rate": 9.15311135165404e-06,
+      "loss": 0.9454,
+      "step": 1976
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8046001686002016,
+      "learning_rate": 9.152141607105231e-06,
+      "loss": 1.0538,
+      "step": 1977
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7880007130694058,
+      "learning_rate": 9.15117135909494e-06,
+      "loss": 0.9394,
+      "step": 1978
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.176333230222553,
+      "learning_rate": 9.15020060774081e-06,
+      "loss": 1.0574,
+      "step": 1979
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7154907261379015,
+      "learning_rate": 9.149229353160546e-06,
+      "loss": 0.9249,
+      "step": 1980
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.711133456808274,
+      "learning_rate": 9.14825759547192e-06,
+      "loss": 0.9297,
+      "step": 1981
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7338962056504033,
+      "learning_rate": 9.14728533479276e-06,
+      "loss": 0.8757,
+      "step": 1982
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8605516292058175,
+      "learning_rate": 9.146312571240955e-06,
+      "loss": 1.016,
+      "step": 1983
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9102413862835839,
+      "learning_rate": 9.145339304934453e-06,
+      "loss": 0.9666,
+      "step": 1984
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6929354084248291,
+      "learning_rate": 9.144365535991273e-06,
+      "loss": 0.9328,
+      "step": 1985
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8152299667650476,
+      "learning_rate": 9.143391264529482e-06,
+      "loss": 1.0348,
+      "step": 1986
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7466243149743401,
+      "learning_rate": 9.142416490667217e-06,
+      "loss": 0.8595,
+      "step": 1987
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8326824901628382,
+      "learning_rate": 9.14144121452267e-06,
+      "loss": 0.9404,
+      "step": 1988
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7347266573221758,
+      "learning_rate": 9.1404654362141e-06,
+      "loss": 0.911,
+      "step": 1989
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6922171787383444,
+      "learning_rate": 9.13948915585982e-06,
+      "loss": 0.9196,
+      "step": 1990
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7867279211598206,
+      "learning_rate": 9.13851237357821e-06,
+      "loss": 0.9403,
+      "step": 1991
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9622757548302667,
+      "learning_rate": 9.137535089487705e-06,
+      "loss": 1.08,
+      "step": 1992
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.794744899221232,
+      "learning_rate": 9.13655730370681e-06,
+      "loss": 0.9066,
+      "step": 1993
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7114966031589056,
+      "learning_rate": 9.135579016354077e-06,
+      "loss": 0.9057,
+      "step": 1994
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.783678781052789,
+      "learning_rate": 9.134600227548133e-06,
+      "loss": 0.9339,
+      "step": 1995
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8616079528022769,
+      "learning_rate": 9.133620937407656e-06,
+      "loss": 1.0765,
+      "step": 1996
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.806231891121564,
+      "learning_rate": 9.13264114605139e-06,
+      "loss": 0.9005,
+      "step": 1997
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6599378469707955,
+      "learning_rate": 9.131660853598139e-06,
+      "loss": 0.9987,
+      "step": 1998
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8407034843126884,
+      "learning_rate": 9.130680060166764e-06,
+      "loss": 0.9262,
+      "step": 1999
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7616703572122319,
+      "learning_rate": 9.129698765876191e-06,
+      "loss": 0.9467,
+      "step": 2000
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7704948726994363,
+      "learning_rate": 9.128716970845407e-06,
+      "loss": 0.9364,
+      "step": 2001
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6724763735565322,
+      "learning_rate": 9.127734675193454e-06,
+      "loss": 0.9046,
+      "step": 2002
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9252039183448373,
+      "learning_rate": 9.126751879039442e-06,
+      "loss": 0.913,
+      "step": 2003
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.807972552966048,
+      "learning_rate": 9.12576858250254e-06,
+      "loss": 0.9207,
+      "step": 2004
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8972406342722112,
+      "learning_rate": 9.12478478570197e-06,
+      "loss": 1.0629,
+      "step": 2005
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7374363063749585,
+      "learning_rate": 9.123800488757027e-06,
+      "loss": 0.9665,
+      "step": 2006
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7645876656194983,
+      "learning_rate": 9.122815691787056e-06,
+      "loss": 0.9274,
+      "step": 2007
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7849554629199056,
+      "learning_rate": 9.121830394911471e-06,
+      "loss": 0.8234,
+      "step": 2008
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7489132991385636,
+      "learning_rate": 9.120844598249738e-06,
+      "loss": 0.9712,
+      "step": 2009
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7689031761572482,
+      "learning_rate": 9.119858301921391e-06,
+      "loss": 0.9483,
+      "step": 2010
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.875902364873694,
+      "learning_rate": 9.118871506046025e-06,
+      "loss": 0.9142,
+      "step": 2011
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.800079356393919,
+      "learning_rate": 9.117884210743287e-06,
+      "loss": 0.9266,
+      "step": 2012
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.806287813430833,
+      "learning_rate": 9.11689641613289e-06,
+      "loss": 0.8733,
+      "step": 2013
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9654706768499086,
+      "learning_rate": 9.115908122334611e-06,
+      "loss": 1.0616,
+      "step": 2014
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8285384671339544,
+      "learning_rate": 9.114919329468283e-06,
+      "loss": 0.9547,
+      "step": 2015
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6156231840262498,
+      "learning_rate": 9.113930037653799e-06,
+      "loss": 0.991,
+      "step": 2016
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8602906061650213,
+      "learning_rate": 9.112940247011116e-06,
+      "loss": 0.9468,
+      "step": 2017
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8690035661700586,
+      "learning_rate": 9.111949957660248e-06,
+      "loss": 0.9612,
+      "step": 2018
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7556032481202462,
+      "learning_rate": 9.110959169721272e-06,
+      "loss": 0.9621,
+      "step": 2019
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8119238569987048,
+      "learning_rate": 9.109967883314323e-06,
+      "loss": 1.0198,
+      "step": 2020
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.781791863960439,
+      "learning_rate": 9.108976098559601e-06,
+      "loss": 0.9422,
+      "step": 2021
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6706407494460285,
+      "learning_rate": 9.107983815577359e-06,
+      "loss": 0.8778,
+      "step": 2022
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5926693854778404,
+      "learning_rate": 9.106991034487917e-06,
+      "loss": 0.8268,
+      "step": 2023
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8895972079387284,
+      "learning_rate": 9.105997755411656e-06,
+      "loss": 1.011,
+      "step": 2024
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7668174462851103,
+      "learning_rate": 9.10500397846901e-06,
+      "loss": 0.8454,
+      "step": 2025
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7614907238180986,
+      "learning_rate": 9.104009703780478e-06,
+      "loss": 1.0237,
+      "step": 2026
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.773006796288227,
+      "learning_rate": 9.103014931466622e-06,
+      "loss": 1.019,
+      "step": 2027
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7470113057396002,
+      "learning_rate": 9.102019661648061e-06,
+      "loss": 0.953,
+      "step": 2028
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7763088829541076,
+      "learning_rate": 9.101023894445474e-06,
+      "loss": 0.915,
+      "step": 2029
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8139989681481696,
+      "learning_rate": 9.100027629979599e-06,
+      "loss": 0.8918,
+      "step": 2030
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8097044932966238,
+      "learning_rate": 9.099030868371241e-06,
+      "loss": 0.987,
+      "step": 2031
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8458615203568167,
+      "learning_rate": 9.09803360974126e-06,
+      "loss": 0.9091,
+      "step": 2032
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7841846436001092,
+      "learning_rate": 9.097035854210574e-06,
+      "loss": 0.8944,
+      "step": 2033
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.903974254590973,
+      "learning_rate": 9.096037601900166e-06,
+      "loss": 1.0455,
+      "step": 2034
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.808336384004518,
+      "learning_rate": 9.095038852931077e-06,
+      "loss": 0.9046,
+      "step": 2035
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7916841903918468,
+      "learning_rate": 9.09403960742441e-06,
+      "loss": 0.9039,
+      "step": 2036
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7569978091041596,
+      "learning_rate": 9.093039865501328e-06,
+      "loss": 0.9101,
+      "step": 2037
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.733811857902193,
+      "learning_rate": 9.09203962728305e-06,
+      "loss": 0.9681,
+      "step": 2038
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.763277805303069,
+      "learning_rate": 9.09103889289086e-06,
+      "loss": 0.8328,
+      "step": 2039
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7230424703294562,
+      "learning_rate": 9.0900376624461e-06,
+      "loss": 0.8593,
+      "step": 2040
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6363534489276021,
+      "learning_rate": 9.089035936070172e-06,
+      "loss": 0.9459,
+      "step": 2041
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6782533624285076,
+      "learning_rate": 9.08803371388454e-06,
+      "loss": 0.9391,
+      "step": 2042
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6826036915294544,
+      "learning_rate": 9.087030996010728e-06,
+      "loss": 0.9393,
+      "step": 2043
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7950544393713717,
+      "learning_rate": 9.086027782570318e-06,
+      "loss": 0.9637,
+      "step": 2044
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7276078090214835,
+      "learning_rate": 9.085024073684952e-06,
+      "loss": 0.9055,
+      "step": 2045
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.766277783126572,
+      "learning_rate": 9.084019869476332e-06,
+      "loss": 0.9954,
+      "step": 2046
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8146453698645704,
+      "learning_rate": 9.083015170066225e-06,
+      "loss": 0.8676,
+      "step": 2047
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7566042710347756,
+      "learning_rate": 9.082009975576452e-06,
+      "loss": 0.8732,
+      "step": 2048
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.878409513251426,
+      "learning_rate": 9.081004286128896e-06,
+      "loss": 1.0147,
+      "step": 2049
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8089996758901536,
+      "learning_rate": 9.079998101845502e-06,
+      "loss": 0.895,
+      "step": 2050
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8572021115348551,
+      "learning_rate": 9.07899142284827e-06,
+      "loss": 0.9963,
+      "step": 2051
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7620703972493956,
+      "learning_rate": 9.077984249259268e-06,
+      "loss": 0.9656,
+      "step": 2052
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6591946235462511,
+      "learning_rate": 9.076976581200616e-06,
+      "loss": 0.8536,
+      "step": 2053
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8187573154498395,
+      "learning_rate": 9.075968418794499e-06,
+      "loss": 0.9794,
+      "step": 2054
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7846316936584588,
+      "learning_rate": 9.074959762163158e-06,
+      "loss": 0.8393,
+      "step": 2055
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7213299987690351,
+      "learning_rate": 9.073950611428896e-06,
+      "loss": 0.9702,
+      "step": 2056
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8510982197896972,
+      "learning_rate": 9.07294096671408e-06,
+      "loss": 0.9566,
+      "step": 2057
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7188407516943063,
+      "learning_rate": 9.071930828141128e-06,
+      "loss": 0.893,
+      "step": 2058
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8063694802017927,
+      "learning_rate": 9.070920195832528e-06,
+      "loss": 0.8768,
+      "step": 2059
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7589722686088451,
+      "learning_rate": 9.069909069910817e-06,
+      "loss": 0.9648,
+      "step": 2060
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6220508022338267,
+      "learning_rate": 9.068897450498602e-06,
+      "loss": 0.9474,
+      "step": 2061
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9559677324400258,
+      "learning_rate": 9.067885337718542e-06,
+      "loss": 0.9136,
+      "step": 2062
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7432672554334079,
+      "learning_rate": 9.066872731693362e-06,
+      "loss": 0.8915,
+      "step": 2063
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9504174095244735,
+      "learning_rate": 9.065859632545842e-06,
+      "loss": 0.9169,
+      "step": 2064
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6984129844659754,
+      "learning_rate": 9.064846040398822e-06,
+      "loss": 0.9181,
+      "step": 2065
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.680348236582505,
+      "learning_rate": 9.06383195537521e-06,
+      "loss": 0.9209,
+      "step": 2066
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8328575974696761,
+      "learning_rate": 9.062817377597961e-06,
+      "loss": 0.9219,
+      "step": 2067
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.641925455201439,
+      "learning_rate": 9.061802307190099e-06,
+      "loss": 0.8707,
+      "step": 2068
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6851491869100046,
+      "learning_rate": 9.060786744274703e-06,
+      "loss": 0.887,
+      "step": 2069
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7619893985132553,
+      "learning_rate": 9.059770688974916e-06,
+      "loss": 0.9873,
+      "step": 2070
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8489254786450469,
+      "learning_rate": 9.058754141413936e-06,
+      "loss": 0.9401,
+      "step": 2071
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6614983260873726,
+      "learning_rate": 9.057737101715024e-06,
+      "loss": 0.9123,
+      "step": 2072
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.791029374629987,
+      "learning_rate": 9.056719570001499e-06,
+      "loss": 0.9578,
+      "step": 2073
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7554113422704898,
+      "learning_rate": 9.05570154639674e-06,
+      "loss": 0.8831,
+      "step": 2074
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6998267906889954,
+      "learning_rate": 9.054683031024188e-06,
+      "loss": 0.9713,
+      "step": 2075
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7483819430015746,
+      "learning_rate": 9.053664024007339e-06,
+      "loss": 0.9672,
+      "step": 2076
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7850456123159555,
+      "learning_rate": 9.052644525469751e-06,
+      "loss": 0.9354,
+      "step": 2077
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5983164920077244,
+      "learning_rate": 9.051624535535043e-06,
+      "loss": 0.9331,
+      "step": 2078
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6997536678097385,
+      "learning_rate": 9.050604054326893e-06,
+      "loss": 0.8764,
+      "step": 2079
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7968568874869546,
+      "learning_rate": 9.049583081969038e-06,
+      "loss": 0.9694,
+      "step": 2080
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.806067178908236,
+      "learning_rate": 9.04856161858527e-06,
+      "loss": 0.9116,
+      "step": 2081
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7534743600104805,
+      "learning_rate": 9.04753966429945e-06,
+      "loss": 0.885,
+      "step": 2082
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9586778850174237,
+      "learning_rate": 9.046517219235492e-06,
+      "loss": 1.07,
+      "step": 2083
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6958623745899681,
+      "learning_rate": 9.045494283517371e-06,
+      "loss": 0.9016,
+      "step": 2084
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6524888632441994,
+      "learning_rate": 9.044470857269121e-06,
+      "loss": 0.8771,
+      "step": 2085
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6904209334105549,
+      "learning_rate": 9.043446940614834e-06,
+      "loss": 0.9209,
+      "step": 2086
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7714665580818523,
+      "learning_rate": 9.042422533678667e-06,
+      "loss": 0.9213,
+      "step": 2087
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.936371201160908,
+      "learning_rate": 9.041397636584833e-06,
+      "loss": 0.9659,
+      "step": 2088
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7471584111185459,
+      "learning_rate": 9.0403722494576e-06,
+      "loss": 0.9888,
+      "step": 2089
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6799661517763989,
+      "learning_rate": 9.039346372421304e-06,
+      "loss": 0.9365,
+      "step": 2090
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0128783018313607,
+      "learning_rate": 9.038320005600335e-06,
+      "loss": 1.0722,
+      "step": 2091
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7459752076535546,
+      "learning_rate": 9.037293149119144e-06,
+      "loss": 0.9468,
+      "step": 2092
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.887321762699653,
+      "learning_rate": 9.036265803102238e-06,
+      "loss": 0.8913,
+      "step": 2093
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7741390027371502,
+      "learning_rate": 9.035237967674189e-06,
+      "loss": 0.9597,
+      "step": 2094
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7559332648108528,
+      "learning_rate": 9.034209642959624e-06,
+      "loss": 0.9659,
+      "step": 2095
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8585174170935415,
+      "learning_rate": 9.033180829083232e-06,
+      "loss": 0.9336,
+      "step": 2096
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7308241463426848,
+      "learning_rate": 9.032151526169761e-06,
+      "loss": 0.9714,
+      "step": 2097
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7173827351268125,
+      "learning_rate": 9.031121734344016e-06,
+      "loss": 0.8977,
+      "step": 2098
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6551805718104564,
+      "learning_rate": 9.030091453730863e-06,
+      "loss": 0.9542,
+      "step": 2099
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0841225339396494,
+      "learning_rate": 9.02906068445523e-06,
+      "loss": 1.0792,
+      "step": 2100
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9216218954203206,
+      "learning_rate": 9.028029426642095e-06,
+      "loss": 1.0552,
+      "step": 2101
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7474996409752583,
+      "learning_rate": 9.026997680416507e-06,
+      "loss": 1.002,
+      "step": 2102
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.759584802480323,
+      "learning_rate": 9.025965445903567e-06,
+      "loss": 0.9576,
+      "step": 2103
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8453118935501323,
+      "learning_rate": 9.024932723228436e-06,
+      "loss": 0.8601,
+      "step": 2104
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7879436207752712,
+      "learning_rate": 9.023899512516337e-06,
+      "loss": 0.9901,
+      "step": 2105
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7646454769840962,
+      "learning_rate": 9.02286581389255e-06,
+      "loss": 1.0398,
+      "step": 2106
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7243040240545167,
+      "learning_rate": 9.021831627482414e-06,
+      "loss": 0.9319,
+      "step": 2107
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7775521629013917,
+      "learning_rate": 9.020796953411328e-06,
+      "loss": 0.9728,
+      "step": 2108
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7250758034844178,
+      "learning_rate": 9.019761791804749e-06,
+      "loss": 0.9096,
+      "step": 2109
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7942744140564892,
+      "learning_rate": 9.018726142788195e-06,
+      "loss": 0.9436,
+      "step": 2110
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7832292171728314,
+      "learning_rate": 9.01769000648724e-06,
+      "loss": 1.0329,
+      "step": 2111
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7327836267277128,
+      "learning_rate": 9.016653383027522e-06,
+      "loss": 0.9155,
+      "step": 2112
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5968188961320082,
+      "learning_rate": 9.015616272534734e-06,
+      "loss": 1.0114,
+      "step": 2113
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7276855931267041,
+      "learning_rate": 9.014578675134629e-06,
+      "loss": 0.9979,
+      "step": 2114
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7234218679489934,
+      "learning_rate": 9.013540590953019e-06,
+      "loss": 0.8758,
+      "step": 2115
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6934949774508288,
+      "learning_rate": 9.012502020115777e-06,
+      "loss": 0.9696,
+      "step": 2116
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9363392076716504,
+      "learning_rate": 9.01146296274883e-06,
+      "loss": 0.9088,
+      "step": 2117
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.552152781994723,
+      "learning_rate": 9.01042341897817e-06,
+      "loss": 1.086,
+      "step": 2118
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7350917338251892,
+      "learning_rate": 9.009383388929842e-06,
+      "loss": 0.9481,
+      "step": 2119
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8608947230585158,
+      "learning_rate": 9.008342872729958e-06,
+      "loss": 0.9233,
+      "step": 2120
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9054266052237752,
+      "learning_rate": 9.007301870504681e-06,
+      "loss": 0.9518,
+      "step": 2121
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.810662070860422,
+      "learning_rate": 9.006260382380238e-06,
+      "loss": 0.9116,
+      "step": 2122
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8604225000856063,
+      "learning_rate": 9.005218408482912e-06,
+      "loss": 0.9395,
+      "step": 2123
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7441463572900258,
+      "learning_rate": 9.004175948939045e-06,
+      "loss": 0.9513,
+      "step": 2124
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7303479468702503,
+      "learning_rate": 9.003133003875039e-06,
+      "loss": 0.9418,
+      "step": 2125
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7903215283249656,
+      "learning_rate": 9.002089573417356e-06,
+      "loss": 0.9055,
+      "step": 2126
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.824790192114977,
+      "learning_rate": 9.001045657692517e-06,
+      "loss": 0.9084,
+      "step": 2127
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.860049981688993,
+      "learning_rate": 9.000001256827096e-06,
+      "loss": 0.9311,
+      "step": 2128
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7360003938096522,
+      "learning_rate": 8.998956370947733e-06,
+      "loss": 0.9355,
+      "step": 2129
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8818083172510638,
+      "learning_rate": 8.997911000181124e-06,
+      "loss": 0.9076,
+      "step": 2130
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8108231129664885,
+      "learning_rate": 8.996865144654023e-06,
+      "loss": 0.9495,
+      "step": 2131
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6721434721062056,
+      "learning_rate": 8.995818804493244e-06,
+      "loss": 0.9506,
+      "step": 2132
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.848871127699889,
+      "learning_rate": 8.994771979825658e-06,
+      "loss": 0.8994,
+      "step": 2133
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7026413944525383,
+      "learning_rate": 8.993724670778199e-06,
+      "loss": 0.9674,
+      "step": 2134
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7815576377227917,
+      "learning_rate": 8.992676877477855e-06,
+      "loss": 0.8743,
+      "step": 2135
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7753246405369383,
+      "learning_rate": 8.991628600051674e-06,
+      "loss": 0.8371,
+      "step": 2136
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7502814874125072,
+      "learning_rate": 8.990579838626764e-06,
+      "loss": 0.9025,
+      "step": 2137
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7896942210522537,
+      "learning_rate": 8.989530593330292e-06,
+      "loss": 0.8898,
+      "step": 2138
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7328745875524927,
+      "learning_rate": 8.988480864289482e-06,
+      "loss": 0.9533,
+      "step": 2139
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7507694595789036,
+      "learning_rate": 8.987430651631615e-06,
+      "loss": 0.8727,
+      "step": 2140
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8049676064187659,
+      "learning_rate": 8.986379955484036e-06,
+      "loss": 0.9354,
+      "step": 2141
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.255981423307433,
+      "learning_rate": 8.985328775974142e-06,
+      "loss": 0.8853,
+      "step": 2142
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.71076923993676,
+      "learning_rate": 8.984277113229398e-06,
+      "loss": 0.9674,
+      "step": 2143
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0173392267794297,
+      "learning_rate": 8.983224967377314e-06,
+      "loss": 1.0709,
+      "step": 2144
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.793846484961734,
+      "learning_rate": 8.982172338545474e-06,
+      "loss": 0.9486,
+      "step": 2145
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8252315401520471,
+      "learning_rate": 8.981119226861508e-06,
+      "loss": 0.8972,
+      "step": 2146
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.765182270604645,
+      "learning_rate": 8.980065632453111e-06,
+      "loss": 1.0373,
+      "step": 2147
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7745222321813605,
+      "learning_rate": 8.979011555448035e-06,
+      "loss": 1.0042,
+      "step": 2148
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9356399896124354,
+      "learning_rate": 8.97795699597409e-06,
+      "loss": 1.0646,
+      "step": 2149
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.740709576278291,
+      "learning_rate": 8.976901954159144e-06,
+      "loss": 0.9292,
+      "step": 2150
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.758730698911262,
+      "learning_rate": 8.975846430131127e-06,
+      "loss": 0.9706,
+      "step": 2151
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6656641973076396,
+      "learning_rate": 8.974790424018022e-06,
+      "loss": 0.9083,
+      "step": 2152
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.735611933034405,
+      "learning_rate": 8.973733935947877e-06,
+      "loss": 0.9401,
+      "step": 2153
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8150238137943122,
+      "learning_rate": 8.97267696604879e-06,
+      "loss": 0.9469,
+      "step": 2154
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7231737559138791,
+      "learning_rate": 8.971619514448928e-06,
+      "loss": 0.97,
+      "step": 2155
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.856552177901469,
+      "learning_rate": 8.970561581276506e-06,
+      "loss": 0.9753,
+      "step": 2156
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.623287975917867,
+      "learning_rate": 8.969503166659803e-06,
+      "loss": 0.941,
+      "step": 2157
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7123576570853698,
+      "learning_rate": 8.968444270727157e-06,
+      "loss": 0.94,
+      "step": 2158
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7712283067389945,
+      "learning_rate": 8.967384893606962e-06,
+      "loss": 0.9153,
+      "step": 2159
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9578754036238484,
+      "learning_rate": 8.96632503542767e-06,
+      "loss": 0.9497,
+      "step": 2160
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8453629765570332,
+      "learning_rate": 8.965264696317795e-06,
+      "loss": 0.938,
+      "step": 2161
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6434096741656588,
+      "learning_rate": 8.964203876405903e-06,
+      "loss": 0.9218,
+      "step": 2162
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.819508938774416,
+      "learning_rate": 8.963142575820626e-06,
+      "loss": 0.9825,
+      "step": 2163
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9039588724434431,
+      "learning_rate": 8.962080794690648e-06,
+      "loss": 1.0866,
+      "step": 2164
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8498672100002458,
+      "learning_rate": 8.961018533144716e-06,
+      "loss": 0.9592,
+      "step": 2165
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8033775233866292,
+      "learning_rate": 8.95995579131163e-06,
+      "loss": 0.9483,
+      "step": 2166
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8882628379740451,
+      "learning_rate": 8.95889256932025e-06,
+      "loss": 1.0734,
+      "step": 2167
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7831838646891607,
+      "learning_rate": 8.9578288672995e-06,
+      "loss": 0.891,
+      "step": 2168
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8119052984443256,
+      "learning_rate": 8.956764685378356e-06,
+      "loss": 0.9477,
+      "step": 2169
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.1028235003862004,
+      "learning_rate": 8.955700023685851e-06,
+      "loss": 0.8349,
+      "step": 2170
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7286313124458272,
+      "learning_rate": 8.95463488235108e-06,
+      "loss": 0.9291,
+      "step": 2171
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6443897611646197,
+      "learning_rate": 8.953569261503198e-06,
+      "loss": 0.8461,
+      "step": 2172
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0381723298649999,
+      "learning_rate": 8.952503161271413e-06,
+      "loss": 1.0719,
+      "step": 2173
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7971667244067868,
+      "learning_rate": 8.951436581784992e-06,
+      "loss": 0.9064,
+      "step": 2174
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7310115421921137,
+      "learning_rate": 8.950369523173263e-06,
+      "loss": 0.9677,
+      "step": 2175
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9047382301973255,
+      "learning_rate": 8.949301985565611e-06,
+      "loss": 0.9465,
+      "step": 2176
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8415883032187619,
+      "learning_rate": 8.948233969091478e-06,
+      "loss": 0.9414,
+      "step": 2177
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7537268845957987,
+      "learning_rate": 8.947165473880364e-06,
+      "loss": 0.9174,
+      "step": 2178
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8619044409735235,
+      "learning_rate": 8.946096500061828e-06,
+      "loss": 0.9995,
+      "step": 2179
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8880796159636029,
+      "learning_rate": 8.94502704776549e-06,
+      "loss": 1.005,
+      "step": 2180
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7544394320130334,
+      "learning_rate": 8.943957117121018e-06,
+      "loss": 0.851,
+      "step": 2181
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7558704694232716,
+      "learning_rate": 8.94288670825815e-06,
+      "loss": 0.8879,
+      "step": 2182
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.6765868586316195,
+      "learning_rate": 8.941815821306675e-06,
+      "loss": 0.9509,
+      "step": 2183
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6785842619919114,
+      "learning_rate": 8.940744456396444e-06,
+      "loss": 0.9352,
+      "step": 2184
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.883867740364058,
+      "learning_rate": 8.93967261365736e-06,
+      "loss": 0.9202,
+      "step": 2185
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7952090744667841,
+      "learning_rate": 8.93860029321939e-06,
+      "loss": 0.9989,
+      "step": 2186
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7271422275196493,
+      "learning_rate": 8.937527495212555e-06,
+      "loss": 0.9294,
+      "step": 2187
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7966281354314397,
+      "learning_rate": 8.93645421976694e-06,
+      "loss": 0.9011,
+      "step": 2188
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9856080334819769,
+      "learning_rate": 8.935380467012675e-06,
+      "loss": 1.0742,
+      "step": 2189
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9822700747764161,
+      "learning_rate": 8.934306237079963e-06,
+      "loss": 1.018,
+      "step": 2190
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8381191098744205,
+      "learning_rate": 8.933231530099057e-06,
+      "loss": 0.8594,
+      "step": 2191
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9040242930042737,
+      "learning_rate": 8.932156346200269e-06,
+      "loss": 0.9122,
+      "step": 2192
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7807746758322613,
+      "learning_rate": 8.931080685513966e-06,
+      "loss": 0.9001,
+      "step": 2193
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.80857488676619,
+      "learning_rate": 8.930004548170577e-06,
+      "loss": 0.9608,
+      "step": 2194
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6873435435612767,
+      "learning_rate": 8.92892793430059e-06,
+      "loss": 0.9221,
+      "step": 2195
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8172995397303782,
+      "learning_rate": 8.927850844034545e-06,
+      "loss": 0.9213,
+      "step": 2196
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8645279967908626,
+      "learning_rate": 8.926773277503042e-06,
+      "loss": 1.0087,
+      "step": 2197
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6585159467473778,
+      "learning_rate": 8.925695234836743e-06,
+      "loss": 0.8852,
+      "step": 2198
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.730985045790014,
+      "learning_rate": 8.924616716166363e-06,
+      "loss": 0.9172,
+      "step": 2199
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.11338070948821,
+      "learning_rate": 8.923537721622675e-06,
+      "loss": 0.9652,
+      "step": 2200
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7301506169110117,
+      "learning_rate": 8.922458251336512e-06,
+      "loss": 0.9122,
+      "step": 2201
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7745369243752658,
+      "learning_rate": 8.921378305438763e-06,
+      "loss": 1.0371,
+      "step": 2202
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.1350645871432334,
+      "learning_rate": 8.920297884060378e-06,
+      "loss": 0.9403,
+      "step": 2203
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0013453371618901,
+      "learning_rate": 8.919216987332357e-06,
+      "loss": 1.0655,
+      "step": 2204
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9107170376434253,
+      "learning_rate": 8.918135615385763e-06,
+      "loss": 0.9088,
+      "step": 2205
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8112727008068048,
+      "learning_rate": 8.917053768351719e-06,
+      "loss": 0.9997,
+      "step": 2206
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6455256574911037,
+      "learning_rate": 8.915971446361404e-06,
+      "loss": 0.9264,
+      "step": 2207
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6939464274422178,
+      "learning_rate": 8.914888649546049e-06,
+      "loss": 0.9247,
+      "step": 2208
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7985990296214152,
+      "learning_rate": 8.913805378036948e-06,
+      "loss": 0.9887,
+      "step": 2209
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7686874411135423,
+      "learning_rate": 8.912721631965454e-06,
+      "loss": 0.9194,
+      "step": 2210
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7477330536976754,
+      "learning_rate": 8.91163741146297e-06,
+      "loss": 0.8807,
+      "step": 2211
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6603041142877848,
+      "learning_rate": 8.910552716660966e-06,
+      "loss": 0.8424,
+      "step": 2212
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8243212030836031,
+      "learning_rate": 8.909467547690961e-06,
+      "loss": 0.9215,
+      "step": 2213
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7388751180644533,
+      "learning_rate": 8.908381904684543e-06,
+      "loss": 0.8704,
+      "step": 2214
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7711008724226498,
+      "learning_rate": 8.90729578777334e-06,
+      "loss": 0.9716,
+      "step": 2215
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.675477382868681,
+      "learning_rate": 8.906209197089055e-06,
+      "loss": 0.9167,
+      "step": 2216
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.874648272016351,
+      "learning_rate": 8.905122132763437e-06,
+      "loss": 0.9242,
+      "step": 2217
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7775010363714228,
+      "learning_rate": 8.904034594928296e-06,
+      "loss": 0.899,
+      "step": 2218
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7882546177916347,
+      "learning_rate": 8.902946583715503e-06,
+      "loss": 0.9617,
+      "step": 2219
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7242692666366477,
+      "learning_rate": 8.901858099256981e-06,
+      "loss": 0.9334,
+      "step": 2220
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6973173672750974,
+      "learning_rate": 8.900769141684714e-06,
+      "loss": 0.9146,
+      "step": 2221
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5744583914823165,
+      "learning_rate": 8.899679711130737e-06,
+      "loss": 0.8809,
+      "step": 2222
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8411325563423053,
+      "learning_rate": 8.898589807727154e-06,
+      "loss": 0.9371,
+      "step": 2223
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.793718776578932,
+      "learning_rate": 8.897499431606116e-06,
+      "loss": 0.9927,
+      "step": 2224
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.745469146363477,
+      "learning_rate": 8.896408582899835e-06,
+      "loss": 0.9432,
+      "step": 2225
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9943174017975462,
+      "learning_rate": 8.89531726174058e-06,
+      "loss": 0.9935,
+      "step": 2226
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8533089556410962,
+      "learning_rate": 8.894225468260676e-06,
+      "loss": 0.9283,
+      "step": 2227
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.737238992237244,
+      "learning_rate": 8.89313320259251e-06,
+      "loss": 0.9106,
+      "step": 2228
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6821878368741026,
+      "learning_rate": 8.89204046486852e-06,
+      "loss": 0.9088,
+      "step": 2229
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7721624805253644,
+      "learning_rate": 8.89094725522121e-06,
+      "loss": 0.945,
+      "step": 2230
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6789599450277066,
+      "learning_rate": 8.889853573783128e-06,
+      "loss": 0.9395,
+      "step": 2231
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9865132905899449,
+      "learning_rate": 8.88875942068689e-06,
+      "loss": 1.0907,
+      "step": 2232
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7707237563927818,
+      "learning_rate": 8.887664796065167e-06,
+      "loss": 1.007,
+      "step": 2233
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7327321338903077,
+      "learning_rate": 8.886569700050682e-06,
+      "loss": 0.9499,
+      "step": 2234
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6609941785317417,
+      "learning_rate": 8.885474132776223e-06,
+      "loss": 1.0144,
+      "step": 2235
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7971466485026215,
+      "learning_rate": 8.884378094374633e-06,
+      "loss": 0.9257,
+      "step": 2236
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6732007998062515,
+      "learning_rate": 8.883281584978805e-06,
+      "loss": 0.8977,
+      "step": 2237
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5950829934834445,
+      "learning_rate": 8.882184604721697e-06,
+      "loss": 0.8744,
+      "step": 2238
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.796367796687254,
+      "learning_rate": 8.88108715373632e-06,
+      "loss": 0.8753,
+      "step": 2239
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6761346694556256,
+      "learning_rate": 8.879989232155749e-06,
+      "loss": 0.8674,
+      "step": 2240
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6334798257469063,
+      "learning_rate": 8.878890840113106e-06,
+      "loss": 0.935,
+      "step": 2241
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6653832277270355,
+      "learning_rate": 8.877791977741575e-06,
+      "loss": 0.9177,
+      "step": 2242
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.630908451655982,
+      "learning_rate": 8.8766926451744e-06,
+      "loss": 0.9233,
+      "step": 2243
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.738640881366043,
+      "learning_rate": 8.875592842544875e-06,
+      "loss": 0.971,
+      "step": 2244
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7911522860601667,
+      "learning_rate": 8.874492569986357e-06,
+      "loss": 0.8742,
+      "step": 2245
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8520767902506714,
+      "learning_rate": 8.873391827632258e-06,
+      "loss": 0.8793,
+      "step": 2246
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7904641727825066,
+      "learning_rate": 8.872290615616045e-06,
+      "loss": 0.9416,
+      "step": 2247
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7471378974494611,
+      "learning_rate": 8.871188934071247e-06,
+      "loss": 1.0072,
+      "step": 2248
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7063788848449455,
+      "learning_rate": 8.870086783131443e-06,
+      "loss": 0.9087,
+      "step": 2249
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.865822919827851,
+      "learning_rate": 8.868984162930276e-06,
+      "loss": 0.8438,
+      "step": 2250
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7674532944386452,
+      "learning_rate": 8.86788107360144e-06,
+      "loss": 0.8918,
+      "step": 2251
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7627097684871929,
+      "learning_rate": 8.866777515278688e-06,
+      "loss": 0.9112,
+      "step": 2252
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7430478952958695,
+      "learning_rate": 8.865673488095833e-06,
+      "loss": 0.9864,
+      "step": 2253
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6258436351749288,
+      "learning_rate": 8.86456899218674e-06,
+      "loss": 0.9776,
+      "step": 2254
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8216854377307679,
+      "learning_rate": 8.863464027685333e-06,
+      "loss": 0.857,
+      "step": 2255
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.710743894586576,
+      "learning_rate": 8.862358594725595e-06,
+      "loss": 0.9913,
+      "step": 2256
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7745092078239628,
+      "learning_rate": 8.86125269344156e-06,
+      "loss": 0.8389,
+      "step": 2257
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8375142355389107,
+      "learning_rate": 8.860146323967325e-06,
+      "loss": 0.9403,
+      "step": 2258
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7574870962597215,
+      "learning_rate": 8.85903948643704e-06,
+      "loss": 0.9456,
+      "step": 2259
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.710870843020024,
+      "learning_rate": 8.857932180984914e-06,
+      "loss": 0.9384,
+      "step": 2260
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8836820207829077,
+      "learning_rate": 8.85682440774521e-06,
+      "loss": 0.9122,
+      "step": 2261
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8551566865436915,
+      "learning_rate": 8.855716166852249e-06,
+      "loss": 0.9094,
+      "step": 2262
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8092320466854968,
+      "learning_rate": 8.854607458440412e-06,
+      "loss": 0.8936,
+      "step": 2263
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7866078645718164,
+      "learning_rate": 8.85349828264413e-06,
+      "loss": 0.8796,
+      "step": 2264
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6754175410174497,
+      "learning_rate": 8.852388639597898e-06,
+      "loss": 0.9656,
+      "step": 2265
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8444137830722658,
+      "learning_rate": 8.85127852943626e-06,
+      "loss": 0.9579,
+      "step": 2266
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8515803788838923,
+      "learning_rate": 8.850167952293825e-06,
+      "loss": 0.9207,
+      "step": 2267
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8050500412271848,
+      "learning_rate": 8.849056908305252e-06,
+      "loss": 0.9088,
+      "step": 2268
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9557694686164702,
+      "learning_rate": 8.847945397605258e-06,
+      "loss": 0.9472,
+      "step": 2269
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7012857790340064,
+      "learning_rate": 8.84683342032862e-06,
+      "loss": 0.8648,
+      "step": 2270
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7057368969309457,
+      "learning_rate": 8.845720976610168e-06,
+      "loss": 0.8667,
+      "step": 2271
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8645723851443787,
+      "learning_rate": 8.844608066584788e-06,
+      "loss": 0.8871,
+      "step": 2272
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6654056404436466,
+      "learning_rate": 8.843494690387427e-06,
+      "loss": 0.9505,
+      "step": 2273
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5977966249949151,
+      "learning_rate": 8.842380848153083e-06,
+      "loss": 0.8786,
+      "step": 2274
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.840432615825486,
+      "learning_rate": 8.841266540016814e-06,
+      "loss": 0.983,
+      "step": 2275
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.82319319116273,
+      "learning_rate": 8.840151766113735e-06,
+      "loss": 0.9379,
+      "step": 2276
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6957639057599092,
+      "learning_rate": 8.839036526579014e-06,
+      "loss": 0.9221,
+      "step": 2277
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.724156847600573,
+      "learning_rate": 8.837920821547881e-06,
+      "loss": 0.8796,
+      "step": 2278
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.700005610858219,
+      "learning_rate": 8.836804651155617e-06,
+      "loss": 0.8831,
+      "step": 2279
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7041111868199266,
+      "learning_rate": 8.83568801553756e-06,
+      "loss": 0.8353,
+      "step": 2280
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9304349656851505,
+      "learning_rate": 8.834570914829108e-06,
+      "loss": 0.9357,
+      "step": 2281
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.163751313698539,
+      "learning_rate": 8.833453349165713e-06,
+      "loss": 0.9612,
+      "step": 2282
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7484077900047175,
+      "learning_rate": 8.832335318682884e-06,
+      "loss": 0.9553,
+      "step": 2283
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9779020881919162,
+      "learning_rate": 8.831216823516185e-06,
+      "loss": 1.0956,
+      "step": 2284
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7345772492850835,
+      "learning_rate": 8.830097863801239e-06,
+      "loss": 0.9192,
+      "step": 2285
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.960960288587253,
+      "learning_rate": 8.828978439673721e-06,
+      "loss": 0.946,
+      "step": 2286
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7716092125926726,
+      "learning_rate": 8.827858551269368e-06,
+      "loss": 0.928,
+      "step": 2287
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6934791683510433,
+      "learning_rate": 8.826738198723967e-06,
+      "loss": 0.9356,
+      "step": 2288
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7063578381591806,
+      "learning_rate": 8.825617382173369e-06,
+      "loss": 0.9722,
+      "step": 2289
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6723775729772732,
+      "learning_rate": 8.824496101753474e-06,
+      "loss": 0.8904,
+      "step": 2290
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8696009597020529,
+      "learning_rate": 8.82337435760024e-06,
+      "loss": 0.9402,
+      "step": 2291
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7405233584444173,
+      "learning_rate": 8.822252149849687e-06,
+      "loss": 0.8992,
+      "step": 2292
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8417353069930344,
+      "learning_rate": 8.821129478637882e-06,
+      "loss": 0.8839,
+      "step": 2293
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7516243429425546,
+      "learning_rate": 8.820006344100953e-06,
+      "loss": 0.9211,
+      "step": 2294
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7762148214051394,
+      "learning_rate": 8.818882746375086e-06,
+      "loss": 0.8858,
+      "step": 2295
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9219412158292026,
+      "learning_rate": 8.81775868559652e-06,
+      "loss": 0.898,
+      "step": 2296
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9138765907741295,
+      "learning_rate": 8.816634161901553e-06,
+      "loss": 0.8745,
+      "step": 2297
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.735940231667222,
+      "learning_rate": 8.815509175426536e-06,
+      "loss": 0.8589,
+      "step": 2298
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8042703679364345,
+      "learning_rate": 8.814383726307876e-06,
+      "loss": 0.9251,
+      "step": 2299
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.819621580351289,
+      "learning_rate": 8.81325781468204e-06,
+      "loss": 0.8698,
+      "step": 2300
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6657412622335976,
+      "learning_rate": 8.812131440685545e-06,
+      "loss": 0.9488,
+      "step": 2301
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7417957221424745,
+      "learning_rate": 8.811004604454973e-06,
+      "loss": 0.8801,
+      "step": 2302
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7855652671323816,
+      "learning_rate": 8.809877306126954e-06,
+      "loss": 0.9496,
+      "step": 2303
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7373590624228923,
+      "learning_rate": 8.808749545838177e-06,
+      "loss": 0.8986,
+      "step": 2304
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6973826252175284,
+      "learning_rate": 8.807621323725387e-06,
+      "loss": 0.8971,
+      "step": 2305
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7037311780425466,
+      "learning_rate": 8.806492639925383e-06,
+      "loss": 0.8962,
+      "step": 2306
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8521089654363712,
+      "learning_rate": 8.805363494575026e-06,
+      "loss": 0.9171,
+      "step": 2307
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7426144935249974,
+      "learning_rate": 8.804233887811224e-06,
+      "loss": 1.0505,
+      "step": 2308
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7186398819920323,
+      "learning_rate": 8.803103819770948e-06,
+      "loss": 0.9894,
+      "step": 2309
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.86370773412969,
+      "learning_rate": 8.801973290591223e-06,
+      "loss": 0.9554,
+      "step": 2310
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.101021544905293,
+      "learning_rate": 8.800842300409129e-06,
+      "loss": 1.0241,
+      "step": 2311
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9089690733672087,
+      "learning_rate": 8.799710849361804e-06,
+      "loss": 1.0643,
+      "step": 2312
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.690893662155264,
+      "learning_rate": 8.798578937586437e-06,
+      "loss": 0.9375,
+      "step": 2313
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8019177766880603,
+      "learning_rate": 8.797446565220279e-06,
+      "loss": 0.9265,
+      "step": 2314
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1378893770859821,
+      "learning_rate": 8.796313732400634e-06,
+      "loss": 1.0682,
+      "step": 2315
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8797171241528554,
+      "learning_rate": 8.79518043926486e-06,
+      "loss": 0.9612,
+      "step": 2316
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7808559573716805,
+      "learning_rate": 8.794046685950374e-06,
+      "loss": 0.9891,
+      "step": 2317
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7792274347163295,
+      "learning_rate": 8.792912472594646e-06,
+      "loss": 0.8401,
+      "step": 2318
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7220294707984976,
+      "learning_rate": 8.791777799335206e-06,
+      "loss": 0.9237,
+      "step": 2319
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.824346193633782,
+      "learning_rate": 8.790642666309637e-06,
+      "loss": 0.8912,
+      "step": 2320
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7348168957884444,
+      "learning_rate": 8.789507073655575e-06,
+      "loss": 0.8449,
+      "step": 2321
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8747735352943342,
+      "learning_rate": 8.788371021510713e-06,
+      "loss": 0.9125,
+      "step": 2322
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8443540259073414,
+      "learning_rate": 8.787234510012808e-06,
+      "loss": 0.9179,
+      "step": 2323
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6345214924799176,
+      "learning_rate": 8.786097539299661e-06,
+      "loss": 0.9267,
+      "step": 2324
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7255855353514613,
+      "learning_rate": 8.784960109509134e-06,
+      "loss": 0.9207,
+      "step": 2325
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0979906977760512,
+      "learning_rate": 8.783822220779145e-06,
+      "loss": 1.0681,
+      "step": 2326
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7348962102969565,
+      "learning_rate": 8.782683873247668e-06,
+      "loss": 0.9084,
+      "step": 2327
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8441644700609707,
+      "learning_rate": 8.781545067052731e-06,
+      "loss": 0.987,
+      "step": 2328
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8384333832232267,
+      "learning_rate": 8.780405802332415e-06,
+      "loss": 1.0594,
+      "step": 2329
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7279788907648446,
+      "learning_rate": 8.779266079224865e-06,
+      "loss": 0.8792,
+      "step": 2330
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.880889557595136,
+      "learning_rate": 8.778125897868273e-06,
+      "loss": 0.9369,
+      "step": 2331
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.770534084427665,
+      "learning_rate": 8.77698525840089e-06,
+      "loss": 0.9948,
+      "step": 2332
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.818618539897555,
+      "learning_rate": 8.775844160961024e-06,
+      "loss": 1.0091,
+      "step": 2333
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.817389357221285,
+      "learning_rate": 8.774702605687036e-06,
+      "loss": 0.9243,
+      "step": 2334
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8010495590084818,
+      "learning_rate": 8.773560592717343e-06,
+      "loss": 0.9478,
+      "step": 2335
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7394516123365835,
+      "learning_rate": 8.772418122190418e-06,
+      "loss": 1.001,
+      "step": 2336
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.711819195945204,
+      "learning_rate": 8.771275194244792e-06,
+      "loss": 0.9386,
+      "step": 2337
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7259075524906125,
+      "learning_rate": 8.770131809019046e-06,
+      "loss": 0.9906,
+      "step": 2338
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7382129719012371,
+      "learning_rate": 8.768987966651822e-06,
+      "loss": 0.9466,
+      "step": 2339
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8146802931959587,
+      "learning_rate": 8.767843667281812e-06,
+      "loss": 0.9404,
+      "step": 2340
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7317072043893071,
+      "learning_rate": 8.766698911047769e-06,
+      "loss": 0.9411,
+      "step": 2341
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7813389007301232,
+      "learning_rate": 8.765553698088496e-06,
+      "loss": 0.9892,
+      "step": 2342
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7206874526220564,
+      "learning_rate": 8.764408028542854e-06,
+      "loss": 0.8401,
+      "step": 2343
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8952718015961223,
+      "learning_rate": 8.763261902549763e-06,
+      "loss": 0.938,
+      "step": 2344
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7574914750095616,
+      "learning_rate": 8.762115320248192e-06,
+      "loss": 1.0005,
+      "step": 2345
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8309393443195099,
+      "learning_rate": 8.760968281777167e-06,
+      "loss": 0.936,
+      "step": 2346
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7996672331657386,
+      "learning_rate": 8.759820787275774e-06,
+      "loss": 0.9611,
+      "step": 2347
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8383364494066772,
+      "learning_rate": 8.758672836883146e-06,
+      "loss": 1.0341,
+      "step": 2348
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7977294833673378,
+      "learning_rate": 8.75752443073848e-06,
+      "loss": 0.8896,
+      "step": 2349
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8740986537909834,
+      "learning_rate": 8.756375568981023e-06,
+      "loss": 0.9897,
+      "step": 2350
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7708249750936327,
+      "learning_rate": 8.755226251750077e-06,
+      "loss": 0.8813,
+      "step": 2351
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6598312294614797,
+      "learning_rate": 8.754076479185003e-06,
+      "loss": 0.952,
+      "step": 2352
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8258352363325507,
+      "learning_rate": 8.752926251425212e-06,
+      "loss": 0.8532,
+      "step": 2353
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6479451241116814,
+      "learning_rate": 8.751775568610175e-06,
+      "loss": 0.8961,
+      "step": 2354
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7806631689790196,
+      "learning_rate": 8.750624430879417e-06,
+      "loss": 1.0017,
+      "step": 2355
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8487523174115212,
+      "learning_rate": 8.749472838372515e-06,
+      "loss": 0.9091,
+      "step": 2356
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6742383741062217,
+      "learning_rate": 8.748320791229106e-06,
+      "loss": 0.8938,
+      "step": 2357
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.67708277650289,
+      "learning_rate": 8.747168289588878e-06,
+      "loss": 1.0039,
+      "step": 2358
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6917890812979146,
+      "learning_rate": 8.746015333591578e-06,
+      "loss": 0.927,
+      "step": 2359
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.832077258992601,
+      "learning_rate": 8.744861923377e-06,
+      "loss": 0.8814,
+      "step": 2360
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7527850331884436,
+      "learning_rate": 8.74370805908501e-06,
+      "loss": 0.9525,
+      "step": 2361
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.635377042651314,
+      "learning_rate": 8.742553740855507e-06,
+      "loss": 0.917,
+      "step": 2362
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7561452525927792,
+      "learning_rate": 8.74139896882846e-06,
+      "loss": 0.8933,
+      "step": 2363
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7629114708971814,
+      "learning_rate": 8.74024374314389e-06,
+      "loss": 0.8581,
+      "step": 2364
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7560154093972253,
+      "learning_rate": 8.739088063941875e-06,
+      "loss": 1.0191,
+      "step": 2365
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.663945976342402,
+      "learning_rate": 8.737931931362537e-06,
+      "loss": 0.8691,
+      "step": 2366
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.809476637438118,
+      "learning_rate": 8.736775345546067e-06,
+      "loss": 0.9411,
+      "step": 2367
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.865570988133656,
+      "learning_rate": 8.735618306632706e-06,
+      "loss": 0.94,
+      "step": 2368
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.775128519795085,
+      "learning_rate": 8.734460814762743e-06,
+      "loss": 0.8632,
+      "step": 2369
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8551373112201106,
+      "learning_rate": 8.733302870076535e-06,
+      "loss": 0.9393,
+      "step": 2370
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7972050672680964,
+      "learning_rate": 8.732144472714481e-06,
+      "loss": 0.9851,
+      "step": 2371
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7491128139640266,
+      "learning_rate": 8.730985622817043e-06,
+      "loss": 0.8879,
+      "step": 2372
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.297385513623497,
+      "learning_rate": 8.729826320524737e-06,
+      "loss": 1.0685,
+      "step": 2373
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.714010666604908,
+      "learning_rate": 8.72866656597813e-06,
+      "loss": 0.959,
+      "step": 2374
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9010578853874043,
+      "learning_rate": 8.727506359317848e-06,
+      "loss": 0.8628,
+      "step": 2375
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8073689963075528,
+      "learning_rate": 8.72634570068457e-06,
+      "loss": 0.9923,
+      "step": 2376
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6556452906935304,
+      "learning_rate": 8.725184590219028e-06,
+      "loss": 0.9122,
+      "step": 2377
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7001238568006234,
+      "learning_rate": 8.724023028062011e-06,
+      "loss": 0.8997,
+      "step": 2378
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7823354709004577,
+      "learning_rate": 8.722861014354363e-06,
+      "loss": 0.9451,
+      "step": 2379
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7115430386256802,
+      "learning_rate": 8.721698549236982e-06,
+      "loss": 0.8489,
+      "step": 2380
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9572711424428628,
+      "learning_rate": 8.720535632850824e-06,
+      "loss": 0.8814,
+      "step": 2381
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6934963814752633,
+      "learning_rate": 8.719372265336893e-06,
+      "loss": 0.9849,
+      "step": 2382
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6763720959230195,
+      "learning_rate": 8.718208446836252e-06,
+      "loss": 0.912,
+      "step": 2383
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8438502631010527,
+      "learning_rate": 8.717044177490016e-06,
+      "loss": 0.9256,
+      "step": 2384
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7492248495930325,
+      "learning_rate": 8.715879457439362e-06,
+      "loss": 0.9181,
+      "step": 2385
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7102807679431347,
+      "learning_rate": 8.714714286825512e-06,
+      "loss": 0.8908,
+      "step": 2386
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7879939358125525,
+      "learning_rate": 8.713548665789748e-06,
+      "loss": 0.9611,
+      "step": 2387
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.646028710925101,
+      "learning_rate": 8.712382594473405e-06,
+      "loss": 0.8294,
+      "step": 2388
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.2280230785073463,
+      "learning_rate": 8.711216073017875e-06,
+      "loss": 0.9631,
+      "step": 2389
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8447117162571527,
+      "learning_rate": 8.710049101564602e-06,
+      "loss": 0.9885,
+      "step": 2390
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6017146206312893,
+      "learning_rate": 8.708881680255084e-06,
+      "loss": 0.9062,
+      "step": 2391
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7302911117537718,
+      "learning_rate": 8.707713809230875e-06,
+      "loss": 0.9538,
+      "step": 2392
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1646342553194156,
+      "learning_rate": 8.706545488633587e-06,
+      "loss": 1.1047,
+      "step": 2393
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7444288806812078,
+      "learning_rate": 8.705376718604877e-06,
+      "loss": 0.9527,
+      "step": 2394
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7061999019572438,
+      "learning_rate": 8.704207499286467e-06,
+      "loss": 0.9304,
+      "step": 2395
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7700908359470797,
+      "learning_rate": 8.703037830820128e-06,
+      "loss": 0.9077,
+      "step": 2396
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7387988253035824,
+      "learning_rate": 8.701867713347686e-06,
+      "loss": 0.9201,
+      "step": 2397
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6939284969817738,
+      "learning_rate": 8.70069714701102e-06,
+      "loss": 0.885,
+      "step": 2398
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7729418488467943,
+      "learning_rate": 8.699526131952065e-06,
+      "loss": 0.9034,
+      "step": 2399
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.927350897791296,
+      "learning_rate": 8.698354668312816e-06,
+      "loss": 1.0159,
+      "step": 2400
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7415217059364518,
+      "learning_rate": 8.697182756235311e-06,
+      "loss": 0.9468,
+      "step": 2401
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6778799843881669,
+      "learning_rate": 8.696010395861651e-06,
+      "loss": 0.9927,
+      "step": 2402
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6326139638711061,
+      "learning_rate": 8.694837587333989e-06,
+      "loss": 0.8399,
+      "step": 2403
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6936960579103433,
+      "learning_rate": 8.69366433079453e-06,
+      "loss": 0.9573,
+      "step": 2404
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8013399609731544,
+      "learning_rate": 8.69249062638554e-06,
+      "loss": 0.9319,
+      "step": 2405
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6746773493408487,
+      "learning_rate": 8.691316474249329e-06,
+      "loss": 0.8991,
+      "step": 2406
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5792474222323174,
+      "learning_rate": 8.69014187452827e-06,
+      "loss": 0.9551,
+      "step": 2407
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.760524958958359,
+      "learning_rate": 8.688966827364788e-06,
+      "loss": 0.9171,
+      "step": 2408
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.789919512915395,
+      "learning_rate": 8.68779133290136e-06,
+      "loss": 0.961,
+      "step": 2409
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7480722007484824,
+      "learning_rate": 8.686615391280519e-06,
+      "loss": 0.8728,
+      "step": 2410
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9431781705892988,
+      "learning_rate": 8.68543900264485e-06,
+      "loss": 1.1135,
+      "step": 2411
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7175193055801685,
+      "learning_rate": 8.684262167136999e-06,
+      "loss": 0.9835,
+      "step": 2412
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8171572262563251,
+      "learning_rate": 8.683084884899658e-06,
+      "loss": 0.9108,
+      "step": 2413
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7062935530704395,
+      "learning_rate": 8.681907156075578e-06,
+      "loss": 0.9795,
+      "step": 2414
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9111796229774067,
+      "learning_rate": 8.68072898080756e-06,
+      "loss": 1.0238,
+      "step": 2415
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7438556891838364,
+      "learning_rate": 8.679550359238464e-06,
+      "loss": 0.8529,
+      "step": 2416
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7239260046899318,
+      "learning_rate": 8.678371291511202e-06,
+      "loss": 0.8956,
+      "step": 2417
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.851184388358164,
+      "learning_rate": 8.67719177776874e-06,
+      "loss": 0.8644,
+      "step": 2418
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.763477375099109,
+      "learning_rate": 8.676011818154097e-06,
+      "loss": 0.9083,
+      "step": 2419
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7367579075925281,
+      "learning_rate": 8.67483141281035e-06,
+      "loss": 0.9629,
+      "step": 2420
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7886986941192753,
+      "learning_rate": 8.673650561880621e-06,
+      "loss": 0.9144,
+      "step": 2421
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7694078922774432,
+      "learning_rate": 8.6724692655081e-06,
+      "loss": 0.8746,
+      "step": 2422
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7176793700985777,
+      "learning_rate": 8.671287523836019e-06,
+      "loss": 0.8601,
+      "step": 2423
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7498294114916535,
+      "learning_rate": 8.670105337007668e-06,
+      "loss": 0.8916,
+      "step": 2424
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8011530259450104,
+      "learning_rate": 8.668922705166392e-06,
+      "loss": 0.9213,
+      "step": 2425
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7455938948884329,
+      "learning_rate": 8.667739628455592e-06,
+      "loss": 0.8951,
+      "step": 2426
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8462481157794397,
+      "learning_rate": 8.666556107018714e-06,
+      "loss": 0.8605,
+      "step": 2427
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6585592695060376,
+      "learning_rate": 8.665372140999268e-06,
+      "loss": 0.936,
+      "step": 2428
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7887363994059737,
+      "learning_rate": 8.664187730540815e-06,
+      "loss": 0.9371,
+      "step": 2429
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8361394118161676,
+      "learning_rate": 8.663002875786965e-06,
+      "loss": 0.8979,
+      "step": 2430
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8026266836105669,
+      "learning_rate": 8.661817576881391e-06,
+      "loss": 0.9943,
+      "step": 2431
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9391349801098203,
+      "learning_rate": 8.660631833967809e-06,
+      "loss": 0.9364,
+      "step": 2432
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7429623059001227,
+      "learning_rate": 8.65944564719e-06,
+      "loss": 0.8566,
+      "step": 2433
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.686506180334736,
+      "learning_rate": 8.658259016691786e-06,
+      "loss": 0.9001,
+      "step": 2434
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9811582285349567,
+      "learning_rate": 8.657071942617057e-06,
+      "loss": 1.0456,
+      "step": 2435
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6710423876887421,
+      "learning_rate": 8.655884425109747e-06,
+      "loss": 0.8858,
+      "step": 2436
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.612682423384455,
+      "learning_rate": 8.654696464313847e-06,
+      "loss": 0.9394,
+      "step": 2437
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6892512202744252,
+      "learning_rate": 8.6535080603734e-06,
+      "loss": 0.8818,
+      "step": 2438
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7457146443398053,
+      "learning_rate": 8.652319213432505e-06,
+      "loss": 0.9272,
+      "step": 2439
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7850855460225186,
+      "learning_rate": 8.651129923635316e-06,
+      "loss": 0.9046,
+      "step": 2440
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8276663082594697,
+      "learning_rate": 8.649940191126033e-06,
+      "loss": 0.9186,
+      "step": 2441
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8439766995236482,
+      "learning_rate": 8.64875001604892e-06,
+      "loss": 0.9325,
+      "step": 2442
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7460498939797564,
+      "learning_rate": 8.647559398548292e-06,
+      "loss": 0.8839,
+      "step": 2443
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7622776491958172,
+      "learning_rate": 8.646368338768508e-06,
+      "loss": 0.9733,
+      "step": 2444
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6988092882535963,
+      "learning_rate": 8.645176836853992e-06,
+      "loss": 1.0066,
+      "step": 2445
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8219745254671376,
+      "learning_rate": 8.643984892949218e-06,
+      "loss": 0.9651,
+      "step": 2446
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7165600497102744,
+      "learning_rate": 8.642792507198714e-06,
+      "loss": 0.9566,
+      "step": 2447
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8769730299455667,
+      "learning_rate": 8.64159967974706e-06,
+      "loss": 0.9096,
+      "step": 2448
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8225456688865855,
+      "learning_rate": 8.64040641073889e-06,
+      "loss": 0.9469,
+      "step": 2449
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7023033948943453,
+      "learning_rate": 8.63921270031889e-06,
+      "loss": 0.9382,
+      "step": 2450
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7620911993920305,
+      "learning_rate": 8.638018548631805e-06,
+      "loss": 0.941,
+      "step": 2451
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.825480318983913,
+      "learning_rate": 8.636823955822432e-06,
+      "loss": 0.9322,
+      "step": 2452
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7925792214649459,
+      "learning_rate": 8.635628922035613e-06,
+      "loss": 0.9164,
+      "step": 2453
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6741443424863673,
+      "learning_rate": 8.634433447416254e-06,
+      "loss": 0.905,
+      "step": 2454
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9030140317010513,
+      "learning_rate": 8.63323753210931e-06,
+      "loss": 0.9305,
+      "step": 2455
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7988515425464195,
+      "learning_rate": 8.63204117625979e-06,
+      "loss": 0.9945,
+      "step": 2456
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.776995206313872,
+      "learning_rate": 8.630844380012756e-06,
+      "loss": 0.8569,
+      "step": 2457
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8373709401308416,
+      "learning_rate": 8.629647143513322e-06,
+      "loss": 0.9302,
+      "step": 2458
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6393368507149435,
+      "learning_rate": 8.628449466906659e-06,
+      "loss": 0.8511,
+      "step": 2459
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7752240529930936,
+      "learning_rate": 8.627251350337988e-06,
+      "loss": 0.9113,
+      "step": 2460
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7078088782028154,
+      "learning_rate": 8.626052793952588e-06,
+      "loss": 0.921,
+      "step": 2461
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9985290487041949,
+      "learning_rate": 8.624853797895784e-06,
+      "loss": 1.0759,
+      "step": 2462
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.803737083695028,
+      "learning_rate": 8.623654362312962e-06,
+      "loss": 0.9,
+      "step": 2463
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6602419193079645,
+      "learning_rate": 8.622454487349556e-06,
+      "loss": 0.8751,
+      "step": 2464
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8216009125841937,
+      "learning_rate": 8.621254173151052e-06,
+      "loss": 0.8768,
+      "step": 2465
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6428471822120607,
+      "learning_rate": 8.620053419862997e-06,
+      "loss": 0.8774,
+      "step": 2466
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7330307936764666,
+      "learning_rate": 8.618852227630986e-06,
+      "loss": 0.9113,
+      "step": 2467
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7681556544352635,
+      "learning_rate": 8.617650596600665e-06,
+      "loss": 0.9404,
+      "step": 2468
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6500920082390649,
+      "learning_rate": 8.616448526917737e-06,
+      "loss": 0.9437,
+      "step": 2469
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.65131610204214,
+      "learning_rate": 8.615246018727957e-06,
+      "loss": 0.9913,
+      "step": 2470
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8256431686893348,
+      "learning_rate": 8.614043072177134e-06,
+      "loss": 0.8844,
+      "step": 2471
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8814306174777384,
+      "learning_rate": 8.61283968741113e-06,
+      "loss": 0.9532,
+      "step": 2472
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.0197438215628654,
+      "learning_rate": 8.611635864575858e-06,
+      "loss": 0.9209,
+      "step": 2473
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7703169583836382,
+      "learning_rate": 8.610431603817287e-06,
+      "loss": 0.9773,
+      "step": 2474
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7188444903311781,
+      "learning_rate": 8.609226905281434e-06,
+      "loss": 0.9365,
+      "step": 2475
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.819938953155006,
+      "learning_rate": 8.608021769114378e-06,
+      "loss": 0.989,
+      "step": 2476
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9910546052503112,
+      "learning_rate": 8.606816195462244e-06,
+      "loss": 1.0919,
+      "step": 2477
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7452015374875418,
+      "learning_rate": 8.605610184471211e-06,
+      "loss": 0.8936,
+      "step": 2478
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.1070025151721157,
+      "learning_rate": 8.604403736287513e-06,
+      "loss": 0.9244,
+      "step": 2479
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8328090029537525,
+      "learning_rate": 8.603196851057435e-06,
+      "loss": 0.9097,
+      "step": 2480
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.804737255921941,
+      "learning_rate": 8.601989528927317e-06,
+      "loss": 0.9181,
+      "step": 2481
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.745638928341168,
+      "learning_rate": 8.600781770043552e-06,
+      "loss": 0.9713,
+      "step": 2482
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6144904681570982,
+      "learning_rate": 8.599573574552581e-06,
+      "loss": 0.9593,
+      "step": 2483
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7376525515275763,
+      "learning_rate": 8.598364942600907e-06,
+      "loss": 0.9063,
+      "step": 2484
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6888803369834962,
+      "learning_rate": 8.597155874335077e-06,
+      "loss": 0.9492,
+      "step": 2485
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5743218824327403,
+      "learning_rate": 8.595946369901696e-06,
+      "loss": 1.0023,
+      "step": 2486
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7296140973810867,
+      "learning_rate": 8.594736429447421e-06,
+      "loss": 0.9606,
+      "step": 2487
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.69745796803244,
+      "learning_rate": 8.593526053118962e-06,
+      "loss": 0.8518,
+      "step": 2488
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.721398944632299,
+      "learning_rate": 8.592315241063078e-06,
+      "loss": 0.9289,
+      "step": 2489
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7899866306878855,
+      "learning_rate": 8.591103993426588e-06,
+      "loss": 0.8511,
+      "step": 2490
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6502318137308107,
+      "learning_rate": 8.589892310356359e-06,
+      "loss": 0.9624,
+      "step": 2491
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.729151913684393,
+      "learning_rate": 8.588680191999308e-06,
+      "loss": 0.9343,
+      "step": 2492
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7524639362629248,
+      "learning_rate": 8.587467638502414e-06,
+      "loss": 0.9633,
+      "step": 2493
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7146310600428751,
+      "learning_rate": 8.5862546500127e-06,
+      "loss": 0.8997,
+      "step": 2494
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7449427125454655,
+      "learning_rate": 8.585041226677247e-06,
+      "loss": 0.9698,
+      "step": 2495
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6653126773387807,
+      "learning_rate": 8.583827368643185e-06,
+      "loss": 0.9545,
+      "step": 2496
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7528122431961486,
+      "learning_rate": 8.5826130760577e-06,
+      "loss": 0.9571,
+      "step": 2497
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7761618676346362,
+      "learning_rate": 8.581398349068029e-06,
+      "loss": 0.8903,
+      "step": 2498
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7683761331634082,
+      "learning_rate": 8.58018318782146e-06,
+      "loss": 0.924,
+      "step": 2499
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7588087591206654,
+      "learning_rate": 8.578967592465337e-06,
+      "loss": 0.9075,
+      "step": 2500
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9804192363450144,
+      "learning_rate": 8.577751563147055e-06,
+      "loss": 1.0642,
+      "step": 2501
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8547548659192321,
+      "learning_rate": 8.576535100014061e-06,
+      "loss": 1.0702,
+      "step": 2502
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6639268978327129,
+      "learning_rate": 8.575318203213856e-06,
+      "loss": 0.9383,
+      "step": 2503
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7278945023287724,
+      "learning_rate": 8.574100872893993e-06,
+      "loss": 1.0086,
+      "step": 2504
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6940903583135796,
+      "learning_rate": 8.572883109202077e-06,
+      "loss": 0.9676,
+      "step": 2505
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9869919268163219,
+      "learning_rate": 8.571664912285766e-06,
+      "loss": 1.0709,
+      "step": 2506
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8090535589569063,
+      "learning_rate": 8.570446282292772e-06,
+      "loss": 0.9351,
+      "step": 2507
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8912921598615418,
+      "learning_rate": 8.569227219370857e-06,
+      "loss": 0.8556,
+      "step": 2508
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7051853885029447,
+      "learning_rate": 8.568007723667837e-06,
+      "loss": 0.9056,
+      "step": 2509
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8980951280342635,
+      "learning_rate": 8.56678779533158e-06,
+      "loss": 0.9061,
+      "step": 2510
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7708773045604274,
+      "learning_rate": 8.565567434510005e-06,
+      "loss": 0.916,
+      "step": 2511
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8556942782473356,
+      "learning_rate": 8.564346641351087e-06,
+      "loss": 0.8957,
+      "step": 2512
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.909167355487353,
+      "learning_rate": 8.56312541600285e-06,
+      "loss": 1.0551,
+      "step": 2513
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6711757428283902,
+      "learning_rate": 8.561903758613372e-06,
+      "loss": 0.9073,
+      "step": 2514
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7133018292570101,
+      "learning_rate": 8.560681669330784e-06,
+      "loss": 0.887,
+      "step": 2515
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7689822897776997,
+      "learning_rate": 8.559459148303268e-06,
+      "loss": 0.9585,
+      "step": 2516
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7096580997932758,
+      "learning_rate": 8.55823619567906e-06,
+      "loss": 0.961,
+      "step": 2517
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6829407649497672,
+      "learning_rate": 8.557012811606444e-06,
+      "loss": 0.8537,
+      "step": 2518
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7099731029499927,
+      "learning_rate": 8.555788996233765e-06,
+      "loss": 0.8999,
+      "step": 2519
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8852470774460868,
+      "learning_rate": 8.554564749709408e-06,
+      "loss": 1.1142,
+      "step": 2520
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7981583572550466,
+      "learning_rate": 8.553340072181823e-06,
+      "loss": 0.9203,
+      "step": 2521
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8324423951383861,
+      "learning_rate": 8.552114963799503e-06,
+      "loss": 0.864,
+      "step": 2522
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7864038851540567,
+      "learning_rate": 8.550889424710998e-06,
+      "loss": 0.8824,
+      "step": 2523
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7558540844592265,
+      "learning_rate": 8.549663455064908e-06,
+      "loss": 0.9754,
+      "step": 2524
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9061760803937868,
+      "learning_rate": 8.548437055009886e-06,
+      "loss": 0.9671,
+      "step": 2525
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7413427651256974,
+      "learning_rate": 8.54721022469464e-06,
+      "loss": 0.8393,
+      "step": 2526
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7560792264343383,
+      "learning_rate": 8.545982964267923e-06,
+      "loss": 0.9143,
+      "step": 2527
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.188509281740642,
+      "learning_rate": 8.544755273878547e-06,
+      "loss": 0.8766,
+      "step": 2528
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.745137770544269,
+      "learning_rate": 8.543527153675374e-06,
+      "loss": 1.0131,
+      "step": 2529
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7448957256431548,
+      "learning_rate": 8.542298603807317e-06,
+      "loss": 0.9939,
+      "step": 2530
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8167657441770644,
+      "learning_rate": 8.541069624423343e-06,
+      "loss": 0.9531,
+      "step": 2531
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7079958232553756,
+      "learning_rate": 8.539840215672468e-06,
+      "loss": 0.9118,
+      "step": 2532
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7073337206425023,
+      "learning_rate": 8.538610377703764e-06,
+      "loss": 0.9325,
+      "step": 2533
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8936395381168936,
+      "learning_rate": 8.537380110666351e-06,
+      "loss": 0.9747,
+      "step": 2534
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6904594663124592,
+      "learning_rate": 8.536149414709405e-06,
+      "loss": 0.972,
+      "step": 2535
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7202107880289506,
+      "learning_rate": 8.534918289982152e-06,
+      "loss": 0.9147,
+      "step": 2536
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7480324379164305,
+      "learning_rate": 8.533686736633869e-06,
+      "loss": 0.9063,
+      "step": 2537
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7520262915967066,
+      "learning_rate": 8.532454754813888e-06,
+      "loss": 0.9765,
+      "step": 2538
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9210625485925725,
+      "learning_rate": 8.531222344671588e-06,
+      "loss": 1.0435,
+      "step": 2539
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.797578736103574,
+      "learning_rate": 8.529989506356407e-06,
+      "loss": 0.9493,
+      "step": 2540
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.25241054458183,
+      "learning_rate": 8.528756240017826e-06,
+      "loss": 0.948,
+      "step": 2541
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7160886596987808,
+      "learning_rate": 8.527522545805386e-06,
+      "loss": 0.8814,
+      "step": 2542
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7440870045135477,
+      "learning_rate": 8.526288423868676e-06,
+      "loss": 0.9733,
+      "step": 2543
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6975794019065396,
+      "learning_rate": 8.525053874357338e-06,
+      "loss": 0.9614,
+      "step": 2544
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8456087674186978,
+      "learning_rate": 8.523818897421066e-06,
+      "loss": 1.0155,
+      "step": 2545
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.798447431066723,
+      "learning_rate": 8.522583493209605e-06,
+      "loss": 1.0064,
+      "step": 2546
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7991590782940976,
+      "learning_rate": 8.521347661872749e-06,
+      "loss": 0.9288,
+      "step": 2547
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7909268845913582,
+      "learning_rate": 8.520111403560349e-06,
+      "loss": 0.8793,
+      "step": 2548
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7337252748638132,
+      "learning_rate": 8.518874718422307e-06,
+      "loss": 0.8773,
+      "step": 2549
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.712702095182618,
+      "learning_rate": 8.517637606608574e-06,
+      "loss": 0.8562,
+      "step": 2550
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9531910222578568,
+      "learning_rate": 8.516400068269153e-06,
+      "loss": 1.082,
+      "step": 2551
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.848233379286405,
+      "learning_rate": 8.515162103554102e-06,
+      "loss": 0.9054,
+      "step": 2552
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6528794441443069,
+      "learning_rate": 8.513923712613526e-06,
+      "loss": 1.0199,
+      "step": 2553
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8384480951643789,
+      "learning_rate": 8.512684895597587e-06,
+      "loss": 1.0706,
+      "step": 2554
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7960463627806293,
+      "learning_rate": 8.511445652656493e-06,
+      "loss": 0.8862,
+      "step": 2555
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.755474264901654,
+      "learning_rate": 8.510205983940508e-06,
+      "loss": 0.9448,
+      "step": 2556
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.720770044125286,
+      "learning_rate": 8.508965889599948e-06,
+      "loss": 0.9122,
+      "step": 2557
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7639823782774813,
+      "learning_rate": 8.507725369785174e-06,
+      "loss": 0.9164,
+      "step": 2558
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5653401031216214,
+      "learning_rate": 8.506484424646606e-06,
+      "loss": 0.9464,
+      "step": 2559
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7079759477726433,
+      "learning_rate": 8.505243054334714e-06,
+      "loss": 0.9168,
+      "step": 2560
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.738999459501983,
+      "learning_rate": 8.504001259000017e-06,
+      "loss": 0.8935,
+      "step": 2561
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8371456695428081,
+      "learning_rate": 8.502759038793086e-06,
+      "loss": 0.9721,
+      "step": 2562
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5854288018038207,
+      "learning_rate": 8.501516393864549e-06,
+      "loss": 0.8508,
+      "step": 2563
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7172489600826966,
+      "learning_rate": 8.500273324365074e-06,
+      "loss": 0.8782,
+      "step": 2564
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7219230345632563,
+      "learning_rate": 8.499029830445393e-06,
+      "loss": 1.0533,
+      "step": 2565
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7330296440402628,
+      "learning_rate": 8.497785912256282e-06,
+      "loss": 0.9504,
+      "step": 2566
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6955362249412649,
+      "learning_rate": 8.496541569948571e-06,
+      "loss": 0.9245,
+      "step": 2567
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.718202823065509,
+      "learning_rate": 8.495296803673138e-06,
+      "loss": 0.8904,
+      "step": 2568
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6827154076704776,
+      "learning_rate": 8.49405161358092e-06,
+      "loss": 0.8937,
+      "step": 2569
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7376731151980855,
+      "learning_rate": 8.492805999822898e-06,
+      "loss": 0.9812,
+      "step": 2570
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9386421883754967,
+      "learning_rate": 8.491559962550104e-06,
+      "loss": 0.9138,
+      "step": 2571
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8945497265588094,
+      "learning_rate": 8.49031350191363e-06,
+      "loss": 0.9849,
+      "step": 2572
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7956170935595999,
+      "learning_rate": 8.48906661806461e-06,
+      "loss": 0.9508,
+      "step": 2573
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7754365932910854,
+      "learning_rate": 8.487819311154233e-06,
+      "loss": 0.9535,
+      "step": 2574
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8794382384300188,
+      "learning_rate": 8.486571581333743e-06,
+      "loss": 0.8962,
+      "step": 2575
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7937174306410917,
+      "learning_rate": 8.485323428754427e-06,
+      "loss": 0.9768,
+      "step": 2576
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7773282609429824,
+      "learning_rate": 8.48407485356763e-06,
+      "loss": 0.9078,
+      "step": 2577
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8646867368635935,
+      "learning_rate": 8.482825855924743e-06,
+      "loss": 0.8525,
+      "step": 2578
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6901706847202045,
+      "learning_rate": 8.481576435977217e-06,
+      "loss": 0.9423,
+      "step": 2579
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7628442634298063,
+      "learning_rate": 8.480326593876545e-06,
+      "loss": 0.9239,
+      "step": 2580
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1094219963095178,
+      "learning_rate": 8.479076329774274e-06,
+      "loss": 1.0517,
+      "step": 2581
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6102205279164097,
+      "learning_rate": 8.477825643822003e-06,
+      "loss": 0.9379,
+      "step": 2582
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7385477973856034,
+      "learning_rate": 8.476574536171385e-06,
+      "loss": 0.9449,
+      "step": 2583
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6785735707479192,
+      "learning_rate": 8.475323006974116e-06,
+      "loss": 0.9834,
+      "step": 2584
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7157062630553985,
+      "learning_rate": 8.474071056381953e-06,
+      "loss": 0.8337,
+      "step": 2585
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7320922808786041,
+      "learning_rate": 8.472818684546697e-06,
+      "loss": 0.9843,
+      "step": 2586
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.773813496638385,
+      "learning_rate": 8.471565891620204e-06,
+      "loss": 0.923,
+      "step": 2587
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6820726665076502,
+      "learning_rate": 8.470312677754377e-06,
+      "loss": 0.8845,
+      "step": 2588
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0732325247699444,
+      "learning_rate": 8.469059043101176e-06,
+      "loss": 0.8959,
+      "step": 2589
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9483009257529063,
+      "learning_rate": 8.467804987812603e-06,
+      "loss": 0.9284,
+      "step": 2590
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6916363965829946,
+      "learning_rate": 8.466550512040722e-06,
+      "loss": 0.9187,
+      "step": 2591
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6975984320722732,
+      "learning_rate": 8.465295615937641e-06,
+      "loss": 0.8892,
+      "step": 2592
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7753475867397381,
+      "learning_rate": 8.464040299655518e-06,
+      "loss": 0.8567,
+      "step": 2593
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.747386784806457,
+      "learning_rate": 8.462784563346567e-06,
+      "loss": 0.8873,
+      "step": 2594
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6508356519570626,
+      "learning_rate": 8.461528407163051e-06,
+      "loss": 0.9508,
+      "step": 2595
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9552568315655434,
+      "learning_rate": 8.46027183125728e-06,
+      "loss": 1.0809,
+      "step": 2596
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7127428165651504,
+      "learning_rate": 8.459014835781622e-06,
+      "loss": 0.9326,
+      "step": 2597
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7751405432908212,
+      "learning_rate": 8.457757420888487e-06,
+      "loss": 0.8169,
+      "step": 2598
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7591020817304277,
+      "learning_rate": 8.456499586730346e-06,
+      "loss": 0.9029,
+      "step": 2599
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7550604614597436,
+      "learning_rate": 8.455241333459714e-06,
+      "loss": 0.9044,
+      "step": 2600
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8192353465068831,
+      "learning_rate": 8.453982661229158e-06,
+      "loss": 0.911,
+      "step": 2601
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8249859401089377,
+      "learning_rate": 8.452723570191298e-06,
+      "loss": 0.8746,
+      "step": 2602
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.703184538855868,
+      "learning_rate": 8.4514640604988e-06,
+      "loss": 0.9262,
+      "step": 2603
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7641586546249024,
+      "learning_rate": 8.450204132304387e-06,
+      "loss": 0.8108,
+      "step": 2604
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7372717243764828,
+      "learning_rate": 8.448943785760827e-06,
+      "loss": 0.9382,
+      "step": 2605
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8409697637621698,
+      "learning_rate": 8.447683021020943e-06,
+      "loss": 0.941,
+      "step": 2606
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6909961113464358,
+      "learning_rate": 8.446421838237606e-06,
+      "loss": 0.8555,
+      "step": 2607
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7745001724058014,
+      "learning_rate": 8.445160237563742e-06,
+      "loss": 0.9677,
+      "step": 2608
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7799225062312554,
+      "learning_rate": 8.44389821915232e-06,
+      "loss": 0.8638,
+      "step": 2609
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6954566451504058,
+      "learning_rate": 8.442635783156365e-06,
+      "loss": 0.8486,
+      "step": 2610
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7981409186172237,
+      "learning_rate": 8.441372929728955e-06,
+      "loss": 0.8815,
+      "step": 2611
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0482956565179307,
+      "learning_rate": 8.44010965902321e-06,
+      "loss": 1.0879,
+      "step": 2612
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7584222387367217,
+      "learning_rate": 8.438845971192313e-06,
+      "loss": 0.937,
+      "step": 2613
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9447779892074077,
+      "learning_rate": 8.437581866389483e-06,
+      "loss": 0.9686,
+      "step": 2614
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.715258059400726,
+      "learning_rate": 8.436317344768004e-06,
+      "loss": 0.9098,
+      "step": 2615
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6744470877283355,
+      "learning_rate": 8.435052406481196e-06,
+      "loss": 0.8853,
+      "step": 2616
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9504950314102734,
+      "learning_rate": 8.433787051682444e-06,
+      "loss": 0.9599,
+      "step": 2617
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6897174407134756,
+      "learning_rate": 8.432521280525174e-06,
+      "loss": 0.9069,
+      "step": 2618
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6534023091249135,
+      "learning_rate": 8.431255093162864e-06,
+      "loss": 0.9883,
+      "step": 2619
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6663760692599767,
+      "learning_rate": 8.429988489749045e-06,
+      "loss": 0.9209,
+      "step": 2620
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8502706037166818,
+      "learning_rate": 8.428721470437296e-06,
+      "loss": 0.9751,
+      "step": 2621
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.855344492547839,
+      "learning_rate": 8.42745403538125e-06,
+      "loss": 0.8843,
+      "step": 2622
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1316661399333625,
+      "learning_rate": 8.426186184734585e-06,
+      "loss": 1.0403,
+      "step": 2623
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8005642837293718,
+      "learning_rate": 8.424917918651032e-06,
+      "loss": 0.8829,
+      "step": 2624
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8743843790242753,
+      "learning_rate": 8.423649237284375e-06,
+      "loss": 0.9562,
+      "step": 2625
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.797817086744795,
+      "learning_rate": 8.422380140788446e-06,
+      "loss": 0.9994,
+      "step": 2626
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7396316483399437,
+      "learning_rate": 8.421110629317123e-06,
+      "loss": 0.9243,
+      "step": 2627
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8260718730516479,
+      "learning_rate": 8.419840703024344e-06,
+      "loss": 0.918,
+      "step": 2628
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7147826556676402,
+      "learning_rate": 8.418570362064092e-06,
+      "loss": 1.0168,
+      "step": 2629
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7155964979321885,
+      "learning_rate": 8.417299606590394e-06,
+      "loss": 0.9469,
+      "step": 2630
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9421527119116255,
+      "learning_rate": 8.41602843675734e-06,
+      "loss": 0.8977,
+      "step": 2631
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6452464874127608,
+      "learning_rate": 8.41475685271906e-06,
+      "loss": 0.9006,
+      "step": 2632
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7015816933350523,
+      "learning_rate": 8.413484854629738e-06,
+      "loss": 0.9426,
+      "step": 2633
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8851094271960538,
+      "learning_rate": 8.412212442643611e-06,
+      "loss": 0.8894,
+      "step": 2634
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.634158953089184,
+      "learning_rate": 8.41093961691496e-06,
+      "loss": 0.9167,
+      "step": 2635
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7876957785658818,
+      "learning_rate": 8.409666377598125e-06,
+      "loss": 0.7817,
+      "step": 2636
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.849030635849866,
+      "learning_rate": 8.408392724847483e-06,
+      "loss": 0.9658,
+      "step": 2637
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8439699668543514,
+      "learning_rate": 8.407118658817475e-06,
+      "loss": 1.0209,
+      "step": 2638
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7982761649350658,
+      "learning_rate": 8.405844179662582e-06,
+      "loss": 1.0246,
+      "step": 2639
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.695499513511908,
+      "learning_rate": 8.40456928753734e-06,
+      "loss": 0.9044,
+      "step": 2640
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1288580516136915,
+      "learning_rate": 8.403293982596336e-06,
+      "loss": 1.0903,
+      "step": 2641
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.835467699708121,
+      "learning_rate": 8.402018264994203e-06,
+      "loss": 0.9588,
+      "step": 2642
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7370235041992217,
+      "learning_rate": 8.400742134885627e-06,
+      "loss": 0.8996,
+      "step": 2643
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6527946909522542,
+      "learning_rate": 8.399465592425342e-06,
+      "loss": 0.9015,
+      "step": 2644
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6916102380545264,
+      "learning_rate": 8.398188637768136e-06,
+      "loss": 0.8555,
+      "step": 2645
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.709860675499942,
+      "learning_rate": 8.396911271068842e-06,
+      "loss": 0.912,
+      "step": 2646
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7404770532970635,
+      "learning_rate": 8.395633492482345e-06,
+      "loss": 0.937,
+      "step": 2647
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.859996815597986,
+      "learning_rate": 8.394355302163578e-06,
+      "loss": 0.9668,
+      "step": 2648
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7359267131818361,
+      "learning_rate": 8.393076700267534e-06,
+      "loss": 0.9,
+      "step": 2649
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7100969269928719,
+      "learning_rate": 8.391797686949238e-06,
+      "loss": 0.9554,
+      "step": 2650
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2409941394393897,
+      "learning_rate": 8.390518262363781e-06,
+      "loss": 1.051,
+      "step": 2651
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.743061164349289,
+      "learning_rate": 8.389238426666294e-06,
+      "loss": 0.8602,
+      "step": 2652
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7255311351340976,
+      "learning_rate": 8.387958180011966e-06,
+      "loss": 0.952,
+      "step": 2653
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.840780779020878,
+      "learning_rate": 8.386677522556026e-06,
+      "loss": 0.9251,
+      "step": 2654
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.818074267855949,
+      "learning_rate": 8.385396454453763e-06,
+      "loss": 0.9531,
+      "step": 2655
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8241573473407016,
+      "learning_rate": 8.384114975860507e-06,
+      "loss": 0.8904,
+      "step": 2656
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8143130407628905,
+      "learning_rate": 8.382833086931643e-06,
+      "loss": 0.955,
+      "step": 2657
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7082477282834667,
+      "learning_rate": 8.381550787822606e-06,
+      "loss": 0.9442,
+      "step": 2658
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.813769931004606,
+      "learning_rate": 8.380268078688877e-06,
+      "loss": 0.9379,
+      "step": 2659
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9471968912168555,
+      "learning_rate": 8.378984959685992e-06,
+      "loss": 1.0636,
+      "step": 2660
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8823170383535623,
+      "learning_rate": 8.377701430969529e-06,
+      "loss": 1.0113,
+      "step": 2661
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.782832459748016,
+      "learning_rate": 8.376417492695123e-06,
+      "loss": 0.9564,
+      "step": 2662
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9258122328249447,
+      "learning_rate": 8.375133145018457e-06,
+      "loss": 0.9549,
+      "step": 2663
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9066642277122932,
+      "learning_rate": 8.37384838809526e-06,
+      "loss": 0.9075,
+      "step": 2664
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7743454664416303,
+      "learning_rate": 8.372563222081316e-06,
+      "loss": 0.9824,
+      "step": 2665
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7473475147276225,
+      "learning_rate": 8.371277647132454e-06,
+      "loss": 0.9312,
+      "step": 2666
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6788261542554566,
+      "learning_rate": 8.369991663404556e-06,
+      "loss": 0.9184,
+      "step": 2667
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.683602206538167,
+      "learning_rate": 8.368705271053547e-06,
+      "loss": 0.9634,
+      "step": 2668
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9017150511640872,
+      "learning_rate": 8.367418470235412e-06,
+      "loss": 0.8989,
+      "step": 2669
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6083489522985317,
+      "learning_rate": 8.366131261106179e-06,
+      "loss": 0.7707,
+      "step": 2670
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6876181223498112,
+      "learning_rate": 8.364843643821928e-06,
+      "loss": 0.9048,
+      "step": 2671
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.647664941227452,
+      "learning_rate": 8.36355561853878e-06,
+      "loss": 0.9269,
+      "step": 2672
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7622096331806898,
+      "learning_rate": 8.36226718541292e-06,
+      "loss": 0.9392,
+      "step": 2673
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.665938051692417,
+      "learning_rate": 8.360978344600572e-06,
+      "loss": 0.9564,
+      "step": 2674
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7172261005916067,
+      "learning_rate": 8.35968909625801e-06,
+      "loss": 0.9068,
+      "step": 2675
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.678394513621128,
+      "learning_rate": 8.358399440541567e-06,
+      "loss": 0.8992,
+      "step": 2676
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3715419765077594,
+      "learning_rate": 8.357109377607611e-06,
+      "loss": 1.0465,
+      "step": 2677
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7960200173193683,
+      "learning_rate": 8.355818907612569e-06,
+      "loss": 0.9591,
+      "step": 2678
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.763986996707925,
+      "learning_rate": 8.354528030712916e-06,
+      "loss": 0.9313,
+      "step": 2679
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8399837515510276,
+      "learning_rate": 8.353236747065175e-06,
+      "loss": 0.9687,
+      "step": 2680
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7314591098805385,
+      "learning_rate": 8.351945056825918e-06,
+      "loss": 0.9257,
+      "step": 2681
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6782753335080047,
+      "learning_rate": 8.350652960151766e-06,
+      "loss": 0.9402,
+      "step": 2682
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.746513737952812,
+      "learning_rate": 8.349360457199392e-06,
+      "loss": 0.8909,
+      "step": 2683
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8061205911221019,
+      "learning_rate": 8.348067548125515e-06,
+      "loss": 0.9114,
+      "step": 2684
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7596610277221119,
+      "learning_rate": 8.346774233086906e-06,
+      "loss": 0.9239,
+      "step": 2685
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5327229850771662,
+      "learning_rate": 8.34548051224038e-06,
+      "loss": 0.8992,
+      "step": 2686
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6502873466957853,
+      "learning_rate": 8.344186385742811e-06,
+      "loss": 0.9044,
+      "step": 2687
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9395286084165593,
+      "learning_rate": 8.342891853751114e-06,
+      "loss": 0.9138,
+      "step": 2688
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7330353443147073,
+      "learning_rate": 8.341596916422255e-06,
+      "loss": 0.9409,
+      "step": 2689
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6875653507488733,
+      "learning_rate": 8.34030157391325e-06,
+      "loss": 0.922,
+      "step": 2690
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1327084874262654,
+      "learning_rate": 8.33900582638116e-06,
+      "loss": 1.0953,
+      "step": 2691
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7702519974024964,
+      "learning_rate": 8.337709673983106e-06,
+      "loss": 0.8898,
+      "step": 2692
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7994040928248924,
+      "learning_rate": 8.336413116876245e-06,
+      "loss": 0.8392,
+      "step": 2693
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.912264281259712,
+      "learning_rate": 8.335116155217793e-06,
+      "loss": 0.9281,
+      "step": 2694
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7735028590372974,
+      "learning_rate": 8.333818789165009e-06,
+      "loss": 0.9877,
+      "step": 2695
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7630128282071538,
+      "learning_rate": 8.332521018875202e-06,
+      "loss": 1.0353,
+      "step": 2696
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.2983693167968013,
+      "learning_rate": 8.331222844505732e-06,
+      "loss": 0.8948,
+      "step": 2697
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7637670354178085,
+      "learning_rate": 8.32992426621401e-06,
+      "loss": 0.9057,
+      "step": 2698
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7155931075709128,
+      "learning_rate": 8.32862528415749e-06,
+      "loss": 0.9213,
+      "step": 2699
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7810209448407306,
+      "learning_rate": 8.327325898493677e-06,
+      "loss": 0.9166,
+      "step": 2700
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8805425341692593,
+      "learning_rate": 8.32602610938013e-06,
+      "loss": 1.0723,
+      "step": 2701
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7871306816141626,
+      "learning_rate": 8.32472591697445e-06,
+      "loss": 0.8776,
+      "step": 2702
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7662755858184496,
+      "learning_rate": 8.323425321434291e-06,
+      "loss": 0.9523,
+      "step": 2703
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7951216627466384,
+      "learning_rate": 8.322124322917354e-06,
+      "loss": 0.8704,
+      "step": 2704
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7425694502880882,
+      "learning_rate": 8.320822921581388e-06,
+      "loss": 0.9157,
+      "step": 2705
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7295089647466184,
+      "learning_rate": 8.319521117584195e-06,
+      "loss": 0.8467,
+      "step": 2706
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6785852328999458,
+      "learning_rate": 8.318218911083622e-06,
+      "loss": 0.8991,
+      "step": 2707
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.985076371177191,
+      "learning_rate": 8.316916302237568e-06,
+      "loss": 0.9487,
+      "step": 2708
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8982798217976038,
+      "learning_rate": 8.315613291203977e-06,
+      "loss": 1.0459,
+      "step": 2709
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7353343579690588,
+      "learning_rate": 8.314309878140843e-06,
+      "loss": 0.9358,
+      "step": 2710
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7953583492918699,
+      "learning_rate": 8.31300606320621e-06,
+      "loss": 0.9766,
+      "step": 2711
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.718497371982298,
+      "learning_rate": 8.311701846558172e-06,
+      "loss": 0.8877,
+      "step": 2712
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8574492711897579,
+      "learning_rate": 8.310397228354864e-06,
+      "loss": 1.0984,
+      "step": 2713
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6380167005396855,
+      "learning_rate": 8.309092208754483e-06,
+      "loss": 0.9213,
+      "step": 2714
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.782591668735152,
+      "learning_rate": 8.307786787915262e-06,
+      "loss": 0.8366,
+      "step": 2715
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6679433014847516,
+      "learning_rate": 8.306480965995488e-06,
+      "loss": 0.9905,
+      "step": 2716
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6163176954143799,
+      "learning_rate": 8.3051747431535e-06,
+      "loss": 0.9058,
+      "step": 2717
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7018227590720274,
+      "learning_rate": 8.30386811954768e-06,
+      "loss": 0.8824,
+      "step": 2718
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6957234767748872,
+      "learning_rate": 8.30256109533646e-06,
+      "loss": 0.8786,
+      "step": 2719
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8324622108291506,
+      "learning_rate": 8.30125367067832e-06,
+      "loss": 1.0019,
+      "step": 2720
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7154783053720553,
+      "learning_rate": 8.29994584573179e-06,
+      "loss": 0.9289,
+      "step": 2721
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6988482463980301,
+      "learning_rate": 8.298637620655454e-06,
+      "loss": 0.9085,
+      "step": 2722
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7942469465855428,
+      "learning_rate": 8.297328995607931e-06,
+      "loss": 0.9633,
+      "step": 2723
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6785819255676184,
+      "learning_rate": 8.296019970747901e-06,
+      "loss": 0.9594,
+      "step": 2724
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.746380807543437,
+      "learning_rate": 8.294710546234086e-06,
+      "loss": 0.8939,
+      "step": 2725
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6461634906873621,
+      "learning_rate": 8.29340072222526e-06,
+      "loss": 0.8934,
+      "step": 2726
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.929523603310582,
+      "learning_rate": 8.292090498880242e-06,
+      "loss": 1.0736,
+      "step": 2727
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7228344104052555,
+      "learning_rate": 8.2907798763579e-06,
+      "loss": 0.8662,
+      "step": 2728
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.203611096269428,
+      "learning_rate": 8.289468854817153e-06,
+      "loss": 0.9843,
+      "step": 2729
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8476155680000359,
+      "learning_rate": 8.288157434416966e-06,
+      "loss": 1.1097,
+      "step": 2730
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.092711603233642,
+      "learning_rate": 8.286845615316357e-06,
+      "loss": 0.8866,
+      "step": 2731
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.631884870900226,
+      "learning_rate": 8.285533397674384e-06,
+      "loss": 0.8052,
+      "step": 2732
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.808410889766534,
+      "learning_rate": 8.284220781650158e-06,
+      "loss": 0.9161,
+      "step": 2733
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6264433908190032,
+      "learning_rate": 8.282907767402842e-06,
+      "loss": 0.9334,
+      "step": 2734
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7557369368203102,
+      "learning_rate": 8.281594355091642e-06,
+      "loss": 0.9406,
+      "step": 2735
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8772571145021079,
+      "learning_rate": 8.280280544875812e-06,
+      "loss": 1.06,
+      "step": 2736
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7095666854816451,
+      "learning_rate": 8.278966336914656e-06,
+      "loss": 0.933,
+      "step": 2737
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7816063460096223,
+      "learning_rate": 8.277651731367528e-06,
+      "loss": 0.8956,
+      "step": 2738
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.788961895320356,
+      "learning_rate": 8.276336728393827e-06,
+      "loss": 0.931,
+      "step": 2739
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7373480983017346,
+      "learning_rate": 8.275021328153005e-06,
+      "loss": 0.978,
+      "step": 2740
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7042862134644112,
+      "learning_rate": 8.273705530804554e-06,
+      "loss": 1.0133,
+      "step": 2741
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.899507680529035,
+      "learning_rate": 8.272389336508022e-06,
+      "loss": 0.9805,
+      "step": 2742
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8309833873559225,
+      "learning_rate": 8.271072745423001e-06,
+      "loss": 0.9816,
+      "step": 2743
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.699465801321826,
+      "learning_rate": 8.269755757709132e-06,
+      "loss": 0.9308,
+      "step": 2744
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9192797505869506,
+      "learning_rate": 8.268438373526106e-06,
+      "loss": 1.046,
+      "step": 2745
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6999154119244861,
+      "learning_rate": 8.267120593033659e-06,
+      "loss": 0.891,
+      "step": 2746
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8664738352806673,
+      "learning_rate": 8.265802416391577e-06,
+      "loss": 0.8752,
+      "step": 2747
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7808979863105727,
+      "learning_rate": 8.264483843759692e-06,
+      "loss": 0.9879,
+      "step": 2748
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8965778774583174,
+      "learning_rate": 8.263164875297888e-06,
+      "loss": 1.0868,
+      "step": 2749
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6631241854303789,
+      "learning_rate": 8.261845511166092e-06,
+      "loss": 0.9544,
+      "step": 2750
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8049196326966428,
+      "learning_rate": 8.260525751524282e-06,
+      "loss": 0.9211,
+      "step": 2751
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7160576619567764,
+      "learning_rate": 8.259205596532485e-06,
+      "loss": 0.9124,
+      "step": 2752
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.686019164171783,
+      "learning_rate": 8.257885046350773e-06,
+      "loss": 0.9157,
+      "step": 2753
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7221077898374724,
+      "learning_rate": 8.256564101139266e-06,
+      "loss": 0.8766,
+      "step": 2754
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8231759610097789,
+      "learning_rate": 8.255242761058136e-06,
+      "loss": 0.957,
+      "step": 2755
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.759682193603437,
+      "learning_rate": 8.2539210262676e-06,
+      "loss": 0.945,
+      "step": 2756
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8793164505391022,
+      "learning_rate": 8.252598896927918e-06,
+      "loss": 0.952,
+      "step": 2757
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.822262619268538,
+      "learning_rate": 8.25127637319941e-06,
+      "loss": 0.9576,
+      "step": 2758
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7788753680637348,
+      "learning_rate": 8.24995345524243e-06,
+      "loss": 0.9522,
+      "step": 2759
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6838462749940153,
+      "learning_rate": 8.248630143217391e-06,
+      "loss": 0.9646,
+      "step": 2760
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7959803781454455,
+      "learning_rate": 8.247306437284748e-06,
+      "loss": 1.0449,
+      "step": 2761
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.880919203775911,
+      "learning_rate": 8.245982337605003e-06,
+      "loss": 1.0622,
+      "step": 2762
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.622618891982184,
+      "learning_rate": 8.244657844338708e-06,
+      "loss": 0.8846,
+      "step": 2763
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7643856711592325,
+      "learning_rate": 8.243332957646465e-06,
+      "loss": 0.8855,
+      "step": 2764
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9391942248676322,
+      "learning_rate": 8.242007677688919e-06,
+      "loss": 0.931,
+      "step": 2765
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7505159058787862,
+      "learning_rate": 8.240682004626765e-06,
+      "loss": 0.9939,
+      "step": 2766
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7870261653742308,
+      "learning_rate": 8.239355938620746e-06,
+      "loss": 0.945,
+      "step": 2767
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7457823445503122,
+      "learning_rate": 8.238029479831651e-06,
+      "loss": 0.9106,
+      "step": 2768
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6643852951084417,
+      "learning_rate": 8.236702628420319e-06,
+      "loss": 0.9082,
+      "step": 2769
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7248900881640263,
+      "learning_rate": 8.235375384547634e-06,
+      "loss": 0.9332,
+      "step": 2770
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6714804928938063,
+      "learning_rate": 8.23404774837453e-06,
+      "loss": 0.8496,
+      "step": 2771
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.751678055902103,
+      "learning_rate": 8.232719720061989e-06,
+      "loss": 0.8999,
+      "step": 2772
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9305454620126178,
+      "learning_rate": 8.231391299771034e-06,
+      "loss": 1.0055,
+      "step": 2773
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7260134700929628,
+      "learning_rate": 8.230062487662745e-06,
+      "loss": 0.9145,
+      "step": 2774
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.677395840109699,
+      "learning_rate": 8.228733283898244e-06,
+      "loss": 0.8984,
+      "step": 2775
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7529293727432356,
+      "learning_rate": 8.2274036886387e-06,
+      "loss": 0.9731,
+      "step": 2776
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6944433358630042,
+      "learning_rate": 8.226073702045334e-06,
+      "loss": 0.8633,
+      "step": 2777
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8480141334615585,
+      "learning_rate": 8.224743324279407e-06,
+      "loss": 1.0811,
+      "step": 2778
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7254054131813572,
+      "learning_rate": 8.223412555502238e-06,
+      "loss": 0.9959,
+      "step": 2779
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7424225148224677,
+      "learning_rate": 8.222081395875182e-06,
+      "loss": 0.8616,
+      "step": 2780
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8218760512356698,
+      "learning_rate": 8.220749845559648e-06,
+      "loss": 0.9797,
+      "step": 2781
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6962925290936426,
+      "learning_rate": 8.219417904717091e-06,
+      "loss": 0.9344,
+      "step": 2782
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6802446339008914,
+      "learning_rate": 8.218085573509016e-06,
+      "loss": 1.0103,
+      "step": 2783
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.720263468485671,
+      "learning_rate": 8.21675285209697e-06,
+      "loss": 0.91,
+      "step": 2784
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8081122650441672,
+      "learning_rate": 8.21541974064255e-06,
+      "loss": 0.8891,
+      "step": 2785
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6801263846602599,
+      "learning_rate": 8.214086239307402e-06,
+      "loss": 0.9617,
+      "step": 2786
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7513200525295334,
+      "learning_rate": 8.212752348253216e-06,
+      "loss": 0.9889,
+      "step": 2787
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6632321991080574,
+      "learning_rate": 8.211418067641734e-06,
+      "loss": 0.9137,
+      "step": 2788
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6936958150935835,
+      "learning_rate": 8.210083397634738e-06,
+      "loss": 0.9442,
+      "step": 2789
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.632199028005076,
+      "learning_rate": 8.208748338394063e-06,
+      "loss": 0.8623,
+      "step": 2790
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8791645381117066,
+      "learning_rate": 8.20741289008159e-06,
+      "loss": 0.9696,
+      "step": 2791
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7403213382327436,
+      "learning_rate": 8.206077052859247e-06,
+      "loss": 0.9606,
+      "step": 2792
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.700611017803361,
+      "learning_rate": 8.204740826889009e-06,
+      "loss": 0.9472,
+      "step": 2793
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6728539355543606,
+      "learning_rate": 8.203404212332897e-06,
+      "loss": 0.9327,
+      "step": 2794
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9094569990041427,
+      "learning_rate": 8.20206720935298e-06,
+      "loss": 1.0646,
+      "step": 2795
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7263232240622515,
+      "learning_rate": 8.200729818111372e-06,
+      "loss": 0.8977,
+      "step": 2796
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7276002853362364,
+      "learning_rate": 8.199392038770243e-06,
+      "loss": 0.9594,
+      "step": 2797
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8271428048832092,
+      "learning_rate": 8.198053871491798e-06,
+      "loss": 0.8812,
+      "step": 2798
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8169464200185546,
+      "learning_rate": 8.196715316438295e-06,
+      "loss": 0.8627,
+      "step": 2799
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7537746440979876,
+      "learning_rate": 8.19537637377204e-06,
+      "loss": 0.948,
+      "step": 2800
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7916949844352335,
+      "learning_rate": 8.194037043655383e-06,
+      "loss": 0.9641,
+      "step": 2801
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7616545752731128,
+      "learning_rate": 8.192697326250722e-06,
+      "loss": 0.9124,
+      "step": 2802
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6936851652706455,
+      "learning_rate": 8.191357221720506e-06,
+      "loss": 0.9201,
+      "step": 2803
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9099671891053467,
+      "learning_rate": 8.190016730227224e-06,
+      "loss": 0.9585,
+      "step": 2804
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7850895794432966,
+      "learning_rate": 8.188675851933414e-06,
+      "loss": 0.9881,
+      "step": 2805
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8391320391949635,
+      "learning_rate": 8.187334587001664e-06,
+      "loss": 0.955,
+      "step": 2806
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.023557484890109,
+      "learning_rate": 8.185992935594608e-06,
+      "loss": 1.0914,
+      "step": 2807
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8077328220047488,
+      "learning_rate": 8.184650897874924e-06,
+      "loss": 0.961,
+      "step": 2808
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7982462090793578,
+      "learning_rate": 8.18330847400534e-06,
+      "loss": 0.9113,
+      "step": 2809
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7523813734378182,
+      "learning_rate": 8.181965664148629e-06,
+      "loss": 0.9471,
+      "step": 2810
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7926691801718706,
+      "learning_rate": 8.180622468467612e-06,
+      "loss": 0.8779,
+      "step": 2811
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7603223730040563,
+      "learning_rate": 8.179278887125152e-06,
+      "loss": 0.9832,
+      "step": 2812
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0126722844195852,
+      "learning_rate": 8.17793492028417e-06,
+      "loss": 1.0702,
+      "step": 2813
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.743455940782216,
+      "learning_rate": 8.176590568107623e-06,
+      "loss": 0.8831,
+      "step": 2814
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7810972964787686,
+      "learning_rate": 8.175245830758516e-06,
+      "loss": 0.9399,
+      "step": 2815
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6576693510758571,
+      "learning_rate": 8.173900708399907e-06,
+      "loss": 0.8413,
+      "step": 2816
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6759733707638447,
+      "learning_rate": 8.172555201194895e-06,
+      "loss": 0.8622,
+      "step": 2817
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7167791207919927,
+      "learning_rate": 8.171209309306626e-06,
+      "loss": 0.853,
+      "step": 2818
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6837904562189485,
+      "learning_rate": 8.169863032898297e-06,
+      "loss": 0.8865,
+      "step": 2819
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8210809224454656,
+      "learning_rate": 8.168516372133145e-06,
+      "loss": 0.929,
+      "step": 2820
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7329153160651818,
+      "learning_rate": 8.16716932717446e-06,
+      "loss": 0.9093,
+      "step": 2821
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8513229929891395,
+      "learning_rate": 8.165821898185577e-06,
+      "loss": 0.9891,
+      "step": 2822
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7944468354068013,
+      "learning_rate": 8.164474085329872e-06,
+      "loss": 0.8756,
+      "step": 2823
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.956797991942685,
+      "learning_rate": 8.163125888770777e-06,
+      "loss": 1.0806,
+      "step": 2824
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7871940791188323,
+      "learning_rate": 8.161777308671763e-06,
+      "loss": 0.9273,
+      "step": 2825
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7556344359549003,
+      "learning_rate": 8.160428345196347e-06,
+      "loss": 0.9543,
+      "step": 2826
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8228769957219948,
+      "learning_rate": 8.159078998508101e-06,
+      "loss": 0.8812,
+      "step": 2827
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.740974131494849,
+      "learning_rate": 8.157729268770636e-06,
+      "loss": 0.9684,
+      "step": 2828
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6956669865645169,
+      "learning_rate": 8.156379156147608e-06,
+      "loss": 0.9403,
+      "step": 2829
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7755018651352146,
+      "learning_rate": 8.155028660802729e-06,
+      "loss": 0.9671,
+      "step": 2830
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7838456917116254,
+      "learning_rate": 8.153677782899744e-06,
+      "loss": 0.9101,
+      "step": 2831
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6607161078193295,
+      "learning_rate": 8.152326522602458e-06,
+      "loss": 0.9342,
+      "step": 2832
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9845393792026735,
+      "learning_rate": 8.150974880074714e-06,
+      "loss": 1.0851,
+      "step": 2833
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7426784699647768,
+      "learning_rate": 8.149622855480401e-06,
+      "loss": 0.8844,
+      "step": 2834
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7423948808625718,
+      "learning_rate": 8.14827044898346e-06,
+      "loss": 0.9315,
+      "step": 2835
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.676680525491925,
+      "learning_rate": 8.146917660747872e-06,
+      "loss": 0.9833,
+      "step": 2836
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7294639602853639,
+      "learning_rate": 8.145564490937669e-06,
+      "loss": 0.8541,
+      "step": 2837
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7865593454650484,
+      "learning_rate": 8.144210939716929e-06,
+      "loss": 0.9438,
+      "step": 2838
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.601630273824032,
+      "learning_rate": 8.142857007249769e-06,
+      "loss": 0.8159,
+      "step": 2839
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7672434005824056,
+      "learning_rate": 8.141502693700364e-06,
+      "loss": 0.9836,
+      "step": 2840
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6876311468679848,
+      "learning_rate": 8.140147999232925e-06,
+      "loss": 0.8766,
+      "step": 2841
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.675170392806443,
+      "learning_rate": 8.138792924011719e-06,
+      "loss": 0.9813,
+      "step": 2842
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6803400543037053,
+      "learning_rate": 8.137437468201047e-06,
+      "loss": 0.9217,
+      "step": 2843
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9569786048851022,
+      "learning_rate": 8.136081631965268e-06,
+      "loss": 1.0825,
+      "step": 2844
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8166448450707677,
+      "learning_rate": 8.134725415468777e-06,
+      "loss": 0.8285,
+      "step": 2845
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7387706074150941,
+      "learning_rate": 8.133368818876024e-06,
+      "loss": 0.8705,
+      "step": 2846
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.85656786405602,
+      "learning_rate": 8.132011842351497e-06,
+      "loss": 1.0582,
+      "step": 2847
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6475130628248664,
+      "learning_rate": 8.130654486059739e-06,
+      "loss": 0.9116,
+      "step": 2848
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8119435398900445,
+      "learning_rate": 8.12929675016533e-06,
+      "loss": 0.9517,
+      "step": 2849
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6826668565577119,
+      "learning_rate": 8.127938634832901e-06,
+      "loss": 0.9666,
+      "step": 2850
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.729977879929465,
+      "learning_rate": 8.12658014022713e-06,
+      "loss": 0.9478,
+      "step": 2851
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7033297083711196,
+      "learning_rate": 8.125221266512739e-06,
+      "loss": 0.918,
+      "step": 2852
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7893971628280614,
+      "learning_rate": 8.123862013854496e-06,
+      "loss": 0.949,
+      "step": 2853
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7702862410351738,
+      "learning_rate": 8.122502382417211e-06,
+      "loss": 0.8918,
+      "step": 2854
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8759427568499567,
+      "learning_rate": 8.121142372365749e-06,
+      "loss": 0.9856,
+      "step": 2855
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.765828197522128,
+      "learning_rate": 8.119781983865014e-06,
+      "loss": 0.9213,
+      "step": 2856
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.846691765707053,
+      "learning_rate": 8.118421217079958e-06,
+      "loss": 0.9246,
+      "step": 2857
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7082448168913487,
+      "learning_rate": 8.117060072175577e-06,
+      "loss": 0.9159,
+      "step": 2858
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9814860035968066,
+      "learning_rate": 8.11569854931692e-06,
+      "loss": 1.0534,
+      "step": 2859
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.638293813068346,
+      "learning_rate": 8.114336648669069e-06,
+      "loss": 0.9557,
+      "step": 2860
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8078328195382445,
+      "learning_rate": 8.112974370397163e-06,
+      "loss": 1.0433,
+      "step": 2861
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.75571520895568,
+      "learning_rate": 8.111611714666382e-06,
+      "loss": 0.8815,
+      "step": 2862
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6332668262539578,
+      "learning_rate": 8.110248681641957e-06,
+      "loss": 0.9062,
+      "step": 2863
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7128781504148496,
+      "learning_rate": 8.108885271489152e-06,
+      "loss": 0.907,
+      "step": 2864
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7604821249341827,
+      "learning_rate": 8.107521484373293e-06,
+      "loss": 0.8669,
+      "step": 2865
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.666299475466886,
+      "learning_rate": 8.10615732045974e-06,
+      "loss": 0.8494,
+      "step": 2866
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.66312537039253,
+      "learning_rate": 8.104792779913903e-06,
+      "loss": 0.9181,
+      "step": 2867
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6796549361441824,
+      "learning_rate": 8.103427862901237e-06,
+      "loss": 0.9127,
+      "step": 2868
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9634481405300964,
+      "learning_rate": 8.102062569587245e-06,
+      "loss": 1.0632,
+      "step": 2869
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6989288504714395,
+      "learning_rate": 8.10069690013747e-06,
+      "loss": 0.9578,
+      "step": 2870
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7012700594309373,
+      "learning_rate": 8.099330854717507e-06,
+      "loss": 0.9138,
+      "step": 2871
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8266361267117774,
+      "learning_rate": 8.097964433492995e-06,
+      "loss": 1.0602,
+      "step": 2872
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9204189545693313,
+      "learning_rate": 8.096597636629612e-06,
+      "loss": 0.9602,
+      "step": 2873
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6538042611961672,
+      "learning_rate": 8.09523046429309e-06,
+      "loss": 0.7947,
+      "step": 2874
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.863480943022291,
+      "learning_rate": 8.093862916649206e-06,
+      "loss": 1.0999,
+      "step": 2875
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6469037531301154,
+      "learning_rate": 8.092494993863775e-06,
+      "loss": 0.9348,
+      "step": 2876
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7379967267179726,
+      "learning_rate": 8.091126696102666e-06,
+      "loss": 0.9316,
+      "step": 2877
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6474141358016106,
+      "learning_rate": 8.089758023531788e-06,
+      "loss": 0.8903,
+      "step": 2878
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.663640849162866,
+      "learning_rate": 8.088388976317096e-06,
+      "loss": 0.9303,
+      "step": 2879
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8730585986589986,
+      "learning_rate": 8.087019554624595e-06,
+      "loss": 1.0487,
+      "step": 2880
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8736874861150625,
+      "learning_rate": 8.085649758620331e-06,
+      "loss": 1.0197,
+      "step": 2881
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.711567417552114,
+      "learning_rate": 8.084279588470395e-06,
+      "loss": 0.8856,
+      "step": 2882
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9085160248260284,
+      "learning_rate": 8.082909044340926e-06,
+      "loss": 0.8919,
+      "step": 2883
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7127941738661436,
+      "learning_rate": 8.081538126398105e-06,
+      "loss": 0.8509,
+      "step": 2884
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8730991458744477,
+      "learning_rate": 8.080166834808166e-06,
+      "loss": 0.9342,
+      "step": 2885
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7544563738481065,
+      "learning_rate": 8.078795169737377e-06,
+      "loss": 0.8991,
+      "step": 2886
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.694982914036666,
+      "learning_rate": 8.077423131352058e-06,
+      "loss": 0.8669,
+      "step": 2887
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.263359502537878,
+      "learning_rate": 8.076050719818577e-06,
+      "loss": 0.9559,
+      "step": 2888
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7157128411106437,
+      "learning_rate": 8.074677935303343e-06,
+      "loss": 0.9383,
+      "step": 2889
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6557470052233758,
+      "learning_rate": 8.073304777972806e-06,
+      "loss": 0.9248,
+      "step": 2890
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7197185685971936,
+      "learning_rate": 8.071931247993472e-06,
+      "loss": 0.8272,
+      "step": 2891
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.763888708393084,
+      "learning_rate": 8.070557345531883e-06,
+      "loss": 0.9762,
+      "step": 2892
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7765761119169388,
+      "learning_rate": 8.06918307075463e-06,
+      "loss": 0.9896,
+      "step": 2893
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.740050352569229,
+      "learning_rate": 8.067808423828347e-06,
+      "loss": 0.9552,
+      "step": 2894
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7920915483073587,
+      "learning_rate": 8.066433404919718e-06,
+      "loss": 0.8713,
+      "step": 2895
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8601944826417283,
+      "learning_rate": 8.065058014195466e-06,
+      "loss": 0.9619,
+      "step": 2896
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7543588636993248,
+      "learning_rate": 8.063682251822363e-06,
+      "loss": 0.8736,
+      "step": 2897
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8395533613622328,
+      "learning_rate": 8.062306117967224e-06,
+      "loss": 0.861,
+      "step": 2898
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8337220037637945,
+      "learning_rate": 8.060929612796914e-06,
+      "loss": 0.8865,
+      "step": 2899
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7254885376233706,
+      "learning_rate": 8.059552736478333e-06,
+      "loss": 0.9309,
+      "step": 2900
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6879788062038426,
+      "learning_rate": 8.058175489178437e-06,
+      "loss": 0.931,
+      "step": 2901
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.744595021716238,
+      "learning_rate": 8.056797871064216e-06,
+      "loss": 0.8438,
+      "step": 2902
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6959573630851648,
+      "learning_rate": 8.05541988230272e-06,
+      "loss": 0.9046,
+      "step": 2903
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7382972432655535,
+      "learning_rate": 8.054041523061027e-06,
+      "loss": 0.9145,
+      "step": 2904
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7967584866206283,
+      "learning_rate": 8.05266279350627e-06,
+      "loss": 1.0066,
+      "step": 2905
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.726117619950246,
+      "learning_rate": 8.051283693805624e-06,
+      "loss": 0.8254,
+      "step": 2906
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6305839145964693,
+      "learning_rate": 8.049904224126312e-06,
+      "loss": 0.8947,
+      "step": 2907
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.736890606645137,
+      "learning_rate": 8.048524384635598e-06,
+      "loss": 0.8799,
+      "step": 2908
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7348075553905002,
+      "learning_rate": 8.047144175500793e-06,
+      "loss": 0.9313,
+      "step": 2909
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0197252195445976,
+      "learning_rate": 8.045763596889253e-06,
+      "loss": 1.0553,
+      "step": 2910
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8911318602830792,
+      "learning_rate": 8.044382648968375e-06,
+      "loss": 1.0697,
+      "step": 2911
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8648565050352723,
+      "learning_rate": 8.043001331905605e-06,
+      "loss": 0.8657,
+      "step": 2912
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7451521569135031,
+      "learning_rate": 8.041619645868434e-06,
+      "loss": 0.8611,
+      "step": 2913
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.673333981687425,
+      "learning_rate": 8.040237591024394e-06,
+      "loss": 0.8989,
+      "step": 2914
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7488349593500943,
+      "learning_rate": 8.038855167541064e-06,
+      "loss": 0.925,
+      "step": 2915
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6510866101260666,
+      "learning_rate": 8.037472375586068e-06,
+      "loss": 0.874,
+      "step": 2916
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8626299467868987,
+      "learning_rate": 8.036089215327077e-06,
+      "loss": 1.0116,
+      "step": 2917
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.739515187137894,
+      "learning_rate": 8.0347056869318e-06,
+      "loss": 0.9359,
+      "step": 2918
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7360371705141284,
+      "learning_rate": 8.033321790567997e-06,
+      "loss": 0.8768,
+      "step": 2919
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7295647459792023,
+      "learning_rate": 8.03193752640347e-06,
+      "loss": 0.8981,
+      "step": 2920
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2310059802688398,
+      "learning_rate": 8.030552894606064e-06,
+      "loss": 1.0667,
+      "step": 2921
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7298853145476654,
+      "learning_rate": 8.029167895343672e-06,
+      "loss": 0.928,
+      "step": 2922
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6233897064856069,
+      "learning_rate": 8.02778252878423e-06,
+      "loss": 0.8812,
+      "step": 2923
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9131666267390448,
+      "learning_rate": 8.026396795095717e-06,
+      "loss": 0.9546,
+      "step": 2924
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.730614456620812,
+      "learning_rate": 8.02501069444616e-06,
+      "loss": 0.9281,
+      "step": 2925
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7764591196803095,
+      "learning_rate": 8.023624227003627e-06,
+      "loss": 0.8615,
+      "step": 2926
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.712387957857228,
+      "learning_rate": 8.022237392936232e-06,
+      "loss": 1.0304,
+      "step": 2927
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6311130112057033,
+      "learning_rate": 8.020850192412134e-06,
+      "loss": 0.8603,
+      "step": 2928
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.810281519117069,
+      "learning_rate": 8.019462625599537e-06,
+      "loss": 0.9525,
+      "step": 2929
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6888513375698708,
+      "learning_rate": 8.018074692666686e-06,
+      "loss": 0.9356,
+      "step": 2930
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7988212060366002,
+      "learning_rate": 8.016686393781874e-06,
+      "loss": 0.9487,
+      "step": 2931
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6967369267536736,
+      "learning_rate": 8.015297729113436e-06,
+      "loss": 0.9826,
+      "step": 2932
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6849113499071326,
+      "learning_rate": 8.013908698829754e-06,
+      "loss": 0.8889,
+      "step": 2933
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7808594686281816,
+      "learning_rate": 8.012519303099251e-06,
+      "loss": 1.0021,
+      "step": 2934
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.089876439198362,
+      "learning_rate": 8.011129542090396e-06,
+      "loss": 0.9069,
+      "step": 2935
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8380134174696203,
+      "learning_rate": 8.009739415971704e-06,
+      "loss": 0.9077,
+      "step": 2936
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7667444970933812,
+      "learning_rate": 8.008348924911734e-06,
+      "loss": 0.9719,
+      "step": 2937
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0309818756778706,
+      "learning_rate": 8.006958069079082e-06,
+      "loss": 1.0557,
+      "step": 2938
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9260482450569845,
+      "learning_rate": 8.005566848642398e-06,
+      "loss": 1.06,
+      "step": 2939
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7180721071682272,
+      "learning_rate": 8.004175263770373e-06,
+      "loss": 0.8919,
+      "step": 2940
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8725319571481909,
+      "learning_rate": 8.00278331463174e-06,
+      "loss": 1.0494,
+      "step": 2941
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8256862139733991,
+      "learning_rate": 8.001391001395278e-06,
+      "loss": 0.9382,
+      "step": 2942
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8326475006450833,
+      "learning_rate": 7.99999832422981e-06,
+      "loss": 0.9035,
+      "step": 2943
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7115561834170654,
+      "learning_rate": 7.998605283304201e-06,
+      "loss": 0.9935,
+      "step": 2944
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.823027935733167,
+      "learning_rate": 7.997211878787366e-06,
+      "loss": 1.0149,
+      "step": 2945
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.85341726597383,
+      "learning_rate": 7.995818110848256e-06,
+      "loss": 0.8772,
+      "step": 2946
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7159463726841102,
+      "learning_rate": 7.994423979655873e-06,
+      "loss": 0.8606,
+      "step": 2947
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7203228487758915,
+      "learning_rate": 7.993029485379258e-06,
+      "loss": 0.9061,
+      "step": 2948
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7505587927625392,
+      "learning_rate": 7.991634628187499e-06,
+      "loss": 0.9761,
+      "step": 2949
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6611388646421599,
+      "learning_rate": 7.99023940824973e-06,
+      "loss": 0.8543,
+      "step": 2950
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.727963538352937,
+      "learning_rate": 7.98884382573512e-06,
+      "loss": 0.9496,
+      "step": 2951
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6618549244602812,
+      "learning_rate": 7.987447880812895e-06,
+      "loss": 0.9276,
+      "step": 2952
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7408533363420053,
+      "learning_rate": 7.986051573652315e-06,
+      "loss": 0.9183,
+      "step": 2953
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6505203951835345,
+      "learning_rate": 7.984654904422686e-06,
+      "loss": 0.8685,
+      "step": 2954
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.833688639920115,
+      "learning_rate": 7.983257873293362e-06,
+      "loss": 0.8817,
+      "step": 2955
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.755261701533233,
+      "learning_rate": 7.981860480433735e-06,
+      "loss": 0.8418,
+      "step": 2956
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7657951827138199,
+      "learning_rate": 7.980462726013246e-06,
+      "loss": 0.9233,
+      "step": 2957
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9123686441348768,
+      "learning_rate": 7.979064610201372e-06,
+      "loss": 0.9375,
+      "step": 2958
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8504300875096165,
+      "learning_rate": 7.977666133167647e-06,
+      "loss": 1.0314,
+      "step": 2959
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.767709834502355,
+      "learning_rate": 7.976267295081637e-06,
+      "loss": 0.9717,
+      "step": 2960
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7015052437190794,
+      "learning_rate": 7.974868096112957e-06,
+      "loss": 0.9007,
+      "step": 2961
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.76733128058154,
+      "learning_rate": 7.973468536431266e-06,
+      "loss": 0.8969,
+      "step": 2962
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7987822606299657,
+      "learning_rate": 7.972068616206261e-06,
+      "loss": 0.8863,
+      "step": 2963
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6685030764179183,
+      "learning_rate": 7.970668335607692e-06,
+      "loss": 0.9207,
+      "step": 2964
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3498620727331005,
+      "learning_rate": 7.969267694805345e-06,
+      "loss": 1.0685,
+      "step": 2965
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8507719264599134,
+      "learning_rate": 7.967866693969053e-06,
+      "loss": 0.9016,
+      "step": 2966
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6764312063865234,
+      "learning_rate": 7.966465333268693e-06,
+      "loss": 0.9398,
+      "step": 2967
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9455703125873531,
+      "learning_rate": 7.965063612874184e-06,
+      "loss": 0.8896,
+      "step": 2968
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7060832490941273,
+      "learning_rate": 7.963661532955492e-06,
+      "loss": 0.891,
+      "step": 2969
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7374244426753016,
+      "learning_rate": 7.962259093682618e-06,
+      "loss": 0.9264,
+      "step": 2970
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7637257042302563,
+      "learning_rate": 7.960856295225619e-06,
+      "loss": 0.9111,
+      "step": 2971
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7471803798653538,
+      "learning_rate": 7.959453137754586e-06,
+      "loss": 0.8759,
+      "step": 2972
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7099133771672819,
+      "learning_rate": 7.958049621439659e-06,
+      "loss": 0.8675,
+      "step": 2973
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.01931654962637,
+      "learning_rate": 7.956645746451014e-06,
+      "loss": 1.0558,
+      "step": 2974
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8118374219059752,
+      "learning_rate": 7.95524151295888e-06,
+      "loss": 0.8937,
+      "step": 2975
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7442891052139102,
+      "learning_rate": 7.953836921133526e-06,
+      "loss": 0.859,
+      "step": 2976
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.7443088721289954,
+      "learning_rate": 7.952431971145261e-06,
+      "loss": 0.8601,
+      "step": 2977
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8486025808592623,
+      "learning_rate": 7.951026663164441e-06,
+      "loss": 0.9132,
+      "step": 2978
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7627401199208774,
+      "learning_rate": 7.949620997361465e-06,
+      "loss": 0.9483,
+      "step": 2979
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6070348988286416,
+      "learning_rate": 7.948214973906775e-06,
+      "loss": 0.807,
+      "step": 2980
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9076659717284925,
+      "learning_rate": 7.946808592970851e-06,
+      "loss": 0.958,
+      "step": 2981
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6559070754735314,
+      "learning_rate": 7.945401854724231e-06,
+      "loss": 0.9406,
+      "step": 2982
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.705407010193344,
+      "learning_rate": 7.943994759337479e-06,
+      "loss": 1.0243,
+      "step": 2983
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8206831405419486,
+      "learning_rate": 7.942587306981213e-06,
+      "loss": 0.9184,
+      "step": 2984
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.953593550419817,
+      "learning_rate": 7.941179497826093e-06,
+      "loss": 0.9542,
+      "step": 2985
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.766722529972484,
+      "learning_rate": 7.939771332042818e-06,
+      "loss": 0.9201,
+      "step": 2986
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7906824996398598,
+      "learning_rate": 7.938362809802134e-06,
+      "loss": 0.9624,
+      "step": 2987
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6741816641776495,
+      "learning_rate": 7.936953931274828e-06,
+      "loss": 0.9067,
+      "step": 2988
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.759505670868694,
+      "learning_rate": 7.935544696631733e-06,
+      "loss": 0.93,
+      "step": 2989
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8037476091131501,
+      "learning_rate": 7.934135106043726e-06,
+      "loss": 0.8611,
+      "step": 2990
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7422269021038563,
+      "learning_rate": 7.93272515968172e-06,
+      "loss": 0.9718,
+      "step": 2991
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7453539235313276,
+      "learning_rate": 7.931314857716677e-06,
+      "loss": 0.8847,
+      "step": 2992
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.072127211437519,
+      "learning_rate": 7.929904200319602e-06,
+      "loss": 0.7712,
+      "step": 2993
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7586598412861487,
+      "learning_rate": 7.928493187661543e-06,
+      "loss": 0.9442,
+      "step": 2994
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7792507858132671,
+      "learning_rate": 7.927081819913589e-06,
+      "loss": 1.0287,
+      "step": 2995
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.851795925412325,
+      "learning_rate": 7.92567009724687e-06,
+      "loss": 0.8984,
+      "step": 2996
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6501559489377569,
+      "learning_rate": 7.92425801983257e-06,
+      "loss": 0.8418,
+      "step": 2997
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.187925619145151,
+      "learning_rate": 7.9228455878419e-06,
+      "loss": 0.9444,
+      "step": 2998
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.694693462816271,
+      "learning_rate": 7.921432801446128e-06,
+      "loss": 0.9787,
+      "step": 2999
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.821450310522034,
+      "learning_rate": 7.920019660816556e-06,
+      "loss": 0.8504,
+      "step": 3000
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5975235472576421,
+      "learning_rate": 7.918606166124534e-06,
+      "loss": 0.87,
+      "step": 3001
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9426950529524436,
+      "learning_rate": 7.917192317541452e-06,
+      "loss": 0.9729,
+      "step": 3002
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.710853465313771,
+      "learning_rate": 7.915778115238744e-06,
+      "loss": 0.9292,
+      "step": 3003
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6917003902401524,
+      "learning_rate": 7.914363559387887e-06,
+      "loss": 0.9098,
+      "step": 3004
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6791101081119189,
+      "learning_rate": 7.912948650160403e-06,
+      "loss": 0.8839,
+      "step": 3005
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7293874954884805,
+      "learning_rate": 7.911533387727852e-06,
+      "loss": 0.8854,
+      "step": 3006
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6405422315090548,
+      "learning_rate": 7.91011777226184e-06,
+      "loss": 0.881,
+      "step": 3007
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.709493422018036,
+      "learning_rate": 7.908701803934014e-06,
+      "loss": 0.8996,
+      "step": 3008
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8026355525597326,
+      "learning_rate": 7.907285482916067e-06,
+      "loss": 0.9754,
+      "step": 3009
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7377275085786918,
+      "learning_rate": 7.905868809379735e-06,
+      "loss": 0.9349,
+      "step": 3010
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.814802883664392,
+      "learning_rate": 7.904451783496789e-06,
+      "loss": 1.001,
+      "step": 3011
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7756232754016974,
+      "learning_rate": 7.903034405439052e-06,
+      "loss": 0.9522,
+      "step": 3012
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6208715011103083,
+      "learning_rate": 7.901616675378384e-06,
+      "loss": 0.915,
+      "step": 3013
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9284894481767172,
+      "learning_rate": 7.900198593486691e-06,
+      "loss": 0.9175,
+      "step": 3014
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8205622467514608,
+      "learning_rate": 7.89878015993592e-06,
+      "loss": 0.9242,
+      "step": 3015
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7856714094061996,
+      "learning_rate": 7.897361374898063e-06,
+      "loss": 0.951,
+      "step": 3016
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7139198487546219,
+      "learning_rate": 7.895942238545149e-06,
+      "loss": 0.976,
+      "step": 3017
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8101574280143737,
+      "learning_rate": 7.894522751049255e-06,
+      "loss": 0.9676,
+      "step": 3018
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.662349386177393,
+      "learning_rate": 7.893102912582497e-06,
+      "loss": 0.8866,
+      "step": 3019
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.85795273393601,
+      "learning_rate": 7.89168272331704e-06,
+      "loss": 0.9625,
+      "step": 3020
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.673722099552431,
+      "learning_rate": 7.890262183425082e-06,
+      "loss": 0.8752,
+      "step": 3021
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8929166558111947,
+      "learning_rate": 7.888841293078868e-06,
+      "loss": 0.9555,
+      "step": 3022
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8435378458037104,
+      "learning_rate": 7.887420052450689e-06,
+      "loss": 1.0,
+      "step": 3023
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7014365406945766,
+      "learning_rate": 7.885998461712875e-06,
+      "loss": 0.9127,
+      "step": 3024
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7433840655630795,
+      "learning_rate": 7.8845765210378e-06,
+      "loss": 0.9037,
+      "step": 3025
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7595089236613992,
+      "learning_rate": 7.883154230597874e-06,
+      "loss": 0.879,
+      "step": 3026
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7863327264883566,
+      "learning_rate": 7.88173159056556e-06,
+      "loss": 0.9323,
+      "step": 3027
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.558519662020936,
+      "learning_rate": 7.880308601113353e-06,
+      "loss": 0.9308,
+      "step": 3028
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7096900760487257,
+      "learning_rate": 7.8788852624138e-06,
+      "loss": 0.9328,
+      "step": 3029
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6583016778587245,
+      "learning_rate": 7.877461574639485e-06,
+      "loss": 0.9276,
+      "step": 3030
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6991975447417853,
+      "learning_rate": 7.876037537963032e-06,
+      "loss": 0.8942,
+      "step": 3031
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9099292348863601,
+      "learning_rate": 7.874613152557113e-06,
+      "loss": 1.0536,
+      "step": 3032
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8116677911213934,
+      "learning_rate": 7.873188418594438e-06,
+      "loss": 0.9387,
+      "step": 3033
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8340986946491094,
+      "learning_rate": 7.871763336247764e-06,
+      "loss": 0.9085,
+      "step": 3034
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6757022910849841,
+      "learning_rate": 7.870337905689882e-06,
+      "loss": 0.9367,
+      "step": 3035
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.75404006273646,
+      "learning_rate": 7.868912127093638e-06,
+      "loss": 1.0033,
+      "step": 3036
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.738293970179308,
+      "learning_rate": 7.867486000631902e-06,
+      "loss": 0.9429,
+      "step": 3037
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8313258671858774,
+      "learning_rate": 7.866059526477608e-06,
+      "loss": 0.9487,
+      "step": 3038
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7814679939797695,
+      "learning_rate": 7.864632704803712e-06,
+      "loss": 0.8702,
+      "step": 3039
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9006363824207952,
+      "learning_rate": 7.863205535783227e-06,
+      "loss": 0.9519,
+      "step": 3040
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7216243912667661,
+      "learning_rate": 7.861778019589198e-06,
+      "loss": 0.9792,
+      "step": 3041
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9321408733901384,
+      "learning_rate": 7.86035015639472e-06,
+      "loss": 0.9333,
+      "step": 3042
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6825535336425743,
+      "learning_rate": 7.858921946372923e-06,
+      "loss": 0.8985,
+      "step": 3043
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.676270907219242,
+      "learning_rate": 7.857493389696984e-06,
+      "loss": 0.8921,
+      "step": 3044
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6987940422330285,
+      "learning_rate": 7.85606448654012e-06,
+      "loss": 0.9349,
+      "step": 3045
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7525260128011788,
+      "learning_rate": 7.854635237075593e-06,
+      "loss": 0.8494,
+      "step": 3046
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8138621179583427,
+      "learning_rate": 7.8532056414767e-06,
+      "loss": 0.9907,
+      "step": 3047
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.710892024979379,
+      "learning_rate": 7.851775699916785e-06,
+      "loss": 0.9481,
+      "step": 3048
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8192404818735401,
+      "learning_rate": 7.850345412569237e-06,
+      "loss": 0.9347,
+      "step": 3049
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6641725503331903,
+      "learning_rate": 7.84891477960748e-06,
+      "loss": 0.9245,
+      "step": 3050
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7644571012578258,
+      "learning_rate": 7.847483801204984e-06,
+      "loss": 0.8661,
+      "step": 3051
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7231555570385668,
+      "learning_rate": 7.846052477535263e-06,
+      "loss": 1.0261,
+      "step": 3052
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.839528574825421,
+      "learning_rate": 7.844620808771865e-06,
+      "loss": 0.8672,
+      "step": 3053
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.81248959634371,
+      "learning_rate": 7.843188795088386e-06,
+      "loss": 0.9689,
+      "step": 3054
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7149268407978189,
+      "learning_rate": 7.841756436658466e-06,
+      "loss": 0.8044,
+      "step": 3055
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8091462278896084,
+      "learning_rate": 7.84032373365578e-06,
+      "loss": 0.9909,
+      "step": 3056
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6670002264795896,
+      "learning_rate": 7.838890686254049e-06,
+      "loss": 0.8881,
+      "step": 3057
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7291509427913374,
+      "learning_rate": 7.837457294627036e-06,
+      "loss": 0.9315,
+      "step": 3058
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.811722329245556,
+      "learning_rate": 7.836023558948542e-06,
+      "loss": 0.96,
+      "step": 3059
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8580763249830132,
+      "learning_rate": 7.834589479392416e-06,
+      "loss": 0.9669,
+      "step": 3060
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9421614663913902,
+      "learning_rate": 7.833155056132543e-06,
+      "loss": 1.1014,
+      "step": 3061
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9030249652092395,
+      "learning_rate": 7.831720289342853e-06,
+      "loss": 1.0775,
+      "step": 3062
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6742759831384806,
+      "learning_rate": 7.830285179197314e-06,
+      "loss": 0.872,
+      "step": 3063
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7974940983109295,
+      "learning_rate": 7.828849725869939e-06,
+      "loss": 0.902,
+      "step": 3064
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8254239885255714,
+      "learning_rate": 7.827413929534784e-06,
+      "loss": 0.8035,
+      "step": 3065
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8127040522227016,
+      "learning_rate": 7.825977790365941e-06,
+      "loss": 1.0319,
+      "step": 3066
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7526067857957455,
+      "learning_rate": 7.824541308537548e-06,
+      "loss": 0.9501,
+      "step": 3067
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1699302195623185,
+      "learning_rate": 7.823104484223785e-06,
+      "loss": 1.1008,
+      "step": 3068
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6239989971280733,
+      "learning_rate": 7.821667317598871e-06,
+      "loss": 0.9383,
+      "step": 3069
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7431801886636598,
+      "learning_rate": 7.820229808837065e-06,
+      "loss": 0.922,
+      "step": 3070
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8553771505080254,
+      "learning_rate": 7.818791958112673e-06,
+      "loss": 0.8473,
+      "step": 3071
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7421454771045228,
+      "learning_rate": 7.817353765600037e-06,
+      "loss": 0.9114,
+      "step": 3072
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6926811912076758,
+      "learning_rate": 7.815915231473547e-06,
+      "loss": 0.9939,
+      "step": 3073
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7751188139799767,
+      "learning_rate": 7.814476355907626e-06,
+      "loss": 0.9232,
+      "step": 3074
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.679711572446368,
+      "learning_rate": 7.813037139076743e-06,
+      "loss": 0.8926,
+      "step": 3075
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7646301580774613,
+      "learning_rate": 7.811597581155407e-06,
+      "loss": 0.9114,
+      "step": 3076
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7108214067086627,
+      "learning_rate": 7.810157682318174e-06,
+      "loss": 0.9328,
+      "step": 3077
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6909959319833827,
+      "learning_rate": 7.808717442739634e-06,
+      "loss": 0.9166,
+      "step": 3078
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9853421847781072,
+      "learning_rate": 7.807276862594421e-06,
+      "loss": 1.0383,
+      "step": 3079
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6316838634500315,
+      "learning_rate": 7.80583594205721e-06,
+      "loss": 0.9612,
+      "step": 3080
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8448612655194305,
+      "learning_rate": 7.804394681302716e-06,
+      "loss": 0.9005,
+      "step": 3081
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7132070117470934,
+      "learning_rate": 7.8029530805057e-06,
+      "loss": 0.9552,
+      "step": 3082
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.706202567616441,
+      "learning_rate": 7.80151113984096e-06,
+      "loss": 0.9922,
+      "step": 3083
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8007173450210207,
+      "learning_rate": 7.800068859483336e-06,
+      "loss": 1.0346,
+      "step": 3084
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8236116896977254,
+      "learning_rate": 7.798626239607709e-06,
+      "loss": 1.0701,
+      "step": 3085
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7874491025239632,
+      "learning_rate": 7.797183280389002e-06,
+      "loss": 0.9589,
+      "step": 3086
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8064951285402403,
+      "learning_rate": 7.79573998200218e-06,
+      "loss": 1.067,
+      "step": 3087
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8823639928384595,
+      "learning_rate": 7.794296344622246e-06,
+      "loss": 0.8493,
+      "step": 3088
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8391652492898654,
+      "learning_rate": 7.792852368424246e-06,
+      "loss": 1.0862,
+      "step": 3089
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.74349621240828,
+      "learning_rate": 7.791408053583269e-06,
+      "loss": 0.8731,
+      "step": 3090
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.683579091777873,
+      "learning_rate": 7.789963400274442e-06,
+      "loss": 1.0047,
+      "step": 3091
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7869768442234644,
+      "learning_rate": 7.788518408672935e-06,
+      "loss": 0.9433,
+      "step": 3092
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.628110287054426,
+      "learning_rate": 7.787073078953955e-06,
+      "loss": 0.824,
+      "step": 3093
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7954317453355313,
+      "learning_rate": 7.785627411292758e-06,
+      "loss": 0.8562,
+      "step": 3094
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8072226907099973,
+      "learning_rate": 7.784181405864634e-06,
+      "loss": 0.9221,
+      "step": 3095
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.791551113166244,
+      "learning_rate": 7.782735062844915e-06,
+      "loss": 0.9127,
+      "step": 3096
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7477268436088218,
+      "learning_rate": 7.781288382408975e-06,
+      "loss": 0.8178,
+      "step": 3097
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0290862247952075,
+      "learning_rate": 7.779841364732233e-06,
+      "loss": 1.0893,
+      "step": 3098
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7883825532855406,
+      "learning_rate": 7.778394009990139e-06,
+      "loss": 0.9025,
+      "step": 3099
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7288902112124946,
+      "learning_rate": 7.776946318358195e-06,
+      "loss": 0.9436,
+      "step": 3100
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6044119336879408,
+      "learning_rate": 7.775498290011935e-06,
+      "loss": 0.9,
+      "step": 3101
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7781942949351532,
+      "learning_rate": 7.774049925126938e-06,
+      "loss": 0.9097,
+      "step": 3102
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7255158475959271,
+      "learning_rate": 7.772601223878824e-06,
+      "loss": 0.8739,
+      "step": 3103
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7577469498160256,
+      "learning_rate": 7.771152186443254e-06,
+      "loss": 0.9593,
+      "step": 3104
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7554665635875506,
+      "learning_rate": 7.76970281299593e-06,
+      "loss": 0.8615,
+      "step": 3105
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8543523837070472,
+      "learning_rate": 7.768253103712589e-06,
+      "loss": 0.9263,
+      "step": 3106
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.738250216364438,
+      "learning_rate": 7.766803058769014e-06,
+      "loss": 0.9904,
+      "step": 3107
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7029036888315305,
+      "learning_rate": 7.765352678341032e-06,
+      "loss": 0.8262,
+      "step": 3108
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8144149750045406,
+      "learning_rate": 7.763901962604502e-06,
+      "loss": 0.9704,
+      "step": 3109
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7120251507802666,
+      "learning_rate": 7.762450911735333e-06,
+      "loss": 0.9475,
+      "step": 3110
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9764443285987277,
+      "learning_rate": 7.760999525909465e-06,
+      "loss": 1.0597,
+      "step": 3111
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7435416313148284,
+      "learning_rate": 7.759547805302888e-06,
+      "loss": 0.8576,
+      "step": 3112
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8228140515521871,
+      "learning_rate": 7.758095750091625e-06,
+      "loss": 0.8968,
+      "step": 3113
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7465910484954197,
+      "learning_rate": 7.756643360451744e-06,
+      "loss": 1.0077,
+      "step": 3114
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7478058870889237,
+      "learning_rate": 7.75519063655935e-06,
+      "loss": 0.8903,
+      "step": 3115
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9053925793106998,
+      "learning_rate": 7.753737578590594e-06,
+      "loss": 0.9959,
+      "step": 3116
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7503647211441604,
+      "learning_rate": 7.752284186721664e-06,
+      "loss": 0.9205,
+      "step": 3117
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.835444213382237,
+      "learning_rate": 7.750830461128786e-06,
+      "loss": 0.9286,
+      "step": 3118
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7430450891325475,
+      "learning_rate": 7.749376401988232e-06,
+      "loss": 0.9379,
+      "step": 3119
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7059204447083476,
+      "learning_rate": 7.747922009476313e-06,
+      "loss": 0.9411,
+      "step": 3120
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.63132569719763,
+      "learning_rate": 7.746467283769374e-06,
+      "loss": 0.9053,
+      "step": 3121
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7386382057297571,
+      "learning_rate": 7.745012225043809e-06,
+      "loss": 0.9353,
+      "step": 3122
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6416360045814167,
+      "learning_rate": 7.743556833476047e-06,
+      "loss": 0.9523,
+      "step": 3123
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6876012725087925,
+      "learning_rate": 7.742101109242562e-06,
+      "loss": 0.9283,
+      "step": 3124
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6750258492957932,
+      "learning_rate": 7.740645052519863e-06,
+      "loss": 0.9386,
+      "step": 3125
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6636591690049638,
+      "learning_rate": 7.739188663484505e-06,
+      "loss": 0.9364,
+      "step": 3126
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7665127980314514,
+      "learning_rate": 7.737731942313077e-06,
+      "loss": 0.8939,
+      "step": 3127
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7402437497321308,
+      "learning_rate": 7.736274889182211e-06,
+      "loss": 0.9665,
+      "step": 3128
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9464497607908496,
+      "learning_rate": 7.734817504268582e-06,
+      "loss": 1.0719,
+      "step": 3129
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8924382248481586,
+      "learning_rate": 7.733359787748904e-06,
+      "loss": 1.0356,
+      "step": 3130
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7947140282999494,
+      "learning_rate": 7.731901739799927e-06,
+      "loss": 0.8639,
+      "step": 3131
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.772435946349754,
+      "learning_rate": 7.730443360598447e-06,
+      "loss": 0.9656,
+      "step": 3132
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9380020134859561,
+      "learning_rate": 7.728984650321293e-06,
+      "loss": 1.0508,
+      "step": 3133
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8998805782205812,
+      "learning_rate": 7.727525609145345e-06,
+      "loss": 0.9303,
+      "step": 3134
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.671496285415158,
+      "learning_rate": 7.726066237247514e-06,
+      "loss": 0.8983,
+      "step": 3135
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8500614052473616,
+      "learning_rate": 7.72460653480475e-06,
+      "loss": 0.961,
+      "step": 3136
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.742155677864123,
+      "learning_rate": 7.723146501994054e-06,
+      "loss": 0.9057,
+      "step": 3137
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6955271258626596,
+      "learning_rate": 7.721686138992456e-06,
+      "loss": 0.8517,
+      "step": 3138
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.707611666828549,
+      "learning_rate": 7.72022544597703e-06,
+      "loss": 0.9337,
+      "step": 3139
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8852183420617958,
+      "learning_rate": 7.718764423124892e-06,
+      "loss": 0.9356,
+      "step": 3140
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7420148000011606,
+      "learning_rate": 7.717303070613193e-06,
+      "loss": 0.983,
+      "step": 3141
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7522962192274052,
+      "learning_rate": 7.71584138861913e-06,
+      "loss": 0.9595,
+      "step": 3142
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8059922443969585,
+      "learning_rate": 7.714379377319933e-06,
+      "loss": 0.8687,
+      "step": 3143
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6540211210751483,
+      "learning_rate": 7.712917036892881e-06,
+      "loss": 0.8285,
+      "step": 3144
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0732036612055784,
+      "learning_rate": 7.711454367515285e-06,
+      "loss": 1.0854,
+      "step": 3145
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7982009959934304,
+      "learning_rate": 7.709991369364498e-06,
+      "loss": 0.9898,
+      "step": 3146
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7128714391638864,
+      "learning_rate": 7.708528042617915e-06,
+      "loss": 0.9621,
+      "step": 3147
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7909820910298118,
+      "learning_rate": 7.707064387452971e-06,
+      "loss": 0.9693,
+      "step": 3148
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6552508897388354,
+      "learning_rate": 7.705600404047136e-06,
+      "loss": 0.8178,
+      "step": 3149
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8304325673145305,
+      "learning_rate": 7.704136092577924e-06,
+      "loss": 0.9607,
+      "step": 3150
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7694254272745213,
+      "learning_rate": 7.70267145322289e-06,
+      "loss": 0.8832,
+      "step": 3151
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.786653127151144,
+      "learning_rate": 7.701206486159622e-06,
+      "loss": 0.9359,
+      "step": 3152
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7727991978862936,
+      "learning_rate": 7.699741191565758e-06,
+      "loss": 0.9387,
+      "step": 3153
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7958468161483594,
+      "learning_rate": 7.698275569618965e-06,
+      "loss": 0.9005,
+      "step": 3154
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.712626754340491,
+      "learning_rate": 7.696809620496957e-06,
+      "loss": 0.894,
+      "step": 3155
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.873992530855375,
+      "learning_rate": 7.695343344377486e-06,
+      "loss": 0.9205,
+      "step": 3156
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5934773951072805,
+      "learning_rate": 7.693876741438342e-06,
+      "loss": 0.8493,
+      "step": 3157
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7390521007988509,
+      "learning_rate": 7.692409811857357e-06,
+      "loss": 0.9531,
+      "step": 3158
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9037258265048332,
+      "learning_rate": 7.690942555812398e-06,
+      "loss": 1.0875,
+      "step": 3159
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6715584526933467,
+      "learning_rate": 7.689474973481379e-06,
+      "loss": 0.9829,
+      "step": 3160
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7727309253045949,
+      "learning_rate": 7.688007065042246e-06,
+      "loss": 0.9582,
+      "step": 3161
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7665136086372766,
+      "learning_rate": 7.686538830672988e-06,
+      "loss": 0.8438,
+      "step": 3162
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6549521823296243,
+      "learning_rate": 7.685070270551637e-06,
+      "loss": 0.9558,
+      "step": 3163
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8109040193701325,
+      "learning_rate": 7.68360138485626e-06,
+      "loss": 0.8631,
+      "step": 3164
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8521614987644577,
+      "learning_rate": 7.682132173764962e-06,
+      "loss": 0.8477,
+      "step": 3165
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9669715784235244,
+      "learning_rate": 7.68066263745589e-06,
+      "loss": 0.9767,
+      "step": 3166
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7045841187736768,
+      "learning_rate": 7.679192776107232e-06,
+      "loss": 0.8847,
+      "step": 3167
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7104221102304642,
+      "learning_rate": 7.677722589897214e-06,
+      "loss": 0.8685,
+      "step": 3168
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6007350121258022,
+      "learning_rate": 7.676252079004101e-06,
+      "loss": 0.9054,
+      "step": 3169
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8539976120946018,
+      "learning_rate": 7.674781243606197e-06,
+      "loss": 0.908,
+      "step": 3170
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.679287217833275,
+      "learning_rate": 7.673310083881844e-06,
+      "loss": 0.9406,
+      "step": 3171
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6239845540371451,
+      "learning_rate": 7.67183860000943e-06,
+      "loss": 0.9131,
+      "step": 3172
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6800973748451429,
+      "learning_rate": 7.670366792167371e-06,
+      "loss": 0.9436,
+      "step": 3173
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8545596638374644,
+      "learning_rate": 7.668894660534135e-06,
+      "loss": 0.9884,
+      "step": 3174
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6730062293627086,
+      "learning_rate": 7.66742220528822e-06,
+      "loss": 0.7663,
+      "step": 3175
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7288454628087877,
+      "learning_rate": 7.665949426608164e-06,
+      "loss": 0.907,
+      "step": 3176
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7209141364750085,
+      "learning_rate": 7.664476324672552e-06,
+      "loss": 0.9329,
+      "step": 3177
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.818573308681355,
+      "learning_rate": 7.663002899659999e-06,
+      "loss": 0.997,
+      "step": 3178
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.700003323472943,
+      "learning_rate": 7.661529151749163e-06,
+      "loss": 0.8711,
+      "step": 3179
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8419405226913856,
+      "learning_rate": 7.660055081118743e-06,
+      "loss": 0.8962,
+      "step": 3180
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7436673021173064,
+      "learning_rate": 7.658580687947473e-06,
+      "loss": 0.8893,
+      "step": 3181
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.0749518377331198,
+      "learning_rate": 7.657105972414129e-06,
+      "loss": 0.8989,
+      "step": 3182
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5839894292687773,
+      "learning_rate": 7.655630934697524e-06,
+      "loss": 0.9468,
+      "step": 3183
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9005557202354577,
+      "learning_rate": 7.654155574976516e-06,
+      "loss": 0.9651,
+      "step": 3184
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7745447881853755,
+      "learning_rate": 7.652679893429993e-06,
+      "loss": 0.9408,
+      "step": 3185
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.694981590177117,
+      "learning_rate": 7.651203890236886e-06,
+      "loss": 0.9999,
+      "step": 3186
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7053717191532503,
+      "learning_rate": 7.64972756557617e-06,
+      "loss": 0.9355,
+      "step": 3187
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8275926477329423,
+      "learning_rate": 7.64825091962685e-06,
+      "loss": 0.8875,
+      "step": 3188
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6957161672310734,
+      "learning_rate": 7.646773952567976e-06,
+      "loss": 0.9064,
+      "step": 3189
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6682651430677493,
+      "learning_rate": 7.645296664578635e-06,
+      "loss": 0.8883,
+      "step": 3190
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8167660103225869,
+      "learning_rate": 7.643819055837955e-06,
+      "loss": 0.9698,
+      "step": 3191
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8470024637425095,
+      "learning_rate": 7.6423411265251e-06,
+      "loss": 0.8651,
+      "step": 3192
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6943686315074165,
+      "learning_rate": 7.640862876819272e-06,
+      "loss": 0.8562,
+      "step": 3193
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7079257970414865,
+      "learning_rate": 7.639384306899716e-06,
+      "loss": 0.9185,
+      "step": 3194
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9606980082804817,
+      "learning_rate": 7.637905416945716e-06,
+      "loss": 1.0351,
+      "step": 3195
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9095499546603583,
+      "learning_rate": 7.636426207136588e-06,
+      "loss": 0.9605,
+      "step": 3196
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7839643222703667,
+      "learning_rate": 7.634946677651694e-06,
+      "loss": 0.8723,
+      "step": 3197
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.737502651847597,
+      "learning_rate": 7.633466828670431e-06,
+      "loss": 0.8967,
+      "step": 3198
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8221707882765172,
+      "learning_rate": 7.631986660372235e-06,
+      "loss": 0.9221,
+      "step": 3199
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8719886660041634,
+      "learning_rate": 7.630506172936583e-06,
+      "loss": 1.0559,
+      "step": 3200
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6636267825452606,
+      "learning_rate": 7.6290253665429905e-06,
+      "loss": 0.961,
+      "step": 3201
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8605510747131204,
+      "learning_rate": 7.627544241371006e-06,
+      "loss": 1.0321,
+      "step": 3202
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6999781964765814,
+      "learning_rate": 7.6260627976002255e-06,
+      "loss": 0.8288,
+      "step": 3203
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8029288377285624,
+      "learning_rate": 7.624581035410277e-06,
+      "loss": 0.8991,
+      "step": 3204
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8623788673407502,
+      "learning_rate": 7.6230989549808296e-06,
+      "loss": 0.9363,
+      "step": 3205
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7993634371626448,
+      "learning_rate": 7.621616556491591e-06,
+      "loss": 0.9252,
+      "step": 3206
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7789923061228887,
+      "learning_rate": 7.620133840122306e-06,
+      "loss": 0.9313,
+      "step": 3207
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7690237035386838,
+      "learning_rate": 7.618650806052761e-06,
+      "loss": 0.9551,
+      "step": 3208
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.733476776135675,
+      "learning_rate": 7.617167454462777e-06,
+      "loss": 0.8977,
+      "step": 3209
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7521930159293178,
+      "learning_rate": 7.615683785532218e-06,
+      "loss": 0.9803,
+      "step": 3210
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6794386201277403,
+      "learning_rate": 7.614199799440981e-06,
+      "loss": 0.9543,
+      "step": 3211
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8941148988641795,
+      "learning_rate": 7.612715496369007e-06,
+      "loss": 1.0675,
+      "step": 3212
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.788758815051161,
+      "learning_rate": 7.61123087649627e-06,
+      "loss": 1.0082,
+      "step": 3213
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7540886054378615,
+      "learning_rate": 7.609745940002786e-06,
+      "loss": 0.8425,
+      "step": 3214
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8595054254276828,
+      "learning_rate": 7.608260687068611e-06,
+      "loss": 0.8342,
+      "step": 3215
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6745912398987448,
+      "learning_rate": 7.606775117873837e-06,
+      "loss": 0.9142,
+      "step": 3216
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7913121430235512,
+      "learning_rate": 7.605289232598592e-06,
+      "loss": 0.8834,
+      "step": 3217
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.818131893843938,
+      "learning_rate": 7.603803031423046e-06,
+      "loss": 0.954,
+      "step": 3218
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.879072223539242,
+      "learning_rate": 7.602316514527405e-06,
+      "loss": 0.9507,
+      "step": 3219
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7415104572545848,
+      "learning_rate": 7.600829682091915e-06,
+      "loss": 0.9244,
+      "step": 3220
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8296540841332827,
+      "learning_rate": 7.599342534296861e-06,
+      "loss": 0.88,
+      "step": 3221
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6980749837098486,
+      "learning_rate": 7.597855071322562e-06,
+      "loss": 0.9143,
+      "step": 3222
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8152405207758098,
+      "learning_rate": 7.59636729334938e-06,
+      "loss": 0.9121,
+      "step": 3223
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7965257454625034,
+      "learning_rate": 7.594879200557712e-06,
+      "loss": 0.8219,
+      "step": 3224
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6844182892925543,
+      "learning_rate": 7.593390793127997e-06,
+      "loss": 0.8775,
+      "step": 3225
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7020023385444039,
+      "learning_rate": 7.591902071240705e-06,
+      "loss": 0.8523,
+      "step": 3226
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7076335555207474,
+      "learning_rate": 7.590413035076354e-06,
+      "loss": 0.9194,
+      "step": 3227
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7963068200928247,
+      "learning_rate": 7.5889236848154905e-06,
+      "loss": 0.8944,
+      "step": 3228
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7777189488158993,
+      "learning_rate": 7.587434020638704e-06,
+      "loss": 0.8987,
+      "step": 3229
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8386141432948673,
+      "learning_rate": 7.585944042726623e-06,
+      "loss": 0.8552,
+      "step": 3230
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8038523386424994,
+      "learning_rate": 7.584453751259912e-06,
+      "loss": 0.9377,
+      "step": 3231
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6768570747188016,
+      "learning_rate": 7.582963146419273e-06,
+      "loss": 0.888,
+      "step": 3232
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.811970696976311,
+      "learning_rate": 7.581472228385448e-06,
+      "loss": 0.9752,
+      "step": 3233
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7468139178931152,
+      "learning_rate": 7.579980997339215e-06,
+      "loss": 0.9799,
+      "step": 3234
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9462443790179987,
+      "learning_rate": 7.5784894534613925e-06,
+      "loss": 1.0922,
+      "step": 3235
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.752540486288468,
+      "learning_rate": 7.576997596932834e-06,
+      "loss": 0.866,
+      "step": 3236
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8034097010203642,
+      "learning_rate": 7.575505427934433e-06,
+      "loss": 0.9398,
+      "step": 3237
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7524342166948448,
+      "learning_rate": 7.5740129466471204e-06,
+      "loss": 0.9187,
+      "step": 3238
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6865178965199388,
+      "learning_rate": 7.572520153251862e-06,
+      "loss": 0.9509,
+      "step": 3239
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7637744852630872,
+      "learning_rate": 7.571027047929667e-06,
+      "loss": 0.9478,
+      "step": 3240
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7002997387038798,
+      "learning_rate": 7.5695336308615785e-06,
+      "loss": 0.958,
+      "step": 3241
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7321206683135355,
+      "learning_rate": 7.56803990222868e-06,
+      "loss": 0.8852,
+      "step": 3242
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.75358728466011,
+      "learning_rate": 7.56654586221209e-06,
+      "loss": 0.9307,
+      "step": 3243
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7529641677278303,
+      "learning_rate": 7.565051510992964e-06,
+      "loss": 0.8728,
+      "step": 3244
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.608867549877374,
+      "learning_rate": 7.563556848752502e-06,
+      "loss": 0.9105,
+      "step": 3245
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7456707325625758,
+      "learning_rate": 7.5620618756719325e-06,
+      "loss": 0.9402,
+      "step": 3246
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7613599751227094,
+      "learning_rate": 7.560566591932526e-06,
+      "loss": 0.8894,
+      "step": 3247
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7069245997247633,
+      "learning_rate": 7.559070997715596e-06,
+      "loss": 0.916,
+      "step": 3248
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8425068464616408,
+      "learning_rate": 7.557575093202484e-06,
+      "loss": 0.8975,
+      "step": 3249
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8165796166336952,
+      "learning_rate": 7.556078878574573e-06,
+      "loss": 0.9078,
+      "step": 3250
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.752575378113844,
+      "learning_rate": 7.554582354013288e-06,
+      "loss": 0.9979,
+      "step": 3251
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6976734361297443,
+      "learning_rate": 7.553085519700085e-06,
+      "loss": 0.8914,
+      "step": 3252
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.685476970027444,
+      "learning_rate": 7.551588375816461e-06,
+      "loss": 0.9057,
+      "step": 3253
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6646624663466865,
+      "learning_rate": 7.550090922543949e-06,
+      "loss": 0.9255,
+      "step": 3254
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.671410598278676,
+      "learning_rate": 7.54859316006412e-06,
+      "loss": 0.9241,
+      "step": 3255
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6576226515766672,
+      "learning_rate": 7.547095088558586e-06,
+      "loss": 0.8211,
+      "step": 3256
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6704685267339927,
+      "learning_rate": 7.545596708208991e-06,
+      "loss": 1.0216,
+      "step": 3257
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.69542926329254,
+      "learning_rate": 7.544098019197018e-06,
+      "loss": 0.8835,
+      "step": 3258
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5936605632516918,
+      "learning_rate": 7.542599021704391e-06,
+      "loss": 0.8803,
+      "step": 3259
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.770487899666137,
+      "learning_rate": 7.541099715912867e-06,
+      "loss": 0.8214,
+      "step": 3260
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7477233064981108,
+      "learning_rate": 7.539600102004241e-06,
+      "loss": 0.9945,
+      "step": 3261
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7934872611365564,
+      "learning_rate": 7.538100180160347e-06,
+      "loss": 1.0322,
+      "step": 3262
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7590684351205144,
+      "learning_rate": 7.5365999505630575e-06,
+      "loss": 0.9198,
+      "step": 3263
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7633529656090035,
+      "learning_rate": 7.5350994133942764e-06,
+      "loss": 0.8568,
+      "step": 3264
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9301144476852095,
+      "learning_rate": 7.533598568835954e-06,
+      "loss": 0.8365,
+      "step": 3265
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.791926145009191,
+      "learning_rate": 7.53209741707007e-06,
+      "loss": 0.891,
+      "step": 3266
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7380834991297436,
+      "learning_rate": 7.530595958278644e-06,
+      "loss": 0.9147,
+      "step": 3267
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6494701761725241,
+      "learning_rate": 7.529094192643733e-06,
+      "loss": 0.8571,
+      "step": 3268
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7473442852751488,
+      "learning_rate": 7.527592120347433e-06,
+      "loss": 0.9325,
+      "step": 3269
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.75144451661381,
+      "learning_rate": 7.526089741571876e-06,
+      "loss": 0.9449,
+      "step": 3270
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7281097186425467,
+      "learning_rate": 7.524587056499227e-06,
+      "loss": 0.9041,
+      "step": 3271
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6590454150434881,
+      "learning_rate": 7.523084065311695e-06,
+      "loss": 0.8983,
+      "step": 3272
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8579740345934426,
+      "learning_rate": 7.521580768191521e-06,
+      "loss": 0.9053,
+      "step": 3273
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7470045275489403,
+      "learning_rate": 7.520077165320986e-06,
+      "loss": 0.887,
+      "step": 3274
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6379848040178568,
+      "learning_rate": 7.518573256882407e-06,
+      "loss": 0.9262,
+      "step": 3275
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7174552648963053,
+      "learning_rate": 7.517069043058136e-06,
+      "loss": 0.9276,
+      "step": 3276
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.580852153288313,
+      "learning_rate": 7.515564524030568e-06,
+      "loss": 0.8904,
+      "step": 3277
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7244884890473644,
+      "learning_rate": 7.514059699982128e-06,
+      "loss": 0.9085,
+      "step": 3278
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6069249931455998,
+      "learning_rate": 7.512554571095282e-06,
+      "loss": 0.8914,
+      "step": 3279
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0577805211931215,
+      "learning_rate": 7.511049137552533e-06,
+      "loss": 1.0919,
+      "step": 3280
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7551151519644463,
+      "learning_rate": 7.509543399536419e-06,
+      "loss": 0.8525,
+      "step": 3281
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7273119192195896,
+      "learning_rate": 7.508037357229515e-06,
+      "loss": 0.9829,
+      "step": 3282
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.697972290513672,
+      "learning_rate": 7.506531010814435e-06,
+      "loss": 0.887,
+      "step": 3283
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.722066702915343,
+      "learning_rate": 7.505024360473829e-06,
+      "loss": 0.8507,
+      "step": 3284
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9046665262633369,
+      "learning_rate": 7.503517406390385e-06,
+      "loss": 1.0987,
+      "step": 3285
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8928219859526223,
+      "learning_rate": 7.502010148746821e-06,
+      "loss": 1.0615,
+      "step": 3286
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7376419223983317,
+      "learning_rate": 7.5005025877259045e-06,
+      "loss": 0.9652,
+      "step": 3287
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8014024314675128,
+      "learning_rate": 7.4989947235104264e-06,
+      "loss": 0.9315,
+      "step": 3288
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9117966804886433,
+      "learning_rate": 7.497486556283222e-06,
+      "loss": 0.9107,
+      "step": 3289
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.73640326340925,
+      "learning_rate": 7.495978086227164e-06,
+      "loss": 0.9514,
+      "step": 3290
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6915311699069033,
+      "learning_rate": 7.494469313525159e-06,
+      "loss": 0.8593,
+      "step": 3291
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7450022154061327,
+      "learning_rate": 7.4929602383601475e-06,
+      "loss": 0.8731,
+      "step": 3292
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7102579278956707,
+      "learning_rate": 7.491450860915115e-06,
+      "loss": 0.9368,
+      "step": 3293
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.715594217374372,
+      "learning_rate": 7.489941181373075e-06,
+      "loss": 1.0199,
+      "step": 3294
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8003577081979956,
+      "learning_rate": 7.488431199917081e-06,
+      "loss": 0.9638,
+      "step": 3295
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.751998317382199,
+      "learning_rate": 7.486920916730228e-06,
+      "loss": 0.868,
+      "step": 3296
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.756928767150894,
+      "learning_rate": 7.48541033199564e-06,
+      "loss": 1.0205,
+      "step": 3297
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6550424964521924,
+      "learning_rate": 7.4838994458964785e-06,
+      "loss": 0.9479,
+      "step": 3298
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6883032413304706,
+      "learning_rate": 7.482388258615947e-06,
+      "loss": 0.91,
+      "step": 3299
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7481837818496226,
+      "learning_rate": 7.480876770337281e-06,
+      "loss": 0.9805,
+      "step": 3300
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8263325115939446,
+      "learning_rate": 7.479364981243754e-06,
+      "loss": 0.8588,
+      "step": 3301
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7572162938301292,
+      "learning_rate": 7.477852891518675e-06,
+      "loss": 0.8628,
+      "step": 3302
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8270750998523413,
+      "learning_rate": 7.4763405013453906e-06,
+      "loss": 0.9744,
+      "step": 3303
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0534308597746713,
+      "learning_rate": 7.474827810907284e-06,
+      "loss": 1.0448,
+      "step": 3304
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7846339277389076,
+      "learning_rate": 7.473314820387772e-06,
+      "loss": 0.7901,
+      "step": 3305
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7110302417972643,
+      "learning_rate": 7.471801529970312e-06,
+      "loss": 0.9333,
+      "step": 3306
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8347411262180946,
+      "learning_rate": 7.470287939838394e-06,
+      "loss": 0.9581,
+      "step": 3307
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6973523481093706,
+      "learning_rate": 7.468774050175548e-06,
+      "loss": 0.9127,
+      "step": 3308
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8961071483694014,
+      "learning_rate": 7.467259861165335e-06,
+      "loss": 0.954,
+      "step": 3309
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7261615033362823,
+      "learning_rate": 7.46574537299136e-06,
+      "loss": 0.8639,
+      "step": 3310
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6380139865200996,
+      "learning_rate": 7.464230585837257e-06,
+      "loss": 0.888,
+      "step": 3311
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6555589446320618,
+      "learning_rate": 7.462715499886701e-06,
+      "loss": 0.8883,
+      "step": 3312
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6983118008345057,
+      "learning_rate": 7.4612001153234e-06,
+      "loss": 0.9286,
+      "step": 3313
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7605505240987802,
+      "learning_rate": 7.4596844323311e-06,
+      "loss": 0.8708,
+      "step": 3314
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6871151868612113,
+      "learning_rate": 7.458168451093583e-06,
+      "loss": 0.8482,
+      "step": 3315
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7518828427566988,
+      "learning_rate": 7.456652171794666e-06,
+      "loss": 0.8993,
+      "step": 3316
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6085552723407484,
+      "learning_rate": 7.455135594618204e-06,
+      "loss": 0.9141,
+      "step": 3317
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9195026846731007,
+      "learning_rate": 7.453618719748087e-06,
+      "loss": 0.8732,
+      "step": 3318
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7029781721289585,
+      "learning_rate": 7.452101547368242e-06,
+      "loss": 0.998,
+      "step": 3319
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8748497300245488,
+      "learning_rate": 7.450584077662628e-06,
+      "loss": 0.962,
+      "step": 3320
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8186491422533853,
+      "learning_rate": 7.449066310815249e-06,
+      "loss": 0.9288,
+      "step": 3321
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7461666374926603,
+      "learning_rate": 7.447548247010137e-06,
+      "loss": 0.9137,
+      "step": 3322
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7520519312904304,
+      "learning_rate": 7.4460298864313606e-06,
+      "loss": 0.9748,
+      "step": 3323
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.685064117045246,
+      "learning_rate": 7.444511229263027e-06,
+      "loss": 0.8902,
+      "step": 3324
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7676762665251464,
+      "learning_rate": 7.442992275689281e-06,
+      "loss": 0.9702,
+      "step": 3325
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9013027778275644,
+      "learning_rate": 7.441473025894298e-06,
+      "loss": 0.9272,
+      "step": 3326
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6556882740715895,
+      "learning_rate": 7.439953480062294e-06,
+      "loss": 0.9084,
+      "step": 3327
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8016120860563096,
+      "learning_rate": 7.43843363837752e-06,
+      "loss": 1.0024,
+      "step": 3328
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0857740614042077,
+      "learning_rate": 7.436913501024259e-06,
+      "loss": 1.0428,
+      "step": 3329
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7906442762242527,
+      "learning_rate": 7.435393068186836e-06,
+      "loss": 0.921,
+      "step": 3330
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.785233540155477,
+      "learning_rate": 7.4338723400496075e-06,
+      "loss": 0.8981,
+      "step": 3331
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6413606470544102,
+      "learning_rate": 7.432351316796964e-06,
+      "loss": 0.9126,
+      "step": 3332
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7899005083175799,
+      "learning_rate": 7.430829998613342e-06,
+      "loss": 0.8799,
+      "step": 3333
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7768760339940515,
+      "learning_rate": 7.429308385683199e-06,
+      "loss": 0.7923,
+      "step": 3334
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8885875206253508,
+      "learning_rate": 7.427786478191041e-06,
+      "loss": 1.0668,
+      "step": 3335
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.759935724746908,
+      "learning_rate": 7.426264276321402e-06,
+      "loss": 0.9477,
+      "step": 3336
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7616315467701726,
+      "learning_rate": 7.424741780258855e-06,
+      "loss": 0.8702,
+      "step": 3337
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.830856127209951,
+      "learning_rate": 7.423218990188008e-06,
+      "loss": 0.955,
+      "step": 3338
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8049959649239131,
+      "learning_rate": 7.421695906293504e-06,
+      "loss": 0.9023,
+      "step": 3339
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8838853878118114,
+      "learning_rate": 7.420172528760022e-06,
+      "loss": 1.0727,
+      "step": 3340
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7480358579933728,
+      "learning_rate": 7.418648857772279e-06,
+      "loss": 0.9406,
+      "step": 3341
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6461169657610413,
+      "learning_rate": 7.417124893515022e-06,
+      "loss": 0.9234,
+      "step": 3342
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.763522855230025,
+      "learning_rate": 7.4156006361730395e-06,
+      "loss": 0.9633,
+      "step": 3343
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8635987906876557,
+      "learning_rate": 7.414076085931152e-06,
+      "loss": 0.8921,
+      "step": 3344
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6694079108516462,
+      "learning_rate": 7.4125512429742166e-06,
+      "loss": 0.8498,
+      "step": 3345
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9387554904060442,
+      "learning_rate": 7.411026107487123e-06,
+      "loss": 0.9155,
+      "step": 3346
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8332840469538487,
+      "learning_rate": 7.409500679654805e-06,
+      "loss": 0.9309,
+      "step": 3347
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6873794724323066,
+      "learning_rate": 7.407974959662222e-06,
+      "loss": 0.8686,
+      "step": 3348
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7726669245130953,
+      "learning_rate": 7.406448947694374e-06,
+      "loss": 1.0045,
+      "step": 3349
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5891647834478793,
+      "learning_rate": 7.404922643936294e-06,
+      "loss": 0.8308,
+      "step": 3350
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8354347368271844,
+      "learning_rate": 7.403396048573052e-06,
+      "loss": 0.896,
+      "step": 3351
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7301188378539172,
+      "learning_rate": 7.4018691617897545e-06,
+      "loss": 0.8815,
+      "step": 3352
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7468554946972616,
+      "learning_rate": 7.40034198377154e-06,
+      "loss": 0.9513,
+      "step": 3353
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.903649293058784,
+      "learning_rate": 7.398814514703585e-06,
+      "loss": 0.9608,
+      "step": 3354
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.867873737546824,
+      "learning_rate": 7.397286754771099e-06,
+      "loss": 0.9427,
+      "step": 3355
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7182649002592478,
+      "learning_rate": 7.39575870415933e-06,
+      "loss": 0.9124,
+      "step": 3356
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7457813898873142,
+      "learning_rate": 7.394230363053558e-06,
+      "loss": 0.8758,
+      "step": 3357
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.777804667996553,
+      "learning_rate": 7.3927017316391024e-06,
+      "loss": 0.9732,
+      "step": 3358
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0199925982213331,
+      "learning_rate": 7.391172810101311e-06,
+      "loss": 1.0812,
+      "step": 3359
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7367887303573601,
+      "learning_rate": 7.389643598625574e-06,
+      "loss": 0.852,
+      "step": 3360
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6509855094140875,
+      "learning_rate": 7.388114097397312e-06,
+      "loss": 0.9238,
+      "step": 3361
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8070664565500492,
+      "learning_rate": 7.386584306601982e-06,
+      "loss": 0.8389,
+      "step": 3362
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.692140573828619,
+      "learning_rate": 7.385054226425077e-06,
+      "loss": 0.8075,
+      "step": 3363
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.836891511694827,
+      "learning_rate": 7.383523857052125e-06,
+      "loss": 0.9303,
+      "step": 3364
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.807655871308822,
+      "learning_rate": 7.381993198668689e-06,
+      "loss": 0.9394,
+      "step": 3365
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9761758390026731,
+      "learning_rate": 7.380462251460364e-06,
+      "loss": 1.0469,
+      "step": 3366
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.774982463810186,
+      "learning_rate": 7.378931015612786e-06,
+      "loss": 0.8895,
+      "step": 3367
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7597933204675105,
+      "learning_rate": 7.377399491311619e-06,
+      "loss": 0.9243,
+      "step": 3368
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7683208501624577,
+      "learning_rate": 7.375867678742567e-06,
+      "loss": 0.8965,
+      "step": 3369
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7933521337735707,
+      "learning_rate": 7.374335578091372e-06,
+      "loss": 0.9665,
+      "step": 3370
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8697269842382845,
+      "learning_rate": 7.372803189543799e-06,
+      "loss": 1.036,
+      "step": 3371
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6992990261014262,
+      "learning_rate": 7.37127051328566e-06,
+      "loss": 0.8811,
+      "step": 3372
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6990455614411435,
+      "learning_rate": 7.369737549502797e-06,
+      "loss": 0.9821,
+      "step": 3373
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7864955568474217,
+      "learning_rate": 7.368204298381085e-06,
+      "loss": 0.8734,
+      "step": 3374
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.604090882580611,
+      "learning_rate": 7.366670760106439e-06,
+      "loss": 0.8986,
+      "step": 3375
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6866980777488114,
+      "learning_rate": 7.365136934864804e-06,
+      "loss": 0.9234,
+      "step": 3376
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7553036531217971,
+      "learning_rate": 7.3636028228421595e-06,
+      "loss": 0.982,
+      "step": 3377
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8151438967238653,
+      "learning_rate": 7.3620684242245264e-06,
+      "loss": 1.0423,
+      "step": 3378
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.730724790126759,
+      "learning_rate": 7.360533739197952e-06,
+      "loss": 0.9448,
+      "step": 3379
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6941573262485696,
+      "learning_rate": 7.358998767948525e-06,
+      "loss": 0.9559,
+      "step": 3380
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6157727713930206,
+      "learning_rate": 7.357463510662365e-06,
+      "loss": 0.8885,
+      "step": 3381
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6743865004593845,
+      "learning_rate": 7.355927967525626e-06,
+      "loss": 0.923,
+      "step": 3382
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6269144548954249,
+      "learning_rate": 7.354392138724499e-06,
+      "loss": 0.8998,
+      "step": 3383
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7212943410745358,
+      "learning_rate": 7.352856024445208e-06,
+      "loss": 0.8217,
+      "step": 3384
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7356861877373753,
+      "learning_rate": 7.351319624874013e-06,
+      "loss": 0.9484,
+      "step": 3385
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.795463863054259,
+      "learning_rate": 7.349782940197208e-06,
+      "loss": 0.9545,
+      "step": 3386
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.683754431552234,
+      "learning_rate": 7.34824597060112e-06,
+      "loss": 0.896,
+      "step": 3387
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.680330292599506,
+      "learning_rate": 7.34670871627211e-06,
+      "loss": 0.9006,
+      "step": 3388
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9670086825680536,
+      "learning_rate": 7.34517117739658e-06,
+      "loss": 1.0633,
+      "step": 3389
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7798192906324743,
+      "learning_rate": 7.343633354160961e-06,
+      "loss": 1.0078,
+      "step": 3390
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7490574833579196,
+      "learning_rate": 7.342095246751717e-06,
+      "loss": 0.9926,
+      "step": 3391
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6377152849727812,
+      "learning_rate": 7.340556855355349e-06,
+      "loss": 0.9025,
+      "step": 3392
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.768482964825861,
+      "learning_rate": 7.339018180158395e-06,
+      "loss": 0.9355,
+      "step": 3393
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6231583351312866,
+      "learning_rate": 7.33747922134742e-06,
+      "loss": 0.9198,
+      "step": 3394
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8476246861894299,
+      "learning_rate": 7.335939979109032e-06,
+      "loss": 0.8514,
+      "step": 3395
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6178812077358735,
+      "learning_rate": 7.334400453629869e-06,
+      "loss": 0.8521,
+      "step": 3396
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6684122014573253,
+      "learning_rate": 7.332860645096604e-06,
+      "loss": 0.8306,
+      "step": 3397
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9369820314050638,
+      "learning_rate": 7.331320553695942e-06,
+      "loss": 1.0569,
+      "step": 3398
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6376415899790333,
+      "learning_rate": 7.329780179614625e-06,
+      "loss": 0.8951,
+      "step": 3399
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7392488393014507,
+      "learning_rate": 7.328239523039431e-06,
+      "loss": 0.9804,
+      "step": 3400
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7282752591342359,
+      "learning_rate": 7.3266985841571666e-06,
+      "loss": 0.9047,
+      "step": 3401
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7167152986144596,
+      "learning_rate": 7.325157363154678e-06,
+      "loss": 0.8986,
+      "step": 3402
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7826895304474335,
+      "learning_rate": 7.323615860218844e-06,
+      "loss": 0.9639,
+      "step": 3403
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.649060803257622,
+      "learning_rate": 7.322074075536575e-06,
+      "loss": 0.8502,
+      "step": 3404
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7382089958096956,
+      "learning_rate": 7.320532009294818e-06,
+      "loss": 0.8667,
+      "step": 3405
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.750635600304196,
+      "learning_rate": 7.3189896616805565e-06,
+      "loss": 0.9503,
+      "step": 3406
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7102984071284362,
+      "learning_rate": 7.317447032880804e-06,
+      "loss": 0.9423,
+      "step": 3407
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7459375123075902,
+      "learning_rate": 7.315904123082609e-06,
+      "loss": 0.9179,
+      "step": 3408
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8225423335822546,
+      "learning_rate": 7.314360932473054e-06,
+      "loss": 0.9606,
+      "step": 3409
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9239747391040125,
+      "learning_rate": 7.312817461239259e-06,
+      "loss": 0.898,
+      "step": 3410
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7327283654368293,
+      "learning_rate": 7.311273709568371e-06,
+      "loss": 0.9005,
+      "step": 3411
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9483837994623496,
+      "learning_rate": 7.309729677647579e-06,
+      "loss": 0.9513,
+      "step": 3412
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9164921195129382,
+      "learning_rate": 7.3081853656641e-06,
+      "loss": 1.026,
+      "step": 3413
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8797180029944156,
+      "learning_rate": 7.306640773805188e-06,
+      "loss": 1.0507,
+      "step": 3414
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6616115847168296,
+      "learning_rate": 7.30509590225813e-06,
+      "loss": 0.9331,
+      "step": 3415
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.705659155990624,
+      "learning_rate": 7.303550751210247e-06,
+      "loss": 0.8546,
+      "step": 3416
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6976483916926663,
+      "learning_rate": 7.302005320848894e-06,
+      "loss": 0.9457,
+      "step": 3417
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7172659978139384,
+      "learning_rate": 7.300459611361461e-06,
+      "loss": 0.8174,
+      "step": 3418
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.050136016099378,
+      "learning_rate": 7.298913622935366e-06,
+      "loss": 1.0902,
+      "step": 3419
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.776453849163029,
+      "learning_rate": 7.297367355758072e-06,
+      "loss": 0.9521,
+      "step": 3420
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8291316049357584,
+      "learning_rate": 7.295820810017065e-06,
+      "loss": 0.8535,
+      "step": 3421
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.801293992009186,
+      "learning_rate": 7.294273985899868e-06,
+      "loss": 0.9103,
+      "step": 3422
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6750016380200354,
+      "learning_rate": 7.292726883594043e-06,
+      "loss": 0.8761,
+      "step": 3423
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7792656006823968,
+      "learning_rate": 7.291179503287178e-06,
+      "loss": 0.8825,
+      "step": 3424
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7017782707414377,
+      "learning_rate": 7.2896318451668975e-06,
+      "loss": 0.9076,
+      "step": 3425
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6493662991980782,
+      "learning_rate": 7.288083909420866e-06,
+      "loss": 0.9552,
+      "step": 3426
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9728495995493363,
+      "learning_rate": 7.28653569623677e-06,
+      "loss": 1.0732,
+      "step": 3427
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6989569772599058,
+      "learning_rate": 7.284987205802338e-06,
+      "loss": 0.8616,
+      "step": 3428
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7825258667492783,
+      "learning_rate": 7.28343843830533e-06,
+      "loss": 0.9139,
+      "step": 3429
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7364359931910722,
+      "learning_rate": 7.281889393933539e-06,
+      "loss": 0.8987,
+      "step": 3430
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6970828888922183,
+      "learning_rate": 7.280340072874792e-06,
+      "loss": 0.9766,
+      "step": 3431
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7681464261981494,
+      "learning_rate": 7.27879047531695e-06,
+      "loss": 0.9512,
+      "step": 3432
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7880879198559911,
+      "learning_rate": 7.277240601447908e-06,
+      "loss": 0.8701,
+      "step": 3433
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7315916656125943,
+      "learning_rate": 7.27569045145559e-06,
+      "loss": 0.9284,
+      "step": 3434
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8191070206057929,
+      "learning_rate": 7.27414002552796e-06,
+      "loss": 0.9299,
+      "step": 3435
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9153661427624924,
+      "learning_rate": 7.272589323853012e-06,
+      "loss": 0.8971,
+      "step": 3436
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6866005279199747,
+      "learning_rate": 7.271038346618774e-06,
+      "loss": 0.8737,
+      "step": 3437
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7364114128552566,
+      "learning_rate": 7.2694870940133065e-06,
+      "loss": 0.8989,
+      "step": 3438
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8170618560555738,
+      "learning_rate": 7.267935566224705e-06,
+      "loss": 0.9275,
+      "step": 3439
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6438185755896468,
+      "learning_rate": 7.2663837634410976e-06,
+      "loss": 0.85,
+      "step": 3440
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8391216782632975,
+      "learning_rate": 7.2648316858506455e-06,
+      "loss": 0.9178,
+      "step": 3441
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.763562830263875,
+      "learning_rate": 7.263279333641542e-06,
+      "loss": 0.8728,
+      "step": 3442
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9055553768871036,
+      "learning_rate": 7.261726707002017e-06,
+      "loss": 0.9763,
+      "step": 3443
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7158823421334461,
+      "learning_rate": 7.260173806120332e-06,
+      "loss": 0.9151,
+      "step": 3444
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8181105950998127,
+      "learning_rate": 7.258620631184782e-06,
+      "loss": 0.9461,
+      "step": 3445
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6162058789610094,
+      "learning_rate": 7.2570671823836925e-06,
+      "loss": 0.8986,
+      "step": 3446
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8027282026510234,
+      "learning_rate": 7.255513459905426e-06,
+      "loss": 0.9112,
+      "step": 3447
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.818396226187163,
+      "learning_rate": 7.253959463938376e-06,
+      "loss": 0.9245,
+      "step": 3448
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.716609130054721,
+      "learning_rate": 7.252405194670973e-06,
+      "loss": 0.8662,
+      "step": 3449
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7486366638058954,
+      "learning_rate": 7.250850652291671e-06,
+      "loss": 0.9078,
+      "step": 3450
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7590519895637964,
+      "learning_rate": 7.24929583698897e-06,
+      "loss": 0.8997,
+      "step": 3451
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7400557813578819,
+      "learning_rate": 7.247740748951394e-06,
+      "loss": 0.8994,
+      "step": 3452
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.611342829961703,
+      "learning_rate": 7.246185388367502e-06,
+      "loss": 0.9758,
+      "step": 3453
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6746883681659026,
+      "learning_rate": 7.2446297554258895e-06,
+      "loss": 0.8587,
+      "step": 3454
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8189898540297689,
+      "learning_rate": 7.24307385031518e-06,
+      "loss": 0.89,
+      "step": 3455
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8961803962971986,
+      "learning_rate": 7.2415176732240335e-06,
+      "loss": 0.9717,
+      "step": 3456
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.865525383092902,
+      "learning_rate": 7.239961224341141e-06,
+      "loss": 1.0283,
+      "step": 3457
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6677886797848644,
+      "learning_rate": 7.238404503855228e-06,
+      "loss": 0.8762,
+      "step": 3458
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.885745128908834,
+      "learning_rate": 7.236847511955051e-06,
+      "loss": 0.9203,
+      "step": 3459
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7486665106804422,
+      "learning_rate": 7.235290248829403e-06,
+      "loss": 0.9285,
+      "step": 3460
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.953345513921867,
+      "learning_rate": 7.233732714667104e-06,
+      "loss": 1.0971,
+      "step": 3461
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8798253475424074,
+      "learning_rate": 7.232174909657015e-06,
+      "loss": 0.9326,
+      "step": 3462
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8255256954074066,
+      "learning_rate": 7.230616833988021e-06,
+      "loss": 0.8921,
+      "step": 3463
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6810973696398843,
+      "learning_rate": 7.229058487849045e-06,
+      "loss": 0.8748,
+      "step": 3464
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.699431875022201,
+      "learning_rate": 7.227499871429047e-06,
+      "loss": 0.9374,
+      "step": 3465
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7002883843124876,
+      "learning_rate": 7.225940984917007e-06,
+      "loss": 0.8776,
+      "step": 3466
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9270130161864797,
+      "learning_rate": 7.224381828501947e-06,
+      "loss": 1.064,
+      "step": 3467
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7414476434478512,
+      "learning_rate": 7.222822402372924e-06,
+      "loss": 1.0389,
+      "step": 3468
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8091634335373168,
+      "learning_rate": 7.221262706719021e-06,
+      "loss": 0.9602,
+      "step": 3469
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7427205375852077,
+      "learning_rate": 7.219702741729357e-06,
+      "loss": 0.8409,
+      "step": 3470
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7690577900169668,
+      "learning_rate": 7.2181425075930845e-06,
+      "loss": 0.9345,
+      "step": 3471
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6724166838587162,
+      "learning_rate": 7.2165820044993835e-06,
+      "loss": 0.9216,
+      "step": 3472
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7634310383908485,
+      "learning_rate": 7.215021232637474e-06,
+      "loss": 0.8481,
+      "step": 3473
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7473934818222827,
+      "learning_rate": 7.213460192196603e-06,
+      "loss": 0.9625,
+      "step": 3474
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7296853384106563,
+      "learning_rate": 7.211898883366053e-06,
+      "loss": 0.9293,
+      "step": 3475
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.651065852169429,
+      "learning_rate": 7.210337306335138e-06,
+      "loss": 0.9326,
+      "step": 3476
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6860875418686192,
+      "learning_rate": 7.208775461293205e-06,
+      "loss": 0.845,
+      "step": 3477
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7629839148290547,
+      "learning_rate": 7.20721334842963e-06,
+      "loss": 0.9358,
+      "step": 3478
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7231969885581089,
+      "learning_rate": 7.205650967933829e-06,
+      "loss": 0.9448,
+      "step": 3479
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9721639078528476,
+      "learning_rate": 7.204088319995245e-06,
+      "loss": 1.0673,
+      "step": 3480
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6975784893916537,
+      "learning_rate": 7.202525404803352e-06,
+      "loss": 0.9274,
+      "step": 3481
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6695056825090684,
+      "learning_rate": 7.2009622225476616e-06,
+      "loss": 0.832,
+      "step": 3482
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6816281471161412,
+      "learning_rate": 7.199398773417714e-06,
+      "loss": 0.8584,
+      "step": 3483
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7921014854808943,
+      "learning_rate": 7.197835057603079e-06,
+      "loss": 0.9024,
+      "step": 3484
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7698940163372854,
+      "learning_rate": 7.196271075293371e-06,
+      "loss": 1.0021,
+      "step": 3485
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6747633777943998,
+      "learning_rate": 7.194706826678222e-06,
+      "loss": 0.9277,
+      "step": 3486
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7409962166720245,
+      "learning_rate": 7.1931423119473035e-06,
+      "loss": 0.8582,
+      "step": 3487
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7812764980601534,
+      "learning_rate": 7.191577531290318e-06,
+      "loss": 0.8233,
+      "step": 3488
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.767316632801242,
+      "learning_rate": 7.190012484897002e-06,
+      "loss": 0.9279,
+      "step": 3489
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7692264559225885,
+      "learning_rate": 7.188447172957121e-06,
+      "loss": 0.9139,
+      "step": 3490
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.748707694661078,
+      "learning_rate": 7.186881595660477e-06,
+      "loss": 0.9114,
+      "step": 3491
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.711610896625979,
+      "learning_rate": 7.1853157531969e-06,
+      "loss": 0.9543,
+      "step": 3492
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6984804499354145,
+      "learning_rate": 7.183749645756254e-06,
+      "loss": 0.893,
+      "step": 3493
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6722394039213344,
+      "learning_rate": 7.182183273528436e-06,
+      "loss": 0.9737,
+      "step": 3494
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.0040722591900275,
+      "learning_rate": 7.18061663670337e-06,
+      "loss": 0.9287,
+      "step": 3495
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.75404960183231,
+      "learning_rate": 7.1790497354710205e-06,
+      "loss": 0.8583,
+      "step": 3496
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7967910722965834,
+      "learning_rate": 7.17748257002138e-06,
+      "loss": 0.9323,
+      "step": 3497
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8916596683543612,
+      "learning_rate": 7.175915140544469e-06,
+      "loss": 0.924,
+      "step": 3498
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7454463320794427,
+      "learning_rate": 7.174347447230346e-06,
+      "loss": 0.9463,
+      "step": 3499
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8483780455521097,
+      "learning_rate": 7.172779490269099e-06,
+      "loss": 0.9119,
+      "step": 3500
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8976118256026944,
+      "learning_rate": 7.171211269850847e-06,
+      "loss": 0.8887,
+      "step": 3501
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.769158991198243,
+      "learning_rate": 7.169642786165747e-06,
+      "loss": 0.9119,
+      "step": 3502
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8277474191183596,
+      "learning_rate": 7.168074039403975e-06,
+      "loss": 0.8757,
+      "step": 3503
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6958585951090037,
+      "learning_rate": 7.166505029755753e-06,
+      "loss": 0.8831,
+      "step": 3504
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7207825956886669,
+      "learning_rate": 7.1649357574113275e-06,
+      "loss": 0.8843,
+      "step": 3505
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7815528148208544,
+      "learning_rate": 7.163366222560977e-06,
+      "loss": 0.8855,
+      "step": 3506
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7192400789337186,
+      "learning_rate": 7.161796425395014e-06,
+      "loss": 0.8877,
+      "step": 3507
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9783996606398329,
+      "learning_rate": 7.160226366103782e-06,
+      "loss": 1.0543,
+      "step": 3508
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6270732648384407,
+      "learning_rate": 7.158656044877654e-06,
+      "loss": 0.8499,
+      "step": 3509
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8241045020053068,
+      "learning_rate": 7.15708546190704e-06,
+      "loss": 0.996,
+      "step": 3510
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8199280208426227,
+      "learning_rate": 7.155514617382377e-06,
+      "loss": 0.8553,
+      "step": 3511
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7553806407382988,
+      "learning_rate": 7.153943511494135e-06,
+      "loss": 0.9134,
+      "step": 3512
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8600354417877532,
+      "learning_rate": 7.152372144432818e-06,
+      "loss": 0.9632,
+      "step": 3513
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7725834857013472,
+      "learning_rate": 7.1508005163889575e-06,
+      "loss": 0.9559,
+      "step": 3514
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7069239121104403,
+      "learning_rate": 7.1492286275531175e-06,
+      "loss": 0.9619,
+      "step": 3515
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7024056934719856,
+      "learning_rate": 7.147656478115899e-06,
+      "loss": 0.9334,
+      "step": 3516
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7481645300916133,
+      "learning_rate": 7.146084068267928e-06,
+      "loss": 0.9439,
+      "step": 3517
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7559461091840354,
+      "learning_rate": 7.144511398199865e-06,
+      "loss": 0.8736,
+      "step": 3518
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8077910504595287,
+      "learning_rate": 7.1429384681024e-06,
+      "loss": 0.7897,
+      "step": 3519
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6310237060081199,
+      "learning_rate": 7.141365278166261e-06,
+      "loss": 0.9498,
+      "step": 3520
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7134204298878384,
+      "learning_rate": 7.1397918285821965e-06,
+      "loss": 0.9041,
+      "step": 3521
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7598912789986503,
+      "learning_rate": 7.138218119540998e-06,
+      "loss": 0.8775,
+      "step": 3522
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8203018974359149,
+      "learning_rate": 7.1366441512334806e-06,
+      "loss": 0.9972,
+      "step": 3523
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6901003079524397,
+      "learning_rate": 7.135069923850493e-06,
+      "loss": 0.9122,
+      "step": 3524
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7067658173682547,
+      "learning_rate": 7.133495437582917e-06,
+      "loss": 0.8729,
+      "step": 3525
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.739467279926299,
+      "learning_rate": 7.131920692621663e-06,
+      "loss": 0.8889,
+      "step": 3526
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.712169742091463,
+      "learning_rate": 7.130345689157676e-06,
+      "loss": 0.949,
+      "step": 3527
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8493548837059157,
+      "learning_rate": 7.12877042738193e-06,
+      "loss": 0.948,
+      "step": 3528
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.760063625206818,
+      "learning_rate": 7.127194907485431e-06,
+      "loss": 0.9048,
+      "step": 3529
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6929367389757077,
+      "learning_rate": 7.125619129659215e-06,
+      "loss": 0.9421,
+      "step": 3530
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7601751255694287,
+      "learning_rate": 7.1240430940943516e-06,
+      "loss": 0.8783,
+      "step": 3531
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6053897818718008,
+      "learning_rate": 7.12246680098194e-06,
+      "loss": 0.9588,
+      "step": 3532
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7603438518916497,
+      "learning_rate": 7.120890250513111e-06,
+      "loss": 0.9433,
+      "step": 3533
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.220126126823655,
+      "learning_rate": 7.119313442879029e-06,
+      "loss": 0.9106,
+      "step": 3534
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6541489141711925,
+      "learning_rate": 7.1177363782708855e-06,
+      "loss": 0.9885,
+      "step": 3535
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9399622569272348,
+      "learning_rate": 7.116159056879904e-06,
+      "loss": 0.8974,
+      "step": 3536
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.794128294565982,
+      "learning_rate": 7.114581478897342e-06,
+      "loss": 0.9918,
+      "step": 3537
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6915538964619918,
+      "learning_rate": 7.113003644514486e-06,
+      "loss": 0.8626,
+      "step": 3538
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7960142506956844,
+      "learning_rate": 7.111425553922653e-06,
+      "loss": 0.8887,
+      "step": 3539
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6931387091978012,
+      "learning_rate": 7.109847207313192e-06,
+      "loss": 0.968,
+      "step": 3540
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.773033659021171,
+      "learning_rate": 7.108268604877484e-06,
+      "loss": 0.9888,
+      "step": 3541
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7818874216305898,
+      "learning_rate": 7.1066897468069394e-06,
+      "loss": 0.9943,
+      "step": 3542
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7319535847860126,
+      "learning_rate": 7.105110633292999e-06,
+      "loss": 0.8553,
+      "step": 3543
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9386846611193761,
+      "learning_rate": 7.103531264527138e-06,
+      "loss": 1.0136,
+      "step": 3544
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8573797711984961,
+      "learning_rate": 7.10195164070086e-06,
+      "loss": 0.9914,
+      "step": 3545
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7146029394373403,
+      "learning_rate": 7.100371762005698e-06,
+      "loss": 0.8788,
+      "step": 3546
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.714033874303207,
+      "learning_rate": 7.098791628633218e-06,
+      "loss": 0.9106,
+      "step": 3547
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6715255869318555,
+      "learning_rate": 7.097211240775018e-06,
+      "loss": 0.9186,
+      "step": 3548
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6075976494129403,
+      "learning_rate": 7.095630598622725e-06,
+      "loss": 0.9878,
+      "step": 3549
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6510073819639206,
+      "learning_rate": 7.094049702367998e-06,
+      "loss": 0.9107,
+      "step": 3550
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6418862968378467,
+      "learning_rate": 7.092468552202524e-06,
+      "loss": 0.8797,
+      "step": 3551
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8735806336054408,
+      "learning_rate": 7.090887148318024e-06,
+      "loss": 0.9168,
+      "step": 3552
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.735907317119064,
+      "learning_rate": 7.089305490906251e-06,
+      "loss": 0.9539,
+      "step": 3553
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7198969613419666,
+      "learning_rate": 7.0877235801589825e-06,
+      "loss": 0.9026,
+      "step": 3554
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7414039773427819,
+      "learning_rate": 7.086141416268033e-06,
+      "loss": 0.9493,
+      "step": 3555
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3458528378871621,
+      "learning_rate": 7.084558999425245e-06,
+      "loss": 1.0523,
+      "step": 3556
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7280329297943133,
+      "learning_rate": 7.082976329822491e-06,
+      "loss": 0.8749,
+      "step": 3557
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7705729272248942,
+      "learning_rate": 7.081393407651676e-06,
+      "loss": 0.9312,
+      "step": 3558
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.4045948670533135,
+      "learning_rate": 7.0798102331047345e-06,
+      "loss": 0.8842,
+      "step": 3559
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6168674531526348,
+      "learning_rate": 7.078226806373632e-06,
+      "loss": 0.9396,
+      "step": 3560
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7586232755342333,
+      "learning_rate": 7.076643127650367e-06,
+      "loss": 0.9196,
+      "step": 3561
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6949599209699688,
+      "learning_rate": 7.075059197126961e-06,
+      "loss": 0.8395,
+      "step": 3562
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.655052520846647,
+      "learning_rate": 7.073475014995472e-06,
+      "loss": 0.912,
+      "step": 3563
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8686108262185473,
+      "learning_rate": 7.071890581447992e-06,
+      "loss": 1.0668,
+      "step": 3564
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.66861172282927,
+      "learning_rate": 7.070305896676635e-06,
+      "loss": 0.9405,
+      "step": 3565
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.708452050627153,
+      "learning_rate": 7.068720960873552e-06,
+      "loss": 0.9237,
+      "step": 3566
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6503561574224743,
+      "learning_rate": 7.067135774230919e-06,
+      "loss": 0.8877,
+      "step": 3567
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8480827476985013,
+      "learning_rate": 7.065550336940947e-06,
+      "loss": 0.8971,
+      "step": 3568
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7398193950071896,
+      "learning_rate": 7.063964649195874e-06,
+      "loss": 0.8831,
+      "step": 3569
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.764079176478315,
+      "learning_rate": 7.0623787111879734e-06,
+      "loss": 0.9852,
+      "step": 3570
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.794083433980299,
+      "learning_rate": 7.060792523109545e-06,
+      "loss": 0.871,
+      "step": 3571
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6832595331538067,
+      "learning_rate": 7.059206085152918e-06,
+      "loss": 0.8388,
+      "step": 3572
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6982910753805576,
+      "learning_rate": 7.057619397510453e-06,
+      "loss": 0.9367,
+      "step": 3573
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7589671389912003,
+      "learning_rate": 7.056032460374542e-06,
+      "loss": 0.862,
+      "step": 3574
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7350236387515476,
+      "learning_rate": 7.054445273937609e-06,
+      "loss": 0.8671,
+      "step": 3575
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.852659575495065,
+      "learning_rate": 7.052857838392104e-06,
+      "loss": 0.9906,
+      "step": 3576
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6926339876220031,
+      "learning_rate": 7.0512701539305065e-06,
+      "loss": 0.823,
+      "step": 3577
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6561098253224362,
+      "learning_rate": 7.049682220745332e-06,
+      "loss": 0.9211,
+      "step": 3578
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8232130648782472,
+      "learning_rate": 7.048094039029123e-06,
+      "loss": 0.9338,
+      "step": 3579
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8125939761170358,
+      "learning_rate": 7.046505608974448e-06,
+      "loss": 0.9743,
+      "step": 3580
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7408365118567453,
+      "learning_rate": 7.044916930773915e-06,
+      "loss": 0.8947,
+      "step": 3581
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7737129317931215,
+      "learning_rate": 7.043328004620154e-06,
+      "loss": 0.9902,
+      "step": 3582
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7646357180579426,
+      "learning_rate": 7.041738830705828e-06,
+      "loss": 0.9835,
+      "step": 3583
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.926701917454396,
+      "learning_rate": 7.040149409223629e-06,
+      "loss": 1.0542,
+      "step": 3584
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7128493588012803,
+      "learning_rate": 7.038559740366282e-06,
+      "loss": 0.9246,
+      "step": 3585
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7843757328494119,
+      "learning_rate": 7.036969824326536e-06,
+      "loss": 0.8649,
+      "step": 3586
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7125016556112236,
+      "learning_rate": 7.035379661297179e-06,
+      "loss": 0.9618,
+      "step": 3587
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6121399830975427,
+      "learning_rate": 7.033789251471019e-06,
+      "loss": 0.9234,
+      "step": 3588
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8622084126529141,
+      "learning_rate": 7.032198595040902e-06,
+      "loss": 1.0326,
+      "step": 3589
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6648160108726295,
+      "learning_rate": 7.0306076921997e-06,
+      "loss": 0.9494,
+      "step": 3590
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7262735458229541,
+      "learning_rate": 7.029016543140312e-06,
+      "loss": 0.908,
+      "step": 3591
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7035821157578581,
+      "learning_rate": 7.027425148055676e-06,
+      "loss": 1.0045,
+      "step": 3592
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8150667815822368,
+      "learning_rate": 7.025833507138751e-06,
+      "loss": 0.9564,
+      "step": 3593
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.564818670908862,
+      "learning_rate": 7.024241620582527e-06,
+      "loss": 0.8421,
+      "step": 3594
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8292583117408145,
+      "learning_rate": 7.022649488580029e-06,
+      "loss": 0.8817,
+      "step": 3595
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.700825587067748,
+      "learning_rate": 7.021057111324307e-06,
+      "loss": 0.8728,
+      "step": 3596
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7051856044142253,
+      "learning_rate": 7.019464489008443e-06,
+      "loss": 0.8248,
+      "step": 3597
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.69796490832553,
+      "learning_rate": 7.017871621825549e-06,
+      "loss": 0.8725,
+      "step": 3598
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7509850983263118,
+      "learning_rate": 7.016278509968762e-06,
+      "loss": 0.9395,
+      "step": 3599
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.74357438942604,
+      "learning_rate": 7.014685153631255e-06,
+      "loss": 0.9511,
+      "step": 3600
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6876935005557043,
+      "learning_rate": 7.013091553006228e-06,
+      "loss": 0.9238,
+      "step": 3601
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8300493357983656,
+      "learning_rate": 7.01149770828691e-06,
+      "loss": 0.9737,
+      "step": 3602
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.606938984260058,
+      "learning_rate": 7.009903619666561e-06,
+      "loss": 0.8865,
+      "step": 3603
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6960378197072044,
+      "learning_rate": 7.008309287338467e-06,
+      "loss": 0.8711,
+      "step": 3604
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.908480662872319,
+      "learning_rate": 7.006714711495949e-06,
+      "loss": 1.0797,
+      "step": 3605
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7229790956755302,
+      "learning_rate": 7.0051198923323535e-06,
+      "loss": 0.8886,
+      "step": 3606
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7752920922468276,
+      "learning_rate": 7.0035248300410595e-06,
+      "loss": 0.9011,
+      "step": 3607
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7733270479359085,
+      "learning_rate": 7.0019295248154714e-06,
+      "loss": 0.8575,
+      "step": 3608
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6264233153332548,
+      "learning_rate": 7.000333976849029e-06,
+      "loss": 0.8651,
+      "step": 3609
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6092462203998321,
+      "learning_rate": 6.998738186335194e-06,
+      "loss": 0.8352,
+      "step": 3610
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.818348322029172,
+      "learning_rate": 6.9971421534674625e-06,
+      "loss": 1.0008,
+      "step": 3611
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8252915006400738,
+      "learning_rate": 6.99554587843936e-06,
+      "loss": 0.9738,
+      "step": 3612
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.628535703389244,
+      "learning_rate": 6.993949361444441e-06,
+      "loss": 0.9075,
+      "step": 3613
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.973083286903235,
+      "learning_rate": 6.992352602676288e-06,
+      "loss": 0.9911,
+      "step": 3614
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.747209253765597,
+      "learning_rate": 6.990755602328511e-06,
+      "loss": 0.9526,
+      "step": 3615
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8398221279292295,
+      "learning_rate": 6.9891583605947565e-06,
+      "loss": 0.9031,
+      "step": 3616
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7261923784510973,
+      "learning_rate": 6.987560877668692e-06,
+      "loss": 0.9351,
+      "step": 3617
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6943700548873455,
+      "learning_rate": 6.9859631537440195e-06,
+      "loss": 0.8773,
+      "step": 3618
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6774115725503271,
+      "learning_rate": 6.984365189014468e-06,
+      "loss": 0.9466,
+      "step": 3619
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7872511198569294,
+      "learning_rate": 6.982766983673796e-06,
+      "loss": 0.9934,
+      "step": 3620
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7808734424974293,
+      "learning_rate": 6.9811685379157916e-06,
+      "loss": 0.9493,
+      "step": 3621
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.861382861203167,
+      "learning_rate": 6.979569851934271e-06,
+      "loss": 1.0723,
+      "step": 3622
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.891245373566788,
+      "learning_rate": 6.977970925923081e-06,
+      "loss": 0.929,
+      "step": 3623
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8010821627236944,
+      "learning_rate": 6.976371760076099e-06,
+      "loss": 0.9234,
+      "step": 3624
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6923243397254573,
+      "learning_rate": 6.974772354587226e-06,
+      "loss": 0.9559,
+      "step": 3625
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7710853243633347,
+      "learning_rate": 6.973172709650397e-06,
+      "loss": 0.8709,
+      "step": 3626
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6971943867156254,
+      "learning_rate": 6.971572825459576e-06,
+      "loss": 0.9049,
+      "step": 3627
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.735043608150645,
+      "learning_rate": 6.96997270220875e-06,
+      "loss": 0.8544,
+      "step": 3628
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6541145806329642,
+      "learning_rate": 6.9683723400919455e-06,
+      "loss": 0.9346,
+      "step": 3629
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8080198751977745,
+      "learning_rate": 6.9667717393032065e-06,
+      "loss": 1.0205,
+      "step": 3630
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8772116250763214,
+      "learning_rate": 6.965170900036614e-06,
+      "loss": 1.0918,
+      "step": 3631
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.65482459642941,
+      "learning_rate": 6.963569822486275e-06,
+      "loss": 0.8737,
+      "step": 3632
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7010884636340624,
+      "learning_rate": 6.961968506846327e-06,
+      "loss": 0.8542,
+      "step": 3633
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7063148031770505,
+      "learning_rate": 6.960366953310931e-06,
+      "loss": 0.8811,
+      "step": 3634
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7668854618493877,
+      "learning_rate": 6.958765162074287e-06,
+      "loss": 0.9082,
+      "step": 3635
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7923528247129772,
+      "learning_rate": 6.957163133330611e-06,
+      "loss": 0.9789,
+      "step": 3636
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.609014366666203,
+      "learning_rate": 6.9555608672741596e-06,
+      "loss": 0.9025,
+      "step": 3637
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8641406077462145,
+      "learning_rate": 6.9539583640992096e-06,
+      "loss": 1.0668,
+      "step": 3638
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7531199417175816,
+      "learning_rate": 6.9523556240000725e-06,
+      "loss": 0.906,
+      "step": 3639
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6857874244839477,
+      "learning_rate": 6.9507526471710865e-06,
+      "loss": 0.8895,
+      "step": 3640
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7908512647272818,
+      "learning_rate": 6.949149433806614e-06,
+      "loss": 0.8874,
+      "step": 3641
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7418529250113568,
+      "learning_rate": 6.947545984101054e-06,
+      "loss": 0.9729,
+      "step": 3642
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8714058693541644,
+      "learning_rate": 6.94594229824883e-06,
+      "loss": 1.0396,
+      "step": 3643
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8071138673207257,
+      "learning_rate": 6.944338376444394e-06,
+      "loss": 1.0111,
+      "step": 3644
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.759354803754439,
+      "learning_rate": 6.942734218882225e-06,
+      "loss": 0.8957,
+      "step": 3645
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.689225846792078,
+      "learning_rate": 6.941129825756836e-06,
+      "loss": 0.979,
+      "step": 3646
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.130986509949675,
+      "learning_rate": 6.939525197262762e-06,
+      "loss": 0.889,
+      "step": 3647
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.639713783453456,
+      "learning_rate": 6.937920333594572e-06,
+      "loss": 0.8984,
+      "step": 3648
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7097915488843474,
+      "learning_rate": 6.936315234946861e-06,
+      "loss": 0.9067,
+      "step": 3649
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8087871128239699,
+      "learning_rate": 6.934709901514252e-06,
+      "loss": 0.9976,
+      "step": 3650
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7255688648281826,
+      "learning_rate": 6.933104333491399e-06,
+      "loss": 0.9276,
+      "step": 3651
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9005083373032283,
+      "learning_rate": 6.931498531072977e-06,
+      "loss": 0.9214,
+      "step": 3652
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.666958058830387,
+      "learning_rate": 6.929892494453703e-06,
+      "loss": 0.8921,
+      "step": 3653
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.695282635440867,
+      "learning_rate": 6.928286223828309e-06,
+      "loss": 0.9089,
+      "step": 3654
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7638583604870655,
+      "learning_rate": 6.926679719391562e-06,
+      "loss": 0.9392,
+      "step": 3655
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8289464413422243,
+      "learning_rate": 6.9250729813382585e-06,
+      "loss": 0.9443,
+      "step": 3656
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7496657726868867,
+      "learning_rate": 6.923466009863219e-06,
+      "loss": 1.0295,
+      "step": 3657
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6332571260722488,
+      "learning_rate": 6.921858805161295e-06,
+      "loss": 0.8893,
+      "step": 3658
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6852687531407198,
+      "learning_rate": 6.920251367427362e-06,
+      "loss": 1.0155,
+      "step": 3659
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6986753239597217,
+      "learning_rate": 6.918643696856333e-06,
+      "loss": 0.8528,
+      "step": 3660
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7148259665793568,
+      "learning_rate": 6.917035793643142e-06,
+      "loss": 0.8369,
+      "step": 3661
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5803507473511513,
+      "learning_rate": 6.915427657982752e-06,
+      "loss": 0.8683,
+      "step": 3662
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.783435142997351,
+      "learning_rate": 6.913819290070154e-06,
+      "loss": 0.8671,
+      "step": 3663
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.705001770528061,
+      "learning_rate": 6.912210690100369e-06,
+      "loss": 0.9762,
+      "step": 3664
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7800298536374306,
+      "learning_rate": 6.910601858268444e-06,
+      "loss": 0.8176,
+      "step": 3665
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.67537335112625,
+      "learning_rate": 6.90899279476946e-06,
+      "loss": 0.9066,
+      "step": 3666
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.812873194083026,
+      "learning_rate": 6.907383499798517e-06,
+      "loss": 0.8636,
+      "step": 3667
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7521746834407605,
+      "learning_rate": 6.905773973550748e-06,
+      "loss": 0.9094,
+      "step": 3668
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.745678422759384,
+      "learning_rate": 6.904164216221315e-06,
+      "loss": 0.8547,
+      "step": 3669
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7778083272687444,
+      "learning_rate": 6.902554228005406e-06,
+      "loss": 0.9038,
+      "step": 3670
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.688256945852924,
+      "learning_rate": 6.900944009098238e-06,
+      "loss": 0.9702,
+      "step": 3671
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7208661560683851,
+      "learning_rate": 6.899333559695056e-06,
+      "loss": 0.8875,
+      "step": 3672
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.629997570936904,
+      "learning_rate": 6.897722879991131e-06,
+      "loss": 0.9192,
+      "step": 3673
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7847764791928995,
+      "learning_rate": 6.896111970181765e-06,
+      "loss": 0.9028,
+      "step": 3674
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7153027589835654,
+      "learning_rate": 6.894500830462285e-06,
+      "loss": 0.9097,
+      "step": 3675
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7074904038794971,
+      "learning_rate": 6.892889461028048e-06,
+      "loss": 0.9425,
+      "step": 3676
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.705484829404824,
+      "learning_rate": 6.891277862074439e-06,
+      "loss": 0.8809,
+      "step": 3677
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0074110379284535,
+      "learning_rate": 6.889666033796869e-06,
+      "loss": 1.0331,
+      "step": 3678
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.726853823249063,
+      "learning_rate": 6.888053976390776e-06,
+      "loss": 0.9379,
+      "step": 3679
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7687340792811335,
+      "learning_rate": 6.886441690051631e-06,
+      "loss": 0.9329,
+      "step": 3680
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6423933673616542,
+      "learning_rate": 6.884829174974927e-06,
+      "loss": 0.8403,
+      "step": 3681
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8623918017004886,
+      "learning_rate": 6.883216431356188e-06,
+      "loss": 0.9943,
+      "step": 3682
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.871937056500414,
+      "learning_rate": 6.881603459390964e-06,
+      "loss": 0.9839,
+      "step": 3683
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.743798689550891,
+      "learning_rate": 6.879990259274832e-06,
+      "loss": 0.9017,
+      "step": 3684
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8666613682805704,
+      "learning_rate": 6.878376831203402e-06,
+      "loss": 1.0712,
+      "step": 3685
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.879389760472025,
+      "learning_rate": 6.876763175372306e-06,
+      "loss": 1.0401,
+      "step": 3686
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7632567345522692,
+      "learning_rate": 6.875149291977201e-06,
+      "loss": 0.8705,
+      "step": 3687
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.864786728366429,
+      "learning_rate": 6.873535181213784e-06,
+      "loss": 0.8552,
+      "step": 3688
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6820992011268312,
+      "learning_rate": 6.871920843277765e-06,
+      "loss": 0.8614,
+      "step": 3689
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8495160183213671,
+      "learning_rate": 6.87030627836489e-06,
+      "loss": 1.0403,
+      "step": 3690
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.899335357881142,
+      "learning_rate": 6.868691486670932e-06,
+      "loss": 0.9286,
+      "step": 3691
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7439479097153436,
+      "learning_rate": 6.867076468391688e-06,
+      "loss": 0.9474,
+      "step": 3692
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8186853891340966,
+      "learning_rate": 6.865461223722987e-06,
+      "loss": 0.9392,
+      "step": 3693
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7642438398249785,
+      "learning_rate": 6.86384575286068e-06,
+      "loss": 0.877,
+      "step": 3694
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.147982875723684,
+      "learning_rate": 6.862230056000648e-06,
+      "loss": 0.9931,
+      "step": 3695
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7402840310387402,
+      "learning_rate": 6.860614133338804e-06,
+      "loss": 0.8747,
+      "step": 3696
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7634381472093945,
+      "learning_rate": 6.858997985071082e-06,
+      "loss": 1.0319,
+      "step": 3697
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8313981938744797,
+      "learning_rate": 6.857381611393445e-06,
+      "loss": 1.0703,
+      "step": 3698
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.876817613513853,
+      "learning_rate": 6.855765012501884e-06,
+      "loss": 0.96,
+      "step": 3699
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6995553086207764,
+      "learning_rate": 6.8541481885924174e-06,
+      "loss": 0.9551,
+      "step": 3700
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7693112625411371,
+      "learning_rate": 6.852531139861091e-06,
+      "loss": 0.9888,
+      "step": 3701
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8254403350141695,
+      "learning_rate": 6.850913866503976e-06,
+      "loss": 1.0925,
+      "step": 3702
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.680283346881603,
+      "learning_rate": 6.849296368717176e-06,
+      "loss": 0.8809,
+      "step": 3703
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7844627933581727,
+      "learning_rate": 6.847678646696814e-06,
+      "loss": 0.8799,
+      "step": 3704
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.697828403326104,
+      "learning_rate": 6.846060700639047e-06,
+      "loss": 0.9546,
+      "step": 3705
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7573637171302605,
+      "learning_rate": 6.844442530740055e-06,
+      "loss": 0.9018,
+      "step": 3706
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7455961203347996,
+      "learning_rate": 6.842824137196047e-06,
+      "loss": 0.921,
+      "step": 3707
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6999668014104816,
+      "learning_rate": 6.84120552020326e-06,
+      "loss": 0.9121,
+      "step": 3708
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8956969697481791,
+      "learning_rate": 6.839586679957957e-06,
+      "loss": 1.058,
+      "step": 3709
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7541173121499185,
+      "learning_rate": 6.837967616656426e-06,
+      "loss": 0.9004,
+      "step": 3710
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8015058039687593,
+      "learning_rate": 6.836348330494985e-06,
+      "loss": 0.9553,
+      "step": 3711
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5840961512474818,
+      "learning_rate": 6.834728821669978e-06,
+      "loss": 0.8442,
+      "step": 3712
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8937536444855703,
+      "learning_rate": 6.8331090903777754e-06,
+      "loss": 1.0761,
+      "step": 3713
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8280253509182,
+      "learning_rate": 6.831489136814777e-06,
+      "loss": 0.9211,
+      "step": 3714
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.636831188915033,
+      "learning_rate": 6.829868961177406e-06,
+      "loss": 0.8298,
+      "step": 3715
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7076738444979378,
+      "learning_rate": 6.8282485636621165e-06,
+      "loss": 0.8862,
+      "step": 3716
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.836120942194683,
+      "learning_rate": 6.826627944465384e-06,
+      "loss": 0.8648,
+      "step": 3717
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6527220215417386,
+      "learning_rate": 6.825007103783716e-06,
+      "loss": 0.8867,
+      "step": 3718
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7162667018186137,
+      "learning_rate": 6.8233860418136466e-06,
+      "loss": 0.8734,
+      "step": 3719
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.709710387424364,
+      "learning_rate": 6.821764758751732e-06,
+      "loss": 0.9524,
+      "step": 3720
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6790190215837812,
+      "learning_rate": 6.820143254794559e-06,
+      "loss": 0.871,
+      "step": 3721
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.034573991525158,
+      "learning_rate": 6.818521530138743e-06,
+      "loss": 1.0798,
+      "step": 3722
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7075812638114842,
+      "learning_rate": 6.816899584980922e-06,
+      "loss": 0.8439,
+      "step": 3723
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7981849960547498,
+      "learning_rate": 6.815277419517761e-06,
+      "loss": 0.9813,
+      "step": 3724
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8486270231115924,
+      "learning_rate": 6.8136550339459566e-06,
+      "loss": 0.8872,
+      "step": 3725
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7568373994344684,
+      "learning_rate": 6.8120324284622255e-06,
+      "loss": 0.9481,
+      "step": 3726
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7144978788693175,
+      "learning_rate": 6.810409603263315e-06,
+      "loss": 0.8744,
+      "step": 3727
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7387026891378985,
+      "learning_rate": 6.808786558546e-06,
+      "loss": 0.9416,
+      "step": 3728
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8033905382000865,
+      "learning_rate": 6.807163294507078e-06,
+      "loss": 0.8616,
+      "step": 3729
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.673079297335638,
+      "learning_rate": 6.805539811343377e-06,
+      "loss": 0.7971,
+      "step": 3730
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8261397689690728,
+      "learning_rate": 6.803916109251749e-06,
+      "loss": 0.957,
+      "step": 3731
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7506908498476363,
+      "learning_rate": 6.802292188429072e-06,
+      "loss": 0.9321,
+      "step": 3732
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7071283542021913,
+      "learning_rate": 6.8006680490722566e-06,
+      "loss": 0.9477,
+      "step": 3733
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.737197466964932,
+      "learning_rate": 6.799043691378231e-06,
+      "loss": 0.8985,
+      "step": 3734
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.710616059311271,
+      "learning_rate": 6.797419115543955e-06,
+      "loss": 0.8771,
+      "step": 3735
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5396238290682722,
+      "learning_rate": 6.795794321766416e-06,
+      "loss": 0.8686,
+      "step": 3736
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.750814874095692,
+      "learning_rate": 6.794169310242624e-06,
+      "loss": 0.9697,
+      "step": 3737
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.677299708022049,
+      "learning_rate": 6.7925440811696165e-06,
+      "loss": 0.9001,
+      "step": 3738
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8570951216973905,
+      "learning_rate": 6.79091863474446e-06,
+      "loss": 0.898,
+      "step": 3739
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.713846810234748,
+      "learning_rate": 6.789292971164245e-06,
+      "loss": 0.9287,
+      "step": 3740
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8020721146555998,
+      "learning_rate": 6.787667090626089e-06,
+      "loss": 0.9547,
+      "step": 3741
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.626897982308909,
+      "learning_rate": 6.786040993327135e-06,
+      "loss": 0.846,
+      "step": 3742
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6455487210144222,
+      "learning_rate": 6.784414679464553e-06,
+      "loss": 0.8831,
+      "step": 3743
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.663410696042311,
+      "learning_rate": 6.7827881492355385e-06,
+      "loss": 0.8859,
+      "step": 3744
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8617414882032935,
+      "learning_rate": 6.7811614028373166e-06,
+      "loss": 1.0618,
+      "step": 3745
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8125551621194524,
+      "learning_rate": 6.779534440467132e-06,
+      "loss": 0.9492,
+      "step": 3746
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6218212009674808,
+      "learning_rate": 6.777907262322263e-06,
+      "loss": 0.8959,
+      "step": 3747
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8343375643107682,
+      "learning_rate": 6.7762798686000086e-06,
+      "loss": 0.9681,
+      "step": 3748
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.696881416147612,
+      "learning_rate": 6.774652259497696e-06,
+      "loss": 0.9193,
+      "step": 3749
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7826004206483415,
+      "learning_rate": 6.773024435212678e-06,
+      "loss": 0.9302,
+      "step": 3750
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8319228669230851,
+      "learning_rate": 6.7713963959423375e-06,
+      "loss": 1.0393,
+      "step": 3751
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5836593806567643,
+      "learning_rate": 6.769768141884074e-06,
+      "loss": 0.8792,
+      "step": 3752
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8505623136624468,
+      "learning_rate": 6.768139673235323e-06,
+      "loss": 1.0475,
+      "step": 3753
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7150614047050985,
+      "learning_rate": 6.7665109901935426e-06,
+      "loss": 0.9164,
+      "step": 3754
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7979499570366428,
+      "learning_rate": 6.764882092956212e-06,
+      "loss": 0.8846,
+      "step": 3755
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7078001607472872,
+      "learning_rate": 6.763252981720845e-06,
+      "loss": 0.9538,
+      "step": 3756
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.33144393160548,
+      "learning_rate": 6.7616236566849736e-06,
+      "loss": 0.9049,
+      "step": 3757
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6731505535393738,
+      "learning_rate": 6.759994118046161e-06,
+      "loss": 0.9553,
+      "step": 3758
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6550027864088952,
+      "learning_rate": 6.758364366001995e-06,
+      "loss": 0.883,
+      "step": 3759
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8331334553070108,
+      "learning_rate": 6.756734400750087e-06,
+      "loss": 1.0632,
+      "step": 3760
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.757811778120651,
+      "learning_rate": 6.755104222488076e-06,
+      "loss": 0.9113,
+      "step": 3761
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.615295593456092,
+      "learning_rate": 6.7534738314136285e-06,
+      "loss": 0.8752,
+      "step": 3762
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.775132222748083,
+      "learning_rate": 6.751843227724433e-06,
+      "loss": 0.8604,
+      "step": 3763
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6472073846059767,
+      "learning_rate": 6.7502124116182066e-06,
+      "loss": 0.8993,
+      "step": 3764
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8549509127193766,
+      "learning_rate": 6.7485813832926905e-06,
+      "loss": 1.0843,
+      "step": 3765
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6136547802069838,
+      "learning_rate": 6.746950142945654e-06,
+      "loss": 0.9533,
+      "step": 3766
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8084535781462003,
+      "learning_rate": 6.7453186907748914e-06,
+      "loss": 0.9102,
+      "step": 3767
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7109587127042984,
+      "learning_rate": 6.743687026978219e-06,
+      "loss": 1.0018,
+      "step": 3768
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6303294676907938,
+      "learning_rate": 6.742055151753484e-06,
+      "loss": 0.8704,
+      "step": 3769
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6104343620415118,
+      "learning_rate": 6.740423065298556e-06,
+      "loss": 0.853,
+      "step": 3770
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6996613994771286,
+      "learning_rate": 6.7387907678113295e-06,
+      "loss": 1.0272,
+      "step": 3771
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.81173269378801,
+      "learning_rate": 6.73715825948973e-06,
+      "loss": 0.8517,
+      "step": 3772
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7244021989865903,
+      "learning_rate": 6.735525540531702e-06,
+      "loss": 0.9635,
+      "step": 3773
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.685113080814913,
+      "learning_rate": 6.733892611135217e-06,
+      "loss": 0.8163,
+      "step": 3774
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6550017451405235,
+      "learning_rate": 6.732259471498278e-06,
+      "loss": 0.9632,
+      "step": 3775
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7606791252224414,
+      "learning_rate": 6.730626121818906e-06,
+      "loss": 0.9009,
+      "step": 3776
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.674070988963595,
+      "learning_rate": 6.728992562295149e-06,
+      "loss": 0.944,
+      "step": 3777
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8175621407834393,
+      "learning_rate": 6.727358793125084e-06,
+      "loss": 0.9644,
+      "step": 3778
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.642129249192693,
+      "learning_rate": 6.72572481450681e-06,
+      "loss": 0.9267,
+      "step": 3779
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6857643110821325,
+      "learning_rate": 6.724090626638452e-06,
+      "loss": 0.906,
+      "step": 3780
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.693399275075128,
+      "learning_rate": 6.722456229718163e-06,
+      "loss": 0.9085,
+      "step": 3781
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8320490535834988,
+      "learning_rate": 6.720821623944117e-06,
+      "loss": 0.8646,
+      "step": 3782
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7590343990620032,
+      "learning_rate": 6.7191868095145165e-06,
+      "loss": 0.8571,
+      "step": 3783
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7106374493728695,
+      "learning_rate": 6.71755178662759e-06,
+      "loss": 0.8587,
+      "step": 3784
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7451363724212081,
+      "learning_rate": 6.715916555481585e-06,
+      "loss": 0.8942,
+      "step": 3785
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.689887042497531,
+      "learning_rate": 6.714281116274783e-06,
+      "loss": 0.9061,
+      "step": 3786
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8080863228122521,
+      "learning_rate": 6.712645469205488e-06,
+      "loss": 0.9156,
+      "step": 3787
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7354201692908116,
+      "learning_rate": 6.711009614472023e-06,
+      "loss": 0.8805,
+      "step": 3788
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.696291411565936,
+      "learning_rate": 6.709373552272744e-06,
+      "loss": 0.9623,
+      "step": 3789
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7908329405629066,
+      "learning_rate": 6.7077372828060294e-06,
+      "loss": 0.8562,
+      "step": 3790
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7902477224563222,
+      "learning_rate": 6.70610080627028e-06,
+      "loss": 0.8979,
+      "step": 3791
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.857914093764839,
+      "learning_rate": 6.704464122863929e-06,
+      "loss": 0.9804,
+      "step": 3792
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9384256381126161,
+      "learning_rate": 6.702827232785426e-06,
+      "loss": 1.0803,
+      "step": 3793
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6929236781205865,
+      "learning_rate": 6.701190136233248e-06,
+      "loss": 0.8658,
+      "step": 3794
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7619154198679348,
+      "learning_rate": 6.699552833405904e-06,
+      "loss": 0.8731,
+      "step": 3795
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7524271704097114,
+      "learning_rate": 6.697915324501918e-06,
+      "loss": 0.8959,
+      "step": 3796
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8373574567558173,
+      "learning_rate": 6.696277609719846e-06,
+      "loss": 1.0541,
+      "step": 3797
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.892572661822782,
+      "learning_rate": 6.694639689258265e-06,
+      "loss": 0.8571,
+      "step": 3798
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7228868993933018,
+      "learning_rate": 6.6930015633157815e-06,
+      "loss": 0.941,
+      "step": 3799
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7706629058604655,
+      "learning_rate": 6.691363232091019e-06,
+      "loss": 0.9637,
+      "step": 3800
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7442359043668563,
+      "learning_rate": 6.6897246957826356e-06,
+      "loss": 0.8838,
+      "step": 3801
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7168805736997106,
+      "learning_rate": 6.688085954589307e-06,
+      "loss": 0.941,
+      "step": 3802
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.687654845367884,
+      "learning_rate": 6.686447008709735e-06,
+      "loss": 0.9362,
+      "step": 3803
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9724080773294006,
+      "learning_rate": 6.684807858342652e-06,
+      "loss": 1.0228,
+      "step": 3804
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7950247486428166,
+      "learning_rate": 6.683168503686807e-06,
+      "loss": 0.8895,
+      "step": 3805
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8668694818664551,
+      "learning_rate": 6.681528944940977e-06,
+      "loss": 1.0589,
+      "step": 3806
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.65610554971664,
+      "learning_rate": 6.6798891823039655e-06,
+      "loss": 0.9491,
+      "step": 3807
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.849518928219065,
+      "learning_rate": 6.6782492159746005e-06,
+      "loss": 1.0995,
+      "step": 3808
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.845130575358366,
+      "learning_rate": 6.676609046151732e-06,
+      "loss": 0.8472,
+      "step": 3809
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7800059664355221,
+      "learning_rate": 6.674968673034235e-06,
+      "loss": 0.9232,
+      "step": 3810
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.773124242462841,
+      "learning_rate": 6.6733280968210125e-06,
+      "loss": 0.8427,
+      "step": 3811
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.653175895595962,
+      "learning_rate": 6.67168731771099e-06,
+      "loss": 0.8532,
+      "step": 3812
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7440775503380563,
+      "learning_rate": 6.670046335903116e-06,
+      "loss": 0.8924,
+      "step": 3813
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8321456922135435,
+      "learning_rate": 6.668405151596366e-06,
+      "loss": 0.9871,
+      "step": 3814
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.919470457788189,
+      "learning_rate": 6.6667637649897424e-06,
+      "loss": 1.0025,
+      "step": 3815
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.708580923251713,
+      "learning_rate": 6.6651221762822635e-06,
+      "loss": 0.8246,
+      "step": 3816
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6280585666348544,
+      "learning_rate": 6.6634803856729805e-06,
+      "loss": 0.8471,
+      "step": 3817
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7434820802522482,
+      "learning_rate": 6.661838393360967e-06,
+      "loss": 0.9325,
+      "step": 3818
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7436636412597648,
+      "learning_rate": 6.660196199545318e-06,
+      "loss": 0.9595,
+      "step": 3819
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8173168222149345,
+      "learning_rate": 6.658553804425156e-06,
+      "loss": 0.8969,
+      "step": 3820
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7442794783196698,
+      "learning_rate": 6.656911208199628e-06,
+      "loss": 0.9485,
+      "step": 3821
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0920182408093924,
+      "learning_rate": 6.655268411067904e-06,
+      "loss": 1.0554,
+      "step": 3822
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8046815800837301,
+      "learning_rate": 6.6536254132291765e-06,
+      "loss": 0.8693,
+      "step": 3823
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.846421799904277,
+      "learning_rate": 6.651982214882669e-06,
+      "loss": 0.9414,
+      "step": 3824
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.832973563914968,
+      "learning_rate": 6.650338816227623e-06,
+      "loss": 0.9422,
+      "step": 3825
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8462046300419594,
+      "learning_rate": 6.6486952174633045e-06,
+      "loss": 0.811,
+      "step": 3826
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9099738734580007,
+      "learning_rate": 6.647051418789008e-06,
+      "loss": 0.8904,
+      "step": 3827
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8745610623910935,
+      "learning_rate": 6.645407420404048e-06,
+      "loss": 1.0305,
+      "step": 3828
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.950037508660796,
+      "learning_rate": 6.643763222507765e-06,
+      "loss": 0.9034,
+      "step": 3829
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7010749976426751,
+      "learning_rate": 6.642118825299526e-06,
+      "loss": 0.8652,
+      "step": 3830
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7935358777836763,
+      "learning_rate": 6.640474228978717e-06,
+      "loss": 0.9422,
+      "step": 3831
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8387611459340544,
+      "learning_rate": 6.638829433744753e-06,
+      "loss": 1.0706,
+      "step": 3832
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7736813098131814,
+      "learning_rate": 6.63718443979707e-06,
+      "loss": 0.8283,
+      "step": 3833
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7613412868740619,
+      "learning_rate": 6.635539247335129e-06,
+      "loss": 0.877,
+      "step": 3834
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.697444426360806,
+      "learning_rate": 6.633893856558416e-06,
+      "loss": 0.9267,
+      "step": 3835
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8319397947035014,
+      "learning_rate": 6.63224826766644e-06,
+      "loss": 0.9298,
+      "step": 3836
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7039154392053841,
+      "learning_rate": 6.630602480858734e-06,
+      "loss": 0.8515,
+      "step": 3837
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7803712196467671,
+      "learning_rate": 6.628956496334856e-06,
+      "loss": 0.9085,
+      "step": 3838
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8640364675587728,
+      "learning_rate": 6.627310314294386e-06,
+      "loss": 0.9888,
+      "step": 3839
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7934083882854839,
+      "learning_rate": 6.6256639349369326e-06,
+      "loss": 0.8424,
+      "step": 3840
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7860277104017066,
+      "learning_rate": 6.624017358462122e-06,
+      "loss": 0.8804,
+      "step": 3841
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.662337117575978,
+      "learning_rate": 6.622370585069605e-06,
+      "loss": 0.9006,
+      "step": 3842
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9107321234658777,
+      "learning_rate": 6.6207236149590635e-06,
+      "loss": 1.0424,
+      "step": 3843
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7304844065830798,
+      "learning_rate": 6.619076448330197e-06,
+      "loss": 0.8863,
+      "step": 3844
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7110270107540906,
+      "learning_rate": 6.6174290853827275e-06,
+      "loss": 0.8523,
+      "step": 3845
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7947723638006057,
+      "learning_rate": 6.615781526316407e-06,
+      "loss": 0.9608,
+      "step": 3846
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8628670940753461,
+      "learning_rate": 6.614133771331006e-06,
+      "loss": 1.0333,
+      "step": 3847
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.053019135918526,
+      "learning_rate": 6.612485820626318e-06,
+      "loss": 0.9534,
+      "step": 3848
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7457243225121548,
+      "learning_rate": 6.610837674402167e-06,
+      "loss": 0.8667,
+      "step": 3849
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.68558955804271,
+      "learning_rate": 6.609189332858394e-06,
+      "loss": 0.8522,
+      "step": 3850
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6722147837025172,
+      "learning_rate": 6.607540796194867e-06,
+      "loss": 0.8847,
+      "step": 3851
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8269058601456925,
+      "learning_rate": 6.605892064611478e-06,
+      "loss": 0.9067,
+      "step": 3852
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6669734213477019,
+      "learning_rate": 6.604243138308138e-06,
+      "loss": 0.8944,
+      "step": 3853
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6931219282539276,
+      "learning_rate": 6.602594017484786e-06,
+      "loss": 0.893,
+      "step": 3854
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8834788772576254,
+      "learning_rate": 6.600944702341385e-06,
+      "loss": 0.9463,
+      "step": 3855
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7777244843479938,
+      "learning_rate": 6.599295193077921e-06,
+      "loss": 0.9048,
+      "step": 3856
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8806687181074289,
+      "learning_rate": 6.5976454898944e-06,
+      "loss": 1.047,
+      "step": 3857
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.837146852289688,
+      "learning_rate": 6.595995592990856e-06,
+      "loss": 0.9498,
+      "step": 3858
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.628324910723718,
+      "learning_rate": 6.594345502567343e-06,
+      "loss": 0.9036,
+      "step": 3859
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6883453585095207,
+      "learning_rate": 6.592695218823943e-06,
+      "loss": 0.948,
+      "step": 3860
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6154040961610314,
+      "learning_rate": 6.591044741960755e-06,
+      "loss": 0.8979,
+      "step": 3861
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6721900192409,
+      "learning_rate": 6.589394072177908e-06,
+      "loss": 0.8374,
+      "step": 3862
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.653179880318606,
+      "learning_rate": 6.58774320967555e-06,
+      "loss": 0.8875,
+      "step": 3863
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6942456114237976,
+      "learning_rate": 6.586092154653855e-06,
+      "loss": 0.957,
+      "step": 3864
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.78447349787036,
+      "learning_rate": 6.584440907313017e-06,
+      "loss": 0.9114,
+      "step": 3865
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5381441869584935,
+      "learning_rate": 6.582789467853257e-06,
+      "loss": 0.9568,
+      "step": 3866
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6679679679252688,
+      "learning_rate": 6.581137836474817e-06,
+      "loss": 0.9593,
+      "step": 3867
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7474066060043723,
+      "learning_rate": 6.579486013377963e-06,
+      "loss": 0.904,
+      "step": 3868
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.730546087814758,
+      "learning_rate": 6.577833998762986e-06,
+      "loss": 0.8936,
+      "step": 3869
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.059028743171172,
+      "learning_rate": 6.576181792830194e-06,
+      "loss": 0.8219,
+      "step": 3870
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7583059057886148,
+      "learning_rate": 6.574529395779928e-06,
+      "loss": 1.01,
+      "step": 3871
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7226902256103813,
+      "learning_rate": 6.572876807812545e-06,
+      "loss": 0.9866,
+      "step": 3872
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.711169858080197,
+      "learning_rate": 6.571224029128426e-06,
+      "loss": 0.9674,
+      "step": 3873
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.760698996191674,
+      "learning_rate": 6.569571059927977e-06,
+      "loss": 0.8762,
+      "step": 3874
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.724970697887914,
+      "learning_rate": 6.567917900411625e-06,
+      "loss": 0.8974,
+      "step": 3875
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.686425209175494,
+      "learning_rate": 6.566264550779821e-06,
+      "loss": 0.9049,
+      "step": 3876
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6879831993005174,
+      "learning_rate": 6.564611011233043e-06,
+      "loss": 0.8525,
+      "step": 3877
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7284367575849662,
+      "learning_rate": 6.562957281971786e-06,
+      "loss": 0.8917,
+      "step": 3878
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8029067057431556,
+      "learning_rate": 6.561303363196568e-06,
+      "loss": 0.8967,
+      "step": 3879
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6411528920854073,
+      "learning_rate": 6.559649255107936e-06,
+      "loss": 0.9489,
+      "step": 3880
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.799262702319796,
+      "learning_rate": 6.557994957906455e-06,
+      "loss": 0.8672,
+      "step": 3881
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7653157234886843,
+      "learning_rate": 6.556340471792713e-06,
+      "loss": 0.9603,
+      "step": 3882
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6652517212523792,
+      "learning_rate": 6.554685796967324e-06,
+      "loss": 0.9147,
+      "step": 3883
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0330163361934335,
+      "learning_rate": 6.553030933630922e-06,
+      "loss": 1.0568,
+      "step": 3884
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8135812050180837,
+      "learning_rate": 6.551375881984166e-06,
+      "loss": 0.8764,
+      "step": 3885
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7775067955689776,
+      "learning_rate": 6.549720642227735e-06,
+      "loss": 0.8721,
+      "step": 3886
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6771465572452495,
+      "learning_rate": 6.548065214562333e-06,
+      "loss": 0.8979,
+      "step": 3887
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6996344054774135,
+      "learning_rate": 6.546409599188687e-06,
+      "loss": 0.891,
+      "step": 3888
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.897490996977252,
+      "learning_rate": 6.5447537963075485e-06,
+      "loss": 0.9201,
+      "step": 3889
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9159743141321777,
+      "learning_rate": 6.543097806119682e-06,
+      "loss": 0.9008,
+      "step": 3890
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.800614194161361,
+      "learning_rate": 6.54144162882589e-06,
+      "loss": 0.8942,
+      "step": 3891
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.674415503151337,
+      "learning_rate": 6.539785264626985e-06,
+      "loss": 0.8977,
+      "step": 3892
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7435893311053547,
+      "learning_rate": 6.538128713723808e-06,
+      "loss": 0.9838,
+      "step": 3893
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.612309517443607,
+      "learning_rate": 6.536471976317223e-06,
+      "loss": 0.7666,
+      "step": 3894
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.738024756971578,
+      "learning_rate": 6.534815052608114e-06,
+      "loss": 0.9103,
+      "step": 3895
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6013460627046625,
+      "learning_rate": 6.533157942797387e-06,
+      "loss": 0.9597,
+      "step": 3896
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8089269631135607,
+      "learning_rate": 6.531500647085974e-06,
+      "loss": 0.9431,
+      "step": 3897
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8385650563692855,
+      "learning_rate": 6.529843165674828e-06,
+      "loss": 0.8698,
+      "step": 3898
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6575087445033134,
+      "learning_rate": 6.528185498764925e-06,
+      "loss": 0.9394,
+      "step": 3899
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6187725412650043,
+      "learning_rate": 6.526527646557261e-06,
+      "loss": 0.9373,
+      "step": 3900
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8889445948591954,
+      "learning_rate": 6.5248696092528565e-06,
+      "loss": 0.8645,
+      "step": 3901
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6768692053843621,
+      "learning_rate": 6.523211387052756e-06,
+      "loss": 0.8966,
+      "step": 3902
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6433509210642596,
+      "learning_rate": 6.521552980158024e-06,
+      "loss": 0.8593,
+      "step": 3903
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7480503388626276,
+      "learning_rate": 6.519894388769745e-06,
+      "loss": 0.9619,
+      "step": 3904
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0974724697563039,
+      "learning_rate": 6.518235613089034e-06,
+      "loss": 1.0788,
+      "step": 3905
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.745344751096273,
+      "learning_rate": 6.516576653317019e-06,
+      "loss": 0.9584,
+      "step": 3906
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8307907422565648,
+      "learning_rate": 6.514917509654856e-06,
+      "loss": 1.0218,
+      "step": 3907
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.720183018886266,
+      "learning_rate": 6.513258182303724e-06,
+      "loss": 0.8331,
+      "step": 3908
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7199463788782183,
+      "learning_rate": 6.511598671464821e-06,
+      "loss": 0.9171,
+      "step": 3909
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6756481394455491,
+      "learning_rate": 6.509938977339366e-06,
+      "loss": 0.9113,
+      "step": 3910
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.721321574618392,
+      "learning_rate": 6.508279100128605e-06,
+      "loss": 0.8495,
+      "step": 3911
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9797567203308861,
+      "learning_rate": 6.5066190400338035e-06,
+      "loss": 0.8256,
+      "step": 3912
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8634837856969013,
+      "learning_rate": 6.5049587972562485e-06,
+      "loss": 0.9505,
+      "step": 3913
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7727178916276758,
+      "learning_rate": 6.503298371997252e-06,
+      "loss": 0.8117,
+      "step": 3914
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6939664743156453,
+      "learning_rate": 6.501637764458145e-06,
+      "loss": 0.9484,
+      "step": 3915
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.866707787904892,
+      "learning_rate": 6.499976974840282e-06,
+      "loss": 0.9027,
+      "step": 3916
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7007876571985032,
+      "learning_rate": 6.4983160033450395e-06,
+      "loss": 0.8728,
+      "step": 3917
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7444909663733197,
+      "learning_rate": 6.496654850173816e-06,
+      "loss": 0.8781,
+      "step": 3918
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8385110408137166,
+      "learning_rate": 6.494993515528032e-06,
+      "loss": 0.9384,
+      "step": 3919
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.67044490482577,
+      "learning_rate": 6.493331999609132e-06,
+      "loss": 0.9758,
+      "step": 3920
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7188206759933016,
+      "learning_rate": 6.4916703026185755e-06,
+      "loss": 0.9466,
+      "step": 3921
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0032155462418202,
+      "learning_rate": 6.490008424757855e-06,
+      "loss": 1.0679,
+      "step": 3922
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8985243074379652,
+      "learning_rate": 6.488346366228475e-06,
+      "loss": 1.0765,
+      "step": 3923
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.3451240214287137,
+      "learning_rate": 6.4866841272319665e-06,
+      "loss": 0.925,
+      "step": 3924
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7369663048698247,
+      "learning_rate": 6.485021707969883e-06,
+      "loss": 0.9186,
+      "step": 3925
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7967996895582161,
+      "learning_rate": 6.483359108643798e-06,
+      "loss": 0.8643,
+      "step": 3926
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.0005325574915815,
+      "learning_rate": 6.481696329455307e-06,
+      "loss": 1.0052,
+      "step": 3927
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7420558287588246,
+      "learning_rate": 6.480033370606027e-06,
+      "loss": 0.9119,
+      "step": 3928
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6268550922206093,
+      "learning_rate": 6.478370232297599e-06,
+      "loss": 0.858,
+      "step": 3929
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6912582964924663,
+      "learning_rate": 6.476706914731683e-06,
+      "loss": 0.8588,
+      "step": 3930
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6199594314717038,
+      "learning_rate": 6.475043418109966e-06,
+      "loss": 0.9028,
+      "step": 3931
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6886780898064158,
+      "learning_rate": 6.473379742634145e-06,
+      "loss": 0.9625,
+      "step": 3932
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9870782148964425,
+      "learning_rate": 6.471715888505952e-06,
+      "loss": 0.9951,
+      "step": 3933
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6640620305644407,
+      "learning_rate": 6.470051855927134e-06,
+      "loss": 0.9394,
+      "step": 3934
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8339189250132935,
+      "learning_rate": 6.46838764509946e-06,
+      "loss": 0.9783,
+      "step": 3935
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.0192965646508627,
+      "learning_rate": 6.466723256224723e-06,
+      "loss": 0.9479,
+      "step": 3936
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6930254999573584,
+      "learning_rate": 6.465058689504734e-06,
+      "loss": 0.8802,
+      "step": 3937
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.76282222904526,
+      "learning_rate": 6.463393945141328e-06,
+      "loss": 0.9176,
+      "step": 3938
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.839147944027083,
+      "learning_rate": 6.4617290233363604e-06,
+      "loss": 0.9054,
+      "step": 3939
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7003596365584817,
+      "learning_rate": 6.46006392429171e-06,
+      "loss": 0.9559,
+      "step": 3940
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5580418437253907,
+      "learning_rate": 6.458398648209274e-06,
+      "loss": 1.0624,
+      "step": 3941
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7760285415949977,
+      "learning_rate": 6.456733195290976e-06,
+      "loss": 0.9325,
+      "step": 3942
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7665553398276648,
+      "learning_rate": 6.455067565738755e-06,
+      "loss": 0.9125,
+      "step": 3943
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.697421218056344,
+      "learning_rate": 6.453401759754574e-06,
+      "loss": 0.9738,
+      "step": 3944
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8520398869277461,
+      "learning_rate": 6.451735777540421e-06,
+      "loss": 0.8812,
+      "step": 3945
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7470787357341448,
+      "learning_rate": 6.450069619298299e-06,
+      "loss": 0.8844,
+      "step": 3946
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7516565544068061,
+      "learning_rate": 6.448403285230238e-06,
+      "loss": 1.0351,
+      "step": 3947
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8413578513050175,
+      "learning_rate": 6.446736775538284e-06,
+      "loss": 0.945,
+      "step": 3948
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7074259037795683,
+      "learning_rate": 6.4450700904245076e-06,
+      "loss": 0.953,
+      "step": 3949
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.116914206446147,
+      "learning_rate": 6.443403230091002e-06,
+      "loss": 0.8957,
+      "step": 3950
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8465286713590245,
+      "learning_rate": 6.4417361947398785e-06,
+      "loss": 0.8642,
+      "step": 3951
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8075240989910308,
+      "learning_rate": 6.440068984573271e-06,
+      "loss": 0.9479,
+      "step": 3952
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7044332375628606,
+      "learning_rate": 6.438401599793333e-06,
+      "loss": 0.8821,
+      "step": 3953
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7741657267084037,
+      "learning_rate": 6.436734040602244e-06,
+      "loss": 1.0103,
+      "step": 3954
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7326726654355806,
+      "learning_rate": 6.435066307202198e-06,
+      "loss": 0.8898,
+      "step": 3955
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7888952608162794,
+      "learning_rate": 6.4333983997954155e-06,
+      "loss": 0.9306,
+      "step": 3956
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9756505815040522,
+      "learning_rate": 6.431730318584135e-06,
+      "loss": 1.0298,
+      "step": 3957
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7784356931637022,
+      "learning_rate": 6.430062063770619e-06,
+      "loss": 0.9815,
+      "step": 3958
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7857967490961197,
+      "learning_rate": 6.428393635557146e-06,
+      "loss": 0.8673,
+      "step": 3959
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6484432981785642,
+      "learning_rate": 6.426725034146021e-06,
+      "loss": 0.8174,
+      "step": 3960
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7222272117479835,
+      "learning_rate": 6.425056259739566e-06,
+      "loss": 0.9717,
+      "step": 3961
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7242847296163148,
+      "learning_rate": 6.4233873125401266e-06,
+      "loss": 0.9055,
+      "step": 3962
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7856650571672923,
+      "learning_rate": 6.4217181927500695e-06,
+      "loss": 0.8802,
+      "step": 3963
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5675209487734652,
+      "learning_rate": 6.4200489005717806e-06,
+      "loss": 0.842,
+      "step": 3964
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6629797683005096,
+      "learning_rate": 6.418379436207664e-06,
+      "loss": 0.9133,
+      "step": 3965
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9033203261972595,
+      "learning_rate": 6.416709799860152e-06,
+      "loss": 0.8796,
+      "step": 3966
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7324247175128238,
+      "learning_rate": 6.4150399917316945e-06,
+      "loss": 0.9172,
+      "step": 3967
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8699551433515686,
+      "learning_rate": 6.41337001202476e-06,
+      "loss": 1.0365,
+      "step": 3968
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8835222291898363,
+      "learning_rate": 6.411699860941836e-06,
+      "loss": 1.0763,
+      "step": 3969
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8684088455439272,
+      "learning_rate": 6.410029538685438e-06,
+      "loss": 0.9139,
+      "step": 3970
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7258885343994286,
+      "learning_rate": 6.408359045458099e-06,
+      "loss": 0.9413,
+      "step": 3971
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8320453269636148,
+      "learning_rate": 6.4066883814623674e-06,
+      "loss": 1.042,
+      "step": 3972
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8506221638523055,
+      "learning_rate": 6.4050175469008225e-06,
+      "loss": 0.9586,
+      "step": 3973
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7195702062108829,
+      "learning_rate": 6.403346541976057e-06,
+      "loss": 0.9008,
+      "step": 3974
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7074178320775972,
+      "learning_rate": 6.401675366890682e-06,
+      "loss": 0.8398,
+      "step": 3975
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8912538092868443,
+      "learning_rate": 6.400004021847338e-06,
+      "loss": 0.9583,
+      "step": 3976
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7486165996178227,
+      "learning_rate": 6.39833250704868e-06,
+      "loss": 0.887,
+      "step": 3977
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7044796776542033,
+      "learning_rate": 6.396660822697384e-06,
+      "loss": 0.923,
+      "step": 3978
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7619137996741225,
+      "learning_rate": 6.39498896899615e-06,
+      "loss": 0.8927,
+      "step": 3979
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.739431671006542,
+      "learning_rate": 6.393316946147692e-06,
+      "loss": 0.8563,
+      "step": 3980
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6361367810204508,
+      "learning_rate": 6.391644754354751e-06,
+      "loss": 0.8372,
+      "step": 3981
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7833271201274037,
+      "learning_rate": 6.389972393820088e-06,
+      "loss": 0.9515,
+      "step": 3982
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.963016328471912,
+      "learning_rate": 6.388299864746476e-06,
+      "loss": 0.8727,
+      "step": 3983
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.838679891182311,
+      "learning_rate": 6.386627167336724e-06,
+      "loss": 1.0075,
+      "step": 3984
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7743819031502654,
+      "learning_rate": 6.384954301793647e-06,
+      "loss": 0.9329,
+      "step": 3985
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.1073906245492458,
+      "learning_rate": 6.3832812683200825e-06,
+      "loss": 0.9673,
+      "step": 3986
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7631329856644022,
+      "learning_rate": 6.381608067118899e-06,
+      "loss": 0.8673,
+      "step": 3987
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9778189708453292,
+      "learning_rate": 6.379934698392973e-06,
+      "loss": 1.0379,
+      "step": 3988
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7192378265329664,
+      "learning_rate": 6.378261162345207e-06,
+      "loss": 0.8959,
+      "step": 3989
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8995153084526877,
+      "learning_rate": 6.3765874591785245e-06,
+      "loss": 1.0392,
+      "step": 3990
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.648183731366141,
+      "learning_rate": 6.374913589095866e-06,
+      "loss": 0.9137,
+      "step": 3991
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8193897880654242,
+      "learning_rate": 6.373239552300194e-06,
+      "loss": 0.8234,
+      "step": 3992
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.976808648283257,
+      "learning_rate": 6.371565348994493e-06,
+      "loss": 1.0109,
+      "step": 3993
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.71385962557775,
+      "learning_rate": 6.3698909793817656e-06,
+      "loss": 0.9742,
+      "step": 3994
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8390663291313467,
+      "learning_rate": 6.3682164436650326e-06,
+      "loss": 0.878,
+      "step": 3995
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6588951781326622,
+      "learning_rate": 6.366541742047341e-06,
+      "loss": 0.8933,
+      "step": 3996
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6787808450692152,
+      "learning_rate": 6.364866874731749e-06,
+      "loss": 0.8814,
+      "step": 3997
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.593626293404446,
+      "learning_rate": 6.363191841921345e-06,
+      "loss": 0.923,
+      "step": 3998
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8482915144525396,
+      "learning_rate": 6.36151664381923e-06,
+      "loss": 0.9699,
+      "step": 3999
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7151242217610918,
+      "learning_rate": 6.359841280628529e-06,
+      "loss": 0.8672,
+      "step": 4000
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.787131297260499,
+      "learning_rate": 6.358165752552383e-06,
+      "loss": 0.8739,
+      "step": 4001
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7099209625632075,
+      "learning_rate": 6.356490059793959e-06,
+      "loss": 0.8755,
+      "step": 4002
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6391885664671784,
+      "learning_rate": 6.354814202556437e-06,
+      "loss": 0.9039,
+      "step": 4003
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.740366418645429,
+      "learning_rate": 6.353138181043024e-06,
+      "loss": 0.8881,
+      "step": 4004
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7480268695522394,
+      "learning_rate": 6.351461995456941e-06,
+      "loss": 0.9565,
+      "step": 4005
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.0009328208351045,
+      "learning_rate": 6.349785646001431e-06,
+      "loss": 0.9025,
+      "step": 4006
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8009717122738471,
+      "learning_rate": 6.348109132879758e-06,
+      "loss": 0.903,
+      "step": 4007
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8504021963425317,
+      "learning_rate": 6.346432456295206e-06,
+      "loss": 0.9411,
+      "step": 4008
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.706813212320307,
+      "learning_rate": 6.344755616451075e-06,
+      "loss": 0.8389,
+      "step": 4009
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7051543665825062,
+      "learning_rate": 6.343078613550691e-06,
+      "loss": 0.8774,
+      "step": 4010
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7052958775982534,
+      "learning_rate": 6.341401447797395e-06,
+      "loss": 0.9244,
+      "step": 4011
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7093526731171986,
+      "learning_rate": 6.339724119394548e-06,
+      "loss": 0.9053,
+      "step": 4012
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1870874952329995,
+      "learning_rate": 6.3380466285455336e-06,
+      "loss": 1.0854,
+      "step": 4013
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7625268872432795,
+      "learning_rate": 6.336368975453752e-06,
+      "loss": 0.8794,
+      "step": 4014
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.879391858440186,
+      "learning_rate": 6.3346911603226245e-06,
+      "loss": 0.9488,
+      "step": 4015
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.681810652228696,
+      "learning_rate": 6.333013183355594e-06,
+      "loss": 0.8792,
+      "step": 4016
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8575135165372507,
+      "learning_rate": 6.331335044756119e-06,
+      "loss": 0.9689,
+      "step": 4017
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.663544587754942,
+      "learning_rate": 6.329656744727679e-06,
+      "loss": 0.9284,
+      "step": 4018
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.740260647611781,
+      "learning_rate": 6.3279782834737755e-06,
+      "loss": 0.9442,
+      "step": 4019
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8368563061691017,
+      "learning_rate": 6.326299661197926e-06,
+      "loss": 0.8632,
+      "step": 4020
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8817559943663245,
+      "learning_rate": 6.324620878103671e-06,
+      "loss": 0.8644,
+      "step": 4021
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7738441166883678,
+      "learning_rate": 6.322941934394568e-06,
+      "loss": 0.9868,
+      "step": 4022
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7011821433322205,
+      "learning_rate": 6.321262830274193e-06,
+      "loss": 0.8193,
+      "step": 4023
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7111022798620577,
+      "learning_rate": 6.319583565946147e-06,
+      "loss": 0.8641,
+      "step": 4024
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8569972600601368,
+      "learning_rate": 6.317904141614043e-06,
+      "loss": 0.8861,
+      "step": 4025
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7422377554277533,
+      "learning_rate": 6.3162245574815186e-06,
+      "loss": 0.946,
+      "step": 4026
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.1030757388947032,
+      "learning_rate": 6.314544813752229e-06,
+      "loss": 0.8271,
+      "step": 4027
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7798087861937413,
+      "learning_rate": 6.312864910629848e-06,
+      "loss": 0.9319,
+      "step": 4028
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7006027908633905,
+      "learning_rate": 6.311184848318072e-06,
+      "loss": 0.9137,
+      "step": 4029
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9191937380322668,
+      "learning_rate": 6.309504627020612e-06,
+      "loss": 1.0761,
+      "step": 4030
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8511165826895926,
+      "learning_rate": 6.3078242469412e-06,
+      "loss": 0.9306,
+      "step": 4031
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6776837616127325,
+      "learning_rate": 6.306143708283592e-06,
+      "loss": 0.8846,
+      "step": 4032
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.739803937852351,
+      "learning_rate": 6.3044630112515545e-06,
+      "loss": 0.9539,
+      "step": 4033
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6866745276215105,
+      "learning_rate": 6.30278215604888e-06,
+      "loss": 0.8769,
+      "step": 4034
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6633010456857806,
+      "learning_rate": 6.3011011428793776e-06,
+      "loss": 0.9749,
+      "step": 4035
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7736829050694976,
+      "learning_rate": 6.299419971946876e-06,
+      "loss": 0.9624,
+      "step": 4036
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7377674810961996,
+      "learning_rate": 6.297738643455225e-06,
+      "loss": 0.8626,
+      "step": 4037
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8335939291285382,
+      "learning_rate": 6.2960571576082875e-06,
+      "loss": 1.0489,
+      "step": 4038
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.811280321684043,
+      "learning_rate": 6.294375514609952e-06,
+      "loss": 0.8904,
+      "step": 4039
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6895227813165294,
+      "learning_rate": 6.292693714664122e-06,
+      "loss": 0.8624,
+      "step": 4040
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6843977924332796,
+      "learning_rate": 6.291011757974723e-06,
+      "loss": 0.9358,
+      "step": 4041
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.840856944242189,
+      "learning_rate": 6.289329644745699e-06,
+      "loss": 0.984,
+      "step": 4042
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.698903197468234,
+      "learning_rate": 6.287647375181009e-06,
+      "loss": 0.8886,
+      "step": 4043
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7307850139654541,
+      "learning_rate": 6.285964949484636e-06,
+      "loss": 0.9188,
+      "step": 4044
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7722243210098503,
+      "learning_rate": 6.284282367860579e-06,
+      "loss": 0.9404,
+      "step": 4045
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7644428767719205,
+      "learning_rate": 6.282599630512858e-06,
+      "loss": 0.8908,
+      "step": 4046
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7023937339725024,
+      "learning_rate": 6.280916737645511e-06,
+      "loss": 0.9174,
+      "step": 4047
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.755403687189088,
+      "learning_rate": 6.2792336894625915e-06,
+      "loss": 0.929,
+      "step": 4048
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.721359899662644,
+      "learning_rate": 6.2775504861681775e-06,
+      "loss": 0.883,
+      "step": 4049
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6627270644213084,
+      "learning_rate": 6.275867127966364e-06,
+      "loss": 0.9124,
+      "step": 4050
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7445372397088001,
+      "learning_rate": 6.27418361506126e-06,
+      "loss": 0.887,
+      "step": 4051
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7162927245855721,
+      "learning_rate": 6.272499947657002e-06,
+      "loss": 0.976,
+      "step": 4052
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.829345682744822,
+      "learning_rate": 6.270816125957739e-06,
+      "loss": 1.0366,
+      "step": 4053
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.840423253429475,
+      "learning_rate": 6.269132150167639e-06,
+      "loss": 1.0693,
+      "step": 4054
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6864759518847927,
+      "learning_rate": 6.267448020490889e-06,
+      "loss": 0.859,
+      "step": 4055
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7011504529370403,
+      "learning_rate": 6.265763737131698e-06,
+      "loss": 0.8868,
+      "step": 4056
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8716944692251618,
+      "learning_rate": 6.26407930029429e-06,
+      "loss": 0.9416,
+      "step": 4057
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8751610096717009,
+      "learning_rate": 6.26239471018291e-06,
+      "loss": 1.067,
+      "step": 4058
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7107574710553786,
+      "learning_rate": 6.260709967001816e-06,
+      "loss": 0.931,
+      "step": 4059
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8801280309424162,
+      "learning_rate": 6.259025070955296e-06,
+      "loss": 1.0653,
+      "step": 4060
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7437068473879977,
+      "learning_rate": 6.2573400222476435e-06,
+      "loss": 0.9227,
+      "step": 4061
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7658704965340868,
+      "learning_rate": 6.255654821083178e-06,
+      "loss": 0.8665,
+      "step": 4062
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6529819362262912,
+      "learning_rate": 6.253969467666239e-06,
+      "loss": 0.8847,
+      "step": 4063
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8001116189743072,
+      "learning_rate": 6.252283962201177e-06,
+      "loss": 0.9323,
+      "step": 4064
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8697944340743636,
+      "learning_rate": 6.2505983048923675e-06,
+      "loss": 1.0275,
+      "step": 4065
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7934903149312482,
+      "learning_rate": 6.2489124959442026e-06,
+      "loss": 0.8545,
+      "step": 4066
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6538995393255496,
+      "learning_rate": 6.247226535561092e-06,
+      "loss": 0.8296,
+      "step": 4067
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.754819005550529,
+      "learning_rate": 6.245540423947463e-06,
+      "loss": 0.9827,
+      "step": 4068
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8111568447737512,
+      "learning_rate": 6.243854161307765e-06,
+      "loss": 0.9299,
+      "step": 4069
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7887977407009261,
+      "learning_rate": 6.242167747846461e-06,
+      "loss": 1.0473,
+      "step": 4070
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7272270334963726,
+      "learning_rate": 6.240481183768036e-06,
+      "loss": 0.9524,
+      "step": 4071
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7026660414866062,
+      "learning_rate": 6.2387944692769896e-06,
+      "loss": 0.887,
+      "step": 4072
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7984945734698372,
+      "learning_rate": 6.237107604577843e-06,
+      "loss": 1.0392,
+      "step": 4073
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.65285112778998,
+      "learning_rate": 6.235420589875136e-06,
+      "loss": 0.8028,
+      "step": 4074
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7043596660159106,
+      "learning_rate": 6.233733425373423e-06,
+      "loss": 0.8184,
+      "step": 4075
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5769506612191302,
+      "learning_rate": 6.232046111277277e-06,
+      "loss": 0.931,
+      "step": 4076
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7743485144190552,
+      "learning_rate": 6.230358647791293e-06,
+      "loss": 0.905,
+      "step": 4077
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.684035043513552,
+      "learning_rate": 6.228671035120083e-06,
+      "loss": 0.9109,
+      "step": 4078
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.643274481846152,
+      "learning_rate": 6.226983273468273e-06,
+      "loss": 0.9696,
+      "step": 4079
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6374746363121848,
+      "learning_rate": 6.225295363040511e-06,
+      "loss": 0.9106,
+      "step": 4080
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6788832780940153,
+      "learning_rate": 6.2236073040414625e-06,
+      "loss": 0.973,
+      "step": 4081
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7705298119470898,
+      "learning_rate": 6.221919096675808e-06,
+      "loss": 0.9792,
+      "step": 4082
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.721295931251655,
+      "learning_rate": 6.220230741148253e-06,
+      "loss": 0.836,
+      "step": 4083
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.755348691293493,
+      "learning_rate": 6.218542237663513e-06,
+      "loss": 0.9281,
+      "step": 4084
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.644138339325869,
+      "learning_rate": 6.216853586426326e-06,
+      "loss": 0.9661,
+      "step": 4085
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8759942270728294,
+      "learning_rate": 6.215164787641446e-06,
+      "loss": 0.8915,
+      "step": 4086
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6908116683292438,
+      "learning_rate": 6.2134758415136465e-06,
+      "loss": 0.9063,
+      "step": 4087
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.628252682535255,
+      "learning_rate": 6.211786748247717e-06,
+      "loss": 0.8381,
+      "step": 4088
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.53893147283309,
+      "learning_rate": 6.210097508048467e-06,
+      "loss": 0.9505,
+      "step": 4089
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9076972425766632,
+      "learning_rate": 6.208408121120722e-06,
+      "loss": 0.9477,
+      "step": 4090
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8303579836340071,
+      "learning_rate": 6.206718587669327e-06,
+      "loss": 0.8087,
+      "step": 4091
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7217632112574377,
+      "learning_rate": 6.205028907899143e-06,
+      "loss": 0.9272,
+      "step": 4092
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7062790110504575,
+      "learning_rate": 6.203339082015049e-06,
+      "loss": 0.9777,
+      "step": 4093
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8178178035978312,
+      "learning_rate": 6.2016491102219435e-06,
+      "loss": 0.9908,
+      "step": 4094
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9373185022399888,
+      "learning_rate": 6.19995899272474e-06,
+      "loss": 1.0453,
+      "step": 4095
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6949696724126513,
+      "learning_rate": 6.198268729728372e-06,
+      "loss": 0.8657,
+      "step": 4096
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.764931852583288,
+      "learning_rate": 6.1965783214377895e-06,
+      "loss": 0.9579,
+      "step": 4097
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8072555675100022,
+      "learning_rate": 6.19488776805796e-06,
+      "loss": 0.8671,
+      "step": 4098
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7560189720097226,
+      "learning_rate": 6.1931970697938695e-06,
+      "loss": 0.81,
+      "step": 4099
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7797001723841435,
+      "learning_rate": 6.191506226850521e-06,
+      "loss": 0.9307,
+      "step": 4100
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7368902863373312,
+      "learning_rate": 6.189815239432935e-06,
+      "loss": 0.9122,
+      "step": 4101
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7307633906267719,
+      "learning_rate": 6.188124107746148e-06,
+      "loss": 0.8597,
+      "step": 4102
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8966760810622938,
+      "learning_rate": 6.186432831995218e-06,
+      "loss": 0.8888,
+      "step": 4103
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7357836786170073,
+      "learning_rate": 6.184741412385217e-06,
+      "loss": 0.8524,
+      "step": 4104
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9123475037629505,
+      "learning_rate": 6.183049849121233e-06,
+      "loss": 0.9417,
+      "step": 4105
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6856157918660948,
+      "learning_rate": 6.18135814240838e-06,
+      "loss": 0.9272,
+      "step": 4106
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7840470728861872,
+      "learning_rate": 6.179666292451776e-06,
+      "loss": 0.9444,
+      "step": 4107
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6803240122309366,
+      "learning_rate": 6.1779742994565686e-06,
+      "loss": 0.9686,
+      "step": 4108
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9627885123668279,
+      "learning_rate": 6.176282163627917e-06,
+      "loss": 1.0448,
+      "step": 4109
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7829164959392731,
+      "learning_rate": 6.174589885170996e-06,
+      "loss": 0.8701,
+      "step": 4110
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.755903774850548,
+      "learning_rate": 6.172897464291005e-06,
+      "loss": 0.9063,
+      "step": 4111
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7555859358773085,
+      "learning_rate": 6.171204901193151e-06,
+      "loss": 0.8769,
+      "step": 4112
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8686098413489176,
+      "learning_rate": 6.169512196082663e-06,
+      "loss": 1.0658,
+      "step": 4113
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7992630422196227,
+      "learning_rate": 6.167819349164792e-06,
+      "loss": 0.9995,
+      "step": 4114
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.710193442257645,
+      "learning_rate": 6.166126360644797e-06,
+      "loss": 0.899,
+      "step": 4115
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.674506640978442,
+      "learning_rate": 6.1644332307279616e-06,
+      "loss": 0.8869,
+      "step": 4116
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.719516833122571,
+      "learning_rate": 6.1627399596195835e-06,
+      "loss": 0.9083,
+      "step": 4117
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8047172641033573,
+      "learning_rate": 6.161046547524976e-06,
+      "loss": 0.8978,
+      "step": 4118
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6801457428529358,
+      "learning_rate": 6.15935299464947e-06,
+      "loss": 0.8551,
+      "step": 4119
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.742807902931998,
+      "learning_rate": 6.157659301198418e-06,
+      "loss": 0.9632,
+      "step": 4120
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.762897988294096,
+      "learning_rate": 6.1559654673771865e-06,
+      "loss": 0.9121,
+      "step": 4121
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8643169353518152,
+      "learning_rate": 6.154271493391155e-06,
+      "loss": 0.8988,
+      "step": 4122
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9079408210279591,
+      "learning_rate": 6.152577379445725e-06,
+      "loss": 0.8817,
+      "step": 4123
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.70948683843766,
+      "learning_rate": 6.150883125746314e-06,
+      "loss": 0.902,
+      "step": 4124
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8590094273255577,
+      "learning_rate": 6.149188732498358e-06,
+      "loss": 0.9021,
+      "step": 4125
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7757050964141583,
+      "learning_rate": 6.147494199907306e-06,
+      "loss": 0.9212,
+      "step": 4126
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7140371471382743,
+      "learning_rate": 6.145799528178625e-06,
+      "loss": 0.9023,
+      "step": 4127
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7531331343336856,
+      "learning_rate": 6.1441047175178025e-06,
+      "loss": 0.8785,
+      "step": 4128
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9679592964718644,
+      "learning_rate": 6.142409768130339e-06,
+      "loss": 1.0468,
+      "step": 4129
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7086181600491628,
+      "learning_rate": 6.14071468022175e-06,
+      "loss": 0.8966,
+      "step": 4130
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7193607118884877,
+      "learning_rate": 6.1390194539975754e-06,
+      "loss": 0.8933,
+      "step": 4131
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7001835640469714,
+      "learning_rate": 6.137324089663365e-06,
+      "loss": 1.0109,
+      "step": 4132
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7059096047730586,
+      "learning_rate": 6.135628587424688e-06,
+      "loss": 0.9374,
+      "step": 4133
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.775002580371089,
+      "learning_rate": 6.133932947487129e-06,
+      "loss": 0.9947,
+      "step": 4134
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7486233352766245,
+      "learning_rate": 6.132237170056291e-06,
+      "loss": 0.9187,
+      "step": 4135
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.655032451748233,
+      "learning_rate": 6.130541255337792e-06,
+      "loss": 0.8396,
+      "step": 4136
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6641662513128037,
+      "learning_rate": 6.128845203537269e-06,
+      "loss": 0.9251,
+      "step": 4137
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6735195311240925,
+      "learning_rate": 6.127149014860374e-06,
+      "loss": 0.9133,
+      "step": 4138
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.212836225302568,
+      "learning_rate": 6.125452689512774e-06,
+      "loss": 0.8031,
+      "step": 4139
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7329577904516487,
+      "learning_rate": 6.123756227700156e-06,
+      "loss": 0.9344,
+      "step": 4140
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8415839705011045,
+      "learning_rate": 6.122059629628221e-06,
+      "loss": 1.0838,
+      "step": 4141
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.825704230416759,
+      "learning_rate": 6.1203628955026874e-06,
+      "loss": 0.9437,
+      "step": 4142
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6848501693058413,
+      "learning_rate": 6.118666025529292e-06,
+      "loss": 0.8848,
+      "step": 4143
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.687868731574758,
+      "learning_rate": 6.116969019913781e-06,
+      "loss": 0.8752,
+      "step": 4144
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8149817159463055,
+      "learning_rate": 6.115271878861928e-06,
+      "loss": 0.9666,
+      "step": 4145
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7316397426823136,
+      "learning_rate": 6.113574602579515e-06,
+      "loss": 0.9504,
+      "step": 4146
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.763981974194585,
+      "learning_rate": 6.1118771912723396e-06,
+      "loss": 0.8779,
+      "step": 4147
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8126288377661568,
+      "learning_rate": 6.110179645146225e-06,
+      "loss": 0.9693,
+      "step": 4148
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7585475078131667,
+      "learning_rate": 6.1084819644070004e-06,
+      "loss": 0.8359,
+      "step": 4149
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7328965246481918,
+      "learning_rate": 6.106784149260514e-06,
+      "loss": 0.8959,
+      "step": 4150
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.788033507189026,
+      "learning_rate": 6.1050861999126355e-06,
+      "loss": 0.9727,
+      "step": 4151
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.763627202658937,
+      "learning_rate": 6.103388116569245e-06,
+      "loss": 0.8772,
+      "step": 4152
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6889138416481313,
+      "learning_rate": 6.101689899436241e-06,
+      "loss": 0.9229,
+      "step": 4153
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8048592328971018,
+      "learning_rate": 6.0999915487195395e-06,
+      "loss": 0.915,
+      "step": 4154
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8384004378930034,
+      "learning_rate": 6.09829306462507e-06,
+      "loss": 1.0008,
+      "step": 4155
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8101276452294761,
+      "learning_rate": 6.096594447358779e-06,
+      "loss": 0.8511,
+      "step": 4156
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7904662449676756,
+      "learning_rate": 6.094895697126631e-06,
+      "loss": 0.92,
+      "step": 4157
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.790616532908676,
+      "learning_rate": 6.0931968141346054e-06,
+      "loss": 0.9694,
+      "step": 4158
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7711627198551483,
+      "learning_rate": 6.091497798588699e-06,
+      "loss": 0.9177,
+      "step": 4159
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.625545222489051,
+      "learning_rate": 6.08979865069492e-06,
+      "loss": 0.7992,
+      "step": 4160
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9112025291482904,
+      "learning_rate": 6.088099370659297e-06,
+      "loss": 1.0273,
+      "step": 4161
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8204381485970937,
+      "learning_rate": 6.086399958687875e-06,
+      "loss": 0.9607,
+      "step": 4162
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6948075399900262,
+      "learning_rate": 6.0847004149867125e-06,
+      "loss": 0.9564,
+      "step": 4163
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8259800858785558,
+      "learning_rate": 6.083000739761885e-06,
+      "loss": 1.0227,
+      "step": 4164
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7173192602465996,
+      "learning_rate": 6.081300933219485e-06,
+      "loss": 0.9705,
+      "step": 4165
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7482220741144299,
+      "learning_rate": 6.079600995565618e-06,
+      "loss": 0.9117,
+      "step": 4166
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8599253965464237,
+      "learning_rate": 6.077900927006408e-06,
+      "loss": 1.0675,
+      "step": 4167
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7689279574854704,
+      "learning_rate": 6.0762007277479955e-06,
+      "loss": 1.0033,
+      "step": 4168
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.898672475557904,
+      "learning_rate": 6.074500397996533e-06,
+      "loss": 0.8752,
+      "step": 4169
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6568772165901031,
+      "learning_rate": 6.072799937958195e-06,
+      "loss": 0.8991,
+      "step": 4170
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6931204364974504,
+      "learning_rate": 6.0710993478391645e-06,
+      "loss": 0.9325,
+      "step": 4171
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6199188526151513,
+      "learning_rate": 6.0693986278456456e-06,
+      "loss": 0.9271,
+      "step": 4172
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6720757280955523,
+      "learning_rate": 6.067697778183857e-06,
+      "loss": 0.9944,
+      "step": 4173
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6776850702841017,
+      "learning_rate": 6.065996799060031e-06,
+      "loss": 0.8823,
+      "step": 4174
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.864801747844051,
+      "learning_rate": 6.064295690680418e-06,
+      "loss": 0.9424,
+      "step": 4175
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.770172612206569,
+      "learning_rate": 6.0625944532512845e-06,
+      "loss": 0.9347,
+      "step": 4176
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.605435623035622,
+      "learning_rate": 6.06089308697891e-06,
+      "loss": 0.8525,
+      "step": 4177
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.74093231927897,
+      "learning_rate": 6.05919159206959e-06,
+      "loss": 0.8645,
+      "step": 4178
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.83362869926186,
+      "learning_rate": 6.0574899687296385e-06,
+      "loss": 0.9486,
+      "step": 4179
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6872494288783138,
+      "learning_rate": 6.055788217165384e-06,
+      "loss": 0.902,
+      "step": 4180
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8352697746365823,
+      "learning_rate": 6.054086337583166e-06,
+      "loss": 0.9087,
+      "step": 4181
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8479977272584176,
+      "learning_rate": 6.052384330189347e-06,
+      "loss": 0.8954,
+      "step": 4182
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8314958137444897,
+      "learning_rate": 6.0506821951903e-06,
+      "loss": 0.8738,
+      "step": 4183
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9445605441243102,
+      "learning_rate": 6.0489799327924125e-06,
+      "loss": 1.0654,
+      "step": 4184
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9012167092560271,
+      "learning_rate": 6.0472775432020945e-06,
+      "loss": 1.0733,
+      "step": 4185
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6887057100296337,
+      "learning_rate": 6.0455750266257625e-06,
+      "loss": 0.9363,
+      "step": 4186
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.0889512013702944,
+      "learning_rate": 6.043872383269855e-06,
+      "loss": 0.8935,
+      "step": 4187
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6822171667548373,
+      "learning_rate": 6.04216961334082e-06,
+      "loss": 0.859,
+      "step": 4188
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.991753969880893,
+      "learning_rate": 6.040466717045127e-06,
+      "loss": 0.912,
+      "step": 4189
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7243561098013682,
+      "learning_rate": 6.038763694589259e-06,
+      "loss": 0.9024,
+      "step": 4190
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.667798355190635,
+      "learning_rate": 6.0370605461797124e-06,
+      "loss": 0.9364,
+      "step": 4191
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.702556719045791,
+      "learning_rate": 6.035357272022997e-06,
+      "loss": 0.9056,
+      "step": 4192
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6750542875694823,
+      "learning_rate": 6.033653872325644e-06,
+      "loss": 0.8831,
+      "step": 4193
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8223344963739267,
+      "learning_rate": 6.031950347294196e-06,
+      "loss": 0.9441,
+      "step": 4194
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.696875915901807,
+      "learning_rate": 6.030246697135209e-06,
+      "loss": 0.9297,
+      "step": 4195
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.734544743142106,
+      "learning_rate": 6.028542922055261e-06,
+      "loss": 0.8738,
+      "step": 4196
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1163995622255767,
+      "learning_rate": 6.026839022260936e-06,
+      "loss": 1.0798,
+      "step": 4197
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6709604116812578,
+      "learning_rate": 6.0251349979588395e-06,
+      "loss": 0.8527,
+      "step": 4198
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7063395823550993,
+      "learning_rate": 6.0234308493555906e-06,
+      "loss": 0.8034,
+      "step": 4199
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6817934376184094,
+      "learning_rate": 6.021726576657823e-06,
+      "loss": 0.8811,
+      "step": 4200
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8639491505348974,
+      "learning_rate": 6.0200221800721845e-06,
+      "loss": 1.0423,
+      "step": 4201
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7097166220123987,
+      "learning_rate": 6.018317659805341e-06,
+      "loss": 0.9341,
+      "step": 4202
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8025752522178218,
+      "learning_rate": 6.01661301606397e-06,
+      "loss": 0.8928,
+      "step": 4203
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8899662195314362,
+      "learning_rate": 6.014908249054767e-06,
+      "loss": 0.8807,
+      "step": 4204
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6547038985566964,
+      "learning_rate": 6.0132033589844395e-06,
+      "loss": 0.8809,
+      "step": 4205
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7031743476214085,
+      "learning_rate": 6.011498346059712e-06,
+      "loss": 0.9122,
+      "step": 4206
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6879928488440026,
+      "learning_rate": 6.009793210487323e-06,
+      "loss": 0.9195,
+      "step": 4207
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.776632234515072,
+      "learning_rate": 6.008087952474025e-06,
+      "loss": 0.9295,
+      "step": 4208
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7543257191893757,
+      "learning_rate": 6.006382572226587e-06,
+      "loss": 0.9405,
+      "step": 4209
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8462332946543254,
+      "learning_rate": 6.004677069951793e-06,
+      "loss": 0.8343,
+      "step": 4210
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7117119613914764,
+      "learning_rate": 6.002971445856441e-06,
+      "loss": 0.9286,
+      "step": 4211
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9022704130135109,
+      "learning_rate": 6.001265700147344e-06,
+      "loss": 0.8858,
+      "step": 4212
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.739402805122118,
+      "learning_rate": 5.999559833031328e-06,
+      "loss": 0.8501,
+      "step": 4213
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7626646900527327,
+      "learning_rate": 5.997853844715237e-06,
+      "loss": 0.9089,
+      "step": 4214
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8759975857440898,
+      "learning_rate": 5.996147735405926e-06,
+      "loss": 0.9175,
+      "step": 4215
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7941165170561362,
+      "learning_rate": 5.994441505310269e-06,
+      "loss": 0.9319,
+      "step": 4216
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8850093931232688,
+      "learning_rate": 5.992735154635151e-06,
+      "loss": 0.8897,
+      "step": 4217
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8003724313625709,
+      "learning_rate": 5.991028683587472e-06,
+      "loss": 0.862,
+      "step": 4218
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5674610925265062,
+      "learning_rate": 5.98932209237415e-06,
+      "loss": 0.9377,
+      "step": 4219
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6836362719644644,
+      "learning_rate": 5.987615381202112e-06,
+      "loss": 0.8879,
+      "step": 4220
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8092850972538288,
+      "learning_rate": 5.985908550278306e-06,
+      "loss": 0.8558,
+      "step": 4221
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7639997714852444,
+      "learning_rate": 5.98420159980969e-06,
+      "loss": 0.9654,
+      "step": 4222
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.769710424022008,
+      "learning_rate": 5.982494530003233e-06,
+      "loss": 0.8426,
+      "step": 4223
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7734298219549138,
+      "learning_rate": 5.98078734106593e-06,
+      "loss": 0.9465,
+      "step": 4224
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.740687329906866,
+      "learning_rate": 5.97908003320478e-06,
+      "loss": 0.8633,
+      "step": 4225
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.781047897234036,
+      "learning_rate": 5.9773726066268e-06,
+      "loss": 0.972,
+      "step": 4226
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8462053547948605,
+      "learning_rate": 5.9756650615390226e-06,
+      "loss": 0.8841,
+      "step": 4227
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9740545298295629,
+      "learning_rate": 5.973957398148493e-06,
+      "loss": 1.0412,
+      "step": 4228
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6773555959092912,
+      "learning_rate": 5.972249616662271e-06,
+      "loss": 0.9932,
+      "step": 4229
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7233561777747886,
+      "learning_rate": 5.970541717287432e-06,
+      "loss": 0.9499,
+      "step": 4230
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.776659568007338,
+      "learning_rate": 5.968833700231062e-06,
+      "loss": 0.9525,
+      "step": 4231
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7306482699067538,
+      "learning_rate": 5.967125565700266e-06,
+      "loss": 0.9211,
+      "step": 4232
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6726863561119891,
+      "learning_rate": 5.965417313902162e-06,
+      "loss": 0.9478,
+      "step": 4233
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7176739152729557,
+      "learning_rate": 5.9637089450438805e-06,
+      "loss": 0.9949,
+      "step": 4234
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.717804756136136,
+      "learning_rate": 5.962000459332566e-06,
+      "loss": 0.9186,
+      "step": 4235
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7704816125365117,
+      "learning_rate": 5.96029185697538e-06,
+      "loss": 0.9405,
+      "step": 4236
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8219912416901443,
+      "learning_rate": 5.958583138179494e-06,
+      "loss": 1.0529,
+      "step": 4237
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.823968776385327,
+      "learning_rate": 5.9568743031521e-06,
+      "loss": 0.93,
+      "step": 4238
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8856193668218744,
+      "learning_rate": 5.955165352100399e-06,
+      "loss": 0.8721,
+      "step": 4239
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.718614387197515,
+      "learning_rate": 5.953456285231602e-06,
+      "loss": 0.8777,
+      "step": 4240
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7663012580521227,
+      "learning_rate": 5.951747102752946e-06,
+      "loss": 0.877,
+      "step": 4241
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7048474312706776,
+      "learning_rate": 5.950037804871673e-06,
+      "loss": 0.8573,
+      "step": 4242
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7539695336925303,
+      "learning_rate": 5.948328391795038e-06,
+      "loss": 0.9076,
+      "step": 4243
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7562892889825772,
+      "learning_rate": 5.9466188637303206e-06,
+      "loss": 0.9824,
+      "step": 4244
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6773661563568087,
+      "learning_rate": 5.9449092208848e-06,
+      "loss": 0.8925,
+      "step": 4245
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6752701090936328,
+      "learning_rate": 5.943199463465779e-06,
+      "loss": 0.9111,
+      "step": 4246
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7486793123120292,
+      "learning_rate": 5.9414895916805705e-06,
+      "loss": 0.9232,
+      "step": 4247
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6908071587137605,
+      "learning_rate": 5.939779605736504e-06,
+      "loss": 0.8137,
+      "step": 4248
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7473477930003563,
+      "learning_rate": 5.93806950584092e-06,
+      "loss": 0.8317,
+      "step": 4249
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6156932394585615,
+      "learning_rate": 5.936359292201175e-06,
+      "loss": 0.8703,
+      "step": 4250
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8078277218393115,
+      "learning_rate": 5.934648965024636e-06,
+      "loss": 0.8293,
+      "step": 4251
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6384893496248998,
+      "learning_rate": 5.93293852451869e-06,
+      "loss": 0.9121,
+      "step": 4252
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.575033611289935,
+      "learning_rate": 5.931227970890732e-06,
+      "loss": 0.9256,
+      "step": 4253
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7964313342505334,
+      "learning_rate": 5.92951730434817e-06,
+      "loss": 0.8793,
+      "step": 4254
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7706353809561772,
+      "learning_rate": 5.927806525098432e-06,
+      "loss": 0.8601,
+      "step": 4255
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.682682420705676,
+      "learning_rate": 5.926095633348953e-06,
+      "loss": 0.8827,
+      "step": 4256
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7227003483172998,
+      "learning_rate": 5.9243846293071846e-06,
+      "loss": 0.8989,
+      "step": 4257
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7027894467509042,
+      "learning_rate": 5.922673513180596e-06,
+      "loss": 0.9562,
+      "step": 4258
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7394861850297625,
+      "learning_rate": 5.9209622851766615e-06,
+      "loss": 0.9081,
+      "step": 4259
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.853175846105959,
+      "learning_rate": 5.919250945502875e-06,
+      "loss": 0.857,
+      "step": 4260
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9562459031184998,
+      "learning_rate": 5.9175394943667415e-06,
+      "loss": 0.8616,
+      "step": 4261
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7369450636923218,
+      "learning_rate": 5.915827931975782e-06,
+      "loss": 0.9047,
+      "step": 4262
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8864642680739013,
+      "learning_rate": 5.914116258537528e-06,
+      "loss": 1.0643,
+      "step": 4263
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8781345712906389,
+      "learning_rate": 5.912404474259525e-06,
+      "loss": 1.0344,
+      "step": 4264
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7446787164869402,
+      "learning_rate": 5.910692579349336e-06,
+      "loss": 0.8563,
+      "step": 4265
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6737713352974184,
+      "learning_rate": 5.908980574014532e-06,
+      "loss": 0.9091,
+      "step": 4266
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7718889829964044,
+      "learning_rate": 5.907268458462699e-06,
+      "loss": 0.9253,
+      "step": 4267
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7780767141320226,
+      "learning_rate": 5.905556232901437e-06,
+      "loss": 0.9019,
+      "step": 4268
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8431436385208149,
+      "learning_rate": 5.90384389753836e-06,
+      "loss": 1.034,
+      "step": 4269
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.881313864104935,
+      "learning_rate": 5.902131452581097e-06,
+      "loss": 0.8625,
+      "step": 4270
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7501680225121614,
+      "learning_rate": 5.900418898237281e-06,
+      "loss": 0.9292,
+      "step": 4271
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8245879395407776,
+      "learning_rate": 5.8987062347145726e-06,
+      "loss": 0.8894,
+      "step": 4272
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7745937977611446,
+      "learning_rate": 5.896993462220634e-06,
+      "loss": 0.8966,
+      "step": 4273
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6352873467373654,
+      "learning_rate": 5.895280580963144e-06,
+      "loss": 0.8852,
+      "step": 4274
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.751777238923234,
+      "learning_rate": 5.8935675911498e-06,
+      "loss": 0.958,
+      "step": 4275
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.848430217786073,
+      "learning_rate": 5.891854492988303e-06,
+      "loss": 0.8734,
+      "step": 4276
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6723795175432834,
+      "learning_rate": 5.890141286686375e-06,
+      "loss": 0.8855,
+      "step": 4277
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8169520772246268,
+      "learning_rate": 5.888427972451747e-06,
+      "loss": 0.8997,
+      "step": 4278
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.703262007909982,
+      "learning_rate": 5.886714550492164e-06,
+      "loss": 0.9783,
+      "step": 4279
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.734149021382422,
+      "learning_rate": 5.8850010210153844e-06,
+      "loss": 0.8702,
+      "step": 4280
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7747630420125693,
+      "learning_rate": 5.883287384229182e-06,
+      "loss": 0.9107,
+      "step": 4281
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.788716356591265,
+      "learning_rate": 5.881573640341336e-06,
+      "loss": 0.906,
+      "step": 4282
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.729814698189558,
+      "learning_rate": 5.879859789559649e-06,
+      "loss": 0.9761,
+      "step": 4283
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6354842820215787,
+      "learning_rate": 5.878145832091929e-06,
+      "loss": 0.8346,
+      "step": 4284
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8784759212292081,
+      "learning_rate": 5.876431768145998e-06,
+      "loss": 0.8792,
+      "step": 4285
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.821416325371338,
+      "learning_rate": 5.874717597929697e-06,
+      "loss": 0.9151,
+      "step": 4286
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6923328757063765,
+      "learning_rate": 5.873003321650869e-06,
+      "loss": 0.9621,
+      "step": 4287
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8835304073228,
+      "learning_rate": 5.871288939517378e-06,
+      "loss": 1.0568,
+      "step": 4288
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6401395836667843,
+      "learning_rate": 5.8695744517371e-06,
+      "loss": 0.8235,
+      "step": 4289
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.767899945410958,
+      "learning_rate": 5.8678598585179226e-06,
+      "loss": 0.8806,
+      "step": 4290
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.733540800078038,
+      "learning_rate": 5.866145160067746e-06,
+      "loss": 0.8872,
+      "step": 4291
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.725118061574199,
+      "learning_rate": 5.8644303565944805e-06,
+      "loss": 0.9158,
+      "step": 4292
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6000106266898613,
+      "learning_rate": 5.862715448306056e-06,
+      "loss": 0.8362,
+      "step": 4293
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7659261023299522,
+      "learning_rate": 5.861000435410406e-06,
+      "loss": 0.9669,
+      "step": 4294
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6543448765509794,
+      "learning_rate": 5.859285318115488e-06,
+      "loss": 0.8886,
+      "step": 4295
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6954925266652585,
+      "learning_rate": 5.857570096629263e-06,
+      "loss": 0.9006,
+      "step": 4296
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6785283965901858,
+      "learning_rate": 5.855854771159706e-06,
+      "loss": 0.9192,
+      "step": 4297
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.687206935586285,
+      "learning_rate": 5.854139341914808e-06,
+      "loss": 0.8308,
+      "step": 4298
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7646740196625315,
+      "learning_rate": 5.8524238091025684e-06,
+      "loss": 0.906,
+      "step": 4299
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8273256207386464,
+      "learning_rate": 5.850708172931005e-06,
+      "loss": 0.8837,
+      "step": 4300
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6832299428973034,
+      "learning_rate": 5.848992433608143e-06,
+      "loss": 0.9461,
+      "step": 4301
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7575096572383964,
+      "learning_rate": 5.847276591342021e-06,
+      "loss": 0.9255,
+      "step": 4302
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.751066348809859,
+      "learning_rate": 5.845560646340691e-06,
+      "loss": 0.9738,
+      "step": 4303
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7217843812268712,
+      "learning_rate": 5.843844598812218e-06,
+      "loss": 0.8337,
+      "step": 4304
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7914325915333456,
+      "learning_rate": 5.842128448964677e-06,
+      "loss": 0.9462,
+      "step": 4305
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7017776142339078,
+      "learning_rate": 5.84041219700616e-06,
+      "loss": 0.9586,
+      "step": 4306
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7140065350238547,
+      "learning_rate": 5.838695843144766e-06,
+      "loss": 0.9231,
+      "step": 4307
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7225392057018107,
+      "learning_rate": 5.836979387588611e-06,
+      "loss": 0.9659,
+      "step": 4308
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8647052960340152,
+      "learning_rate": 5.835262830545817e-06,
+      "loss": 1.0681,
+      "step": 4309
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8192044238230678,
+      "learning_rate": 5.833546172224527e-06,
+      "loss": 0.9821,
+      "step": 4310
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.67182539202832,
+      "learning_rate": 5.8318294128328885e-06,
+      "loss": 0.9422,
+      "step": 4311
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.720461234822713,
+      "learning_rate": 5.830112552579068e-06,
+      "loss": 0.8885,
+      "step": 4312
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8347582402011187,
+      "learning_rate": 5.828395591671238e-06,
+      "loss": 0.9903,
+      "step": 4313
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6966756007422459,
+      "learning_rate": 5.826678530317585e-06,
+      "loss": 0.9046,
+      "step": 4314
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5648275193457388,
+      "learning_rate": 5.824961368726312e-06,
+      "loss": 0.9417,
+      "step": 4315
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7062030958926075,
+      "learning_rate": 5.823244107105628e-06,
+      "loss": 0.8618,
+      "step": 4316
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.933248628902336,
+      "learning_rate": 5.821526745663758e-06,
+      "loss": 0.9097,
+      "step": 4317
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7658419677548072,
+      "learning_rate": 5.8198092846089396e-06,
+      "loss": 0.863,
+      "step": 4318
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7273904409940526,
+      "learning_rate": 5.818091724149417e-06,
+      "loss": 0.9419,
+      "step": 4319
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6376488801928495,
+      "learning_rate": 5.816374064493453e-06,
+      "loss": 0.881,
+      "step": 4320
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7075981065594557,
+      "learning_rate": 5.81465630584932e-06,
+      "loss": 0.964,
+      "step": 4321
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8815099912296813,
+      "learning_rate": 5.812938448425299e-06,
+      "loss": 0.9162,
+      "step": 4322
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6225405223737397,
+      "learning_rate": 5.811220492429692e-06,
+      "loss": 0.9306,
+      "step": 4323
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6361454733396816,
+      "learning_rate": 5.809502438070801e-06,
+      "loss": 0.7772,
+      "step": 4324
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7562619549347187,
+      "learning_rate": 5.80778428555695e-06,
+      "loss": 0.9236,
+      "step": 4325
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7720787192198995,
+      "learning_rate": 5.8060660350964685e-06,
+      "loss": 0.9124,
+      "step": 4326
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6743200460432013,
+      "learning_rate": 5.8043476868977025e-06,
+      "loss": 0.8344,
+      "step": 4327
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8811800336967499,
+      "learning_rate": 5.802629241169006e-06,
+      "loss": 1.0546,
+      "step": 4328
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.690157621779071,
+      "learning_rate": 5.800910698118746e-06,
+      "loss": 0.9243,
+      "step": 4329
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7268529082398683,
+      "learning_rate": 5.799192057955303e-06,
+      "loss": 0.8579,
+      "step": 4330
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7267972393192101,
+      "learning_rate": 5.7974733208870686e-06,
+      "loss": 0.9086,
+      "step": 4331
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7468968471027715,
+      "learning_rate": 5.795754487122444e-06,
+      "loss": 0.9707,
+      "step": 4332
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6661456279089482,
+      "learning_rate": 5.794035556869843e-06,
+      "loss": 0.9082,
+      "step": 4333
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7209165527445007,
+      "learning_rate": 5.792316530337696e-06,
+      "loss": 0.9397,
+      "step": 4334
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7664896586594365,
+      "learning_rate": 5.790597407734437e-06,
+      "loss": 0.9268,
+      "step": 4335
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7738362724752272,
+      "learning_rate": 5.788878189268516e-06,
+      "loss": 0.8993,
+      "step": 4336
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6874772805153853,
+      "learning_rate": 5.787158875148396e-06,
+      "loss": 0.8693,
+      "step": 4337
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9704488355707166,
+      "learning_rate": 5.785439465582549e-06,
+      "loss": 1.0663,
+      "step": 4338
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7014149290465737,
+      "learning_rate": 5.783719960779458e-06,
+      "loss": 0.8799,
+      "step": 4339
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7803285993184608,
+      "learning_rate": 5.78200036094762e-06,
+      "loss": 0.8408,
+      "step": 4340
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6676713181013065,
+      "learning_rate": 5.780280666295544e-06,
+      "loss": 0.8609,
+      "step": 4341
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6392816024672245,
+      "learning_rate": 5.778560877031744e-06,
+      "loss": 0.8944,
+      "step": 4342
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6731555632694153,
+      "learning_rate": 5.7768409933647555e-06,
+      "loss": 0.8302,
+      "step": 4343
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6890219271176474,
+      "learning_rate": 5.775121015503118e-06,
+      "loss": 0.8722,
+      "step": 4344
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.701784821704374,
+      "learning_rate": 5.773400943655385e-06,
+      "loss": 1.0016,
+      "step": 4345
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6413016385059154,
+      "learning_rate": 5.771680778030122e-06,
+      "loss": 0.9035,
+      "step": 4346
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7642405622325543,
+      "learning_rate": 5.769960518835902e-06,
+      "loss": 0.99,
+      "step": 4347
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.855919933081425,
+      "learning_rate": 5.768240166281317e-06,
+      "loss": 0.9586,
+      "step": 4348
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.758492355538492,
+      "learning_rate": 5.7665197205749636e-06,
+      "loss": 0.9111,
+      "step": 4349
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8175318259710618,
+      "learning_rate": 5.764799181925449e-06,
+      "loss": 0.866,
+      "step": 4350
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.721235413612071,
+      "learning_rate": 5.763078550541399e-06,
+      "loss": 0.9331,
+      "step": 4351
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8006451985343097,
+      "learning_rate": 5.761357826631444e-06,
+      "loss": 0.895,
+      "step": 4352
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7103740870150868,
+      "learning_rate": 5.759637010404226e-06,
+      "loss": 1.0018,
+      "step": 4353
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5894063217712353,
+      "learning_rate": 5.757916102068403e-06,
+      "loss": 0.8712,
+      "step": 4354
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.023462572910606,
+      "learning_rate": 5.75619510183264e-06,
+      "loss": 0.8934,
+      "step": 4355
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6190923394457406,
+      "learning_rate": 5.754474009905613e-06,
+      "loss": 0.9186,
+      "step": 4356
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.809752420592461,
+      "learning_rate": 5.752752826496012e-06,
+      "loss": 0.8934,
+      "step": 4357
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7547080230966816,
+      "learning_rate": 5.751031551812536e-06,
+      "loss": 0.9264,
+      "step": 4358
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.848610475092547,
+      "learning_rate": 5.749310186063892e-06,
+      "loss": 0.9773,
+      "step": 4359
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8259705643302377,
+      "learning_rate": 5.74758872945881e-06,
+      "loss": 0.8971,
+      "step": 4360
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7016342958289672,
+      "learning_rate": 5.745867182206013e-06,
+      "loss": 0.9768,
+      "step": 4361
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.743427085760458,
+      "learning_rate": 5.7441455445142505e-06,
+      "loss": 0.8538,
+      "step": 4362
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8119013095299743,
+      "learning_rate": 5.7424238165922745e-06,
+      "loss": 0.9652,
+      "step": 4363
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7107975318307735,
+      "learning_rate": 5.74070199864885e-06,
+      "loss": 0.9011,
+      "step": 4364
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6979437689962997,
+      "learning_rate": 5.7389800908927575e-06,
+      "loss": 0.8717,
+      "step": 4365
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8613498956215577,
+      "learning_rate": 5.737258093532781e-06,
+      "loss": 0.961,
+      "step": 4366
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7767167747869528,
+      "learning_rate": 5.735536006777717e-06,
+      "loss": 0.9113,
+      "step": 4367
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8073277804076218,
+      "learning_rate": 5.733813830836377e-06,
+      "loss": 0.9033,
+      "step": 4368
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6884017835066574,
+      "learning_rate": 5.73209156591758e-06,
+      "loss": 0.8494,
+      "step": 4369
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.091243913649683,
+      "learning_rate": 5.7303692122301566e-06,
+      "loss": 0.8579,
+      "step": 4370
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9874037224395742,
+      "learning_rate": 5.728646769982951e-06,
+      "loss": 0.9752,
+      "step": 4371
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.89495750280681,
+      "learning_rate": 5.726924239384809e-06,
+      "loss": 1.0992,
+      "step": 4372
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6973270287082058,
+      "learning_rate": 5.725201620644598e-06,
+      "loss": 0.9101,
+      "step": 4373
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8500862858025551,
+      "learning_rate": 5.7234789139711914e-06,
+      "loss": 1.0827,
+      "step": 4374
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.701599527327432,
+      "learning_rate": 5.721756119573471e-06,
+      "loss": 0.9428,
+      "step": 4375
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.022255303985465,
+      "learning_rate": 5.720033237660334e-06,
+      "loss": 0.8931,
+      "step": 4376
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.808461280304135,
+      "learning_rate": 5.718310268440684e-06,
+      "loss": 0.9132,
+      "step": 4377
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7909915563865106,
+      "learning_rate": 5.716587212123436e-06,
+      "loss": 0.8678,
+      "step": 4378
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7272557967321531,
+      "learning_rate": 5.7148640689175194e-06,
+      "loss": 0.8893,
+      "step": 4379
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8429627517791256,
+      "learning_rate": 5.713140839031871e-06,
+      "loss": 0.8227,
+      "step": 4380
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7817679701255957,
+      "learning_rate": 5.711417522675437e-06,
+      "loss": 0.9816,
+      "step": 4381
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.718625975627871,
+      "learning_rate": 5.709694120057174e-06,
+      "loss": 0.8872,
+      "step": 4382
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.726136968685944,
+      "learning_rate": 5.707970631386054e-06,
+      "loss": 0.9388,
+      "step": 4383
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.744502240291231,
+      "learning_rate": 5.706247056871052e-06,
+      "loss": 0.8295,
+      "step": 4384
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.740402938248681,
+      "learning_rate": 5.704523396721162e-06,
+      "loss": 0.9304,
+      "step": 4385
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7095260944011252,
+      "learning_rate": 5.702799651145381e-06,
+      "loss": 0.8848,
+      "step": 4386
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7219652715732094,
+      "learning_rate": 5.701075820352718e-06,
+      "loss": 0.9658,
+      "step": 4387
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6797244914817266,
+      "learning_rate": 5.699351904552196e-06,
+      "loss": 0.8687,
+      "step": 4388
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6895829540513125,
+      "learning_rate": 5.697627903952845e-06,
+      "loss": 0.8195,
+      "step": 4389
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7694739505913462,
+      "learning_rate": 5.695903818763703e-06,
+      "loss": 0.8834,
+      "step": 4390
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6748294607599246,
+      "learning_rate": 5.694179649193826e-06,
+      "loss": 0.8251,
+      "step": 4391
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.634224447186001,
+      "learning_rate": 5.692455395452272e-06,
+      "loss": 0.9045,
+      "step": 4392
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.841948887860489,
+      "learning_rate": 5.6907310577481155e-06,
+      "loss": 0.8925,
+      "step": 4393
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.804045643199622,
+      "learning_rate": 5.6890066362904374e-06,
+      "loss": 0.8938,
+      "step": 4394
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8375508023946332,
+      "learning_rate": 5.687282131288326e-06,
+      "loss": 0.9708,
+      "step": 4395
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.758431757504343,
+      "learning_rate": 5.68555754295089e-06,
+      "loss": 0.8539,
+      "step": 4396
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.801633553963666,
+      "learning_rate": 5.683832871487238e-06,
+      "loss": 0.8795,
+      "step": 4397
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6897793223317072,
+      "learning_rate": 5.682108117106491e-06,
+      "loss": 0.8703,
+      "step": 4398
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7173681280307052,
+      "learning_rate": 5.680383280017785e-06,
+      "loss": 0.9006,
+      "step": 4399
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9233956985962328,
+      "learning_rate": 5.678658360430261e-06,
+      "loss": 1.0783,
+      "step": 4400
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8057696683195357,
+      "learning_rate": 5.676933358553068e-06,
+      "loss": 0.9528,
+      "step": 4401
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7694931840829837,
+      "learning_rate": 5.6752082745953765e-06,
+      "loss": 0.8859,
+      "step": 4402
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7382345113535036,
+      "learning_rate": 5.673483108766349e-06,
+      "loss": 0.9795,
+      "step": 4403
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8703638602604409,
+      "learning_rate": 5.671757861275176e-06,
+      "loss": 0.9482,
+      "step": 4404
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6826634622998033,
+      "learning_rate": 5.670032532331045e-06,
+      "loss": 0.8856,
+      "step": 4405
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7224051876415518,
+      "learning_rate": 5.668307122143159e-06,
+      "loss": 0.9104,
+      "step": 4406
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.78867284449983,
+      "learning_rate": 5.666581630920731e-06,
+      "loss": 0.8498,
+      "step": 4407
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8895480265320632,
+      "learning_rate": 5.664856058872984e-06,
+      "loss": 1.1357,
+      "step": 4408
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8126786374096393,
+      "learning_rate": 5.6631304062091456e-06,
+      "loss": 0.9882,
+      "step": 4409
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6491924368194437,
+      "learning_rate": 5.661404673138459e-06,
+      "loss": 0.9223,
+      "step": 4410
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.69417106399307,
+      "learning_rate": 5.659678859870177e-06,
+      "loss": 0.9057,
+      "step": 4411
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.752861565874866,
+      "learning_rate": 5.657952966613558e-06,
+      "loss": 0.9491,
+      "step": 4412
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8398795317246768,
+      "learning_rate": 5.656226993577875e-06,
+      "loss": 1.0682,
+      "step": 4413
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7757416164905326,
+      "learning_rate": 5.654500940972405e-06,
+      "loss": 0.9402,
+      "step": 4414
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.692296412796347,
+      "learning_rate": 5.652774809006439e-06,
+      "loss": 0.9632,
+      "step": 4415
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8158247882279208,
+      "learning_rate": 5.6510485978892785e-06,
+      "loss": 0.9612,
+      "step": 4416
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.79296525299831,
+      "learning_rate": 5.64932230783023e-06,
+      "loss": 0.8147,
+      "step": 4417
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6683349046646079,
+      "learning_rate": 5.647595939038615e-06,
+      "loss": 0.8932,
+      "step": 4418
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7532381813383346,
+      "learning_rate": 5.6458694917237575e-06,
+      "loss": 0.8804,
+      "step": 4419
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.875324981687024,
+      "learning_rate": 5.644142966094997e-06,
+      "loss": 0.9121,
+      "step": 4420
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7250095444930669,
+      "learning_rate": 5.642416362361683e-06,
+      "loss": 0.8858,
+      "step": 4421
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7477866025864333,
+      "learning_rate": 5.640689680733171e-06,
+      "loss": 0.8414,
+      "step": 4422
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8346747913763409,
+      "learning_rate": 5.638962921418825e-06,
+      "loss": 1.0483,
+      "step": 4423
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7822262298579257,
+      "learning_rate": 5.6372360846280224e-06,
+      "loss": 0.9589,
+      "step": 4424
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8399044312559669,
+      "learning_rate": 5.635509170570148e-06,
+      "loss": 1.037,
+      "step": 4425
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8897345073343004,
+      "learning_rate": 5.633782179454594e-06,
+      "loss": 0.8922,
+      "step": 4426
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.787888350903323,
+      "learning_rate": 5.632055111490768e-06,
+      "loss": 0.9482,
+      "step": 4427
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8152861013273506,
+      "learning_rate": 5.6303279668880794e-06,
+      "loss": 1.0713,
+      "step": 4428
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6874950696346482,
+      "learning_rate": 5.6286007458559525e-06,
+      "loss": 0.9487,
+      "step": 4429
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9156831539560288,
+      "learning_rate": 5.6268734486038164e-06,
+      "loss": 0.8808,
+      "step": 4430
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7602698730568485,
+      "learning_rate": 5.6251460753411155e-06,
+      "loss": 0.8671,
+      "step": 4431
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6472782155053773,
+      "learning_rate": 5.623418626277296e-06,
+      "loss": 0.8815,
+      "step": 4432
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7150741769902929,
+      "learning_rate": 5.62169110162182e-06,
+      "loss": 0.9352,
+      "step": 4433
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8308043921206636,
+      "learning_rate": 5.619963501584155e-06,
+      "loss": 0.8969,
+      "step": 4434
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7231722052151708,
+      "learning_rate": 5.618235826373777e-06,
+      "loss": 0.9038,
+      "step": 4435
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6684172732091316,
+      "learning_rate": 5.616508076200175e-06,
+      "loss": 0.7979,
+      "step": 4436
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7005244987061152,
+      "learning_rate": 5.6147802512728425e-06,
+      "loss": 0.9377,
+      "step": 4437
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8403292853601343,
+      "learning_rate": 5.613052351801284e-06,
+      "loss": 1.0668,
+      "step": 4438
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.726947377021382,
+      "learning_rate": 5.611324377995017e-06,
+      "loss": 0.9467,
+      "step": 4439
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8087873443779541,
+      "learning_rate": 5.6095963300635585e-06,
+      "loss": 1.0241,
+      "step": 4440
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7533901335270679,
+      "learning_rate": 5.607868208216445e-06,
+      "loss": 0.9842,
+      "step": 4441
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7716509060196968,
+      "learning_rate": 5.606140012663215e-06,
+      "loss": 0.8546,
+      "step": 4442
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7499975141440463,
+      "learning_rate": 5.604411743613418e-06,
+      "loss": 0.96,
+      "step": 4443
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7689560460983862,
+      "learning_rate": 5.6026834012766155e-06,
+      "loss": 0.8245,
+      "step": 4444
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.740357221417932,
+      "learning_rate": 5.6009549858623735e-06,
+      "loss": 0.8817,
+      "step": 4445
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.848031965802933,
+      "learning_rate": 5.5992264975802646e-06,
+      "loss": 0.9813,
+      "step": 4446
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8391077751190138,
+      "learning_rate": 5.597497936639879e-06,
+      "loss": 1.0452,
+      "step": 4447
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8202903954942881,
+      "learning_rate": 5.595769303250809e-06,
+      "loss": 1.0657,
+      "step": 4448
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7870355975503178,
+      "learning_rate": 5.594040597622654e-06,
+      "loss": 0.8703,
+      "step": 4449
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7269904186260898,
+      "learning_rate": 5.592311819965033e-06,
+      "loss": 0.942,
+      "step": 4450
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8432462187731478,
+      "learning_rate": 5.590582970487558e-06,
+      "loss": 1.0937,
+      "step": 4451
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8230030569635873,
+      "learning_rate": 5.588854049399864e-06,
+      "loss": 0.9654,
+      "step": 4452
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6979856613558493,
+      "learning_rate": 5.587125056911587e-06,
+      "loss": 0.9216,
+      "step": 4453
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8310370532320412,
+      "learning_rate": 5.585395993232371e-06,
+      "loss": 0.9739,
+      "step": 4454
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7245828142409558,
+      "learning_rate": 5.583666858571874e-06,
+      "loss": 0.8603,
+      "step": 4455
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.68867402529456,
+      "learning_rate": 5.581937653139757e-06,
+      "loss": 0.9371,
+      "step": 4456
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7943148084305938,
+      "learning_rate": 5.580208377145693e-06,
+      "loss": 0.8823,
+      "step": 4457
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.70423497123324,
+      "learning_rate": 5.578479030799363e-06,
+      "loss": 0.9329,
+      "step": 4458
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.759876556134152,
+      "learning_rate": 5.576749614310457e-06,
+      "loss": 0.8506,
+      "step": 4459
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7584029333198166,
+      "learning_rate": 5.575020127888672e-06,
+      "loss": 0.9239,
+      "step": 4460
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.754783794346986,
+      "learning_rate": 5.573290571743715e-06,
+      "loss": 0.9784,
+      "step": 4461
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8732674872020809,
+      "learning_rate": 5.5715609460852985e-06,
+      "loss": 1.0161,
+      "step": 4462
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8346460293774912,
+      "learning_rate": 5.5698312511231454e-06,
+      "loss": 1.0411,
+      "step": 4463
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7025912552605986,
+      "learning_rate": 5.568101487066991e-06,
+      "loss": 0.8691,
+      "step": 4464
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7238269248882718,
+      "learning_rate": 5.566371654126573e-06,
+      "loss": 0.8562,
+      "step": 4465
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.674195755641617,
+      "learning_rate": 5.564641752511637e-06,
+      "loss": 0.9839,
+      "step": 4466
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8666056456666086,
+      "learning_rate": 5.562911782431943e-06,
+      "loss": 1.0398,
+      "step": 4467
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7062201295774624,
+      "learning_rate": 5.561181744097255e-06,
+      "loss": 0.895,
+      "step": 4468
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.640549685719731,
+      "learning_rate": 5.559451637717346e-06,
+      "loss": 0.9691,
+      "step": 4469
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8607872043163429,
+      "learning_rate": 5.557721463501997e-06,
+      "loss": 1.0335,
+      "step": 4470
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6426565399455073,
+      "learning_rate": 5.555991221660998e-06,
+      "loss": 0.8849,
+      "step": 4471
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8158690942758193,
+      "learning_rate": 5.554260912404146e-06,
+      "loss": 1.0659,
+      "step": 4472
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6770186589003653,
+      "learning_rate": 5.5525305359412485e-06,
+      "loss": 0.9103,
+      "step": 4473
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8198258533851601,
+      "learning_rate": 5.550800092482117e-06,
+      "loss": 0.9469,
+      "step": 4474
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.875535777489499,
+      "learning_rate": 5.549069582236577e-06,
+      "loss": 0.8955,
+      "step": 4475
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6917452485260298,
+      "learning_rate": 5.547339005414457e-06,
+      "loss": 0.8596,
+      "step": 4476
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7767330298346917,
+      "learning_rate": 5.545608362225594e-06,
+      "loss": 0.8949,
+      "step": 4477
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6559786149165756,
+      "learning_rate": 5.543877652879838e-06,
+      "loss": 0.9589,
+      "step": 4478
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6823744541613697,
+      "learning_rate": 5.542146877587042e-06,
+      "loss": 0.8964,
+      "step": 4479
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9399703813497159,
+      "learning_rate": 5.540416036557064e-06,
+      "loss": 1.0678,
+      "step": 4480
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8958916218514894,
+      "learning_rate": 5.538685129999782e-06,
+      "loss": 0.9084,
+      "step": 4481
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.835581180906324,
+      "learning_rate": 5.536954158125071e-06,
+      "loss": 0.8757,
+      "step": 4482
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7588361818462706,
+      "learning_rate": 5.535223121142817e-06,
+      "loss": 0.8766,
+      "step": 4483
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.693264456882381,
+      "learning_rate": 5.5334920192629135e-06,
+      "loss": 0.8754,
+      "step": 4484
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8138872771006682,
+      "learning_rate": 5.531760852695264e-06,
+      "loss": 0.9052,
+      "step": 4485
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7071780273815909,
+      "learning_rate": 5.530029621649777e-06,
+      "loss": 0.8715,
+      "step": 4486
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7696725679921885,
+      "learning_rate": 5.528298326336372e-06,
+      "loss": 0.9761,
+      "step": 4487
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.77021995286334,
+      "learning_rate": 5.5265669669649725e-06,
+      "loss": 0.9308,
+      "step": 4488
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8538213495987061,
+      "learning_rate": 5.524835543745515e-06,
+      "loss": 0.9196,
+      "step": 4489
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6822729512115473,
+      "learning_rate": 5.523104056887937e-06,
+      "loss": 0.8081,
+      "step": 4490
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7688068196747344,
+      "learning_rate": 5.521372506602187e-06,
+      "loss": 0.8839,
+      "step": 4491
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8136981048524516,
+      "learning_rate": 5.519640893098227e-06,
+      "loss": 0.9514,
+      "step": 4492
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7286030736407574,
+      "learning_rate": 5.517909216586015e-06,
+      "loss": 0.9201,
+      "step": 4493
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7697904809857206,
+      "learning_rate": 5.516177477275524e-06,
+      "loss": 0.8694,
+      "step": 4494
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8336896661387727,
+      "learning_rate": 5.514445675376736e-06,
+      "loss": 0.884,
+      "step": 4495
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7412022300923362,
+      "learning_rate": 5.512713811099636e-06,
+      "loss": 0.8835,
+      "step": 4496
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7070109001882205,
+      "learning_rate": 5.510981884654217e-06,
+      "loss": 0.9104,
+      "step": 4497
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5698683202437849,
+      "learning_rate": 5.509249896250486e-06,
+      "loss": 0.7937,
+      "step": 4498
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6364746365179967,
+      "learning_rate": 5.507517846098447e-06,
+      "loss": 0.9448,
+      "step": 4499
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7903048684370588,
+      "learning_rate": 5.505785734408121e-06,
+      "loss": 0.9341,
+      "step": 4500
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6832840485687173,
+      "learning_rate": 5.504053561389531e-06,
+      "loss": 0.852,
+      "step": 4501
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9081902846469121,
+      "learning_rate": 5.502321327252708e-06,
+      "loss": 1.0367,
+      "step": 4502
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7330104159153432,
+      "learning_rate": 5.500589032207693e-06,
+      "loss": 0.9159,
+      "step": 4503
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5751818060098317,
+      "learning_rate": 5.4988566764645315e-06,
+      "loss": 0.9309,
+      "step": 4504
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7095080923292634,
+      "learning_rate": 5.497124260233278e-06,
+      "loss": 0.9391,
+      "step": 4505
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8726116025603265,
+      "learning_rate": 5.495391783723993e-06,
+      "loss": 0.9125,
+      "step": 4506
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7277559920204868,
+      "learning_rate": 5.493659247146749e-06,
+      "loss": 1.0202,
+      "step": 4507
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8453142056947867,
+      "learning_rate": 5.491926650711618e-06,
+      "loss": 0.9223,
+      "step": 4508
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7134062460809283,
+      "learning_rate": 5.490193994628685e-06,
+      "loss": 0.9147,
+      "step": 4509
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6763299486919163,
+      "learning_rate": 5.488461279108041e-06,
+      "loss": 0.931,
+      "step": 4510
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6719689144519583,
+      "learning_rate": 5.486728504359781e-06,
+      "loss": 0.7988,
+      "step": 4511
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8806268342551522,
+      "learning_rate": 5.484995670594012e-06,
+      "loss": 1.0604,
+      "step": 4512
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8528297403604825,
+      "learning_rate": 5.483262778020847e-06,
+      "loss": 0.9903,
+      "step": 4513
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7233122080116507,
+      "learning_rate": 5.481529826850404e-06,
+      "loss": 0.9204,
+      "step": 4514
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8724917390101656,
+      "learning_rate": 5.479796817292809e-06,
+      "loss": 0.9492,
+      "step": 4515
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7332733474100304,
+      "learning_rate": 5.478063749558195e-06,
+      "loss": 0.9193,
+      "step": 4516
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5131083453592364,
+      "learning_rate": 5.476330623856705e-06,
+      "loss": 0.8557,
+      "step": 4517
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8462212356994248,
+      "learning_rate": 5.474597440398483e-06,
+      "loss": 0.971,
+      "step": 4518
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6641438366282582,
+      "learning_rate": 5.472864199393687e-06,
+      "loss": 0.9228,
+      "step": 4519
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.61193354212265,
+      "learning_rate": 5.4711309010524764e-06,
+      "loss": 0.8517,
+      "step": 4520
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5819466838445126,
+      "learning_rate": 5.469397545585019e-06,
+      "loss": 0.9544,
+      "step": 4521
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7745831685862068,
+      "learning_rate": 5.46766413320149e-06,
+      "loss": 0.9264,
+      "step": 4522
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7479392398203222,
+      "learning_rate": 5.465930664112073e-06,
+      "loss": 0.947,
+      "step": 4523
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7699435927339686,
+      "learning_rate": 5.46419713852696e-06,
+      "loss": 0.9375,
+      "step": 4524
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.697043144119612,
+      "learning_rate": 5.46246355665634e-06,
+      "loss": 0.8739,
+      "step": 4525
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6160032936047493,
+      "learning_rate": 5.4607299187104205e-06,
+      "loss": 0.9482,
+      "step": 4526
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8449851611342056,
+      "learning_rate": 5.458996224899409e-06,
+      "loss": 0.9069,
+      "step": 4527
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.791852077255882,
+      "learning_rate": 5.457262475433523e-06,
+      "loss": 1.0281,
+      "step": 4528
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.709995209549197,
+      "learning_rate": 5.455528670522987e-06,
+      "loss": 0.9235,
+      "step": 4529
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.791368373247366,
+      "learning_rate": 5.453794810378028e-06,
+      "loss": 0.9479,
+      "step": 4530
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7743050131258544,
+      "learning_rate": 5.452060895208883e-06,
+      "loss": 0.9537,
+      "step": 4531
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.698317462349543,
+      "learning_rate": 5.450326925225798e-06,
+      "loss": 0.9319,
+      "step": 4532
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8034785590687042,
+      "learning_rate": 5.44859290063902e-06,
+      "loss": 1.0419,
+      "step": 4533
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6673383191199331,
+      "learning_rate": 5.446858821658806e-06,
+      "loss": 0.8256,
+      "step": 4534
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8114801495635167,
+      "learning_rate": 5.44512468849542e-06,
+      "loss": 1.0723,
+      "step": 4535
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7819432992158732,
+      "learning_rate": 5.44339050135913e-06,
+      "loss": 0.9138,
+      "step": 4536
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7496730093423345,
+      "learning_rate": 5.441656260460213e-06,
+      "loss": 0.9306,
+      "step": 4537
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.897373896804874,
+      "learning_rate": 5.439921966008953e-06,
+      "loss": 0.9549,
+      "step": 4538
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8681262733048793,
+      "learning_rate": 5.438187618215636e-06,
+      "loss": 0.8661,
+      "step": 4539
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.721289786065305,
+      "learning_rate": 5.436453217290562e-06,
+      "loss": 0.8695,
+      "step": 4540
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6557071408330994,
+      "learning_rate": 5.43471876344403e-06,
+      "loss": 0.9017,
+      "step": 4541
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6925655724312247,
+      "learning_rate": 5.432984256886348e-06,
+      "loss": 0.9014,
+      "step": 4542
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7838152948451536,
+      "learning_rate": 5.431249697827833e-06,
+      "loss": 0.8776,
+      "step": 4543
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6716142034827348,
+      "learning_rate": 5.429515086478805e-06,
+      "loss": 0.9445,
+      "step": 4544
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.629113029553762,
+      "learning_rate": 5.4277804230495935e-06,
+      "loss": 0.8494,
+      "step": 4545
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7549392390176415,
+      "learning_rate": 5.426045707750529e-06,
+      "loss": 0.9589,
+      "step": 4546
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7831024032422127,
+      "learning_rate": 5.424310940791954e-06,
+      "loss": 0.93,
+      "step": 4547
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7903773691550644,
+      "learning_rate": 5.422576122384216e-06,
+      "loss": 0.8886,
+      "step": 4548
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7453412087419156,
+      "learning_rate": 5.420841252737664e-06,
+      "loss": 0.8657,
+      "step": 4549
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6574610935402623,
+      "learning_rate": 5.419106332062661e-06,
+      "loss": 0.9364,
+      "step": 4550
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7551367727458511,
+      "learning_rate": 5.41737136056957e-06,
+      "loss": 0.9112,
+      "step": 4551
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7209471093287045,
+      "learning_rate": 5.415636338468763e-06,
+      "loss": 0.9443,
+      "step": 4552
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.808813854554889,
+      "learning_rate": 5.413901265970616e-06,
+      "loss": 0.9397,
+      "step": 4553
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7447499149616428,
+      "learning_rate": 5.412166143285515e-06,
+      "loss": 0.8807,
+      "step": 4554
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7024341385040747,
+      "learning_rate": 5.410430970623847e-06,
+      "loss": 0.9243,
+      "step": 4555
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8156177987813094,
+      "learning_rate": 5.408695748196009e-06,
+      "loss": 0.9272,
+      "step": 4556
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7028552537428028,
+      "learning_rate": 5.406960476212403e-06,
+      "loss": 0.8886,
+      "step": 4557
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6560708558979378,
+      "learning_rate": 5.405225154883436e-06,
+      "loss": 0.902,
+      "step": 4558
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.782467321274546,
+      "learning_rate": 5.403489784419521e-06,
+      "loss": 0.8634,
+      "step": 4559
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7021959633169952,
+      "learning_rate": 5.40175436503108e-06,
+      "loss": 0.8631,
+      "step": 4560
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8941363983000871,
+      "learning_rate": 5.400018896928537e-06,
+      "loss": 1.0344,
+      "step": 4561
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7703053327094236,
+      "learning_rate": 5.398283380322323e-06,
+      "loss": 0.8696,
+      "step": 4562
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8130593245448887,
+      "learning_rate": 5.396547815422876e-06,
+      "loss": 0.9531,
+      "step": 4563
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6952485660615186,
+      "learning_rate": 5.3948122024406405e-06,
+      "loss": 0.93,
+      "step": 4564
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.585048755078583,
+      "learning_rate": 5.393076541586063e-06,
+      "loss": 0.9373,
+      "step": 4565
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6633059819815623,
+      "learning_rate": 5.391340833069601e-06,
+      "loss": 0.8809,
+      "step": 4566
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6165995737699843,
+      "learning_rate": 5.389605077101713e-06,
+      "loss": 1.0072,
+      "step": 4567
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7032233371267007,
+      "learning_rate": 5.387869273892866e-06,
+      "loss": 0.9313,
+      "step": 4568
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7533925224009796,
+      "learning_rate": 5.386133423653532e-06,
+      "loss": 0.9002,
+      "step": 4569
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6612093620113988,
+      "learning_rate": 5.3843975265941896e-06,
+      "loss": 0.9982,
+      "step": 4570
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6652564903991782,
+      "learning_rate": 5.382661582925322e-06,
+      "loss": 0.9214,
+      "step": 4571
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7069809071106985,
+      "learning_rate": 5.38092559285742e-06,
+      "loss": 0.9141,
+      "step": 4572
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7712882949730695,
+      "learning_rate": 5.379189556600974e-06,
+      "loss": 0.9097,
+      "step": 4573
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.748173324651378,
+      "learning_rate": 5.3774534743664885e-06,
+      "loss": 0.8223,
+      "step": 4574
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7099773912089573,
+      "learning_rate": 5.375717346364467e-06,
+      "loss": 0.9251,
+      "step": 4575
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7386085556116726,
+      "learning_rate": 5.373981172805421e-06,
+      "loss": 0.9596,
+      "step": 4576
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8256980004521808,
+      "learning_rate": 5.3722449538998725e-06,
+      "loss": 0.8583,
+      "step": 4577
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7020209597649143,
+      "learning_rate": 5.370508689858336e-06,
+      "loss": 0.8723,
+      "step": 4578
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7029699950435835,
+      "learning_rate": 5.368772380891345e-06,
+      "loss": 0.9415,
+      "step": 4579
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7523055066111035,
+      "learning_rate": 5.367036027209431e-06,
+      "loss": 0.9341,
+      "step": 4580
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8471106687703471,
+      "learning_rate": 5.365299629023134e-06,
+      "loss": 0.9037,
+      "step": 4581
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.727985401031493,
+      "learning_rate": 5.363563186542997e-06,
+      "loss": 0.7947,
+      "step": 4582
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8157869200489005,
+      "learning_rate": 5.361826699979571e-06,
+      "loss": 0.9399,
+      "step": 4583
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7338578116199364,
+      "learning_rate": 5.36009016954341e-06,
+      "loss": 0.8736,
+      "step": 4584
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7494954011122297,
+      "learning_rate": 5.358353595445074e-06,
+      "loss": 0.9028,
+      "step": 4585
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6984709997572656,
+      "learning_rate": 5.35661697789513e-06,
+      "loss": 0.929,
+      "step": 4586
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7399188352381914,
+      "learning_rate": 5.354880317104145e-06,
+      "loss": 0.9626,
+      "step": 4587
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6890655330285105,
+      "learning_rate": 5.353143613282702e-06,
+      "loss": 0.8139,
+      "step": 4588
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.853510247665804,
+      "learning_rate": 5.351406866641378e-06,
+      "loss": 0.8917,
+      "step": 4589
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8575042507183054,
+      "learning_rate": 5.349670077390757e-06,
+      "loss": 1.0534,
+      "step": 4590
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.73827405305754,
+      "learning_rate": 5.347933245741435e-06,
+      "loss": 0.9241,
+      "step": 4591
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8380803560071384,
+      "learning_rate": 5.346196371904009e-06,
+      "loss": 0.9246,
+      "step": 4592
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8254939727833401,
+      "learning_rate": 5.344459456089078e-06,
+      "loss": 0.9106,
+      "step": 4593
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6816637952038043,
+      "learning_rate": 5.342722498507251e-06,
+      "loss": 0.8356,
+      "step": 4594
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.783708588022164,
+      "learning_rate": 5.340985499369138e-06,
+      "loss": 0.984,
+      "step": 4595
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.781469591974205,
+      "learning_rate": 5.33924845888536e-06,
+      "loss": 0.9861,
+      "step": 4596
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6732049514752332,
+      "learning_rate": 5.337511377266535e-06,
+      "loss": 0.885,
+      "step": 4597
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8156133969968213,
+      "learning_rate": 5.335774254723293e-06,
+      "loss": 0.8519,
+      "step": 4598
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8270018084042678,
+      "learning_rate": 5.334037091466265e-06,
+      "loss": 0.9226,
+      "step": 4599
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6076775701121917,
+      "learning_rate": 5.332299887706087e-06,
+      "loss": 0.9099,
+      "step": 4600
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8093160941863644,
+      "learning_rate": 5.330562643653402e-06,
+      "loss": 0.9659,
+      "step": 4601
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7685595039435356,
+      "learning_rate": 5.3288253595188575e-06,
+      "loss": 0.9433,
+      "step": 4602
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7669769730839444,
+      "learning_rate": 5.327088035513105e-06,
+      "loss": 0.9388,
+      "step": 4603
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7542731992182852,
+      "learning_rate": 5.325350671846801e-06,
+      "loss": 0.9532,
+      "step": 4604
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.280128039724611,
+      "learning_rate": 5.323613268730605e-06,
+      "loss": 0.9025,
+      "step": 4605
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8616513211245294,
+      "learning_rate": 5.321875826375185e-06,
+      "loss": 0.8365,
+      "step": 4606
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.0563842223177935,
+      "learning_rate": 5.32013834499121e-06,
+      "loss": 0.9732,
+      "step": 4607
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8761448601317794,
+      "learning_rate": 5.3184008247893595e-06,
+      "loss": 1.006,
+      "step": 4608
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.686803324532385,
+      "learning_rate": 5.31666326598031e-06,
+      "loss": 0.8534,
+      "step": 4609
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7103775041712825,
+      "learning_rate": 5.314925668774747e-06,
+      "loss": 0.9691,
+      "step": 4610
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7327069199147556,
+      "learning_rate": 5.313188033383363e-06,
+      "loss": 0.9939,
+      "step": 4611
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6494755203400253,
+      "learning_rate": 5.311450360016848e-06,
+      "loss": 0.8311,
+      "step": 4612
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7483451268583148,
+      "learning_rate": 5.3097126488859044e-06,
+      "loss": 0.8875,
+      "step": 4613
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6990428584893351,
+      "learning_rate": 5.3079749002012345e-06,
+      "loss": 0.8689,
+      "step": 4614
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8658343120710026,
+      "learning_rate": 5.306237114173546e-06,
+      "loss": 0.8778,
+      "step": 4615
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6350843307691532,
+      "learning_rate": 5.3044992910135515e-06,
+      "loss": 0.8677,
+      "step": 4616
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7249692060686297,
+      "learning_rate": 5.302761430931969e-06,
+      "loss": 0.8798,
+      "step": 4617
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.825229117102335,
+      "learning_rate": 5.301023534139516e-06,
+      "loss": 0.8657,
+      "step": 4618
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6677073739588832,
+      "learning_rate": 5.299285600846926e-06,
+      "loss": 0.8882,
+      "step": 4619
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8222687665777388,
+      "learning_rate": 5.297547631264923e-06,
+      "loss": 0.9094,
+      "step": 4620
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7832214996826337,
+      "learning_rate": 5.2958096256042435e-06,
+      "loss": 0.9459,
+      "step": 4621
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8134629872960856,
+      "learning_rate": 5.294071584075628e-06,
+      "loss": 1.0521,
+      "step": 4622
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9240228333849383,
+      "learning_rate": 5.2923335068898195e-06,
+      "loss": 0.9827,
+      "step": 4623
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7323451618256709,
+      "learning_rate": 5.290595394257564e-06,
+      "loss": 0.8631,
+      "step": 4624
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7335165381453945,
+      "learning_rate": 5.288857246389617e-06,
+      "loss": 0.8743,
+      "step": 4625
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6816715027045794,
+      "learning_rate": 5.287119063496731e-06,
+      "loss": 0.8159,
+      "step": 4626
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7455824876020585,
+      "learning_rate": 5.2853808457896695e-06,
+      "loss": 0.8407,
+      "step": 4627
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8105196590728867,
+      "learning_rate": 5.2836425934791965e-06,
+      "loss": 0.9235,
+      "step": 4628
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6026037352297327,
+      "learning_rate": 5.2819043067760824e-06,
+      "loss": 0.9288,
+      "step": 4629
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7512980273128977,
+      "learning_rate": 5.280165985891098e-06,
+      "loss": 0.926,
+      "step": 4630
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.672702842411126,
+      "learning_rate": 5.278427631035022e-06,
+      "loss": 0.8206,
+      "step": 4631
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7513664316094502,
+      "learning_rate": 5.276689242418635e-06,
+      "loss": 0.8668,
+      "step": 4632
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7326789867438002,
+      "learning_rate": 5.274950820252725e-06,
+      "loss": 0.8822,
+      "step": 4633
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8908274588421048,
+      "learning_rate": 5.27321236474808e-06,
+      "loss": 0.9884,
+      "step": 4634
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6917190503349,
+      "learning_rate": 5.271473876115494e-06,
+      "loss": 0.832,
+      "step": 4635
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7061425485270514,
+      "learning_rate": 5.269735354565764e-06,
+      "loss": 0.8352,
+      "step": 4636
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8046833829344417,
+      "learning_rate": 5.267996800309693e-06,
+      "loss": 0.892,
+      "step": 4637
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.879725768424691,
+      "learning_rate": 5.266258213558084e-06,
+      "loss": 0.957,
+      "step": 4638
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6466577504153728,
+      "learning_rate": 5.264519594521751e-06,
+      "loss": 0.8427,
+      "step": 4639
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.715273852841163,
+      "learning_rate": 5.262780943411504e-06,
+      "loss": 0.874,
+      "step": 4640
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.902334733830288,
+      "learning_rate": 5.261042260438163e-06,
+      "loss": 0.8975,
+      "step": 4641
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8035319549729754,
+      "learning_rate": 5.259303545812546e-06,
+      "loss": 0.8855,
+      "step": 4642
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7085984408582833,
+      "learning_rate": 5.257564799745481e-06,
+      "loss": 1.0221,
+      "step": 4643
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6900538493223507,
+      "learning_rate": 5.255826022447796e-06,
+      "loss": 0.9411,
+      "step": 4644
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8055883979717955,
+      "learning_rate": 5.254087214130324e-06,
+      "loss": 0.9634,
+      "step": 4645
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7806593542176847,
+      "learning_rate": 5.252348375003902e-06,
+      "loss": 0.909,
+      "step": 4646
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7534707635130935,
+      "learning_rate": 5.2506095052793695e-06,
+      "loss": 0.8806,
+      "step": 4647
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7935282121799816,
+      "learning_rate": 5.24887060516757e-06,
+      "loss": 0.868,
+      "step": 4648
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8346726882323847,
+      "learning_rate": 5.24713167487935e-06,
+      "loss": 1.0241,
+      "step": 4649
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7183385991024804,
+      "learning_rate": 5.2453927146255646e-06,
+      "loss": 0.9566,
+      "step": 4650
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7032994750572283,
+      "learning_rate": 5.2436537246170675e-06,
+      "loss": 0.8738,
+      "step": 4651
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7741417968228432,
+      "learning_rate": 5.241914705064713e-06,
+      "loss": 0.9429,
+      "step": 4652
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8091718569413904,
+      "learning_rate": 5.2401756561793686e-06,
+      "loss": 1.0352,
+      "step": 4653
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5473360986156506,
+      "learning_rate": 5.238436578171899e-06,
+      "loss": 0.8129,
+      "step": 4654
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7037449023925755,
+      "learning_rate": 5.236697471253168e-06,
+      "loss": 0.9233,
+      "step": 4655
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8282079654270955,
+      "learning_rate": 5.2349583356340575e-06,
+      "loss": 1.0523,
+      "step": 4656
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7171287861296636,
+      "learning_rate": 5.233219171525435e-06,
+      "loss": 0.9306,
+      "step": 4657
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6948430386018638,
+      "learning_rate": 5.231479979138187e-06,
+      "loss": 0.8593,
+      "step": 4658
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.679841879965499,
+      "learning_rate": 5.229740758683191e-06,
+      "loss": 0.8747,
+      "step": 4659
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.684122506833001,
+      "learning_rate": 5.228001510371336e-06,
+      "loss": 0.9217,
+      "step": 4660
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6755217982249582,
+      "learning_rate": 5.226262234413514e-06,
+      "loss": 0.8043,
+      "step": 4661
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8689141853034956,
+      "learning_rate": 5.224522931020616e-06,
+      "loss": 0.7967,
+      "step": 4662
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7211060850798299,
+      "learning_rate": 5.222783600403536e-06,
+      "loss": 1.0156,
+      "step": 4663
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6723564027846356,
+      "learning_rate": 5.2210442427731774e-06,
+      "loss": 0.9582,
+      "step": 4664
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6867185025040994,
+      "learning_rate": 5.2193048583404435e-06,
+      "loss": 0.8332,
+      "step": 4665
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6855970312046424,
+      "learning_rate": 5.217565447316238e-06,
+      "loss": 0.8853,
+      "step": 4666
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7247923440963406,
+      "learning_rate": 5.215826009911474e-06,
+      "loss": 0.9259,
+      "step": 4667
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.865404951427475,
+      "learning_rate": 5.214086546337061e-06,
+      "loss": 0.9305,
+      "step": 4668
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7124016475459252,
+      "learning_rate": 5.212347056803916e-06,
+      "loss": 0.9411,
+      "step": 4669
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6782666019229384,
+      "learning_rate": 5.210607541522959e-06,
+      "loss": 0.9349,
+      "step": 4670
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6806949785879604,
+      "learning_rate": 5.208868000705109e-06,
+      "loss": 0.9263,
+      "step": 4671
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.673224801467606,
+      "learning_rate": 5.2071284345612975e-06,
+      "loss": 0.8365,
+      "step": 4672
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8106465989183727,
+      "learning_rate": 5.205388843302447e-06,
+      "loss": 0.9023,
+      "step": 4673
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.659037687953835,
+      "learning_rate": 5.2036492271394915e-06,
+      "loss": 0.8482,
+      "step": 4674
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8107280254431233,
+      "learning_rate": 5.201909586283365e-06,
+      "loss": 0.8612,
+      "step": 4675
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6390588979871388,
+      "learning_rate": 5.200169920945005e-06,
+      "loss": 0.8622,
+      "step": 4676
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7561413646389288,
+      "learning_rate": 5.198430231335353e-06,
+      "loss": 0.9812,
+      "step": 4677
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7206455174387931,
+      "learning_rate": 5.19669051766535e-06,
+      "loss": 0.8775,
+      "step": 4678
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.664345175163039,
+      "learning_rate": 5.1949507801459455e-06,
+      "loss": 0.8536,
+      "step": 4679
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.685887386860053,
+      "learning_rate": 5.193211018988085e-06,
+      "loss": 0.8213,
+      "step": 4680
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7387999861424646,
+      "learning_rate": 5.191471234402723e-06,
+      "loss": 0.9391,
+      "step": 4681
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.822313074342098,
+      "learning_rate": 5.189731426600814e-06,
+      "loss": 0.8715,
+      "step": 4682
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7007484205570726,
+      "learning_rate": 5.187991595793314e-06,
+      "loss": 0.799,
+      "step": 4683
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7467479723325923,
+      "learning_rate": 5.1862517421911875e-06,
+      "loss": 0.8908,
+      "step": 4684
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6794355854376535,
+      "learning_rate": 5.184511866005393e-06,
+      "loss": 0.8693,
+      "step": 4685
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6369118875533104,
+      "learning_rate": 5.182771967446899e-06,
+      "loss": 0.9464,
+      "step": 4686
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.722970626712947,
+      "learning_rate": 5.181032046726674e-06,
+      "loss": 0.9361,
+      "step": 4687
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7125789688149462,
+      "learning_rate": 5.17929210405569e-06,
+      "loss": 0.9724,
+      "step": 4688
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6530629707486482,
+      "learning_rate": 5.177552139644919e-06,
+      "loss": 0.9455,
+      "step": 4689
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.685861485122584,
+      "learning_rate": 5.17581215370534e-06,
+      "loss": 0.9516,
+      "step": 4690
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.692791703795029,
+      "learning_rate": 5.17407214644793e-06,
+      "loss": 0.8825,
+      "step": 4691
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7719263972228192,
+      "learning_rate": 5.172332118083673e-06,
+      "loss": 0.9329,
+      "step": 4692
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7086818876422143,
+      "learning_rate": 5.170592068823554e-06,
+      "loss": 0.939,
+      "step": 4693
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6999018699786816,
+      "learning_rate": 5.168851998878556e-06,
+      "loss": 0.8727,
+      "step": 4694
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.676477408944589,
+      "learning_rate": 5.1671119084596725e-06,
+      "loss": 0.9673,
+      "step": 4695
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8857400581083027,
+      "learning_rate": 5.165371797777894e-06,
+      "loss": 1.068,
+      "step": 4696
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7767791450218338,
+      "learning_rate": 5.163631667044213e-06,
+      "loss": 0.8858,
+      "step": 4697
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6537987324388463,
+      "learning_rate": 5.1618915164696304e-06,
+      "loss": 0.8885,
+      "step": 4698
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7750384008407645,
+      "learning_rate": 5.160151346265143e-06,
+      "loss": 0.867,
+      "step": 4699
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6805440214214908,
+      "learning_rate": 5.158411156641752e-06,
+      "loss": 0.8823,
+      "step": 4700
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.650033070846011,
+      "learning_rate": 5.156670947810463e-06,
+      "loss": 0.8647,
+      "step": 4701
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8096955169928213,
+      "learning_rate": 5.154930719982281e-06,
+      "loss": 0.9324,
+      "step": 4702
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.698529259645741,
+      "learning_rate": 5.153190473368213e-06,
+      "loss": 0.934,
+      "step": 4703
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6363590350479256,
+      "learning_rate": 5.151450208179276e-06,
+      "loss": 0.7885,
+      "step": 4704
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.799676340877873,
+      "learning_rate": 5.149709924626476e-06,
+      "loss": 0.8338,
+      "step": 4705
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7272070122381717,
+      "learning_rate": 5.147969622920833e-06,
+      "loss": 0.8736,
+      "step": 4706
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7504478431687218,
+      "learning_rate": 5.146229303273364e-06,
+      "loss": 0.9128,
+      "step": 4707
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7287306899698407,
+      "learning_rate": 5.144488965895084e-06,
+      "loss": 0.9403,
+      "step": 4708
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.900063501428234,
+      "learning_rate": 5.142748610997024e-06,
+      "loss": 0.8998,
+      "step": 4709
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6854663458028285,
+      "learning_rate": 5.1410082387902e-06,
+      "loss": 0.8287,
+      "step": 4710
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6510741504136797,
+      "learning_rate": 5.139267849485639e-06,
+      "loss": 0.8786,
+      "step": 4711
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8432290695556817,
+      "learning_rate": 5.1375274432943745e-06,
+      "loss": 1.0081,
+      "step": 4712
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.685499381294145,
+      "learning_rate": 5.135787020427431e-06,
+      "loss": 0.8827,
+      "step": 4713
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7189958328087633,
+      "learning_rate": 5.134046581095844e-06,
+      "loss": 0.8548,
+      "step": 4714
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7616004313609683,
+      "learning_rate": 5.132306125510649e-06,
+      "loss": 0.807,
+      "step": 4715
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7868691143168831,
+      "learning_rate": 5.130565653882877e-06,
+      "loss": 0.9133,
+      "step": 4716
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7772789080039024,
+      "learning_rate": 5.12882516642357e-06,
+      "loss": 0.9395,
+      "step": 4717
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7178137006406249,
+      "learning_rate": 5.127084663343769e-06,
+      "loss": 0.9299,
+      "step": 4718
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6934601053346077,
+      "learning_rate": 5.125344144854514e-06,
+      "loss": 0.8726,
+      "step": 4719
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7345311422523366,
+      "learning_rate": 5.123603611166849e-06,
+      "loss": 0.8394,
+      "step": 4720
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8912498098835481,
+      "learning_rate": 5.121863062491819e-06,
+      "loss": 1.097,
+      "step": 4721
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7426498045054293,
+      "learning_rate": 5.120122499040473e-06,
+      "loss": 0.8008,
+      "step": 4722
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9433080814967136,
+      "learning_rate": 5.118381921023859e-06,
+      "loss": 0.9114,
+      "step": 4723
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7814654948141198,
+      "learning_rate": 5.116641328653031e-06,
+      "loss": 0.9349,
+      "step": 4724
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8476925083307308,
+      "learning_rate": 5.1149007221390386e-06,
+      "loss": 1.0228,
+      "step": 4725
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.710626564060376,
+      "learning_rate": 5.113160101692939e-06,
+      "loss": 0.9077,
+      "step": 4726
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6932430183902245,
+      "learning_rate": 5.111419467525786e-06,
+      "loss": 0.9194,
+      "step": 4727
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7425876796972077,
+      "learning_rate": 5.109678819848637e-06,
+      "loss": 0.8282,
+      "step": 4728
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7239156448404551,
+      "learning_rate": 5.1079381588725544e-06,
+      "loss": 0.934,
+      "step": 4729
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7197050816190993,
+      "learning_rate": 5.106197484808598e-06,
+      "loss": 0.9461,
+      "step": 4730
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6989920769219904,
+      "learning_rate": 5.104456797867831e-06,
+      "loss": 0.9028,
+      "step": 4731
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8248161967967698,
+      "learning_rate": 5.102716098261316e-06,
+      "loss": 0.9419,
+      "step": 4732
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.819220088292649,
+      "learning_rate": 5.10097538620012e-06,
+      "loss": 1.0306,
+      "step": 4733
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7244734188930948,
+      "learning_rate": 5.09923466189531e-06,
+      "loss": 1.0134,
+      "step": 4734
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7259640537279293,
+      "learning_rate": 5.097493925557956e-06,
+      "loss": 0.8928,
+      "step": 4735
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7209419001011022,
+      "learning_rate": 5.095753177399127e-06,
+      "loss": 0.9136,
+      "step": 4736
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7071871727002306,
+      "learning_rate": 5.0940124176298955e-06,
+      "loss": 0.8989,
+      "step": 4737
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8070499432221345,
+      "learning_rate": 5.092271646461334e-06,
+      "loss": 0.9953,
+      "step": 4738
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7517178080971165,
+      "learning_rate": 5.090530864104517e-06,
+      "loss": 0.9309,
+      "step": 4739
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6154290679092247,
+      "learning_rate": 5.0887900707705205e-06,
+      "loss": 0.9123,
+      "step": 4740
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.634135685952796,
+      "learning_rate": 5.087049266670424e-06,
+      "loss": 0.9218,
+      "step": 4741
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7362638849291596,
+      "learning_rate": 5.085308452015301e-06,
+      "loss": 0.8782,
+      "step": 4742
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.653567802905575,
+      "learning_rate": 5.0835676270162346e-06,
+      "loss": 0.9128,
+      "step": 4743
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.787941006730624,
+      "learning_rate": 5.081826791884307e-06,
+      "loss": 0.8422,
+      "step": 4744
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6975393713476306,
+      "learning_rate": 5.080085946830597e-06,
+      "loss": 0.8667,
+      "step": 4745
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7709590275934095,
+      "learning_rate": 5.078345092066191e-06,
+      "loss": 0.9821,
+      "step": 4746
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6851676586736077,
+      "learning_rate": 5.076604227802172e-06,
+      "loss": 0.8617,
+      "step": 4747
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6780966903850907,
+      "learning_rate": 5.074863354249625e-06,
+      "loss": 0.9429,
+      "step": 4748
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7694564947855016,
+      "learning_rate": 5.073122471619641e-06,
+      "loss": 0.9326,
+      "step": 4749
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9192236529921621,
+      "learning_rate": 5.0713815801233025e-06,
+      "loss": 1.0806,
+      "step": 4750
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8094256937574864,
+      "learning_rate": 5.069640679971702e-06,
+      "loss": 0.9047,
+      "step": 4751
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.617099179600719,
+      "learning_rate": 5.0678997713759305e-06,
+      "loss": 0.9366,
+      "step": 4752
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6585948603398726,
+      "learning_rate": 5.066158854547076e-06,
+      "loss": 0.9135,
+      "step": 4753
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8328440620211266,
+      "learning_rate": 5.064417929696233e-06,
+      "loss": 0.9406,
+      "step": 4754
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7884813054603999,
+      "learning_rate": 5.062676997034494e-06,
+      "loss": 0.9033,
+      "step": 4755
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7596241395064007,
+      "learning_rate": 5.060936056772952e-06,
+      "loss": 0.8833,
+      "step": 4756
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7707263034468876,
+      "learning_rate": 5.059195109122705e-06,
+      "loss": 0.905,
+      "step": 4757
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8458093726400087,
+      "learning_rate": 5.057454154294846e-06,
+      "loss": 0.8706,
+      "step": 4758
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6759211096121591,
+      "learning_rate": 5.055713192500472e-06,
+      "loss": 0.932,
+      "step": 4759
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6801193331167659,
+      "learning_rate": 5.053972223950682e-06,
+      "loss": 0.9218,
+      "step": 4760
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6539917691599473,
+      "learning_rate": 5.052231248856574e-06,
+      "loss": 0.885,
+      "step": 4761
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7045298975200525,
+      "learning_rate": 5.050490267429246e-06,
+      "loss": 0.9031,
+      "step": 4762
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7972579993698374,
+      "learning_rate": 5.0487492798798e-06,
+      "loss": 0.9407,
+      "step": 4763
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9486738941212485,
+      "learning_rate": 5.047008286419336e-06,
+      "loss": 0.8981,
+      "step": 4764
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6839135486027523,
+      "learning_rate": 5.0452672872589535e-06,
+      "loss": 0.9293,
+      "step": 4765
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.769080031595128,
+      "learning_rate": 5.043526282609757e-06,
+      "loss": 0.9376,
+      "step": 4766
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6648809732260106,
+      "learning_rate": 5.04178527268285e-06,
+      "loss": 0.8499,
+      "step": 4767
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9298363149288023,
+      "learning_rate": 5.040044257689334e-06,
+      "loss": 1.0565,
+      "step": 4768
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8062610957216028,
+      "learning_rate": 5.038303237840314e-06,
+      "loss": 0.9375,
+      "step": 4769
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6786032960274666,
+      "learning_rate": 5.036562213346892e-06,
+      "loss": 0.8815,
+      "step": 4770
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7761196458151665,
+      "learning_rate": 5.0348211844201786e-06,
+      "loss": 0.9342,
+      "step": 4771
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6903762677850032,
+      "learning_rate": 5.033080151271276e-06,
+      "loss": 0.9238,
+      "step": 4772
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.745540257783605,
+      "learning_rate": 5.03133911411129e-06,
+      "loss": 0.9402,
+      "step": 4773
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6773748049526107,
+      "learning_rate": 5.029598073151329e-06,
+      "loss": 0.8583,
+      "step": 4774
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8351588029202397,
+      "learning_rate": 5.0278570286025e-06,
+      "loss": 0.9198,
+      "step": 4775
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.822199635499785,
+      "learning_rate": 5.0261159806759086e-06,
+      "loss": 0.972,
+      "step": 4776
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7550983007511867,
+      "learning_rate": 5.024374929582665e-06,
+      "loss": 0.8906,
+      "step": 4777
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6938467824753478,
+      "learning_rate": 5.022633875533879e-06,
+      "loss": 0.9406,
+      "step": 4778
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7262467488750828,
+      "learning_rate": 5.020892818740655e-06,
+      "loss": 0.9384,
+      "step": 4779
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.702656062676979,
+      "learning_rate": 5.019151759414107e-06,
+      "loss": 0.9074,
+      "step": 4780
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.796300295115342,
+      "learning_rate": 5.0174106977653415e-06,
+      "loss": 0.9345,
+      "step": 4781
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8993377898522714,
+      "learning_rate": 5.015669634005467e-06,
+      "loss": 0.9306,
+      "step": 4782
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6354457280224923,
+      "learning_rate": 5.013928568345598e-06,
+      "loss": 0.8325,
+      "step": 4783
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9218539352515713,
+      "learning_rate": 5.01218750099684e-06,
+      "loss": 1.0467,
+      "step": 4784
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6287606711371216,
+      "learning_rate": 5.010446432170306e-06,
+      "loss": 0.9311,
+      "step": 4785
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7029420064127943,
+      "learning_rate": 5.008705362077108e-06,
+      "loss": 0.919,
+      "step": 4786
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7368254245738373,
+      "learning_rate": 5.006964290928351e-06,
+      "loss": 0.8858,
+      "step": 4787
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6024921740491467,
+      "learning_rate": 5.0052232189351524e-06,
+      "loss": 0.9305,
+      "step": 4788
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6385307225849608,
+      "learning_rate": 5.003482146308621e-06,
+      "loss": 0.8846,
+      "step": 4789
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.626391623580546,
+      "learning_rate": 5.001741073259866e-06,
+      "loss": 0.9018,
+      "step": 4790
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7115494651105547,
+      "learning_rate": 5e-06,
+      "loss": 0.8457,
+      "step": 4791
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7034067908295565,
+      "learning_rate": 4.998258926740135e-06,
+      "loss": 0.8892,
+      "step": 4792
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7617499344892245,
+      "learning_rate": 4.99651785369138e-06,
+      "loss": 0.9169,
+      "step": 4793
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6863659439187473,
+      "learning_rate": 4.9947767810648475e-06,
+      "loss": 0.8515,
+      "step": 4794
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7052613127497251,
+      "learning_rate": 4.993035709071649e-06,
+      "loss": 0.895,
+      "step": 4795
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7535963883285903,
+      "learning_rate": 4.991294637922893e-06,
+      "loss": 0.9175,
+      "step": 4796
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7451774280943328,
+      "learning_rate": 4.989553567829696e-06,
+      "loss": 0.9251,
+      "step": 4797
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7221937410329757,
+      "learning_rate": 4.987812499003162e-06,
+      "loss": 0.8727,
+      "step": 4798
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7688201644805153,
+      "learning_rate": 4.986071431654405e-06,
+      "loss": 0.852,
+      "step": 4799
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7273363863419482,
+      "learning_rate": 4.984330365994535e-06,
+      "loss": 0.9188,
+      "step": 4800
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8887239502477813,
+      "learning_rate": 4.98258930223466e-06,
+      "loss": 0.858,
+      "step": 4801
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9072582683342727,
+      "learning_rate": 4.980848240585895e-06,
+      "loss": 0.8942,
+      "step": 4802
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.655868837620404,
+      "learning_rate": 4.979107181259346e-06,
+      "loss": 0.9452,
+      "step": 4803
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6628130420812606,
+      "learning_rate": 4.977366124466122e-06,
+      "loss": 0.9178,
+      "step": 4804
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7746133207574712,
+      "learning_rate": 4.975625070417336e-06,
+      "loss": 0.9363,
+      "step": 4805
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7922534236053538,
+      "learning_rate": 4.973884019324092e-06,
+      "loss": 0.8927,
+      "step": 4806
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6460441713949103,
+      "learning_rate": 4.972142971397504e-06,
+      "loss": 0.8145,
+      "step": 4807
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.917648631524272,
+      "learning_rate": 4.970401926848674e-06,
+      "loss": 1.0628,
+      "step": 4808
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.708967594211526,
+      "learning_rate": 4.968660885888711e-06,
+      "loss": 0.9043,
+      "step": 4809
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.730400383382931,
+      "learning_rate": 4.966919848728726e-06,
+      "loss": 0.8849,
+      "step": 4810
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6496946654061142,
+      "learning_rate": 4.965178815579823e-06,
+      "loss": 0.8932,
+      "step": 4811
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8227242241555114,
+      "learning_rate": 4.963437786653108e-06,
+      "loss": 1.0611,
+      "step": 4812
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7230147049972075,
+      "learning_rate": 4.961696762159688e-06,
+      "loss": 0.9324,
+      "step": 4813
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7401100799768645,
+      "learning_rate": 4.959955742310668e-06,
+      "loss": 0.86,
+      "step": 4814
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6954359506760037,
+      "learning_rate": 4.958214727317152e-06,
+      "loss": 1.0114,
+      "step": 4815
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7431264884551454,
+      "learning_rate": 4.956473717390243e-06,
+      "loss": 0.9806,
+      "step": 4816
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8429996612992005,
+      "learning_rate": 4.9547327127410465e-06,
+      "loss": 1.0187,
+      "step": 4817
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8435627514073136,
+      "learning_rate": 4.952991713580666e-06,
+      "loss": 1.0812,
+      "step": 4818
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7850147919658506,
+      "learning_rate": 4.951250720120203e-06,
+      "loss": 0.9662,
+      "step": 4819
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7801700603157813,
+      "learning_rate": 4.949509732570756e-06,
+      "loss": 0.8399,
+      "step": 4820
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8506915405232696,
+      "learning_rate": 4.947768751143428e-06,
+      "loss": 1.0901,
+      "step": 4821
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7591150263682396,
+      "learning_rate": 4.94602777604932e-06,
+      "loss": 0.8969,
+      "step": 4822
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7194445729195134,
+      "learning_rate": 4.9442868074995296e-06,
+      "loss": 0.8947,
+      "step": 4823
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6439758898773533,
+      "learning_rate": 4.942545845705156e-06,
+      "loss": 0.8358,
+      "step": 4824
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6857027545425474,
+      "learning_rate": 4.9408048908772965e-06,
+      "loss": 0.8942,
+      "step": 4825
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6991767035652792,
+      "learning_rate": 4.939063943227049e-06,
+      "loss": 0.863,
+      "step": 4826
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8350409648325308,
+      "learning_rate": 4.937323002965507e-06,
+      "loss": 0.9225,
+      "step": 4827
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6774764008375693,
+      "learning_rate": 4.935582070303767e-06,
+      "loss": 0.8259,
+      "step": 4828
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7192534155454902,
+      "learning_rate": 4.933841145452925e-06,
+      "loss": 0.9671,
+      "step": 4829
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7724993459495595,
+      "learning_rate": 4.932100228624072e-06,
+      "loss": 0.928,
+      "step": 4830
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7446309758462493,
+      "learning_rate": 4.9303593200282995e-06,
+      "loss": 0.9082,
+      "step": 4831
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6614335614089262,
+      "learning_rate": 4.928618419876698e-06,
+      "loss": 0.978,
+      "step": 4832
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.727420708655698,
+      "learning_rate": 4.926877528380362e-06,
+      "loss": 0.9348,
+      "step": 4833
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6845516467933916,
+      "learning_rate": 4.925136645750377e-06,
+      "loss": 0.8845,
+      "step": 4834
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7590560227274192,
+      "learning_rate": 4.923395772197829e-06,
+      "loss": 0.9784,
+      "step": 4835
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.705534386820642,
+      "learning_rate": 4.9216549079338105e-06,
+      "loss": 0.9163,
+      "step": 4836
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7662771158122483,
+      "learning_rate": 4.919914053169404e-06,
+      "loss": 0.8164,
+      "step": 4837
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6706093813394591,
+      "learning_rate": 4.918173208115695e-06,
+      "loss": 0.9098,
+      "step": 4838
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6863149987741377,
+      "learning_rate": 4.916432372983768e-06,
+      "loss": 0.9285,
+      "step": 4839
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8406799270566048,
+      "learning_rate": 4.914691547984701e-06,
+      "loss": 0.982,
+      "step": 4840
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9635601251887823,
+      "learning_rate": 4.912950733329579e-06,
+      "loss": 1.0319,
+      "step": 4841
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.753867279481688,
+      "learning_rate": 4.91120992922948e-06,
+      "loss": 0.9833,
+      "step": 4842
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9315705758611811,
+      "learning_rate": 4.909469135895485e-06,
+      "loss": 1.0183,
+      "step": 4843
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7676605937450491,
+      "learning_rate": 4.9077283535386675e-06,
+      "loss": 0.853,
+      "step": 4844
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7042795112689766,
+      "learning_rate": 4.905987582370106e-06,
+      "loss": 0.8703,
+      "step": 4845
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7850443844228427,
+      "learning_rate": 4.904246822600874e-06,
+      "loss": 0.9188,
+      "step": 4846
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7426662083404159,
+      "learning_rate": 4.902506074442044e-06,
+      "loss": 0.8057,
+      "step": 4847
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7208155566425132,
+      "learning_rate": 4.900765338104691e-06,
+      "loss": 0.9724,
+      "step": 4848
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.729412391628358,
+      "learning_rate": 4.899024613799881e-06,
+      "loss": 0.8888,
+      "step": 4849
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7370049990550247,
+      "learning_rate": 4.897283901738687e-06,
+      "loss": 0.9518,
+      "step": 4850
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6483186413808462,
+      "learning_rate": 4.8955432021321725e-06,
+      "loss": 0.8348,
+      "step": 4851
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7788386554899722,
+      "learning_rate": 4.893802515191404e-06,
+      "loss": 0.875,
+      "step": 4852
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7471770332821512,
+      "learning_rate": 4.892061841127446e-06,
+      "loss": 1.0048,
+      "step": 4853
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.665553389328553,
+      "learning_rate": 4.8903211801513645e-06,
+      "loss": 0.8034,
+      "step": 4854
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0009174307895428,
+      "learning_rate": 4.888580532474216e-06,
+      "loss": 1.0408,
+      "step": 4855
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7172313399960755,
+      "learning_rate": 4.886839898307062e-06,
+      "loss": 1.0074,
+      "step": 4856
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.914026532487705,
+      "learning_rate": 4.885099277860962e-06,
+      "loss": 0.85,
+      "step": 4857
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7021606974594914,
+      "learning_rate": 4.883358671346969e-06,
+      "loss": 0.8934,
+      "step": 4858
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6670179251155857,
+      "learning_rate": 4.88161807897614e-06,
+      "loss": 0.8868,
+      "step": 4859
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6802884049022515,
+      "learning_rate": 4.879877500959529e-06,
+      "loss": 0.9294,
+      "step": 4860
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6750226440965872,
+      "learning_rate": 4.878136937508183e-06,
+      "loss": 0.8543,
+      "step": 4861
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7285709086271681,
+      "learning_rate": 4.876396388833155e-06,
+      "loss": 0.9236,
+      "step": 4862
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.810884848401018,
+      "learning_rate": 4.874655855145488e-06,
+      "loss": 0.983,
+      "step": 4863
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7688546435412285,
+      "learning_rate": 4.872915336656233e-06,
+      "loss": 0.9332,
+      "step": 4864
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6929533595809845,
+      "learning_rate": 4.871174833576431e-06,
+      "loss": 0.9325,
+      "step": 4865
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6924147825152458,
+      "learning_rate": 4.869434346117124e-06,
+      "loss": 0.8823,
+      "step": 4866
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.721282898115799,
+      "learning_rate": 4.867693874489353e-06,
+      "loss": 0.8598,
+      "step": 4867
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5959027853042433,
+      "learning_rate": 4.865953418904156e-06,
+      "loss": 0.9318,
+      "step": 4868
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6625208677345755,
+      "learning_rate": 4.864212979572569e-06,
+      "loss": 0.9147,
+      "step": 4869
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7730603584715678,
+      "learning_rate": 4.862472556705626e-06,
+      "loss": 0.9592,
+      "step": 4870
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.752387856832597,
+      "learning_rate": 4.8607321505143614e-06,
+      "loss": 0.8357,
+      "step": 4871
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7462785312499836,
+      "learning_rate": 4.858991761209803e-06,
+      "loss": 0.9323,
+      "step": 4872
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.757173767268217,
+      "learning_rate": 4.85725138900298e-06,
+      "loss": 0.9003,
+      "step": 4873
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.81650699236033,
+      "learning_rate": 4.855511034104916e-06,
+      "loss": 0.8852,
+      "step": 4874
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.0143147896129596,
+      "learning_rate": 4.853770696726639e-06,
+      "loss": 0.9466,
+      "step": 4875
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6719593650470217,
+      "learning_rate": 4.852030377079169e-06,
+      "loss": 0.9374,
+      "step": 4876
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.861831702530439,
+      "learning_rate": 4.850290075373525e-06,
+      "loss": 0.9821,
+      "step": 4877
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7047592148532476,
+      "learning_rate": 4.848549791820725e-06,
+      "loss": 0.8953,
+      "step": 4878
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8044570263578867,
+      "learning_rate": 4.846809526631787e-06,
+      "loss": 0.919,
+      "step": 4879
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7041214123114021,
+      "learning_rate": 4.845069280017721e-06,
+      "loss": 0.9141,
+      "step": 4880
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.740102159523388,
+      "learning_rate": 4.843329052189538e-06,
+      "loss": 0.8715,
+      "step": 4881
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6555008551474355,
+      "learning_rate": 4.841588843358251e-06,
+      "loss": 0.8648,
+      "step": 4882
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.688443291249926,
+      "learning_rate": 4.839848653734859e-06,
+      "loss": 0.9483,
+      "step": 4883
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.675964683963416,
+      "learning_rate": 4.838108483530371e-06,
+      "loss": 0.9117,
+      "step": 4884
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7733940112563926,
+      "learning_rate": 4.8363683329557885e-06,
+      "loss": 0.8392,
+      "step": 4885
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.723104068905473,
+      "learning_rate": 4.834628202222107e-06,
+      "loss": 0.879,
+      "step": 4886
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7076784268290062,
+      "learning_rate": 4.832888091540328e-06,
+      "loss": 0.9032,
+      "step": 4887
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6965639441495983,
+      "learning_rate": 4.831148001121445e-06,
+      "loss": 0.8856,
+      "step": 4888
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7562527913751769,
+      "learning_rate": 4.829407931176447e-06,
+      "loss": 0.8882,
+      "step": 4889
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.805934885301784,
+      "learning_rate": 4.827667881916327e-06,
+      "loss": 0.9803,
+      "step": 4890
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.655831784148836,
+      "learning_rate": 4.825927853552071e-06,
+      "loss": 0.8735,
+      "step": 4891
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7267496587135869,
+      "learning_rate": 4.824187846294662e-06,
+      "loss": 0.8875,
+      "step": 4892
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6267473470805156,
+      "learning_rate": 4.822447860355083e-06,
+      "loss": 0.9006,
+      "step": 4893
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6723025215625729,
+      "learning_rate": 4.820707895944312e-06,
+      "loss": 0.8885,
+      "step": 4894
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6853589333631322,
+      "learning_rate": 4.818967953273327e-06,
+      "loss": 0.9058,
+      "step": 4895
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7831704948342737,
+      "learning_rate": 4.817228032553103e-06,
+      "loss": 0.9068,
+      "step": 4896
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6910273452221423,
+      "learning_rate": 4.815488133994608e-06,
+      "loss": 0.8924,
+      "step": 4897
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7380444877150196,
+      "learning_rate": 4.813748257808814e-06,
+      "loss": 0.8337,
+      "step": 4898
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.765943202911277,
+      "learning_rate": 4.8120084042066865e-06,
+      "loss": 0.9243,
+      "step": 4899
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.710470832650906,
+      "learning_rate": 4.810268573399187e-06,
+      "loss": 0.8523,
+      "step": 4900
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7174725698469502,
+      "learning_rate": 4.8085287655972775e-06,
+      "loss": 0.9295,
+      "step": 4901
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.940075942342574,
+      "learning_rate": 4.806788981011916e-06,
+      "loss": 0.8977,
+      "step": 4902
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8416865749041946,
+      "learning_rate": 4.805049219854058e-06,
+      "loss": 1.0693,
+      "step": 4903
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6555694395362475,
+      "learning_rate": 4.803309482334652e-06,
+      "loss": 0.9145,
+      "step": 4904
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6922666094176584,
+      "learning_rate": 4.801569768664649e-06,
+      "loss": 0.8611,
+      "step": 4905
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8759528360746873,
+      "learning_rate": 4.799830079054996e-06,
+      "loss": 0.9331,
+      "step": 4906
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.853245051424717,
+      "learning_rate": 4.798090413716637e-06,
+      "loss": 0.9654,
+      "step": 4907
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7261593599767453,
+      "learning_rate": 4.796350772860511e-06,
+      "loss": 0.9421,
+      "step": 4908
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7399779419358394,
+      "learning_rate": 4.7946111566975545e-06,
+      "loss": 1.0012,
+      "step": 4909
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.667344544063853,
+      "learning_rate": 4.792871565438705e-06,
+      "loss": 0.866,
+      "step": 4910
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7048439457956874,
+      "learning_rate": 4.7911319992948914e-06,
+      "loss": 0.8638,
+      "step": 4911
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6487921953177145,
+      "learning_rate": 4.789392458477042e-06,
+      "loss": 0.8658,
+      "step": 4912
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7510075508692577,
+      "learning_rate": 4.787652943196087e-06,
+      "loss": 0.8571,
+      "step": 4913
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7178197112208113,
+      "learning_rate": 4.785913453662941e-06,
+      "loss": 0.8907,
+      "step": 4914
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7542090324121329,
+      "learning_rate": 4.784173990088529e-06,
+      "loss": 0.8424,
+      "step": 4915
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7654776563926982,
+      "learning_rate": 4.782434552683763e-06,
+      "loss": 0.8811,
+      "step": 4916
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7584989473836496,
+      "learning_rate": 4.780695141659557e-06,
+      "loss": 0.9332,
+      "step": 4917
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7691730439629336,
+      "learning_rate": 4.778955757226823e-06,
+      "loss": 0.9158,
+      "step": 4918
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7138177429477168,
+      "learning_rate": 4.777216399596466e-06,
+      "loss": 0.9147,
+      "step": 4919
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7723941864952462,
+      "learning_rate": 4.775477068979386e-06,
+      "loss": 0.9024,
+      "step": 4920
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8676529433827483,
+      "learning_rate": 4.773737765586487e-06,
+      "loss": 1.0353,
+      "step": 4921
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6702229850783805,
+      "learning_rate": 4.771998489628664e-06,
+      "loss": 0.8439,
+      "step": 4922
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6606929942531037,
+      "learning_rate": 4.7702592413168095e-06,
+      "loss": 0.8956,
+      "step": 4923
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.807287816449294,
+      "learning_rate": 4.768520020861817e-06,
+      "loss": 0.9506,
+      "step": 4924
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8094032645768986,
+      "learning_rate": 4.766780828474566e-06,
+      "loss": 1.0188,
+      "step": 4925
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7171466898882788,
+      "learning_rate": 4.765041664365945e-06,
+      "loss": 0.8827,
+      "step": 4926
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8413143291446661,
+      "learning_rate": 4.763302528746833e-06,
+      "loss": 1.0625,
+      "step": 4927
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7792857919709386,
+      "learning_rate": 4.761563421828104e-06,
+      "loss": 0.8941,
+      "step": 4928
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6568057029664993,
+      "learning_rate": 4.759824343820632e-06,
+      "loss": 0.9548,
+      "step": 4929
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7821034969480372,
+      "learning_rate": 4.758085294935288e-06,
+      "loss": 0.9084,
+      "step": 4930
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8216061055780877,
+      "learning_rate": 4.756346275382934e-06,
+      "loss": 0.9277,
+      "step": 4931
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.728434840077039,
+      "learning_rate": 4.754607285374435e-06,
+      "loss": 0.9187,
+      "step": 4932
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6379028606385058,
+      "learning_rate": 4.75286832512065e-06,
+      "loss": 0.8911,
+      "step": 4933
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.773831547618212,
+      "learning_rate": 4.7511293948324325e-06,
+      "loss": 0.937,
+      "step": 4934
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6832624291615406,
+      "learning_rate": 4.749390494720633e-06,
+      "loss": 0.8854,
+      "step": 4935
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7084327394601138,
+      "learning_rate": 4.7476516249961e-06,
+      "loss": 0.8857,
+      "step": 4936
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8848238275333606,
+      "learning_rate": 4.7459127858696765e-06,
+      "loss": 1.0375,
+      "step": 4937
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7727857331064247,
+      "learning_rate": 4.744173977552205e-06,
+      "loss": 0.9043,
+      "step": 4938
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7041458754670333,
+      "learning_rate": 4.74243520025452e-06,
+      "loss": 0.9586,
+      "step": 4939
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6784417678457635,
+      "learning_rate": 4.7406964541874546e-06,
+      "loss": 0.8888,
+      "step": 4940
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6991211169308387,
+      "learning_rate": 4.7389577395618395e-06,
+      "loss": 0.9102,
+      "step": 4941
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.386887547925124,
+      "learning_rate": 4.737219056588497e-06,
+      "loss": 0.8723,
+      "step": 4942
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6572630967689768,
+      "learning_rate": 4.735480405478249e-06,
+      "loss": 0.9218,
+      "step": 4943
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6887765341586247,
+      "learning_rate": 4.733741786441916e-06,
+      "loss": 0.8702,
+      "step": 4944
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.666158809144583,
+      "learning_rate": 4.73200319969031e-06,
+      "loss": 0.9591,
+      "step": 4945
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.093306662233018,
+      "learning_rate": 4.730264645434238e-06,
+      "loss": 0.8523,
+      "step": 4946
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6553631036949914,
+      "learning_rate": 4.728526123884509e-06,
+      "loss": 0.9287,
+      "step": 4947
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7872925202747005,
+      "learning_rate": 4.7267876352519215e-06,
+      "loss": 0.9384,
+      "step": 4948
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.745842003773778,
+      "learning_rate": 4.725049179747276e-06,
+      "loss": 0.9495,
+      "step": 4949
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7608330213275287,
+      "learning_rate": 4.723310757581365e-06,
+      "loss": 0.8772,
+      "step": 4950
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8632210263611498,
+      "learning_rate": 4.721572368964979e-06,
+      "loss": 1.0656,
+      "step": 4951
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7111283778911603,
+      "learning_rate": 4.719834014108903e-06,
+      "loss": 0.9473,
+      "step": 4952
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.647769329340415,
+      "learning_rate": 4.718095693223919e-06,
+      "loss": 0.8684,
+      "step": 4953
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6686118973624922,
+      "learning_rate": 4.7163574065208034e-06,
+      "loss": 0.8148,
+      "step": 4954
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.717168280452624,
+      "learning_rate": 4.7146191542103305e-06,
+      "loss": 0.8644,
+      "step": 4955
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8039591716309638,
+      "learning_rate": 4.712880936503271e-06,
+      "loss": 0.9153,
+      "step": 4956
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8243455927199925,
+      "learning_rate": 4.7111427536103856e-06,
+      "loss": 0.9185,
+      "step": 4957
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8277846328534417,
+      "learning_rate": 4.709404605742438e-06,
+      "loss": 1.0335,
+      "step": 4958
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7584402968009591,
+      "learning_rate": 4.707666493110182e-06,
+      "loss": 0.9273,
+      "step": 4959
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.724181600053812,
+      "learning_rate": 4.7059284159243725e-06,
+      "loss": 0.8672,
+      "step": 4960
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6629145513539565,
+      "learning_rate": 4.704190374395757e-06,
+      "loss": 0.9043,
+      "step": 4961
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7094265380277436,
+      "learning_rate": 4.7024523687350775e-06,
+      "loss": 0.9478,
+      "step": 4962
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.805608639238089,
+      "learning_rate": 4.700714399153075e-06,
+      "loss": 0.8706,
+      "step": 4963
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7805025011859075,
+      "learning_rate": 4.698976465860484e-06,
+      "loss": 0.8466,
+      "step": 4964
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7025785117051342,
+      "learning_rate": 4.697238569068033e-06,
+      "loss": 0.9355,
+      "step": 4965
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6578604041334952,
+      "learning_rate": 4.695500708986451e-06,
+      "loss": 0.7994,
+      "step": 4966
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6642505181482334,
+      "learning_rate": 4.693762885826456e-06,
+      "loss": 0.9109,
+      "step": 4967
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8076434816373579,
+      "learning_rate": 4.692025099798767e-06,
+      "loss": 0.867,
+      "step": 4968
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8817693193863512,
+      "learning_rate": 4.690287351114097e-06,
+      "loss": 0.9833,
+      "step": 4969
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8551260413633194,
+      "learning_rate": 4.688549639983154e-06,
+      "loss": 1.0675,
+      "step": 4970
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7052222427364945,
+      "learning_rate": 4.686811966616639e-06,
+      "loss": 0.9821,
+      "step": 4971
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5530615187398884,
+      "learning_rate": 4.685074331225254e-06,
+      "loss": 0.8332,
+      "step": 4972
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8035401098222605,
+      "learning_rate": 4.683336734019692e-06,
+      "loss": 0.8522,
+      "step": 4973
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.781242263206731,
+      "learning_rate": 4.681599175210641e-06,
+      "loss": 0.9823,
+      "step": 4974
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6392703495582157,
+      "learning_rate": 4.67986165500879e-06,
+      "loss": 0.8032,
+      "step": 4975
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.859291988007837,
+      "learning_rate": 4.678124173624816e-06,
+      "loss": 0.8722,
+      "step": 4976
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9749289010487534,
+      "learning_rate": 4.676386731269398e-06,
+      "loss": 0.934,
+      "step": 4977
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6749928775437437,
+      "learning_rate": 4.674649328153202e-06,
+      "loss": 0.9057,
+      "step": 4978
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.735243114327548,
+      "learning_rate": 4.672911964486896e-06,
+      "loss": 0.8714,
+      "step": 4979
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8147422547627723,
+      "learning_rate": 4.671174640481143e-06,
+      "loss": 0.9172,
+      "step": 4980
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7732669645356542,
+      "learning_rate": 4.669437356346599e-06,
+      "loss": 0.9292,
+      "step": 4981
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6300231384242516,
+      "learning_rate": 4.6677001122939134e-06,
+      "loss": 0.8704,
+      "step": 4982
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.669247919210026,
+      "learning_rate": 4.665962908533736e-06,
+      "loss": 0.8896,
+      "step": 4983
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6589565461227995,
+      "learning_rate": 4.664225745276709e-06,
+      "loss": 0.9298,
+      "step": 4984
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6685161508745137,
+      "learning_rate": 4.662488622733466e-06,
+      "loss": 0.8821,
+      "step": 4985
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8721036708805519,
+      "learning_rate": 4.660751541114641e-06,
+      "loss": 1.028,
+      "step": 4986
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6353546693453442,
+      "learning_rate": 4.659014500630863e-06,
+      "loss": 0.8398,
+      "step": 4987
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.742223477508306,
+      "learning_rate": 4.657277501492751e-06,
+      "loss": 0.9114,
+      "step": 4988
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.730970515906384,
+      "learning_rate": 4.655540543910924e-06,
+      "loss": 0.8551,
+      "step": 4989
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8031535527938525,
+      "learning_rate": 4.653803628095993e-06,
+      "loss": 0.9488,
+      "step": 4990
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.698132700217416,
+      "learning_rate": 4.6520667542585655e-06,
+      "loss": 0.8638,
+      "step": 4991
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7247778728320484,
+      "learning_rate": 4.650329922609244e-06,
+      "loss": 0.8925,
+      "step": 4992
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7127755718926827,
+      "learning_rate": 4.648593133358625e-06,
+      "loss": 0.8779,
+      "step": 4993
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.743547321415513,
+      "learning_rate": 4.646856386717299e-06,
+      "loss": 0.8683,
+      "step": 4994
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6791253428177764,
+      "learning_rate": 4.645119682895855e-06,
+      "loss": 0.9073,
+      "step": 4995
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7796117744745876,
+      "learning_rate": 4.643383022104871e-06,
+      "loss": 0.8133,
+      "step": 4996
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7069230659059722,
+      "learning_rate": 4.641646404554927e-06,
+      "loss": 0.8602,
+      "step": 4997
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8301916717115407,
+      "learning_rate": 4.639909830456593e-06,
+      "loss": 0.9255,
+      "step": 4998
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7804799526353505,
+      "learning_rate": 4.638173300020431e-06,
+      "loss": 0.8825,
+      "step": 4999
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6784487326405346,
+      "learning_rate": 4.6364368134570046e-06,
+      "loss": 0.9191,
+      "step": 5000
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7723975731958774,
+      "learning_rate": 4.634700370976867e-06,
+      "loss": 0.9004,
+      "step": 5001
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7325358022086093,
+      "learning_rate": 4.63296397279057e-06,
+      "loss": 0.8276,
+      "step": 5002
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7925479346536586,
+      "learning_rate": 4.631227619108657e-06,
+      "loss": 0.9254,
+      "step": 5003
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7001621334201689,
+      "learning_rate": 4.629491310141666e-06,
+      "loss": 0.9171,
+      "step": 5004
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7739891625559263,
+      "learning_rate": 4.62775504610013e-06,
+      "loss": 0.9894,
+      "step": 5005
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6571003259366335,
+      "learning_rate": 4.626018827194579e-06,
+      "loss": 0.8741,
+      "step": 5006
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.525887234208154,
+      "learning_rate": 4.624282653635534e-06,
+      "loss": 0.8256,
+      "step": 5007
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7663074789739956,
+      "learning_rate": 4.622546525633512e-06,
+      "loss": 0.9049,
+      "step": 5008
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.805854410033661,
+      "learning_rate": 4.6208104433990285e-06,
+      "loss": 0.8538,
+      "step": 5009
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.711185162345477,
+      "learning_rate": 4.619074407142582e-06,
+      "loss": 0.9004,
+      "step": 5010
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6145315307261079,
+      "learning_rate": 4.617338417074679e-06,
+      "loss": 0.8928,
+      "step": 5011
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8348426907860208,
+      "learning_rate": 4.615602473405812e-06,
+      "loss": 1.0594,
+      "step": 5012
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8929071157458481,
+      "learning_rate": 4.6138665763464685e-06,
+      "loss": 0.9084,
+      "step": 5013
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6558015470080174,
+      "learning_rate": 4.612130726107135e-06,
+      "loss": 0.8653,
+      "step": 5014
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7837267933952008,
+      "learning_rate": 4.610394922898289e-06,
+      "loss": 1.0622,
+      "step": 5015
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7689384364313019,
+      "learning_rate": 4.6086591669303995e-06,
+      "loss": 0.8927,
+      "step": 5016
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7657955137912844,
+      "learning_rate": 4.6069234584139374e-06,
+      "loss": 0.9039,
+      "step": 5017
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7075077938714043,
+      "learning_rate": 4.60518779755936e-06,
+      "loss": 0.8896,
+      "step": 5018
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7203469676313992,
+      "learning_rate": 4.603452184577126e-06,
+      "loss": 0.8656,
+      "step": 5019
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8109647475423194,
+      "learning_rate": 4.6017166196776795e-06,
+      "loss": 0.9596,
+      "step": 5020
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7598116607320102,
+      "learning_rate": 4.599981103071465e-06,
+      "loss": 0.9201,
+      "step": 5021
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9637712383167458,
+      "learning_rate": 4.5982456349689214e-06,
+      "loss": 0.8877,
+      "step": 5022
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7302577737454343,
+      "learning_rate": 4.59651021558048e-06,
+      "loss": 0.8767,
+      "step": 5023
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7615645976647443,
+      "learning_rate": 4.594774845116565e-06,
+      "loss": 0.9147,
+      "step": 5024
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7234266576942547,
+      "learning_rate": 4.5930395237875986e-06,
+      "loss": 0.9438,
+      "step": 5025
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6561627694704253,
+      "learning_rate": 4.5913042518039925e-06,
+      "loss": 0.9181,
+      "step": 5026
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7235208764288112,
+      "learning_rate": 4.589569029376153e-06,
+      "loss": 0.9335,
+      "step": 5027
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7372581847311495,
+      "learning_rate": 4.587833856714486e-06,
+      "loss": 0.8573,
+      "step": 5028
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7861011274551326,
+      "learning_rate": 4.586098734029384e-06,
+      "loss": 0.8677,
+      "step": 5029
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.753058950019016,
+      "learning_rate": 4.5843636615312395e-06,
+      "loss": 0.8736,
+      "step": 5030
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6487596252175827,
+      "learning_rate": 4.582628639430432e-06,
+      "loss": 0.8226,
+      "step": 5031
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.926340413471015,
+      "learning_rate": 4.58089366793734e-06,
+      "loss": 0.9699,
+      "step": 5032
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9302532562829554,
+      "learning_rate": 4.579158747262337e-06,
+      "loss": 1.0428,
+      "step": 5033
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7067057246929194,
+      "learning_rate": 4.577423877615787e-06,
+      "loss": 0.8932,
+      "step": 5034
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.72422945801201,
+      "learning_rate": 4.575689059208048e-06,
+      "loss": 0.9247,
+      "step": 5035
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.669732234196792,
+      "learning_rate": 4.573954292249472e-06,
+      "loss": 0.9102,
+      "step": 5036
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6499808331322126,
+      "learning_rate": 4.572219576950408e-06,
+      "loss": 0.9287,
+      "step": 5037
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7047732945569647,
+      "learning_rate": 4.570484913521196e-06,
+      "loss": 0.9072,
+      "step": 5038
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7275377215426253,
+      "learning_rate": 4.568750302172168e-06,
+      "loss": 0.9277,
+      "step": 5039
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8483489524248907,
+      "learning_rate": 4.5670157431136546e-06,
+      "loss": 1.0403,
+      "step": 5040
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.777735679661856,
+      "learning_rate": 4.565281236555973e-06,
+      "loss": 0.8819,
+      "step": 5041
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.039740631661275,
+      "learning_rate": 4.56354678270944e-06,
+      "loss": 0.8272,
+      "step": 5042
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7523403278141951,
+      "learning_rate": 4.561812381784366e-06,
+      "loss": 0.9625,
+      "step": 5043
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8388913211727835,
+      "learning_rate": 4.560078033991049e-06,
+      "loss": 0.8953,
+      "step": 5044
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7789843053556418,
+      "learning_rate": 4.558343739539788e-06,
+      "loss": 0.9269,
+      "step": 5045
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7654709345697694,
+      "learning_rate": 4.556609498640872e-06,
+      "loss": 0.8727,
+      "step": 5046
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7818276966848035,
+      "learning_rate": 4.554875311504581e-06,
+      "loss": 0.8723,
+      "step": 5047
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.697932740364964,
+      "learning_rate": 4.553141178341195e-06,
+      "loss": 0.9863,
+      "step": 5048
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6824305633795558,
+      "learning_rate": 4.551407099360981e-06,
+      "loss": 0.9704,
+      "step": 5049
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7135854110353326,
+      "learning_rate": 4.549673074774203e-06,
+      "loss": 0.9283,
+      "step": 5050
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9700543804637186,
+      "learning_rate": 4.547939104791118e-06,
+      "loss": 1.086,
+      "step": 5051
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7852315824655227,
+      "learning_rate": 4.546205189621974e-06,
+      "loss": 0.9471,
+      "step": 5052
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.859643354853795,
+      "learning_rate": 4.544471329477015e-06,
+      "loss": 0.9284,
+      "step": 5053
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6878635022987745,
+      "learning_rate": 4.542737524566479e-06,
+      "loss": 0.8686,
+      "step": 5054
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8612253151187004,
+      "learning_rate": 4.541003775100592e-06,
+      "loss": 0.8968,
+      "step": 5055
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7984457437776935,
+      "learning_rate": 4.539270081289581e-06,
+      "loss": 0.8612,
+      "step": 5056
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7529287772734765,
+      "learning_rate": 4.537536443343662e-06,
+      "loss": 0.9172,
+      "step": 5057
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9105040137339048,
+      "learning_rate": 4.5358028614730424e-06,
+      "loss": 1.057,
+      "step": 5058
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8573565328816046,
+      "learning_rate": 4.534069335887927e-06,
+      "loss": 0.8773,
+      "step": 5059
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6647024113720141,
+      "learning_rate": 4.53233586679851e-06,
+      "loss": 0.8925,
+      "step": 5060
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6437603938106602,
+      "learning_rate": 4.530602454414983e-06,
+      "loss": 0.8836,
+      "step": 5061
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6790031076641545,
+      "learning_rate": 4.528869098947526e-06,
+      "loss": 0.9063,
+      "step": 5062
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7009224807102945,
+      "learning_rate": 4.527135800606314e-06,
+      "loss": 0.9241,
+      "step": 5063
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8088811832890863,
+      "learning_rate": 4.5254025596015175e-06,
+      "loss": 0.937,
+      "step": 5064
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.796524387875531,
+      "learning_rate": 4.523669376143296e-06,
+      "loss": 0.9108,
+      "step": 5065
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6868482056734595,
+      "learning_rate": 4.5219362504418055e-06,
+      "loss": 0.8425,
+      "step": 5066
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7925136584320065,
+      "learning_rate": 4.5202031827071915e-06,
+      "loss": 0.8644,
+      "step": 5067
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8086046454500984,
+      "learning_rate": 4.518470173149597e-06,
+      "loss": 0.8587,
+      "step": 5068
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7350504706157457,
+      "learning_rate": 4.5167372219791545e-06,
+      "loss": 0.8693,
+      "step": 5069
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7731682982512782,
+      "learning_rate": 4.515004329405988e-06,
+      "loss": 0.9511,
+      "step": 5070
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8003066529177185,
+      "learning_rate": 4.51327149564022e-06,
+      "loss": 0.9252,
+      "step": 5071
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7340718783537605,
+      "learning_rate": 4.511538720891963e-06,
+      "loss": 0.8949,
+      "step": 5072
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8489959949332184,
+      "learning_rate": 4.509806005371317e-06,
+      "loss": 1.0348,
+      "step": 5073
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.704508161067822,
+      "learning_rate": 4.508073349288384e-06,
+      "loss": 0.9494,
+      "step": 5074
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6918237325462866,
+      "learning_rate": 4.506340752853253e-06,
+      "loss": 0.9074,
+      "step": 5075
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8087203927931619,
+      "learning_rate": 4.5046082162760076e-06,
+      "loss": 0.9719,
+      "step": 5076
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9106539094256585,
+      "learning_rate": 4.502875739766725e-06,
+      "loss": 0.9012,
+      "step": 5077
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.691077191373585,
+      "learning_rate": 4.50114332353547e-06,
+      "loss": 0.9144,
+      "step": 5078
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6532460182826971,
+      "learning_rate": 4.499410967792309e-06,
+      "loss": 0.8378,
+      "step": 5079
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7335520151124657,
+      "learning_rate": 4.497678672747294e-06,
+      "loss": 0.884,
+      "step": 5080
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8326739353913888,
+      "learning_rate": 4.49594643861047e-06,
+      "loss": 0.9138,
+      "step": 5081
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.641661747628651,
+      "learning_rate": 4.49421426559188e-06,
+      "loss": 0.8142,
+      "step": 5082
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.801576561749422,
+      "learning_rate": 4.492482153901554e-06,
+      "loss": 0.9036,
+      "step": 5083
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8114245631584247,
+      "learning_rate": 4.490750103749516e-06,
+      "loss": 0.9086,
+      "step": 5084
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.79377208638708,
+      "learning_rate": 4.489018115345784e-06,
+      "loss": 0.8882,
+      "step": 5085
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8734219073654627,
+      "learning_rate": 4.487286188900365e-06,
+      "loss": 1.0784,
+      "step": 5086
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7494986004574615,
+      "learning_rate": 4.485554324623266e-06,
+      "loss": 0.947,
+      "step": 5087
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8168018163742186,
+      "learning_rate": 4.483822522724477e-06,
+      "loss": 0.8659,
+      "step": 5088
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8926026892436059,
+      "learning_rate": 4.4820907834139865e-06,
+      "loss": 0.9076,
+      "step": 5089
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.856481874122231,
+      "learning_rate": 4.480359106901775e-06,
+      "loss": 1.0331,
+      "step": 5090
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6718827012367594,
+      "learning_rate": 4.478627493397813e-06,
+      "loss": 0.8972,
+      "step": 5091
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7502436528718952,
+      "learning_rate": 4.476895943112064e-06,
+      "loss": 0.8984,
+      "step": 5092
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.726810174230518,
+      "learning_rate": 4.475164456254488e-06,
+      "loss": 0.8437,
+      "step": 5093
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8159787995132556,
+      "learning_rate": 4.473433033035028e-06,
+      "loss": 1.0656,
+      "step": 5094
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.790334279965691,
+      "learning_rate": 4.4717016736636295e-06,
+      "loss": 0.9884,
+      "step": 5095
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7655706367790471,
+      "learning_rate": 4.469970378350225e-06,
+      "loss": 0.9905,
+      "step": 5096
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7998595206612015,
+      "learning_rate": 4.468239147304737e-06,
+      "loss": 1.0056,
+      "step": 5097
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6842333121518265,
+      "learning_rate": 4.466507980737087e-06,
+      "loss": 0.8824,
+      "step": 5098
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6970587530288117,
+      "learning_rate": 4.464776878857184e-06,
+      "loss": 0.9723,
+      "step": 5099
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6795141726325977,
+      "learning_rate": 4.46304584187493e-06,
+      "loss": 0.8496,
+      "step": 5100
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.1365707729525605,
+      "learning_rate": 4.461314870000218e-06,
+      "loss": 0.9222,
+      "step": 5101
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7199809337766478,
+      "learning_rate": 4.459583963442935e-06,
+      "loss": 0.8673,
+      "step": 5102
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7547711751982245,
+      "learning_rate": 4.45785312241296e-06,
+      "loss": 0.9287,
+      "step": 5103
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6895454250056332,
+      "learning_rate": 4.456122347120165e-06,
+      "loss": 0.8547,
+      "step": 5104
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6872180361089455,
+      "learning_rate": 4.454391637774408e-06,
+      "loss": 0.8644,
+      "step": 5105
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6061034337759328,
+      "learning_rate": 4.452660994585546e-06,
+      "loss": 0.8151,
+      "step": 5106
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.615189873116892,
+      "learning_rate": 4.450930417763425e-06,
+      "loss": 0.8286,
+      "step": 5107
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7012834249162005,
+      "learning_rate": 4.449199907517885e-06,
+      "loss": 0.9111,
+      "step": 5108
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7212794677857668,
+      "learning_rate": 4.447469464058753e-06,
+      "loss": 0.8681,
+      "step": 5109
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7867010508905863,
+      "learning_rate": 4.445739087595855e-06,
+      "loss": 0.8291,
+      "step": 5110
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8086509906893826,
+      "learning_rate": 4.444008778339003e-06,
+      "loss": 1.0169,
+      "step": 5111
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7267090826304217,
+      "learning_rate": 4.4422785364980036e-06,
+      "loss": 1.0024,
+      "step": 5112
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7105640728394493,
+      "learning_rate": 4.440548362282655e-06,
+      "loss": 0.9026,
+      "step": 5113
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8023257963659827,
+      "learning_rate": 4.438818255902746e-06,
+      "loss": 0.82,
+      "step": 5114
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7501616753182352,
+      "learning_rate": 4.437088217568059e-06,
+      "loss": 0.9549,
+      "step": 5115
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.794981692531099,
+      "learning_rate": 4.435358247488365e-06,
+      "loss": 0.8704,
+      "step": 5116
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9421394918310673,
+      "learning_rate": 4.433628345873429e-06,
+      "loss": 0.9005,
+      "step": 5117
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7184590208842059,
+      "learning_rate": 4.431898512933011e-06,
+      "loss": 0.8938,
+      "step": 5118
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.766872714664663,
+      "learning_rate": 4.430168748876855e-06,
+      "loss": 0.9176,
+      "step": 5119
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7305551056954995,
+      "learning_rate": 4.428439053914702e-06,
+      "loss": 0.9304,
+      "step": 5120
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7510315489012667,
+      "learning_rate": 4.426709428256287e-06,
+      "loss": 0.9097,
+      "step": 5121
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6834922885973587,
+      "learning_rate": 4.424979872111329e-06,
+      "loss": 0.8831,
+      "step": 5122
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.717667872654713,
+      "learning_rate": 4.423250385689543e-06,
+      "loss": 0.9342,
+      "step": 5123
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6060837868231885,
+      "learning_rate": 4.421520969200637e-06,
+      "loss": 0.8526,
+      "step": 5124
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8606167282444606,
+      "learning_rate": 4.419791622854309e-06,
+      "loss": 0.8937,
+      "step": 5125
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.636562296784056,
+      "learning_rate": 4.4180623468602455e-06,
+      "loss": 0.8981,
+      "step": 5126
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9406668728685015,
+      "learning_rate": 4.416333141428129e-06,
+      "loss": 0.9289,
+      "step": 5127
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8103236328544385,
+      "learning_rate": 4.414604006767631e-06,
+      "loss": 0.9228,
+      "step": 5128
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7720775706426872,
+      "learning_rate": 4.412874943088416e-06,
+      "loss": 0.866,
+      "step": 5129
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5842441371802278,
+      "learning_rate": 4.411145950600137e-06,
+      "loss": 0.8817,
+      "step": 5130
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.743180288580314,
+      "learning_rate": 4.4094170295124426e-06,
+      "loss": 0.8885,
+      "step": 5131
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6612511777810255,
+      "learning_rate": 4.407688180034969e-06,
+      "loss": 0.9288,
+      "step": 5132
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.003867326308389,
+      "learning_rate": 4.405959402377345e-06,
+      "loss": 0.8728,
+      "step": 5133
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7689019387066054,
+      "learning_rate": 4.4042306967491935e-06,
+      "loss": 0.8969,
+      "step": 5134
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6747953164820495,
+      "learning_rate": 4.402502063360121e-06,
+      "loss": 0.9634,
+      "step": 5135
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8065807282744992,
+      "learning_rate": 4.400773502419738e-06,
+      "loss": 0.8834,
+      "step": 5136
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8643115266123131,
+      "learning_rate": 4.39904501413763e-06,
+      "loss": 0.9427,
+      "step": 5137
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7071078207940897,
+      "learning_rate": 4.397316598723385e-06,
+      "loss": 0.9812,
+      "step": 5138
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8716453798864132,
+      "learning_rate": 4.395588256386583e-06,
+      "loss": 1.0674,
+      "step": 5139
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6851216270952745,
+      "learning_rate": 4.393859987336786e-06,
+      "loss": 0.8752,
+      "step": 5140
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.588699914484462,
+      "learning_rate": 4.392131791783557e-06,
+      "loss": 0.9368,
+      "step": 5141
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7913208980145345,
+      "learning_rate": 4.390403669936443e-06,
+      "loss": 0.8956,
+      "step": 5142
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7413994973170699,
+      "learning_rate": 4.3886756220049855e-06,
+      "loss": 0.9145,
+      "step": 5143
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7590427397202708,
+      "learning_rate": 4.386947648198717e-06,
+      "loss": 0.9279,
+      "step": 5144
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6774553615689793,
+      "learning_rate": 4.38521974872716e-06,
+      "loss": 0.8735,
+      "step": 5145
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7683266627589713,
+      "learning_rate": 4.383491923799828e-06,
+      "loss": 0.9771,
+      "step": 5146
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7835347423256822,
+      "learning_rate": 4.3817641736262255e-06,
+      "loss": 0.943,
+      "step": 5147
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7335684566555338,
+      "learning_rate": 4.380036498415847e-06,
+      "loss": 0.9028,
+      "step": 5148
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7623361422228212,
+      "learning_rate": 4.3783088983781815e-06,
+      "loss": 0.9006,
+      "step": 5149
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.667910454355491,
+      "learning_rate": 4.376581373722705e-06,
+      "loss": 0.9683,
+      "step": 5150
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6890897833543494,
+      "learning_rate": 4.374853924658886e-06,
+      "loss": 0.9259,
+      "step": 5151
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8324155765875638,
+      "learning_rate": 4.373126551396184e-06,
+      "loss": 0.9931,
+      "step": 5152
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7760150839671658,
+      "learning_rate": 4.37139925414405e-06,
+      "loss": 0.9249,
+      "step": 5153
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7687630531632523,
+      "learning_rate": 4.369672033111921e-06,
+      "loss": 0.8591,
+      "step": 5154
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8361678030774675,
+      "learning_rate": 4.367944888509233e-06,
+      "loss": 1.0678,
+      "step": 5155
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7425093412631631,
+      "learning_rate": 4.366217820545407e-06,
+      "loss": 0.9046,
+      "step": 5156
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7248029458744072,
+      "learning_rate": 4.364490829429855e-06,
+      "loss": 0.9018,
+      "step": 5157
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7307422116002074,
+      "learning_rate": 4.36276391537198e-06,
+      "loss": 0.8693,
+      "step": 5158
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8169200886809083,
+      "learning_rate": 4.361037078581176e-06,
+      "loss": 0.9446,
+      "step": 5159
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6424990056199789,
+      "learning_rate": 4.35931031926683e-06,
+      "loss": 0.8716,
+      "step": 5160
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7146585119890985,
+      "learning_rate": 4.3575836376383175e-06,
+      "loss": 0.9342,
+      "step": 5161
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8211842445113322,
+      "learning_rate": 4.3558570339050034e-06,
+      "loss": 0.8531,
+      "step": 5162
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7643228658517895,
+      "learning_rate": 4.354130508276243e-06,
+      "loss": 0.8839,
+      "step": 5163
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7203456285505647,
+      "learning_rate": 4.352404060961387e-06,
+      "loss": 0.9007,
+      "step": 5164
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8287165920418696,
+      "learning_rate": 4.3506776921697705e-06,
+      "loss": 1.0459,
+      "step": 5165
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7961573720899304,
+      "learning_rate": 4.3489514021107215e-06,
+      "loss": 0.9529,
+      "step": 5166
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7896551508395213,
+      "learning_rate": 4.3472251909935635e-06,
+      "loss": 0.9251,
+      "step": 5167
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.729208088159677,
+      "learning_rate": 4.3454990590275966e-06,
+      "loss": 0.8767,
+      "step": 5168
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7266909598363769,
+      "learning_rate": 4.343773006422128e-06,
+      "loss": 0.8664,
+      "step": 5169
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.807781761334988,
+      "learning_rate": 4.342047033386444e-06,
+      "loss": 1.0587,
+      "step": 5170
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.670830287439024,
+      "learning_rate": 4.3403211401298245e-06,
+      "loss": 0.928,
+      "step": 5171
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8339197640703069,
+      "learning_rate": 4.338595326861542e-06,
+      "loss": 0.9487,
+      "step": 5172
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8627351602846731,
+      "learning_rate": 4.336869593790857e-06,
+      "loss": 0.9384,
+      "step": 5173
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8474917374652085,
+      "learning_rate": 4.335143941127018e-06,
+      "loss": 0.905,
+      "step": 5174
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8206866551215968,
+      "learning_rate": 4.3334183690792695e-06,
+      "loss": 0.8765,
+      "step": 5175
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7065313109958729,
+      "learning_rate": 4.3316928778568414e-06,
+      "loss": 0.9527,
+      "step": 5176
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6633947106582345,
+      "learning_rate": 4.329967467668955e-06,
+      "loss": 0.8799,
+      "step": 5177
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7616176335129183,
+      "learning_rate": 4.328242138724827e-06,
+      "loss": 0.9171,
+      "step": 5178
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7436189515475435,
+      "learning_rate": 4.326516891233652e-06,
+      "loss": 0.8841,
+      "step": 5179
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8051649463451014,
+      "learning_rate": 4.324791725404627e-06,
+      "loss": 0.8725,
+      "step": 5180
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7442729286939758,
+      "learning_rate": 4.3230666414469325e-06,
+      "loss": 0.925,
+      "step": 5181
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8548427192518901,
+      "learning_rate": 4.321341639569741e-06,
+      "loss": 1.0145,
+      "step": 5182
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7691357110891661,
+      "learning_rate": 4.3196167199822155e-06,
+      "loss": 0.8764,
+      "step": 5183
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.557931660663016,
+      "learning_rate": 4.317891882893509e-06,
+      "loss": 0.8624,
+      "step": 5184
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7411703071822364,
+      "learning_rate": 4.316167128512763e-06,
+      "loss": 0.8814,
+      "step": 5185
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7381190135608173,
+      "learning_rate": 4.314442457049111e-06,
+      "loss": 0.8998,
+      "step": 5186
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8231275008124754,
+      "learning_rate": 4.312717868711674e-06,
+      "loss": 0.8584,
+      "step": 5187
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8584119714062202,
+      "learning_rate": 4.310993363709563e-06,
+      "loss": 0.9113,
+      "step": 5188
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8427930562903077,
+      "learning_rate": 4.309268942251887e-06,
+      "loss": 0.8443,
+      "step": 5189
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7063623795224583,
+      "learning_rate": 4.307544604547729e-06,
+      "loss": 0.9114,
+      "step": 5190
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.681711518505076,
+      "learning_rate": 4.305820350806176e-06,
+      "loss": 0.82,
+      "step": 5191
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8787990996744685,
+      "learning_rate": 4.304096181236299e-06,
+      "loss": 1.0272,
+      "step": 5192
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8592710712841123,
+      "learning_rate": 4.302372096047157e-06,
+      "loss": 0.8113,
+      "step": 5193
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.834612918126684,
+      "learning_rate": 4.300648095447806e-06,
+      "loss": 0.9349,
+      "step": 5194
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8008788753913036,
+      "learning_rate": 4.298924179647283e-06,
+      "loss": 0.9652,
+      "step": 5195
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7702943002910667,
+      "learning_rate": 4.297200348854621e-06,
+      "loss": 0.9184,
+      "step": 5196
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7864370409296657,
+      "learning_rate": 4.295476603278839e-06,
+      "loss": 0.9319,
+      "step": 5197
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7360170364285181,
+      "learning_rate": 4.293752943128948e-06,
+      "loss": 0.7534,
+      "step": 5198
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.761449119024601,
+      "learning_rate": 4.2920293686139485e-06,
+      "loss": 0.8241,
+      "step": 5199
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.787800964764351,
+      "learning_rate": 4.290305879942828e-06,
+      "loss": 0.8716,
+      "step": 5200
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.768048893719742,
+      "learning_rate": 4.288582477324567e-06,
+      "loss": 0.8628,
+      "step": 5201
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7747744929116274,
+      "learning_rate": 4.286859160968131e-06,
+      "loss": 0.961,
+      "step": 5202
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7490486130252179,
+      "learning_rate": 4.285135931082481e-06,
+      "loss": 0.8395,
+      "step": 5203
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6255006554610352,
+      "learning_rate": 4.283412787876565e-06,
+      "loss": 0.8423,
+      "step": 5204
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7152960916147142,
+      "learning_rate": 4.281689731559318e-06,
+      "loss": 0.9065,
+      "step": 5205
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5900430902540859,
+      "learning_rate": 4.279966762339668e-06,
+      "loss": 0.8583,
+      "step": 5206
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.793893928445111,
+      "learning_rate": 4.27824388042653e-06,
+      "loss": 1.045,
+      "step": 5207
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7378023974690229,
+      "learning_rate": 4.276521086028809e-06,
+      "loss": 0.8462,
+      "step": 5208
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.713279240195725,
+      "learning_rate": 4.274798379355403e-06,
+      "loss": 0.9101,
+      "step": 5209
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6802263666887662,
+      "learning_rate": 4.273075760615193e-06,
+      "loss": 0.8951,
+      "step": 5210
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8694385822561155,
+      "learning_rate": 4.271353230017053e-06,
+      "loss": 0.9101,
+      "step": 5211
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8160688246512284,
+      "learning_rate": 4.269630787769845e-06,
+      "loss": 0.8517,
+      "step": 5212
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.760416692276652,
+      "learning_rate": 4.2679084340824215e-06,
+      "loss": 0.9636,
+      "step": 5213
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7697291866185991,
+      "learning_rate": 4.266186169163625e-06,
+      "loss": 0.8994,
+      "step": 5214
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5815943711868883,
+      "learning_rate": 4.264463993222286e-06,
+      "loss": 0.8798,
+      "step": 5215
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8059696951057306,
+      "learning_rate": 4.262741906467221e-06,
+      "loss": 1.0339,
+      "step": 5216
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7588830107980258,
+      "learning_rate": 4.261019909107243e-06,
+      "loss": 0.9098,
+      "step": 5217
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6149628206228688,
+      "learning_rate": 4.25929800135115e-06,
+      "loss": 0.8407,
+      "step": 5218
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.679149652430536,
+      "learning_rate": 4.257576183407726e-06,
+      "loss": 0.8474,
+      "step": 5219
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.742732141119367,
+      "learning_rate": 4.255854455485753e-06,
+      "loss": 0.8758,
+      "step": 5220
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7666383634761624,
+      "learning_rate": 4.254132817793989e-06,
+      "loss": 0.9133,
+      "step": 5221
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7756314083981366,
+      "learning_rate": 4.252411270541194e-06,
+      "loss": 0.9224,
+      "step": 5222
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5902194035921124,
+      "learning_rate": 4.250689813936109e-06,
+      "loss": 0.9309,
+      "step": 5223
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6980404182673787,
+      "learning_rate": 4.248968448187466e-06,
+      "loss": 0.9574,
+      "step": 5224
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7621367950409685,
+      "learning_rate": 4.247247173503989e-06,
+      "loss": 0.9441,
+      "step": 5225
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8136494441005555,
+      "learning_rate": 4.245525990094388e-06,
+      "loss": 0.9271,
+      "step": 5226
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.672629940935499,
+      "learning_rate": 4.2438048981673615e-06,
+      "loss": 0.8321,
+      "step": 5227
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7013877565767463,
+      "learning_rate": 4.242083897931597e-06,
+      "loss": 0.9009,
+      "step": 5228
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6390775331908634,
+      "learning_rate": 4.240362989595775e-06,
+      "loss": 0.7961,
+      "step": 5229
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8584400988238313,
+      "learning_rate": 4.2386421733685575e-06,
+      "loss": 1.0467,
+      "step": 5230
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7093473916504491,
+      "learning_rate": 4.236921449458603e-06,
+      "loss": 0.9069,
+      "step": 5231
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8003714341785333,
+      "learning_rate": 4.235200818074553e-06,
+      "loss": 0.8582,
+      "step": 5232
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.708074517742616,
+      "learning_rate": 4.233480279425039e-06,
+      "loss": 0.9642,
+      "step": 5233
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6392092686386484,
+      "learning_rate": 4.231759833718685e-06,
+      "loss": 0.8667,
+      "step": 5234
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.893148663871309,
+      "learning_rate": 4.2300394811641e-06,
+      "loss": 0.9856,
+      "step": 5235
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6562070741872026,
+      "learning_rate": 4.22831922196988e-06,
+      "loss": 0.9097,
+      "step": 5236
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6410693515599672,
+      "learning_rate": 4.2265990563446165e-06,
+      "loss": 0.829,
+      "step": 5237
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8069994993695149,
+      "learning_rate": 4.224878984496884e-06,
+      "loss": 0.9431,
+      "step": 5238
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6388830579051825,
+      "learning_rate": 4.223159006635245e-06,
+      "loss": 0.8409,
+      "step": 5239
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7438704241435903,
+      "learning_rate": 4.221439122968256e-06,
+      "loss": 0.8999,
+      "step": 5240
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.682974406713379,
+      "learning_rate": 4.219719333704459e-06,
+      "loss": 0.8671,
+      "step": 5241
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6509525831181275,
+      "learning_rate": 4.217999639052382e-06,
+      "loss": 0.9191,
+      "step": 5242
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7075361037311823,
+      "learning_rate": 4.216280039220544e-06,
+      "loss": 0.947,
+      "step": 5243
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7635450525790939,
+      "learning_rate": 4.214560534417453e-06,
+      "loss": 0.8893,
+      "step": 5244
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8270582442764346,
+      "learning_rate": 4.2128411248516055e-06,
+      "loss": 1.0316,
+      "step": 5245
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7337356955959509,
+      "learning_rate": 4.211121810731485e-06,
+      "loss": 0.9692,
+      "step": 5246
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7301010623673383,
+      "learning_rate": 4.209402592265564e-06,
+      "loss": 0.8359,
+      "step": 5247
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7427658572906044,
+      "learning_rate": 4.207683469662306e-06,
+      "loss": 0.8248,
+      "step": 5248
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7475329097881946,
+      "learning_rate": 4.2059644431301575e-06,
+      "loss": 0.8939,
+      "step": 5249
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8781578509375443,
+      "learning_rate": 4.204245512877557e-06,
+      "loss": 0.9115,
+      "step": 5250
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.706616050941255,
+      "learning_rate": 4.202526679112932e-06,
+      "loss": 0.9613,
+      "step": 5251
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 6.825120608831123,
+      "learning_rate": 4.200807942044699e-06,
+      "loss": 1.0615,
+      "step": 5252
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8224766992448589,
+      "learning_rate": 4.199089301881256e-06,
+      "loss": 0.9045,
+      "step": 5253
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.830492760241334,
+      "learning_rate": 4.197370758830998e-06,
+      "loss": 1.0746,
+      "step": 5254
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6651264855953978,
+      "learning_rate": 4.1956523131023e-06,
+      "loss": 0.9037,
+      "step": 5255
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7690287648234613,
+      "learning_rate": 4.193933964903532e-06,
+      "loss": 1.0082,
+      "step": 5256
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7066958286915905,
+      "learning_rate": 4.192215714443052e-06,
+      "loss": 0.8032,
+      "step": 5257
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7812540058642454,
+      "learning_rate": 4.1904975619292e-06,
+      "loss": 1.083,
+      "step": 5258
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6889118183485556,
+      "learning_rate": 4.18877950757031e-06,
+      "loss": 0.9161,
+      "step": 5259
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7413941419707901,
+      "learning_rate": 4.187061551574701e-06,
+      "loss": 0.8918,
+      "step": 5260
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.71968704328864,
+      "learning_rate": 4.185343694150682e-06,
+      "loss": 0.912,
+      "step": 5261
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7374996491477825,
+      "learning_rate": 4.1836259355065475e-06,
+      "loss": 0.8983,
+      "step": 5262
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.68618223079258,
+      "learning_rate": 4.181908275850586e-06,
+      "loss": 0.8588,
+      "step": 5263
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.742417569120698,
+      "learning_rate": 4.180190715391063e-06,
+      "loss": 0.8876,
+      "step": 5264
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7940490575975223,
+      "learning_rate": 4.178473254336243e-06,
+      "loss": 0.9172,
+      "step": 5265
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7589492760254664,
+      "learning_rate": 4.176755892894374e-06,
+      "loss": 0.9245,
+      "step": 5266
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7816306368295989,
+      "learning_rate": 4.175038631273689e-06,
+      "loss": 0.8826,
+      "step": 5267
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7210342204333662,
+      "learning_rate": 4.173321469682416e-06,
+      "loss": 0.8767,
+      "step": 5268
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7770498719753853,
+      "learning_rate": 4.171604408328765e-06,
+      "loss": 0.8854,
+      "step": 5269
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6502633300170177,
+      "learning_rate": 4.1698874474209326e-06,
+      "loss": 0.8413,
+      "step": 5270
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.774845626693493,
+      "learning_rate": 4.1681705871671114e-06,
+      "loss": 0.9341,
+      "step": 5271
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6143898512427932,
+      "learning_rate": 4.166453827775474e-06,
+      "loss": 0.8462,
+      "step": 5272
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.639064417317491,
+      "learning_rate": 4.164737169454185e-06,
+      "loss": 0.9215,
+      "step": 5273
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7448598031634466,
+      "learning_rate": 4.1630206124113925e-06,
+      "loss": 0.9223,
+      "step": 5274
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6707938299038103,
+      "learning_rate": 4.161304156855235e-06,
+      "loss": 0.8026,
+      "step": 5275
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.809860532579585,
+      "learning_rate": 4.159587802993842e-06,
+      "loss": 1.0788,
+      "step": 5276
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7048574201050153,
+      "learning_rate": 4.157871551035324e-06,
+      "loss": 0.837,
+      "step": 5277
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.612646737377047,
+      "learning_rate": 4.156155401187783e-06,
+      "loss": 0.8403,
+      "step": 5278
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7293804228530885,
+      "learning_rate": 4.15443935365931e-06,
+      "loss": 1.0311,
+      "step": 5279
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6142847174363417,
+      "learning_rate": 4.1527234086579806e-06,
+      "loss": 0.8549,
+      "step": 5280
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6839282269428308,
+      "learning_rate": 4.151007566391857e-06,
+      "loss": 0.9223,
+      "step": 5281
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7869393887568636,
+      "learning_rate": 4.149291827068995e-06,
+      "loss": 0.8674,
+      "step": 5282
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6733283506405559,
+      "learning_rate": 4.1475761908974315e-06,
+      "loss": 0.906,
+      "step": 5283
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7413237377571351,
+      "learning_rate": 4.145860658085194e-06,
+      "loss": 0.8485,
+      "step": 5284
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6651980843571728,
+      "learning_rate": 4.1441452288402965e-06,
+      "loss": 0.9071,
+      "step": 5285
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8281197660592983,
+      "learning_rate": 4.142429903370739e-06,
+      "loss": 0.8432,
+      "step": 5286
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8828844721689901,
+      "learning_rate": 4.140714681884513e-06,
+      "loss": 1.0507,
+      "step": 5287
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6962764278338047,
+      "learning_rate": 4.1389995645895944e-06,
+      "loss": 0.8533,
+      "step": 5288
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6363315630697153,
+      "learning_rate": 4.137284551693946e-06,
+      "loss": 0.8752,
+      "step": 5289
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.2826618340300784,
+      "learning_rate": 4.13556964340552e-06,
+      "loss": 0.8596,
+      "step": 5290
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5801687066033614,
+      "learning_rate": 4.133854839932257e-06,
+      "loss": 0.8778,
+      "step": 5291
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7752846244560878,
+      "learning_rate": 4.132140141482079e-06,
+      "loss": 0.9377,
+      "step": 5292
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.856282286775335,
+      "learning_rate": 4.1304255482629e-06,
+      "loss": 0.9673,
+      "step": 5293
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7463158389264535,
+      "learning_rate": 4.128711060482624e-06,
+      "loss": 0.8253,
+      "step": 5294
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6896901097649875,
+      "learning_rate": 4.126996678349133e-06,
+      "loss": 0.9309,
+      "step": 5295
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7388497352295709,
+      "learning_rate": 4.125282402070307e-06,
+      "loss": 0.8577,
+      "step": 5296
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.779369865979201,
+      "learning_rate": 4.1235682318540035e-06,
+      "loss": 0.9778,
+      "step": 5297
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7143787873112328,
+      "learning_rate": 4.121854167908072e-06,
+      "loss": 0.8964,
+      "step": 5298
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8205646997135942,
+      "learning_rate": 4.120140210440352e-06,
+      "loss": 0.8473,
+      "step": 5299
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8934999377280766,
+      "learning_rate": 4.1184263596586645e-06,
+      "loss": 0.8857,
+      "step": 5300
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7875284365607609,
+      "learning_rate": 4.11671261577082e-06,
+      "loss": 0.9344,
+      "step": 5301
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7845150020833727,
+      "learning_rate": 4.114998978984616e-06,
+      "loss": 0.9505,
+      "step": 5302
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9029426742163396,
+      "learning_rate": 4.113285449507837e-06,
+      "loss": 0.9654,
+      "step": 5303
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6917276209738963,
+      "learning_rate": 4.111572027548253e-06,
+      "loss": 0.9363,
+      "step": 5304
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6825947045669538,
+      "learning_rate": 4.109858713313628e-06,
+      "loss": 0.9245,
+      "step": 5305
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8093897149213569,
+      "learning_rate": 4.108145507011698e-06,
+      "loss": 1.0505,
+      "step": 5306
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8126412203451248,
+      "learning_rate": 4.1064324088502026e-06,
+      "loss": 0.8707,
+      "step": 5307
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.859055294760676,
+      "learning_rate": 4.104719419036857e-06,
+      "loss": 0.9273,
+      "step": 5308
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6781502075804846,
+      "learning_rate": 4.1030065377793675e-06,
+      "loss": 0.8983,
+      "step": 5309
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.583928067420168,
+      "learning_rate": 4.101293765285429e-06,
+      "loss": 0.8204,
+      "step": 5310
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8910080153479412,
+      "learning_rate": 4.09958110176272e-06,
+      "loss": 0.945,
+      "step": 5311
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8261901627628784,
+      "learning_rate": 4.097868547418905e-06,
+      "loss": 1.0557,
+      "step": 5312
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6867258668078684,
+      "learning_rate": 4.09615610246164e-06,
+      "loss": 0.9062,
+      "step": 5313
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.811598382882989,
+      "learning_rate": 4.0944437670985635e-06,
+      "loss": 0.9409,
+      "step": 5314
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6183886284606244,
+      "learning_rate": 4.0927315415373016e-06,
+      "loss": 0.8833,
+      "step": 5315
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6883529564181532,
+      "learning_rate": 4.091019425985471e-06,
+      "loss": 0.8733,
+      "step": 5316
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7662957969264452,
+      "learning_rate": 4.089307420650665e-06,
+      "loss": 0.9222,
+      "step": 5317
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6543661592522836,
+      "learning_rate": 4.087595525740476e-06,
+      "loss": 0.8671,
+      "step": 5318
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.67514759014006,
+      "learning_rate": 4.085883741462474e-06,
+      "loss": 0.8294,
+      "step": 5319
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6793032705514526,
+      "learning_rate": 4.084172068024219e-06,
+      "loss": 0.9558,
+      "step": 5320
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7903096675862071,
+      "learning_rate": 4.082460505633259e-06,
+      "loss": 0.8494,
+      "step": 5321
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8793282370963615,
+      "learning_rate": 4.080749054497127e-06,
+      "loss": 0.9165,
+      "step": 5322
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6937905572250167,
+      "learning_rate": 4.079037714823341e-06,
+      "loss": 0.8642,
+      "step": 5323
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.830906912396178,
+      "learning_rate": 4.077326486819405e-06,
+      "loss": 0.8994,
+      "step": 5324
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6791987304762914,
+      "learning_rate": 4.075615370692815e-06,
+      "loss": 0.8562,
+      "step": 5325
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8116209696419372,
+      "learning_rate": 4.073904366651049e-06,
+      "loss": 0.8868,
+      "step": 5326
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8543754156506511,
+      "learning_rate": 4.072193474901571e-06,
+      "loss": 1.033,
+      "step": 5327
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7895339809347135,
+      "learning_rate": 4.070482695651832e-06,
+      "loss": 0.8826,
+      "step": 5328
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7299747082524215,
+      "learning_rate": 4.068772029109271e-06,
+      "loss": 0.832,
+      "step": 5329
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8191728701524195,
+      "learning_rate": 4.067061475481311e-06,
+      "loss": 0.8756,
+      "step": 5330
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7044544067464777,
+      "learning_rate": 4.0653510349753645e-06,
+      "loss": 0.9195,
+      "step": 5331
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6615041519299651,
+      "learning_rate": 4.063640707798826e-06,
+      "loss": 0.9459,
+      "step": 5332
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6608012764746711,
+      "learning_rate": 4.061930494159081e-06,
+      "loss": 0.8293,
+      "step": 5333
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8209221601213084,
+      "learning_rate": 4.060220394263497e-06,
+      "loss": 0.9454,
+      "step": 5334
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.774325591506145,
+      "learning_rate": 4.0585104083194295e-06,
+      "loss": 0.9125,
+      "step": 5335
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7477163642703712,
+      "learning_rate": 4.056800536534222e-06,
+      "loss": 0.9657,
+      "step": 5336
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.775826408680486,
+      "learning_rate": 4.0550907791152025e-06,
+      "loss": 0.8844,
+      "step": 5337
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6841843407390913,
+      "learning_rate": 4.053381136269682e-06,
+      "loss": 0.8409,
+      "step": 5338
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8800486690773819,
+      "learning_rate": 4.0516716082049625e-06,
+      "loss": 0.9136,
+      "step": 5339
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.983303803269195,
+      "learning_rate": 4.049962195128328e-06,
+      "loss": 0.9027,
+      "step": 5340
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8097911053361404,
+      "learning_rate": 4.0482528972470545e-06,
+      "loss": 0.8761,
+      "step": 5341
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8364976953788503,
+      "learning_rate": 4.046543714768399e-06,
+      "loss": 0.9019,
+      "step": 5342
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8271791686113197,
+      "learning_rate": 4.044834647899603e-06,
+      "loss": 0.8965,
+      "step": 5343
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8543084115340847,
+      "learning_rate": 4.043125696847901e-06,
+      "loss": 1.0339,
+      "step": 5344
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7247376147945515,
+      "learning_rate": 4.041416861820506e-06,
+      "loss": 0.878,
+      "step": 5345
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7929528337061738,
+      "learning_rate": 4.039708143024621e-06,
+      "loss": 0.9295,
+      "step": 5346
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7207452756492931,
+      "learning_rate": 4.037999540667436e-06,
+      "loss": 0.8749,
+      "step": 5347
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8159327383061363,
+      "learning_rate": 4.036291054956122e-06,
+      "loss": 1.0589,
+      "step": 5348
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.742623459107978,
+      "learning_rate": 4.0345826860978395e-06,
+      "loss": 0.8914,
+      "step": 5349
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7560931571487364,
+      "learning_rate": 4.0328744342997355e-06,
+      "loss": 0.9126,
+      "step": 5350
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.819180724055411,
+      "learning_rate": 4.031166299768939e-06,
+      "loss": 1.0488,
+      "step": 5351
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7520332794232376,
+      "learning_rate": 4.029458282712571e-06,
+      "loss": 0.9031,
+      "step": 5352
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7596683633419,
+      "learning_rate": 4.0277503833377306e-06,
+      "loss": 0.8463,
+      "step": 5353
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.642914650350948,
+      "learning_rate": 4.026042601851509e-06,
+      "loss": 0.8806,
+      "step": 5354
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8289970853285702,
+      "learning_rate": 4.024334938460978e-06,
+      "loss": 1.039,
+      "step": 5355
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8358950126227351,
+      "learning_rate": 4.022627393373201e-06,
+      "loss": 0.865,
+      "step": 5356
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7175001865611605,
+      "learning_rate": 4.020919966795221e-06,
+      "loss": 0.8852,
+      "step": 5357
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6861641380403398,
+      "learning_rate": 4.019212658934073e-06,
+      "loss": 1.0017,
+      "step": 5358
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7723028044433262,
+      "learning_rate": 4.01750546999677e-06,
+      "loss": 0.9758,
+      "step": 5359
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.628643355773677,
+      "learning_rate": 4.015798400190314e-06,
+      "loss": 0.8955,
+      "step": 5360
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7623445209197022,
+      "learning_rate": 4.014091449721697e-06,
+      "loss": 0.9051,
+      "step": 5361
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7360172403103387,
+      "learning_rate": 4.01238461879789e-06,
+      "loss": 0.9131,
+      "step": 5362
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.763367313769858,
+      "learning_rate": 4.0106779076258515e-06,
+      "loss": 0.8855,
+      "step": 5363
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6607262054373768,
+      "learning_rate": 4.0089713164125285e-06,
+      "loss": 0.8264,
+      "step": 5364
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.574827716797458,
+      "learning_rate": 4.007264845364851e-06,
+      "loss": 0.8892,
+      "step": 5365
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8080301399018799,
+      "learning_rate": 4.005558494689732e-06,
+      "loss": 0.869,
+      "step": 5366
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7592629383423783,
+      "learning_rate": 4.0038522645940746e-06,
+      "loss": 1.0576,
+      "step": 5367
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.865349476636154,
+      "learning_rate": 4.002146155284765e-06,
+      "loss": 0.8947,
+      "step": 5368
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6405058178867107,
+      "learning_rate": 4.0004401669686746e-06,
+      "loss": 0.8777,
+      "step": 5369
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6215616543685971,
+      "learning_rate": 3.998734299852659e-06,
+      "loss": 0.7906,
+      "step": 5370
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7190300181569405,
+      "learning_rate": 3.99702855414356e-06,
+      "loss": 0.8627,
+      "step": 5371
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7561891417528654,
+      "learning_rate": 3.9953229300482085e-06,
+      "loss": 0.8421,
+      "step": 5372
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6401439191065865,
+      "learning_rate": 3.993617427773416e-06,
+      "loss": 0.8078,
+      "step": 5373
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7272773878587546,
+      "learning_rate": 3.991912047525976e-06,
+      "loss": 0.8322,
+      "step": 5374
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7008708661661323,
+      "learning_rate": 3.99020678951268e-06,
+      "loss": 0.8927,
+      "step": 5375
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8056917085061484,
+      "learning_rate": 3.9885016539402896e-06,
+      "loss": 0.8634,
+      "step": 5376
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7502036204964766,
+      "learning_rate": 3.986796641015561e-06,
+      "loss": 0.9219,
+      "step": 5377
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7215346983706339,
+      "learning_rate": 3.985091750945234e-06,
+      "loss": 0.936,
+      "step": 5378
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7431472573677849,
+      "learning_rate": 3.983386983936031e-06,
+      "loss": 0.8869,
+      "step": 5379
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7275006394336032,
+      "learning_rate": 3.981682340194661e-06,
+      "loss": 0.8891,
+      "step": 5380
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8336665015895801,
+      "learning_rate": 3.979977819927818e-06,
+      "loss": 0.8744,
+      "step": 5381
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7175307290944628,
+      "learning_rate": 3.97827342334218e-06,
+      "loss": 0.8545,
+      "step": 5382
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7355177904443462,
+      "learning_rate": 3.976569150644411e-06,
+      "loss": 0.9773,
+      "step": 5383
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.764640346382825,
+      "learning_rate": 3.974865002041163e-06,
+      "loss": 0.8943,
+      "step": 5384
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8233890580940408,
+      "learning_rate": 3.973160977739065e-06,
+      "loss": 0.9371,
+      "step": 5385
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6642661120027258,
+      "learning_rate": 3.97145707794474e-06,
+      "loss": 0.8364,
+      "step": 5386
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7242038750971047,
+      "learning_rate": 3.969753302864791e-06,
+      "loss": 0.8774,
+      "step": 5387
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7291887370341168,
+      "learning_rate": 3.9680496527058055e-06,
+      "loss": 0.9539,
+      "step": 5388
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8325325967214428,
+      "learning_rate": 3.966346127674356e-06,
+      "loss": 0.8919,
+      "step": 5389
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9136170314591707,
+      "learning_rate": 3.964642727977004e-06,
+      "loss": 0.9203,
+      "step": 5390
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6997352862715425,
+      "learning_rate": 3.96293945382029e-06,
+      "loss": 0.9209,
+      "step": 5391
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6713290698052294,
+      "learning_rate": 3.9612363054107425e-06,
+      "loss": 0.853,
+      "step": 5392
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6925650053166064,
+      "learning_rate": 3.959533282954874e-06,
+      "loss": 0.8738,
+      "step": 5393
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7189002479504698,
+      "learning_rate": 3.957830386659181e-06,
+      "loss": 0.8171,
+      "step": 5394
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7434379545998133,
+      "learning_rate": 3.956127616730148e-06,
+      "loss": 0.9529,
+      "step": 5395
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7811407780823432,
+      "learning_rate": 3.954424973374239e-06,
+      "loss": 0.8622,
+      "step": 5396
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8258810123332079,
+      "learning_rate": 3.952722456797906e-06,
+      "loss": 1.0486,
+      "step": 5397
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.635994821493495,
+      "learning_rate": 3.9510200672075874e-06,
+      "loss": 0.8598,
+      "step": 5398
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7817125727015353,
+      "learning_rate": 3.949317804809701e-06,
+      "loss": 0.8878,
+      "step": 5399
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9376377275647474,
+      "learning_rate": 3.9476156698106555e-06,
+      "loss": 0.9785,
+      "step": 5400
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.139339617380846,
+      "learning_rate": 3.945913662416836e-06,
+      "loss": 0.8852,
+      "step": 5401
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6865921111538404,
+      "learning_rate": 3.944211782834618e-06,
+      "loss": 0.8676,
+      "step": 5402
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7657886015420805,
+      "learning_rate": 3.942510031270363e-06,
+      "loss": 0.891,
+      "step": 5403
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.768086639322376,
+      "learning_rate": 3.940808407930412e-06,
+      "loss": 0.8701,
+      "step": 5404
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7567879271773947,
+      "learning_rate": 3.939106913021091e-06,
+      "loss": 0.9116,
+      "step": 5405
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7287134320510633,
+      "learning_rate": 3.937405546748716e-06,
+      "loss": 0.8697,
+      "step": 5406
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.721140395896254,
+      "learning_rate": 3.935704309319583e-06,
+      "loss": 0.9731,
+      "step": 5407
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8536309978034724,
+      "learning_rate": 3.934003200939969e-06,
+      "loss": 0.9147,
+      "step": 5408
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8405020214449676,
+      "learning_rate": 3.932302221816144e-06,
+      "loss": 1.0494,
+      "step": 5409
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8269520425725289,
+      "learning_rate": 3.930601372154355e-06,
+      "loss": 0.9956,
+      "step": 5410
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8563436517405723,
+      "learning_rate": 3.928900652160837e-06,
+      "loss": 0.9047,
+      "step": 5411
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6947305189128332,
+      "learning_rate": 3.927200062041808e-06,
+      "loss": 0.9007,
+      "step": 5412
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7031677346707028,
+      "learning_rate": 3.925499602003468e-06,
+      "loss": 0.9535,
+      "step": 5413
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6451226280581113,
+      "learning_rate": 3.923799272252007e-06,
+      "loss": 0.8528,
+      "step": 5414
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7944068765305448,
+      "learning_rate": 3.9220990729935935e-06,
+      "loss": 0.9218,
+      "step": 5415
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.804814683069681,
+      "learning_rate": 3.920399004434383e-06,
+      "loss": 0.896,
+      "step": 5416
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6394907965895702,
+      "learning_rate": 3.918699066780517e-06,
+      "loss": 0.8655,
+      "step": 5417
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.750073308375723,
+      "learning_rate": 3.916999260238116e-06,
+      "loss": 0.8901,
+      "step": 5418
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7459733081088256,
+      "learning_rate": 3.915299585013288e-06,
+      "loss": 0.9197,
+      "step": 5419
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8374589015824432,
+      "learning_rate": 3.913600041312125e-06,
+      "loss": 0.9443,
+      "step": 5420
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.726264835632272,
+      "learning_rate": 3.911900629340703e-06,
+      "loss": 0.8936,
+      "step": 5421
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6633657824665615,
+      "learning_rate": 3.910201349305082e-06,
+      "loss": 0.8277,
+      "step": 5422
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.717638574976483,
+      "learning_rate": 3.908502201411304e-06,
+      "loss": 0.8968,
+      "step": 5423
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8539644854705541,
+      "learning_rate": 3.906803185865395e-06,
+      "loss": 1.0592,
+      "step": 5424
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7960764283126927,
+      "learning_rate": 3.9051043028733696e-06,
+      "loss": 0.8519,
+      "step": 5425
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8158328044882288,
+      "learning_rate": 3.903405552641222e-06,
+      "loss": 1.051,
+      "step": 5426
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7676458568350797,
+      "learning_rate": 3.901706935374933e-06,
+      "loss": 0.8988,
+      "step": 5427
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.796331161895292,
+      "learning_rate": 3.900008451280462e-06,
+      "loss": 0.8494,
+      "step": 5428
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.797479930601746,
+      "learning_rate": 3.89831010056376e-06,
+      "loss": 0.8484,
+      "step": 5429
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6498646801821606,
+      "learning_rate": 3.896611883430756e-06,
+      "loss": 0.9388,
+      "step": 5430
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7101636998447212,
+      "learning_rate": 3.894913800087365e-06,
+      "loss": 0.9332,
+      "step": 5431
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8819357275059887,
+      "learning_rate": 3.8932158507394885e-06,
+      "loss": 0.9042,
+      "step": 5432
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6483786301298522,
+      "learning_rate": 3.891518035593003e-06,
+      "loss": 0.9179,
+      "step": 5433
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7421259523552843,
+      "learning_rate": 3.889820354853777e-06,
+      "loss": 0.8305,
+      "step": 5434
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7466925982565225,
+      "learning_rate": 3.888122808727661e-06,
+      "loss": 0.8568,
+      "step": 5435
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6957028071264761,
+      "learning_rate": 3.886425397420487e-06,
+      "loss": 0.8648,
+      "step": 5436
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7117086125125631,
+      "learning_rate": 3.884728121138073e-06,
+      "loss": 0.8752,
+      "step": 5437
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8003968197385485,
+      "learning_rate": 3.8830309800862196e-06,
+      "loss": 0.9021,
+      "step": 5438
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.756984990195337,
+      "learning_rate": 3.88133397447071e-06,
+      "loss": 0.8687,
+      "step": 5439
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7400245643971342,
+      "learning_rate": 3.879637104497313e-06,
+      "loss": 0.9994,
+      "step": 5440
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7150744571696062,
+      "learning_rate": 3.87794037037178e-06,
+      "loss": 0.9498,
+      "step": 5441
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6773725019316044,
+      "learning_rate": 3.876243772299844e-06,
+      "loss": 0.8855,
+      "step": 5442
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.691879758081048,
+      "learning_rate": 3.8745473104872276e-06,
+      "loss": 0.8908,
+      "step": 5443
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.743186355602486,
+      "learning_rate": 3.872850985139628e-06,
+      "loss": 0.9473,
+      "step": 5444
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.686695104766522,
+      "learning_rate": 3.871154796462732e-06,
+      "loss": 0.8564,
+      "step": 5445
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7376438567262453,
+      "learning_rate": 3.86945874466221e-06,
+      "loss": 0.8553,
+      "step": 5446
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.803152040787705,
+      "learning_rate": 3.86776282994371e-06,
+      "loss": 0.9532,
+      "step": 5447
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.756656787083854,
+      "learning_rate": 3.866067052512872e-06,
+      "loss": 0.9316,
+      "step": 5448
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8172871254978917,
+      "learning_rate": 3.864371412575314e-06,
+      "loss": 0.8864,
+      "step": 5449
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6911606235084264,
+      "learning_rate": 3.862675910336637e-06,
+      "loss": 0.8945,
+      "step": 5450
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7366971283236292,
+      "learning_rate": 3.860980546002425e-06,
+      "loss": 0.8927,
+      "step": 5451
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7013086058548021,
+      "learning_rate": 3.859285319778251e-06,
+      "loss": 0.8436,
+      "step": 5452
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.609641865517127,
+      "learning_rate": 3.857590231869665e-06,
+      "loss": 0.8377,
+      "step": 5453
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7931817207887,
+      "learning_rate": 3.8558952824822e-06,
+      "loss": 0.9228,
+      "step": 5454
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7430235735648119,
+      "learning_rate": 3.854200471821376e-06,
+      "loss": 0.8946,
+      "step": 5455
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7813790399090625,
+      "learning_rate": 3.852505800092696e-06,
+      "loss": 0.9269,
+      "step": 5456
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7354176585258403,
+      "learning_rate": 3.850811267501643e-06,
+      "loss": 0.8851,
+      "step": 5457
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7898293987829987,
+      "learning_rate": 3.8491168742536865e-06,
+      "loss": 0.8902,
+      "step": 5458
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7430770786755125,
+      "learning_rate": 3.847422620554276e-06,
+      "loss": 0.865,
+      "step": 5459
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7290401468512222,
+      "learning_rate": 3.845728506608847e-06,
+      "loss": 0.9232,
+      "step": 5460
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6740039054225269,
+      "learning_rate": 3.844034532622816e-06,
+      "loss": 0.8693,
+      "step": 5461
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7989609740296304,
+      "learning_rate": 3.842340698801581e-06,
+      "loss": 0.9529,
+      "step": 5462
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6472626924425813,
+      "learning_rate": 3.84064700535053e-06,
+      "loss": 0.8408,
+      "step": 5463
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.763357469811722,
+      "learning_rate": 3.838953452475026e-06,
+      "loss": 0.8912,
+      "step": 5464
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7773832782659156,
+      "learning_rate": 3.837260040380419e-06,
+      "loss": 0.8775,
+      "step": 5465
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6866741221911763,
+      "learning_rate": 3.83556676927204e-06,
+      "loss": 0.8425,
+      "step": 5466
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7696767876568005,
+      "learning_rate": 3.833873639355204e-06,
+      "loss": 0.865,
+      "step": 5467
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.723225353754857,
+      "learning_rate": 3.83218065083521e-06,
+      "loss": 0.905,
+      "step": 5468
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.72041010012403,
+      "learning_rate": 3.830487803917338e-06,
+      "loss": 0.7877,
+      "step": 5469
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7185076581387875,
+      "learning_rate": 3.828795098806851e-06,
+      "loss": 0.8756,
+      "step": 5470
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7458230221506337,
+      "learning_rate": 3.827102535708998e-06,
+      "loss": 0.8937,
+      "step": 5471
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7300387753129531,
+      "learning_rate": 3.825410114829005e-06,
+      "loss": 0.92,
+      "step": 5472
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8462507511677857,
+      "learning_rate": 3.823717836372084e-06,
+      "loss": 1.0439,
+      "step": 5473
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7361929077376925,
+      "learning_rate": 3.822025700543431e-06,
+      "loss": 0.9178,
+      "step": 5474
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6459593273342064,
+      "learning_rate": 3.820333707548225e-06,
+      "loss": 0.9242,
+      "step": 5475
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7069191863913327,
+      "learning_rate": 3.818641857591623e-06,
+      "loss": 0.932,
+      "step": 5476
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7296906975506454,
+      "learning_rate": 3.8169501508787686e-06,
+      "loss": 0.8599,
+      "step": 5477
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6021456098522204,
+      "learning_rate": 3.815258587614785e-06,
+      "loss": 0.8898,
+      "step": 5478
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.686541510676499,
+      "learning_rate": 3.8135671680047837e-06,
+      "loss": 0.8729,
+      "step": 5479
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6598196541498524,
+      "learning_rate": 3.8118758922538533e-06,
+      "loss": 0.8557,
+      "step": 5480
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7353164560017316,
+      "learning_rate": 3.8101847605670663e-06,
+      "loss": 0.9466,
+      "step": 5481
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9020871229345766,
+      "learning_rate": 3.80849377314948e-06,
+      "loss": 1.0161,
+      "step": 5482
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.661237256577535,
+      "learning_rate": 3.806802930206131e-06,
+      "loss": 0.9389,
+      "step": 5483
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7926240697092408,
+      "learning_rate": 3.8051122319420406e-06,
+      "loss": 0.8601,
+      "step": 5484
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6725787600985047,
+      "learning_rate": 3.803421678562213e-06,
+      "loss": 0.8131,
+      "step": 5485
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7037017640939962,
+      "learning_rate": 3.801731270271629e-06,
+      "loss": 0.8705,
+      "step": 5486
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.733240109372076,
+      "learning_rate": 3.8000410072752614e-06,
+      "loss": 0.9752,
+      "step": 5487
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6740980854060834,
+      "learning_rate": 3.7983508897780586e-06,
+      "loss": 0.9261,
+      "step": 5488
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6828176414591034,
+      "learning_rate": 3.7966609179849528e-06,
+      "loss": 0.8448,
+      "step": 5489
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7853829025214263,
+      "learning_rate": 3.794971092100858e-06,
+      "loss": 0.8776,
+      "step": 5490
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7461555151667107,
+      "learning_rate": 3.7932814123306737e-06,
+      "loss": 0.9061,
+      "step": 5491
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7014574257530117,
+      "learning_rate": 3.7915918788792793e-06,
+      "loss": 0.8533,
+      "step": 5492
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8514063147977666,
+      "learning_rate": 3.7899024919515337e-06,
+      "loss": 1.0698,
+      "step": 5493
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.654335551874766,
+      "learning_rate": 3.788213251752284e-06,
+      "loss": 0.8629,
+      "step": 5494
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.786657676875161,
+      "learning_rate": 3.7865241584863547e-06,
+      "loss": 0.9714,
+      "step": 5495
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6828499502069274,
+      "learning_rate": 3.7848352123585562e-06,
+      "loss": 0.9156,
+      "step": 5496
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7540698700250033,
+      "learning_rate": 3.7831464135736767e-06,
+      "loss": 0.8709,
+      "step": 5497
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7795811022332515,
+      "learning_rate": 3.7814577623364885e-06,
+      "loss": 0.9121,
+      "step": 5498
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7274507817122284,
+      "learning_rate": 3.7797692588517487e-06,
+      "loss": 0.9814,
+      "step": 5499
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7486240768528967,
+      "learning_rate": 3.7780809033241927e-06,
+      "loss": 0.8676,
+      "step": 5500
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5966822051917842,
+      "learning_rate": 3.776392695958539e-06,
+      "loss": 0.8885,
+      "step": 5501
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.80495886006849,
+      "learning_rate": 3.7747046369594905e-06,
+      "loss": 0.9237,
+      "step": 5502
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6213563534859778,
+      "learning_rate": 3.7730167265317288e-06,
+      "loss": 0.8893,
+      "step": 5503
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6999256338477609,
+      "learning_rate": 3.771328964879918e-06,
+      "loss": 0.8059,
+      "step": 5504
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8298462868081797,
+      "learning_rate": 3.7696413522087068e-06,
+      "loss": 1.0512,
+      "step": 5505
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6460561723246052,
+      "learning_rate": 3.7679538887227247e-06,
+      "loss": 0.886,
+      "step": 5506
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8052042822676628,
+      "learning_rate": 3.7662665746265803e-06,
+      "loss": 0.9229,
+      "step": 5507
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7707642311885385,
+      "learning_rate": 3.7645794101248666e-06,
+      "loss": 0.9157,
+      "step": 5508
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.700475976841979,
+      "learning_rate": 3.7628923954221575e-06,
+      "loss": 0.8968,
+      "step": 5509
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7952408126455726,
+      "learning_rate": 3.7612055307230117e-06,
+      "loss": 0.8726,
+      "step": 5510
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8081859739198791,
+      "learning_rate": 3.759518816231966e-06,
+      "loss": 0.8969,
+      "step": 5511
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7701824496304486,
+      "learning_rate": 3.7578322521535395e-06,
+      "loss": 0.9249,
+      "step": 5512
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8076081419365964,
+      "learning_rate": 3.7561458386922355e-06,
+      "loss": 0.8981,
+      "step": 5513
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7708094719287937,
+      "learning_rate": 3.754459576052537e-06,
+      "loss": 0.8501,
+      "step": 5514
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6887222515511375,
+      "learning_rate": 3.752773464438909e-06,
+      "loss": 0.9305,
+      "step": 5515
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6817858462321236,
+      "learning_rate": 3.7510875040557974e-06,
+      "loss": 0.8541,
+      "step": 5516
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7238824278525104,
+      "learning_rate": 3.749401695107634e-06,
+      "loss": 0.8835,
+      "step": 5517
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8473853542106831,
+      "learning_rate": 3.7477160377988247e-06,
+      "loss": 1.0516,
+      "step": 5518
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7850614146026114,
+      "learning_rate": 3.746030532333763e-06,
+      "loss": 0.8477,
+      "step": 5519
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7617801947645448,
+      "learning_rate": 3.7443451789168235e-06,
+      "loss": 0.9219,
+      "step": 5520
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7465378366580957,
+      "learning_rate": 3.7426599777523577e-06,
+      "loss": 0.9084,
+      "step": 5521
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6941672962295955,
+      "learning_rate": 3.7409749290447062e-06,
+      "loss": 0.8359,
+      "step": 5522
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7252952322528343,
+      "learning_rate": 3.7392900329981845e-06,
+      "loss": 0.8662,
+      "step": 5523
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9807967411887528,
+      "learning_rate": 3.7376052898170916e-06,
+      "loss": 0.9074,
+      "step": 5524
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6577665830496298,
+      "learning_rate": 3.735920699705711e-06,
+      "loss": 0.8602,
+      "step": 5525
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.814942787120635,
+      "learning_rate": 3.734236262868303e-06,
+      "loss": 0.9772,
+      "step": 5526
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6371620702206282,
+      "learning_rate": 3.732551979509113e-06,
+      "loss": 0.8678,
+      "step": 5527
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7816612387157116,
+      "learning_rate": 3.7308678498323648e-06,
+      "loss": 0.8986,
+      "step": 5528
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6849919158264808,
+      "learning_rate": 3.7291838740422627e-06,
+      "loss": 0.9234,
+      "step": 5529
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8378444394254156,
+      "learning_rate": 3.7275000523429993e-06,
+      "loss": 1.0636,
+      "step": 5530
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7395234485226971,
+      "learning_rate": 3.725816384938741e-06,
+      "loss": 0.9207,
+      "step": 5531
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6559729038878623,
+      "learning_rate": 3.7241328720336377e-06,
+      "loss": 0.8375,
+      "step": 5532
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6969503654704081,
+      "learning_rate": 3.722449513831823e-06,
+      "loss": 0.8953,
+      "step": 5533
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8623147189327178,
+      "learning_rate": 3.7207663105374098e-06,
+      "loss": 0.8894,
+      "step": 5534
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.824315878473531,
+      "learning_rate": 3.7190832623544902e-06,
+      "loss": 0.889,
+      "step": 5535
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7786135998824093,
+      "learning_rate": 3.717400369487142e-06,
+      "loss": 0.9342,
+      "step": 5536
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8191369854770012,
+      "learning_rate": 3.715717632139421e-06,
+      "loss": 1.0468,
+      "step": 5537
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8434284793520319,
+      "learning_rate": 3.714035050515366e-06,
+      "loss": 1.0813,
+      "step": 5538
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8337268659602917,
+      "learning_rate": 3.7123526248189935e-06,
+      "loss": 1.0859,
+      "step": 5539
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.85344794231511,
+      "learning_rate": 3.7106703552543026e-06,
+      "loss": 0.9078,
+      "step": 5540
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8296879315320331,
+      "learning_rate": 3.7089882420252775e-06,
+      "loss": 0.8333,
+      "step": 5541
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8093035167059648,
+      "learning_rate": 3.7073062853358794e-06,
+      "loss": 0.9443,
+      "step": 5542
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7924929683229873,
+      "learning_rate": 3.7056244853900493e-06,
+      "loss": 0.8867,
+      "step": 5543
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6732644454675774,
+      "learning_rate": 3.7039428423917138e-06,
+      "loss": 0.9382,
+      "step": 5544
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6566454971099709,
+      "learning_rate": 3.7022613565447774e-06,
+      "loss": 0.8017,
+      "step": 5545
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6955677673990874,
+      "learning_rate": 3.7005800280531244e-06,
+      "loss": 0.8959,
+      "step": 5546
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7453767405157152,
+      "learning_rate": 3.6988988571206224e-06,
+      "loss": 0.896,
+      "step": 5547
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8263251412075266,
+      "learning_rate": 3.697217843951121e-06,
+      "loss": 1.0443,
+      "step": 5548
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7947877778623598,
+      "learning_rate": 3.6955369887484476e-06,
+      "loss": 0.9294,
+      "step": 5549
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6611706798070793,
+      "learning_rate": 3.693856291716411e-06,
+      "loss": 0.8967,
+      "step": 5550
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7536699755198983,
+      "learning_rate": 3.692175753058802e-06,
+      "loss": 0.9205,
+      "step": 5551
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7144336632932475,
+      "learning_rate": 3.69049537297939e-06,
+      "loss": 0.842,
+      "step": 5552
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6686949893356477,
+      "learning_rate": 3.6888151516819303e-06,
+      "loss": 1.0578,
+      "step": 5553
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8039866198202736,
+      "learning_rate": 3.6871350893701533e-06,
+      "loss": 0.9516,
+      "step": 5554
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7297261461546214,
+      "learning_rate": 3.6854551862477724e-06,
+      "loss": 0.8363,
+      "step": 5555
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6216737965885455,
+      "learning_rate": 3.6837754425184823e-06,
+      "loss": 0.9557,
+      "step": 5556
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7194931369470228,
+      "learning_rate": 3.6820958583859585e-06,
+      "loss": 0.9312,
+      "step": 5557
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8466179637176745,
+      "learning_rate": 3.680416434053854e-06,
+      "loss": 0.8714,
+      "step": 5558
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.358048567269164,
+      "learning_rate": 3.678737169725809e-06,
+      "loss": 0.8741,
+      "step": 5559
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.807373975030623,
+      "learning_rate": 3.6770580656054343e-06,
+      "loss": 1.0392,
+      "step": 5560
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7940286785973945,
+      "learning_rate": 3.6753791218963306e-06,
+      "loss": 0.9257,
+      "step": 5561
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6461386631210442,
+      "learning_rate": 3.673700338802076e-06,
+      "loss": 0.8772,
+      "step": 5562
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8128366282386044,
+      "learning_rate": 3.672021716526226e-06,
+      "loss": 0.9334,
+      "step": 5563
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6670290088021622,
+      "learning_rate": 3.670343255272322e-06,
+      "loss": 0.8808,
+      "step": 5564
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8422077333538174,
+      "learning_rate": 3.668664955243883e-06,
+      "loss": 1.0345,
+      "step": 5565
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7020559194372507,
+      "learning_rate": 3.666986816644407e-06,
+      "loss": 0.8807,
+      "step": 5566
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7722858973293232,
+      "learning_rate": 3.665308839677375e-06,
+      "loss": 0.8448,
+      "step": 5567
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6751062232187195,
+      "learning_rate": 3.663631024546249e-06,
+      "loss": 0.8203,
+      "step": 5568
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8437965099377183,
+      "learning_rate": 3.6619533714544664e-06,
+      "loss": 0.9149,
+      "step": 5569
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9265827081628075,
+      "learning_rate": 3.6602758806054537e-06,
+      "loss": 0.9376,
+      "step": 5570
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8125044001298872,
+      "learning_rate": 3.6585985522026067e-06,
+      "loss": 1.0411,
+      "step": 5571
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.809003710362506,
+      "learning_rate": 3.6569213864493103e-06,
+      "loss": 0.9056,
+      "step": 5572
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7366758178844799,
+      "learning_rate": 3.655244383548926e-06,
+      "loss": 0.8897,
+      "step": 5573
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8422687025962967,
+      "learning_rate": 3.6535675437047955e-06,
+      "loss": 1.0362,
+      "step": 5574
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8131825489629856,
+      "learning_rate": 3.651890867120243e-06,
+      "loss": 0.888,
+      "step": 5575
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6846898702464828,
+      "learning_rate": 3.6502143539985706e-06,
+      "loss": 0.9992,
+      "step": 5576
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7520949171665252,
+      "learning_rate": 3.64853800454306e-06,
+      "loss": 0.9715,
+      "step": 5577
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6527210380942445,
+      "learning_rate": 3.646861818956977e-06,
+      "loss": 0.9576,
+      "step": 5578
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.889509295455256,
+      "learning_rate": 3.6451857974435635e-06,
+      "loss": 0.9303,
+      "step": 5579
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8204522035795471,
+      "learning_rate": 3.6435099402060437e-06,
+      "loss": 1.0416,
+      "step": 5580
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8544432294865565,
+      "learning_rate": 3.6418342474476186e-06,
+      "loss": 1.0411,
+      "step": 5581
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.858222558500551,
+      "learning_rate": 3.640158719371473e-06,
+      "loss": 0.9392,
+      "step": 5582
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8077654630402038,
+      "learning_rate": 3.638483356180771e-06,
+      "loss": 1.0232,
+      "step": 5583
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7065771527257185,
+      "learning_rate": 3.636808158078656e-06,
+      "loss": 0.9575,
+      "step": 5584
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.602905210611717,
+      "learning_rate": 3.635133125268252e-06,
+      "loss": 0.7753,
+      "step": 5585
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8023345408645265,
+      "learning_rate": 3.633458257952661e-06,
+      "loss": 1.0294,
+      "step": 5586
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0471990877139636,
+      "learning_rate": 3.631783556334968e-06,
+      "loss": 0.8164,
+      "step": 5587
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6967693393904264,
+      "learning_rate": 3.6301090206182365e-06,
+      "loss": 0.8765,
+      "step": 5588
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7100124040188183,
+      "learning_rate": 3.628434651005507e-06,
+      "loss": 0.9467,
+      "step": 5589
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7987208693960448,
+      "learning_rate": 3.6267604476998063e-06,
+      "loss": 0.8698,
+      "step": 5590
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.850664274180435,
+      "learning_rate": 3.625086410904136e-06,
+      "loss": 0.9044,
+      "step": 5591
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.702221572553094,
+      "learning_rate": 3.623412540821478e-06,
+      "loss": 0.8585,
+      "step": 5592
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6014603960380234,
+      "learning_rate": 3.621738837654795e-06,
+      "loss": 0.8303,
+      "step": 5593
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6383275631227547,
+      "learning_rate": 3.620065301607029e-06,
+      "loss": 0.9033,
+      "step": 5594
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6856015433999665,
+      "learning_rate": 3.618391932881102e-06,
+      "loss": 0.9159,
+      "step": 5595
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8184283673595089,
+      "learning_rate": 3.616718731679918e-06,
+      "loss": 1.0658,
+      "step": 5596
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6326251370703713,
+      "learning_rate": 3.6150456982063555e-06,
+      "loss": 0.8641,
+      "step": 5597
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7077566976961511,
+      "learning_rate": 3.6133728326632765e-06,
+      "loss": 0.8769,
+      "step": 5598
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8102771075159586,
+      "learning_rate": 3.611700135253523e-06,
+      "loss": 1.0335,
+      "step": 5599
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7561055856307544,
+      "learning_rate": 3.6100276061799133e-06,
+      "loss": 0.9163,
+      "step": 5600
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7921173651756581,
+      "learning_rate": 3.6083552456452487e-06,
+      "loss": 0.9182,
+      "step": 5601
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.87667579763137,
+      "learning_rate": 3.60668305385231e-06,
+      "loss": 0.8589,
+      "step": 5602
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8322814743914788,
+      "learning_rate": 3.6050110310038533e-06,
+      "loss": 0.9011,
+      "step": 5603
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6058641836457963,
+      "learning_rate": 3.603339177302618e-06,
+      "loss": 0.8726,
+      "step": 5604
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.769992996334562,
+      "learning_rate": 3.6016674929513218e-06,
+      "loss": 0.9335,
+      "step": 5605
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7324418397678591,
+      "learning_rate": 3.599995978152664e-06,
+      "loss": 0.8575,
+      "step": 5606
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.822640472887243,
+      "learning_rate": 3.5983246331093196e-06,
+      "loss": 0.9061,
+      "step": 5607
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8600725223371645,
+      "learning_rate": 3.5966534580239454e-06,
+      "loss": 1.0364,
+      "step": 5608
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7365507643690683,
+      "learning_rate": 3.5949824530991783e-06,
+      "loss": 0.8715,
+      "step": 5609
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7477268774541044,
+      "learning_rate": 3.5933116185376325e-06,
+      "loss": 0.9084,
+      "step": 5610
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.696317442501295,
+      "learning_rate": 3.591640954541903e-06,
+      "loss": 0.9295,
+      "step": 5611
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.891740035280818,
+      "learning_rate": 3.5899704613145637e-06,
+      "loss": 0.8756,
+      "step": 5612
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.729968150102001,
+      "learning_rate": 3.588300139058165e-06,
+      "loss": 0.8106,
+      "step": 5613
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7773379505569598,
+      "learning_rate": 3.586629987975243e-06,
+      "loss": 0.8996,
+      "step": 5614
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7604489080700891,
+      "learning_rate": 3.5849600082683068e-06,
+      "loss": 0.8485,
+      "step": 5615
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8666187768677037,
+      "learning_rate": 3.5832902001398486e-06,
+      "loss": 1.0497,
+      "step": 5616
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.737951671921557,
+      "learning_rate": 3.581620563792336e-06,
+      "loss": 0.8886,
+      "step": 5617
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7730476964235946,
+      "learning_rate": 3.5799510994282215e-06,
+      "loss": 0.9036,
+      "step": 5618
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7986216097863812,
+      "learning_rate": 3.5782818072499313e-06,
+      "loss": 1.0632,
+      "step": 5619
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.753989187941031,
+      "learning_rate": 3.5766126874598734e-06,
+      "loss": 0.8624,
+      "step": 5620
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6413928296415943,
+      "learning_rate": 3.574943740260435e-06,
+      "loss": 0.9417,
+      "step": 5621
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.810872734070383,
+      "learning_rate": 3.57327496585398e-06,
+      "loss": 1.0286,
+      "step": 5622
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.771685237740851,
+      "learning_rate": 3.571606364442857e-06,
+      "loss": 0.9652,
+      "step": 5623
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.69409231725767,
+      "learning_rate": 3.569937936229384e-06,
+      "loss": 0.8889,
+      "step": 5624
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8137739382316567,
+      "learning_rate": 3.568269681415866e-06,
+      "loss": 0.9252,
+      "step": 5625
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6934537123820026,
+      "learning_rate": 3.5666016002045857e-06,
+      "loss": 0.8559,
+      "step": 5626
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.69622061966645,
+      "learning_rate": 3.5649336927978036e-06,
+      "loss": 0.8786,
+      "step": 5627
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7769454870897883,
+      "learning_rate": 3.563265959397757e-06,
+      "loss": 0.9224,
+      "step": 5628
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6697460944034928,
+      "learning_rate": 3.561598400206667e-06,
+      "loss": 0.8459,
+      "step": 5629
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7976141428353352,
+      "learning_rate": 3.559931015426731e-06,
+      "loss": 0.8522,
+      "step": 5630
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7936161488417472,
+      "learning_rate": 3.5582638052601227e-06,
+      "loss": 0.9381,
+      "step": 5631
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.601610348015723,
+      "learning_rate": 3.556596769908999e-06,
+      "loss": 0.9252,
+      "step": 5632
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.633213386122276,
+      "learning_rate": 3.554929909575494e-06,
+      "loss": 0.8121,
+      "step": 5633
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.658230050768033,
+      "learning_rate": 3.5532632244617184e-06,
+      "loss": 0.9114,
+      "step": 5634
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7586741954537217,
+      "learning_rate": 3.551596714769765e-06,
+      "loss": 0.814,
+      "step": 5635
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7818636521949829,
+      "learning_rate": 3.5499303807017018e-06,
+      "loss": 0.8759,
+      "step": 5636
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.779964203245575,
+      "learning_rate": 3.54826422245958e-06,
+      "loss": 0.8645,
+      "step": 5637
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7685497562787866,
+      "learning_rate": 3.546598240245427e-06,
+      "loss": 0.8949,
+      "step": 5638
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7985707910537034,
+      "learning_rate": 3.5449324342612464e-06,
+      "loss": 0.7795,
+      "step": 5639
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.858739349362007,
+      "learning_rate": 3.5432668047090258e-06,
+      "loss": 0.9002,
+      "step": 5640
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6748499532546481,
+      "learning_rate": 3.5416013517907265e-06,
+      "loss": 0.8971,
+      "step": 5641
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7903718507677502,
+      "learning_rate": 3.5399360757082918e-06,
+      "loss": 0.8844,
+      "step": 5642
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8907651267994066,
+      "learning_rate": 3.5382709766636404e-06,
+      "loss": 1.0004,
+      "step": 5643
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7316705781084434,
+      "learning_rate": 3.5366060548586745e-06,
+      "loss": 0.8748,
+      "step": 5644
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7529696579026233,
+      "learning_rate": 3.5349413104952686e-06,
+      "loss": 0.8693,
+      "step": 5645
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7223377312626231,
+      "learning_rate": 3.5332767437752792e-06,
+      "loss": 0.8427,
+      "step": 5646
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8367606486825332,
+      "learning_rate": 3.531612354900542e-06,
+      "loss": 0.87,
+      "step": 5647
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7233465693958416,
+      "learning_rate": 3.5299481440728673e-06,
+      "loss": 0.9799,
+      "step": 5648
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8272221517110389,
+      "learning_rate": 3.528284111494049e-06,
+      "loss": 1.044,
+      "step": 5649
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7710311878509273,
+      "learning_rate": 3.526620257365857e-06,
+      "loss": 0.9254,
+      "step": 5650
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8145597223891708,
+      "learning_rate": 3.5249565818900365e-06,
+      "loss": 1.0627,
+      "step": 5651
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9221674932841766,
+      "learning_rate": 3.5232930852683168e-06,
+      "loss": 0.8769,
+      "step": 5652
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.634025118892234,
+      "learning_rate": 3.521629767702401e-06,
+      "loss": 0.9536,
+      "step": 5653
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6873157883503735,
+      "learning_rate": 3.5199666293939725e-06,
+      "loss": 0.8699,
+      "step": 5654
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8680320553975331,
+      "learning_rate": 3.518303670544696e-06,
+      "loss": 0.9231,
+      "step": 5655
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8092927869223665,
+      "learning_rate": 3.5166408913562033e-06,
+      "loss": 1.0694,
+      "step": 5656
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8384536799090528,
+      "learning_rate": 3.5149782920301178e-06,
+      "loss": 1.0536,
+      "step": 5657
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6646442299263833,
+      "learning_rate": 3.513315872768035e-06,
+      "loss": 0.894,
+      "step": 5658
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6578216398771763,
+      "learning_rate": 3.5116536337715255e-06,
+      "loss": 0.917,
+      "step": 5659
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7214383106650173,
+      "learning_rate": 3.5099915752421467e-06,
+      "loss": 0.8678,
+      "step": 5660
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7684468400330795,
+      "learning_rate": 3.5083296973814253e-06,
+      "loss": 0.8944,
+      "step": 5661
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7835645607360697,
+      "learning_rate": 3.5066680003908695e-06,
+      "loss": 0.8845,
+      "step": 5662
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7929621316165107,
+      "learning_rate": 3.5050064844719677e-06,
+      "loss": 0.871,
+      "step": 5663
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6647261201456773,
+      "learning_rate": 3.503345149826185e-06,
+      "loss": 0.8678,
+      "step": 5664
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8889024873909521,
+      "learning_rate": 3.5016839966549626e-06,
+      "loss": 0.9782,
+      "step": 5665
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7755315105945544,
+      "learning_rate": 3.50002302515972e-06,
+      "loss": 0.9652,
+      "step": 5666
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6269696659026092,
+      "learning_rate": 3.498362235541856e-06,
+      "loss": 0.8726,
+      "step": 5667
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.723174528423652,
+      "learning_rate": 3.4967016280027494e-06,
+      "loss": 0.8852,
+      "step": 5668
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7534294896856653,
+      "learning_rate": 3.4950412027437523e-06,
+      "loss": 0.875,
+      "step": 5669
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8394842374752671,
+      "learning_rate": 3.4933809599661973e-06,
+      "loss": 1.0248,
+      "step": 5670
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8354183268277448,
+      "learning_rate": 3.491720899871396e-06,
+      "loss": 1.0445,
+      "step": 5671
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.725845834623862,
+      "learning_rate": 3.4900610226606346e-06,
+      "loss": 0.8397,
+      "step": 5672
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7689199007674459,
+      "learning_rate": 3.4884013285351796e-06,
+      "loss": 0.8842,
+      "step": 5673
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7047701766087395,
+      "learning_rate": 3.4867418176962755e-06,
+      "loss": 0.8868,
+      "step": 5674
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7106233871873942,
+      "learning_rate": 3.485082490345143e-06,
+      "loss": 0.9266,
+      "step": 5675
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8601754239861215,
+      "learning_rate": 3.4834233466829827e-06,
+      "loss": 0.9526,
+      "step": 5676
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6917923766884189,
+      "learning_rate": 3.4817643869109683e-06,
+      "loss": 0.9166,
+      "step": 5677
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7247029501181677,
+      "learning_rate": 3.4801056112302554e-06,
+      "loss": 0.9771,
+      "step": 5678
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7376642320691873,
+      "learning_rate": 3.4784470198419783e-06,
+      "loss": 0.8287,
+      "step": 5679
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6852546240687505,
+      "learning_rate": 3.4767886129472455e-06,
+      "loss": 0.9039,
+      "step": 5680
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7082633040690438,
+      "learning_rate": 3.4751303907471444e-06,
+      "loss": 0.9118,
+      "step": 5681
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7633879171618096,
+      "learning_rate": 3.47347235344274e-06,
+      "loss": 1.0027,
+      "step": 5682
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9377356455620571,
+      "learning_rate": 3.471814501235076e-06,
+      "loss": 0.8162,
+      "step": 5683
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.626445069527124,
+      "learning_rate": 3.4701568343251723e-06,
+      "loss": 0.8798,
+      "step": 5684
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7281177915022998,
+      "learning_rate": 3.468499352914026e-06,
+      "loss": 0.8811,
+      "step": 5685
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8457188744192228,
+      "learning_rate": 3.4668420572026158e-06,
+      "loss": 0.867,
+      "step": 5686
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7344652576529356,
+      "learning_rate": 3.4651849473918887e-06,
+      "loss": 0.8594,
+      "step": 5687
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.766703886054441,
+      "learning_rate": 3.463528023682779e-06,
+      "loss": 0.9157,
+      "step": 5688
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6969499748763655,
+      "learning_rate": 3.461871286276194e-06,
+      "loss": 0.9073,
+      "step": 5689
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7195402374126778,
+      "learning_rate": 3.4602147353730165e-06,
+      "loss": 0.9796,
+      "step": 5690
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6957301264341562,
+      "learning_rate": 3.4585583711741115e-06,
+      "loss": 0.9358,
+      "step": 5691
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6187384947830747,
+      "learning_rate": 3.456902193880319e-06,
+      "loss": 0.9426,
+      "step": 5692
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.634320839798652,
+      "learning_rate": 3.455246203692454e-06,
+      "loss": 0.8466,
+      "step": 5693
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7978086798186974,
+      "learning_rate": 3.453590400811313e-06,
+      "loss": 0.924,
+      "step": 5694
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7418443981846683,
+      "learning_rate": 3.4519347854376677e-06,
+      "loss": 0.8565,
+      "step": 5695
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7197169201939806,
+      "learning_rate": 3.450279357772266e-06,
+      "loss": 0.9083,
+      "step": 5696
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7474437701082768,
+      "learning_rate": 3.4486241180158373e-06,
+      "loss": 0.9903,
+      "step": 5697
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.635727934254354,
+      "learning_rate": 3.44696906636908e-06,
+      "loss": 0.9267,
+      "step": 5698
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.766418775765097,
+      "learning_rate": 3.445314203032678e-06,
+      "loss": 1.0015,
+      "step": 5699
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6472864912368657,
+      "learning_rate": 3.443659528207289e-06,
+      "loss": 0.8065,
+      "step": 5700
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8452692700400021,
+      "learning_rate": 3.442005042093547e-06,
+      "loss": 1.0823,
+      "step": 5701
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8802698939429319,
+      "learning_rate": 3.440350744892066e-06,
+      "loss": 0.9812,
+      "step": 5702
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6986588522164532,
+      "learning_rate": 3.4386966368034335e-06,
+      "loss": 0.897,
+      "step": 5703
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7856493007613126,
+      "learning_rate": 3.437042718028215e-06,
+      "loss": 0.9253,
+      "step": 5704
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.75474866167282,
+      "learning_rate": 3.435388988766958e-06,
+      "loss": 0.8866,
+      "step": 5705
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7445865708488477,
+      "learning_rate": 3.4337354492201786e-06,
+      "loss": 0.8907,
+      "step": 5706
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.622242984414676,
+      "learning_rate": 3.4320820995883776e-06,
+      "loss": 0.989,
+      "step": 5707
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7533171983478932,
+      "learning_rate": 3.430428940072026e-06,
+      "loss": 0.8435,
+      "step": 5708
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7757032678984757,
+      "learning_rate": 3.4287759708715753e-06,
+      "loss": 0.934,
+      "step": 5709
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7217221974408432,
+      "learning_rate": 3.427123192187456e-06,
+      "loss": 0.9206,
+      "step": 5710
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8145076883315294,
+      "learning_rate": 3.4254706042200724e-06,
+      "loss": 0.8656,
+      "step": 5711
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7386062100273125,
+      "learning_rate": 3.4238182071698066e-06,
+      "loss": 0.9234,
+      "step": 5712
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7669339235805144,
+      "learning_rate": 3.422166001237016e-06,
+      "loss": 0.8319,
+      "step": 5713
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7099214016120043,
+      "learning_rate": 3.4205139866220384e-06,
+      "loss": 0.8891,
+      "step": 5714
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7658374518382718,
+      "learning_rate": 3.418862163525185e-06,
+      "loss": 0.9116,
+      "step": 5715
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7625864898946653,
+      "learning_rate": 3.417210532146744e-06,
+      "loss": 0.8948,
+      "step": 5716
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.710508732239102,
+      "learning_rate": 3.415559092686984e-06,
+      "loss": 0.9357,
+      "step": 5717
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6971155993836997,
+      "learning_rate": 3.4139078453461472e-06,
+      "loss": 0.947,
+      "step": 5718
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.72881275206995,
+      "learning_rate": 3.412256790324452e-06,
+      "loss": 0.8991,
+      "step": 5719
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.934393845436041,
+      "learning_rate": 3.4106059278220938e-06,
+      "loss": 0.8343,
+      "step": 5720
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.757278858994872,
+      "learning_rate": 3.408955258039246e-06,
+      "loss": 0.8983,
+      "step": 5721
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7240995504068202,
+      "learning_rate": 3.407304781176059e-06,
+      "loss": 0.8783,
+      "step": 5722
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7134395421524382,
+      "learning_rate": 3.4056544974326584e-06,
+      "loss": 0.9777,
+      "step": 5723
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6101468793947178,
+      "learning_rate": 3.404004407009145e-06,
+      "loss": 0.8154,
+      "step": 5724
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6959294039152442,
+      "learning_rate": 3.402354510105601e-06,
+      "loss": 0.9545,
+      "step": 5725
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7402670397680935,
+      "learning_rate": 3.4007048069220807e-06,
+      "loss": 0.9122,
+      "step": 5726
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6855463297375057,
+      "learning_rate": 3.3990552976586144e-06,
+      "loss": 0.8455,
+      "step": 5727
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7361777118985908,
+      "learning_rate": 3.397405982515214e-06,
+      "loss": 0.8388,
+      "step": 5728
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7871447523058728,
+      "learning_rate": 3.3957568616918644e-06,
+      "loss": 1.0172,
+      "step": 5729
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6660135005365655,
+      "learning_rate": 3.3941079353885255e-06,
+      "loss": 0.8818,
+      "step": 5730
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6981787455385082,
+      "learning_rate": 3.3924592038051345e-06,
+      "loss": 0.8864,
+      "step": 5731
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7832170232417044,
+      "learning_rate": 3.390810667141606e-06,
+      "loss": 0.9094,
+      "step": 5732
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6923568125066233,
+      "learning_rate": 3.3891623255978347e-06,
+      "loss": 0.9375,
+      "step": 5733
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8724790286602109,
+      "learning_rate": 3.387514179373683e-06,
+      "loss": 0.8947,
+      "step": 5734
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8291648810467946,
+      "learning_rate": 3.385866228668996e-06,
+      "loss": 0.8931,
+      "step": 5735
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8135187869317142,
+      "learning_rate": 3.3842184736835943e-06,
+      "loss": 0.8801,
+      "step": 5736
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.765389729239931,
+      "learning_rate": 3.382570914617273e-06,
+      "loss": 0.9826,
+      "step": 5737
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.904392359900347,
+      "learning_rate": 3.3809235516698045e-06,
+      "loss": 0.9555,
+      "step": 5738
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7379480911842708,
+      "learning_rate": 3.379276385040938e-06,
+      "loss": 0.9911,
+      "step": 5739
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7360384975289973,
+      "learning_rate": 3.3776294149303956e-06,
+      "loss": 0.9569,
+      "step": 5740
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7899757211305796,
+      "learning_rate": 3.375982641537881e-06,
+      "loss": 0.8751,
+      "step": 5741
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8244468884082404,
+      "learning_rate": 3.3743360650630695e-06,
+      "loss": 1.0447,
+      "step": 5742
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8264718254936257,
+      "learning_rate": 3.3726896857056145e-06,
+      "loss": 1.0654,
+      "step": 5743
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.669230663865067,
+      "learning_rate": 3.3710435036651453e-06,
+      "loss": 0.9431,
+      "step": 5744
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.741028265127649,
+      "learning_rate": 3.369397519141267e-06,
+      "loss": 0.8751,
+      "step": 5745
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7405595169561119,
+      "learning_rate": 3.3677517323335616e-06,
+      "loss": 0.8688,
+      "step": 5746
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.79414594356946,
+      "learning_rate": 3.3661061434415847e-06,
+      "loss": 1.0066,
+      "step": 5747
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7589790136098502,
+      "learning_rate": 3.3644607526648722e-06,
+      "loss": 0.9194,
+      "step": 5748
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7254000271549643,
+      "learning_rate": 3.3628155602029312e-06,
+      "loss": 0.9111,
+      "step": 5749
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7068781784892262,
+      "learning_rate": 3.3611705662552494e-06,
+      "loss": 0.8894,
+      "step": 5750
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6418302022950935,
+      "learning_rate": 3.3595257710212853e-06,
+      "loss": 0.8876,
+      "step": 5751
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7475609693421148,
+      "learning_rate": 3.3578811747004757e-06,
+      "loss": 0.9347,
+      "step": 5752
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7267591491952203,
+      "learning_rate": 3.356236777492236e-06,
+      "loss": 0.8496,
+      "step": 5753
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6164003727443874,
+      "learning_rate": 3.3545925795959545e-06,
+      "loss": 0.8285,
+      "step": 5754
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.722020446152228,
+      "learning_rate": 3.3529485812109934e-06,
+      "loss": 0.9692,
+      "step": 5755
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.706702860051739,
+      "learning_rate": 3.3513047825366968e-06,
+      "loss": 0.8232,
+      "step": 5756
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7650500657551864,
+      "learning_rate": 3.349661183772379e-06,
+      "loss": 0.9264,
+      "step": 5757
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7956699121435815,
+      "learning_rate": 3.3480177851173314e-06,
+      "loss": 0.8948,
+      "step": 5758
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7025879829427009,
+      "learning_rate": 3.346374586770823e-06,
+      "loss": 0.9015,
+      "step": 5759
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.82375281035598,
+      "learning_rate": 3.3447315889320985e-06,
+      "loss": 0.918,
+      "step": 5760
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.632183415309447,
+      "learning_rate": 3.343088791800374e-06,
+      "loss": 0.8812,
+      "step": 5761
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.811219170301259,
+      "learning_rate": 3.341446195574846e-06,
+      "loss": 0.8749,
+      "step": 5762
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7520473981541287,
+      "learning_rate": 3.339803800454684e-06,
+      "loss": 0.9729,
+      "step": 5763
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7889797391941364,
+      "learning_rate": 3.3381616066390355e-06,
+      "loss": 0.9018,
+      "step": 5764
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7494296061428083,
+      "learning_rate": 3.3365196143270207e-06,
+      "loss": 0.8926,
+      "step": 5765
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7628053667965524,
+      "learning_rate": 3.334877823717737e-06,
+      "loss": 0.8574,
+      "step": 5766
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.751832616726469,
+      "learning_rate": 3.3332362350102592e-06,
+      "loss": 0.8969,
+      "step": 5767
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7195571774843181,
+      "learning_rate": 3.3315948484036344e-06,
+      "loss": 0.9042,
+      "step": 5768
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.762227091301854,
+      "learning_rate": 3.3299536640968844e-06,
+      "loss": 0.925,
+      "step": 5769
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9110697913040866,
+      "learning_rate": 3.328312682289011e-06,
+      "loss": 0.955,
+      "step": 5770
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.700169285435469,
+      "learning_rate": 3.3266719031789896e-06,
+      "loss": 0.8472,
+      "step": 5771
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6896652932168508,
+      "learning_rate": 3.3250313269657673e-06,
+      "loss": 0.8414,
+      "step": 5772
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7818536432657286,
+      "learning_rate": 3.3233909538482712e-06,
+      "loss": 0.854,
+      "step": 5773
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6985143627535695,
+      "learning_rate": 3.321750784025401e-06,
+      "loss": 0.93,
+      "step": 5774
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6279143683199,
+      "learning_rate": 3.320110817696035e-06,
+      "loss": 0.9123,
+      "step": 5775
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9555096071316398,
+      "learning_rate": 3.3184710550590244e-06,
+      "loss": 0.9382,
+      "step": 5776
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.103852717527289,
+      "learning_rate": 3.316831496313195e-06,
+      "loss": 0.8883,
+      "step": 5777
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.971856442751516,
+      "learning_rate": 3.3151921416573486e-06,
+      "loss": 0.8985,
+      "step": 5778
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8829336693496748,
+      "learning_rate": 3.3135529912902644e-06,
+      "loss": 1.0066,
+      "step": 5779
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.786030582197453,
+      "learning_rate": 3.3119140454106945e-06,
+      "loss": 0.8347,
+      "step": 5780
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8193071348438765,
+      "learning_rate": 3.310275304217365e-06,
+      "loss": 0.8452,
+      "step": 5781
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7117178894294423,
+      "learning_rate": 3.3086367679089824e-06,
+      "loss": 0.9706,
+      "step": 5782
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6892220202739219,
+      "learning_rate": 3.306998436684221e-06,
+      "loss": 0.8568,
+      "step": 5783
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8330533094692414,
+      "learning_rate": 3.3053603107417365e-06,
+      "loss": 0.9201,
+      "step": 5784
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6459899130515607,
+      "learning_rate": 3.3037223902801565e-06,
+      "loss": 0.803,
+      "step": 5785
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8012086481251026,
+      "learning_rate": 3.302084675498083e-06,
+      "loss": 0.9313,
+      "step": 5786
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.637243855471105,
+      "learning_rate": 3.3004471665940975e-06,
+      "loss": 0.8225,
+      "step": 5787
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7829961251308677,
+      "learning_rate": 3.2988098637667524e-06,
+      "loss": 0.8458,
+      "step": 5788
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.613536358794607,
+      "learning_rate": 3.297172767214576e-06,
+      "loss": 0.883,
+      "step": 5789
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.670723223885938,
+      "learning_rate": 3.2955358771360725e-06,
+      "loss": 0.8946,
+      "step": 5790
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6456907278972213,
+      "learning_rate": 3.293899193729719e-06,
+      "loss": 0.9107,
+      "step": 5791
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.723710093010037,
+      "learning_rate": 3.2922627171939726e-06,
+      "loss": 0.897,
+      "step": 5792
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.748523795570614,
+      "learning_rate": 3.2906264477272576e-06,
+      "loss": 0.872,
+      "step": 5793
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7301139740521405,
+      "learning_rate": 3.288990385527978e-06,
+      "loss": 0.9024,
+      "step": 5794
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7835934813411236,
+      "learning_rate": 3.2873545307945144e-06,
+      "loss": 0.9094,
+      "step": 5795
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9027777499611487,
+      "learning_rate": 3.2857188837252175e-06,
+      "loss": 0.9486,
+      "step": 5796
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7222171748702142,
+      "learning_rate": 3.2840834445184156e-06,
+      "loss": 0.883,
+      "step": 5797
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.75808612779035,
+      "learning_rate": 3.2824482133724124e-06,
+      "loss": 0.8763,
+      "step": 5798
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7571828199912405,
+      "learning_rate": 3.2808131904854843e-06,
+      "loss": 0.8693,
+      "step": 5799
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.638493796332804,
+      "learning_rate": 3.2791783760558836e-06,
+      "loss": 0.9599,
+      "step": 5800
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.817029349532796,
+      "learning_rate": 3.277543770281838e-06,
+      "loss": 1.0665,
+      "step": 5801
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7667230253082553,
+      "learning_rate": 3.2759093733615482e-06,
+      "loss": 0.9491,
+      "step": 5802
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7888898816856202,
+      "learning_rate": 3.274275185493192e-06,
+      "loss": 1.0475,
+      "step": 5803
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7130336450099488,
+      "learning_rate": 3.272641206874918e-06,
+      "loss": 0.9359,
+      "step": 5804
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.660348242138406,
+      "learning_rate": 3.2710074377048516e-06,
+      "loss": 0.8718,
+      "step": 5805
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7042370200609216,
+      "learning_rate": 3.269373878181096e-06,
+      "loss": 0.9358,
+      "step": 5806
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7189607946490364,
+      "learning_rate": 3.2677405285017227e-06,
+      "loss": 0.9056,
+      "step": 5807
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7544762190067664,
+      "learning_rate": 3.266107388864783e-06,
+      "loss": 0.965,
+      "step": 5808
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.804614382010377,
+      "learning_rate": 3.264474459468299e-06,
+      "loss": 0.9378,
+      "step": 5809
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7687082694966068,
+      "learning_rate": 3.262841740510271e-06,
+      "loss": 0.8667,
+      "step": 5810
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6952843148052628,
+      "learning_rate": 3.2612092321886713e-06,
+      "loss": 0.8818,
+      "step": 5811
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7620733424580628,
+      "learning_rate": 3.259576934701445e-06,
+      "loss": 0.8665,
+      "step": 5812
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8298126042883063,
+      "learning_rate": 3.2579448482465194e-06,
+      "loss": 1.0597,
+      "step": 5813
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.638757599668099,
+      "learning_rate": 3.256312973021783e-06,
+      "loss": 0.8826,
+      "step": 5814
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8382741582075028,
+      "learning_rate": 3.254681309225111e-06,
+      "loss": 0.8486,
+      "step": 5815
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8396038098906063,
+      "learning_rate": 3.2530498570543477e-06,
+      "loss": 1.0856,
+      "step": 5816
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.778140043055113,
+      "learning_rate": 3.2514186167073103e-06,
+      "loss": 0.9371,
+      "step": 5817
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7919930539222075,
+      "learning_rate": 3.2497875883817955e-06,
+      "loss": 0.9046,
+      "step": 5818
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7955919999162238,
+      "learning_rate": 3.248156772275569e-06,
+      "loss": 0.9549,
+      "step": 5819
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.696499604060048,
+      "learning_rate": 3.2465261685863723e-06,
+      "loss": 0.915,
+      "step": 5820
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7711061301234048,
+      "learning_rate": 3.244895777511925e-06,
+      "loss": 0.8534,
+      "step": 5821
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6958458087679902,
+      "learning_rate": 3.243265599249914e-06,
+      "loss": 0.9142,
+      "step": 5822
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8918054880134345,
+      "learning_rate": 3.2416356339980056e-06,
+      "loss": 0.8632,
+      "step": 5823
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7062613504988868,
+      "learning_rate": 3.240005881953841e-06,
+      "loss": 0.8859,
+      "step": 5824
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8250422472016088,
+      "learning_rate": 3.2383763433150277e-06,
+      "loss": 0.9737,
+      "step": 5825
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6476670911839033,
+      "learning_rate": 3.236747018279157e-06,
+      "loss": 0.8399,
+      "step": 5826
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.681962846605396,
+      "learning_rate": 3.23511790704379e-06,
+      "loss": 0.8563,
+      "step": 5827
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5896146058912484,
+      "learning_rate": 3.23348900980646e-06,
+      "loss": 0.8947,
+      "step": 5828
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8609636379948766,
+      "learning_rate": 3.2318603267646774e-06,
+      "loss": 0.8716,
+      "step": 5829
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.734706335680236,
+      "learning_rate": 3.2302318581159268e-06,
+      "loss": 0.8395,
+      "step": 5830
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7055075314447223,
+      "learning_rate": 3.228603604057664e-06,
+      "loss": 0.8816,
+      "step": 5831
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.661771339457933,
+      "learning_rate": 3.226975564787322e-06,
+      "loss": 0.9296,
+      "step": 5832
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8081030255513815,
+      "learning_rate": 3.2253477405023047e-06,
+      "loss": 1.0326,
+      "step": 5833
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7450600295028296,
+      "learning_rate": 3.2237201313999927e-06,
+      "loss": 0.9199,
+      "step": 5834
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7860641091909126,
+      "learning_rate": 3.2220927376777393e-06,
+      "loss": 0.9594,
+      "step": 5835
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.718613978015559,
+      "learning_rate": 3.2204655595328693e-06,
+      "loss": 0.88,
+      "step": 5836
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8418347140608273,
+      "learning_rate": 3.2188385971626855e-06,
+      "loss": 0.9128,
+      "step": 5837
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7116959615346439,
+      "learning_rate": 3.2172118507644624e-06,
+      "loss": 0.9798,
+      "step": 5838
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.644165060120402,
+      "learning_rate": 3.215585320535449e-06,
+      "loss": 0.9366,
+      "step": 5839
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.682685171140966,
+      "learning_rate": 3.2139590066728667e-06,
+      "loss": 0.8643,
+      "step": 5840
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.714710437004348,
+      "learning_rate": 3.212332909373912e-06,
+      "loss": 0.9016,
+      "step": 5841
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6193632366649306,
+      "learning_rate": 3.2107070288357557e-06,
+      "loss": 0.8802,
+      "step": 5842
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8139165637713215,
+      "learning_rate": 3.2090813652555396e-06,
+      "loss": 0.8576,
+      "step": 5843
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7101282848964754,
+      "learning_rate": 3.207455918830384e-06,
+      "loss": 0.9138,
+      "step": 5844
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.773345262055934,
+      "learning_rate": 3.205830689757379e-06,
+      "loss": 0.9478,
+      "step": 5845
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8002754430276613,
+      "learning_rate": 3.204205678233586e-06,
+      "loss": 0.9529,
+      "step": 5846
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8139947586285003,
+      "learning_rate": 3.202580884456047e-06,
+      "loss": 0.9228,
+      "step": 5847
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.766318365203739,
+      "learning_rate": 3.2009563086217703e-06,
+      "loss": 0.9484,
+      "step": 5848
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8162479348117928,
+      "learning_rate": 3.1993319509277455e-06,
+      "loss": 0.897,
+      "step": 5849
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7993666149127705,
+      "learning_rate": 3.1977078115709285e-06,
+      "loss": 0.9509,
+      "step": 5850
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8798044213271798,
+      "learning_rate": 3.1960838907482524e-06,
+      "loss": 0.9063,
+      "step": 5851
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7897726830714662,
+      "learning_rate": 3.194460188656624e-06,
+      "loss": 0.8724,
+      "step": 5852
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.693502055303962,
+      "learning_rate": 3.1928367054929234e-06,
+      "loss": 0.9001,
+      "step": 5853
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7186852925788019,
+      "learning_rate": 3.1912134414540008e-06,
+      "loss": 0.9553,
+      "step": 5854
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6270424406216109,
+      "learning_rate": 3.1895903967366847e-06,
+      "loss": 0.8204,
+      "step": 5855
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.759408938722275,
+      "learning_rate": 3.1879675715377766e-06,
+      "loss": 0.9269,
+      "step": 5856
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7325589827577368,
+      "learning_rate": 3.186344966054046e-06,
+      "loss": 0.8269,
+      "step": 5857
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6812588213180517,
+      "learning_rate": 3.1847225804822408e-06,
+      "loss": 0.8368,
+      "step": 5858
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7020839215752526,
+      "learning_rate": 3.18310041501908e-06,
+      "loss": 0.9014,
+      "step": 5859
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8185721354319937,
+      "learning_rate": 3.181478469861259e-06,
+      "loss": 0.9388,
+      "step": 5860
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6911778680923761,
+      "learning_rate": 3.179856745205442e-06,
+      "loss": 0.8233,
+      "step": 5861
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7263573971030368,
+      "learning_rate": 3.1782352412482693e-06,
+      "loss": 0.8779,
+      "step": 5862
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6619367258931648,
+      "learning_rate": 3.176613958186355e-06,
+      "loss": 0.8337,
+      "step": 5863
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7507500152488946,
+      "learning_rate": 3.1749928962162845e-06,
+      "loss": 0.8808,
+      "step": 5864
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.82462849945111,
+      "learning_rate": 3.1733720555346158e-06,
+      "loss": 0.9555,
+      "step": 5865
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8058423835082655,
+      "learning_rate": 3.1717514363378864e-06,
+      "loss": 0.9253,
+      "step": 5866
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.809882810738741,
+      "learning_rate": 3.170131038822595e-06,
+      "loss": 1.068,
+      "step": 5867
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6765038888351997,
+      "learning_rate": 3.1685108631852244e-06,
+      "loss": 0.9453,
+      "step": 5868
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6872054348538064,
+      "learning_rate": 3.1668909096222254e-06,
+      "loss": 0.8332,
+      "step": 5869
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6314366293187808,
+      "learning_rate": 3.1652711783300234e-06,
+      "loss": 0.9802,
+      "step": 5870
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7319075843527971,
+      "learning_rate": 3.163651669505017e-06,
+      "loss": 0.9108,
+      "step": 5871
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6905214576276417,
+      "learning_rate": 3.1620323833435763e-06,
+      "loss": 0.9665,
+      "step": 5872
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6689876058058062,
+      "learning_rate": 3.160413320042045e-06,
+      "loss": 0.8354,
+      "step": 5873
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7242511033724275,
+      "learning_rate": 3.15879447979674e-06,
+      "loss": 0.9233,
+      "step": 5874
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7616438454034957,
+      "learning_rate": 3.157175862803953e-06,
+      "loss": 0.8893,
+      "step": 5875
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7910306081815668,
+      "learning_rate": 3.155557469259946e-06,
+      "loss": 0.8419,
+      "step": 5876
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8121446284266336,
+      "learning_rate": 3.153939299360956e-06,
+      "loss": 0.9873,
+      "step": 5877
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6781771244354162,
+      "learning_rate": 3.1523213533031884e-06,
+      "loss": 0.897,
+      "step": 5878
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7116553834123547,
+      "learning_rate": 3.150703631282826e-06,
+      "loss": 0.8951,
+      "step": 5879
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8551495244832301,
+      "learning_rate": 3.149086133496025e-06,
+      "loss": 0.9029,
+      "step": 5880
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7034610374779258,
+      "learning_rate": 3.1474688601389113e-06,
+      "loss": 0.8916,
+      "step": 5881
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7071151458351785,
+      "learning_rate": 3.145851811407584e-06,
+      "loss": 0.8459,
+      "step": 5882
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7485908135064498,
+      "learning_rate": 3.144234987498117e-06,
+      "loss": 0.9286,
+      "step": 5883
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.977034554165767,
+      "learning_rate": 3.1426183886065565e-06,
+      "loss": 0.947,
+      "step": 5884
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.702660258914151,
+      "learning_rate": 3.141002014928918e-06,
+      "loss": 0.8267,
+      "step": 5885
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6175068476557635,
+      "learning_rate": 3.1393858666611958e-06,
+      "loss": 0.8594,
+      "step": 5886
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7277996217375142,
+      "learning_rate": 3.1377699439993525e-06,
+      "loss": 0.9464,
+      "step": 5887
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6841837664110364,
+      "learning_rate": 3.1361542471393232e-06,
+      "loss": 0.8136,
+      "step": 5888
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6172030697683588,
+      "learning_rate": 3.1345387762770165e-06,
+      "loss": 0.9168,
+      "step": 5889
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7724043498871818,
+      "learning_rate": 3.132923531608313e-06,
+      "loss": 0.9275,
+      "step": 5890
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8871575303102432,
+      "learning_rate": 3.13130851332907e-06,
+      "loss": 1.0896,
+      "step": 5891
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8027824178855458,
+      "learning_rate": 3.1296937216351115e-06,
+      "loss": 0.8817,
+      "step": 5892
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6955780874868664,
+      "learning_rate": 3.128079156722236e-06,
+      "loss": 0.9195,
+      "step": 5893
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8026061750144518,
+      "learning_rate": 3.126464818786218e-06,
+      "loss": 0.878,
+      "step": 5894
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8456182732961544,
+      "learning_rate": 3.124850708022799e-06,
+      "loss": 0.9758,
+      "step": 5895
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8926715493755302,
+      "learning_rate": 3.1232368246276956e-06,
+      "loss": 0.8442,
+      "step": 5896
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8223375152386583,
+      "learning_rate": 3.121623168796598e-06,
+      "loss": 0.8765,
+      "step": 5897
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7724458824086222,
+      "learning_rate": 3.120009740725169e-06,
+      "loss": 0.904,
+      "step": 5898
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7367399672833368,
+      "learning_rate": 3.118396540609038e-06,
+      "loss": 0.9019,
+      "step": 5899
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6642235151955018,
+      "learning_rate": 3.116783568643814e-06,
+      "loss": 0.8635,
+      "step": 5900
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7740724111473123,
+      "learning_rate": 3.115170825025074e-06,
+      "loss": 0.8207,
+      "step": 5901
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7005533921918339,
+      "learning_rate": 3.1135583099483703e-06,
+      "loss": 0.8672,
+      "step": 5902
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8503859378528282,
+      "learning_rate": 3.1119460236092246e-06,
+      "loss": 0.8917,
+      "step": 5903
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8095082068170109,
+      "learning_rate": 3.1103339662031328e-06,
+      "loss": 1.0401,
+      "step": 5904
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7662652192247699,
+      "learning_rate": 3.1087221379255615e-06,
+      "loss": 0.9392,
+      "step": 5905
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9375254254121423,
+      "learning_rate": 3.1071105389719523e-06,
+      "loss": 0.9165,
+      "step": 5906
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7984812071496903,
+      "learning_rate": 3.1054991695377156e-06,
+      "loss": 0.8915,
+      "step": 5907
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.707407525132162,
+      "learning_rate": 3.103888029818235e-06,
+      "loss": 0.8661,
+      "step": 5908
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8170398232572122,
+      "learning_rate": 3.102277120008871e-06,
+      "loss": 0.9539,
+      "step": 5909
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6988510130241157,
+      "learning_rate": 3.100666440304946e-06,
+      "loss": 0.9478,
+      "step": 5910
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5817312829675299,
+      "learning_rate": 3.0990559909017635e-06,
+      "loss": 0.8782,
+      "step": 5911
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8304123532334345,
+      "learning_rate": 3.0974457719945953e-06,
+      "loss": 1.0577,
+      "step": 5912
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7057524585909356,
+      "learning_rate": 3.095835783778686e-06,
+      "loss": 0.9172,
+      "step": 5913
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.780023688449989,
+      "learning_rate": 3.0942260264492537e-06,
+      "loss": 0.8883,
+      "step": 5914
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8276461051412023,
+      "learning_rate": 3.0926165002014854e-06,
+      "loss": 1.0582,
+      "step": 5915
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6817030771110815,
+      "learning_rate": 3.091007205230541e-06,
+      "loss": 0.8155,
+      "step": 5916
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.71056639917109,
+      "learning_rate": 3.0893981417315556e-06,
+      "loss": 0.9186,
+      "step": 5917
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7902469432423358,
+      "learning_rate": 3.0877893098996324e-06,
+      "loss": 0.9168,
+      "step": 5918
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8247880020586374,
+      "learning_rate": 3.086180709929849e-06,
+      "loss": 1.0401,
+      "step": 5919
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8095971890401104,
+      "learning_rate": 3.0845723420172514e-06,
+      "loss": 0.9764,
+      "step": 5920
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6950450251188678,
+      "learning_rate": 3.08296420635686e-06,
+      "loss": 0.9544,
+      "step": 5921
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0910220658295766,
+      "learning_rate": 3.0813563031436676e-06,
+      "loss": 0.9251,
+      "step": 5922
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6838286193708225,
+      "learning_rate": 3.079748632572639e-06,
+      "loss": 0.9142,
+      "step": 5923
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7032773692498633,
+      "learning_rate": 3.0781411948387074e-06,
+      "loss": 0.9118,
+      "step": 5924
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.918452891098777,
+      "learning_rate": 3.0765339901367826e-06,
+      "loss": 0.9139,
+      "step": 5925
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7036606501229774,
+      "learning_rate": 3.074927018661743e-06,
+      "loss": 0.8939,
+      "step": 5926
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.616987712187354,
+      "learning_rate": 3.073320280608437e-06,
+      "loss": 0.9406,
+      "step": 5927
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8117324214979824,
+      "learning_rate": 3.071713776171692e-06,
+      "loss": 0.8702,
+      "step": 5928
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.879351501104517,
+      "learning_rate": 3.070107505546298e-06,
+      "loss": 0.9429,
+      "step": 5929
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.731158276582898,
+      "learning_rate": 3.0685014689270244e-06,
+      "loss": 0.7974,
+      "step": 5930
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6794122072728714,
+      "learning_rate": 3.0668956665086047e-06,
+      "loss": 0.8496,
+      "step": 5931
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8182068221685521,
+      "learning_rate": 3.0652900984857492e-06,
+      "loss": 0.9865,
+      "step": 5932
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7515375859336404,
+      "learning_rate": 3.0636847650531404e-06,
+      "loss": 0.8685,
+      "step": 5933
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6439952511456322,
+      "learning_rate": 3.0620796664054294e-06,
+      "loss": 0.8575,
+      "step": 5934
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7981703670205724,
+      "learning_rate": 3.0604748027372393e-06,
+      "loss": 0.9268,
+      "step": 5935
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7069045138522576,
+      "learning_rate": 3.0588701742431654e-06,
+      "loss": 0.9177,
+      "step": 5936
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7098471859973892,
+      "learning_rate": 3.057265781117776e-06,
+      "loss": 0.9167,
+      "step": 5937
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7173485875129368,
+      "learning_rate": 3.055661623555608e-06,
+      "loss": 0.861,
+      "step": 5938
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.69203311469032,
+      "learning_rate": 3.0540577017511703e-06,
+      "loss": 0.8739,
+      "step": 5939
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8165688591674226,
+      "learning_rate": 3.0524540158989478e-06,
+      "loss": 1.0237,
+      "step": 5940
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7626798376281505,
+      "learning_rate": 3.0508505661933873e-06,
+      "loss": 0.9922,
+      "step": 5941
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.725256495051611,
+      "learning_rate": 3.049247352828917e-06,
+      "loss": 0.8986,
+      "step": 5942
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8081005236479097,
+      "learning_rate": 3.0476443759999296e-06,
+      "loss": 0.9254,
+      "step": 5943
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8188434452971831,
+      "learning_rate": 3.0460416359007917e-06,
+      "loss": 0.967,
+      "step": 5944
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6943497285145859,
+      "learning_rate": 3.0444391327258425e-06,
+      "loss": 0.867,
+      "step": 5945
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7926859645603526,
+      "learning_rate": 3.0428368666693907e-06,
+      "loss": 1.0175,
+      "step": 5946
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7407891648600382,
+      "learning_rate": 3.041234837925715e-06,
+      "loss": 0.937,
+      "step": 5947
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7844941206071652,
+      "learning_rate": 3.039633046689069e-06,
+      "loss": 0.8568,
+      "step": 5948
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6509290964576664,
+      "learning_rate": 3.038031493153675e-06,
+      "loss": 0.8999,
+      "step": 5949
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8123154619729146,
+      "learning_rate": 3.0364301775137246e-06,
+      "loss": 1.046,
+      "step": 5950
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7764822099963673,
+      "learning_rate": 3.0348290999633877e-06,
+      "loss": 0.9426,
+      "step": 5951
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6771879028974497,
+      "learning_rate": 3.033228260696795e-06,
+      "loss": 0.9158,
+      "step": 5952
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6875834981263063,
+      "learning_rate": 3.031627659908057e-06,
+      "loss": 0.8935,
+      "step": 5953
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7973580216136049,
+      "learning_rate": 3.0300272977912516e-06,
+      "loss": 1.0494,
+      "step": 5954
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6045989277864892,
+      "learning_rate": 3.028427174540426e-06,
+      "loss": 0.8222,
+      "step": 5955
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8550055334470514,
+      "learning_rate": 3.0268272903496036e-06,
+      "loss": 0.9832,
+      "step": 5956
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7728818648426052,
+      "learning_rate": 3.0252276454127753e-06,
+      "loss": 0.8825,
+      "step": 5957
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7010165986880494,
+      "learning_rate": 3.023628239923902e-06,
+      "loss": 0.817,
+      "step": 5958
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8348257350184076,
+      "learning_rate": 3.0220290740769193e-06,
+      "loss": 1.0267,
+      "step": 5959
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7380103659541342,
+      "learning_rate": 3.0204301480657306e-06,
+      "loss": 0.8772,
+      "step": 5960
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7714315071128273,
+      "learning_rate": 3.0188314620842097e-06,
+      "loss": 0.9352,
+      "step": 5961
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8143390478487336,
+      "learning_rate": 3.0172330163262072e-06,
+      "loss": 1.016,
+      "step": 5962
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7505343289147601,
+      "learning_rate": 3.0156348109855337e-06,
+      "loss": 0.9167,
+      "step": 5963
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8094075966124814,
+      "learning_rate": 3.014036846255982e-06,
+      "loss": 0.8797,
+      "step": 5964
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7562080751011817,
+      "learning_rate": 3.0124391223313093e-06,
+      "loss": 0.9049,
+      "step": 5965
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7706505604968563,
+      "learning_rate": 3.0108416394052443e-06,
+      "loss": 0.9225,
+      "step": 5966
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7200430098579997,
+      "learning_rate": 3.009244397671489e-06,
+      "loss": 0.822,
+      "step": 5967
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8031389733107772,
+      "learning_rate": 3.007647397323714e-06,
+      "loss": 1.0237,
+      "step": 5968
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7894576419525712,
+      "learning_rate": 3.00605063855556e-06,
+      "loss": 0.8858,
+      "step": 5969
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7302050680397307,
+      "learning_rate": 3.0044541215606398e-06,
+      "loss": 0.8412,
+      "step": 5970
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7829716701145696,
+      "learning_rate": 3.002857846532538e-06,
+      "loss": 1.0205,
+      "step": 5971
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.679517764147183,
+      "learning_rate": 3.0012618136648085e-06,
+      "loss": 0.9287,
+      "step": 5972
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.840580883876344,
+      "learning_rate": 2.9996660231509742e-06,
+      "loss": 0.8858,
+      "step": 5973
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6860806230385499,
+      "learning_rate": 2.9980704751845302e-06,
+      "loss": 0.8518,
+      "step": 5974
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7400871044752215,
+      "learning_rate": 2.996475169958942e-06,
+      "loss": 0.94,
+      "step": 5975
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7440812409815565,
+      "learning_rate": 2.9948801076676477e-06,
+      "loss": 0.9361,
+      "step": 5976
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.737919860365998,
+      "learning_rate": 2.993285288504053e-06,
+      "loss": 0.9233,
+      "step": 5977
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7584902462603629,
+      "learning_rate": 2.9916907126615343e-06,
+      "loss": 0.9503,
+      "step": 5978
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6395309564306122,
+      "learning_rate": 2.9900963803334417e-06,
+      "loss": 0.8621,
+      "step": 5979
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7413119372710852,
+      "learning_rate": 2.9885022917130914e-06,
+      "loss": 0.8919,
+      "step": 5980
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6809644726035522,
+      "learning_rate": 2.986908446993772e-06,
+      "loss": 0.8671,
+      "step": 5981
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6494030906710535,
+      "learning_rate": 2.9853148463687455e-06,
+      "loss": 0.9215,
+      "step": 5982
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.738825957280495,
+      "learning_rate": 2.9837214900312396e-06,
+      "loss": 0.8628,
+      "step": 5983
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8068360252179506,
+      "learning_rate": 2.9821283781744537e-06,
+      "loss": 1.0391,
+      "step": 5984
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.625075188469242,
+      "learning_rate": 2.980535510991558e-06,
+      "loss": 0.8903,
+      "step": 5985
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6781251703528512,
+      "learning_rate": 2.978942888675693e-06,
+      "loss": 0.8902,
+      "step": 5986
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7113889524109467,
+      "learning_rate": 2.977350511419972e-06,
+      "loss": 0.8454,
+      "step": 5987
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8034136801238725,
+      "learning_rate": 2.9757583794174744e-06,
+      "loss": 0.8418,
+      "step": 5988
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6942994468917052,
+      "learning_rate": 2.9741664928612506e-06,
+      "loss": 0.8373,
+      "step": 5989
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.759702015861477,
+      "learning_rate": 2.972574851944325e-06,
+      "loss": 0.9098,
+      "step": 5990
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.73186605271841,
+      "learning_rate": 2.9709834568596873e-06,
+      "loss": 0.8729,
+      "step": 5991
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6970701558848489,
+      "learning_rate": 2.9693923078003005e-06,
+      "loss": 0.899,
+      "step": 5992
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7270454880222068,
+      "learning_rate": 2.9678014049591e-06,
+      "loss": 0.9201,
+      "step": 5993
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.789339017907301,
+      "learning_rate": 2.9662107485289815e-06,
+      "loss": 0.8291,
+      "step": 5994
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7770931892178092,
+      "learning_rate": 2.964620338702823e-06,
+      "loss": 0.874,
+      "step": 5995
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8974762030576109,
+      "learning_rate": 2.9630301756734646e-06,
+      "loss": 0.9253,
+      "step": 5996
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.740669981591448,
+      "learning_rate": 2.96144025963372e-06,
+      "loss": 0.8837,
+      "step": 5997
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6977449563437372,
+      "learning_rate": 2.959850590776372e-06,
+      "loss": 0.9595,
+      "step": 5998
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7593274925550832,
+      "learning_rate": 2.958261169294174e-06,
+      "loss": 0.8789,
+      "step": 5999
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7578438580062556,
+      "learning_rate": 2.9566719953798474e-06,
+      "loss": 0.8369,
+      "step": 6000
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7524473715110551,
+      "learning_rate": 2.9550830692260856e-06,
+      "loss": 0.8574,
+      "step": 6001
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7793799471767124,
+      "learning_rate": 2.953494391025552e-06,
+      "loss": 0.9759,
+      "step": 6002
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7594154263664985,
+      "learning_rate": 2.951905960970879e-06,
+      "loss": 0.9248,
+      "step": 6003
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7603678001473364,
+      "learning_rate": 2.95031777925467e-06,
+      "loss": 0.8848,
+      "step": 6004
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.809113419279355,
+      "learning_rate": 2.9487298460694955e-06,
+      "loss": 0.9269,
+      "step": 6005
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.746570422010187,
+      "learning_rate": 2.9471421616078987e-06,
+      "loss": 0.8938,
+      "step": 6006
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6448328629558564,
+      "learning_rate": 2.9455547260623926e-06,
+      "loss": 0.8646,
+      "step": 6007
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7160312740831634,
+      "learning_rate": 2.9439675396254586e-06,
+      "loss": 0.9686,
+      "step": 6008
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7851820131759302,
+      "learning_rate": 2.9423806024895475e-06,
+      "loss": 0.9191,
+      "step": 6009
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7940048909883966,
+      "learning_rate": 2.9407939148470834e-06,
+      "loss": 0.9155,
+      "step": 6010
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8216422592687966,
+      "learning_rate": 2.939207476890456e-06,
+      "loss": 1.0329,
+      "step": 6011
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7354790876734008,
+      "learning_rate": 2.937621288812026e-06,
+      "loss": 0.9656,
+      "step": 6012
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7888611400271068,
+      "learning_rate": 2.9360353508041257e-06,
+      "loss": 0.8988,
+      "step": 6013
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7453972062441918,
+      "learning_rate": 2.9344496630590548e-06,
+      "loss": 0.8861,
+      "step": 6014
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.618093670826447,
+      "learning_rate": 2.9328642257690835e-06,
+      "loss": 0.8745,
+      "step": 6015
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7286287431528895,
+      "learning_rate": 2.931279039126451e-06,
+      "loss": 0.9263,
+      "step": 6016
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7158371445740197,
+      "learning_rate": 2.9296941033233662e-06,
+      "loss": 0.8541,
+      "step": 6017
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7616907820192784,
+      "learning_rate": 2.9281094185520097e-06,
+      "loss": 0.9177,
+      "step": 6018
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8819569321560612,
+      "learning_rate": 2.9265249850045284e-06,
+      "loss": 0.9371,
+      "step": 6019
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.74306855507922,
+      "learning_rate": 2.924940802873041e-06,
+      "loss": 0.8426,
+      "step": 6020
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6661061955898464,
+      "learning_rate": 2.9233568723496354e-06,
+      "loss": 0.9155,
+      "step": 6021
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6578804687695754,
+      "learning_rate": 2.9217731936263683e-06,
+      "loss": 0.8616,
+      "step": 6022
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7127656192611829,
+      "learning_rate": 2.9201897668952655e-06,
+      "loss": 0.912,
+      "step": 6023
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8235003911092991,
+      "learning_rate": 2.918606592348324e-06,
+      "loss": 1.0367,
+      "step": 6024
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8349585148487906,
+      "learning_rate": 2.9170236701775113e-06,
+      "loss": 1.0247,
+      "step": 6025
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7789840478179613,
+      "learning_rate": 2.9154410005747586e-06,
+      "loss": 0.8222,
+      "step": 6026
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.765453500414764,
+      "learning_rate": 2.9138585837319694e-06,
+      "loss": 0.8439,
+      "step": 6027
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8536606188868114,
+      "learning_rate": 2.912276419841019e-06,
+      "loss": 0.9478,
+      "step": 6028
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8497212071649534,
+      "learning_rate": 2.9106945090937524e-06,
+      "loss": 0.913,
+      "step": 6029
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.024634847031988,
+      "learning_rate": 2.9091128516819767e-06,
+      "loss": 0.8534,
+      "step": 6030
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8227162172449467,
+      "learning_rate": 2.9075314477974764e-06,
+      "loss": 0.9863,
+      "step": 6031
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6940789843880182,
+      "learning_rate": 2.905950297632004e-06,
+      "loss": 0.8558,
+      "step": 6032
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6851608904491584,
+      "learning_rate": 2.9043694013772756e-06,
+      "loss": 0.8513,
+      "step": 6033
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7066822500640777,
+      "learning_rate": 2.9027887592249817e-06,
+      "loss": 0.9334,
+      "step": 6034
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7286416496596715,
+      "learning_rate": 2.9012083713667833e-06,
+      "loss": 0.911,
+      "step": 6035
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7863654302441503,
+      "learning_rate": 2.8996282379943053e-06,
+      "loss": 0.9281,
+      "step": 6036
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7557832930313828,
+      "learning_rate": 2.8980483592991426e-06,
+      "loss": 0.9111,
+      "step": 6037
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7098172894974268,
+      "learning_rate": 2.896468735472863e-06,
+      "loss": 0.86,
+      "step": 6038
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7804800660102624,
+      "learning_rate": 2.894889366707002e-06,
+      "loss": 0.9099,
+      "step": 6039
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7691636988629234,
+      "learning_rate": 2.8933102531930622e-06,
+      "loss": 1.0019,
+      "step": 6040
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8042255133971408,
+      "learning_rate": 2.8917313951225164e-06,
+      "loss": 0.8375,
+      "step": 6041
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7254406629344132,
+      "learning_rate": 2.8901527926868088e-06,
+      "loss": 0.9047,
+      "step": 6042
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.775340304987951,
+      "learning_rate": 2.888574446077348e-06,
+      "loss": 0.8659,
+      "step": 6043
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8143593539774812,
+      "learning_rate": 2.886996355485514e-06,
+      "loss": 0.9103,
+      "step": 6044
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7762774877316694,
+      "learning_rate": 2.885418521102659e-06,
+      "loss": 0.8831,
+      "step": 6045
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.61466674779323,
+      "learning_rate": 2.8838409431200976e-06,
+      "loss": 0.8773,
+      "step": 6046
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9349153091569657,
+      "learning_rate": 2.882263621729116e-06,
+      "loss": 0.8956,
+      "step": 6047
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.858944405783447,
+      "learning_rate": 2.880686557120973e-06,
+      "loss": 0.8307,
+      "step": 6048
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.628088203636189,
+      "learning_rate": 2.8791097494868896e-06,
+      "loss": 0.8353,
+      "step": 6049
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5899853853130876,
+      "learning_rate": 2.877533199018061e-06,
+      "loss": 0.8783,
+      "step": 6050
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8280499929353247,
+      "learning_rate": 2.875956905905651e-06,
+      "loss": 0.8927,
+      "step": 6051
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7767763436892685,
+      "learning_rate": 2.8743808703407866e-06,
+      "loss": 0.9425,
+      "step": 6052
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8142903115228977,
+      "learning_rate": 2.87280509251457e-06,
+      "loss": 0.9859,
+      "step": 6053
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6492030419701122,
+      "learning_rate": 2.8712295726180717e-06,
+      "loss": 0.8989,
+      "step": 6054
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8452905252490812,
+      "learning_rate": 2.8696543108423246e-06,
+      "loss": 0.8865,
+      "step": 6055
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.832295767403557,
+      "learning_rate": 2.8680793073783363e-06,
+      "loss": 1.0378,
+      "step": 6056
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.70554044412064,
+      "learning_rate": 2.866504562417086e-06,
+      "loss": 0.8911,
+      "step": 6057
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.778417207166715,
+      "learning_rate": 2.864930076149509e-06,
+      "loss": 0.9217,
+      "step": 6058
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7198575500235072,
+      "learning_rate": 2.8633558487665215e-06,
+      "loss": 0.8713,
+      "step": 6059
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.634632637012075,
+      "learning_rate": 2.861781880459005e-06,
+      "loss": 0.8268,
+      "step": 6060
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9047645191316849,
+      "learning_rate": 2.8602081714178043e-06,
+      "loss": 0.8769,
+      "step": 6061
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7232663563317052,
+      "learning_rate": 2.8586347218337406e-06,
+      "loss": 0.9415,
+      "step": 6062
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6429387289573003,
+      "learning_rate": 2.8570615318976013e-06,
+      "loss": 0.8474,
+      "step": 6063
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.738225941033339,
+      "learning_rate": 2.855488601800137e-06,
+      "loss": 0.8276,
+      "step": 6064
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.630078377791851,
+      "learning_rate": 2.853915931732073e-06,
+      "loss": 0.875,
+      "step": 6065
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6769466404880535,
+      "learning_rate": 2.852343521884103e-06,
+      "loss": 0.9314,
+      "step": 6066
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8019745524330946,
+      "learning_rate": 2.850771372446884e-06,
+      "loss": 0.9682,
+      "step": 6067
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.664919311853383,
+      "learning_rate": 2.8491994836110454e-06,
+      "loss": 0.9121,
+      "step": 6068
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8311633682934464,
+      "learning_rate": 2.8476278555671834e-06,
+      "loss": 1.0422,
+      "step": 6069
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9697373497803674,
+      "learning_rate": 2.846056488505866e-06,
+      "loss": 0.8938,
+      "step": 6070
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6921412618593161,
+      "learning_rate": 2.844485382617624e-06,
+      "loss": 0.8738,
+      "step": 6071
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.672874282792823,
+      "learning_rate": 2.84291453809296e-06,
+      "loss": 0.9525,
+      "step": 6072
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7222164872068368,
+      "learning_rate": 2.8413439551223472e-06,
+      "loss": 0.8762,
+      "step": 6073
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7768337113099122,
+      "learning_rate": 2.839773633896219e-06,
+      "loss": 0.9038,
+      "step": 6074
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6849617734951094,
+      "learning_rate": 2.8382035746049863e-06,
+      "loss": 0.865,
+      "step": 6075
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.775290582768773,
+      "learning_rate": 2.836633777439024e-06,
+      "loss": 0.9265,
+      "step": 6076
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6670501314048238,
+      "learning_rate": 2.8350642425886734e-06,
+      "loss": 0.9609,
+      "step": 6077
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.749364522715793,
+      "learning_rate": 2.833494970244248e-06,
+      "loss": 0.8145,
+      "step": 6078
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6736627553678913,
+      "learning_rate": 2.831925960596027e-06,
+      "loss": 0.9642,
+      "step": 6079
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8997364502248302,
+      "learning_rate": 2.830357213834256e-06,
+      "loss": 0.8854,
+      "step": 6080
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7107816666517686,
+      "learning_rate": 2.8287887301491522e-06,
+      "loss": 0.9282,
+      "step": 6081
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6829469479402708,
+      "learning_rate": 2.827220509730903e-06,
+      "loss": 0.9168,
+      "step": 6082
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.686013133551253,
+      "learning_rate": 2.8256525527696554e-06,
+      "loss": 0.8921,
+      "step": 6083
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6752289618941414,
+      "learning_rate": 2.8240848594555314e-06,
+      "loss": 0.9276,
+      "step": 6084
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5764214290328076,
+      "learning_rate": 2.822517429978622e-06,
+      "loss": 0.8378,
+      "step": 6085
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6422650037291362,
+      "learning_rate": 2.8209502645289795e-06,
+      "loss": 0.8236,
+      "step": 6086
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7487957223460724,
+      "learning_rate": 2.8193833632966296e-06,
+      "loss": 0.8647,
+      "step": 6087
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8092256444254193,
+      "learning_rate": 2.817816726471566e-06,
+      "loss": 0.9285,
+      "step": 6088
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8236141328707458,
+      "learning_rate": 2.816250354243748e-06,
+      "loss": 1.0566,
+      "step": 6089
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6693660660672482,
+      "learning_rate": 2.814684246803101e-06,
+      "loss": 0.8262,
+      "step": 6090
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7125257832949223,
+      "learning_rate": 2.8131184043395256e-06,
+      "loss": 0.8362,
+      "step": 6091
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7759666143815545,
+      "learning_rate": 2.8115528270428798e-06,
+      "loss": 0.9318,
+      "step": 6092
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.589969089918983,
+      "learning_rate": 2.8099875151029987e-06,
+      "loss": 0.8193,
+      "step": 6093
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8961872838356477,
+      "learning_rate": 2.808422468709684e-06,
+      "loss": 0.8918,
+      "step": 6094
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6670071855491393,
+      "learning_rate": 2.8068576880526986e-06,
+      "loss": 0.8798,
+      "step": 6095
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7368612222942021,
+      "learning_rate": 2.805293173321779e-06,
+      "loss": 0.9518,
+      "step": 6096
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8193291967282579,
+      "learning_rate": 2.8037289247066307e-06,
+      "loss": 0.9421,
+      "step": 6097
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7154085092020732,
+      "learning_rate": 2.8021649423969198e-06,
+      "loss": 0.8796,
+      "step": 6098
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8348735251047712,
+      "learning_rate": 2.800601226582289e-06,
+      "loss": 0.9242,
+      "step": 6099
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6148475010514527,
+      "learning_rate": 2.7990377774523397e-06,
+      "loss": 0.8657,
+      "step": 6100
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7385997736522063,
+      "learning_rate": 2.79747459519665e-06,
+      "loss": 0.8558,
+      "step": 6101
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7598897283801989,
+      "learning_rate": 2.795911680004757e-06,
+      "loss": 0.8774,
+      "step": 6102
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6602601522277274,
+      "learning_rate": 2.7943490320661703e-06,
+      "loss": 0.8586,
+      "step": 6103
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.635301117132657,
+      "learning_rate": 2.7927866515703705e-06,
+      "loss": 0.7736,
+      "step": 6104
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7754370969891193,
+      "learning_rate": 2.7912245387067972e-06,
+      "loss": 0.9574,
+      "step": 6105
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7642830933516673,
+      "learning_rate": 2.7896626936648622e-06,
+      "loss": 0.8804,
+      "step": 6106
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6572626116735314,
+      "learning_rate": 2.7881011166339485e-06,
+      "loss": 0.8855,
+      "step": 6107
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7311138507231523,
+      "learning_rate": 2.7865398078033976e-06,
+      "loss": 0.9892,
+      "step": 6108
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7606384982214118,
+      "learning_rate": 2.784978767362528e-06,
+      "loss": 0.9203,
+      "step": 6109
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6820993322318152,
+      "learning_rate": 2.783417995500619e-06,
+      "loss": 0.9415,
+      "step": 6110
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7114924009000194,
+      "learning_rate": 2.7818574924069184e-06,
+      "loss": 0.8152,
+      "step": 6111
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.75787421656962,
+      "learning_rate": 2.7802972582706434e-06,
+      "loss": 0.8778,
+      "step": 6112
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7979792417167832,
+      "learning_rate": 2.778737293280981e-06,
+      "loss": 1.012,
+      "step": 6113
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7866699431767055,
+      "learning_rate": 2.777177597627077e-06,
+      "loss": 0.9008,
+      "step": 6114
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8327443518825703,
+      "learning_rate": 2.7756181714980525e-06,
+      "loss": 1.0168,
+      "step": 6115
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.828361947255035,
+      "learning_rate": 2.774059015082996e-06,
+      "loss": 0.8671,
+      "step": 6116
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7077204830703197,
+      "learning_rate": 2.772500128570955e-06,
+      "loss": 0.8736,
+      "step": 6117
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.780330384818927,
+      "learning_rate": 2.770941512150953e-06,
+      "loss": 0.9395,
+      "step": 6118
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8020072770312363,
+      "learning_rate": 2.76938316601198e-06,
+      "loss": 0.8437,
+      "step": 6119
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.830277331522214,
+      "learning_rate": 2.7678250903429875e-06,
+      "loss": 0.8965,
+      "step": 6120
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7055387144444438,
+      "learning_rate": 2.7662672853328966e-06,
+      "loss": 0.8392,
+      "step": 6121
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6717683683712818,
+      "learning_rate": 2.7647097511706e-06,
+      "loss": 0.8614,
+      "step": 6122
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.702798164971416,
+      "learning_rate": 2.7631524880449513e-06,
+      "loss": 0.9627,
+      "step": 6123
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.831077564777693,
+      "learning_rate": 2.7615954961447733e-06,
+      "loss": 0.8586,
+      "step": 6124
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6853325022880683,
+      "learning_rate": 2.7600387756588616e-06,
+      "loss": 0.9364,
+      "step": 6125
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8018666698121288,
+      "learning_rate": 2.7584823267759686e-06,
+      "loss": 0.799,
+      "step": 6126
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7595746619995463,
+      "learning_rate": 2.7569261496848203e-06,
+      "loss": 0.925,
+      "step": 6127
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8196257907160969,
+      "learning_rate": 2.7553702445741125e-06,
+      "loss": 1.0466,
+      "step": 6128
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.806424488256776,
+      "learning_rate": 2.753814611632498e-06,
+      "loss": 0.8055,
+      "step": 6129
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.700399070990565,
+      "learning_rate": 2.752259251048606e-06,
+      "loss": 0.8753,
+      "step": 6130
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6877855745986927,
+      "learning_rate": 2.750704163011031e-06,
+      "loss": 0.8591,
+      "step": 6131
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7878932804134957,
+      "learning_rate": 2.749149347708331e-06,
+      "loss": 0.9575,
+      "step": 6132
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8296910250790603,
+      "learning_rate": 2.7475948053290302e-06,
+      "loss": 0.9094,
+      "step": 6133
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.660713281541635,
+      "learning_rate": 2.7460405360616243e-06,
+      "loss": 0.9381,
+      "step": 6134
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7339096091046362,
+      "learning_rate": 2.7444865400945763e-06,
+      "loss": 0.9092,
+      "step": 6135
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7005716654999283,
+      "learning_rate": 2.7429328176163088e-06,
+      "loss": 0.8896,
+      "step": 6136
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7424268511699303,
+      "learning_rate": 2.741379368815219e-06,
+      "loss": 0.8578,
+      "step": 6137
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7134153301420674,
+      "learning_rate": 2.7398261938796688e-06,
+      "loss": 0.8535,
+      "step": 6138
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7039167623203721,
+      "learning_rate": 2.7382732929979833e-06,
+      "loss": 0.846,
+      "step": 6139
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6881739499677193,
+      "learning_rate": 2.7367206663584578e-06,
+      "loss": 0.9552,
+      "step": 6140
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7171640333496894,
+      "learning_rate": 2.7351683141493566e-06,
+      "loss": 0.8075,
+      "step": 6141
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8379567687547802,
+      "learning_rate": 2.733616236558905e-06,
+      "loss": 0.8996,
+      "step": 6142
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6275172951458954,
+      "learning_rate": 2.7320644337752966e-06,
+      "loss": 0.8432,
+      "step": 6143
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6712431521403892,
+      "learning_rate": 2.7305129059866956e-06,
+      "loss": 0.8203,
+      "step": 6144
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7396737297790197,
+      "learning_rate": 2.728961653381228e-06,
+      "loss": 0.9509,
+      "step": 6145
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8209484307315307,
+      "learning_rate": 2.7274106761469887e-06,
+      "loss": 0.9298,
+      "step": 6146
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9527494138888886,
+      "learning_rate": 2.7258599744720416e-06,
+      "loss": 0.9489,
+      "step": 6147
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7011918620207098,
+      "learning_rate": 2.7243095485444103e-06,
+      "loss": 0.8481,
+      "step": 6148
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8191168547401848,
+      "learning_rate": 2.722759398552093e-06,
+      "loss": 0.8755,
+      "step": 6149
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7166966132940573,
+      "learning_rate": 2.7212095246830505e-06,
+      "loss": 0.8506,
+      "step": 6150
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7393814112851602,
+      "learning_rate": 2.719659927125208e-06,
+      "loss": 0.8868,
+      "step": 6151
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6628590426973615,
+      "learning_rate": 2.718110606066462e-06,
+      "loss": 0.8747,
+      "step": 6152
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8767285961531108,
+      "learning_rate": 2.7165615616946726e-06,
+      "loss": 0.9128,
+      "step": 6153
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8381295079296902,
+      "learning_rate": 2.715012794197664e-06,
+      "loss": 0.899,
+      "step": 6154
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8012285743553142,
+      "learning_rate": 2.7134643037632315e-06,
+      "loss": 0.8646,
+      "step": 6155
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7164834442786616,
+      "learning_rate": 2.711916090579137e-06,
+      "loss": 0.8701,
+      "step": 6156
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6802485651173003,
+      "learning_rate": 2.710368154833103e-06,
+      "loss": 0.8475,
+      "step": 6157
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.701378978988291,
+      "learning_rate": 2.7088204967128236e-06,
+      "loss": 0.8761,
+      "step": 6158
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8182565638409682,
+      "learning_rate": 2.7072731164059594e-06,
+      "loss": 0.9362,
+      "step": 6159
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6926862301343832,
+      "learning_rate": 2.7057260141001328e-06,
+      "loss": 0.9149,
+      "step": 6160
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6524880193801237,
+      "learning_rate": 2.704179189982936e-06,
+      "loss": 0.856,
+      "step": 6161
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7574193113715393,
+      "learning_rate": 2.70263264424193e-06,
+      "loss": 0.9467,
+      "step": 6162
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7510876589350153,
+      "learning_rate": 2.7010863770646357e-06,
+      "loss": 0.8388,
+      "step": 6163
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9298244623026277,
+      "learning_rate": 2.699540388638542e-06,
+      "loss": 0.9799,
+      "step": 6164
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6913001211307064,
+      "learning_rate": 2.6979946791511066e-06,
+      "loss": 0.8799,
+      "step": 6165
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7412542077277038,
+      "learning_rate": 2.696449248789754e-06,
+      "loss": 0.8942,
+      "step": 6166
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8776618189144318,
+      "learning_rate": 2.6949040977418706e-06,
+      "loss": 0.9305,
+      "step": 6167
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6211607056577284,
+      "learning_rate": 2.693359226194812e-06,
+      "loss": 0.9374,
+      "step": 6168
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6755559823174764,
+      "learning_rate": 2.691814634335902e-06,
+      "loss": 0.8806,
+      "step": 6169
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8616153380947458,
+      "learning_rate": 2.6902703223524216e-06,
+      "loss": 0.9068,
+      "step": 6170
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7474974771450928,
+      "learning_rate": 2.688726290431629e-06,
+      "loss": 0.9237,
+      "step": 6171
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6975096027185144,
+      "learning_rate": 2.687182538760743e-06,
+      "loss": 0.848,
+      "step": 6172
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7077989609981274,
+      "learning_rate": 2.6856390675269478e-06,
+      "loss": 0.86,
+      "step": 6173
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.609554740471953,
+      "learning_rate": 2.684095876917393e-06,
+      "loss": 0.9436,
+      "step": 6174
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6832056388017014,
+      "learning_rate": 2.6825529671191984e-06,
+      "loss": 0.9061,
+      "step": 6175
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.725019550304898,
+      "learning_rate": 2.681010338319444e-06,
+      "loss": 0.8383,
+      "step": 6176
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.678435866897065,
+      "learning_rate": 2.6794679907051817e-06,
+      "loss": 0.9535,
+      "step": 6177
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6690741678490348,
+      "learning_rate": 2.6779259244634272e-06,
+      "loss": 0.8366,
+      "step": 6178
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7510164623022888,
+      "learning_rate": 2.6763841397811576e-06,
+      "loss": 0.9418,
+      "step": 6179
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7566829317472894,
+      "learning_rate": 2.674842636845322e-06,
+      "loss": 0.9314,
+      "step": 6180
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.813481297400205,
+      "learning_rate": 2.673301415842835e-06,
+      "loss": 1.046,
+      "step": 6181
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6435635367913084,
+      "learning_rate": 2.67176047696057e-06,
+      "loss": 0.8794,
+      "step": 6182
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6996318007317546,
+      "learning_rate": 2.670219820385374e-06,
+      "loss": 0.8346,
+      "step": 6183
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7106808405689775,
+      "learning_rate": 2.6686794463040613e-06,
+      "loss": 0.9316,
+      "step": 6184
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.627998981607846,
+      "learning_rate": 2.6671393549033984e-06,
+      "loss": 0.8758,
+      "step": 6185
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6915832936319433,
+      "learning_rate": 2.665599546370131e-06,
+      "loss": 0.8529,
+      "step": 6186
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.715970477488259,
+      "learning_rate": 2.6640600208909697e-06,
+      "loss": 0.9359,
+      "step": 6187
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.1142375692527695,
+      "learning_rate": 2.662520778652581e-06,
+      "loss": 0.8449,
+      "step": 6188
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.781565197401211,
+      "learning_rate": 2.660981819841607e-06,
+      "loss": 0.8362,
+      "step": 6189
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6902480019670985,
+      "learning_rate": 2.6594431446446528e-06,
+      "loss": 0.8901,
+      "step": 6190
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6496129533624062,
+      "learning_rate": 2.657904753248285e-06,
+      "loss": 0.7987,
+      "step": 6191
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7316616197052805,
+      "learning_rate": 2.6563666458390396e-06,
+      "loss": 0.908,
+      "step": 6192
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7553495351245245,
+      "learning_rate": 2.65482882260342e-06,
+      "loss": 0.9397,
+      "step": 6193
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6934216274031424,
+      "learning_rate": 2.6532912837278896e-06,
+      "loss": 0.9114,
+      "step": 6194
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.823232650728939,
+      "learning_rate": 2.6517540293988832e-06,
+      "loss": 1.0912,
+      "step": 6195
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8743201258744652,
+      "learning_rate": 2.650217059802794e-06,
+      "loss": 1.0428,
+      "step": 6196
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.728273483205488,
+      "learning_rate": 2.648680375125989e-06,
+      "loss": 0.8976,
+      "step": 6197
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7852192510222362,
+      "learning_rate": 2.6471439755547934e-06,
+      "loss": 0.8204,
+      "step": 6198
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7876489853246944,
+      "learning_rate": 2.645607861275502e-06,
+      "loss": 0.9146,
+      "step": 6199
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7495815991025765,
+      "learning_rate": 2.644072032474376e-06,
+      "loss": 0.9586,
+      "step": 6200
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6563216523725741,
+      "learning_rate": 2.642536489337636e-06,
+      "loss": 0.9673,
+      "step": 6201
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8265955239181466,
+      "learning_rate": 2.6410012320514742e-06,
+      "loss": 0.8628,
+      "step": 6202
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7571581376497385,
+      "learning_rate": 2.639466260802048e-06,
+      "loss": 0.9667,
+      "step": 6203
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7925343909729858,
+      "learning_rate": 2.6379315757754744e-06,
+      "loss": 0.9068,
+      "step": 6204
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6086819683171814,
+      "learning_rate": 2.6363971771578417e-06,
+      "loss": 0.8885,
+      "step": 6205
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8424749613468725,
+      "learning_rate": 2.6348630651351997e-06,
+      "loss": 0.8389,
+      "step": 6206
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.854209122636025,
+      "learning_rate": 2.6333292398935636e-06,
+      "loss": 0.9361,
+      "step": 6207
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7259709078000067,
+      "learning_rate": 2.6317957016189155e-06,
+      "loss": 0.9704,
+      "step": 6208
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7515886533757585,
+      "learning_rate": 2.6302624504972054e-06,
+      "loss": 0.8861,
+      "step": 6209
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7301643400860627,
+      "learning_rate": 2.628729486714341e-06,
+      "loss": 0.8962,
+      "step": 6210
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6698217474603771,
+      "learning_rate": 2.627196810456201e-06,
+      "loss": 0.9055,
+      "step": 6211
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8099011066801103,
+      "learning_rate": 2.62566442190863e-06,
+      "loss": 1.037,
+      "step": 6212
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6395785259050835,
+      "learning_rate": 2.624132321257432e-06,
+      "loss": 0.9171,
+      "step": 6213
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6848556822820735,
+      "learning_rate": 2.622600508688381e-06,
+      "loss": 0.932,
+      "step": 6214
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7889213707058302,
+      "learning_rate": 2.621068984387216e-06,
+      "loss": 0.8899,
+      "step": 6215
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7345693881479076,
+      "learning_rate": 2.6195377485396377e-06,
+      "loss": 0.8741,
+      "step": 6216
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7477254963087112,
+      "learning_rate": 2.6180068013313133e-06,
+      "loss": 0.8369,
+      "step": 6217
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7926610085563496,
+      "learning_rate": 2.616476142947877e-06,
+      "loss": 1.0007,
+      "step": 6218
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6474256170837194,
+      "learning_rate": 2.614945773574924e-06,
+      "loss": 0.8531,
+      "step": 6219
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6661381948890501,
+      "learning_rate": 2.6134156933980186e-06,
+      "loss": 0.8473,
+      "step": 6220
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.813394203858817,
+      "learning_rate": 2.61188590260269e-06,
+      "loss": 0.9301,
+      "step": 6221
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.668248392688055,
+      "learning_rate": 2.610356401374427e-06,
+      "loss": 0.8517,
+      "step": 6222
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9081147575290032,
+      "learning_rate": 2.608827189898688e-06,
+      "loss": 0.9288,
+      "step": 6223
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.696113848464344,
+      "learning_rate": 2.6072982683608992e-06,
+      "loss": 0.9001,
+      "step": 6224
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7235785910801638,
+      "learning_rate": 2.6057696369464412e-06,
+      "loss": 0.8589,
+      "step": 6225
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6694403220797782,
+      "learning_rate": 2.6042412958406717e-06,
+      "loss": 0.8371,
+      "step": 6226
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.674340469936745,
+      "learning_rate": 2.6027132452289015e-06,
+      "loss": 0.8662,
+      "step": 6227
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5955827785615304,
+      "learning_rate": 2.601185485296418e-06,
+      "loss": 0.9183,
+      "step": 6228
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.698844018872994,
+      "learning_rate": 2.5996580162284618e-06,
+      "loss": 0.8316,
+      "step": 6229
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7184580310036721,
+      "learning_rate": 2.5981308382102467e-06,
+      "loss": 0.9349,
+      "step": 6230
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8048176256362345,
+      "learning_rate": 2.5966039514269494e-06,
+      "loss": 0.9832,
+      "step": 6231
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6756674030837515,
+      "learning_rate": 2.595077356063708e-06,
+      "loss": 0.919,
+      "step": 6232
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.668349327487681,
+      "learning_rate": 2.5935510523056268e-06,
+      "loss": 0.8892,
+      "step": 6233
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7747301334217365,
+      "learning_rate": 2.592025040337779e-06,
+      "loss": 0.8672,
+      "step": 6234
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8341675821986696,
+      "learning_rate": 2.5904993203451957e-06,
+      "loss": 0.8769,
+      "step": 6235
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6727155224561752,
+      "learning_rate": 2.5889738925128753e-06,
+      "loss": 0.8948,
+      "step": 6236
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6194755173972684,
+      "learning_rate": 2.587448757025787e-06,
+      "loss": 0.8658,
+      "step": 6237
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7618496047116412,
+      "learning_rate": 2.585923914068851e-06,
+      "loss": 0.9491,
+      "step": 6238
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.768186687704413,
+      "learning_rate": 2.5843993638269618e-06,
+      "loss": 0.9313,
+      "step": 6239
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9522179335423024,
+      "learning_rate": 2.58287510648498e-06,
+      "loss": 0.835,
+      "step": 6240
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8617733968542577,
+      "learning_rate": 2.5813511422277227e-06,
+      "loss": 1.0341,
+      "step": 6241
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7929784821512722,
+      "learning_rate": 2.579827471239978e-06,
+      "loss": 0.8373,
+      "step": 6242
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8244143928671318,
+      "learning_rate": 2.5783040937064973e-06,
+      "loss": 1.0385,
+      "step": 6243
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.746841577584406,
+      "learning_rate": 2.5767810098119925e-06,
+      "loss": 0.9423,
+      "step": 6244
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7203119142976462,
+      "learning_rate": 2.575258219741145e-06,
+      "loss": 0.8609,
+      "step": 6245
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8166916344841809,
+      "learning_rate": 2.573735723678599e-06,
+      "loss": 1.0295,
+      "step": 6246
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7848081194818683,
+      "learning_rate": 2.5722135218089596e-06,
+      "loss": 0.8538,
+      "step": 6247
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7623487772581938,
+      "learning_rate": 2.570691614316802e-06,
+      "loss": 0.9245,
+      "step": 6248
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6404582625312074,
+      "learning_rate": 2.5691700013866615e-06,
+      "loss": 0.8274,
+      "step": 6249
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7058493750051824,
+      "learning_rate": 2.5676486832030366e-06,
+      "loss": 0.8657,
+      "step": 6250
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.665627661706924,
+      "learning_rate": 2.5661276599503946e-06,
+      "loss": 0.9219,
+      "step": 6251
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6173258466095692,
+      "learning_rate": 2.5646069318131663e-06,
+      "loss": 0.8614,
+      "step": 6252
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8056869964177008,
+      "learning_rate": 2.563086498975742e-06,
+      "loss": 1.0238,
+      "step": 6253
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.5857994653785266,
+      "learning_rate": 2.5615663616224807e-06,
+      "loss": 0.8853,
+      "step": 6254
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7033636318659988,
+      "learning_rate": 2.560046519937707e-06,
+      "loss": 0.9227,
+      "step": 6255
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7328194119138727,
+      "learning_rate": 2.558526974105703e-06,
+      "loss": 0.8902,
+      "step": 6256
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7651588127763196,
+      "learning_rate": 2.5570077243107195e-06,
+      "loss": 0.8991,
+      "step": 6257
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8233859792312979,
+      "learning_rate": 2.5554887707369734e-06,
+      "loss": 0.8671,
+      "step": 6258
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7853192415757095,
+      "learning_rate": 2.5539701135686424e-06,
+      "loss": 0.9258,
+      "step": 6259
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8833438065441415,
+      "learning_rate": 2.552451752989865e-06,
+      "loss": 0.9093,
+      "step": 6260
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7033180450457077,
+      "learning_rate": 2.5509336891847516e-06,
+      "loss": 1.0048,
+      "step": 6261
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7738970481845326,
+      "learning_rate": 2.549415922337373e-06,
+      "loss": 0.8613,
+      "step": 6262
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8354950668127237,
+      "learning_rate": 2.5478984526317596e-06,
+      "loss": 0.9139,
+      "step": 6263
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7193073421244744,
+      "learning_rate": 2.5463812802519133e-06,
+      "loss": 0.9214,
+      "step": 6264
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8390743602576256,
+      "learning_rate": 2.544864405381797e-06,
+      "loss": 1.0755,
+      "step": 6265
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.694436913237053,
+      "learning_rate": 2.543347828205335e-06,
+      "loss": 0.9062,
+      "step": 6266
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7571316115903421,
+      "learning_rate": 2.541831548906417e-06,
+      "loss": 0.9149,
+      "step": 6267
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7909768759227815,
+      "learning_rate": 2.5403155676689005e-06,
+      "loss": 0.9436,
+      "step": 6268
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6163432616387976,
+      "learning_rate": 2.5387998846766016e-06,
+      "loss": 1.0191,
+      "step": 6269
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7445657368293812,
+      "learning_rate": 2.5372845001132995e-06,
+      "loss": 0.9784,
+      "step": 6270
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7309628117703755,
+      "learning_rate": 2.5357694141627444e-06,
+      "loss": 0.8894,
+      "step": 6271
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.71367086546903,
+      "learning_rate": 2.534254627008641e-06,
+      "loss": 0.8659,
+      "step": 6272
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8782553533182014,
+      "learning_rate": 2.532740138834665e-06,
+      "loss": 0.9021,
+      "step": 6273
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.659536796246654,
+      "learning_rate": 2.5312259498244544e-06,
+      "loss": 0.8797,
+      "step": 6274
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7219781013255586,
+      "learning_rate": 2.5297120601616078e-06,
+      "loss": 0.9111,
+      "step": 6275
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7275326697380633,
+      "learning_rate": 2.5281984700296893e-06,
+      "loss": 0.8625,
+      "step": 6276
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.751038908019539,
+      "learning_rate": 2.52668517961223e-06,
+      "loss": 0.8456,
+      "step": 6277
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7850071231354494,
+      "learning_rate": 2.5251721890927173e-06,
+      "loss": 0.8947,
+      "step": 6278
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7948236548349568,
+      "learning_rate": 2.5236594986546115e-06,
+      "loss": 0.8916,
+      "step": 6279
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7790145071690193,
+      "learning_rate": 2.5221471084813275e-06,
+      "loss": 1.0667,
+      "step": 6280
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6694011775816162,
+      "learning_rate": 2.5206350187562478e-06,
+      "loss": 0.8233,
+      "step": 6281
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7924469931292124,
+      "learning_rate": 2.51912322966272e-06,
+      "loss": 0.8215,
+      "step": 6282
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7382225125943191,
+      "learning_rate": 2.517611741384055e-06,
+      "loss": 0.8745,
+      "step": 6283
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.784656675374681,
+      "learning_rate": 2.516100554103523e-06,
+      "loss": 0.8867,
+      "step": 6284
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7235178279457042,
+      "learning_rate": 2.514589668004362e-06,
+      "loss": 0.9218,
+      "step": 6285
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7003293711459908,
+      "learning_rate": 2.513079083269774e-06,
+      "loss": 0.87,
+      "step": 6286
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7622962016753754,
+      "learning_rate": 2.511568800082919e-06,
+      "loss": 0.8675,
+      "step": 6287
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7634359918302613,
+      "learning_rate": 2.5100588186269257e-06,
+      "loss": 0.9789,
+      "step": 6288
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7062418159122503,
+      "learning_rate": 2.508549139084887e-06,
+      "loss": 0.8949,
+      "step": 6289
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8258182315515412,
+      "learning_rate": 2.5070397616398546e-06,
+      "loss": 0.8839,
+      "step": 6290
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7253363598707276,
+      "learning_rate": 2.5055306864748434e-06,
+      "loss": 0.8195,
+      "step": 6291
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8035688670621712,
+      "learning_rate": 2.504021913772836e-06,
+      "loss": 1.0734,
+      "step": 6292
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8141134361941402,
+      "learning_rate": 2.502513443716779e-06,
+      "loss": 0.945,
+      "step": 6293
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.702694864436999,
+      "learning_rate": 2.5010052764895752e-06,
+      "loss": 0.9183,
+      "step": 6294
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6997325147671798,
+      "learning_rate": 2.4994974122740968e-06,
+      "loss": 0.9226,
+      "step": 6295
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7691628706769729,
+      "learning_rate": 2.4979898512531796e-06,
+      "loss": 0.9788,
+      "step": 6296
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8579654144770366,
+      "learning_rate": 2.4964825936096164e-06,
+      "loss": 0.8455,
+      "step": 6297
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.734242582088059,
+      "learning_rate": 2.4949756395261706e-06,
+      "loss": 0.8553,
+      "step": 6298
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.72127889455071,
+      "learning_rate": 2.493468989185566e-06,
+      "loss": 0.9023,
+      "step": 6299
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8492916319245993,
+      "learning_rate": 2.4919626427704873e-06,
+      "loss": 0.8853,
+      "step": 6300
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7081866454464063,
+      "learning_rate": 2.4904566004635835e-06,
+      "loss": 0.8857,
+      "step": 6301
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8092648036106815,
+      "learning_rate": 2.4889508624474694e-06,
+      "loss": 0.96,
+      "step": 6302
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.734881060649047,
+      "learning_rate": 2.4874454289047194e-06,
+      "loss": 0.8888,
+      "step": 6303
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6413113648280484,
+      "learning_rate": 2.4859403000178727e-06,
+      "loss": 0.8875,
+      "step": 6304
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8255599126601714,
+      "learning_rate": 2.484435475969434e-06,
+      "loss": 1.0331,
+      "step": 6305
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5936253508054463,
+      "learning_rate": 2.4829309569418647e-06,
+      "loss": 0.8797,
+      "step": 6306
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6760817727923698,
+      "learning_rate": 2.4814267431175942e-06,
+      "loss": 0.8882,
+      "step": 6307
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7049749146747466,
+      "learning_rate": 2.4799228346790157e-06,
+      "loss": 0.8693,
+      "step": 6308
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7938586693913281,
+      "learning_rate": 2.47841923180848e-06,
+      "loss": 1.0454,
+      "step": 6309
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.659684506317678,
+      "learning_rate": 2.4769159346883053e-06,
+      "loss": 0.8767,
+      "step": 6310
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.759498728230593,
+      "learning_rate": 2.4754129435007757e-06,
+      "loss": 0.7966,
+      "step": 6311
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7735586144994937,
+      "learning_rate": 2.4739102584281268e-06,
+      "loss": 0.8261,
+      "step": 6312
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6222596364976718,
+      "learning_rate": 2.4724078796525674e-06,
+      "loss": 0.8279,
+      "step": 6313
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7357147757004663,
+      "learning_rate": 2.4709058073562688e-06,
+      "loss": 0.896,
+      "step": 6314
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6736312475252992,
+      "learning_rate": 2.4694040417213584e-06,
+      "loss": 0.84,
+      "step": 6315
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.712387500991326,
+      "learning_rate": 2.467902582929931e-06,
+      "loss": 0.8809,
+      "step": 6316
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8312300503986002,
+      "learning_rate": 2.4664014311640483e-06,
+      "loss": 1.0661,
+      "step": 6317
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8406427409416657,
+      "learning_rate": 2.464900586605724e-06,
+      "loss": 0.8822,
+      "step": 6318
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6642163541767439,
+      "learning_rate": 2.463400049436944e-06,
+      "loss": 0.8972,
+      "step": 6319
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6403885250513497,
+      "learning_rate": 2.4618998198396544e-06,
+      "loss": 0.8535,
+      "step": 6320
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7939277976489716,
+      "learning_rate": 2.46039989799576e-06,
+      "loss": 0.8749,
+      "step": 6321
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7437980695061905,
+      "learning_rate": 2.4589002840871356e-06,
+      "loss": 0.9086,
+      "step": 6322
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7420456231035646,
+      "learning_rate": 2.4574009782956098e-06,
+      "loss": 0.8592,
+      "step": 6323
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.681034388239921,
+      "learning_rate": 2.455901980802983e-06,
+      "loss": 0.9303,
+      "step": 6324
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7403804341569369,
+      "learning_rate": 2.4544032917910105e-06,
+      "loss": 0.8293,
+      "step": 6325
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7589515871965098,
+      "learning_rate": 2.4529049114414143e-06,
+      "loss": 0.9118,
+      "step": 6326
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6439032633449087,
+      "learning_rate": 2.451406839935881e-06,
+      "loss": 0.8802,
+      "step": 6327
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.70280600615638,
+      "learning_rate": 2.4499090774560525e-06,
+      "loss": 0.8553,
+      "step": 6328
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7901878989428093,
+      "learning_rate": 2.4484116241835406e-06,
+      "loss": 1.0551,
+      "step": 6329
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7773834627293514,
+      "learning_rate": 2.446914480299917e-06,
+      "loss": 0.8611,
+      "step": 6330
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7882227483365227,
+      "learning_rate": 2.4454176459867128e-06,
+      "loss": 0.8776,
+      "step": 6331
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.784805683420405,
+      "learning_rate": 2.4439211214254277e-06,
+      "loss": 0.8604,
+      "step": 6332
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8284797712547111,
+      "learning_rate": 2.4424249067975193e-06,
+      "loss": 1.05,
+      "step": 6333
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7737240075618959,
+      "learning_rate": 2.4409290022844063e-06,
+      "loss": 0.9075,
+      "step": 6334
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.778426538358641,
+      "learning_rate": 2.4394334080674737e-06,
+      "loss": 0.8287,
+      "step": 6335
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7130205874184843,
+      "learning_rate": 2.4379381243280705e-06,
+      "loss": 0.8397,
+      "step": 6336
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9545372288863674,
+      "learning_rate": 2.4364431512475e-06,
+      "loss": 0.9525,
+      "step": 6337
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7313769113621853,
+      "learning_rate": 2.4349484890070357e-06,
+      "loss": 0.9567,
+      "step": 6338
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6923486560772174,
+      "learning_rate": 2.433454137787912e-06,
+      "loss": 0.9243,
+      "step": 6339
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6825799692952383,
+      "learning_rate": 2.4319600977713202e-06,
+      "loss": 0.8907,
+      "step": 6340
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6735757110258604,
+      "learning_rate": 2.4304663691384206e-06,
+      "loss": 0.9446,
+      "step": 6341
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7574880840258724,
+      "learning_rate": 2.4289729520703335e-06,
+      "loss": 0.8246,
+      "step": 6342
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7286652635244675,
+      "learning_rate": 2.4274798467481397e-06,
+      "loss": 0.8394,
+      "step": 6343
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.804380902965081,
+      "learning_rate": 2.4259870533528816e-06,
+      "loss": 0.8157,
+      "step": 6344
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.619433649124254,
+      "learning_rate": 2.4244945720655693e-06,
+      "loss": 0.8605,
+      "step": 6345
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6771545394488558,
+      "learning_rate": 2.4230024030671674e-06,
+      "loss": 0.9685,
+      "step": 6346
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8188077663120488,
+      "learning_rate": 2.421510546538608e-06,
+      "loss": 1.0438,
+      "step": 6347
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7957326103343462,
+      "learning_rate": 2.4200190026607863e-06,
+      "loss": 1.0169,
+      "step": 6348
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6400952560309199,
+      "learning_rate": 2.418527771614553e-06,
+      "loss": 0.9201,
+      "step": 6349
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7626544881990391,
+      "learning_rate": 2.4170368535807276e-06,
+      "loss": 0.9058,
+      "step": 6350
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.795793615419143,
+      "learning_rate": 2.41554624874009e-06,
+      "loss": 0.9046,
+      "step": 6351
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7285805637971123,
+      "learning_rate": 2.414055957273378e-06,
+      "loss": 0.8933,
+      "step": 6352
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7044415277680376,
+      "learning_rate": 2.412565979361298e-06,
+      "loss": 0.8826,
+      "step": 6353
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7692479680389944,
+      "learning_rate": 2.411076315184511e-06,
+      "loss": 1.035,
+      "step": 6354
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.732504776997473,
+      "learning_rate": 2.409586964923649e-06,
+      "loss": 0.8953,
+      "step": 6355
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7191306763940477,
+      "learning_rate": 2.4080979287592956e-06,
+      "loss": 0.8896,
+      "step": 6356
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.679213792696338,
+      "learning_rate": 2.4066092068720044e-06,
+      "loss": 0.9286,
+      "step": 6357
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7370330393150843,
+      "learning_rate": 2.4051207994422886e-06,
+      "loss": 0.9929,
+      "step": 6358
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6569850276375588,
+      "learning_rate": 2.4036327066506204e-06,
+      "loss": 0.8631,
+      "step": 6359
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.684328953303427,
+      "learning_rate": 2.402144928677438e-06,
+      "loss": 0.9678,
+      "step": 6360
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7581061140868317,
+      "learning_rate": 2.4006574657031406e-06,
+      "loss": 0.961,
+      "step": 6361
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6182529012931322,
+      "learning_rate": 2.399170317908085e-06,
+      "loss": 0.9204,
+      "step": 6362
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6800318427721672,
+      "learning_rate": 2.3976834854725945e-06,
+      "loss": 0.8624,
+      "step": 6363
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7326101338727924,
+      "learning_rate": 2.396196968576957e-06,
+      "loss": 0.8201,
+      "step": 6364
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6964426323139046,
+      "learning_rate": 2.3947107674014097e-06,
+      "loss": 0.8008,
+      "step": 6365
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7852349429707048,
+      "learning_rate": 2.393224882126164e-06,
+      "loss": 1.0117,
+      "step": 6366
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7906967346692242,
+      "learning_rate": 2.39173931293139e-06,
+      "loss": 0.8804,
+      "step": 6367
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5991529197963048,
+      "learning_rate": 2.3902540599972145e-06,
+      "loss": 0.8894,
+      "step": 6368
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7571790425556097,
+      "learning_rate": 2.388769123503731e-06,
+      "loss": 0.9482,
+      "step": 6369
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8279520923737917,
+      "learning_rate": 2.387284503630996e-06,
+      "loss": 0.8651,
+      "step": 6370
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7597453783842736,
+      "learning_rate": 2.38580020055902e-06,
+      "loss": 0.8438,
+      "step": 6371
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.649568201050979,
+      "learning_rate": 2.3843162144677828e-06,
+      "loss": 0.8818,
+      "step": 6372
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7651789362003865,
+      "learning_rate": 2.3828325455372235e-06,
+      "loss": 0.8626,
+      "step": 6373
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6981179040575505,
+      "learning_rate": 2.3813491939472395e-06,
+      "loss": 0.9027,
+      "step": 6374
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.857453234191097,
+      "learning_rate": 2.379866159877695e-06,
+      "loss": 0.8993,
+      "step": 6375
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8223534122546445,
+      "learning_rate": 2.3783834435084118e-06,
+      "loss": 0.9394,
+      "step": 6376
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.747991650920335,
+      "learning_rate": 2.376901045019172e-06,
+      "loss": 0.9611,
+      "step": 6377
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7375928943168886,
+      "learning_rate": 2.375418964589724e-06,
+      "loss": 0.8438,
+      "step": 6378
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6699328607708452,
+      "learning_rate": 2.3739372023997766e-06,
+      "loss": 0.9418,
+      "step": 6379
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8601396381478732,
+      "learning_rate": 2.3724557586289948e-06,
+      "loss": 0.9251,
+      "step": 6380
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6598587504071074,
+      "learning_rate": 2.370974633457011e-06,
+      "loss": 0.9022,
+      "step": 6381
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.640404748916299,
+      "learning_rate": 2.3694938270634187e-06,
+      "loss": 0.856,
+      "step": 6382
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.780357684810993,
+      "learning_rate": 2.368013339627766e-06,
+      "loss": 0.9752,
+      "step": 6383
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6577966730502312,
+      "learning_rate": 2.3665331713295702e-06,
+      "loss": 0.8679,
+      "step": 6384
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6394701337816577,
+      "learning_rate": 2.3650533223483087e-06,
+      "loss": 0.8654,
+      "step": 6385
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6092014135355948,
+      "learning_rate": 2.363573792863415e-06,
+      "loss": 0.8398,
+      "step": 6386
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7970457322296525,
+      "learning_rate": 2.3620945830542865e-06,
+      "loss": 1.063,
+      "step": 6387
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7806065649340237,
+      "learning_rate": 2.360615693100284e-06,
+      "loss": 0.8465,
+      "step": 6388
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.83389565021063,
+      "learning_rate": 2.35913712318073e-06,
+      "loss": 0.9759,
+      "step": 6389
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.870641427461459,
+      "learning_rate": 2.3576588734749022e-06,
+      "loss": 0.9264,
+      "step": 6390
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.907716870294996,
+      "learning_rate": 2.3561809441620458e-06,
+      "loss": 0.9462,
+      "step": 6391
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7560314380246362,
+      "learning_rate": 2.3547033354213663e-06,
+      "loss": 0.8711,
+      "step": 6392
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.718082464890501,
+      "learning_rate": 2.3532260474320253e-06,
+      "loss": 0.9093,
+      "step": 6393
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8607681501190125,
+      "learning_rate": 2.3517490803731506e-06,
+      "loss": 0.9109,
+      "step": 6394
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.629149739466703,
+      "learning_rate": 2.350272434423832e-06,
+      "loss": 0.8486,
+      "step": 6395
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7286045547375022,
+      "learning_rate": 2.3487961097631157e-06,
+      "loss": 0.889,
+      "step": 6396
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6485881294266451,
+      "learning_rate": 2.347320106570009e-06,
+      "loss": 0.8467,
+      "step": 6397
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6462412616095465,
+      "learning_rate": 2.3458444250234867e-06,
+      "loss": 0.891,
+      "step": 6398
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7789994729182725,
+      "learning_rate": 2.3443690653024766e-06,
+      "loss": 0.9041,
+      "step": 6399
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7640392222154468,
+      "learning_rate": 2.342894027585872e-06,
+      "loss": 0.8624,
+      "step": 6400
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7648592280305473,
+      "learning_rate": 2.341419312052529e-06,
+      "loss": 0.8123,
+      "step": 6401
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5786820035097453,
+      "learning_rate": 2.3399449188812585e-06,
+      "loss": 0.8792,
+      "step": 6402
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6446761369502103,
+      "learning_rate": 2.338470848250837e-06,
+      "loss": 0.952,
+      "step": 6403
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7453703479620972,
+      "learning_rate": 2.3369971003400023e-06,
+      "loss": 0.8966,
+      "step": 6404
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7360531874124625,
+      "learning_rate": 2.3355236753274485e-06,
+      "loss": 0.8973,
+      "step": 6405
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.682536674756058,
+      "learning_rate": 2.3340505733918366e-06,
+      "loss": 0.8975,
+      "step": 6406
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6906074913766744,
+      "learning_rate": 2.3325777947117834e-06,
+      "loss": 0.9044,
+      "step": 6407
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.763309128437027,
+      "learning_rate": 2.3311053394658673e-06,
+      "loss": 0.9257,
+      "step": 6408
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7794618159630347,
+      "learning_rate": 2.3296332078326295e-06,
+      "loss": 0.8532,
+      "step": 6409
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9017116519705788,
+      "learning_rate": 2.3281613999905735e-06,
+      "loss": 0.8143,
+      "step": 6410
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6167622553043741,
+      "learning_rate": 2.326689916118157e-06,
+      "loss": 0.9034,
+      "step": 6411
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6832448697424574,
+      "learning_rate": 2.3252187563938043e-06,
+      "loss": 0.854,
+      "step": 6412
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8022472792218662,
+      "learning_rate": 2.3237479209959007e-06,
+      "loss": 0.9031,
+      "step": 6413
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6857281545105625,
+      "learning_rate": 2.3222774101027864e-06,
+      "loss": 0.8264,
+      "step": 6414
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8342658948264268,
+      "learning_rate": 2.3208072238927676e-06,
+      "loss": 0.9711,
+      "step": 6415
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6052059113500816,
+      "learning_rate": 2.3193373625441113e-06,
+      "loss": 0.9606,
+      "step": 6416
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9129378632159246,
+      "learning_rate": 2.3178678262350413e-06,
+      "loss": 0.9165,
+      "step": 6417
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7979017153175558,
+      "learning_rate": 2.3163986151437425e-06,
+      "loss": 0.8838,
+      "step": 6418
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8112027081461685,
+      "learning_rate": 2.314929729448363e-06,
+      "loss": 1.0528,
+      "step": 6419
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8016567224249653,
+      "learning_rate": 2.313461169327013e-06,
+      "loss": 0.8438,
+      "step": 6420
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6190363985458798,
+      "learning_rate": 2.311992934957756e-06,
+      "loss": 0.9004,
+      "step": 6421
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7395807127674972,
+      "learning_rate": 2.3105250265186223e-06,
+      "loss": 0.8941,
+      "step": 6422
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7801445618860856,
+      "learning_rate": 2.3090574441876037e-06,
+      "loss": 0.8477,
+      "step": 6423
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.684259725232249,
+      "learning_rate": 2.3075901881426447e-06,
+      "loss": 0.9577,
+      "step": 6424
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6512595673873343,
+      "learning_rate": 2.306123258561658e-06,
+      "loss": 0.876,
+      "step": 6425
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7376056379547626,
+      "learning_rate": 2.3046566556225147e-06,
+      "loss": 0.9278,
+      "step": 6426
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.640799043452101,
+      "learning_rate": 2.303190379503043e-06,
+      "loss": 0.8539,
+      "step": 6427
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.837145620016028,
+      "learning_rate": 2.3017244303810367e-06,
+      "loss": 1.0359,
+      "step": 6428
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.678453617185128,
+      "learning_rate": 2.3002588084342453e-06,
+      "loss": 0.9749,
+      "step": 6429
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7322387791929172,
+      "learning_rate": 2.2987935138403786e-06,
+      "loss": 0.8729,
+      "step": 6430
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.787351980524748,
+      "learning_rate": 2.2973285467771117e-06,
+      "loss": 0.8499,
+      "step": 6431
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7297912063073069,
+      "learning_rate": 2.2958639074220773e-06,
+      "loss": 0.9498,
+      "step": 6432
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7682062023649043,
+      "learning_rate": 2.2943995959528654e-06,
+      "loss": 0.9066,
+      "step": 6433
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6911226896456393,
+      "learning_rate": 2.2929356125470297e-06,
+      "loss": 0.8444,
+      "step": 6434
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7473526372955408,
+      "learning_rate": 2.2914719573820854e-06,
+      "loss": 0.8904,
+      "step": 6435
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.709803939428721,
+      "learning_rate": 2.2900086306355024e-06,
+      "loss": 0.8879,
+      "step": 6436
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.714766430048214,
+      "learning_rate": 2.2885456324847156e-06,
+      "loss": 0.9938,
+      "step": 6437
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7734537824627243,
+      "learning_rate": 2.2870829631071207e-06,
+      "loss": 0.7769,
+      "step": 6438
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6265006960462027,
+      "learning_rate": 2.2856206226800686e-06,
+      "loss": 0.8827,
+      "step": 6439
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7953922639171864,
+      "learning_rate": 2.284158611380873e-06,
+      "loss": 0.8401,
+      "step": 6440
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6539623981731828,
+      "learning_rate": 2.2826969293868103e-06,
+      "loss": 0.8675,
+      "step": 6441
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6732542436388826,
+      "learning_rate": 2.2812355768751106e-06,
+      "loss": 0.8402,
+      "step": 6442
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7148741805280188,
+      "learning_rate": 2.2797745540229703e-06,
+      "loss": 0.969,
+      "step": 6443
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8267873061019673,
+      "learning_rate": 2.2783138610075454e-06,
+      "loss": 0.9714,
+      "step": 6444
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.718420172799168,
+      "learning_rate": 2.2768534980059464e-06,
+      "loss": 0.8913,
+      "step": 6445
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7543979168408128,
+      "learning_rate": 2.2753934651952485e-06,
+      "loss": 0.9188,
+      "step": 6446
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6958023606060988,
+      "learning_rate": 2.273933762752488e-06,
+      "loss": 0.8422,
+      "step": 6447
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6728451610751116,
+      "learning_rate": 2.2724743908546553e-06,
+      "loss": 0.9664,
+      "step": 6448
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7736551340772795,
+      "learning_rate": 2.271015349678707e-06,
+      "loss": 0.8438,
+      "step": 6449
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7341991670853147,
+      "learning_rate": 2.2695566394015552e-06,
+      "loss": 0.8442,
+      "step": 6450
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8223362674719412,
+      "learning_rate": 2.268098260200075e-06,
+      "loss": 0.8811,
+      "step": 6451
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6889992774898281,
+      "learning_rate": 2.2666402122510976e-06,
+      "loss": 0.8354,
+      "step": 6452
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7202953775164767,
+      "learning_rate": 2.2651824957314177e-06,
+      "loss": 0.8754,
+      "step": 6453
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8086863280767134,
+      "learning_rate": 2.2637251108177903e-06,
+      "loss": 0.9026,
+      "step": 6454
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7791433870769235,
+      "learning_rate": 2.262268057686925e-06,
+      "loss": 0.8948,
+      "step": 6455
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.704754465481641,
+      "learning_rate": 2.260811336515496e-06,
+      "loss": 0.911,
+      "step": 6456
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6309859294236584,
+      "learning_rate": 2.259354947480138e-06,
+      "loss": 0.9137,
+      "step": 6457
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7905060768889112,
+      "learning_rate": 2.257898890757439e-06,
+      "loss": 0.9301,
+      "step": 6458
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6854299724254367,
+      "learning_rate": 2.2564431665239545e-06,
+      "loss": 0.9584,
+      "step": 6459
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8071680285834402,
+      "learning_rate": 2.2549877749561942e-06,
+      "loss": 0.8846,
+      "step": 6460
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.966384103576259,
+      "learning_rate": 2.2535327162306286e-06,
+      "loss": 0.9001,
+      "step": 6461
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.644441686769002,
+      "learning_rate": 2.2520779905236895e-06,
+      "loss": 0.8767,
+      "step": 6462
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7641531725257853,
+      "learning_rate": 2.2506235980117695e-06,
+      "loss": 0.8134,
+      "step": 6463
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6698888397631382,
+      "learning_rate": 2.2491695388712147e-06,
+      "loss": 0.8875,
+      "step": 6464
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7140474633689036,
+      "learning_rate": 2.2477158132783373e-06,
+      "loss": 0.8336,
+      "step": 6465
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6292820705781816,
+      "learning_rate": 2.2462624214094074e-06,
+      "loss": 0.8117,
+      "step": 6466
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.703032164635888,
+      "learning_rate": 2.244809363440651e-06,
+      "loss": 0.8762,
+      "step": 6467
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.822290143682642,
+      "learning_rate": 2.2433566395482577e-06,
+      "loss": 1.0459,
+      "step": 6468
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8433594189442188,
+      "learning_rate": 2.241904249908377e-06,
+      "loss": 0.9836,
+      "step": 6469
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8341475622017183,
+      "learning_rate": 2.2404521946971153e-06,
+      "loss": 0.9301,
+      "step": 6470
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7042718432280926,
+      "learning_rate": 2.2390004740905363e-06,
+      "loss": 0.8782,
+      "step": 6471
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7864534021631688,
+      "learning_rate": 2.2375490882646698e-06,
+      "loss": 0.9144,
+      "step": 6472
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6414715419715076,
+      "learning_rate": 2.2360980373954988e-06,
+      "loss": 0.8805,
+      "step": 6473
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6337315145944105,
+      "learning_rate": 2.2346473216589694e-06,
+      "loss": 0.8905,
+      "step": 6474
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6777478932148686,
+      "learning_rate": 2.233196941230988e-06,
+      "loss": 0.8884,
+      "step": 6475
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7425050400721716,
+      "learning_rate": 2.2317468962874136e-06,
+      "loss": 0.9187,
+      "step": 6476
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7734889528466828,
+      "learning_rate": 2.230297187004072e-06,
+      "loss": 0.9496,
+      "step": 6477
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9224113032142296,
+      "learning_rate": 2.2288478135567466e-06,
+      "loss": 0.9179,
+      "step": 6478
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8311997805937364,
+      "learning_rate": 2.2273987761211755e-06,
+      "loss": 1.0378,
+      "step": 6479
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.850258514323672,
+      "learning_rate": 2.225950074873064e-06,
+      "loss": 0.9207,
+      "step": 6480
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7809591364767459,
+      "learning_rate": 2.2245017099880667e-06,
+      "loss": 0.9104,
+      "step": 6481
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7240391872122136,
+      "learning_rate": 2.223053681641808e-06,
+      "loss": 0.9074,
+      "step": 6482
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6664542788668992,
+      "learning_rate": 2.2216059900098624e-06,
+      "loss": 0.881,
+      "step": 6483
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8632306943209074,
+      "learning_rate": 2.220158635267769e-06,
+      "loss": 0.8161,
+      "step": 6484
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7832573405557359,
+      "learning_rate": 2.2187116175910263e-06,
+      "loss": 1.0424,
+      "step": 6485
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6652644140554596,
+      "learning_rate": 2.2172649371550863e-06,
+      "loss": 0.8183,
+      "step": 6486
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7201456017107313,
+      "learning_rate": 2.2158185941353672e-06,
+      "loss": 0.8584,
+      "step": 6487
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8318461269806292,
+      "learning_rate": 2.2143725887072433e-06,
+      "loss": 0.8774,
+      "step": 6488
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7415917176426496,
+      "learning_rate": 2.212926921046045e-06,
+      "loss": 0.9731,
+      "step": 6489
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8229419483855842,
+      "learning_rate": 2.2114815913270654e-06,
+      "loss": 1.0442,
+      "step": 6490
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7100792690292168,
+      "learning_rate": 2.2100365997255603e-06,
+      "loss": 0.8463,
+      "step": 6491
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7146701057287772,
+      "learning_rate": 2.2085919464167327e-06,
+      "loss": 0.8774,
+      "step": 6492
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7712598130870785,
+      "learning_rate": 2.2071476315757547e-06,
+      "loss": 0.9057,
+      "step": 6493
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7653120402700195,
+      "learning_rate": 2.2057036553777565e-06,
+      "loss": 0.8445,
+      "step": 6494
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8224215311535858,
+      "learning_rate": 2.2042600179978217e-06,
+      "loss": 0.8458,
+      "step": 6495
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.679261289247971,
+      "learning_rate": 2.2028167196109984e-06,
+      "loss": 0.8481,
+      "step": 6496
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7110145221830693,
+      "learning_rate": 2.201373760392293e-06,
+      "loss": 0.8785,
+      "step": 6497
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8873982272738539,
+      "learning_rate": 2.1999311405166653e-06,
+      "loss": 1.0019,
+      "step": 6498
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6345783333742534,
+      "learning_rate": 2.1984888601590402e-06,
+      "loss": 0.8492,
+      "step": 6499
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6164480458479393,
+      "learning_rate": 2.1970469194943013e-06,
+      "loss": 0.8275,
+      "step": 6500
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6168536132642155,
+      "learning_rate": 2.195605318697284e-06,
+      "loss": 0.8976,
+      "step": 6501
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7392008653487319,
+      "learning_rate": 2.194164057942793e-06,
+      "loss": 0.8533,
+      "step": 6502
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8425339924552915,
+      "learning_rate": 2.1927231374055824e-06,
+      "loss": 0.8477,
+      "step": 6503
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7486191085993927,
+      "learning_rate": 2.191282557260368e-06,
+      "loss": 0.892,
+      "step": 6504
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8196822167068794,
+      "learning_rate": 2.1898423176818266e-06,
+      "loss": 1.0546,
+      "step": 6505
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.804972751044285,
+      "learning_rate": 2.188402418844594e-06,
+      "loss": 0.9826,
+      "step": 6506
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7490978399672044,
+      "learning_rate": 2.1869628609232595e-06,
+      "loss": 0.9523,
+      "step": 6507
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6456700300755442,
+      "learning_rate": 2.185523644092376e-06,
+      "loss": 0.8537,
+      "step": 6508
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.747714028169617,
+      "learning_rate": 2.1840847685264553e-06,
+      "loss": 0.8765,
+      "step": 6509
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8602985304062363,
+      "learning_rate": 2.182646234399963e-06,
+      "loss": 0.884,
+      "step": 6510
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6904509006898039,
+      "learning_rate": 2.181208041887327e-06,
+      "loss": 0.8733,
+      "step": 6511
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7180806610484716,
+      "learning_rate": 2.179770191162936e-06,
+      "loss": 0.8612,
+      "step": 6512
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7795506390134943,
+      "learning_rate": 2.1783326824011325e-06,
+      "loss": 0.9185,
+      "step": 6513
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7702820459645796,
+      "learning_rate": 2.1768955157762165e-06,
+      "loss": 0.8839,
+      "step": 6514
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.84986440693234,
+      "learning_rate": 2.1754586914624524e-06,
+      "loss": 0.9714,
+      "step": 6515
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6903287549363535,
+      "learning_rate": 2.174022209634061e-06,
+      "loss": 0.9282,
+      "step": 6516
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7930715009173661,
+      "learning_rate": 2.172586070465218e-06,
+      "loss": 1.0627,
+      "step": 6517
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8065462985657579,
+      "learning_rate": 2.171150274130061e-06,
+      "loss": 0.9436,
+      "step": 6518
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7428430462294333,
+      "learning_rate": 2.169714820802688e-06,
+      "loss": 0.9408,
+      "step": 6519
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7579717672960957,
+      "learning_rate": 2.168279710657149e-06,
+      "loss": 0.8797,
+      "step": 6520
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.704672862183652,
+      "learning_rate": 2.1668449438674573e-06,
+      "loss": 0.8725,
+      "step": 6521
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.608623656331159,
+      "learning_rate": 2.165410520607585e-06,
+      "loss": 0.8339,
+      "step": 6522
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7820718510333096,
+      "learning_rate": 2.1639764410514593e-06,
+      "loss": 0.9533,
+      "step": 6523
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8121246062051826,
+      "learning_rate": 2.162542705372966e-06,
+      "loss": 0.9224,
+      "step": 6524
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7451680080069862,
+      "learning_rate": 2.1611093137459533e-06,
+      "loss": 0.888,
+      "step": 6525
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7877371388730705,
+      "learning_rate": 2.159676266344222e-06,
+      "loss": 0.9034,
+      "step": 6526
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7508309475572,
+      "learning_rate": 2.158243563341535e-06,
+      "loss": 0.8797,
+      "step": 6527
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6283336682703422,
+      "learning_rate": 2.1568112049116152e-06,
+      "loss": 0.868,
+      "step": 6528
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.814573546727918,
+      "learning_rate": 2.155379191228137e-06,
+      "loss": 0.8615,
+      "step": 6529
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7989014016275456,
+      "learning_rate": 2.1539475224647382e-06,
+      "loss": 1.0468,
+      "step": 6530
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6934492282860902,
+      "learning_rate": 2.1525161987950165e-06,
+      "loss": 0.8293,
+      "step": 6531
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7291628375835078,
+      "learning_rate": 2.1510852203925208e-06,
+      "loss": 0.9606,
+      "step": 6532
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7402937626098416,
+      "learning_rate": 2.149654587430765e-06,
+      "loss": 0.8741,
+      "step": 6533
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7789184130482414,
+      "learning_rate": 2.148224300083216e-06,
+      "loss": 0.8279,
+      "step": 6534
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7488139381423606,
+      "learning_rate": 2.1467943585233037e-06,
+      "loss": 0.8515,
+      "step": 6535
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6871851775977171,
+      "learning_rate": 2.14536476292441e-06,
+      "loss": 0.8904,
+      "step": 6536
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8087439925144518,
+      "learning_rate": 2.143935513459882e-06,
+      "loss": 1.0543,
+      "step": 6537
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.869935151028315,
+      "learning_rate": 2.1425066103030174e-06,
+      "loss": 0.8462,
+      "step": 6538
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8089762425976452,
+      "learning_rate": 2.141078053627078e-06,
+      "loss": 1.0616,
+      "step": 6539
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.732904262014079,
+      "learning_rate": 2.1396498436052824e-06,
+      "loss": 0.9287,
+      "step": 6540
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.862995675573417,
+      "learning_rate": 2.1382219804108023e-06,
+      "loss": 0.9363,
+      "step": 6541
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6700893245665551,
+      "learning_rate": 2.136794464216774e-06,
+      "loss": 0.7933,
+      "step": 6542
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8855387731039055,
+      "learning_rate": 2.135367295196289e-06,
+      "loss": 0.9527,
+      "step": 6543
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7674023667883219,
+      "learning_rate": 2.133940473522395e-06,
+      "loss": 0.9696,
+      "step": 6544
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.659872679461199,
+      "learning_rate": 2.132513999368098e-06,
+      "loss": 0.914,
+      "step": 6545
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7660987234609811,
+      "learning_rate": 2.1310878729063645e-06,
+      "loss": 1.0418,
+      "step": 6546
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8053046035028604,
+      "learning_rate": 2.1296620943101187e-06,
+      "loss": 0.8083,
+      "step": 6547
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6437885934811287,
+      "learning_rate": 2.128236663752238e-06,
+      "loss": 0.9065,
+      "step": 6548
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.774525129844558,
+      "learning_rate": 2.1268115814055616e-06,
+      "loss": 0.8546,
+      "step": 6549
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7359510073685382,
+      "learning_rate": 2.1253868474428892e-06,
+      "loss": 0.9053,
+      "step": 6550
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8131639876813324,
+      "learning_rate": 2.123962462036969e-06,
+      "loss": 0.8777,
+      "step": 6551
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8687424577741796,
+      "learning_rate": 2.122538425360516e-06,
+      "loss": 0.9315,
+      "step": 6552
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.4971173166682594,
+      "learning_rate": 2.1211147375862005e-06,
+      "loss": 0.8363,
+      "step": 6553
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.906942881193207,
+      "learning_rate": 2.1196913988866467e-06,
+      "loss": 0.8663,
+      "step": 6554
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6336544506942645,
+      "learning_rate": 2.1182684094344425e-06,
+      "loss": 0.9543,
+      "step": 6555
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8278470624935752,
+      "learning_rate": 2.116845769402128e-06,
+      "loss": 0.9389,
+      "step": 6556
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7816537710535942,
+      "learning_rate": 2.1154234789622023e-06,
+      "loss": 0.9152,
+      "step": 6557
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7250078997103446,
+      "learning_rate": 2.114001538287125e-06,
+      "loss": 0.8605,
+      "step": 6558
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.735779656413698,
+      "learning_rate": 2.1125799475493116e-06,
+      "loss": 0.8959,
+      "step": 6559
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7633399344610248,
+      "learning_rate": 2.1111587069211324e-06,
+      "loss": 0.9171,
+      "step": 6560
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7970626629178774,
+      "learning_rate": 2.1097378165749192e-06,
+      "loss": 0.877,
+      "step": 6561
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8155411533445395,
+      "learning_rate": 2.1083172766829624e-06,
+      "loss": 1.0459,
+      "step": 6562
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7974433490189392,
+      "learning_rate": 2.1068970874175028e-06,
+      "loss": 0.9685,
+      "step": 6563
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.560233100474211,
+      "learning_rate": 2.1054772489507454e-06,
+      "loss": 0.8488,
+      "step": 6564
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.4171123543495603,
+      "learning_rate": 2.1040577614548523e-06,
+      "loss": 0.9033,
+      "step": 6565
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7332220712905924,
+      "learning_rate": 2.1026386251019394e-06,
+      "loss": 0.8648,
+      "step": 6566
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.801269275748386,
+      "learning_rate": 2.1012198400640804e-06,
+      "loss": 1.0624,
+      "step": 6567
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.604394175743742,
+      "learning_rate": 2.099801406513311e-06,
+      "loss": 0.866,
+      "step": 6568
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7125115013387315,
+      "learning_rate": 2.0983833246216177e-06,
+      "loss": 0.9311,
+      "step": 6569
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7415911016110615,
+      "learning_rate": 2.0969655945609496e-06,
+      "loss": 0.9323,
+      "step": 6570
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6713327675086096,
+      "learning_rate": 2.095548216503214e-06,
+      "loss": 0.9172,
+      "step": 6571
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8518350382764888,
+      "learning_rate": 2.0941311906202672e-06,
+      "loss": 0.9405,
+      "step": 6572
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7280119736042032,
+      "learning_rate": 2.0927145170839324e-06,
+      "loss": 0.858,
+      "step": 6573
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7566571361783188,
+      "learning_rate": 2.0912981960659873e-06,
+      "loss": 0.8805,
+      "step": 6574
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6999652566728793,
+      "learning_rate": 2.0898822277381614e-06,
+      "loss": 0.8304,
+      "step": 6575
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8014802545931343,
+      "learning_rate": 2.0884666122721504e-06,
+      "loss": 0.9092,
+      "step": 6576
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.905328512619001,
+      "learning_rate": 2.0870513498395976e-06,
+      "loss": 0.824,
+      "step": 6577
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7224313557461404,
+      "learning_rate": 2.085636440612114e-06,
+      "loss": 0.8988,
+      "step": 6578
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7329489280156887,
+      "learning_rate": 2.0842218847612574e-06,
+      "loss": 0.954,
+      "step": 6579
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.695717943358972,
+      "learning_rate": 2.0828076824585484e-06,
+      "loss": 0.8973,
+      "step": 6580
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7787279887618899,
+      "learning_rate": 2.081393833875468e-06,
+      "loss": 0.9379,
+      "step": 6581
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.861501847671149,
+      "learning_rate": 2.0799803391834444e-06,
+      "loss": 0.906,
+      "step": 6582
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.726323229737557,
+      "learning_rate": 2.0785671985538724e-06,
+      "loss": 0.9899,
+      "step": 6583
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6718726863618756,
+      "learning_rate": 2.0771544121581004e-06,
+      "loss": 0.8178,
+      "step": 6584
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7480456928302555,
+      "learning_rate": 2.075741980167431e-06,
+      "loss": 0.9852,
+      "step": 6585
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7602492031713983,
+      "learning_rate": 2.07432990275313e-06,
+      "loss": 0.8519,
+      "step": 6586
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6764569899839548,
+      "learning_rate": 2.0729181800864147e-06,
+      "loss": 0.8535,
+      "step": 6587
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7853781465766567,
+      "learning_rate": 2.0715068123384593e-06,
+      "loss": 0.9452,
+      "step": 6588
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.804465003638183,
+      "learning_rate": 2.0700957996803983e-06,
+      "loss": 0.871,
+      "step": 6589
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8413857129108644,
+      "learning_rate": 2.0686851422833248e-06,
+      "loss": 0.9544,
+      "step": 6590
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6845093658371615,
+      "learning_rate": 2.067274840318282e-06,
+      "loss": 0.9021,
+      "step": 6591
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9848772686674427,
+      "learning_rate": 2.0658648939562752e-06,
+      "loss": 0.9191,
+      "step": 6592
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8216597560982837,
+      "learning_rate": 2.0644553033682675e-06,
+      "loss": 0.8991,
+      "step": 6593
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7060643321128839,
+      "learning_rate": 2.0630460687251722e-06,
+      "loss": 0.8565,
+      "step": 6594
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.832488231026425,
+      "learning_rate": 2.0616371901978672e-06,
+      "loss": 0.9113,
+      "step": 6595
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7738437949181498,
+      "learning_rate": 2.060228667957184e-06,
+      "loss": 0.924,
+      "step": 6596
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6528044742322285,
+      "learning_rate": 2.05882050217391e-06,
+      "loss": 0.9417,
+      "step": 6597
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8012859691026287,
+      "learning_rate": 2.0574126930187882e-06,
+      "loss": 1.0514,
+      "step": 6598
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6518159959119059,
+      "learning_rate": 2.0560052406625236e-06,
+      "loss": 0.8148,
+      "step": 6599
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7070087920227248,
+      "learning_rate": 2.054598145275772e-06,
+      "loss": 0.8351,
+      "step": 6600
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7212968647519868,
+      "learning_rate": 2.0531914070291485e-06,
+      "loss": 0.8988,
+      "step": 6601
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7681325986910603,
+      "learning_rate": 2.0517850260932288e-06,
+      "loss": 0.847,
+      "step": 6602
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7900270133362173,
+      "learning_rate": 2.050379002638537e-06,
+      "loss": 0.8952,
+      "step": 6603
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9298472715869732,
+      "learning_rate": 2.048973336835559e-06,
+      "loss": 0.886,
+      "step": 6604
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7825152998053173,
+      "learning_rate": 2.04756802885474e-06,
+      "loss": 0.9348,
+      "step": 6605
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7034938940169955,
+      "learning_rate": 2.0461630788664746e-06,
+      "loss": 0.9092,
+      "step": 6606
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8139634288653974,
+      "learning_rate": 2.044758487041119e-06,
+      "loss": 1.0336,
+      "step": 6607
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7927525541022816,
+      "learning_rate": 2.0433542535489868e-06,
+      "loss": 1.0624,
+      "step": 6608
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.741321114110323,
+      "learning_rate": 2.0419503785603444e-06,
+      "loss": 0.8042,
+      "step": 6609
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7929716813261991,
+      "learning_rate": 2.0405468622454157e-06,
+      "loss": 1.0722,
+      "step": 6610
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8319185123593109,
+      "learning_rate": 2.039143704774382e-06,
+      "loss": 1.0648,
+      "step": 6611
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8325420973332729,
+      "learning_rate": 2.037740906317383e-06,
+      "loss": 1.0732,
+      "step": 6612
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.698885943887545,
+      "learning_rate": 2.03633846704451e-06,
+      "loss": 0.9319,
+      "step": 6613
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.793832422527581,
+      "learning_rate": 2.034936387125816e-06,
+      "loss": 0.8815,
+      "step": 6614
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.682269625795327,
+      "learning_rate": 2.033534666731308e-06,
+      "loss": 0.9237,
+      "step": 6615
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7594399177308309,
+      "learning_rate": 2.032133306030948e-06,
+      "loss": 0.8752,
+      "step": 6616
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7832779786521455,
+      "learning_rate": 2.0307323051946554e-06,
+      "loss": 0.8903,
+      "step": 6617
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6436473382487349,
+      "learning_rate": 2.0293316643923112e-06,
+      "loss": 0.8276,
+      "step": 6618
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9230790960183213,
+      "learning_rate": 2.0279313837937407e-06,
+      "loss": 0.8859,
+      "step": 6619
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7079254024673398,
+      "learning_rate": 2.026531463568736e-06,
+      "loss": 0.9078,
+      "step": 6620
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7036267331578427,
+      "learning_rate": 2.0251319038870448e-06,
+      "loss": 0.8147,
+      "step": 6621
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8232668128603995,
+      "learning_rate": 2.023732704918364e-06,
+      "loss": 0.8379,
+      "step": 6622
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8251597426370927,
+      "learning_rate": 2.0223338668323534e-06,
+      "loss": 1.0461,
+      "step": 6623
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6319486597494803,
+      "learning_rate": 2.0209353897986288e-06,
+      "loss": 0.9223,
+      "step": 6624
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8710712585047917,
+      "learning_rate": 2.019537273986757e-06,
+      "loss": 0.8417,
+      "step": 6625
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7435377767995675,
+      "learning_rate": 2.0181395195662658e-06,
+      "loss": 0.928,
+      "step": 6626
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6680285616912485,
+      "learning_rate": 2.0167421267066396e-06,
+      "loss": 0.8637,
+      "step": 6627
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7973981059527666,
+      "learning_rate": 2.015345095577314e-06,
+      "loss": 0.9076,
+      "step": 6628
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.695107518504038,
+      "learning_rate": 2.0139484263476867e-06,
+      "loss": 0.952,
+      "step": 6629
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.691062627197711,
+      "learning_rate": 2.0125521191871054e-06,
+      "loss": 0.8496,
+      "step": 6630
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7017830952546598,
+      "learning_rate": 2.0111561742648807e-06,
+      "loss": 0.8314,
+      "step": 6631
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6914953698486894,
+      "learning_rate": 2.0097605917502722e-06,
+      "loss": 0.9175,
+      "step": 6632
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7496125835090524,
+      "learning_rate": 2.0083653718125028e-06,
+      "loss": 0.8663,
+      "step": 6633
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7200634243282291,
+      "learning_rate": 2.0069705146207443e-06,
+      "loss": 0.8756,
+      "step": 6634
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6758419587361633,
+      "learning_rate": 2.005576020344129e-06,
+      "loss": 0.9793,
+      "step": 6635
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7826588594816537,
+      "learning_rate": 2.004181889151746e-06,
+      "loss": 0.8803,
+      "step": 6636
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7968083372733458,
+      "learning_rate": 2.002788121212636e-06,
+      "loss": 1.059,
+      "step": 6637
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7634950205502977,
+      "learning_rate": 2.0013947166957994e-06,
+      "loss": 1.074,
+      "step": 6638
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7469682647173377,
+      "learning_rate": 2.0000016757701924e-06,
+      "loss": 0.8641,
+      "step": 6639
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7798595833165498,
+      "learning_rate": 1.998608998604725e-06,
+      "loss": 0.8915,
+      "step": 6640
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6317725214390892,
+      "learning_rate": 1.997216685368262e-06,
+      "loss": 0.8831,
+      "step": 6641
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.773682462898771,
+      "learning_rate": 1.995824736229628e-06,
+      "loss": 1.0586,
+      "step": 6642
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6821681318247257,
+      "learning_rate": 1.9944331513576038e-06,
+      "loss": 0.9522,
+      "step": 6643
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.71955930318639,
+      "learning_rate": 1.9930419309209197e-06,
+      "loss": 0.9543,
+      "step": 6644
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.691716800639837,
+      "learning_rate": 1.9916510750882685e-06,
+      "loss": 0.8728,
+      "step": 6645
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7538000774750189,
+      "learning_rate": 1.990260584028297e-06,
+      "loss": 0.8473,
+      "step": 6646
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.743677045300998,
+      "learning_rate": 1.988870457909604e-06,
+      "loss": 1.0128,
+      "step": 6647
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.699347455894714,
+      "learning_rate": 1.987480696900749e-06,
+      "loss": 0.8996,
+      "step": 6648
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.724643892740598,
+      "learning_rate": 1.9860913011702477e-06,
+      "loss": 0.8716,
+      "step": 6649
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6796773451037332,
+      "learning_rate": 1.984702270886566e-06,
+      "loss": 0.9562,
+      "step": 6650
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7955667282605543,
+      "learning_rate": 1.983313606218128e-06,
+      "loss": 0.8918,
+      "step": 6651
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.713753932419462,
+      "learning_rate": 1.981925307333317e-06,
+      "loss": 0.8297,
+      "step": 6652
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7615541066283877,
+      "learning_rate": 1.980537374400465e-06,
+      "loss": 0.904,
+      "step": 6653
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8166905439886167,
+      "learning_rate": 1.979149807587866e-06,
+      "loss": 0.8986,
+      "step": 6654
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6316944122374408,
+      "learning_rate": 1.97776260706377e-06,
+      "loss": 0.8339,
+      "step": 6655
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.802952069216094,
+      "learning_rate": 1.9763757729963746e-06,
+      "loss": 0.8883,
+      "step": 6656
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7396031612093914,
+      "learning_rate": 1.974989305553841e-06,
+      "loss": 0.8402,
+      "step": 6657
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7589201487340693,
+      "learning_rate": 1.9736032049042846e-06,
+      "loss": 0.8751,
+      "step": 6658
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7807386429536298,
+      "learning_rate": 1.9722174712157714e-06,
+      "loss": 0.894,
+      "step": 6659
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7354546466718792,
+      "learning_rate": 1.97083210465633e-06,
+      "loss": 0.9415,
+      "step": 6660
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7986314757071624,
+      "learning_rate": 1.9694471053939375e-06,
+      "loss": 1.0571,
+      "step": 6661
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7314159187374094,
+      "learning_rate": 1.9680624735965327e-06,
+      "loss": 0.898,
+      "step": 6662
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.721440805795135,
+      "learning_rate": 1.966678209432004e-06,
+      "loss": 0.9505,
+      "step": 6663
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6597960366376177,
+      "learning_rate": 1.965294313068202e-06,
+      "loss": 0.8927,
+      "step": 6664
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6118661143688366,
+      "learning_rate": 1.9639107846729246e-06,
+      "loss": 0.9029,
+      "step": 6665
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7697234615120874,
+      "learning_rate": 1.962527624413932e-06,
+      "loss": 0.9011,
+      "step": 6666
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.081322024491878,
+      "learning_rate": 1.961144832458938e-06,
+      "loss": 0.8904,
+      "step": 6667
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.704638803918307,
+      "learning_rate": 1.9597624089756078e-06,
+      "loss": 0.7947,
+      "step": 6668
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8230635386830232,
+      "learning_rate": 1.958380354131567e-06,
+      "loss": 0.8,
+      "step": 6669
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.616479410968544,
+      "learning_rate": 1.9569986680943958e-06,
+      "loss": 0.8914,
+      "step": 6670
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.658407105869247,
+      "learning_rate": 1.9556173510316273e-06,
+      "loss": 0.835,
+      "step": 6671
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6491299562393036,
+      "learning_rate": 1.9542364031107486e-06,
+      "loss": 0.8901,
+      "step": 6672
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7554670452497472,
+      "learning_rate": 1.9528558244992063e-06,
+      "loss": 0.8982,
+      "step": 6673
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8739560360919851,
+      "learning_rate": 1.9514756153644025e-06,
+      "loss": 0.9125,
+      "step": 6674
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7009730922870576,
+      "learning_rate": 1.9500957758736884e-06,
+      "loss": 0.9347,
+      "step": 6675
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8331114646630717,
+      "learning_rate": 1.948716306194376e-06,
+      "loss": 0.769,
+      "step": 6676
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7212746599080477,
+      "learning_rate": 1.947337206493732e-06,
+      "loss": 0.9008,
+      "step": 6677
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7166641488457437,
+      "learning_rate": 1.945958476938975e-06,
+      "loss": 0.8605,
+      "step": 6678
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6224586526106506,
+      "learning_rate": 1.944580117697281e-06,
+      "loss": 0.8318,
+      "step": 6679
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7261616320284379,
+      "learning_rate": 1.9432021289357833e-06,
+      "loss": 0.8132,
+      "step": 6680
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.818350782995743,
+      "learning_rate": 1.941824510821564e-06,
+      "loss": 1.0489,
+      "step": 6681
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.638555482437467,
+      "learning_rate": 1.9404472635216674e-06,
+      "loss": 0.8358,
+      "step": 6682
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7424471071570804,
+      "learning_rate": 1.9390703872030885e-06,
+      "loss": 0.9079,
+      "step": 6683
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8907141059788921,
+      "learning_rate": 1.937693882032776e-06,
+      "loss": 0.8711,
+      "step": 6684
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8659250193383614,
+      "learning_rate": 1.936317748177637e-06,
+      "loss": 0.8603,
+      "step": 6685
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9011114946213044,
+      "learning_rate": 1.934941985804536e-06,
+      "loss": 0.8845,
+      "step": 6686
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6517142975114083,
+      "learning_rate": 1.933566595080284e-06,
+      "loss": 0.9299,
+      "step": 6687
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7111427009698037,
+      "learning_rate": 1.9321915761716536e-06,
+      "loss": 0.8937,
+      "step": 6688
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6586672732435885,
+      "learning_rate": 1.930816929245373e-06,
+      "loss": 0.9177,
+      "step": 6689
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7588969376173822,
+      "learning_rate": 1.9294426544681183e-06,
+      "loss": 1.0057,
+      "step": 6690
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8007735289826785,
+      "learning_rate": 1.9280687520065285e-06,
+      "loss": 1.0275,
+      "step": 6691
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9322877696298502,
+      "learning_rate": 1.926695222027194e-06,
+      "loss": 0.8988,
+      "step": 6692
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8112134553442374,
+      "learning_rate": 1.9253220646966596e-06,
+      "loss": 1.0534,
+      "step": 6693
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7972933628167935,
+      "learning_rate": 1.923949280181423e-06,
+      "loss": 0.9698,
+      "step": 6694
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7053878352988527,
+      "learning_rate": 1.922576868647943e-06,
+      "loss": 0.8591,
+      "step": 6695
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8280487038002564,
+      "learning_rate": 1.921204830262625e-06,
+      "loss": 0.8398,
+      "step": 6696
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7834764977368278,
+      "learning_rate": 1.919833165191836e-06,
+      "loss": 0.8928,
+      "step": 6697
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.738724498861658,
+      "learning_rate": 1.918461873601896e-06,
+      "loss": 0.9567,
+      "step": 6698
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.757378633708845,
+      "learning_rate": 1.917090955659076e-06,
+      "loss": 0.8507,
+      "step": 6699
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6925602036805698,
+      "learning_rate": 1.9157204115296063e-06,
+      "loss": 0.8778,
+      "step": 6700
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7886945433566005,
+      "learning_rate": 1.914350241379671e-06,
+      "loss": 0.9811,
+      "step": 6701
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6705505192461292,
+      "learning_rate": 1.9129804453754053e-06,
+      "loss": 0.8708,
+      "step": 6702
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7300394029316584,
+      "learning_rate": 1.911611023682905e-06,
+      "loss": 0.8864,
+      "step": 6703
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7045684385721467,
+      "learning_rate": 1.9102419764682133e-06,
+      "loss": 0.901,
+      "step": 6704
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.683789075295256,
+      "learning_rate": 1.908873303897336e-06,
+      "loss": 0.8427,
+      "step": 6705
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7609103524366228,
+      "learning_rate": 1.9075050061362255e-06,
+      "loss": 0.8994,
+      "step": 6706
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8884080763692728,
+      "learning_rate": 1.9061370833507947e-06,
+      "loss": 0.9191,
+      "step": 6707
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6898067393719378,
+      "learning_rate": 1.90476953570691e-06,
+      "loss": 0.9479,
+      "step": 6708
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8133154312922324,
+      "learning_rate": 1.9034023633703885e-06,
+      "loss": 1.0248,
+      "step": 6709
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7610835156185274,
+      "learning_rate": 1.9020355665070062e-06,
+      "loss": 1.0294,
+      "step": 6710
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7318741600489804,
+      "learning_rate": 1.9006691452824932e-06,
+      "loss": 0.8938,
+      "step": 6711
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.714713643247803,
+      "learning_rate": 1.8993030998625295e-06,
+      "loss": 0.9123,
+      "step": 6712
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7064021100613893,
+      "learning_rate": 1.8979374304127567e-06,
+      "loss": 0.9344,
+      "step": 6713
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.011661736289325,
+      "learning_rate": 1.896572137098765e-06,
+      "loss": 0.9062,
+      "step": 6714
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7904086199943855,
+      "learning_rate": 1.8952072200860988e-06,
+      "loss": 1.0447,
+      "step": 6715
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7996827818822383,
+      "learning_rate": 1.8938426795402614e-06,
+      "loss": 0.8891,
+      "step": 6716
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7096018071105226,
+      "learning_rate": 1.8924785156267089e-06,
+      "loss": 0.8354,
+      "step": 6717
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.767828307489623,
+      "learning_rate": 1.8911147285108484e-06,
+      "loss": 0.9179,
+      "step": 6718
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7271965404623815,
+      "learning_rate": 1.8897513183580445e-06,
+      "loss": 0.8816,
+      "step": 6719
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1629365776725202,
+      "learning_rate": 1.8883882853336183e-06,
+      "loss": 0.8545,
+      "step": 6720
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.681382392205773,
+      "learning_rate": 1.8870256296028378e-06,
+      "loss": 0.8798,
+      "step": 6721
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.729826254847735,
+      "learning_rate": 1.8856633513309313e-06,
+      "loss": 0.9264,
+      "step": 6722
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7460456052860156,
+      "learning_rate": 1.8843014506830826e-06,
+      "loss": 0.9406,
+      "step": 6723
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7895398124502808,
+      "learning_rate": 1.8829399278244243e-06,
+      "loss": 0.9232,
+      "step": 6724
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7847565912451042,
+      "learning_rate": 1.8815787829200438e-06,
+      "loss": 0.9072,
+      "step": 6725
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7191372852981854,
+      "learning_rate": 1.880218016134987e-06,
+      "loss": 0.9062,
+      "step": 6726
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6997387214582256,
+      "learning_rate": 1.878857627634253e-06,
+      "loss": 0.923,
+      "step": 6727
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.690622040158969,
+      "learning_rate": 1.8774976175827898e-06,
+      "loss": 0.9001,
+      "step": 6728
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7551763542787262,
+      "learning_rate": 1.8761379861455076e-06,
+      "loss": 0.8547,
+      "step": 6729
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9137302223990171,
+      "learning_rate": 1.874778733487262e-06,
+      "loss": 0.8308,
+      "step": 6730
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6328445469449737,
+      "learning_rate": 1.8734198597728698e-06,
+      "loss": 0.9271,
+      "step": 6731
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7170599354966187,
+      "learning_rate": 1.8720613651670999e-06,
+      "loss": 0.8448,
+      "step": 6732
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5388679068838917,
+      "learning_rate": 1.8707032498346711e-06,
+      "loss": 0.913,
+      "step": 6733
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6875461713603581,
+      "learning_rate": 1.8693455139402621e-06,
+      "loss": 0.8645,
+      "step": 6734
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8153761450798231,
+      "learning_rate": 1.8679881576485042e-06,
+      "loss": 0.91,
+      "step": 6735
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7268004723532482,
+      "learning_rate": 1.8666311811239796e-06,
+      "loss": 0.9819,
+      "step": 6736
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8440627641595173,
+      "learning_rate": 1.8652745845312253e-06,
+      "loss": 0.8178,
+      "step": 6737
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.819933123721399,
+      "learning_rate": 1.8639183680347345e-06,
+      "loss": 0.9127,
+      "step": 6738
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7152058404761696,
+      "learning_rate": 1.8625625317989548e-06,
+      "loss": 0.9468,
+      "step": 6739
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.615328647068905,
+      "learning_rate": 1.8612070759882827e-06,
+      "loss": 0.811,
+      "step": 6740
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7946315662042833,
+      "learning_rate": 1.8598520007670744e-06,
+      "loss": 0.9438,
+      "step": 6741
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.727598999381991,
+      "learning_rate": 1.8584973062996382e-06,
+      "loss": 0.8616,
+      "step": 6742
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7560514474056543,
+      "learning_rate": 1.8571429927502322e-06,
+      "loss": 0.8657,
+      "step": 6743
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7969532431497461,
+      "learning_rate": 1.8557890602830731e-06,
+      "loss": 0.8704,
+      "step": 6744
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7846208435150089,
+      "learning_rate": 1.854435509062334e-06,
+      "loss": 1.0643,
+      "step": 6745
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6655694653684243,
+      "learning_rate": 1.8530823392521302e-06,
+      "loss": 0.9401,
+      "step": 6746
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8366621710916666,
+      "learning_rate": 1.8517295510165416e-06,
+      "loss": 0.9651,
+      "step": 6747
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6920321586426823,
+      "learning_rate": 1.8503771445196005e-06,
+      "loss": 0.8207,
+      "step": 6748
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6308593347114158,
+      "learning_rate": 1.849025119925288e-06,
+      "loss": 0.9872,
+      "step": 6749
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.611261217699422,
+      "learning_rate": 1.847673477397542e-06,
+      "loss": 0.9366,
+      "step": 6750
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.735059774018659,
+      "learning_rate": 1.8463222171002565e-06,
+      "loss": 0.8652,
+      "step": 6751
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7260471724143036,
+      "learning_rate": 1.8449713391972734e-06,
+      "loss": 0.9177,
+      "step": 6752
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.77926381564285,
+      "learning_rate": 1.8436208438523916e-06,
+      "loss": 0.9955,
+      "step": 6753
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8449248119200312,
+      "learning_rate": 1.8422707312293663e-06,
+      "loss": 0.8881,
+      "step": 6754
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7865063621936539,
+      "learning_rate": 1.8409210014918994e-06,
+      "loss": 0.9436,
+      "step": 6755
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8172921972911207,
+      "learning_rate": 1.839571654803654e-06,
+      "loss": 0.9105,
+      "step": 6756
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.625792545752456,
+      "learning_rate": 1.8382226913282391e-06,
+      "loss": 0.9343,
+      "step": 6757
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7477379331484049,
+      "learning_rate": 1.8368741112292255e-06,
+      "loss": 0.9182,
+      "step": 6758
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7917617849647247,
+      "learning_rate": 1.8355259146701282e-06,
+      "loss": 0.9491,
+      "step": 6759
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7295379874476284,
+      "learning_rate": 1.8341781018144256e-06,
+      "loss": 0.8652,
+      "step": 6760
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7271874629701593,
+      "learning_rate": 1.8328306728255407e-06,
+      "loss": 0.8769,
+      "step": 6761
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.741468574340731,
+      "learning_rate": 1.8314836278668558e-06,
+      "loss": 0.9108,
+      "step": 6762
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7484048436737174,
+      "learning_rate": 1.830136967101706e-06,
+      "loss": 0.9058,
+      "step": 6763
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7055018806028972,
+      "learning_rate": 1.8287906906933755e-06,
+      "loss": 0.9225,
+      "step": 6764
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6526425958890114,
+      "learning_rate": 1.8274447988051065e-06,
+      "loss": 0.8998,
+      "step": 6765
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7344644028285687,
+      "learning_rate": 1.8260992916000947e-06,
+      "loss": 0.863,
+      "step": 6766
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6894064893115037,
+      "learning_rate": 1.824754169241486e-06,
+      "loss": 0.9949,
+      "step": 6767
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7574064312687918,
+      "learning_rate": 1.8234094318923795e-06,
+      "loss": 0.8673,
+      "step": 6768
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.731909227435955,
+      "learning_rate": 1.82206507971583e-06,
+      "loss": 0.8098,
+      "step": 6769
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.799540911919679,
+      "learning_rate": 1.8207211128748482e-06,
+      "loss": 1.0364,
+      "step": 6770
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8189181789756352,
+      "learning_rate": 1.8193775315323903e-06,
+      "loss": 0.891,
+      "step": 6771
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.734861886118441,
+      "learning_rate": 1.8180343358513714e-06,
+      "loss": 0.8983,
+      "step": 6772
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7261942102713896,
+      "learning_rate": 1.8166915259946617e-06,
+      "loss": 0.947,
+      "step": 6773
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7020924554472685,
+      "learning_rate": 1.8153491021250763e-06,
+      "loss": 0.8972,
+      "step": 6774
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6674441194986105,
+      "learning_rate": 1.814007064405392e-06,
+      "loss": 0.8661,
+      "step": 6775
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8388079950967003,
+      "learning_rate": 1.8126654129983367e-06,
+      "loss": 1.0347,
+      "step": 6776
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7876221362515354,
+      "learning_rate": 1.8113241480665883e-06,
+      "loss": 1.0516,
+      "step": 6777
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.871283718528524,
+      "learning_rate": 1.8099832697727787e-06,
+      "loss": 0.8717,
+      "step": 6778
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7801931243741674,
+      "learning_rate": 1.8086427782794964e-06,
+      "loss": 1.0158,
+      "step": 6779
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.762568277918021,
+      "learning_rate": 1.8073026737492783e-06,
+      "loss": 0.857,
+      "step": 6780
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.688306545191722,
+      "learning_rate": 1.8059629563446173e-06,
+      "loss": 0.8758,
+      "step": 6781
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8380639240245642,
+      "learning_rate": 1.8046236262279615e-06,
+      "loss": 0.8622,
+      "step": 6782
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6784512123660265,
+      "learning_rate": 1.8032846835617056e-06,
+      "loss": 0.8366,
+      "step": 6783
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6943559806938433,
+      "learning_rate": 1.8019461285082023e-06,
+      "loss": 0.8704,
+      "step": 6784
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.799322959335075,
+      "learning_rate": 1.8006079612297583e-06,
+      "loss": 1.043,
+      "step": 6785
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.730418408913277,
+      "learning_rate": 1.7992701818886271e-06,
+      "loss": 0.8881,
+      "step": 6786
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8020664794691135,
+      "learning_rate": 1.7979327906470206e-06,
+      "loss": 0.842,
+      "step": 6787
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.762974201609828,
+      "learning_rate": 1.7965957876671047e-06,
+      "loss": 0.9212,
+      "step": 6788
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7980887024419843,
+      "learning_rate": 1.795259173110993e-06,
+      "loss": 0.8537,
+      "step": 6789
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6637994213221692,
+      "learning_rate": 1.7939229471407532e-06,
+      "loss": 0.9067,
+      "step": 6790
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6649602236552614,
+      "learning_rate": 1.792587109918411e-06,
+      "loss": 0.8135,
+      "step": 6791
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.681264371358223,
+      "learning_rate": 1.7912516616059377e-06,
+      "loss": 0.8342,
+      "step": 6792
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.910552699651834,
+      "learning_rate": 1.7899166023652625e-06,
+      "loss": 0.9898,
+      "step": 6793
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6727610248705524,
+      "learning_rate": 1.788581932358268e-06,
+      "loss": 0.8477,
+      "step": 6794
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7248290584335482,
+      "learning_rate": 1.787247651746784e-06,
+      "loss": 0.8808,
+      "step": 6795
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.769438178138947,
+      "learning_rate": 1.785913760692598e-06,
+      "loss": 0.8693,
+      "step": 6796
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6143237499419798,
+      "learning_rate": 1.7845802593574512e-06,
+      "loss": 0.8722,
+      "step": 6797
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.689145053575692,
+      "learning_rate": 1.7832471479030328e-06,
+      "loss": 0.8824,
+      "step": 6798
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7227798498392426,
+      "learning_rate": 1.781914426490986e-06,
+      "loss": 0.925,
+      "step": 6799
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8114971463745766,
+      "learning_rate": 1.7805820952829095e-06,
+      "loss": 0.785,
+      "step": 6800
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8374163404893544,
+      "learning_rate": 1.7792501544403545e-06,
+      "loss": 0.9134,
+      "step": 6801
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6712388218340632,
+      "learning_rate": 1.7779186041248204e-06,
+      "loss": 0.8822,
+      "step": 6802
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8459740488456966,
+      "learning_rate": 1.7765874444977638e-06,
+      "loss": 0.8959,
+      "step": 6803
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8000833429945955,
+      "learning_rate": 1.7752566757205936e-06,
+      "loss": 1.0394,
+      "step": 6804
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7143934114306176,
+      "learning_rate": 1.7739262979546674e-06,
+      "loss": 0.8324,
+      "step": 6805
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.670417976056517,
+      "learning_rate": 1.7725963113612998e-06,
+      "loss": 0.9585,
+      "step": 6806
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7415364605358754,
+      "learning_rate": 1.771266716101757e-06,
+      "loss": 0.8391,
+      "step": 6807
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6931484007181608,
+      "learning_rate": 1.7699375123372554e-06,
+      "loss": 0.8894,
+      "step": 6808
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6954730000542293,
+      "learning_rate": 1.768608700228967e-06,
+      "loss": 0.9063,
+      "step": 6809
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.690278557438335,
+      "learning_rate": 1.767280279938014e-06,
+      "loss": 0.8457,
+      "step": 6810
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6597553780400327,
+      "learning_rate": 1.7659522516254707e-06,
+      "loss": 0.9008,
+      "step": 6811
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7089649572179197,
+      "learning_rate": 1.764624615452366e-06,
+      "loss": 0.8602,
+      "step": 6812
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7441067620336643,
+      "learning_rate": 1.7632973715796825e-06,
+      "loss": 0.8875,
+      "step": 6813
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6035093273385714,
+      "learning_rate": 1.7619705201683495e-06,
+      "loss": 0.8492,
+      "step": 6814
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5987850700137118,
+      "learning_rate": 1.7606440613792547e-06,
+      "loss": 0.8182,
+      "step": 6815
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7140255392667343,
+      "learning_rate": 1.7593179953732364e-06,
+      "loss": 0.8908,
+      "step": 6816
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.640938883807315,
+      "learning_rate": 1.7579923223110818e-06,
+      "loss": 0.9475,
+      "step": 6817
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8509082320612162,
+      "learning_rate": 1.7566670423535348e-06,
+      "loss": 0.8842,
+      "step": 6818
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7037062969734111,
+      "learning_rate": 1.7553421556612927e-06,
+      "loss": 0.8815,
+      "step": 6819
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6966699520604616,
+      "learning_rate": 1.754017662395e-06,
+      "loss": 0.9197,
+      "step": 6820
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8031852482407524,
+      "learning_rate": 1.7526935627152542e-06,
+      "loss": 0.9265,
+      "step": 6821
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7120420373925513,
+      "learning_rate": 1.75136985678261e-06,
+      "loss": 0.9101,
+      "step": 6822
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7524470028123695,
+      "learning_rate": 1.750046544757571e-06,
+      "loss": 0.9565,
+      "step": 6823
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6446071448235784,
+      "learning_rate": 1.7487236268005919e-06,
+      "loss": 0.8612,
+      "step": 6824
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6151391056585793,
+      "learning_rate": 1.7474011030720834e-06,
+      "loss": 0.9167,
+      "step": 6825
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7168642476755656,
+      "learning_rate": 1.7460789737324025e-06,
+      "loss": 0.8656,
+      "step": 6826
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.764050426994709,
+      "learning_rate": 1.7447572389418643e-06,
+      "loss": 0.8756,
+      "step": 6827
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.832349362677392,
+      "learning_rate": 1.7434358988607353e-06,
+      "loss": 0.9614,
+      "step": 6828
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.751617123529363,
+      "learning_rate": 1.7421149536492282e-06,
+      "loss": 0.8392,
+      "step": 6829
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7247464499008838,
+      "learning_rate": 1.7407944034675173e-06,
+      "loss": 0.861,
+      "step": 6830
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7882554100636091,
+      "learning_rate": 1.7394742484757187e-06,
+      "loss": 1.0422,
+      "step": 6831
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5938085899511738,
+      "learning_rate": 1.7381544888339103e-06,
+      "loss": 0.9301,
+      "step": 6832
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7140139302773292,
+      "learning_rate": 1.7368351247021136e-06,
+      "loss": 0.8918,
+      "step": 6833
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7270086826781677,
+      "learning_rate": 1.7355161562403078e-06,
+      "loss": 0.81,
+      "step": 6834
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5778159210099778,
+      "learning_rate": 1.7341975836084247e-06,
+      "loss": 0.8605,
+      "step": 6835
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6597368886586847,
+      "learning_rate": 1.7328794069663414e-06,
+      "loss": 0.804,
+      "step": 6836
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7034721406773479,
+      "learning_rate": 1.7315616264738933e-06,
+      "loss": 0.9237,
+      "step": 6837
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7029987397866007,
+      "learning_rate": 1.7302442422908677e-06,
+      "loss": 0.8979,
+      "step": 6838
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8591539241036474,
+      "learning_rate": 1.7289272545769987e-06,
+      "loss": 0.9419,
+      "step": 6839
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7861626747357349,
+      "learning_rate": 1.7276106634919775e-06,
+      "loss": 0.8741,
+      "step": 6840
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6828618522195882,
+      "learning_rate": 1.726294469195448e-06,
+      "loss": 0.8585,
+      "step": 6841
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.816664303416701,
+      "learning_rate": 1.7249786718469968e-06,
+      "loss": 1.0151,
+      "step": 6842
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7936861674238076,
+      "learning_rate": 1.723663271606173e-06,
+      "loss": 1.067,
+      "step": 6843
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7433202108712524,
+      "learning_rate": 1.7223482686324738e-06,
+      "loss": 0.9628,
+      "step": 6844
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6835446560491487,
+      "learning_rate": 1.7210336630853453e-06,
+      "loss": 0.8582,
+      "step": 6845
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7244429339364802,
+      "learning_rate": 1.7197194551241896e-06,
+      "loss": 0.8189,
+      "step": 6846
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8230269255135336,
+      "learning_rate": 1.7184056449083603e-06,
+      "loss": 0.9038,
+      "step": 6847
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7982135554042011,
+      "learning_rate": 1.7170922325971585e-06,
+      "loss": 0.8781,
+      "step": 6848
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5882710165572504,
+      "learning_rate": 1.7157792183498413e-06,
+      "loss": 0.9338,
+      "step": 6849
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8340272743340436,
+      "learning_rate": 1.714466602325618e-06,
+      "loss": 0.8819,
+      "step": 6850
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8709551624988427,
+      "learning_rate": 1.7131543846836457e-06,
+      "loss": 0.8402,
+      "step": 6851
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6349161666153629,
+      "learning_rate": 1.7118425655830346e-06,
+      "loss": 0.9073,
+      "step": 6852
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9498820828583627,
+      "learning_rate": 1.7105311451828483e-06,
+      "loss": 0.9421,
+      "step": 6853
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.698467816983888,
+      "learning_rate": 1.709220123642103e-06,
+      "loss": 0.9279,
+      "step": 6854
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6953958669953584,
+      "learning_rate": 1.707909501119761e-06,
+      "loss": 0.8229,
+      "step": 6855
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8425572802114498,
+      "learning_rate": 1.706599277774743e-06,
+      "loss": 0.8765,
+      "step": 6856
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7068201126253246,
+      "learning_rate": 1.705289453765915e-06,
+      "loss": 0.913,
+      "step": 6857
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7751368991901568,
+      "learning_rate": 1.7039800292520997e-06,
+      "loss": 0.8991,
+      "step": 6858
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7273887440098223,
+      "learning_rate": 1.7026710043920702e-06,
+      "loss": 0.9214,
+      "step": 6859
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7727513397365018,
+      "learning_rate": 1.701362379344547e-06,
+      "loss": 0.8671,
+      "step": 6860
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6185274977732638,
+      "learning_rate": 1.7000541542682087e-06,
+      "loss": 0.8384,
+      "step": 6861
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.621562280134425,
+      "learning_rate": 1.6987463293216817e-06,
+      "loss": 0.9192,
+      "step": 6862
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7044073868024594,
+      "learning_rate": 1.697438904663543e-06,
+      "loss": 0.9034,
+      "step": 6863
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8136007489314534,
+      "learning_rate": 1.696131880452322e-06,
+      "loss": 1.0781,
+      "step": 6864
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.582588482545356,
+      "learning_rate": 1.6948252568465001e-06,
+      "loss": 0.7868,
+      "step": 6865
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6796319411133531,
+      "learning_rate": 1.6935190340045121e-06,
+      "loss": 0.8994,
+      "step": 6866
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7769314599767678,
+      "learning_rate": 1.692213212084739e-06,
+      "loss": 0.9106,
+      "step": 6867
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.700661313506012,
+      "learning_rate": 1.690907791245518e-06,
+      "loss": 0.9092,
+      "step": 6868
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7341927765821927,
+      "learning_rate": 1.6896027716451364e-06,
+      "loss": 0.9427,
+      "step": 6869
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.697018346472416,
+      "learning_rate": 1.6882981534418301e-06,
+      "loss": 0.9228,
+      "step": 6870
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.775873086513128,
+      "learning_rate": 1.6869939367937899e-06,
+      "loss": 0.9057,
+      "step": 6871
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8133729458009092,
+      "learning_rate": 1.6856901218591598e-06,
+      "loss": 1.0886,
+      "step": 6872
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0108320401450697,
+      "learning_rate": 1.6843867087960252e-06,
+      "loss": 0.9878,
+      "step": 6873
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7113148831228198,
+      "learning_rate": 1.6830836977624326e-06,
+      "loss": 0.8872,
+      "step": 6874
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8987783013882376,
+      "learning_rate": 1.6817810889163788e-06,
+      "loss": 0.8558,
+      "step": 6875
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8668357817895518,
+      "learning_rate": 1.6804788824158058e-06,
+      "loss": 0.9344,
+      "step": 6876
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6915126402369727,
+      "learning_rate": 1.6791770784186128e-06,
+      "loss": 0.9469,
+      "step": 6877
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8836868755669443,
+      "learning_rate": 1.6778756770826488e-06,
+      "loss": 0.9198,
+      "step": 6878
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.703013353621644,
+      "learning_rate": 1.6765746785657104e-06,
+      "loss": 0.8972,
+      "step": 6879
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8463880418994392,
+      "learning_rate": 1.6752740830255503e-06,
+      "loss": 0.9379,
+      "step": 6880
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.660280151209288,
+      "learning_rate": 1.6739738906198711e-06,
+      "loss": 0.9099,
+      "step": 6881
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6430753569167227,
+      "learning_rate": 1.6726741015063225e-06,
+      "loss": 0.9551,
+      "step": 6882
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7091741672406278,
+      "learning_rate": 1.671374715842512e-06,
+      "loss": 0.902,
+      "step": 6883
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7254690458154878,
+      "learning_rate": 1.6700757337859907e-06,
+      "loss": 0.8829,
+      "step": 6884
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7993757151710104,
+      "learning_rate": 1.668777155494269e-06,
+      "loss": 0.9484,
+      "step": 6885
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.804461091842194,
+      "learning_rate": 1.6674789811247993e-06,
+      "loss": 0.8301,
+      "step": 6886
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7916770419703338,
+      "learning_rate": 1.666181210834994e-06,
+      "loss": 0.9179,
+      "step": 6887
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7870747253228467,
+      "learning_rate": 1.6648838447822084e-06,
+      "loss": 0.8647,
+      "step": 6888
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7699141677728926,
+      "learning_rate": 1.6635868831237551e-06,
+      "loss": 1.0449,
+      "step": 6889
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7952977059862438,
+      "learning_rate": 1.6622903260168954e-06,
+      "loss": 0.9131,
+      "step": 6890
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7777452815092758,
+      "learning_rate": 1.6609941736188395e-06,
+      "loss": 0.939,
+      "step": 6891
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.677261198504098,
+      "learning_rate": 1.6596984260867516e-06,
+      "loss": 0.9256,
+      "step": 6892
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6643533624785818,
+      "learning_rate": 1.6584030835777465e-06,
+      "loss": 0.8062,
+      "step": 6893
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6822830117344607,
+      "learning_rate": 1.6571081462488876e-06,
+      "loss": 0.8812,
+      "step": 6894
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8064495220306743,
+      "learning_rate": 1.655813614257189e-06,
+      "loss": 1.0295,
+      "step": 6895
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8491145921954455,
+      "learning_rate": 1.6545194877596194e-06,
+      "loss": 0.924,
+      "step": 6896
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7273679433025995,
+      "learning_rate": 1.6532257669130968e-06,
+      "loss": 0.9185,
+      "step": 6897
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7384990424684188,
+      "learning_rate": 1.6519324518744867e-06,
+      "loss": 0.881,
+      "step": 6898
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.917735422829965,
+      "learning_rate": 1.650639542800609e-06,
+      "loss": 0.8958,
+      "step": 6899
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.733737055136843,
+      "learning_rate": 1.6493470398482352e-06,
+      "loss": 0.8784,
+      "step": 6900
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7076256448992382,
+      "learning_rate": 1.6480549431740834e-06,
+      "loss": 0.9114,
+      "step": 6901
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6846432331129724,
+      "learning_rate": 1.646763252934825e-06,
+      "loss": 0.804,
+      "step": 6902
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.725923458970688,
+      "learning_rate": 1.645471969287084e-06,
+      "loss": 0.9125,
+      "step": 6903
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7679344107230115,
+      "learning_rate": 1.644181092387432e-06,
+      "loss": 0.8939,
+      "step": 6904
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6542075245429493,
+      "learning_rate": 1.6428906223923902e-06,
+      "loss": 0.8355,
+      "step": 6905
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6529529041152011,
+      "learning_rate": 1.6416005594584355e-06,
+      "loss": 0.9187,
+      "step": 6906
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8035164497075127,
+      "learning_rate": 1.6403109037419895e-06,
+      "loss": 1.0725,
+      "step": 6907
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7814105908416771,
+      "learning_rate": 1.6390216553994292e-06,
+      "loss": 1.0219,
+      "step": 6908
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8286350273943441,
+      "learning_rate": 1.6377328145870824e-06,
+      "loss": 0.7898,
+      "step": 6909
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6227577853108464,
+      "learning_rate": 1.6364443814612207e-06,
+      "loss": 0.8287,
+      "step": 6910
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6417595165842345,
+      "learning_rate": 1.6351563561780742e-06,
+      "loss": 0.8393,
+      "step": 6911
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7840474310224526,
+      "learning_rate": 1.6338687388938218e-06,
+      "loss": 0.9293,
+      "step": 6912
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6835820108095845,
+      "learning_rate": 1.6325815297645875e-06,
+      "loss": 0.9643,
+      "step": 6913
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5435596842207269,
+      "learning_rate": 1.631294728946452e-06,
+      "loss": 0.9091,
+      "step": 6914
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6891954212070406,
+      "learning_rate": 1.6300083365954462e-06,
+      "loss": 0.9482,
+      "step": 6915
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7460184661761975,
+      "learning_rate": 1.6287223528675478e-06,
+      "loss": 0.829,
+      "step": 6916
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8679713008348668,
+      "learning_rate": 1.6274367779186845e-06,
+      "loss": 0.9668,
+      "step": 6917
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7357075312241537,
+      "learning_rate": 1.6261516119047393e-06,
+      "loss": 0.9227,
+      "step": 6918
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7040071331055202,
+      "learning_rate": 1.6248668549815445e-06,
+      "loss": 0.8616,
+      "step": 6919
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6513700612776312,
+      "learning_rate": 1.6235825073048772e-06,
+      "loss": 0.9038,
+      "step": 6920
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7297525587560503,
+      "learning_rate": 1.6222985690304732e-06,
+      "loss": 0.9246,
+      "step": 6921
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7506410967908024,
+      "learning_rate": 1.62101504031401e-06,
+      "loss": 0.8613,
+      "step": 6922
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7940945581054802,
+      "learning_rate": 1.6197319213111235e-06,
+      "loss": 1.0614,
+      "step": 6923
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7397047232705767,
+      "learning_rate": 1.6184492121773958e-06,
+      "loss": 0.882,
+      "step": 6924
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7275506998483345,
+      "learning_rate": 1.6171669130683593e-06,
+      "loss": 0.8368,
+      "step": 6925
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7224315962608217,
+      "learning_rate": 1.6158850241394957e-06,
+      "loss": 0.9116,
+      "step": 6926
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7201018195694147,
+      "learning_rate": 1.6146035455462395e-06,
+      "loss": 0.9029,
+      "step": 6927
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6663574989875263,
+      "learning_rate": 1.613322477443976e-06,
+      "loss": 0.8262,
+      "step": 6928
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7246444915426904,
+      "learning_rate": 1.6120418199880366e-06,
+      "loss": 0.8797,
+      "step": 6929
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7128938946104602,
+      "learning_rate": 1.6107615733337061e-06,
+      "loss": 0.8596,
+      "step": 6930
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7589873126369537,
+      "learning_rate": 1.6094817376362214e-06,
+      "loss": 0.7933,
+      "step": 6931
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6961945299549808,
+      "learning_rate": 1.6082023130507628e-06,
+      "loss": 0.8489,
+      "step": 6932
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7444580569970618,
+      "learning_rate": 1.6069232997324679e-06,
+      "loss": 0.8084,
+      "step": 6933
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7757612283445128,
+      "learning_rate": 1.6056446978364215e-06,
+      "loss": 0.8378,
+      "step": 6934
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7071203088551532,
+      "learning_rate": 1.6043665075176561e-06,
+      "loss": 0.8658,
+      "step": 6935
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6705363177105303,
+      "learning_rate": 1.6030887289311604e-06,
+      "loss": 0.8658,
+      "step": 6936
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6747912177989717,
+      "learning_rate": 1.6018113622318666e-06,
+      "loss": 0.8301,
+      "step": 6937
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9154193765964813,
+      "learning_rate": 1.6005344075746586e-06,
+      "loss": 0.9773,
+      "step": 6938
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.770323721074381,
+      "learning_rate": 1.5992578651143742e-06,
+      "loss": 0.9029,
+      "step": 6939
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8890726432441385,
+      "learning_rate": 1.5979817350057991e-06,
+      "loss": 0.9035,
+      "step": 6940
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6953706263399888,
+      "learning_rate": 1.5967060174036652e-06,
+      "loss": 0.9298,
+      "step": 6941
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6848022856465479,
+      "learning_rate": 1.59543071246266e-06,
+      "loss": 0.8803,
+      "step": 6942
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7171093647064752,
+      "learning_rate": 1.5941558203374198e-06,
+      "loss": 0.9093,
+      "step": 6943
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6756632979798511,
+      "learning_rate": 1.5928813411825267e-06,
+      "loss": 0.9145,
+      "step": 6944
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8209436873978789,
+      "learning_rate": 1.591607275152517e-06,
+      "loss": 1.0502,
+      "step": 6945
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8936097643129102,
+      "learning_rate": 1.5903336224018774e-06,
+      "loss": 0.8349,
+      "step": 6946
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.781170574716261,
+      "learning_rate": 1.5890603830850403e-06,
+      "loss": 0.9187,
+      "step": 6947
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.710306038367912,
+      "learning_rate": 1.58778755735639e-06,
+      "loss": 0.9744,
+      "step": 6948
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8514330108392898,
+      "learning_rate": 1.586515145370262e-06,
+      "loss": 0.9615,
+      "step": 6949
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.700461809792858,
+      "learning_rate": 1.5852431472809426e-06,
+      "loss": 0.9278,
+      "step": 6950
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7220535444817868,
+      "learning_rate": 1.5839715632426623e-06,
+      "loss": 0.834,
+      "step": 6951
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7726919873226807,
+      "learning_rate": 1.582700393409608e-06,
+      "loss": 0.8332,
+      "step": 6952
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7138274532707232,
+      "learning_rate": 1.5814296379359107e-06,
+      "loss": 0.9264,
+      "step": 6953
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6806685161949746,
+      "learning_rate": 1.5801592969756558e-06,
+      "loss": 0.9383,
+      "step": 6954
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7484392381234368,
+      "learning_rate": 1.5788893706828773e-06,
+      "loss": 0.8985,
+      "step": 6955
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.721822778145402,
+      "learning_rate": 1.5776198592115555e-06,
+      "loss": 0.8841,
+      "step": 6956
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.753028902912283,
+      "learning_rate": 1.5763507627156265e-06,
+      "loss": 0.9235,
+      "step": 6957
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6317271525181802,
+      "learning_rate": 1.5750820813489688e-06,
+      "loss": 0.826,
+      "step": 6958
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8122242947217784,
+      "learning_rate": 1.5738138152654175e-06,
+      "loss": 1.0678,
+      "step": 6959
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6383249368385338,
+      "learning_rate": 1.572545964618752e-06,
+      "loss": 0.8806,
+      "step": 6960
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.774295674843163,
+      "learning_rate": 1.5712785295627037e-06,
+      "loss": 0.9169,
+      "step": 6961
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.728543292789014,
+      "learning_rate": 1.5700115102509562e-06,
+      "loss": 0.9564,
+      "step": 6962
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7565814655714278,
+      "learning_rate": 1.5687449068371368e-06,
+      "loss": 0.9207,
+      "step": 6963
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7620906210684442,
+      "learning_rate": 1.5674787194748264e-06,
+      "loss": 0.9111,
+      "step": 6964
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.72802578801936,
+      "learning_rate": 1.566212948317557e-06,
+      "loss": 0.9569,
+      "step": 6965
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8009215546920747,
+      "learning_rate": 1.5649475935188034e-06,
+      "loss": 1.0538,
+      "step": 6966
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7618794239157773,
+      "learning_rate": 1.5636826552319973e-06,
+      "loss": 0.8709,
+      "step": 6967
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6753239312225756,
+      "learning_rate": 1.5624181336105188e-06,
+      "loss": 0.9109,
+      "step": 6968
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7170890255168159,
+      "learning_rate": 1.5611540288076892e-06,
+      "loss": 1.0002,
+      "step": 6969
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.742737651758367,
+      "learning_rate": 1.5598903409767895e-06,
+      "loss": 0.8993,
+      "step": 6970
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7780289816749608,
+      "learning_rate": 1.5586270702710477e-06,
+      "loss": 0.9139,
+      "step": 6971
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6824981207035556,
+      "learning_rate": 1.5573642168436358e-06,
+      "loss": 0.8561,
+      "step": 6972
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7948615409894175,
+      "learning_rate": 1.5561017808476815e-06,
+      "loss": 0.8955,
+      "step": 6973
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8320705521433387,
+      "learning_rate": 1.5548397624362605e-06,
+      "loss": 0.8727,
+      "step": 6974
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6452279104494536,
+      "learning_rate": 1.5535781617623942e-06,
+      "loss": 0.8845,
+      "step": 6975
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7020957105647088,
+      "learning_rate": 1.5523169789790576e-06,
+      "loss": 0.8593,
+      "step": 6976
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8085986244526793,
+      "learning_rate": 1.5510562142391744e-06,
+      "loss": 0.9049,
+      "step": 6977
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7592665236511282,
+      "learning_rate": 1.549795867695616e-06,
+      "loss": 0.9266,
+      "step": 6978
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7920620028604474,
+      "learning_rate": 1.5485359395012012e-06,
+      "loss": 1.0587,
+      "step": 6979
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7050143549083525,
+      "learning_rate": 1.547276429808703e-06,
+      "loss": 0.8679,
+      "step": 6980
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7257886152298951,
+      "learning_rate": 1.5460173387708427e-06,
+      "loss": 0.9222,
+      "step": 6981
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7152734046663634,
+      "learning_rate": 1.544758666540286e-06,
+      "loss": 0.9729,
+      "step": 6982
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7310456787021276,
+      "learning_rate": 1.5435004132696547e-06,
+      "loss": 0.8264,
+      "step": 6983
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.81729238627351,
+      "learning_rate": 1.5422425791115132e-06,
+      "loss": 0.9172,
+      "step": 6984
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8002060094670137,
+      "learning_rate": 1.5409851642183792e-06,
+      "loss": 1.0436,
+      "step": 6985
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9106706803761853,
+      "learning_rate": 1.5397281687427212e-06,
+      "loss": 0.9295,
+      "step": 6986
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7687421945250164,
+      "learning_rate": 1.5384715928369503e-06,
+      "loss": 1.0349,
+      "step": 6987
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7422877874302911,
+      "learning_rate": 1.5372154366534325e-06,
+      "loss": 0.9174,
+      "step": 6988
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8252831550986466,
+      "learning_rate": 1.5359597003444827e-06,
+      "loss": 0.9896,
+      "step": 6989
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7886248774729665,
+      "learning_rate": 1.5347043840623617e-06,
+      "loss": 0.8882,
+      "step": 6990
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.716980050286474,
+      "learning_rate": 1.5334494879592786e-06,
+      "loss": 0.8166,
+      "step": 6991
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6884545459077325,
+      "learning_rate": 1.5321950121873968e-06,
+      "loss": 0.8345,
+      "step": 6992
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6879131541119972,
+      "learning_rate": 1.5309409568988264e-06,
+      "loss": 0.8167,
+      "step": 6993
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6796183951825223,
+      "learning_rate": 1.5296873222456232e-06,
+      "loss": 0.8919,
+      "step": 6994
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7680887274351678,
+      "learning_rate": 1.528434108379796e-06,
+      "loss": 0.8589,
+      "step": 6995
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8847060692184536,
+      "learning_rate": 1.5271813154533033e-06,
+      "loss": 0.9908,
+      "step": 6996
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.931416564405431,
+      "learning_rate": 1.5259289436180468e-06,
+      "loss": 0.9782,
+      "step": 6997
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7510080313844953,
+      "learning_rate": 1.524676993025883e-06,
+      "loss": 0.8821,
+      "step": 6998
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8177369009062037,
+      "learning_rate": 1.5234254638286183e-06,
+      "loss": 0.8809,
+      "step": 6999
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8762523144608232,
+      "learning_rate": 1.5221743561779988e-06,
+      "loss": 0.8537,
+      "step": 7000
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.37520519563173,
+      "learning_rate": 1.520923670225728e-06,
+      "loss": 0.9237,
+      "step": 7001
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7973831831114462,
+      "learning_rate": 1.5196734061234581e-06,
+      "loss": 0.8955,
+      "step": 7002
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7370570396784173,
+      "learning_rate": 1.5184235640227846e-06,
+      "loss": 0.8164,
+      "step": 7003
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8576582802817276,
+      "learning_rate": 1.5171741440752568e-06,
+      "loss": 0.9282,
+      "step": 7004
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.777703248050023,
+      "learning_rate": 1.5159251464323732e-06,
+      "loss": 0.8657,
+      "step": 7005
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.086855378505545,
+      "learning_rate": 1.5146765712455747e-06,
+      "loss": 0.8705,
+      "step": 7006
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7993501356761152,
+      "learning_rate": 1.5134284186662585e-06,
+      "loss": 0.8701,
+      "step": 7007
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6677224561185113,
+      "learning_rate": 1.5121806888457674e-06,
+      "loss": 0.8851,
+      "step": 7008
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7645903764555948,
+      "learning_rate": 1.5109333819353904e-06,
+      "loss": 1.0185,
+      "step": 7009
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7736482499444295,
+      "learning_rate": 1.5096864980863718e-06,
+      "loss": 0.8592,
+      "step": 7010
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.738990667294761,
+      "learning_rate": 1.5084400374498964e-06,
+      "loss": 0.9233,
+      "step": 7011
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8197102317284632,
+      "learning_rate": 1.507194000177105e-06,
+      "loss": 1.0702,
+      "step": 7012
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7982266840924497,
+      "learning_rate": 1.5059483864190817e-06,
+      "loss": 0.8755,
+      "step": 7013
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6627801880163413,
+      "learning_rate": 1.5047031963268617e-06,
+      "loss": 0.9415,
+      "step": 7014
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8317666729043471,
+      "learning_rate": 1.503458430051431e-06,
+      "loss": 0.8968,
+      "step": 7015
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.72987622534171,
+      "learning_rate": 1.5022140877437186e-06,
+      "loss": 0.9175,
+      "step": 7016
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7600079737548184,
+      "learning_rate": 1.500970169554608e-06,
+      "loss": 0.857,
+      "step": 7017
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0117934657736227,
+      "learning_rate": 1.4997266756349265e-06,
+      "loss": 0.8898,
+      "step": 7018
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8087090775029488,
+      "learning_rate": 1.4984836061354524e-06,
+      "loss": 1.025,
+      "step": 7019
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8811138831201464,
+      "learning_rate": 1.497240961206914e-06,
+      "loss": 0.9023,
+      "step": 7020
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7986943883005735,
+      "learning_rate": 1.4959987409999855e-06,
+      "loss": 0.8772,
+      "step": 7021
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7134760992588565,
+      "learning_rate": 1.4947569456652877e-06,
+      "loss": 0.8729,
+      "step": 7022
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7531675057016387,
+      "learning_rate": 1.4935155753533947e-06,
+      "loss": 0.9017,
+      "step": 7023
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6880249967291248,
+      "learning_rate": 1.4922746302148283e-06,
+      "loss": 0.8957,
+      "step": 7024
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8014702835345631,
+      "learning_rate": 1.4910341104000548e-06,
+      "loss": 1.0335,
+      "step": 7025
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7087155617443626,
+      "learning_rate": 1.4897940160594926e-06,
+      "loss": 0.8772,
+      "step": 7026
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8562933028465036,
+      "learning_rate": 1.4885543473435088e-06,
+      "loss": 0.9163,
+      "step": 7027
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6664291551411319,
+      "learning_rate": 1.4873151044024149e-06,
+      "loss": 0.8162,
+      "step": 7028
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6848821856691956,
+      "learning_rate": 1.4860762873864741e-06,
+      "loss": 0.9225,
+      "step": 7029
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.845625665794136,
+      "learning_rate": 1.4848378964458997e-06,
+      "loss": 0.8225,
+      "step": 7030
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6714334849685386,
+      "learning_rate": 1.4835999317308492e-06,
+      "loss": 0.8776,
+      "step": 7031
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5663729766136538,
+      "learning_rate": 1.4823623933914277e-06,
+      "loss": 0.8214,
+      "step": 7032
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7089695380977425,
+      "learning_rate": 1.4811252815776955e-06,
+      "loss": 0.847,
+      "step": 7033
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.633793044996098,
+      "learning_rate": 1.479888596439652e-06,
+      "loss": 0.8537,
+      "step": 7034
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7454438218440775,
+      "learning_rate": 1.4786523381272521e-06,
+      "loss": 0.9549,
+      "step": 7035
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7765962433907524,
+      "learning_rate": 1.4774165067903984e-06,
+      "loss": 1.0476,
+      "step": 7036
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7271359020886792,
+      "learning_rate": 1.4761811025789352e-06,
+      "loss": 0.9075,
+      "step": 7037
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7392364165148841,
+      "learning_rate": 1.4749461256426617e-06,
+      "loss": 0.8996,
+      "step": 7038
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8701356227209658,
+      "learning_rate": 1.4737115761313249e-06,
+      "loss": 0.9331,
+      "step": 7039
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.658571385100619,
+      "learning_rate": 1.4724774541946145e-06,
+      "loss": 0.9195,
+      "step": 7040
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6540504394261784,
+      "learning_rate": 1.4712437599821743e-06,
+      "loss": 0.8746,
+      "step": 7041
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8527748439086306,
+      "learning_rate": 1.4700104936435955e-06,
+      "loss": 1.0221,
+      "step": 7042
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.74978944045805,
+      "learning_rate": 1.468777655328414e-06,
+      "loss": 0.9274,
+      "step": 7043
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.655168189879266,
+      "learning_rate": 1.467545245186114e-06,
+      "loss": 0.8245,
+      "step": 7044
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7013224245299177,
+      "learning_rate": 1.4663132633661315e-06,
+      "loss": 0.9366,
+      "step": 7045
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7937662254172413,
+      "learning_rate": 1.4650817100178494e-06,
+      "loss": 1.0461,
+      "step": 7046
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7950435440689796,
+      "learning_rate": 1.4638505852905955e-06,
+      "loss": 0.9152,
+      "step": 7047
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.825621835578695,
+      "learning_rate": 1.4626198893336507e-06,
+      "loss": 0.8963,
+      "step": 7048
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6285456283843853,
+      "learning_rate": 1.4613896222962376e-06,
+      "loss": 0.8471,
+      "step": 7049
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6962219734541413,
+      "learning_rate": 1.4601597843275328e-06,
+      "loss": 0.9119,
+      "step": 7050
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7539775794789876,
+      "learning_rate": 1.4589303755766586e-06,
+      "loss": 0.8411,
+      "step": 7051
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7764619057409665,
+      "learning_rate": 1.457701396192685e-06,
+      "loss": 0.8666,
+      "step": 7052
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6101294112231987,
+      "learning_rate": 1.4564728463246274e-06,
+      "loss": 0.8836,
+      "step": 7053
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6723515291619309,
+      "learning_rate": 1.4552447261214536e-06,
+      "loss": 1.0025,
+      "step": 7054
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7730777264588926,
+      "learning_rate": 1.4540170357320788e-06,
+      "loss": 0.8679,
+      "step": 7055
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7547192806745964,
+      "learning_rate": 1.4527897753053621e-06,
+      "loss": 0.9042,
+      "step": 7056
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.728649021008933,
+      "learning_rate": 1.4515629449901142e-06,
+      "loss": 0.8777,
+      "step": 7057
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7011515005087847,
+      "learning_rate": 1.4503365449350938e-06,
+      "loss": 0.8479,
+      "step": 7058
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7414315692740892,
+      "learning_rate": 1.4491105752890034e-06,
+      "loss": 0.9368,
+      "step": 7059
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7345481907189801,
+      "learning_rate": 1.4478850362004976e-06,
+      "loss": 0.852,
+      "step": 7060
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6700186698334887,
+      "learning_rate": 1.4466599278181786e-06,
+      "loss": 0.8989,
+      "step": 7061
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6931873727662659,
+      "learning_rate": 1.4454352502905922e-06,
+      "loss": 0.868,
+      "step": 7062
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7775204931903898,
+      "learning_rate": 1.4442110037662377e-06,
+      "loss": 0.9074,
+      "step": 7063
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7828795903853583,
+      "learning_rate": 1.4429871883935577e-06,
+      "loss": 0.8461,
+      "step": 7064
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7814603854286049,
+      "learning_rate": 1.441763804320942e-06,
+      "loss": 0.8962,
+      "step": 7065
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.695409950223556,
+      "learning_rate": 1.440540851696733e-06,
+      "loss": 0.8856,
+      "step": 7066
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.705997000916623,
+      "learning_rate": 1.4393183306692176e-06,
+      "loss": 0.8671,
+      "step": 7067
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7019530298357515,
+      "learning_rate": 1.4380962413866289e-06,
+      "loss": 0.8279,
+      "step": 7068
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7242429954254317,
+      "learning_rate": 1.4368745839971509e-06,
+      "loss": 0.8827,
+      "step": 7069
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6419293694439554,
+      "learning_rate": 1.4356533586489152e-06,
+      "loss": 0.9082,
+      "step": 7070
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7414080129648566,
+      "learning_rate": 1.4344325654899965e-06,
+      "loss": 0.87,
+      "step": 7071
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.721749899847165,
+      "learning_rate": 1.4332122046684211e-06,
+      "loss": 0.9533,
+      "step": 7072
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.865263929000295,
+      "learning_rate": 1.4319922763321641e-06,
+      "loss": 0.9077,
+      "step": 7073
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.677624904590976,
+      "learning_rate": 1.430772780629145e-06,
+      "loss": 0.9316,
+      "step": 7074
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6404621078309713,
+      "learning_rate": 1.4295537177072289e-06,
+      "loss": 0.9339,
+      "step": 7075
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.679548717134846,
+      "learning_rate": 1.4283350877142343e-06,
+      "loss": 0.9049,
+      "step": 7076
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8451165980034023,
+      "learning_rate": 1.4271168907979249e-06,
+      "loss": 1.011,
+      "step": 7077
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.823830692431705,
+      "learning_rate": 1.4258991271060086e-06,
+      "loss": 0.902,
+      "step": 7078
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6514477058876962,
+      "learning_rate": 1.4246817967861464e-06,
+      "loss": 0.9112,
+      "step": 7079
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7877311718636557,
+      "learning_rate": 1.4234648999859413e-06,
+      "loss": 0.7937,
+      "step": 7080
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7102862580260139,
+      "learning_rate": 1.4222484368529472e-06,
+      "loss": 0.8542,
+      "step": 7081
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6522506584506498,
+      "learning_rate": 1.4210324075346654e-06,
+      "loss": 0.8632,
+      "step": 7082
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.685773869177117,
+      "learning_rate": 1.4198168121785416e-06,
+      "loss": 0.8619,
+      "step": 7083
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7058197976865817,
+      "learning_rate": 1.418601650931974e-06,
+      "loss": 0.8104,
+      "step": 7084
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.640643042487272,
+      "learning_rate": 1.417386923942301e-06,
+      "loss": 0.8195,
+      "step": 7085
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6609778740414582,
+      "learning_rate": 1.4161726313568165e-06,
+      "loss": 0.8795,
+      "step": 7086
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8171929364411803,
+      "learning_rate": 1.4149587733227543e-06,
+      "loss": 0.8257,
+      "step": 7087
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7990363265696868,
+      "learning_rate": 1.4137453499872999e-06,
+      "loss": 1.0239,
+      "step": 7088
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7272460015542976,
+      "learning_rate": 1.4125323614975878e-06,
+      "loss": 0.9006,
+      "step": 7089
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6502405281089467,
+      "learning_rate": 1.4113198080006929e-06,
+      "loss": 0.8663,
+      "step": 7090
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6613112569467894,
+      "learning_rate": 1.4101076896436427e-06,
+      "loss": 0.9191,
+      "step": 7091
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.760510031326903,
+      "learning_rate": 1.4088960065734137e-06,
+      "loss": 0.8918,
+      "step": 7092
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8004943788997478,
+      "learning_rate": 1.4076847589369225e-06,
+      "loss": 0.9684,
+      "step": 7093
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7314136892005974,
+      "learning_rate": 1.406473946881039e-06,
+      "loss": 0.939,
+      "step": 7094
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.718912271091685,
+      "learning_rate": 1.4052635705525813e-06,
+      "loss": 0.8655,
+      "step": 7095
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7450747248848246,
+      "learning_rate": 1.4040536300983053e-06,
+      "loss": 0.9333,
+      "step": 7096
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6689581339057178,
+      "learning_rate": 1.4028441256649238e-06,
+      "loss": 0.8489,
+      "step": 7097
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7440651038903439,
+      "learning_rate": 1.401635057399095e-06,
+      "loss": 0.8751,
+      "step": 7098
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8000199991590979,
+      "learning_rate": 1.4004264254474193e-06,
+      "loss": 1.049,
+      "step": 7099
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7741082713580016,
+      "learning_rate": 1.3992182299564493e-06,
+      "loss": 0.8593,
+      "step": 7100
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7437981846703563,
+      "learning_rate": 1.3980104710726843e-06,
+      "loss": 0.852,
+      "step": 7101
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8463664216158018,
+      "learning_rate": 1.3968031489425659e-06,
+      "loss": 0.8804,
+      "step": 7102
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7346995920068333,
+      "learning_rate": 1.395596263712488e-06,
+      "loss": 0.8102,
+      "step": 7103
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8346774280396432,
+      "learning_rate": 1.3943898155287905e-06,
+      "loss": 0.8929,
+      "step": 7104
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.691891691998727,
+      "learning_rate": 1.3931838045377587e-06,
+      "loss": 0.8804,
+      "step": 7105
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7306681765222538,
+      "learning_rate": 1.3919782308856233e-06,
+      "loss": 0.9053,
+      "step": 7106
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7044878452262733,
+      "learning_rate": 1.3907730947185665e-06,
+      "loss": 0.8695,
+      "step": 7107
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6179002405648044,
+      "learning_rate": 1.3895683961827166e-06,
+      "loss": 0.9345,
+      "step": 7108
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7825899374075032,
+      "learning_rate": 1.3883641354241439e-06,
+      "loss": 0.905,
+      "step": 7109
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8071416559410125,
+      "learning_rate": 1.3871603125888705e-06,
+      "loss": 0.8613,
+      "step": 7110
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.818590373318333,
+      "learning_rate": 1.3859569278228668e-06,
+      "loss": 0.8936,
+      "step": 7111
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6924579673666171,
+      "learning_rate": 1.3847539812720434e-06,
+      "loss": 0.79,
+      "step": 7112
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7076327554381066,
+      "learning_rate": 1.3835514730822647e-06,
+      "loss": 0.7969,
+      "step": 7113
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7189194524465523,
+      "learning_rate": 1.3823494033993363e-06,
+      "loss": 0.8821,
+      "step": 7114
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.72156953665311,
+      "learning_rate": 1.3811477723690148e-06,
+      "loss": 0.8289,
+      "step": 7115
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7166403067589227,
+      "learning_rate": 1.3799465801370032e-06,
+      "loss": 0.8031,
+      "step": 7116
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6315823888977337,
+      "learning_rate": 1.378745826848949e-06,
+      "loss": 0.9167,
+      "step": 7117
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8498857251118632,
+      "learning_rate": 1.3775455126504466e-06,
+      "loss": 0.933,
+      "step": 7118
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8869209821413018,
+      "learning_rate": 1.3763456376870387e-06,
+      "loss": 0.9036,
+      "step": 7119
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7205644818815182,
+      "learning_rate": 1.3751462021042167e-06,
+      "loss": 0.8209,
+      "step": 7120
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.670123607955723,
+      "learning_rate": 1.373947206047413e-06,
+      "loss": 0.9136,
+      "step": 7121
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7509639984797516,
+      "learning_rate": 1.3727486496620113e-06,
+      "loss": 0.8942,
+      "step": 7122
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7321404207335946,
+      "learning_rate": 1.3715505330933426e-06,
+      "loss": 0.8957,
+      "step": 7123
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7066066536545388,
+      "learning_rate": 1.3703528564866792e-06,
+      "loss": 0.8687,
+      "step": 7124
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.67954504178895,
+      "learning_rate": 1.3691556199872453e-06,
+      "loss": 0.8247,
+      "step": 7125
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7874813747642309,
+      "learning_rate": 1.367958823740213e-06,
+      "loss": 0.8825,
+      "step": 7126
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.694268184208488,
+      "learning_rate": 1.366762467890692e-06,
+      "loss": 0.9211,
+      "step": 7127
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8146963494786443,
+      "learning_rate": 1.3655665525837475e-06,
+      "loss": 0.8677,
+      "step": 7128
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7000439038511104,
+      "learning_rate": 1.3643710779643892e-06,
+      "loss": 0.8259,
+      "step": 7129
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7527552411412317,
+      "learning_rate": 1.3631760441775704e-06,
+      "loss": 0.956,
+      "step": 7130
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7508120056334895,
+      "learning_rate": 1.3619814513681944e-06,
+      "loss": 0.8902,
+      "step": 7131
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7394477838486113,
+      "learning_rate": 1.3607872996811112e-06,
+      "loss": 0.9244,
+      "step": 7132
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6870177275022338,
+      "learning_rate": 1.3595935892611123e-06,
+      "loss": 0.8536,
+      "step": 7133
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7698462856319925,
+      "learning_rate": 1.3584003202529417e-06,
+      "loss": 1.0608,
+      "step": 7134
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7768129170651232,
+      "learning_rate": 1.3572074928012878e-06,
+      "loss": 0.8246,
+      "step": 7135
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.779485721514264,
+      "learning_rate": 1.3560151070507827e-06,
+      "loss": 0.9046,
+      "step": 7136
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7910171414549776,
+      "learning_rate": 1.3548231631460096e-06,
+      "loss": 1.037,
+      "step": 7137
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7725884225224293,
+      "learning_rate": 1.3536316612314937e-06,
+      "loss": 1.08,
+      "step": 7138
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.780118900891426,
+      "learning_rate": 1.3524406014517116e-06,
+      "loss": 0.9015,
+      "step": 7139
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7316668702437834,
+      "learning_rate": 1.3512499839510795e-06,
+      "loss": 0.9509,
+      "step": 7140
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.649442707837179,
+      "learning_rate": 1.3500598088739664e-06,
+      "loss": 0.8908,
+      "step": 7141
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6940994482813907,
+      "learning_rate": 1.3488700763646862e-06,
+      "loss": 0.87,
+      "step": 7142
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8098702775948423,
+      "learning_rate": 1.3476807865674951e-06,
+      "loss": 1.0477,
+      "step": 7143
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8070855001654088,
+      "learning_rate": 1.3464919396266018e-06,
+      "loss": 1.0366,
+      "step": 7144
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7590045322739305,
+      "learning_rate": 1.3453035356861544e-06,
+      "loss": 0.8908,
+      "step": 7145
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7726707818935807,
+      "learning_rate": 1.3441155748902534e-06,
+      "loss": 0.9473,
+      "step": 7146
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7054039848118565,
+      "learning_rate": 1.342928057382944e-06,
+      "loss": 0.8981,
+      "step": 7147
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7491444779381635,
+      "learning_rate": 1.3417409833082157e-06,
+      "loss": 0.9734,
+      "step": 7148
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7032986220714532,
+      "learning_rate": 1.340554352810003e-06,
+      "loss": 0.8498,
+      "step": 7149
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6138173839538594,
+      "learning_rate": 1.3393681660321917e-06,
+      "loss": 0.8324,
+      "step": 7150
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.787644550916954,
+      "learning_rate": 1.3381824231186113e-06,
+      "loss": 0.8859,
+      "step": 7151
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8044161855049358,
+      "learning_rate": 1.3369971242130354e-06,
+      "loss": 1.0687,
+      "step": 7152
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6363572998328166,
+      "learning_rate": 1.3358122694591862e-06,
+      "loss": 0.962,
+      "step": 7153
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7640402449574797,
+      "learning_rate": 1.3346278590007334e-06,
+      "loss": 0.8958,
+      "step": 7154
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7088826617174575,
+      "learning_rate": 1.333443892981287e-06,
+      "loss": 0.8884,
+      "step": 7155
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6808109325821585,
+      "learning_rate": 1.33226037154441e-06,
+      "loss": 0.9792,
+      "step": 7156
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7671699711691871,
+      "learning_rate": 1.3310772948336086e-06,
+      "loss": 1.0044,
+      "step": 7157
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7491708799811367,
+      "learning_rate": 1.3298946629923337e-06,
+      "loss": 0.8957,
+      "step": 7158
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6872110482674947,
+      "learning_rate": 1.3287124761639824e-06,
+      "loss": 0.916,
+      "step": 7159
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6939223139477135,
+      "learning_rate": 1.327530734491902e-06,
+      "loss": 0.8547,
+      "step": 7160
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7859580856893522,
+      "learning_rate": 1.3263494381193792e-06,
+      "loss": 0.9199,
+      "step": 7161
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6506799585137797,
+      "learning_rate": 1.3251685871896519e-06,
+      "loss": 0.8938,
+      "step": 7162
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7931945275530317,
+      "learning_rate": 1.3239881818459044e-06,
+      "loss": 0.963,
+      "step": 7163
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6349486341012596,
+      "learning_rate": 1.322808222231261e-06,
+      "loss": 0.9378,
+      "step": 7164
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7906801964229425,
+      "learning_rate": 1.3216287084887985e-06,
+      "loss": 1.0656,
+      "step": 7165
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8182989342485274,
+      "learning_rate": 1.3204496407615374e-06,
+      "loss": 0.9266,
+      "step": 7166
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6600695546441302,
+      "learning_rate": 1.3192710191924413e-06,
+      "loss": 0.8201,
+      "step": 7167
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7022127034849772,
+      "learning_rate": 1.3180928439244233e-06,
+      "loss": 0.9085,
+      "step": 7168
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7851094496087841,
+      "learning_rate": 1.3169151151003435e-06,
+      "loss": 1.0329,
+      "step": 7169
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7212717211180324,
+      "learning_rate": 1.3157378328630027e-06,
+      "loss": 0.8333,
+      "step": 7170
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7625920264468666,
+      "learning_rate": 1.3145609973551505e-06,
+      "loss": 0.83,
+      "step": 7171
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7236212450406387,
+      "learning_rate": 1.3133846087194823e-06,
+      "loss": 0.8883,
+      "step": 7172
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7708630358750201,
+      "learning_rate": 1.3122086670986422e-06,
+      "loss": 0.9148,
+      "step": 7173
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7447064767902556,
+      "learning_rate": 1.3110331726352132e-06,
+      "loss": 0.8909,
+      "step": 7174
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.946780965607682,
+      "learning_rate": 1.3098581254717313e-06,
+      "loss": 0.9584,
+      "step": 7175
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7723214645731253,
+      "learning_rate": 1.3086835257506719e-06,
+      "loss": 0.8672,
+      "step": 7176
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.700055720138507,
+      "learning_rate": 1.3075093736144612e-06,
+      "loss": 0.9927,
+      "step": 7177
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5799844376896903,
+      "learning_rate": 1.30633566920547e-06,
+      "loss": 0.8789,
+      "step": 7178
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.75596273495497,
+      "learning_rate": 1.3051624126660134e-06,
+      "loss": 0.952,
+      "step": 7179
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6916138863707135,
+      "learning_rate": 1.3039896041383505e-06,
+      "loss": 0.9306,
+      "step": 7180
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8000096446973306,
+      "learning_rate": 1.3028172437646901e-06,
+      "loss": 1.0416,
+      "step": 7181
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6553935119011225,
+      "learning_rate": 1.3016453316871868e-06,
+      "loss": 0.8004,
+      "step": 7182
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6611702843670078,
+      "learning_rate": 1.3004738680479357e-06,
+      "loss": 1.0027,
+      "step": 7183
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7117314630019973,
+      "learning_rate": 1.2993028529889817e-06,
+      "loss": 0.853,
+      "step": 7184
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.759996530845333,
+      "learning_rate": 1.2981322866523171e-06,
+      "loss": 0.9686,
+      "step": 7185
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8336636413304956,
+      "learning_rate": 1.2969621691798733e-06,
+      "loss": 0.8435,
+      "step": 7186
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7583826312310695,
+      "learning_rate": 1.295792500713533e-06,
+      "loss": 0.917,
+      "step": 7187
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7013166412794325,
+      "learning_rate": 1.2946232813951236e-06,
+      "loss": 0.8528,
+      "step": 7188
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7814776755600937,
+      "learning_rate": 1.2934545113664142e-06,
+      "loss": 1.0522,
+      "step": 7189
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6847704344625294,
+      "learning_rate": 1.2922861907691258e-06,
+      "loss": 0.8978,
+      "step": 7190
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6760121896866862,
+      "learning_rate": 1.2911183197449183e-06,
+      "loss": 0.8732,
+      "step": 7191
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7694806264855285,
+      "learning_rate": 1.2899508984354002e-06,
+      "loss": 0.8387,
+      "step": 7192
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6901784771049233,
+      "learning_rate": 1.2887839269821262e-06,
+      "loss": 0.8889,
+      "step": 7193
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7684566186355903,
+      "learning_rate": 1.2876174055265967e-06,
+      "loss": 0.9128,
+      "step": 7194
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7596035363775542,
+      "learning_rate": 1.2864513342102536e-06,
+      "loss": 0.9033,
+      "step": 7195
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5638625115698919,
+      "learning_rate": 1.285285713174489e-06,
+      "loss": 0.8551,
+      "step": 7196
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.708286391534676,
+      "learning_rate": 1.2841205425606396e-06,
+      "loss": 0.9848,
+      "step": 7197
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7394157424499348,
+      "learning_rate": 1.2829558225099836e-06,
+      "loss": 0.8693,
+      "step": 7198
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.818599205873659,
+      "learning_rate": 1.2817915531637492e-06,
+      "loss": 0.8766,
+      "step": 7199
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6209794548139278,
+      "learning_rate": 1.2806277346631085e-06,
+      "loss": 0.7363,
+      "step": 7200
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7814123130324737,
+      "learning_rate": 1.279464367149178e-06,
+      "loss": 1.0289,
+      "step": 7201
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8237910862608124,
+      "learning_rate": 1.2783014507630175e-06,
+      "loss": 0.7712,
+      "step": 7202
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7439312381746799,
+      "learning_rate": 1.2771389856456374e-06,
+      "loss": 0.9298,
+      "step": 7203
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7132786996767635,
+      "learning_rate": 1.275976971937991e-06,
+      "loss": 0.8762,
+      "step": 7204
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7062495412858127,
+      "learning_rate": 1.2748154097809745e-06,
+      "loss": 0.8416,
+      "step": 7205
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7478705638405603,
+      "learning_rate": 1.2736542993154316e-06,
+      "loss": 0.8572,
+      "step": 7206
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6966726612536829,
+      "learning_rate": 1.2724936406821536e-06,
+      "loss": 0.9788,
+      "step": 7207
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.768531197698784,
+      "learning_rate": 1.2713334340218703e-06,
+      "loss": 0.8916,
+      "step": 7208
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6339292907640433,
+      "learning_rate": 1.2701736794752645e-06,
+      "loss": 0.962,
+      "step": 7209
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7623883346738256,
+      "learning_rate": 1.2690143771829572e-06,
+      "loss": 0.8711,
+      "step": 7210
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7175489762755871,
+      "learning_rate": 1.267855527285521e-06,
+      "loss": 0.9507,
+      "step": 7211
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6672798221942142,
+      "learning_rate": 1.266697129923467e-06,
+      "loss": 0.8898,
+      "step": 7212
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.738222051448157,
+      "learning_rate": 1.2655391852372583e-06,
+      "loss": 0.9,
+      "step": 7213
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7478545988833027,
+      "learning_rate": 1.264381693367297e-06,
+      "loss": 0.8855,
+      "step": 7214
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.892509551969094,
+      "learning_rate": 1.2632246544539333e-06,
+      "loss": 0.9423,
+      "step": 7215
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6997650773381268,
+      "learning_rate": 1.2620680686374649e-06,
+      "loss": 0.7884,
+      "step": 7216
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7560487935186873,
+      "learning_rate": 1.2609119360581278e-06,
+      "loss": 0.8574,
+      "step": 7217
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8371574345628348,
+      "learning_rate": 1.2597562568561095e-06,
+      "loss": 0.9396,
+      "step": 7218
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8417726652325708,
+      "learning_rate": 1.2586010311715408e-06,
+      "loss": 0.8285,
+      "step": 7219
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.706859373676377,
+      "learning_rate": 1.257446259144494e-06,
+      "loss": 0.9043,
+      "step": 7220
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.662084758529687,
+      "learning_rate": 1.2562919409149916e-06,
+      "loss": 0.9641,
+      "step": 7221
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6935184152409044,
+      "learning_rate": 1.2551380766230003e-06,
+      "loss": 0.7848,
+      "step": 7222
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6638082644001209,
+      "learning_rate": 1.253984666408425e-06,
+      "loss": 0.9693,
+      "step": 7223
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.727243088227625,
+      "learning_rate": 1.2528317104111226e-06,
+      "loss": 0.8735,
+      "step": 7224
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8144125837509333,
+      "learning_rate": 1.2516792087708963e-06,
+      "loss": 0.8958,
+      "step": 7225
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6266378421540586,
+      "learning_rate": 1.2505271616274861e-06,
+      "loss": 0.9876,
+      "step": 7226
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8140259703830728,
+      "learning_rate": 1.2493755691205845e-06,
+      "loss": 0.8344,
+      "step": 7227
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6778111520543377,
+      "learning_rate": 1.2482244313898268e-06,
+      "loss": 0.8629,
+      "step": 7228
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7874628393428538,
+      "learning_rate": 1.247073748574789e-06,
+      "loss": 0.8016,
+      "step": 7229
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7162557196400536,
+      "learning_rate": 1.2459235208149984e-06,
+      "loss": 0.8904,
+      "step": 7230
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7612406123528435,
+      "learning_rate": 1.2447737482499245e-06,
+      "loss": 0.8543,
+      "step": 7231
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8234760461342195,
+      "learning_rate": 1.24362443101898e-06,
+      "loss": 1.0432,
+      "step": 7232
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6212559871747572,
+      "learning_rate": 1.2424755692615214e-06,
+      "loss": 0.8596,
+      "step": 7233
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8629887040681894,
+      "learning_rate": 1.2413271631168544e-06,
+      "loss": 0.9488,
+      "step": 7234
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.735056008313184,
+      "learning_rate": 1.2401792127242284e-06,
+      "loss": 0.8535,
+      "step": 7235
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7909709772165332,
+      "learning_rate": 1.2390317182228334e-06,
+      "loss": 0.9436,
+      "step": 7236
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7153653017902766,
+      "learning_rate": 1.2378846797518091e-06,
+      "loss": 0.978,
+      "step": 7237
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7488306602794066,
+      "learning_rate": 1.2367380974502386e-06,
+      "loss": 0.928,
+      "step": 7238
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8985655766062213,
+      "learning_rate": 1.2355919714571457e-06,
+      "loss": 0.8793,
+      "step": 7239
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7847978891670986,
+      "learning_rate": 1.234446301911506e-06,
+      "loss": 0.8729,
+      "step": 7240
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7058192334055207,
+      "learning_rate": 1.2333010889522328e-06,
+      "loss": 0.8262,
+      "step": 7241
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8251233560909994,
+      "learning_rate": 1.2321563327181885e-06,
+      "loss": 0.8885,
+      "step": 7242
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6889597477983211,
+      "learning_rate": 1.2310120333481795e-06,
+      "loss": 0.8388,
+      "step": 7243
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7743877453412082,
+      "learning_rate": 1.229868190980955e-06,
+      "loss": 0.8893,
+      "step": 7244
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9246085897594924,
+      "learning_rate": 1.2287248057552094e-06,
+      "loss": 0.8424,
+      "step": 7245
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6610038855476332,
+      "learning_rate": 1.2275818778095821e-06,
+      "loss": 0.8882,
+      "step": 7246
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6975806634638708,
+      "learning_rate": 1.226439407282659e-06,
+      "loss": 0.9154,
+      "step": 7247
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8990647885360525,
+      "learning_rate": 1.225297394312966e-06,
+      "loss": 0.9109,
+      "step": 7248
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.793848435127848,
+      "learning_rate": 1.224155839038977e-06,
+      "loss": 0.8424,
+      "step": 7249
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.617932472610186,
+      "learning_rate": 1.2230147415991117e-06,
+      "loss": 0.8161,
+      "step": 7250
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7080856336516712,
+      "learning_rate": 1.2218741021317282e-06,
+      "loss": 0.8861,
+      "step": 7251
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7051092833788617,
+      "learning_rate": 1.2207339207751356e-06,
+      "loss": 0.8134,
+      "step": 7252
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.850486419772358,
+      "learning_rate": 1.2195941976675869e-06,
+      "loss": 0.8992,
+      "step": 7253
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.795881537220781,
+      "learning_rate": 1.2184549329472717e-06,
+      "loss": 0.9301,
+      "step": 7254
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6230241615577008,
+      "learning_rate": 1.2173161267523332e-06,
+      "loss": 0.8836,
+      "step": 7255
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.676842288983878,
+      "learning_rate": 1.2161777792208563e-06,
+      "loss": 0.9149,
+      "step": 7256
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8035981244138628,
+      "learning_rate": 1.2150398904908673e-06,
+      "loss": 0.9763,
+      "step": 7257
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.715091564754427,
+      "learning_rate": 1.2139024607003401e-06,
+      "loss": 0.8615,
+      "step": 7258
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.620080625460703,
+      "learning_rate": 1.2127654899871937e-06,
+      "loss": 0.8719,
+      "step": 7259
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6924009158066093,
+      "learning_rate": 1.2116289784892872e-06,
+      "loss": 0.8704,
+      "step": 7260
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8601010629438974,
+      "learning_rate": 1.210492926344427e-06,
+      "loss": 0.8966,
+      "step": 7261
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8179378849969926,
+      "learning_rate": 1.2093573336903652e-06,
+      "loss": 1.0601,
+      "step": 7262
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8592943615146251,
+      "learning_rate": 1.2082222006647942e-06,
+      "loss": 0.9687,
+      "step": 7263
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7073166809431428,
+      "learning_rate": 1.207087527405355e-06,
+      "loss": 0.9387,
+      "step": 7264
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.729277453216935,
+      "learning_rate": 1.2059533140496277e-06,
+      "loss": 0.8445,
+      "step": 7265
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7487542353354095,
+      "learning_rate": 1.2048195607351421e-06,
+      "loss": 0.9179,
+      "step": 7266
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.819094323175927,
+      "learning_rate": 1.2036862675993678e-06,
+      "loss": 0.8479,
+      "step": 7267
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5843550845369212,
+      "learning_rate": 1.2025534347797214e-06,
+      "loss": 0.8669,
+      "step": 7268
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7613961938061535,
+      "learning_rate": 1.201421062413564e-06,
+      "loss": 0.9019,
+      "step": 7269
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8018674362679901,
+      "learning_rate": 1.2002891506381976e-06,
+      "loss": 1.0271,
+      "step": 7270
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.705159139983316,
+      "learning_rate": 1.199157699590872e-06,
+      "loss": 0.8579,
+      "step": 7271
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8121547798380657,
+      "learning_rate": 1.1980267094087777e-06,
+      "loss": 0.8958,
+      "step": 7272
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7082727224614864,
+      "learning_rate": 1.1968961802290524e-06,
+      "loss": 0.848,
+      "step": 7273
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8063511704123212,
+      "learning_rate": 1.1957661121887782e-06,
+      "loss": 0.8267,
+      "step": 7274
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.727891629099468,
+      "learning_rate": 1.1946365054249775e-06,
+      "loss": 0.8356,
+      "step": 7275
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8380143292176646,
+      "learning_rate": 1.1935073600746184e-06,
+      "loss": 1.0375,
+      "step": 7276
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7762505653514151,
+      "learning_rate": 1.1923786762746148e-06,
+      "loss": 0.8985,
+      "step": 7277
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8333118665826367,
+      "learning_rate": 1.1912504541618252e-06,
+      "loss": 0.8809,
+      "step": 7278
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7514522638972865,
+      "learning_rate": 1.1901226938730471e-06,
+      "loss": 0.9362,
+      "step": 7279
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5566338765683723,
+      "learning_rate": 1.1889953955450273e-06,
+      "loss": 0.9149,
+      "step": 7280
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7582517642218802,
+      "learning_rate": 1.1878685593144558e-06,
+      "loss": 0.9387,
+      "step": 7281
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.687905336359075,
+      "learning_rate": 1.1867421853179622e-06,
+      "loss": 0.9652,
+      "step": 7282
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6968170896379013,
+      "learning_rate": 1.185616273692125e-06,
+      "loss": 0.8653,
+      "step": 7283
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6951992931198046,
+      "learning_rate": 1.1844908245734659e-06,
+      "loss": 0.8973,
+      "step": 7284
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6989555865090609,
+      "learning_rate": 1.183365838098449e-06,
+      "loss": 0.9782,
+      "step": 7285
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6391295443282756,
+      "learning_rate": 1.1822413144034806e-06,
+      "loss": 0.8502,
+      "step": 7286
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.813290859458668,
+      "learning_rate": 1.181117253624916e-06,
+      "loss": 0.9087,
+      "step": 7287
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8033032078115374,
+      "learning_rate": 1.1799936558990483e-06,
+      "loss": 1.071,
+      "step": 7288
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7498922904527956,
+      "learning_rate": 1.1788705213621199e-06,
+      "loss": 0.9208,
+      "step": 7289
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6194658761366059,
+      "learning_rate": 1.1777478501503154e-06,
+      "loss": 0.9911,
+      "step": 7290
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6904160307365548,
+      "learning_rate": 1.17662564239976e-06,
+      "loss": 0.905,
+      "step": 7291
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7040061486075442,
+      "learning_rate": 1.1755038982465266e-06,
+      "loss": 0.767,
+      "step": 7292
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6634699534595698,
+      "learning_rate": 1.174382617826632e-06,
+      "loss": 0.8905,
+      "step": 7293
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7286886154560686,
+      "learning_rate": 1.1732618012760327e-06,
+      "loss": 0.8822,
+      "step": 7294
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8541035993595005,
+      "learning_rate": 1.1721414487306326e-06,
+      "loss": 0.8793,
+      "step": 7295
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6430324034485775,
+      "learning_rate": 1.1710215603262798e-06,
+      "loss": 0.9538,
+      "step": 7296
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8254952030840428,
+      "learning_rate": 1.1699021361987634e-06,
+      "loss": 1.0378,
+      "step": 7297
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.764530007115625,
+      "learning_rate": 1.1687831764838158e-06,
+      "loss": 0.8967,
+      "step": 7298
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8164304808463245,
+      "learning_rate": 1.1676646813171166e-06,
+      "loss": 0.8823,
+      "step": 7299
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7690369758376498,
+      "learning_rate": 1.1665466508342876e-06,
+      "loss": 0.8645,
+      "step": 7300
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7461822654313346,
+      "learning_rate": 1.1654290851708921e-06,
+      "loss": 0.901,
+      "step": 7301
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7227374696377924,
+      "learning_rate": 1.16431198446244e-06,
+      "loss": 0.9777,
+      "step": 7302
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7840131721652494,
+      "learning_rate": 1.1631953488443847e-06,
+      "loss": 1.0426,
+      "step": 7303
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7781163577538208,
+      "learning_rate": 1.1620791784521196e-06,
+      "loss": 0.8509,
+      "step": 7304
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6213727986572777,
+      "learning_rate": 1.1609634734209867e-06,
+      "loss": 0.9202,
+      "step": 7305
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.79768087742204,
+      "learning_rate": 1.1598482338862676e-06,
+      "loss": 0.973,
+      "step": 7306
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6748183431560135,
+      "learning_rate": 1.158733459983188e-06,
+      "loss": 0.8894,
+      "step": 7307
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.1918140233694134,
+      "learning_rate": 1.1576191518469193e-06,
+      "loss": 0.9032,
+      "step": 7308
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.683846178775958,
+      "learning_rate": 1.156505309612576e-06,
+      "loss": 0.8945,
+      "step": 7309
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7742832052351933,
+      "learning_rate": 1.1553919334152137e-06,
+      "loss": 1.0328,
+      "step": 7310
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7800821613076367,
+      "learning_rate": 1.1542790233898333e-06,
+      "loss": 1.053,
+      "step": 7311
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8131941173806192,
+      "learning_rate": 1.1531665796713814e-06,
+      "loss": 0.939,
+      "step": 7312
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7111820576782042,
+      "learning_rate": 1.1520546023947421e-06,
+      "loss": 0.9051,
+      "step": 7313
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7934914716172021,
+      "learning_rate": 1.1509430916947483e-06,
+      "loss": 0.9698,
+      "step": 7314
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7324047509609322,
+      "learning_rate": 1.149832047706176e-06,
+      "loss": 0.8245,
+      "step": 7315
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6513134170744688,
+      "learning_rate": 1.1487214705637395e-06,
+      "loss": 0.7149,
+      "step": 7316
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7894830175100631,
+      "learning_rate": 1.1476113604021039e-06,
+      "loss": 1.0215,
+      "step": 7317
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6655822276533985,
+      "learning_rate": 1.1465017173558717e-06,
+      "loss": 0.9106,
+      "step": 7318
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7008368810267882,
+      "learning_rate": 1.1453925415595902e-06,
+      "loss": 0.8867,
+      "step": 7319
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6427844004435166,
+      "learning_rate": 1.1442838331477512e-06,
+      "loss": 0.8575,
+      "step": 7320
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6872034371786135,
+      "learning_rate": 1.143175592254792e-06,
+      "loss": 0.9202,
+      "step": 7321
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5761945999185591,
+      "learning_rate": 1.142067819015088e-06,
+      "loss": 0.8351,
+      "step": 7322
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.68627768381419,
+      "learning_rate": 1.1409605135629603e-06,
+      "loss": 0.865,
+      "step": 7323
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6962375686957685,
+      "learning_rate": 1.1398536760326762e-06,
+      "loss": 0.8315,
+      "step": 7324
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6936647144432988,
+      "learning_rate": 1.1387473065584404e-06,
+      "loss": 0.8806,
+      "step": 7325
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6337509633018175,
+      "learning_rate": 1.1376414052744055e-06,
+      "loss": 0.8792,
+      "step": 7326
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7141847906504069,
+      "learning_rate": 1.1365359723146673e-06,
+      "loss": 0.8596,
+      "step": 7327
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7361731874652355,
+      "learning_rate": 1.1354310078132618e-06,
+      "loss": 0.8996,
+      "step": 7328
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8100217980055973,
+      "learning_rate": 1.1343265119041685e-06,
+      "loss": 0.9241,
+      "step": 7329
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7666386594077366,
+      "learning_rate": 1.1332224847213125e-06,
+      "loss": 0.8797,
+      "step": 7330
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6525452492622426,
+      "learning_rate": 1.1321189263985622e-06,
+      "loss": 0.8439,
+      "step": 7331
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.833807389219447,
+      "learning_rate": 1.1310158370697254e-06,
+      "loss": 0.9075,
+      "step": 7332
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.726151570350535,
+      "learning_rate": 1.1299132168685567e-06,
+      "loss": 0.927,
+      "step": 7333
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6779789018614633,
+      "learning_rate": 1.1288110659287544e-06,
+      "loss": 0.9366,
+      "step": 7334
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.634732437851536,
+      "learning_rate": 1.127709384383955e-06,
+      "loss": 0.8908,
+      "step": 7335
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6429823126696996,
+      "learning_rate": 1.1266081723677435e-06,
+      "loss": 0.8306,
+      "step": 7336
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7934801743396125,
+      "learning_rate": 1.1255074300136437e-06,
+      "loss": 1.0669,
+      "step": 7337
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7582488421103417,
+      "learning_rate": 1.124407157455127e-06,
+      "loss": 0.957,
+      "step": 7338
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.766126857111412,
+      "learning_rate": 1.1233073548256019e-06,
+      "loss": 0.8935,
+      "step": 7339
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7133015421767666,
+      "learning_rate": 1.1222080222584265e-06,
+      "loss": 0.866,
+      "step": 7340
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.788667575980634,
+      "learning_rate": 1.1211091598868956e-06,
+      "loss": 1.006,
+      "step": 7341
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7316929614193102,
+      "learning_rate": 1.1200107678442518e-06,
+      "loss": 0.9014,
+      "step": 7342
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8606435970548227,
+      "learning_rate": 1.11891284626368e-06,
+      "loss": 0.9248,
+      "step": 7343
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.759588217049405,
+      "learning_rate": 1.1178153952783045e-06,
+      "loss": 0.8756,
+      "step": 7344
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.761021070146567,
+      "learning_rate": 1.1167184150211962e-06,
+      "loss": 0.9426,
+      "step": 7345
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7840092602368611,
+      "learning_rate": 1.1156219056253692e-06,
+      "loss": 1.0585,
+      "step": 7346
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7221077545435126,
+      "learning_rate": 1.1145258672237764e-06,
+      "loss": 0.9068,
+      "step": 7347
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7426769487424307,
+      "learning_rate": 1.1134302999493174e-06,
+      "loss": 0.8833,
+      "step": 7348
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.049281492564361,
+      "learning_rate": 1.1123352039348362e-06,
+      "loss": 0.9628,
+      "step": 7349
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7068670059246993,
+      "learning_rate": 1.1112405793131114e-06,
+      "loss": 0.8925,
+      "step": 7350
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7224558564592949,
+      "learning_rate": 1.1101464262168733e-06,
+      "loss": 0.9303,
+      "step": 7351
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7327591815267886,
+      "learning_rate": 1.1090527447787924e-06,
+      "loss": 0.9608,
+      "step": 7352
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8690034051988902,
+      "learning_rate": 1.1079595351314793e-06,
+      "loss": 0.9057,
+      "step": 7353
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6892630725582283,
+      "learning_rate": 1.1068667974074903e-06,
+      "loss": 0.8603,
+      "step": 7354
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6824083324658907,
+      "learning_rate": 1.1057745317393254e-06,
+      "loss": 0.7772,
+      "step": 7355
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7352508705743914,
+      "learning_rate": 1.1046827382594227e-06,
+      "loss": 0.8726,
+      "step": 7356
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7299157342692528,
+      "learning_rate": 1.1035914171001665e-06,
+      "loss": 0.9503,
+      "step": 7357
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9452884007979727,
+      "learning_rate": 1.1025005683938862e-06,
+      "loss": 0.8861,
+      "step": 7358
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.760253546847027,
+      "learning_rate": 1.1014101922728482e-06,
+      "loss": 0.8722,
+      "step": 7359
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7855408769882823,
+      "learning_rate": 1.1003202888692633e-06,
+      "loss": 0.9674,
+      "step": 7360
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8386437135271612,
+      "learning_rate": 1.0992308583152878e-06,
+      "loss": 1.055,
+      "step": 7361
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7650631422308378,
+      "learning_rate": 1.09814190074302e-06,
+      "loss": 0.8728,
+      "step": 7362
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7471761334865932,
+      "learning_rate": 1.0970534162844977e-06,
+      "loss": 0.8929,
+      "step": 7363
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7316805681300618,
+      "learning_rate": 1.0959654050717034e-06,
+      "loss": 0.8333,
+      "step": 7364
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.762640180827757,
+      "learning_rate": 1.0948778672365646e-06,
+      "loss": 0.9034,
+      "step": 7365
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6079687970532943,
+      "learning_rate": 1.0937908029109461e-06,
+      "loss": 0.8815,
+      "step": 7366
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7137084160566847,
+      "learning_rate": 1.0927042122266613e-06,
+      "loss": 0.926,
+      "step": 7367
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8183519820313863,
+      "learning_rate": 1.0916180953154592e-06,
+      "loss": 0.8856,
+      "step": 7368
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7871112391371549,
+      "learning_rate": 1.0905324523090376e-06,
+      "loss": 0.8018,
+      "step": 7369
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6429311404803781,
+      "learning_rate": 1.0894472833390357e-06,
+      "loss": 0.9549,
+      "step": 7370
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6899767084562851,
+      "learning_rate": 1.0883625885370319e-06,
+      "loss": 0.7481,
+      "step": 7371
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7510208597576167,
+      "learning_rate": 1.0872783680345488e-06,
+      "loss": 0.9352,
+      "step": 7372
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8158205445576651,
+      "learning_rate": 1.0861946219630527e-06,
+      "loss": 1.0545,
+      "step": 7373
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.679353629155164,
+      "learning_rate": 1.0851113504539528e-06,
+      "loss": 0.8479,
+      "step": 7374
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7264411555008512,
+      "learning_rate": 1.084028553638597e-06,
+      "loss": 0.8286,
+      "step": 7375
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7223374015031898,
+      "learning_rate": 1.0829462316482798e-06,
+      "loss": 0.8928,
+      "step": 7376
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7830326446248905,
+      "learning_rate": 1.0818643846142373e-06,
+      "loss": 1.0336,
+      "step": 7377
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6518626850335538,
+      "learning_rate": 1.0807830126676444e-06,
+      "loss": 0.882,
+      "step": 7378
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6861574259416512,
+      "learning_rate": 1.0797021159396231e-06,
+      "loss": 0.8796,
+      "step": 7379
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.746967941865538,
+      "learning_rate": 1.0786216945612371e-06,
+      "loss": 0.8542,
+      "step": 7380
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.717896304240222,
+      "learning_rate": 1.0775417486634893e-06,
+      "loss": 0.8743,
+      "step": 7381
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7665092728621261,
+      "learning_rate": 1.076462278377326e-06,
+      "loss": 0.9116,
+      "step": 7382
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8316118946700117,
+      "learning_rate": 1.0753832838336398e-06,
+      "loss": 0.8811,
+      "step": 7383
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7455592113064085,
+      "learning_rate": 1.0743047651632587e-06,
+      "loss": 0.8637,
+      "step": 7384
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.677612755007564,
+      "learning_rate": 1.073226722496959e-06,
+      "loss": 0.8822,
+      "step": 7385
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7643441701564833,
+      "learning_rate": 1.0721491559654579e-06,
+      "loss": 0.8596,
+      "step": 7386
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8068590676367027,
+      "learning_rate": 1.0710720656994116e-06,
+      "loss": 0.9888,
+      "step": 7387
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6973538948144187,
+      "learning_rate": 1.0699954518294225e-06,
+      "loss": 0.8614,
+      "step": 7388
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7135942906969515,
+      "learning_rate": 1.0689193144860356e-06,
+      "loss": 0.8789,
+      "step": 7389
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.850454387646111,
+      "learning_rate": 1.0678436537997321e-06,
+      "loss": 0.9111,
+      "step": 7390
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7407976014197732,
+      "learning_rate": 1.066768469900944e-06,
+      "loss": 0.8601,
+      "step": 7391
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6982264947338424,
+      "learning_rate": 1.065693762920037e-06,
+      "loss": 0.9358,
+      "step": 7392
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8075752207421827,
+      "learning_rate": 1.064619532987326e-06,
+      "loss": 0.9091,
+      "step": 7393
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7552954678345043,
+      "learning_rate": 1.063545780233063e-06,
+      "loss": 0.9515,
+      "step": 7394
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6965134406390194,
+      "learning_rate": 1.0624725047874452e-06,
+      "loss": 0.8763,
+      "step": 7395
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6647411432965047,
+      "learning_rate": 1.061399706780612e-06,
+      "loss": 0.8686,
+      "step": 7396
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8021943512968018,
+      "learning_rate": 1.0603273863426412e-06,
+      "loss": 1.0537,
+      "step": 7397
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6183031982979021,
+      "learning_rate": 1.0592555436035573e-06,
+      "loss": 0.9117,
+      "step": 7398
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8194397155092208,
+      "learning_rate": 1.0581841786933262e-06,
+      "loss": 0.8841,
+      "step": 7399
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7181061387553238,
+      "learning_rate": 1.0571132917418508e-06,
+      "loss": 0.9217,
+      "step": 7400
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7844767536680001,
+      "learning_rate": 1.0560428828789837e-06,
+      "loss": 1.0515,
+      "step": 7401
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.818362061063144,
+      "learning_rate": 1.0549729522345142e-06,
+      "loss": 0.9565,
+      "step": 7402
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8463317569559128,
+      "learning_rate": 1.0539034999381731e-06,
+      "loss": 0.9145,
+      "step": 7403
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8020787141701602,
+      "learning_rate": 1.052834526119637e-06,
+      "loss": 0.8904,
+      "step": 7404
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7138913299900018,
+      "learning_rate": 1.0517660309085242e-06,
+      "loss": 0.8639,
+      "step": 7405
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7680815582276053,
+      "learning_rate": 1.0506980144343898e-06,
+      "loss": 1.063,
+      "step": 7406
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7405837125287273,
+      "learning_rate": 1.0496304768267373e-06,
+      "loss": 0.833,
+      "step": 7407
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.86446043659506,
+      "learning_rate": 1.048563418215009e-06,
+      "loss": 0.8735,
+      "step": 7408
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7140735494978634,
+      "learning_rate": 1.0474968387285884e-06,
+      "loss": 0.9337,
+      "step": 7409
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.769117581804003,
+      "learning_rate": 1.046430738496802e-06,
+      "loss": 1.0563,
+      "step": 7410
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8336734305366362,
+      "learning_rate": 1.04536511764892e-06,
+      "loss": 1.0481,
+      "step": 7411
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7078514003734424,
+      "learning_rate": 1.044299976314151e-06,
+      "loss": 0.8969,
+      "step": 7412
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7906107510259344,
+      "learning_rate": 1.0432353146216455e-06,
+      "loss": 1.0405,
+      "step": 7413
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8252322774369518,
+      "learning_rate": 1.0421711327005014e-06,
+      "loss": 1.0166,
+      "step": 7414
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8049787129305694,
+      "learning_rate": 1.0411074306797502e-06,
+      "loss": 0.9323,
+      "step": 7415
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.737505306210331,
+      "learning_rate": 1.0400442086883717e-06,
+      "loss": 0.845,
+      "step": 7416
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7177254183456578,
+      "learning_rate": 1.038981466855286e-06,
+      "loss": 0.9109,
+      "step": 7417
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6679515561667353,
+      "learning_rate": 1.0379192053093523e-06,
+      "loss": 0.9393,
+      "step": 7418
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6638152849304269,
+      "learning_rate": 1.036857424179374e-06,
+      "loss": 0.916,
+      "step": 7419
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.766203513185313,
+      "learning_rate": 1.0357961235940978e-06,
+      "loss": 0.8481,
+      "step": 7420
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6932190004303862,
+      "learning_rate": 1.0347353036822061e-06,
+      "loss": 0.9329,
+      "step": 7421
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6589993131162466,
+      "learning_rate": 1.0336749645723298e-06,
+      "loss": 0.8272,
+      "step": 7422
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8156555092119184,
+      "learning_rate": 1.0326151063930396e-06,
+      "loss": 0.796,
+      "step": 7423
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6786141058746926,
+      "learning_rate": 1.031555729272845e-06,
+      "loss": 0.9064,
+      "step": 7424
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.716384580083553,
+      "learning_rate": 1.0304968333401983e-06,
+      "loss": 0.8568,
+      "step": 7425
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9242271727566926,
+      "learning_rate": 1.0294384187234952e-06,
+      "loss": 0.8794,
+      "step": 7426
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7847599215961499,
+      "learning_rate": 1.0283804855510744e-06,
+      "loss": 1.0286,
+      "step": 7427
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.713969873637573,
+      "learning_rate": 1.0273230339512103e-06,
+      "loss": 0.8943,
+      "step": 7428
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7959527533467823,
+      "learning_rate": 1.0262660640521245e-06,
+      "loss": 1.0106,
+      "step": 7429
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7005092406688935,
+      "learning_rate": 1.0252095759819785e-06,
+      "loss": 0.9045,
+      "step": 7430
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6279413463219636,
+      "learning_rate": 1.0241535698688742e-06,
+      "loss": 0.8724,
+      "step": 7431
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7374742975204573,
+      "learning_rate": 1.0230980458408574e-06,
+      "loss": 0.8693,
+      "step": 7432
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6512880788237159,
+      "learning_rate": 1.0220430040259116e-06,
+      "loss": 0.8845,
+      "step": 7433
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6915629157491914,
+      "learning_rate": 1.020988444551967e-06,
+      "loss": 0.8173,
+      "step": 7434
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7551003206098652,
+      "learning_rate": 1.0199343675468898e-06,
+      "loss": 0.9321,
+      "step": 7435
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.777347136653578,
+      "learning_rate": 1.0188807731384932e-06,
+      "loss": 0.9098,
+      "step": 7436
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6950701417138496,
+      "learning_rate": 1.0178276614545269e-06,
+      "loss": 0.8939,
+      "step": 7437
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.751724881127983,
+      "learning_rate": 1.016775032622685e-06,
+      "loss": 0.9137,
+      "step": 7438
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7561071905334344,
+      "learning_rate": 1.0157228867706042e-06,
+      "loss": 0.9124,
+      "step": 7439
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6359660680458328,
+      "learning_rate": 1.0146712240258578e-06,
+      "loss": 0.8359,
+      "step": 7440
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6472994814005986,
+      "learning_rate": 1.013620044515965e-06,
+      "loss": 0.8458,
+      "step": 7441
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6671174211023274,
+      "learning_rate": 1.0125693483683863e-06,
+      "loss": 0.8299,
+      "step": 7442
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8525611424949733,
+      "learning_rate": 1.0115191357105192e-06,
+      "loss": 0.854,
+      "step": 7443
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7592724136367586,
+      "learning_rate": 1.010469406669709e-06,
+      "loss": 0.8737,
+      "step": 7444
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8775773936378102,
+      "learning_rate": 1.0094201613732374e-06,
+      "loss": 0.8721,
+      "step": 7445
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8082533765831228,
+      "learning_rate": 1.008371399948327e-06,
+      "loss": 0.9318,
+      "step": 7446
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7821285921448549,
+      "learning_rate": 1.0073231225221457e-06,
+      "loss": 0.8947,
+      "step": 7447
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6384583744927261,
+      "learning_rate": 1.0062753292218025e-06,
+      "loss": 0.8721,
+      "step": 7448
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6549292838909113,
+      "learning_rate": 1.0052280201743426e-06,
+      "loss": 0.9225,
+      "step": 7449
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7209440829804084,
+      "learning_rate": 1.004181195506757e-06,
+      "loss": 0.8835,
+      "step": 7450
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7810106313508058,
+      "learning_rate": 1.0031348553459785e-06,
+      "loss": 0.8917,
+      "step": 7451
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7987806226766205,
+      "learning_rate": 1.002088999818877e-06,
+      "loss": 1.0493,
+      "step": 7452
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.777629513077754,
+      "learning_rate": 1.0010436290522674e-06,
+      "loss": 0.9167,
+      "step": 7453
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6421252603242091,
+      "learning_rate": 9.999987431729052e-07,
+      "loss": 0.8712,
+      "step": 7454
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6940357000473558,
+      "learning_rate": 9.989543423074855e-07,
+      "loss": 0.8568,
+      "step": 7455
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.619756720744302,
+      "learning_rate": 9.979104265826438e-07,
+      "loss": 0.9345,
+      "step": 7456
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6931185788219418,
+      "learning_rate": 9.968669961249611e-07,
+      "loss": 0.8134,
+      "step": 7457
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.697006850818595,
+      "learning_rate": 9.95824051060957e-07,
+      "loss": 0.8873,
+      "step": 7458
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7123681122975698,
+      "learning_rate": 9.947815915170895e-07,
+      "loss": 0.856,
+      "step": 7459
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.709630604862187,
+      "learning_rate": 9.937396176197623e-07,
+      "loss": 0.8706,
+      "step": 7460
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8597341142274448,
+      "learning_rate": 9.926981294953192e-07,
+      "loss": 0.9058,
+      "step": 7461
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.693761452785304,
+      "learning_rate": 9.916571272700421e-07,
+      "loss": 0.9142,
+      "step": 7462
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6494704990828555,
+      "learning_rate": 9.906166110701588e-07,
+      "loss": 0.8724,
+      "step": 7463
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.750118477293998,
+      "learning_rate": 9.895765810218322e-07,
+      "loss": 0.8256,
+      "step": 7464
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.035241925263039,
+      "learning_rate": 9.885370372511727e-07,
+      "loss": 0.8364,
+      "step": 7465
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5941246686554178,
+      "learning_rate": 9.874979798842254e-07,
+      "loss": 0.846,
+      "step": 7466
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7307034688370917,
+      "learning_rate": 9.86459409046983e-07,
+      "loss": 0.9948,
+      "step": 7467
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6733351971333528,
+      "learning_rate": 9.85421324865372e-07,
+      "loss": 0.9172,
+      "step": 7468
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6672178382938263,
+      "learning_rate": 9.843837274652667e-07,
+      "loss": 0.8986,
+      "step": 7469
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.755942214009784,
+      "learning_rate": 9.833466169724792e-07,
+      "loss": 0.9332,
+      "step": 7470
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.778365900496306,
+      "learning_rate": 9.823099935127605e-07,
+      "loss": 0.9592,
+      "step": 7471
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7441445039321022,
+      "learning_rate": 9.812738572118063e-07,
+      "loss": 0.9224,
+      "step": 7472
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7421832119941743,
+      "learning_rate": 9.802382081952528e-07,
+      "loss": 0.9447,
+      "step": 7473
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7454891227623683,
+      "learning_rate": 9.792030465886736e-07,
+      "loss": 0.8702,
+      "step": 7474
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6384831352355456,
+      "learning_rate": 9.781683725175867e-07,
+      "loss": 0.9162,
+      "step": 7475
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7305653279973348,
+      "learning_rate": 9.771341861074523e-07,
+      "loss": 0.9643,
+      "step": 7476
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.757569008966683,
+      "learning_rate": 9.761004874836643e-07,
+      "loss": 0.9591,
+      "step": 7477
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8188570467466174,
+      "learning_rate": 9.750672767715651e-07,
+      "loss": 0.9423,
+      "step": 7478
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7850000952825742,
+      "learning_rate": 9.740345540964359e-07,
+      "loss": 0.9163,
+      "step": 7479
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6809155880382092,
+      "learning_rate": 9.730023195834948e-07,
+      "loss": 0.8204,
+      "step": 7480
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6069458568819561,
+      "learning_rate": 9.71970573357906e-07,
+      "loss": 0.9232,
+      "step": 7481
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.775462146831102,
+      "learning_rate": 9.709393155447734e-07,
+      "loss": 0.8609,
+      "step": 7482
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7943762815321713,
+      "learning_rate": 9.699085462691376e-07,
+      "loss": 0.8485,
+      "step": 7483
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6088010007993747,
+      "learning_rate": 9.688782656559842e-07,
+      "loss": 0.9491,
+      "step": 7484
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6491506624924286,
+      "learning_rate": 9.678484738302401e-07,
+      "loss": 0.8848,
+      "step": 7485
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7396473431974164,
+      "learning_rate": 9.66819170916769e-07,
+      "loss": 0.8493,
+      "step": 7486
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7575140822242683,
+      "learning_rate": 9.65790357040377e-07,
+      "loss": 0.9016,
+      "step": 7487
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8377030982020464,
+      "learning_rate": 9.647620323258123e-07,
+      "loss": 0.9055,
+      "step": 7488
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7688247978885039,
+      "learning_rate": 9.637341968977638e-07,
+      "loss": 1.0315,
+      "step": 7489
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6911437341489899,
+      "learning_rate": 9.62706850880858e-07,
+      "loss": 0.8699,
+      "step": 7490
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6089855880343522,
+      "learning_rate": 9.616799943996652e-07,
+      "loss": 0.9499,
+      "step": 7491
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.752606019649648,
+      "learning_rate": 9.606536275786965e-07,
+      "loss": 0.9248,
+      "step": 7492
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7145776107280355,
+      "learning_rate": 9.596277505423996e-07,
+      "loss": 0.9204,
+      "step": 7493
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7380905271503913,
+      "learning_rate": 9.586023634151675e-07,
+      "loss": 0.8699,
+      "step": 7494
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8850460520328105,
+      "learning_rate": 9.575774663213327e-07,
+      "loss": 0.9142,
+      "step": 7495
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6952259179932236,
+      "learning_rate": 9.565530593851657e-07,
+      "loss": 0.8621,
+      "step": 7496
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8650568778700318,
+      "learning_rate": 9.55529142730881e-07,
+      "loss": 0.865,
+      "step": 7497
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6995718713193269,
+      "learning_rate": 9.545057164826317e-07,
+      "loss": 0.872,
+      "step": 7498
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7737561687512202,
+      "learning_rate": 9.534827807645091e-07,
+      "loss": 0.8579,
+      "step": 7499
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.911708725321387,
+      "learning_rate": 9.524603357005502e-07,
+      "loss": 0.9102,
+      "step": 7500
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8176478444179738,
+      "learning_rate": 9.51438381414731e-07,
+      "loss": 1.0314,
+      "step": 7501
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7911021447058241,
+      "learning_rate": 9.504169180309641e-07,
+      "loss": 1.0475,
+      "step": 7502
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7801279015928657,
+      "learning_rate": 9.493959456731072e-07,
+      "loss": 0.8582,
+      "step": 7503
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8444219055110163,
+      "learning_rate": 9.483754644649573e-07,
+      "loss": 0.8537,
+      "step": 7504
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7777083848354767,
+      "learning_rate": 9.473554745302494e-07,
+      "loss": 0.8392,
+      "step": 7505
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7898371461719043,
+      "learning_rate": 9.463359759926616e-07,
+      "loss": 0.9005,
+      "step": 7506
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6989355218401136,
+      "learning_rate": 9.453169689758135e-07,
+      "loss": 0.8679,
+      "step": 7507
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7952566528706494,
+      "learning_rate": 9.442984536032612e-07,
+      "loss": 1.0325,
+      "step": 7508
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.727238212882518,
+      "learning_rate": 9.432804299985021e-07,
+      "loss": 0.9495,
+      "step": 7509
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7622294009983397,
+      "learning_rate": 9.422628982849786e-07,
+      "loss": 0.9228,
+      "step": 7510
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7350765076741295,
+      "learning_rate": 9.412458585860657e-07,
+      "loss": 0.8686,
+      "step": 7511
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9298925196996435,
+      "learning_rate": 9.402293110250854e-07,
+      "loss": 0.9279,
+      "step": 7512
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6259806623637474,
+      "learning_rate": 9.392132557252986e-07,
+      "loss": 0.8543,
+      "step": 7513
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7627126617419873,
+      "learning_rate": 9.381976928099029e-07,
+      "loss": 0.8303,
+      "step": 7514
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7279499528250697,
+      "learning_rate": 9.371826224020397e-07,
+      "loss": 0.9442,
+      "step": 7515
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8633324024823712,
+      "learning_rate": 9.361680446247923e-07,
+      "loss": 0.8522,
+      "step": 7516
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5828403511894253,
+      "learning_rate": 9.351539596011777e-07,
+      "loss": 0.9122,
+      "step": 7517
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7740955846138671,
+      "learning_rate": 9.341403674541605e-07,
+      "loss": 1.0264,
+      "step": 7518
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6644677108266288,
+      "learning_rate": 9.3312726830664e-07,
+      "loss": 0.8328,
+      "step": 7519
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6700162191088324,
+      "learning_rate": 9.321146622814597e-07,
+      "loss": 0.9813,
+      "step": 7520
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8486260036708304,
+      "learning_rate": 9.311025495013998e-07,
+      "loss": 0.877,
+      "step": 7521
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7762072785453806,
+      "learning_rate": 9.300909300891831e-07,
+      "loss": 0.9383,
+      "step": 7522
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7816728927581338,
+      "learning_rate": 9.290798041674737e-07,
+      "loss": 0.9247,
+      "step": 7523
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6263595234163108,
+      "learning_rate": 9.280691718588713e-07,
+      "loss": 0.9394,
+      "step": 7524
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7951857027251206,
+      "learning_rate": 9.270590332859202e-07,
+      "loss": 1.0641,
+      "step": 7525
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7693634543349892,
+      "learning_rate": 9.260493885711036e-07,
+      "loss": 0.8831,
+      "step": 7526
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8409739002620005,
+      "learning_rate": 9.250402378368428e-07,
+      "loss": 0.9508,
+      "step": 7527
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8049055558744254,
+      "learning_rate": 9.240315812055029e-07,
+      "loss": 0.9077,
+      "step": 7528
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7183665578837524,
+      "learning_rate": 9.230234187993859e-07,
+      "loss": 0.93,
+      "step": 7529
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7504152190428186,
+      "learning_rate": 9.220157507407335e-07,
+      "loss": 0.8819,
+      "step": 7530
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8363921517880848,
+      "learning_rate": 9.210085771517297e-07,
+      "loss": 0.8664,
+      "step": 7531
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6628223863276888,
+      "learning_rate": 9.200018981545001e-07,
+      "loss": 0.8601,
+      "step": 7532
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.556040270231309,
+      "learning_rate": 9.189957138711053e-07,
+      "loss": 0.8952,
+      "step": 7533
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7106669614245364,
+      "learning_rate": 9.17990024423549e-07,
+      "loss": 0.9248,
+      "step": 7534
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6069223506576378,
+      "learning_rate": 9.169848299337763e-07,
+      "loss": 0.8808,
+      "step": 7535
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8078842526553374,
+      "learning_rate": 9.159801305236687e-07,
+      "loss": 0.8823,
+      "step": 7536
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6980569363773421,
+      "learning_rate": 9.149759263150493e-07,
+      "loss": 0.792,
+      "step": 7537
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7711587688808093,
+      "learning_rate": 9.139722174296839e-07,
+      "loss": 0.9041,
+      "step": 7538
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7428206973434905,
+      "learning_rate": 9.129690039892736e-07,
+      "loss": 0.9107,
+      "step": 7539
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7996695531159022,
+      "learning_rate": 9.119662861154599e-07,
+      "loss": 0.9042,
+      "step": 7540
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.689083312126614,
+      "learning_rate": 9.109640639298295e-07,
+      "loss": 0.9111,
+      "step": 7541
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7631554352642633,
+      "learning_rate": 9.099623375539019e-07,
+      "loss": 0.9297,
+      "step": 7542
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7712852959891103,
+      "learning_rate": 9.089611071091414e-07,
+      "loss": 0.9377,
+      "step": 7543
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6318152716621532,
+      "learning_rate": 9.079603727169523e-07,
+      "loss": 0.8848,
+      "step": 7544
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.680134057195605,
+      "learning_rate": 9.069601344986734e-07,
+      "loss": 0.9136,
+      "step": 7545
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8089847754045532,
+      "learning_rate": 9.059603925755895e-07,
+      "loss": 0.9491,
+      "step": 7546
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6743156099227399,
+      "learning_rate": 9.049611470689235e-07,
+      "loss": 0.9003,
+      "step": 7547
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.733316341571945,
+      "learning_rate": 9.039623980998347e-07,
+      "loss": 0.9003,
+      "step": 7548
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6374182962853234,
+      "learning_rate": 9.029641457894267e-07,
+      "loss": 0.8682,
+      "step": 7549
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8365043625738997,
+      "learning_rate": 9.019663902587416e-07,
+      "loss": 0.887,
+      "step": 7550
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9261231977697735,
+      "learning_rate": 9.009691316287599e-07,
+      "loss": 0.9191,
+      "step": 7551
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7966914458141496,
+      "learning_rate": 8.999723700204011e-07,
+      "loss": 0.837,
+      "step": 7552
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.812251504940816,
+      "learning_rate": 8.989761055545276e-07,
+      "loss": 0.9293,
+      "step": 7553
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8338486516434905,
+      "learning_rate": 8.979803383519409e-07,
+      "loss": 1.0231,
+      "step": 7554
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7248498510981238,
+      "learning_rate": 8.969850685333786e-07,
+      "loss": 0.8824,
+      "step": 7555
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5873929539741345,
+      "learning_rate": 8.959902962195222e-07,
+      "loss": 0.8746,
+      "step": 7556
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6070394778845973,
+      "learning_rate": 8.949960215309922e-07,
+      "loss": 0.8782,
+      "step": 7557
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6904616105626713,
+      "learning_rate": 8.940022445883451e-07,
+      "loss": 0.8486,
+      "step": 7558
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6566832365377089,
+      "learning_rate": 8.930089655120833e-07,
+      "loss": 0.903,
+      "step": 7559
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6577357780400859,
+      "learning_rate": 8.920161844226416e-07,
+      "loss": 0.8257,
+      "step": 7560
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7689955987780956,
+      "learning_rate": 8.910239014404016e-07,
+      "loss": 0.8464,
+      "step": 7561
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.665637878029967,
+      "learning_rate": 8.900321166856774e-07,
+      "loss": 0.8585,
+      "step": 7562
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8561626172669419,
+      "learning_rate": 8.890408302787301e-07,
+      "loss": 0.8627,
+      "step": 7563
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7327119902903307,
+      "learning_rate": 8.880500423397531e-07,
+      "loss": 0.9095,
+      "step": 7564
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7095745206556874,
+      "learning_rate": 8.870597529888847e-07,
+      "loss": 0.8933,
+      "step": 7565
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6382550813919294,
+      "learning_rate": 8.860699623462021e-07,
+      "loss": 0.9012,
+      "step": 7566
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.782585531442091,
+      "learning_rate": 8.850806705317183e-07,
+      "loss": 0.9337,
+      "step": 7567
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7589552283248235,
+      "learning_rate": 8.84091877665389e-07,
+      "loss": 0.8269,
+      "step": 7568
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.703979063988604,
+      "learning_rate": 8.83103583867111e-07,
+      "loss": 0.9244,
+      "step": 7569
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7305442823056714,
+      "learning_rate": 8.82115789256715e-07,
+      "loss": 0.896,
+      "step": 7570
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6470216085790148,
+      "learning_rate": 8.811284939539771e-07,
+      "loss": 0.8724,
+      "step": 7571
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6696528080458362,
+      "learning_rate": 8.801416980786099e-07,
+      "loss": 0.8956,
+      "step": 7572
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6175418608078778,
+      "learning_rate": 8.791554017502634e-07,
+      "loss": 0.9221,
+      "step": 7573
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7315639921370438,
+      "learning_rate": 8.781696050885313e-07,
+      "loss": 0.9266,
+      "step": 7574
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6739014901228928,
+      "learning_rate": 8.771843082129455e-07,
+      "loss": 0.861,
+      "step": 7575
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7234351367320297,
+      "learning_rate": 8.76199511242975e-07,
+      "loss": 0.9135,
+      "step": 7576
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7239969873869712,
+      "learning_rate": 8.752152142980308e-07,
+      "loss": 0.9117,
+      "step": 7577
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6260054618043283,
+      "learning_rate": 8.74231417497463e-07,
+      "loss": 0.869,
+      "step": 7578
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8111370010175276,
+      "learning_rate": 8.732481209605587e-07,
+      "loss": 0.8779,
+      "step": 7579
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8087168097332318,
+      "learning_rate": 8.722653248065466e-07,
+      "loss": 0.8277,
+      "step": 7580
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7791708064155636,
+      "learning_rate": 8.712830291545954e-07,
+      "loss": 0.8666,
+      "step": 7581
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6474602914598075,
+      "learning_rate": 8.70301234123811e-07,
+      "loss": 0.9119,
+      "step": 7582
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7460997560935734,
+      "learning_rate": 8.693199398332381e-07,
+      "loss": 0.9421,
+      "step": 7583
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7966163313855352,
+      "learning_rate": 8.683391464018625e-07,
+      "loss": 0.864,
+      "step": 7584
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7671606641699713,
+      "learning_rate": 8.673588539486117e-07,
+      "loss": 0.9212,
+      "step": 7585
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6748575705175561,
+      "learning_rate": 8.663790625923451e-07,
+      "loss": 0.8446,
+      "step": 7586
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7339476893560808,
+      "learning_rate": 8.653997724518682e-07,
+      "loss": 0.8464,
+      "step": 7587
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6716303170084572,
+      "learning_rate": 8.644209836459244e-07,
+      "loss": 0.8786,
+      "step": 7588
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.728454058044721,
+      "learning_rate": 8.634426962931925e-07,
+      "loss": 0.8918,
+      "step": 7589
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.648071284751071,
+      "learning_rate": 8.62464910512294e-07,
+      "loss": 0.8854,
+      "step": 7590
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7520429261716601,
+      "learning_rate": 8.614876264217914e-07,
+      "loss": 0.8482,
+      "step": 7591
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.810777248898115,
+      "learning_rate": 8.60510844140181e-07,
+      "loss": 0.9323,
+      "step": 7592
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7210160077577847,
+      "learning_rate": 8.59534563785901e-07,
+      "loss": 0.9929,
+      "step": 7593
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7303321965522753,
+      "learning_rate": 8.585587854773308e-07,
+      "loss": 0.933,
+      "step": 7594
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.792656412644984,
+      "learning_rate": 8.575835093327845e-07,
+      "loss": 1.05,
+      "step": 7595
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7949874893749755,
+      "learning_rate": 8.566087354705182e-07,
+      "loss": 0.9806,
+      "step": 7596
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8993914634868043,
+      "learning_rate": 8.556344640087289e-07,
+      "loss": 0.8789,
+      "step": 7597
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7297552850615328,
+      "learning_rate": 8.546606950655467e-07,
+      "loss": 0.8528,
+      "step": 7598
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7924984276988014,
+      "learning_rate": 8.536874287590469e-07,
+      "loss": 0.8846,
+      "step": 7599
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.202840029078937,
+      "learning_rate": 8.527146652072421e-07,
+      "loss": 0.8735,
+      "step": 7600
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7175521049308382,
+      "learning_rate": 8.517424045280809e-07,
+      "loss": 0.8858,
+      "step": 7601
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5930632590041474,
+      "learning_rate": 8.507706468394544e-07,
+      "loss": 0.8922,
+      "step": 7602
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7593662082744626,
+      "learning_rate": 8.497993922591935e-07,
+      "loss": 0.892,
+      "step": 7603
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6454711983992356,
+      "learning_rate": 8.488286409050623e-07,
+      "loss": 0.8895,
+      "step": 7604
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.722399914857841,
+      "learning_rate": 8.478583928947692e-07,
+      "loss": 0.8724,
+      "step": 7605
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5900776065208047,
+      "learning_rate": 8.46888648345962e-07,
+      "loss": 0.9121,
+      "step": 7606
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6671855867245031,
+      "learning_rate": 8.459194073762228e-07,
+      "loss": 0.9002,
+      "step": 7607
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.586221036620047,
+      "learning_rate": 8.449506701030774e-07,
+      "loss": 0.9057,
+      "step": 7608
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.747596861845314,
+      "learning_rate": 8.439824366439886e-07,
+      "loss": 0.846,
+      "step": 7609
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8539056989636502,
+      "learning_rate": 8.430147071163558e-07,
+      "loss": 0.7976,
+      "step": 7610
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6543679197811214,
+      "learning_rate": 8.420474816375213e-07,
+      "loss": 0.988,
+      "step": 7611
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.580812976734974,
+      "learning_rate": 8.410807603247656e-07,
+      "loss": 0.8736,
+      "step": 7612
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7004682790652799,
+      "learning_rate": 8.401145432953045e-07,
+      "loss": 0.8903,
+      "step": 7613
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.75099238841902,
+      "learning_rate": 8.391488306662976e-07,
+      "loss": 0.9469,
+      "step": 7614
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.766864429740249,
+      "learning_rate": 8.381836225548384e-07,
+      "loss": 0.8404,
+      "step": 7615
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9115772791078784,
+      "learning_rate": 8.372189190779639e-07,
+      "loss": 0.8333,
+      "step": 7616
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7787133087559813,
+      "learning_rate": 8.362547203526456e-07,
+      "loss": 1.0232,
+      "step": 7617
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.727582878495329,
+      "learning_rate": 8.35291026495797e-07,
+      "loss": 0.8673,
+      "step": 7618
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6727302304855556,
+      "learning_rate": 8.343278376242709e-07,
+      "loss": 0.9213,
+      "step": 7619
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6540936053112212,
+      "learning_rate": 8.333651538548543e-07,
+      "loss": 0.8662,
+      "step": 7620
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7987174808280918,
+      "learning_rate": 8.324029753042773e-07,
+      "loss": 1.0317,
+      "step": 7621
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7367812444208781,
+      "learning_rate": 8.314413020892087e-07,
+      "loss": 0.9754,
+      "step": 7622
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7773400860108092,
+      "learning_rate": 8.304801343262525e-07,
+      "loss": 1.0343,
+      "step": 7623
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6779815990466356,
+      "learning_rate": 8.295194721319555e-07,
+      "loss": 0.8141,
+      "step": 7624
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8139981294878667,
+      "learning_rate": 8.285593156228006e-07,
+      "loss": 0.915,
+      "step": 7625
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.727122596020832,
+      "learning_rate": 8.275996649152085e-07,
+      "loss": 0.9542,
+      "step": 7626
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7495634345084157,
+      "learning_rate": 8.266405201255412e-07,
+      "loss": 0.8811,
+      "step": 7627
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.183515539138076,
+      "learning_rate": 8.256818813701e-07,
+      "loss": 0.9013,
+      "step": 7628
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6403317948729157,
+      "learning_rate": 8.247237487651205e-07,
+      "loss": 0.8305,
+      "step": 7629
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.660066506854636,
+      "learning_rate": 8.237661224267807e-07,
+      "loss": 0.9092,
+      "step": 7630
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7173635046228868,
+      "learning_rate": 8.228090024711977e-07,
+      "loss": 0.9305,
+      "step": 7631
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6790732332000802,
+      "learning_rate": 8.218523890144225e-07,
+      "loss": 0.8504,
+      "step": 7632
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7899474937425524,
+      "learning_rate": 8.208962821724498e-07,
+      "loss": 0.9235,
+      "step": 7633
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7258139307383407,
+      "learning_rate": 8.199406820612115e-07,
+      "loss": 0.8799,
+      "step": 7634
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6482010493468788,
+      "learning_rate": 8.189855887965764e-07,
+      "loss": 0.8391,
+      "step": 7635
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.704140883915139,
+      "learning_rate": 8.180310024943517e-07,
+      "loss": 0.8237,
+      "step": 7636
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.971538319681906,
+      "learning_rate": 8.170769232702869e-07,
+      "loss": 0.8839,
+      "step": 7637
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6987021177148245,
+      "learning_rate": 8.161233512400641e-07,
+      "loss": 0.9097,
+      "step": 7638
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.907528445529689,
+      "learning_rate": 8.15170286519309e-07,
+      "loss": 0.9025,
+      "step": 7639
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6101558918348347,
+      "learning_rate": 8.142177292235854e-07,
+      "loss": 0.8676,
+      "step": 7640
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6558606851998914,
+      "learning_rate": 8.132656794683914e-07,
+      "loss": 0.9027,
+      "step": 7641
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7806888554198398,
+      "learning_rate": 8.123141373691678e-07,
+      "loss": 1.0416,
+      "step": 7642
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7276992420040653,
+      "learning_rate": 8.113631030412932e-07,
+      "loss": 0.9631,
+      "step": 7643
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8310473963281162,
+      "learning_rate": 8.104125766000814e-07,
+      "loss": 0.877,
+      "step": 7644
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6476861104843405,
+      "learning_rate": 8.094625581607901e-07,
+      "loss": 0.8596,
+      "step": 7645
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6449035505441318,
+      "learning_rate": 8.085130478386089e-07,
+      "loss": 0.9145,
+      "step": 7646
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6675686623192103,
+      "learning_rate": 8.075640457486722e-07,
+      "loss": 0.964,
+      "step": 7647
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8523007427580116,
+      "learning_rate": 8.066155520060476e-07,
+      "loss": 0.9197,
+      "step": 7648
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7891302598776317,
+      "learning_rate": 8.056675667257441e-07,
+      "loss": 0.894,
+      "step": 7649
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.801769979750733,
+      "learning_rate": 8.047200900227092e-07,
+      "loss": 0.9074,
+      "step": 7650
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6137443757240406,
+      "learning_rate": 8.037731220118256e-07,
+      "loss": 0.8728,
+      "step": 7651
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6807320722641177,
+      "learning_rate": 8.028266628079173e-07,
+      "loss": 0.8418,
+      "step": 7652
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6853924993509217,
+      "learning_rate": 8.018807125257478e-07,
+      "loss": 0.8588,
+      "step": 7653
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7842750176187387,
+      "learning_rate": 8.009352712800128e-07,
+      "loss": 0.8706,
+      "step": 7654
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.759825912357956,
+      "learning_rate": 7.999903391853547e-07,
+      "loss": 0.8628,
+      "step": 7655
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7990374171003884,
+      "learning_rate": 7.990459163563474e-07,
+      "loss": 0.8726,
+      "step": 7656
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7871925615883618,
+      "learning_rate": 7.981020029075043e-07,
+      "loss": 1.0623,
+      "step": 7657
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.632537567379499,
+      "learning_rate": 7.971585989532793e-07,
+      "loss": 0.894,
+      "step": 7658
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.664499421824791,
+      "learning_rate": 7.962157046080648e-07,
+      "loss": 0.8328,
+      "step": 7659
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9746165672990794,
+      "learning_rate": 7.952733199861879e-07,
+      "loss": 0.8781,
+      "step": 7660
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.801139788687762,
+      "learning_rate": 7.943314452019169e-07,
+      "loss": 0.9396,
+      "step": 7661
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7184413942738999,
+      "learning_rate": 7.933900803694578e-07,
+      "loss": 0.9379,
+      "step": 7662
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7900193814473193,
+      "learning_rate": 7.924492256029531e-07,
+      "loss": 0.8515,
+      "step": 7663
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7872169143727652,
+      "learning_rate": 7.915088810164856e-07,
+      "loss": 0.9278,
+      "step": 7664
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8368366197827093,
+      "learning_rate": 7.905690467240762e-07,
+      "loss": 0.9348,
+      "step": 7665
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7850600128005745,
+      "learning_rate": 7.896297228396821e-07,
+      "loss": 0.8847,
+      "step": 7666
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.752444361923007,
+      "learning_rate": 7.886909094771983e-07,
+      "loss": 0.886,
+      "step": 7667
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7292238430428717,
+      "learning_rate": 7.877526067504615e-07,
+      "loss": 0.8447,
+      "step": 7668
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.753016418858087,
+      "learning_rate": 7.868148147732413e-07,
+      "loss": 0.9036,
+      "step": 7669
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7189331177129563,
+      "learning_rate": 7.858775336592506e-07,
+      "loss": 0.8238,
+      "step": 7670
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8074416941998166,
+      "learning_rate": 7.849407635221378e-07,
+      "loss": 1.0675,
+      "step": 7671
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.743708634926743,
+      "learning_rate": 7.840045044754879e-07,
+      "loss": 0.9145,
+      "step": 7672
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.753232106608159,
+      "learning_rate": 7.830687566328266e-07,
+      "loss": 0.793,
+      "step": 7673
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7407737711978193,
+      "learning_rate": 7.82133520107618e-07,
+      "loss": 0.8413,
+      "step": 7674
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6788612279997834,
+      "learning_rate": 7.811987950132599e-07,
+      "loss": 0.9864,
+      "step": 7675
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7049726908434872,
+      "learning_rate": 7.802645814630921e-07,
+      "loss": 0.8516,
+      "step": 7676
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7354495843714315,
+      "learning_rate": 7.79330879570393e-07,
+      "loss": 0.8664,
+      "step": 7677
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.657748055202429,
+      "learning_rate": 7.783976894483752e-07,
+      "loss": 0.8001,
+      "step": 7678
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7373619010882853,
+      "learning_rate": 7.774650112101911e-07,
+      "loss": 0.8403,
+      "step": 7679
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6857473827878116,
+      "learning_rate": 7.765328449689313e-07,
+      "loss": 0.8987,
+      "step": 7680
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8546243186991298,
+      "learning_rate": 7.756011908376254e-07,
+      "loss": 0.8931,
+      "step": 7681
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7916737711606063,
+      "learning_rate": 7.746700489292381e-07,
+      "loss": 0.8545,
+      "step": 7682
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6916002910570889,
+      "learning_rate": 7.737394193566744e-07,
+      "loss": 0.8788,
+      "step": 7683
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6652998494777649,
+      "learning_rate": 7.72809302232777e-07,
+      "loss": 0.9267,
+      "step": 7684
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6658168320928952,
+      "learning_rate": 7.718796976703241e-07,
+      "loss": 0.9264,
+      "step": 7685
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7983250095518972,
+      "learning_rate": 7.709506057820343e-07,
+      "loss": 0.8967,
+      "step": 7686
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7037808040586886,
+      "learning_rate": 7.700220266805647e-07,
+      "loss": 0.8677,
+      "step": 7687
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6387884239421937,
+      "learning_rate": 7.690939604785069e-07,
+      "loss": 0.8573,
+      "step": 7688
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.745865446843562,
+      "learning_rate": 7.681664072883915e-07,
+      "loss": 0.8585,
+      "step": 7689
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7443588843556945,
+      "learning_rate": 7.672393672226902e-07,
+      "loss": 0.888,
+      "step": 7690
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7502192909681642,
+      "learning_rate": 7.663128403938063e-07,
+      "loss": 0.9636,
+      "step": 7691
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6140759491160124,
+      "learning_rate": 7.653868269140868e-07,
+      "loss": 0.8691,
+      "step": 7692
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7733010970245242,
+      "learning_rate": 7.644613268958146e-07,
+      "loss": 1.0444,
+      "step": 7693
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.631721996678535,
+      "learning_rate": 7.635363404512069e-07,
+      "loss": 0.8333,
+      "step": 7694
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6930548948147035,
+      "learning_rate": 7.626118676924238e-07,
+      "loss": 0.9316,
+      "step": 7695
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.830490619158252,
+      "learning_rate": 7.616879087315615e-07,
+      "loss": 0.9333,
+      "step": 7696
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8203816061363631,
+      "learning_rate": 7.607644636806505e-07,
+      "loss": 1.0275,
+      "step": 7697
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7362689319455866,
+      "learning_rate": 7.59841532651664e-07,
+      "loss": 0.8702,
+      "step": 7698
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8144170470625411,
+      "learning_rate": 7.589191157565102e-07,
+      "loss": 0.9276,
+      "step": 7699
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7748218256434314,
+      "learning_rate": 7.579972131070335e-07,
+      "loss": 1.0423,
+      "step": 7700
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8963666771912087,
+      "learning_rate": 7.570758248150195e-07,
+      "loss": 0.9085,
+      "step": 7701
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.704489755929583,
+      "learning_rate": 7.561549509921901e-07,
+      "loss": 0.9449,
+      "step": 7702
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6917353817943295,
+      "learning_rate": 7.552345917502025e-07,
+      "loss": 1.0182,
+      "step": 7703
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7534234077322564,
+      "learning_rate": 7.543147472006545e-07,
+      "loss": 0.9417,
+      "step": 7704
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6608910408783957,
+      "learning_rate": 7.533954174550817e-07,
+      "loss": 0.9031,
+      "step": 7705
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7831497124393656,
+      "learning_rate": 7.524766026249535e-07,
+      "loss": 0.9568,
+      "step": 7706
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7903010117951579,
+      "learning_rate": 7.515583028216811e-07,
+      "loss": 0.8577,
+      "step": 7707
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7881487043017859,
+      "learning_rate": 7.506405181566123e-07,
+      "loss": 0.9961,
+      "step": 7708
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7429719072442948,
+      "learning_rate": 7.4972324874103e-07,
+      "loss": 0.8588,
+      "step": 7709
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7839743603731708,
+      "learning_rate": 7.488064946861556e-07,
+      "loss": 1.0408,
+      "step": 7710
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7404313576230663,
+      "learning_rate": 7.478902561031504e-07,
+      "loss": 0.8622,
+      "step": 7711
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7938450924002184,
+      "learning_rate": 7.469745331031114e-07,
+      "loss": 1.031,
+      "step": 7712
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7798506293184613,
+      "learning_rate": 7.460593257970716e-07,
+      "loss": 0.8807,
+      "step": 7713
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7359288674529645,
+      "learning_rate": 7.451446342960044e-07,
+      "loss": 0.9059,
+      "step": 7714
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7817997384596359,
+      "learning_rate": 7.442304587108202e-07,
+      "loss": 1.0652,
+      "step": 7715
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8067934781463348,
+      "learning_rate": 7.433167991523632e-07,
+      "loss": 0.8564,
+      "step": 7716
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6671905178644464,
+      "learning_rate": 7.424036557314196e-07,
+      "loss": 0.8003,
+      "step": 7717
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6677455572054083,
+      "learning_rate": 7.414910285587118e-07,
+      "loss": 0.9658,
+      "step": 7718
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.392109792983048,
+      "learning_rate": 7.405789177448986e-07,
+      "loss": 0.8123,
+      "step": 7719
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7120892186654093,
+      "learning_rate": 7.396673234005741e-07,
+      "loss": 0.8959,
+      "step": 7720
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6724411719694086,
+      "learning_rate": 7.387562456362762e-07,
+      "loss": 0.8414,
+      "step": 7721
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7387898802948645,
+      "learning_rate": 7.378456845624726e-07,
+      "loss": 0.8507,
+      "step": 7722
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7744144790886418,
+      "learning_rate": 7.369356402895739e-07,
+      "loss": 0.896,
+      "step": 7723
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7535744184587503,
+      "learning_rate": 7.360261129279272e-07,
+      "loss": 0.8594,
+      "step": 7724
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.660808961354507,
+      "learning_rate": 7.351171025878129e-07,
+      "loss": 0.7881,
+      "step": 7725
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6583237907325992,
+      "learning_rate": 7.342086093794532e-07,
+      "loss": 0.89,
+      "step": 7726
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7523281388918133,
+      "learning_rate": 7.333006334130078e-07,
+      "loss": 0.9534,
+      "step": 7727
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7313692080126075,
+      "learning_rate": 7.323931747985679e-07,
+      "loss": 0.8686,
+      "step": 7728
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7739313161210712,
+      "learning_rate": 7.314862336461692e-07,
+      "loss": 0.8238,
+      "step": 7729
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6780867567610633,
+      "learning_rate": 7.305798100657823e-07,
+      "loss": 0.8913,
+      "step": 7730
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7783599336694502,
+      "learning_rate": 7.2967390416731e-07,
+      "loss": 0.947,
+      "step": 7731
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7563416570546309,
+      "learning_rate": 7.287685160605995e-07,
+      "loss": 0.9241,
+      "step": 7732
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7506834869381207,
+      "learning_rate": 7.278636458554322e-07,
+      "loss": 0.8692,
+      "step": 7733
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6902865974098378,
+      "learning_rate": 7.269592936615249e-07,
+      "loss": 0.8747,
+      "step": 7734
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7772210872584498,
+      "learning_rate": 7.260554595885349e-07,
+      "loss": 1.0479,
+      "step": 7735
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7561431222767219,
+      "learning_rate": 7.251521437460562e-07,
+      "loss": 0.8388,
+      "step": 7736
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7239177936464762,
+      "learning_rate": 7.242493462436162e-07,
+      "loss": 0.8851,
+      "step": 7737
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8262056957339523,
+      "learning_rate": 7.233470671906845e-07,
+      "loss": 0.8667,
+      "step": 7738
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6841975692543876,
+      "learning_rate": 7.224453066966652e-07,
+      "loss": 0.8901,
+      "step": 7739
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7915975259328318,
+      "learning_rate": 7.215440648708983e-07,
+      "loss": 0.8739,
+      "step": 7740
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7362008657223553,
+      "learning_rate": 7.206433418226649e-07,
+      "loss": 0.8805,
+      "step": 7741
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7398175663805546,
+      "learning_rate": 7.197431376611785e-07,
+      "loss": 0.8579,
+      "step": 7742
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.844613770077535,
+      "learning_rate": 7.18843452495594e-07,
+      "loss": 0.8581,
+      "step": 7743
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6726879315024692,
+      "learning_rate": 7.179442864349989e-07,
+      "loss": 0.9121,
+      "step": 7744
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7302548496239603,
+      "learning_rate": 7.170456395884218e-07,
+      "loss": 0.9347,
+      "step": 7745
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7523101409781034,
+      "learning_rate": 7.161475120648276e-07,
+      "loss": 0.843,
+      "step": 7746
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.759028985512471,
+      "learning_rate": 7.152499039731153e-07,
+      "loss": 0.8395,
+      "step": 7747
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.134177032191232,
+      "learning_rate": 7.143528154221236e-07,
+      "loss": 0.8703,
+      "step": 7748
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6894303673661402,
+      "learning_rate": 7.13456246520629e-07,
+      "loss": 0.8832,
+      "step": 7749
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7143994988240272,
+      "learning_rate": 7.125601973773416e-07,
+      "loss": 0.9664,
+      "step": 7750
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7557590162891064,
+      "learning_rate": 7.116646681009121e-07,
+      "loss": 0.9363,
+      "step": 7751
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6484940923910598,
+      "learning_rate": 7.107696587999264e-07,
+      "loss": 0.9637,
+      "step": 7752
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8807956121737608,
+      "learning_rate": 7.098751695829048e-07,
+      "loss": 0.8352,
+      "step": 7753
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7150758448677024,
+      "learning_rate": 7.089812005583097e-07,
+      "loss": 0.8571,
+      "step": 7754
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7293091944638441,
+      "learning_rate": 7.080877518345386e-07,
+      "loss": 0.888,
+      "step": 7755
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8285704812578263,
+      "learning_rate": 7.071948235199227e-07,
+      "loss": 0.9282,
+      "step": 7756
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6840902766428387,
+      "learning_rate": 7.063024157227343e-07,
+      "loss": 0.9191,
+      "step": 7757
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8084632743774316,
+      "learning_rate": 7.054105285511814e-07,
+      "loss": 1.0619,
+      "step": 7758
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.767387931476611,
+      "learning_rate": 7.045191621134067e-07,
+      "loss": 0.9149,
+      "step": 7759
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7672562075077562,
+      "learning_rate": 7.036283165174923e-07,
+      "loss": 0.8776,
+      "step": 7760
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6368711205294566,
+      "learning_rate": 7.027379918714578e-07,
+      "loss": 0.8671,
+      "step": 7761
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6896878887050955,
+      "learning_rate": 7.018481882832562e-07,
+      "loss": 0.963,
+      "step": 7762
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6726730560215048,
+      "learning_rate": 7.009589058607791e-07,
+      "loss": 0.8739,
+      "step": 7763
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5900148558234117,
+      "learning_rate": 7.000701447118563e-07,
+      "loss": 0.7669,
+      "step": 7764
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7609548318208832,
+      "learning_rate": 6.991819049442516e-07,
+      "loss": 0.8467,
+      "step": 7765
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6644074246235667,
+      "learning_rate": 6.982941866656684e-07,
+      "loss": 0.9442,
+      "step": 7766
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.673595463654634,
+      "learning_rate": 6.974069899837465e-07,
+      "loss": 0.8699,
+      "step": 7767
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6805623622188068,
+      "learning_rate": 6.96520315006059e-07,
+      "loss": 0.9298,
+      "step": 7768
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7508270345304147,
+      "learning_rate": 6.956341618401197e-07,
+      "loss": 0.9307,
+      "step": 7769
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8030828713312694,
+      "learning_rate": 6.947485305933787e-07,
+      "loss": 0.9312,
+      "step": 7770
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6886665874151885,
+      "learning_rate": 6.938634213732198e-07,
+      "loss": 0.8978,
+      "step": 7771
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.715262980328026,
+      "learning_rate": 6.929788342869675e-07,
+      "loss": 0.8559,
+      "step": 7772
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6603698556265347,
+      "learning_rate": 6.92094769441879e-07,
+      "loss": 0.9629,
+      "step": 7773
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.798170876604677,
+      "learning_rate": 6.912112269451526e-07,
+      "loss": 1.0498,
+      "step": 7774
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7423056205282816,
+      "learning_rate": 6.903282069039185e-07,
+      "loss": 0.9595,
+      "step": 7775
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8531583556840845,
+      "learning_rate": 6.894457094252471e-07,
+      "loss": 0.8686,
+      "step": 7776
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6618977155343508,
+      "learning_rate": 6.88563734616145e-07,
+      "loss": 0.8046,
+      "step": 7777
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7107939615914105,
+      "learning_rate": 6.876822825835528e-07,
+      "loss": 0.9147,
+      "step": 7778
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7003297204419012,
+      "learning_rate": 6.868013534343509e-07,
+      "loss": 0.8589,
+      "step": 7779
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8023776461057955,
+      "learning_rate": 6.859209472753558e-07,
+      "loss": 1.0354,
+      "step": 7780
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8435252959107828,
+      "learning_rate": 6.850410642133176e-07,
+      "loss": 0.819,
+      "step": 7781
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7567202768305095,
+      "learning_rate": 6.841617043549259e-07,
+      "loss": 0.8827,
+      "step": 7782
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5776943904280234,
+      "learning_rate": 6.832828678068093e-07,
+      "loss": 0.9547,
+      "step": 7783
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5914041671451347,
+      "learning_rate": 6.824045546755243e-07,
+      "loss": 0.9087,
+      "step": 7784
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7872685828415734,
+      "learning_rate": 6.815267650675722e-07,
+      "loss": 1.066,
+      "step": 7785
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8242710110947193,
+      "learning_rate": 6.806494990893892e-07,
+      "loss": 0.8739,
+      "step": 7786
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.0383153332442028,
+      "learning_rate": 6.797727568473444e-07,
+      "loss": 0.9297,
+      "step": 7787
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6980530067068784,
+      "learning_rate": 6.788965384477464e-07,
+      "loss": 0.8814,
+      "step": 7788
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.857856490852591,
+      "learning_rate": 6.780208439968416e-07,
+      "loss": 0.9448,
+      "step": 7789
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7832387571216752,
+      "learning_rate": 6.771456736008086e-07,
+      "loss": 1.0819,
+      "step": 7790
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7201232613596555,
+      "learning_rate": 6.762710273657658e-07,
+      "loss": 0.8644,
+      "step": 7791
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7883479208114924,
+      "learning_rate": 6.753969053977683e-07,
+      "loss": 1.0676,
+      "step": 7792
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.817404287289521,
+      "learning_rate": 6.745233078028041e-07,
+      "loss": 0.911,
+      "step": 7793
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8406251993547607,
+      "learning_rate": 6.736502346868018e-07,
+      "loss": 0.9847,
+      "step": 7794
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7015104907799106,
+      "learning_rate": 6.72777686155624e-07,
+      "loss": 0.9438,
+      "step": 7795
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8575108680595303,
+      "learning_rate": 6.719056623150688e-07,
+      "loss": 0.9116,
+      "step": 7796
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6710221987603033,
+      "learning_rate": 6.710341632708734e-07,
+      "loss": 0.8989,
+      "step": 7797
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6475211377818015,
+      "learning_rate": 6.701631891287108e-07,
+      "loss": 0.8826,
+      "step": 7798
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8001884357317572,
+      "learning_rate": 6.692927399941879e-07,
+      "loss": 0.9042,
+      "step": 7799
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7881848965239275,
+      "learning_rate": 6.6842281597285e-07,
+      "loss": 0.9855,
+      "step": 7800
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7940411516796069,
+      "learning_rate": 6.675534171701803e-07,
+      "loss": 0.8819,
+      "step": 7801
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8609950757713096,
+      "learning_rate": 6.666845436915942e-07,
+      "loss": 0.8528,
+      "step": 7802
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7857443241832727,
+      "learning_rate": 6.658161956424458e-07,
+      "loss": 0.9352,
+      "step": 7803
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7723532933983268,
+      "learning_rate": 6.64948373128027e-07,
+      "loss": 0.9089,
+      "step": 7804
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.739182346768685,
+      "learning_rate": 6.640810762535637e-07,
+      "loss": 0.9433,
+      "step": 7805
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7099544049750004,
+      "learning_rate": 6.632143051242168e-07,
+      "loss": 0.8763,
+      "step": 7806
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7362308400147566,
+      "learning_rate": 6.623480598450865e-07,
+      "loss": 0.8856,
+      "step": 7807
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.638090381610639,
+      "learning_rate": 6.61482340521209e-07,
+      "loss": 0.8458,
+      "step": 7808
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7894078014379493,
+      "learning_rate": 6.606171472575539e-07,
+      "loss": 0.919,
+      "step": 7809
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6562597616812074,
+      "learning_rate": 6.597524801590299e-07,
+      "loss": 0.8493,
+      "step": 7810
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7839685668426282,
+      "learning_rate": 6.588883393304818e-07,
+      "loss": 0.997,
+      "step": 7811
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7414000036486947,
+      "learning_rate": 6.580247248766875e-07,
+      "loss": 0.8344,
+      "step": 7812
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.69601914283341,
+      "learning_rate": 6.571616369023642e-07,
+      "loss": 0.862,
+      "step": 7813
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.697070390605251,
+      "learning_rate": 6.562990755121662e-07,
+      "loss": 0.9495,
+      "step": 7814
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6145975173619047,
+      "learning_rate": 6.554370408106797e-07,
+      "loss": 0.7864,
+      "step": 7815
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7424470108045005,
+      "learning_rate": 6.545755329024295e-07,
+      "loss": 0.8309,
+      "step": 7816
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8234251170996645,
+      "learning_rate": 6.537145518918775e-07,
+      "loss": 0.9436,
+      "step": 7817
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7701724382962414,
+      "learning_rate": 6.528540978834191e-07,
+      "loss": 1.0135,
+      "step": 7818
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7290914003082234,
+      "learning_rate": 6.519941709813881e-07,
+      "loss": 0.8811,
+      "step": 7819
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6961596717780516,
+      "learning_rate": 6.511347712900545e-07,
+      "loss": 0.8553,
+      "step": 7820
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8626118287178162,
+      "learning_rate": 6.502758989136221e-07,
+      "loss": 0.8108,
+      "step": 7821
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8126082532986698,
+      "learning_rate": 6.494175539562326e-07,
+      "loss": 1.0517,
+      "step": 7822
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7187137452564634,
+      "learning_rate": 6.485597365219648e-07,
+      "loss": 0.9059,
+      "step": 7823
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7864708112473484,
+      "learning_rate": 6.47702446714829e-07,
+      "loss": 0.866,
+      "step": 7824
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.682651782942201,
+      "learning_rate": 6.468456846387777e-07,
+      "loss": 0.8549,
+      "step": 7825
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7710435878216572,
+      "learning_rate": 6.459894503976943e-07,
+      "loss": 1.0419,
+      "step": 7826
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7307086018090372,
+      "learning_rate": 6.451337440953997e-07,
+      "loss": 0.8997,
+      "step": 7827
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8047019490359074,
+      "learning_rate": 6.442785658356526e-07,
+      "loss": 1.0562,
+      "step": 7828
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.75444673267608,
+      "learning_rate": 6.434239157221467e-07,
+      "loss": 0.9085,
+      "step": 7829
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6914596953714456,
+      "learning_rate": 6.425697938585091e-07,
+      "loss": 0.8214,
+      "step": 7830
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7377406989013107,
+      "learning_rate": 6.417162003483062e-07,
+      "loss": 0.9251,
+      "step": 7831
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6702037485460557,
+      "learning_rate": 6.408631352950406e-07,
+      "loss": 0.8693,
+      "step": 7832
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7169569390535613,
+      "learning_rate": 6.400105988021471e-07,
+      "loss": 0.8876,
+      "step": 7833
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7723681688272388,
+      "learning_rate": 6.391585909729998e-07,
+      "loss": 0.8933,
+      "step": 7834
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8281165037805127,
+      "learning_rate": 6.38307111910908e-07,
+      "loss": 0.8551,
+      "step": 7835
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7369107429891015,
+      "learning_rate": 6.374561617191166e-07,
+      "loss": 0.975,
+      "step": 7836
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7288818712559797,
+      "learning_rate": 6.366057405008041e-07,
+      "loss": 0.8239,
+      "step": 7837
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7348096738844598,
+      "learning_rate": 6.357558483590887e-07,
+      "loss": 0.8116,
+      "step": 7838
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7610652853270758,
+      "learning_rate": 6.349064853970233e-07,
+      "loss": 0.8702,
+      "step": 7839
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.706565673429808,
+      "learning_rate": 6.340576517175939e-07,
+      "loss": 0.8371,
+      "step": 7840
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8033589397598428,
+      "learning_rate": 6.332093474237266e-07,
+      "loss": 0.945,
+      "step": 7841
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8054077753994097,
+      "learning_rate": 6.323615726182813e-07,
+      "loss": 1.0775,
+      "step": 7842
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6278831953214516,
+      "learning_rate": 6.31514327404052e-07,
+      "loss": 0.8906,
+      "step": 7843
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.711354896686527,
+      "learning_rate": 6.306676118837707e-07,
+      "loss": 0.8573,
+      "step": 7844
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.64182038273929,
+      "learning_rate": 6.29821426160106e-07,
+      "loss": 0.9186,
+      "step": 7845
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.820714873483598,
+      "learning_rate": 6.289757703356597e-07,
+      "loss": 0.8337,
+      "step": 7846
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8819564530819983,
+      "learning_rate": 6.281306445129698e-07,
+      "loss": 0.9314,
+      "step": 7847
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6144000210893918,
+      "learning_rate": 6.27286048794512e-07,
+      "loss": 0.8603,
+      "step": 7848
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7762129548058245,
+      "learning_rate": 6.264419832826946e-07,
+      "loss": 0.9058,
+      "step": 7849
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.74795955103121,
+      "learning_rate": 6.255984480798649e-07,
+      "loss": 0.9265,
+      "step": 7850
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7602707935884465,
+      "learning_rate": 6.247554432883052e-07,
+      "loss": 0.8588,
+      "step": 7851
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7984519460961446,
+      "learning_rate": 6.239129690102308e-07,
+      "loss": 0.8724,
+      "step": 7852
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7326844208148713,
+      "learning_rate": 6.230710253477956e-07,
+      "loss": 0.9146,
+      "step": 7853
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6572665232377881,
+      "learning_rate": 6.222296124030891e-07,
+      "loss": 0.8222,
+      "step": 7854
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.736445141453136,
+      "learning_rate": 6.213887302781335e-07,
+      "loss": 0.8834,
+      "step": 7855
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9065246129167421,
+      "learning_rate": 6.205483790748901e-07,
+      "loss": 0.9383,
+      "step": 7856
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7246369509476258,
+      "learning_rate": 6.197085588952556e-07,
+      "loss": 0.9117,
+      "step": 7857
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6033146934687563,
+      "learning_rate": 6.188692698410576e-07,
+      "loss": 0.8369,
+      "step": 7858
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7719305246759598,
+      "learning_rate": 6.180305120140645e-07,
+      "loss": 0.812,
+      "step": 7859
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.0768196385484505,
+      "learning_rate": 6.171922855159801e-07,
+      "loss": 0.858,
+      "step": 7860
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7635306030902915,
+      "learning_rate": 6.163545904484397e-07,
+      "loss": 0.939,
+      "step": 7861
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6984876584930622,
+      "learning_rate": 6.155174269130182e-07,
+      "loss": 0.9281,
+      "step": 7862
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8909000323797545,
+      "learning_rate": 6.146807950112249e-07,
+      "loss": 0.9912,
+      "step": 7863
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.801336717277184,
+      "learning_rate": 6.138446948445032e-07,
+      "loss": 0.934,
+      "step": 7864
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8630002506827297,
+      "learning_rate": 6.130091265142329e-07,
+      "loss": 0.9332,
+      "step": 7865
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9374952670812275,
+      "learning_rate": 6.121740901217322e-07,
+      "loss": 0.9203,
+      "step": 7866
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6589477516210713,
+      "learning_rate": 6.113395857682486e-07,
+      "loss": 0.8926,
+      "step": 7867
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7763166137758148,
+      "learning_rate": 6.105056135549714e-07,
+      "loss": 0.8664,
+      "step": 7868
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7861660980520329,
+      "learning_rate": 6.096721735830202e-07,
+      "loss": 0.8506,
+      "step": 7869
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7013882394706052,
+      "learning_rate": 6.088392659534547e-07,
+      "loss": 0.9368,
+      "step": 7870
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.743811797414318,
+      "learning_rate": 6.080068907672659e-07,
+      "loss": 1.0021,
+      "step": 7871
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6497088447310508,
+      "learning_rate": 6.071750481253835e-07,
+      "loss": 0.8624,
+      "step": 7872
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7651581393806608,
+      "learning_rate": 6.063437381286719e-07,
+      "loss": 0.905,
+      "step": 7873
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.768843292301885,
+      "learning_rate": 6.055129608779276e-07,
+      "loss": 0.9466,
+      "step": 7874
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6552891612375364,
+      "learning_rate": 6.046827164738872e-07,
+      "loss": 0.8988,
+      "step": 7875
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.0688674203397714,
+      "learning_rate": 6.03853005017222e-07,
+      "loss": 0.944,
+      "step": 7876
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.651855956280592,
+      "learning_rate": 6.030238266085342e-07,
+      "loss": 0.949,
+      "step": 7877
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7843709313063025,
+      "learning_rate": 6.021951813483673e-07,
+      "loss": 1.0395,
+      "step": 7878
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7636916448887057,
+      "learning_rate": 6.013670693371959e-07,
+      "loss": 0.8859,
+      "step": 7879
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6500553374974933,
+      "learning_rate": 6.005394906754309e-07,
+      "loss": 0.8215,
+      "step": 7880
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6994022327112646,
+      "learning_rate": 5.997124454634201e-07,
+      "loss": 0.8624,
+      "step": 7881
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7112742039458833,
+      "learning_rate": 5.988859338014457e-07,
+      "loss": 0.9346,
+      "step": 7882
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7179344729269632,
+      "learning_rate": 5.980599557897243e-07,
+      "loss": 0.9236,
+      "step": 7883
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6388942901543608,
+      "learning_rate": 5.972345115284094e-07,
+      "loss": 0.8503,
+      "step": 7884
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.713928739915737,
+      "learning_rate": 5.96409601117589e-07,
+      "loss": 0.8694,
+      "step": 7885
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7129968476476136,
+      "learning_rate": 5.955852246572852e-07,
+      "loss": 0.8432,
+      "step": 7886
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7667010812511776,
+      "learning_rate": 5.947613822474574e-07,
+      "loss": 0.8318,
+      "step": 7887
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.865419860719586,
+      "learning_rate": 5.939380739880002e-07,
+      "loss": 0.8955,
+      "step": 7888
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6611657557032762,
+      "learning_rate": 5.93115299978741e-07,
+      "loss": 0.8858,
+      "step": 7889
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8856513144596931,
+      "learning_rate": 5.922930603194438e-07,
+      "loss": 0.9238,
+      "step": 7890
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7093377542438546,
+      "learning_rate": 5.9147135510981e-07,
+      "loss": 0.8986,
+      "step": 7891
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.766929537984134,
+      "learning_rate": 5.906501844494716e-07,
+      "loss": 0.851,
+      "step": 7892
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7746675384328152,
+      "learning_rate": 5.898295484379996e-07,
+      "loss": 0.9955,
+      "step": 7893
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.646247497935098,
+      "learning_rate": 5.890094471749003e-07,
+      "loss": 0.8565,
+      "step": 7894
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7517502990098466,
+      "learning_rate": 5.881898807596115e-07,
+      "loss": 0.8728,
+      "step": 7895
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8189811847612463,
+      "learning_rate": 5.873708492915093e-07,
+      "loss": 0.8667,
+      "step": 7896
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7768415306282674,
+      "learning_rate": 5.865523528699051e-07,
+      "loss": 0.9572,
+      "step": 7897
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.697541629888557,
+      "learning_rate": 5.857343915940434e-07,
+      "loss": 0.9526,
+      "step": 7898
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8067957386047127,
+      "learning_rate": 5.849169655631054e-07,
+      "loss": 0.9177,
+      "step": 7899
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6057550101116984,
+      "learning_rate": 5.841000748762055e-07,
+      "loss": 0.8896,
+      "step": 7900
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6345049469928137,
+      "learning_rate": 5.832837196323965e-07,
+      "loss": 0.8056,
+      "step": 7901
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7121396696200009,
+      "learning_rate": 5.824678999306622e-07,
+      "loss": 0.9083,
+      "step": 7902
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7499764787668752,
+      "learning_rate": 5.816526158699243e-07,
+      "loss": 0.949,
+      "step": 7903
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.802574852045122,
+      "learning_rate": 5.8083786754904e-07,
+      "loss": 0.8465,
+      "step": 7904
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7240040921421307,
+      "learning_rate": 5.800236550667992e-07,
+      "loss": 0.8745,
+      "step": 7905
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7497590017391875,
+      "learning_rate": 5.792099785219274e-07,
+      "loss": 0.8233,
+      "step": 7906
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7341430918906509,
+      "learning_rate": 5.783968380130877e-07,
+      "loss": 0.8803,
+      "step": 7907
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6755727790916526,
+      "learning_rate": 5.775842336388742e-07,
+      "loss": 0.8515,
+      "step": 7908
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6778542515605335,
+      "learning_rate": 5.767721654978187e-07,
+      "loss": 0.8473,
+      "step": 7909
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6475146757503896,
+      "learning_rate": 5.759606336883894e-07,
+      "loss": 0.96,
+      "step": 7910
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8076493170651007,
+      "learning_rate": 5.751496383089833e-07,
+      "loss": 0.8496,
+      "step": 7911
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8327267749987748,
+      "learning_rate": 5.743391794579379e-07,
+      "loss": 0.9199,
+      "step": 7912
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9226681358014712,
+      "learning_rate": 5.735292572335261e-07,
+      "loss": 0.9428,
+      "step": 7913
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7628440127150906,
+      "learning_rate": 5.727198717339511e-07,
+      "loss": 0.9225,
+      "step": 7914
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6505130252161535,
+      "learning_rate": 5.719110230573543e-07,
+      "loss": 0.8363,
+      "step": 7915
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7350253381262497,
+      "learning_rate": 5.711027113018136e-07,
+      "loss": 0.8995,
+      "step": 7916
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.783277244904883,
+      "learning_rate": 5.702949365653365e-07,
+      "loss": 0.8799,
+      "step": 7917
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7757796319785997,
+      "learning_rate": 5.694876989458698e-07,
+      "loss": 1.01,
+      "step": 7918
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8000712806215133,
+      "learning_rate": 5.68680998541295e-07,
+      "loss": 0.876,
+      "step": 7919
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7389306840694156,
+      "learning_rate": 5.678748354494251e-07,
+      "loss": 0.825,
+      "step": 7920
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5345556694922722,
+      "learning_rate": 5.670692097680124e-07,
+      "loss": 0.8793,
+      "step": 7921
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8091850847721855,
+      "learning_rate": 5.662641215947406e-07,
+      "loss": 1.036,
+      "step": 7922
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7520169895629933,
+      "learning_rate": 5.654595710272287e-07,
+      "loss": 0.8793,
+      "step": 7923
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.753883343669283,
+      "learning_rate": 5.646555581630319e-07,
+      "loss": 0.8531,
+      "step": 7924
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7719799634592018,
+      "learning_rate": 5.638520830996402e-07,
+      "loss": 1.026,
+      "step": 7925
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.688578778612078,
+      "learning_rate": 5.630491459344766e-07,
+      "loss": 0.7993,
+      "step": 7926
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7145116085454655,
+      "learning_rate": 5.622467467649007e-07,
+      "loss": 0.8536,
+      "step": 7927
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7087940050712067,
+      "learning_rate": 5.614448856882066e-07,
+      "loss": 0.8414,
+      "step": 7928
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6713232539888867,
+      "learning_rate": 5.60643562801621e-07,
+      "loss": 0.9017,
+      "step": 7929
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7137163936424917,
+      "learning_rate": 5.598427782023086e-07,
+      "loss": 0.8653,
+      "step": 7930
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8587799743530178,
+      "learning_rate": 5.590425319873683e-07,
+      "loss": 0.9692,
+      "step": 7931
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9794648227149074,
+      "learning_rate": 5.582428242538307e-07,
+      "loss": 0.8679,
+      "step": 7932
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7824274913346327,
+      "learning_rate": 5.574436550986628e-07,
+      "loss": 0.9178,
+      "step": 7933
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7770633777902116,
+      "learning_rate": 5.566450246187672e-07,
+      "loss": 0.9244,
+      "step": 7934
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.804258750381648,
+      "learning_rate": 5.558469329109823e-07,
+      "loss": 0.8725,
+      "step": 7935
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7220075284791883,
+      "learning_rate": 5.550493800720774e-07,
+      "loss": 0.9412,
+      "step": 7936
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7733129929533094,
+      "learning_rate": 5.542523661987586e-07,
+      "loss": 1.0002,
+      "step": 7937
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7983730296490733,
+      "learning_rate": 5.534558913876686e-07,
+      "loss": 1.002,
+      "step": 7938
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8083137484727841,
+      "learning_rate": 5.526599557353807e-07,
+      "loss": 1.0522,
+      "step": 7939
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7854058358040377,
+      "learning_rate": 5.518645593384047e-07,
+      "loss": 0.8752,
+      "step": 7940
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7636924112774797,
+      "learning_rate": 5.510697022931876e-07,
+      "loss": 1.0185,
+      "step": 7941
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7782722237893485,
+      "learning_rate": 5.502753846961067e-07,
+      "loss": 0.8864,
+      "step": 7942
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7598582535354647,
+      "learning_rate": 5.494816066434749e-07,
+      "loss": 1.0607,
+      "step": 7943
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7367699302688586,
+      "learning_rate": 5.486883682315425e-07,
+      "loss": 0.933,
+      "step": 7944
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6583462118643466,
+      "learning_rate": 5.478956695564902e-07,
+      "loss": 0.8613,
+      "step": 7945
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7753376634860525,
+      "learning_rate": 5.471035107144374e-07,
+      "loss": 0.9057,
+      "step": 7946
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7955977466345596,
+      "learning_rate": 5.46311891801436e-07,
+      "loss": 0.8905,
+      "step": 7947
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7435345691058008,
+      "learning_rate": 5.455208129134704e-07,
+      "loss": 0.8429,
+      "step": 7948
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.838567215552812,
+      "learning_rate": 5.447302741464638e-07,
+      "loss": 0.9159,
+      "step": 7949
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.749712080414966,
+      "learning_rate": 5.439402755962719e-07,
+      "loss": 0.8869,
+      "step": 7950
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6114560141957042,
+      "learning_rate": 5.43150817358683e-07,
+      "loss": 0.8519,
+      "step": 7951
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6649242820896324,
+      "learning_rate": 5.423618995294239e-07,
+      "loss": 0.8999,
+      "step": 7952
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8130109141590365,
+      "learning_rate": 5.415735222041518e-07,
+      "loss": 0.8223,
+      "step": 7953
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8010902159043953,
+      "learning_rate": 5.407856854784599e-07,
+      "loss": 0.957,
+      "step": 7954
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6867213101529297,
+      "learning_rate": 5.399983894478772e-07,
+      "loss": 0.8573,
+      "step": 7955
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.846847216205357,
+      "learning_rate": 5.392116342078662e-07,
+      "loss": 0.8121,
+      "step": 7956
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7286034379292898,
+      "learning_rate": 5.384254198538224e-07,
+      "loss": 0.869,
+      "step": 7957
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6636311259950418,
+      "learning_rate": 5.376397464810779e-07,
+      "loss": 0.82,
+      "step": 7958
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8122530121909868,
+      "learning_rate": 5.36854614184899e-07,
+      "loss": 0.9347,
+      "step": 7959
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.804172766116558,
+      "learning_rate": 5.360700230604842e-07,
+      "loss": 0.9893,
+      "step": 7960
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6640679028720096,
+      "learning_rate": 5.352859732029686e-07,
+      "loss": 0.8365,
+      "step": 7961
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7716065276171995,
+      "learning_rate": 5.345024647074221e-07,
+      "loss": 0.9081,
+      "step": 7962
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7670520956670592,
+      "learning_rate": 5.337194976688464e-07,
+      "loss": 0.9582,
+      "step": 7963
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.599721672768864,
+      "learning_rate": 5.329370721821781e-07,
+      "loss": 0.8467,
+      "step": 7964
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7575973758919392,
+      "learning_rate": 5.321551883422904e-07,
+      "loss": 0.9027,
+      "step": 7965
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.670566651606589,
+      "learning_rate": 5.3137384624399e-07,
+      "loss": 0.8479,
+      "step": 7966
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6620611459241823,
+      "learning_rate": 5.305930459820158e-07,
+      "loss": 0.8464,
+      "step": 7967
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8138824929006832,
+      "learning_rate": 5.298127876510429e-07,
+      "loss": 0.8844,
+      "step": 7968
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7538607543266584,
+      "learning_rate": 5.29033071345681e-07,
+      "loss": 0.9612,
+      "step": 7969
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7459935380197018,
+      "learning_rate": 5.282538971604729e-07,
+      "loss": 0.9495,
+      "step": 7970
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7253545031312323,
+      "learning_rate": 5.274752651898957e-07,
+      "loss": 0.8859,
+      "step": 7971
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7689301395781445,
+      "learning_rate": 5.266971755283628e-07,
+      "loss": 0.9595,
+      "step": 7972
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6964134773527824,
+      "learning_rate": 5.259196282702178e-07,
+      "loss": 0.8575,
+      "step": 7973
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6822730003470077,
+      "learning_rate": 5.25142623509744e-07,
+      "loss": 0.8421,
+      "step": 7974
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6508261834536417,
+      "learning_rate": 5.243661613411544e-07,
+      "loss": 0.8738,
+      "step": 7975
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8144014282122483,
+      "learning_rate": 5.235902418585958e-07,
+      "loss": 0.8259,
+      "step": 7976
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7920097811792606,
+      "learning_rate": 5.228148651561537e-07,
+      "loss": 0.8071,
+      "step": 7977
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.670264366394135,
+      "learning_rate": 5.22040031327845e-07,
+      "loss": 0.8672,
+      "step": 7978
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7020123838195784,
+      "learning_rate": 5.212657404676192e-07,
+      "loss": 0.9179,
+      "step": 7979
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7464520726371668,
+      "learning_rate": 5.204919926693635e-07,
+      "loss": 0.9002,
+      "step": 7980
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.708216445308849,
+      "learning_rate": 5.197187880268973e-07,
+      "loss": 0.9451,
+      "step": 7981
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7912226423136715,
+      "learning_rate": 5.189461266339735e-07,
+      "loss": 0.8775,
+      "step": 7982
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7927089787515248,
+      "learning_rate": 5.181740085842795e-07,
+      "loss": 1.0352,
+      "step": 7983
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5725626564890451,
+      "learning_rate": 5.174024339714401e-07,
+      "loss": 0.9518,
+      "step": 7984
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7957600221265526,
+      "learning_rate": 5.166314028890091e-07,
+      "loss": 0.9612,
+      "step": 7985
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6649172717144423,
+      "learning_rate": 5.158609154304761e-07,
+      "loss": 0.7555,
+      "step": 7986
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6134719274800247,
+      "learning_rate": 5.150909716892671e-07,
+      "loss": 0.855,
+      "step": 7987
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5990768105394935,
+      "learning_rate": 5.143215717587385e-07,
+      "loss": 0.8743,
+      "step": 7988
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7064954520102924,
+      "learning_rate": 5.13552715732184e-07,
+      "loss": 0.8455,
+      "step": 7989
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7383807833040477,
+      "learning_rate": 5.127844037028313e-07,
+      "loss": 0.8494,
+      "step": 7990
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6323407361198568,
+      "learning_rate": 5.120166357638378e-07,
+      "loss": 0.8649,
+      "step": 7991
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.741055547598097,
+      "learning_rate": 5.112494120083e-07,
+      "loss": 0.9117,
+      "step": 7992
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7083230821739126,
+      "learning_rate": 5.104827325292466e-07,
+      "loss": 0.8849,
+      "step": 7993
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7497514562162197,
+      "learning_rate": 5.097165974196389e-07,
+      "loss": 0.8295,
+      "step": 7994
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7816401543559374,
+      "learning_rate": 5.089510067723752e-07,
+      "loss": 1.0412,
+      "step": 7995
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7130542851389865,
+      "learning_rate": 5.081859606802841e-07,
+      "loss": 0.8306,
+      "step": 7996
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.585374111772938,
+      "learning_rate": 5.074214592361315e-07,
+      "loss": 0.8809,
+      "step": 7997
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.704889214866008,
+      "learning_rate": 5.066575025326148e-07,
+      "loss": 1.0048,
+      "step": 7998
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.688307570904051,
+      "learning_rate": 5.05894090662366e-07,
+      "loss": 0.8727,
+      "step": 7999
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7910509254765243,
+      "learning_rate": 5.051312237179539e-07,
+      "loss": 1.0374,
+      "step": 8000
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8153528947878643,
+      "learning_rate": 5.043689017918757e-07,
+      "loss": 0.9174,
+      "step": 8001
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.816183859868071,
+      "learning_rate": 5.036071249765673e-07,
+      "loss": 0.8121,
+      "step": 8002
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7348330039927427,
+      "learning_rate": 5.028458933643976e-07,
+      "loss": 0.9448,
+      "step": 8003
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.70606789961362,
+      "learning_rate": 5.020852070476657e-07,
+      "loss": 0.8248,
+      "step": 8004
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.777563833275154,
+      "learning_rate": 5.013250661186103e-07,
+      "loss": 0.8236,
+      "step": 8005
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7055021383175055,
+      "learning_rate": 5.005654706694002e-07,
+      "loss": 0.9108,
+      "step": 8006
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.719270678615359,
+      "learning_rate": 4.998064207921377e-07,
+      "loss": 0.8781,
+      "step": 8007
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7760079722646493,
+      "learning_rate": 4.990479165788609e-07,
+      "loss": 0.9381,
+      "step": 8008
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8193308033775668,
+      "learning_rate": 4.982899581215428e-07,
+      "loss": 0.9143,
+      "step": 8009
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7383998627765311,
+      "learning_rate": 4.975325455120855e-07,
+      "loss": 0.8934,
+      "step": 8010
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.729471515307609,
+      "learning_rate": 4.9677567884233e-07,
+      "loss": 0.9249,
+      "step": 8011
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8763879913062969,
+      "learning_rate": 4.96019358204049e-07,
+      "loss": 0.822,
+      "step": 8012
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6426464075305853,
+      "learning_rate": 4.952635836889475e-07,
+      "loss": 0.8629,
+      "step": 8013
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6671705124592293,
+      "learning_rate": 4.94508355388667e-07,
+      "loss": 0.8708,
+      "step": 8014
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6947551754001913,
+      "learning_rate": 4.937536733947817e-07,
+      "loss": 0.9263,
+      "step": 8015
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7346097204123276,
+      "learning_rate": 4.92999537798799e-07,
+      "loss": 0.8393,
+      "step": 8016
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8675047846233566,
+      "learning_rate": 4.922459486921594e-07,
+      "loss": 0.8126,
+      "step": 8017
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7063534299477443,
+      "learning_rate": 4.914929061662399e-07,
+      "loss": 0.8365,
+      "step": 8018
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8048899249347088,
+      "learning_rate": 4.907404103123481e-07,
+      "loss": 1.0782,
+      "step": 8019
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7250709764390129,
+      "learning_rate": 4.89988461221727e-07,
+      "loss": 0.82,
+      "step": 8020
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.776156454801782,
+      "learning_rate": 4.89237058985555e-07,
+      "loss": 0.9719,
+      "step": 8021
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6358633380519503,
+      "learning_rate": 4.88486203694939e-07,
+      "loss": 0.8769,
+      "step": 8022
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6784351050334996,
+      "learning_rate": 4.877358954409245e-07,
+      "loss": 0.8654,
+      "step": 8023
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8868439257367278,
+      "learning_rate": 4.869861343144894e-07,
+      "loss": 0.8079,
+      "step": 8024
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7352715391140265,
+      "learning_rate": 4.862369204065437e-07,
+      "loss": 0.8855,
+      "step": 8025
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6783778390905457,
+      "learning_rate": 4.854882538079331e-07,
+      "loss": 0.8556,
+      "step": 8026
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6363815257874872,
+      "learning_rate": 4.847401346094361e-07,
+      "loss": 0.8393,
+      "step": 8027
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7805608358024557,
+      "learning_rate": 4.839925629017638e-07,
+      "loss": 1.0733,
+      "step": 8028
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7247545916252052,
+      "learning_rate": 4.832455387755619e-07,
+      "loss": 0.9149,
+      "step": 8029
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6600147169553279,
+      "learning_rate": 4.824990623214099e-07,
+      "loss": 0.9254,
+      "step": 8030
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.816016405509356,
+      "learning_rate": 4.817531336298221e-07,
+      "loss": 0.8752,
+      "step": 8031
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6513352468162448,
+      "learning_rate": 4.810077527912422e-07,
+      "loss": 0.8738,
+      "step": 8032
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6491773509683738,
+      "learning_rate": 4.802629198960517e-07,
+      "loss": 0.9033,
+      "step": 8033
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6435320237552906,
+      "learning_rate": 4.795186350345648e-07,
+      "loss": 0.8562,
+      "step": 8034
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.776172509483379,
+      "learning_rate": 4.787748982970275e-07,
+      "loss": 0.9996,
+      "step": 8035
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5376376380411116,
+      "learning_rate": 4.780317097736203e-07,
+      "loss": 0.8893,
+      "step": 8036
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7687407000085977,
+      "learning_rate": 4.772890695544602e-07,
+      "loss": 0.9107,
+      "step": 8037
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6244048452696227,
+      "learning_rate": 4.7654697772959057e-07,
+      "loss": 0.9174,
+      "step": 8038
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.81002711906273,
+      "learning_rate": 4.758054343889945e-07,
+      "loss": 0.9022,
+      "step": 8039
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7510325643806377,
+      "learning_rate": 4.750644396225873e-07,
+      "loss": 0.8804,
+      "step": 8040
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6776147876584806,
+      "learning_rate": 4.743239935202165e-07,
+      "loss": 0.8936,
+      "step": 8041
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7775054578373586,
+      "learning_rate": 4.735840961716631e-07,
+      "loss": 0.837,
+      "step": 8042
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7725974036631507,
+      "learning_rate": 4.728447476666442e-07,
+      "loss": 0.8984,
+      "step": 8043
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8185002329916373,
+      "learning_rate": 4.7210594809480647e-07,
+      "loss": 0.956,
+      "step": 8044
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8460188304989817,
+      "learning_rate": 4.713676975457321e-07,
+      "loss": 0.9083,
+      "step": 8045
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7111886197232955,
+      "learning_rate": 4.7062999610893835e-07,
+      "loss": 0.8499,
+      "step": 8046
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7751571878681345,
+      "learning_rate": 4.698928438738715e-07,
+      "loss": 0.9082,
+      "step": 8047
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.802182662732776,
+      "learning_rate": 4.691562409299161e-07,
+      "loss": 0.8759,
+      "step": 8048
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7922927390253922,
+      "learning_rate": 4.6842018736638685e-07,
+      "loss": 1.0281,
+      "step": 8049
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.727938724111258,
+      "learning_rate": 4.6768468327253127e-07,
+      "loss": 0.9053,
+      "step": 8050
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.4594072211268045,
+      "learning_rate": 4.66949728737533e-07,
+      "loss": 0.8201,
+      "step": 8051
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6930996005957646,
+      "learning_rate": 4.6621532385050904e-07,
+      "loss": 0.9343,
+      "step": 8052
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7170146122482113,
+      "learning_rate": 4.654814687005066e-07,
+      "loss": 0.9348,
+      "step": 8053
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.884854534956769,
+      "learning_rate": 4.6474816337650883e-07,
+      "loss": 0.8467,
+      "step": 8054
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7174408997173494,
+      "learning_rate": 4.640154079674325e-07,
+      "loss": 0.8094,
+      "step": 8055
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.87605869127692,
+      "learning_rate": 4.6328320256212534e-07,
+      "loss": 0.8767,
+      "step": 8056
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8648936647214198,
+      "learning_rate": 4.625515472493697e-07,
+      "loss": 0.8017,
+      "step": 8057
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7046386164481417,
+      "learning_rate": 4.618204421178835e-07,
+      "loss": 0.8383,
+      "step": 8058
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7150157750160273,
+      "learning_rate": 4.610898872563135e-07,
+      "loss": 0.9014,
+      "step": 8059
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7393285090018638,
+      "learning_rate": 4.603598827532424e-07,
+      "loss": 0.8787,
+      "step": 8060
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7421368147517011,
+      "learning_rate": 4.596304286971853e-07,
+      "loss": 0.9119,
+      "step": 8061
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7552876524109977,
+      "learning_rate": 4.589015251765933e-07,
+      "loss": 0.886,
+      "step": 8062
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7740399297622256,
+      "learning_rate": 4.58173172279846e-07,
+      "loss": 0.9273,
+      "step": 8063
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6619775429351784,
+      "learning_rate": 4.574453700952591e-07,
+      "loss": 0.8164,
+      "step": 8064
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7436017289728163,
+      "learning_rate": 4.567181187110831e-07,
+      "loss": 0.8232,
+      "step": 8065
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6712952050676817,
+      "learning_rate": 4.5599141821549673e-07,
+      "loss": 0.951,
+      "step": 8066
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6730732802771922,
+      "learning_rate": 4.552652686966169e-07,
+      "loss": 0.9615,
+      "step": 8067
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.770916945578325,
+      "learning_rate": 4.545396702424926e-07,
+      "loss": 0.9703,
+      "step": 8068
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7442625841483845,
+      "learning_rate": 4.5381462294110336e-07,
+      "loss": 0.8779,
+      "step": 8069
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6576426138856928,
+      "learning_rate": 4.5309012688036335e-07,
+      "loss": 0.8649,
+      "step": 8070
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7120236968782254,
+      "learning_rate": 4.5236618214812233e-07,
+      "loss": 0.8468,
+      "step": 8071
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7924976104291408,
+      "learning_rate": 4.516427888321584e-07,
+      "loss": 0.8688,
+      "step": 8072
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7277319154563702,
+      "learning_rate": 4.5091994702018695e-07,
+      "loss": 0.8545,
+      "step": 8073
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6849481146237761,
+      "learning_rate": 4.5019765679985614e-07,
+      "loss": 0.9143,
+      "step": 8074
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6561628558090684,
+      "learning_rate": 4.4947591825874383e-07,
+      "loss": 0.9423,
+      "step": 8075
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6758387354156294,
+      "learning_rate": 4.487547314843643e-07,
+      "loss": 0.9856,
+      "step": 8076
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7187461631143157,
+      "learning_rate": 4.480340965641655e-07,
+      "loss": 0.8725,
+      "step": 8077
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7163908638667729,
+      "learning_rate": 4.473140135855247e-07,
+      "loss": 0.916,
+      "step": 8078
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8488989254824135,
+      "learning_rate": 4.4659448263575544e-07,
+      "loss": 0.9783,
+      "step": 8079
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8034663591798084,
+      "learning_rate": 4.458755038021029e-07,
+      "loss": 0.9117,
+      "step": 8080
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7048816914822535,
+      "learning_rate": 4.451570771717467e-07,
+      "loss": 0.8337,
+      "step": 8081
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.745741296671989,
+      "learning_rate": 4.444392028317968e-07,
+      "loss": 0.8995,
+      "step": 8082
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6921835156004805,
+      "learning_rate": 4.437218808693e-07,
+      "loss": 0.88,
+      "step": 8083
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.618272883163135,
+      "learning_rate": 4.4300511137123193e-07,
+      "loss": 0.905,
+      "step": 8084
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7878331536409675,
+      "learning_rate": 4.4228889442450463e-07,
+      "loss": 1.0481,
+      "step": 8085
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7804815306051947,
+      "learning_rate": 4.4157323011596254e-07,
+      "loss": 0.8395,
+      "step": 8086
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6968938840559376,
+      "learning_rate": 4.408581185323807e-07,
+      "loss": 0.8233,
+      "step": 8087
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6202629040841932,
+      "learning_rate": 4.4014355976046974e-07,
+      "loss": 0.8339,
+      "step": 8088
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8398557056051188,
+      "learning_rate": 4.394295538868731e-07,
+      "loss": 0.881,
+      "step": 8089
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7092206346052676,
+      "learning_rate": 4.3871610099816606e-07,
+      "loss": 0.8661,
+      "step": 8090
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.70297316506603,
+      "learning_rate": 4.3800320118085604e-07,
+      "loss": 0.8999,
+      "step": 8091
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6831996585824591,
+      "learning_rate": 4.37290854521385e-07,
+      "loss": 0.8622,
+      "step": 8092
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.667577789771012,
+      "learning_rate": 4.3657906110612935e-07,
+      "loss": 0.8953,
+      "step": 8093
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.722015589901042,
+      "learning_rate": 4.358678210213935e-07,
+      "loss": 0.9319,
+      "step": 8094
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.721836406912448,
+      "learning_rate": 4.3515713435342e-07,
+      "loss": 0.8599,
+      "step": 8095
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7838899583779377,
+      "learning_rate": 4.3444700118838167e-07,
+      "loss": 0.8881,
+      "step": 8096
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8084587348900927,
+      "learning_rate": 4.337374216123841e-07,
+      "loss": 0.8989,
+      "step": 8097
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8681547234807678,
+      "learning_rate": 4.330283957114656e-07,
+      "loss": 0.8839,
+      "step": 8098
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.0044976851354743,
+      "learning_rate": 4.3231992357160026e-07,
+      "loss": 0.7993,
+      "step": 8099
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6738531083260109,
+      "learning_rate": 4.316120052786904e-07,
+      "loss": 0.8744,
+      "step": 8100
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8382023475712965,
+      "learning_rate": 4.3090464091857576e-07,
+      "loss": 0.9305,
+      "step": 8101
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6836123502412175,
+      "learning_rate": 4.3019783057702493e-07,
+      "loss": 0.875,
+      "step": 8102
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.719647776787755,
+      "learning_rate": 4.29491574339741e-07,
+      "loss": 0.9556,
+      "step": 8103
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.815766205458188,
+      "learning_rate": 4.287858722923605e-07,
+      "loss": 0.8301,
+      "step": 8104
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.755622917717037,
+      "learning_rate": 4.2808072452045377e-07,
+      "loss": 1.0294,
+      "step": 8105
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7210863465768784,
+      "learning_rate": 4.2737613110951924e-07,
+      "loss": 0.8171,
+      "step": 8106
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8334825786884967,
+      "learning_rate": 4.266720921449935e-07,
+      "loss": 0.8408,
+      "step": 8107
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6728134575584162,
+      "learning_rate": 4.2596860771224435e-07,
+      "loss": 0.8195,
+      "step": 8108
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.3156528578368825,
+      "learning_rate": 4.252656778965697e-07,
+      "loss": 0.8667,
+      "step": 8109
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7392662344132004,
+      "learning_rate": 4.2456330278320245e-07,
+      "loss": 0.8534,
+      "step": 8110
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.725682930397735,
+      "learning_rate": 4.238614824573101e-07,
+      "loss": 0.8911,
+      "step": 8111
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8089098699349329,
+      "learning_rate": 4.2316021700398945e-07,
+      "loss": 0.794,
+      "step": 8112
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7532770217872775,
+      "learning_rate": 4.2245950650827037e-07,
+      "loss": 0.9476,
+      "step": 8113
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7745959127619306,
+      "learning_rate": 4.2175935105511767e-07,
+      "loss": 1.0359,
+      "step": 8114
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7831060897720843,
+      "learning_rate": 4.2105975072942674e-07,
+      "loss": 0.8605,
+      "step": 8115
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.760119585753882,
+      "learning_rate": 4.2036070561602695e-07,
+      "loss": 1.0839,
+      "step": 8116
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6661119537642035,
+      "learning_rate": 4.1966221579968103e-07,
+      "loss": 0.949,
+      "step": 8117
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9474873672361783,
+      "learning_rate": 4.1896428136508126e-07,
+      "loss": 0.9884,
+      "step": 8118
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6237346443567215,
+      "learning_rate": 4.182669023968561e-07,
+      "loss": 0.8585,
+      "step": 8119
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.786639003616739,
+      "learning_rate": 4.1757007897956503e-07,
+      "loss": 0.9695,
+      "step": 8120
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7787767312588021,
+      "learning_rate": 4.1687381119769887e-07,
+      "loss": 1.003,
+      "step": 8121
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8079335505777928,
+      "learning_rate": 4.161780991356845e-07,
+      "loss": 1.0626,
+      "step": 8122
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.701481883475435,
+      "learning_rate": 4.154829428778778e-07,
+      "loss": 0.8897,
+      "step": 8123
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.818219366432782,
+      "learning_rate": 4.1478834250857027e-07,
+      "loss": 0.9126,
+      "step": 8124
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.260961382099815,
+      "learning_rate": 4.140942981119833e-07,
+      "loss": 0.9304,
+      "step": 8125
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.72885634959852,
+      "learning_rate": 4.134008097722725e-07,
+      "loss": 0.8211,
+      "step": 8126
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6378426653786742,
+      "learning_rate": 4.127078775735266e-07,
+      "loss": 0.799,
+      "step": 8127
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7102947427228954,
+      "learning_rate": 4.1201550159976454e-07,
+      "loss": 0.9564,
+      "step": 8128
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.824772417577214,
+      "learning_rate": 4.113236819349403e-07,
+      "loss": 0.927,
+      "step": 8129
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.702348057123386,
+      "learning_rate": 4.1063241866293955e-07,
+      "loss": 0.8913,
+      "step": 8130
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8187677372983375,
+      "learning_rate": 4.099417118675797e-07,
+      "loss": 0.8811,
+      "step": 8131
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8205760936554345,
+      "learning_rate": 4.092515616326126e-07,
+      "loss": 0.9485,
+      "step": 8132
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7679182208883465,
+      "learning_rate": 4.0856196804171965e-07,
+      "loss": 1.0463,
+      "step": 8133
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7803805669142488,
+      "learning_rate": 4.078729311785168e-07,
+      "loss": 1.0325,
+      "step": 8134
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8440505103888065,
+      "learning_rate": 4.071844511265527e-07,
+      "loss": 0.9139,
+      "step": 8135
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7882096588328067,
+      "learning_rate": 4.064965279693084e-07,
+      "loss": 0.9549,
+      "step": 8136
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7930667856350735,
+      "learning_rate": 4.058091617901949e-07,
+      "loss": 0.8789,
+      "step": 8137
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7666321976988473,
+      "learning_rate": 4.0512235267256007e-07,
+      "loss": 0.9522,
+      "step": 8138
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6579234455843967,
+      "learning_rate": 4.044361006996811e-07,
+      "loss": 0.8823,
+      "step": 8139
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7610444687758813,
+      "learning_rate": 4.0375040595476753e-07,
+      "loss": 0.9515,
+      "step": 8140
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.779367572345615,
+      "learning_rate": 4.0306526852096296e-07,
+      "loss": 1.0282,
+      "step": 8141
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6695118733110033,
+      "learning_rate": 4.023806884813436e-07,
+      "loss": 0.791,
+      "step": 8142
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7501156902116224,
+      "learning_rate": 4.016966659189159e-07,
+      "loss": 0.8607,
+      "step": 8143
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7310092272451592,
+      "learning_rate": 4.0101320091661954e-07,
+      "loss": 0.9122,
+      "step": 8144
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8176749191318708,
+      "learning_rate": 4.003302935573289e-07,
+      "loss": 0.9541,
+      "step": 8145
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7134905405841765,
+      "learning_rate": 3.996479439238465e-07,
+      "loss": 0.9011,
+      "step": 8146
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7755983719651804,
+      "learning_rate": 3.989661520989102e-07,
+      "loss": 0.8028,
+      "step": 8147
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6696769578005788,
+      "learning_rate": 3.9828491816519153e-07,
+      "loss": 0.8945,
+      "step": 8148
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.777880467216096,
+      "learning_rate": 3.9760424220529e-07,
+      "loss": 1.0282,
+      "step": 8149
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5865532294561977,
+      "learning_rate": 3.969241243017413e-07,
+      "loss": 0.8616,
+      "step": 8150
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7647819676090695,
+      "learning_rate": 3.962445645370122e-07,
+      "loss": 0.9001,
+      "step": 8151
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7510273029426933,
+      "learning_rate": 3.955655629935007e-07,
+      "loss": 0.7952,
+      "step": 8152
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6593281449366233,
+      "learning_rate": 3.9488711975353866e-07,
+      "loss": 0.8933,
+      "step": 8153
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.665852040270858,
+      "learning_rate": 3.9420923489939034e-07,
+      "loss": 0.8471,
+      "step": 8154
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7407635567241004,
+      "learning_rate": 3.935319085132505e-07,
+      "loss": 0.8895,
+      "step": 8155
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6318697124188302,
+      "learning_rate": 3.928551406772468e-07,
+      "loss": 0.9089,
+      "step": 8156
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6389472793304734,
+      "learning_rate": 3.921789314734409e-07,
+      "loss": 0.8909,
+      "step": 8157
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.688808798869453,
+      "learning_rate": 3.9150328098382593e-07,
+      "loss": 0.7868,
+      "step": 8158
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8258002826014934,
+      "learning_rate": 3.908281892903254e-07,
+      "loss": 0.9826,
+      "step": 8159
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.683348948672413,
+      "learning_rate": 3.9015365647479655e-07,
+      "loss": 0.815,
+      "step": 8160
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6876235519214713,
+      "learning_rate": 3.8947968261903055e-07,
+      "loss": 0.8485,
+      "step": 8161
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.722380106101749,
+      "learning_rate": 3.8880626780474717e-07,
+      "loss": 0.8597,
+      "step": 8162
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6771966477792897,
+      "learning_rate": 3.88133412113601e-07,
+      "loss": 0.7686,
+      "step": 8163
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.723507518340281,
+      "learning_rate": 3.874611156271801e-07,
+      "loss": 0.8606,
+      "step": 8164
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.818899301957327,
+      "learning_rate": 3.8678937842699884e-07,
+      "loss": 0.874,
+      "step": 8165
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8647459702706561,
+      "learning_rate": 3.8611820059450913e-07,
+      "loss": 0.9046,
+      "step": 8166
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7928131330933363,
+      "learning_rate": 3.8544758221109534e-07,
+      "loss": 1.075,
+      "step": 8167
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.791917699158972,
+      "learning_rate": 3.8477752335807026e-07,
+      "loss": 0.8587,
+      "step": 8168
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.729018501468878,
+      "learning_rate": 3.841080241166811e-07,
+      "loss": 0.9033,
+      "step": 8169
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7988438949769137,
+      "learning_rate": 3.834390845681091e-07,
+      "loss": 0.8905,
+      "step": 8170
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9008796033054556,
+      "learning_rate": 3.82770704793462e-07,
+      "loss": 0.879,
+      "step": 8171
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7178303932301533,
+      "learning_rate": 3.821028848737857e-07,
+      "loss": 0.8752,
+      "step": 8172
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7143364164150483,
+      "learning_rate": 3.8143562489005525e-07,
+      "loss": 0.8755,
+      "step": 8173
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6951119947358289,
+      "learning_rate": 3.807689249231772e-07,
+      "loss": 0.9141,
+      "step": 8174
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7760369745336935,
+      "learning_rate": 3.801027850539929e-07,
+      "loss": 0.836,
+      "step": 8175
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6641904400138208,
+      "learning_rate": 3.794372053632722e-07,
+      "loss": 0.8125,
+      "step": 8176
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6883503627025096,
+      "learning_rate": 3.787721859317206e-07,
+      "loss": 0.8758,
+      "step": 8177
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.773589767564627,
+      "learning_rate": 3.7810772683997245e-07,
+      "loss": 1.085,
+      "step": 8178
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7734077416308642,
+      "learning_rate": 3.774438281685977e-07,
+      "loss": 0.8774,
+      "step": 8179
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7624409633092848,
+      "learning_rate": 3.767804899980937e-07,
+      "loss": 0.9973,
+      "step": 8180
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6583213643361574,
+      "learning_rate": 3.761177124088944e-07,
+      "loss": 0.9263,
+      "step": 8181
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6559030408712874,
+      "learning_rate": 3.754554954813644e-07,
+      "loss": 0.9087,
+      "step": 8182
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7653366808982298,
+      "learning_rate": 3.747938392957973e-07,
+      "loss": 0.8727,
+      "step": 8183
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8141702277947844,
+      "learning_rate": 3.7413274393242327e-07,
+      "loss": 0.8631,
+      "step": 8184
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7277509683651031,
+      "learning_rate": 3.7347220947140273e-07,
+      "loss": 0.8906,
+      "step": 8185
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7297884823958671,
+      "learning_rate": 3.7281223599282656e-07,
+      "loss": 0.9022,
+      "step": 8186
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6981736781539112,
+      "learning_rate": 3.7215282357671803e-07,
+      "loss": 0.9113,
+      "step": 8187
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6502757275622002,
+      "learning_rate": 3.7149397230303473e-07,
+      "loss": 0.9434,
+      "step": 8188
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7419547375187157,
+      "learning_rate": 3.7083568225166456e-07,
+      "loss": 0.8514,
+      "step": 8189
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.1287243611026656,
+      "learning_rate": 3.701779535024269e-07,
+      "loss": 0.9456,
+      "step": 8190
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7395438612625287,
+      "learning_rate": 3.6952078613507304e-07,
+      "loss": 0.9324,
+      "step": 8191
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7368789779615625,
+      "learning_rate": 3.688641802292892e-07,
+      "loss": 0.8322,
+      "step": 8192
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7343094963255603,
+      "learning_rate": 3.682081358646883e-07,
+      "loss": 0.8903,
+      "step": 8193
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5542959484142138,
+      "learning_rate": 3.67552653120819e-07,
+      "loss": 0.844,
+      "step": 8194
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.79854860654558,
+      "learning_rate": 3.668977320771616e-07,
+      "loss": 0.897,
+      "step": 8195
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7563731980988884,
+      "learning_rate": 3.6624337281312704e-07,
+      "loss": 0.9285,
+      "step": 8196
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.704541133324898,
+      "learning_rate": 3.6558957540805794e-07,
+      "loss": 0.8265,
+      "step": 8197
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.707805061123568,
+      "learning_rate": 3.649363399412309e-07,
+      "loss": 0.9372,
+      "step": 8198
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7370905444092848,
+      "learning_rate": 3.642836664918509e-07,
+      "loss": 0.8409,
+      "step": 8199
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7284720708583472,
+      "learning_rate": 3.6363155513905847e-07,
+      "loss": 0.93,
+      "step": 8200
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7071436315797968,
+      "learning_rate": 3.629800059619248e-07,
+      "loss": 0.9168,
+      "step": 8201
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7540245316243035,
+      "learning_rate": 3.6232901903945075e-07,
+      "loss": 0.8467,
+      "step": 8202
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.741262651953551,
+      "learning_rate": 3.616785944505713e-07,
+      "loss": 0.9054,
+      "step": 8203
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8118644964216342,
+      "learning_rate": 3.61028732274154e-07,
+      "loss": 0.9933,
+      "step": 8204
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6980224099886094,
+      "learning_rate": 3.603794325889953e-07,
+      "loss": 0.9186,
+      "step": 8205
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.713919473476343,
+      "learning_rate": 3.5973069547382556e-07,
+      "loss": 0.7973,
+      "step": 8206
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7141362790152725,
+      "learning_rate": 3.5908252100730735e-07,
+      "loss": 0.912,
+      "step": 8207
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6281356580153694,
+      "learning_rate": 3.584349092680328e-07,
+      "loss": 0.821,
+      "step": 8208
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.621482759413654,
+      "learning_rate": 3.577878603345269e-07,
+      "loss": 0.951,
+      "step": 8209
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7572882205238864,
+      "learning_rate": 3.5714137428524754e-07,
+      "loss": 0.9538,
+      "step": 8210
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6877925181437599,
+      "learning_rate": 3.5649545119858244e-07,
+      "loss": 0.9048,
+      "step": 8211
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6772556997647348,
+      "learning_rate": 3.558500911528523e-07,
+      "loss": 0.9105,
+      "step": 8212
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7774330625583736,
+      "learning_rate": 3.552052942263101e-07,
+      "loss": 1.0366,
+      "step": 8213
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7705531994266366,
+      "learning_rate": 3.5456106049713835e-07,
+      "loss": 0.7759,
+      "step": 8214
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7755764200168127,
+      "learning_rate": 3.5391739004345336e-07,
+      "loss": 1.0619,
+      "step": 8215
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6460772463533957,
+      "learning_rate": 3.5327428294330333e-07,
+      "loss": 0.9031,
+      "step": 8216
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8196801791652817,
+      "learning_rate": 3.5263173927466585e-07,
+      "loss": 1.0199,
+      "step": 8217
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6113002643086793,
+      "learning_rate": 3.5198975911545143e-07,
+      "loss": 0.8825,
+      "step": 8218
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.693174227300383,
+      "learning_rate": 3.5134834254350213e-07,
+      "loss": 0.8883,
+      "step": 8219
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7251567389026323,
+      "learning_rate": 3.5070748963659417e-07,
+      "loss": 0.9188,
+      "step": 8220
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.855696886155795,
+      "learning_rate": 3.5006720047243037e-07,
+      "loss": 0.9299,
+      "step": 8221
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6225522699011063,
+      "learning_rate": 3.494274751286497e-07,
+      "loss": 0.9041,
+      "step": 8222
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9062900312357827,
+      "learning_rate": 3.487883136828213e-07,
+      "loss": 0.8708,
+      "step": 8223
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8826443907434436,
+      "learning_rate": 3.481497162124442e-07,
+      "loss": 0.8647,
+      "step": 8224
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.733252118139951,
+      "learning_rate": 3.4751168279495094e-07,
+      "loss": 0.9538,
+      "step": 8225
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6118799716555123,
+      "learning_rate": 3.468742135077069e-07,
+      "loss": 0.953,
+      "step": 8226
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6836291484866632,
+      "learning_rate": 3.4623730842800573e-07,
+      "loss": 0.8231,
+      "step": 8227
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6650101702332119,
+      "learning_rate": 3.4560096763307516e-07,
+      "loss": 0.887,
+      "step": 8228
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.85128967855895,
+      "learning_rate": 3.4496519120007344e-07,
+      "loss": 0.8312,
+      "step": 8229
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7695420264761204,
+      "learning_rate": 3.4432997920609e-07,
+      "loss": 0.8495,
+      "step": 8230
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8048646045718002,
+      "learning_rate": 3.4369533172814715e-07,
+      "loss": 1.0251,
+      "step": 8231
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8619569597291394,
+      "learning_rate": 3.4306124884319894e-07,
+      "loss": 0.9592,
+      "step": 8232
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6743000221666884,
+      "learning_rate": 3.424277306281282e-07,
+      "loss": 0.8287,
+      "step": 8233
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7458004827712252,
+      "learning_rate": 3.417947771597524e-07,
+      "loss": 0.9071,
+      "step": 8234
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.799389954025595,
+      "learning_rate": 3.411623885148202e-07,
+      "loss": 0.8221,
+      "step": 8235
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7014900791466614,
+      "learning_rate": 3.4053056477000856e-07,
+      "loss": 0.8628,
+      "step": 8236
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6772054513650267,
+      "learning_rate": 3.3989930600192953e-07,
+      "loss": 0.968,
+      "step": 8237
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7471942067528425,
+      "learning_rate": 3.3926861228712637e-07,
+      "loss": 0.9099,
+      "step": 8238
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6797339124180506,
+      "learning_rate": 3.386384837020723e-07,
+      "loss": 0.8827,
+      "step": 8239
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8036985537139214,
+      "learning_rate": 3.3800892032317123e-07,
+      "loss": 0.8261,
+      "step": 8240
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8903091362822204,
+      "learning_rate": 3.373799222267615e-07,
+      "loss": 0.9073,
+      "step": 8241
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7790311194577395,
+      "learning_rate": 3.3675148948910996e-07,
+      "loss": 0.9138,
+      "step": 8242
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8049835051729738,
+      "learning_rate": 3.3612362218641727e-07,
+      "loss": 1.0387,
+      "step": 8243
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7595528933411977,
+      "learning_rate": 3.354963203948147e-07,
+      "loss": 0.8983,
+      "step": 8244
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7930536151247514,
+      "learning_rate": 3.348695841903637e-07,
+      "loss": 0.872,
+      "step": 8245
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7480891087035033,
+      "learning_rate": 3.3424341364905853e-07,
+      "loss": 0.8826,
+      "step": 8246
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7658621787608952,
+      "learning_rate": 3.3361780884682615e-07,
+      "loss": 0.8241,
+      "step": 8247
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7903514472539686,
+      "learning_rate": 3.3299276985952034e-07,
+      "loss": 0.8953,
+      "step": 8248
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7932063030370908,
+      "learning_rate": 3.323682967629321e-07,
+      "loss": 0.9293,
+      "step": 8249
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9294572180963336,
+      "learning_rate": 3.317443896327788e-07,
+      "loss": 0.8946,
+      "step": 8250
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8020539495030725,
+      "learning_rate": 3.311210485447125e-07,
+      "loss": 0.9875,
+      "step": 8251
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8325587832327367,
+      "learning_rate": 3.3049827357431506e-07,
+      "loss": 1.0283,
+      "step": 8252
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7110551367060656,
+      "learning_rate": 3.2987606479709946e-07,
+      "loss": 0.8225,
+      "step": 8253
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7201780684385002,
+      "learning_rate": 3.2925442228851246e-07,
+      "loss": 0.9142,
+      "step": 8254
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.664246445006223,
+      "learning_rate": 3.286333461239288e-07,
+      "loss": 0.8967,
+      "step": 8255
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6771430242901224,
+      "learning_rate": 3.2801283637865613e-07,
+      "loss": 0.7718,
+      "step": 8256
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7287444116149755,
+      "learning_rate": 3.2739289312793467e-07,
+      "loss": 0.9445,
+      "step": 8257
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7586567174474266,
+      "learning_rate": 3.2677351644693323e-07,
+      "loss": 0.9085,
+      "step": 8258
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7149182479309786,
+      "learning_rate": 3.261547064107551e-07,
+      "loss": 0.8462,
+      "step": 8259
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7307791061981495,
+      "learning_rate": 3.2553646309443133e-07,
+      "loss": 0.8548,
+      "step": 8260
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5912068002473503,
+      "learning_rate": 3.2491878657292643e-07,
+      "loss": 0.8272,
+      "step": 8261
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7368098575044884,
+      "learning_rate": 3.24301676921136e-07,
+      "loss": 0.8853,
+      "step": 8262
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7770227770216724,
+      "learning_rate": 3.2368513421388745e-07,
+      "loss": 0.8144,
+      "step": 8263
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.783000880048568,
+      "learning_rate": 3.230691585259371e-07,
+      "loss": 0.9038,
+      "step": 8264
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7657973176558202,
+      "learning_rate": 3.2245374993197577e-07,
+      "loss": 0.8617,
+      "step": 8265
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7646293345475237,
+      "learning_rate": 3.218389085066237e-07,
+      "loss": 0.9276,
+      "step": 8266
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.745825437204345,
+      "learning_rate": 3.212246343244313e-07,
+      "loss": 0.9316,
+      "step": 8267
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7224294316372664,
+      "learning_rate": 3.2061092745988175e-07,
+      "loss": 0.8668,
+      "step": 8268
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7830915441469475,
+      "learning_rate": 3.1999778798739057e-07,
+      "loss": 1.061,
+      "step": 8269
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7398840659731118,
+      "learning_rate": 3.1938521598130213e-07,
+      "loss": 0.8412,
+      "step": 8270
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5959840269064922,
+      "learning_rate": 3.18773211515892e-07,
+      "loss": 0.8359,
+      "step": 8271
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7159229727459169,
+      "learning_rate": 3.181617746653687e-07,
+      "loss": 0.938,
+      "step": 8272
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7288830507637387,
+      "learning_rate": 3.1755090550387167e-07,
+      "loss": 0.8879,
+      "step": 8273
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7225100291805466,
+      "learning_rate": 3.169406041054696e-07,
+      "loss": 0.8606,
+      "step": 8274
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7013841561081522,
+      "learning_rate": 3.163308705441648e-07,
+      "loss": 0.8811,
+      "step": 8275
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6950553143608977,
+      "learning_rate": 3.1572170489388823e-07,
+      "loss": 0.8879,
+      "step": 8276
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6317523985256812,
+      "learning_rate": 3.151131072285041e-07,
+      "loss": 0.8789,
+      "step": 8277
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.781445641160241,
+      "learning_rate": 3.145050776218078e-07,
+      "loss": 0.9204,
+      "step": 8278
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7186599361799402,
+      "learning_rate": 3.1389761614752376e-07,
+      "loss": 0.9153,
+      "step": 8279
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7618082148956338,
+      "learning_rate": 3.132907228793086e-07,
+      "loss": 0.9007,
+      "step": 8280
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.768144975270962,
+      "learning_rate": 3.126843978907518e-07,
+      "loss": 0.9053,
+      "step": 8281
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.701342747693401,
+      "learning_rate": 3.1207864125537167e-07,
+      "loss": 0.9245,
+      "step": 8282
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7650422614169508,
+      "learning_rate": 3.1147345304661736e-07,
+      "loss": 0.927,
+      "step": 8283
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5649051054413041,
+      "learning_rate": 3.108688333378701e-07,
+      "loss": 0.8462,
+      "step": 8284
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6401365902808507,
+      "learning_rate": 3.1026478220244406e-07,
+      "loss": 0.8064,
+      "step": 8285
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.628356480846652,
+      "learning_rate": 3.0966129971358003e-07,
+      "loss": 0.7961,
+      "step": 8286
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7647937697238762,
+      "learning_rate": 3.090583859444535e-07,
+      "loss": 0.8838,
+      "step": 8287
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7059483862094622,
+      "learning_rate": 3.084560409681703e-07,
+      "loss": 0.8476,
+      "step": 8288
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.730200703057485,
+      "learning_rate": 3.078542648577659e-07,
+      "loss": 0.8784,
+      "step": 8289
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.750077237880868,
+      "learning_rate": 3.0725305768620815e-07,
+      "loss": 0.9091,
+      "step": 8290
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7395107309558624,
+      "learning_rate": 3.0665241952639704e-07,
+      "loss": 0.8836,
+      "step": 8291
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6398811306243006,
+      "learning_rate": 3.0605235045115867e-07,
+      "loss": 0.8735,
+      "step": 8292
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7090003297543044,
+      "learning_rate": 3.0545285053325604e-07,
+      "loss": 0.8159,
+      "step": 8293
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7567922594390677,
+      "learning_rate": 3.048539198453798e-07,
+      "loss": 0.9556,
+      "step": 8294
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6303668623500427,
+      "learning_rate": 3.04255558460152e-07,
+      "loss": 0.7677,
+      "step": 8295
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6858998984106728,
+      "learning_rate": 3.036577664501267e-07,
+      "loss": 0.9164,
+      "step": 8296
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.759862512496004,
+      "learning_rate": 3.030605438877882e-07,
+      "loss": 0.9276,
+      "step": 8297
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7814795410738065,
+      "learning_rate": 3.024638908455513e-07,
+      "loss": 0.8408,
+      "step": 8298
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6193404930908049,
+      "learning_rate": 3.0186780739576204e-07,
+      "loss": 0.8317,
+      "step": 8299
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6828976996261342,
+      "learning_rate": 3.012722936106993e-07,
+      "loss": 0.9197,
+      "step": 8300
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7360006352193476,
+      "learning_rate": 3.0067734956256867e-07,
+      "loss": 0.8787,
+      "step": 8301
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.716712007892586,
+      "learning_rate": 3.000829753235118e-07,
+      "loss": 0.8195,
+      "step": 8302
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7279039330397081,
+      "learning_rate": 2.9948917096559617e-07,
+      "loss": 0.9261,
+      "step": 8303
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8128127608369178,
+      "learning_rate": 2.9889593656082405e-07,
+      "loss": 0.9598,
+      "step": 8304
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7257820544096112,
+      "learning_rate": 2.983032721811263e-07,
+      "loss": 0.9341,
+      "step": 8305
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8916615514831667,
+      "learning_rate": 2.9771117789836713e-07,
+      "loss": 0.8638,
+      "step": 8306
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6746276576816008,
+      "learning_rate": 2.971196537843379e-07,
+      "loss": 0.8717,
+      "step": 8307
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.891322867628431,
+      "learning_rate": 2.9652869991076414e-07,
+      "loss": 0.9272,
+      "step": 8308
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6125276485623985,
+      "learning_rate": 2.9593831634930127e-07,
+      "loss": 0.8787,
+      "step": 8309
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5978542956039876,
+      "learning_rate": 2.953485031715342e-07,
+      "loss": 0.8887,
+      "step": 8310
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.701120392388444,
+      "learning_rate": 2.947592604489807e-07,
+      "loss": 0.9508,
+      "step": 8311
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6976429829658428,
+      "learning_rate": 2.9417058825308874e-07,
+      "loss": 0.8175,
+      "step": 8312
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9023824783139522,
+      "learning_rate": 2.935824866552367e-07,
+      "loss": 0.8653,
+      "step": 8313
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7769641107788798,
+      "learning_rate": 2.929949557267331e-07,
+      "loss": 1.0537,
+      "step": 8314
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6354062561497018,
+      "learning_rate": 2.9240799553881813e-07,
+      "loss": 0.8564,
+      "step": 8315
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.756939835427688,
+      "learning_rate": 2.918216061626644e-07,
+      "loss": 0.9193,
+      "step": 8316
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5930890217357605,
+      "learning_rate": 2.912357876693711e-07,
+      "loss": 0.9139,
+      "step": 8317
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7678403777636653,
+      "learning_rate": 2.9065054012997305e-07,
+      "loss": 0.8984,
+      "step": 8318
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7690569541035667,
+      "learning_rate": 2.900658636154324e-07,
+      "loss": 0.999,
+      "step": 8319
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7442246667689518,
+      "learning_rate": 2.894817581966436e-07,
+      "loss": 0.8686,
+      "step": 8320
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.784033595959471,
+      "learning_rate": 2.8889822394443047e-07,
+      "loss": 0.8895,
+      "step": 8321
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7301961162666224,
+      "learning_rate": 2.883152609295503e-07,
+      "loss": 0.8947,
+      "step": 8322
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8060890418602493,
+      "learning_rate": 2.8773286922268885e-07,
+      "loss": 0.9192,
+      "step": 8323
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5734727175261571,
+      "learning_rate": 2.871510488944612e-07,
+      "loss": 0.9217,
+      "step": 8324
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9352344545709945,
+      "learning_rate": 2.8656980001541766e-07,
+      "loss": 0.8876,
+      "step": 8325
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7989544039765482,
+      "learning_rate": 2.859891226560346e-07,
+      "loss": 1.0072,
+      "step": 8326
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7527492240774807,
+      "learning_rate": 2.854090168867224e-07,
+      "loss": 0.8461,
+      "step": 8327
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8067250440673923,
+      "learning_rate": 2.848294827778214e-07,
+      "loss": 0.9115,
+      "step": 8328
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7137661277835317,
+      "learning_rate": 2.8425052039959986e-07,
+      "loss": 0.8397,
+      "step": 8329
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6971533338129419,
+      "learning_rate": 2.836721298222611e-07,
+      "loss": 0.8713,
+      "step": 8330
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.746670294137461,
+      "learning_rate": 2.830943111159368e-07,
+      "loss": 0.8809,
+      "step": 8331
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7232542262427302,
+      "learning_rate": 2.8251706435068806e-07,
+      "loss": 0.8636,
+      "step": 8332
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7228696503616736,
+      "learning_rate": 2.8194038959650895e-07,
+      "loss": 0.9181,
+      "step": 8333
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7863185987799219,
+      "learning_rate": 2.813642869233235e-07,
+      "loss": 1.0583,
+      "step": 8334
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7060508284130371,
+      "learning_rate": 2.8078875640098646e-07,
+      "loss": 0.9059,
+      "step": 8335
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6559340539382477,
+      "learning_rate": 2.8021379809928137e-07,
+      "loss": 0.9174,
+      "step": 8336
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7585103187938036,
+      "learning_rate": 2.796394120879259e-07,
+      "loss": 0.91,
+      "step": 8337
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.8612956859620318,
+      "learning_rate": 2.790655984365642e-07,
+      "loss": 0.8718,
+      "step": 8338
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.651371288112955,
+      "learning_rate": 2.7849235721477407e-07,
+      "loss": 0.9186,
+      "step": 8339
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7312306160003528,
+      "learning_rate": 2.779196884920643e-07,
+      "loss": 0.7828,
+      "step": 8340
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7241040755330146,
+      "learning_rate": 2.7734759233787046e-07,
+      "loss": 0.8543,
+      "step": 8341
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6961230174611481,
+      "learning_rate": 2.7677606882156314e-07,
+      "loss": 0.898,
+      "step": 8342
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7704046932855195,
+      "learning_rate": 2.762051180124414e-07,
+      "loss": 0.8781,
+      "step": 8343
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7873476692264727,
+      "learning_rate": 2.7563473997973434e-07,
+      "loss": 0.8949,
+      "step": 8344
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8984121809831151,
+      "learning_rate": 2.750649347926021e-07,
+      "loss": 0.8325,
+      "step": 8345
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7712463258793013,
+      "learning_rate": 2.7449570252013557e-07,
+      "loss": 1.0244,
+      "step": 8346
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7477946343504673,
+      "learning_rate": 2.7392704323135676e-07,
+      "loss": 0.9346,
+      "step": 8347
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8349171715912473,
+      "learning_rate": 2.733589569952172e-07,
+      "loss": 0.8337,
+      "step": 8348
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6531814841103745,
+      "learning_rate": 2.7279144388059895e-07,
+      "loss": 0.8791,
+      "step": 8349
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6560968214091845,
+      "learning_rate": 2.7222450395631593e-07,
+      "loss": 0.8324,
+      "step": 8350
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.770585030882444,
+      "learning_rate": 2.7165813729111035e-07,
+      "loss": 0.9123,
+      "step": 8351
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9754424032871574,
+      "learning_rate": 2.7109234395365666e-07,
+      "loss": 0.8646,
+      "step": 8352
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7730158791832826,
+      "learning_rate": 2.705271240125601e-07,
+      "loss": 0.8872,
+      "step": 8353
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7241171443052736,
+      "learning_rate": 2.699624775363541e-07,
+      "loss": 0.8521,
+      "step": 8354
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8021656756133565,
+      "learning_rate": 2.6939840459350497e-07,
+      "loss": 1.0402,
+      "step": 8355
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8295368356647344,
+      "learning_rate": 2.688349052524081e-07,
+      "loss": 0.8653,
+      "step": 8356
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8299443167732117,
+      "learning_rate": 2.6827197958138927e-07,
+      "loss": 0.789,
+      "step": 8357
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6279266773442387,
+      "learning_rate": 2.6770962764870625e-07,
+      "loss": 0.879,
+      "step": 8358
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6396356738048945,
+      "learning_rate": 2.6714784952254546e-07,
+      "loss": 0.8332,
+      "step": 8359
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8739378759025611,
+      "learning_rate": 2.665866452710242e-07,
+      "loss": 0.8668,
+      "step": 8360
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7274792989861147,
+      "learning_rate": 2.6602601496219073e-07,
+      "loss": 0.8622,
+      "step": 8361
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8215880602488403,
+      "learning_rate": 2.65465958664024e-07,
+      "loss": 1.0652,
+      "step": 8362
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8149898291187118,
+      "learning_rate": 2.6490647644443145e-07,
+      "loss": 1.0473,
+      "step": 8363
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6967209402491683,
+      "learning_rate": 2.643475683712532e-07,
+      "loss": 0.9706,
+      "step": 8364
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8754854217414039,
+      "learning_rate": 2.6378923451225893e-07,
+      "loss": 0.795,
+      "step": 8365
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7018712744058277,
+      "learning_rate": 2.632314749351483e-07,
+      "loss": 0.9416,
+      "step": 8366
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8964919139136545,
+      "learning_rate": 2.6267428970755124e-07,
+      "loss": 0.9684,
+      "step": 8367
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7738727748780305,
+      "learning_rate": 2.6211767889702813e-07,
+      "loss": 0.868,
+      "step": 8368
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.576380069339271,
+      "learning_rate": 2.61561642571071e-07,
+      "loss": 0.94,
+      "step": 8369
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7461751968087937,
+      "learning_rate": 2.610061807971004e-07,
+      "loss": 0.8751,
+      "step": 8370
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7166950632434579,
+      "learning_rate": 2.6045129364246857e-07,
+      "loss": 0.8049,
+      "step": 8371
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7769283108857284,
+      "learning_rate": 2.598969811744562e-07,
+      "loss": 1.0655,
+      "step": 8372
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7150001076598882,
+      "learning_rate": 2.593432434602766e-07,
+      "loss": 0.9434,
+      "step": 8373
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.672542349634942,
+      "learning_rate": 2.5879008056707286e-07,
+      "loss": 0.9167,
+      "step": 8374
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7018201318770547,
+      "learning_rate": 2.5823749256191686e-07,
+      "loss": 0.824,
+      "step": 8375
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7319519380302282,
+      "learning_rate": 2.5768547951181276e-07,
+      "loss": 0.9228,
+      "step": 8376
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.653157083413519,
+      "learning_rate": 2.5713404148369316e-07,
+      "loss": 0.8738,
+      "step": 8377
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9419672351449841,
+      "learning_rate": 2.5658317854442283e-07,
+      "loss": 0.9794,
+      "step": 8378
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7488285223565594,
+      "learning_rate": 2.56032890760794e-07,
+      "loss": 0.8887,
+      "step": 8379
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7065456769529799,
+      "learning_rate": 2.5548317819953205e-07,
+      "loss": 0.9228,
+      "step": 8380
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7470408461025857,
+      "learning_rate": 2.5493404092729267e-07,
+      "loss": 0.9296,
+      "step": 8381
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8351294771186066,
+      "learning_rate": 2.543854790106587e-07,
+      "loss": 0.8794,
+      "step": 8382
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.719370269714994,
+      "learning_rate": 2.538374925161463e-07,
+      "loss": 0.8651,
+      "step": 8383
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7963835562171078,
+      "learning_rate": 2.5329008151020076e-07,
+      "loss": 1.06,
+      "step": 8384
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8770621158887415,
+      "learning_rate": 2.5274324605919664e-07,
+      "loss": 0.9537,
+      "step": 8385
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9152526542136643,
+      "learning_rate": 2.5219698622944044e-07,
+      "loss": 0.9409,
+      "step": 8386
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7078238432815724,
+      "learning_rate": 2.5165130208716913e-07,
+      "loss": 0.9924,
+      "step": 8387
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6103081463452098,
+      "learning_rate": 2.5110619369854594e-07,
+      "loss": 0.8259,
+      "step": 8388
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.918953951719248,
+      "learning_rate": 2.505616611296685e-07,
+      "loss": 0.8913,
+      "step": 8389
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7893580895071342,
+      "learning_rate": 2.500177044465646e-07,
+      "loss": 0.8251,
+      "step": 8390
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6967829828997167,
+      "learning_rate": 2.494743237151892e-07,
+      "loss": 0.822,
+      "step": 8391
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9384750155627701,
+      "learning_rate": 2.489315190014291e-07,
+      "loss": 0.8577,
+      "step": 8392
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6954704728568617,
+      "learning_rate": 2.483892903711027e-07,
+      "loss": 0.8249,
+      "step": 8393
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.795808404921678,
+      "learning_rate": 2.4784763788995527e-07,
+      "loss": 0.9206,
+      "step": 8394
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7046670857868997,
+      "learning_rate": 2.473065616236647e-07,
+      "loss": 0.846,
+      "step": 8395
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.720535573088482,
+      "learning_rate": 2.467660616378398e-07,
+      "loss": 0.9594,
+      "step": 8396
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7085437143882398,
+      "learning_rate": 2.462261379980163e-07,
+      "loss": 0.925,
+      "step": 8397
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7935701679262799,
+      "learning_rate": 2.4568679076966197e-07,
+      "loss": 0.8651,
+      "step": 8398
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.719374768746955,
+      "learning_rate": 2.451480200181744e-07,
+      "loss": 0.8945,
+      "step": 8399
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7209782763431611,
+      "learning_rate": 2.44609825808883e-07,
+      "loss": 0.891,
+      "step": 8400
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7408933591433935,
+      "learning_rate": 2.44072208207044e-07,
+      "loss": 0.9241,
+      "step": 8401
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7867322790798434,
+      "learning_rate": 2.435351672778463e-07,
+      "loss": 0.8921,
+      "step": 8402
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7211992872557857,
+      "learning_rate": 2.4299870308640726e-07,
+      "loss": 0.8551,
+      "step": 8403
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6660130091027334,
+      "learning_rate": 2.4246281569777485e-07,
+      "loss": 0.8792,
+      "step": 8404
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7707080399922959,
+      "learning_rate": 2.4192750517692877e-07,
+      "loss": 0.8595,
+      "step": 8405
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8222771568290337,
+      "learning_rate": 2.4139277158877537e-07,
+      "loss": 0.8393,
+      "step": 8406
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.741653652942318,
+      "learning_rate": 2.40858614998154e-07,
+      "loss": 0.8889,
+      "step": 8407
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7221904548262505,
+      "learning_rate": 2.403250354698333e-07,
+      "loss": 0.8019,
+      "step": 8408
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.693263474033574,
+      "learning_rate": 2.3979203306851165e-07,
+      "loss": 0.796,
+      "step": 8409
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7828284411830166,
+      "learning_rate": 2.3925960785881554e-07,
+      "loss": 0.934,
+      "step": 8410
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6880230606022133,
+      "learning_rate": 2.3872775990530505e-07,
+      "loss": 0.8203,
+      "step": 8411
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8002601522633312,
+      "learning_rate": 2.381964892724692e-07,
+      "loss": 0.8468,
+      "step": 8412
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7974549445599064,
+      "learning_rate": 2.3766579602472474e-07,
+      "loss": 0.9479,
+      "step": 8413
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.752061782235739,
+      "learning_rate": 2.3713568022642018e-07,
+      "loss": 0.8755,
+      "step": 8414
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.704083552680782,
+      "learning_rate": 2.3660614194183584e-07,
+      "loss": 0.9732,
+      "step": 8415
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7198376069406256,
+      "learning_rate": 2.3607718123517753e-07,
+      "loss": 0.9093,
+      "step": 8416
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6589696253128774,
+      "learning_rate": 2.3554879817058507e-07,
+      "loss": 0.8737,
+      "step": 8417
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7727194899693068,
+      "learning_rate": 2.3502099281212775e-07,
+      "loss": 0.8379,
+      "step": 8418
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6256817978482674,
+      "learning_rate": 2.3449376522380106e-07,
+      "loss": 0.8583,
+      "step": 8419
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7510024964021975,
+      "learning_rate": 2.3396711546953444e-07,
+      "loss": 0.8703,
+      "step": 8420
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8269968374734757,
+      "learning_rate": 2.3344104361318677e-07,
+      "loss": 0.8256,
+      "step": 8421
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8023354534019362,
+      "learning_rate": 2.329155497185448e-07,
+      "loss": 1.0808,
+      "step": 8422
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7643709267344734,
+      "learning_rate": 2.3239063384932703e-07,
+      "loss": 1.0539,
+      "step": 8423
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.721198422373936,
+      "learning_rate": 2.3186629606918198e-07,
+      "loss": 0.8703,
+      "step": 8424
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7807718538188582,
+      "learning_rate": 2.3134253644168604e-07,
+      "loss": 0.9133,
+      "step": 8425
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7074900577158572,
+      "learning_rate": 2.308193550303478e-07,
+      "loss": 0.9227,
+      "step": 8426
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7204427297591909,
+      "learning_rate": 2.3029675189860546e-07,
+      "loss": 0.9105,
+      "step": 8427
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6632258403211164,
+      "learning_rate": 2.297747271098244e-07,
+      "loss": 0.9132,
+      "step": 8428
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7332858661318864,
+      "learning_rate": 2.2925328072730402e-07,
+      "loss": 0.8888,
+      "step": 8429
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7260645347112678,
+      "learning_rate": 2.287324128142704e-07,
+      "loss": 0.9496,
+      "step": 8430
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6612688491627783,
+      "learning_rate": 2.2821212343388077e-07,
+      "loss": 0.9502,
+      "step": 8431
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.678469187492468,
+      "learning_rate": 2.2769241264922194e-07,
+      "loss": 0.8257,
+      "step": 8432
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7129464758033428,
+      "learning_rate": 2.2717328052331122e-07,
+      "loss": 0.8725,
+      "step": 8433
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7036683279732037,
+      "learning_rate": 2.2665472711909387e-07,
+      "loss": 0.9064,
+      "step": 8434
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7184377785281197,
+      "learning_rate": 2.2613675249944677e-07,
+      "loss": 0.8989,
+      "step": 8435
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8476081155428064,
+      "learning_rate": 2.256193567271775e-07,
+      "loss": 0.9201,
+      "step": 8436
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6252869184458472,
+      "learning_rate": 2.251025398650203e-07,
+      "loss": 0.943,
+      "step": 8437
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.832413032139026,
+      "learning_rate": 2.2458630197564225e-07,
+      "loss": 0.9019,
+      "step": 8438
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6138752285384295,
+      "learning_rate": 2.240706431216383e-07,
+      "loss": 0.8514,
+      "step": 8439
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7054899909506267,
+      "learning_rate": 2.2355556336553452e-07,
+      "loss": 0.8949,
+      "step": 8440
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6680582260049517,
+      "learning_rate": 2.2304106276978432e-07,
+      "loss": 0.8939,
+      "step": 8441
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7222854413360502,
+      "learning_rate": 2.2252714139677445e-07,
+      "loss": 0.9155,
+      "step": 8442
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7430794003572796,
+      "learning_rate": 2.2201379930882006e-07,
+      "loss": 0.871,
+      "step": 8443
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7575622928383645,
+      "learning_rate": 2.2150103656816357e-07,
+      "loss": 0.9337,
+      "step": 8444
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7176389137049595,
+      "learning_rate": 2.2098885323698027e-07,
+      "loss": 0.8712,
+      "step": 8445
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.910411707725528,
+      "learning_rate": 2.2047724937737547e-07,
+      "loss": 0.8774,
+      "step": 8446
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.695132597174356,
+      "learning_rate": 2.1996622505138064e-07,
+      "loss": 0.8792,
+      "step": 8447
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8518833280298195,
+      "learning_rate": 2.1945578032096016e-07,
+      "loss": 0.9209,
+      "step": 8448
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8160747466788103,
+      "learning_rate": 2.1894591524800835e-07,
+      "loss": 0.9423,
+      "step": 8449
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7455418439040133,
+      "learning_rate": 2.1843662989434687e-07,
+      "loss": 0.8595,
+      "step": 8450
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6669567059849877,
+      "learning_rate": 2.179279243217286e-07,
+      "loss": 0.912,
+      "step": 8451
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6944599810927288,
+      "learning_rate": 2.1741979859183582e-07,
+      "loss": 0.9169,
+      "step": 8452
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6891837424699985,
+      "learning_rate": 2.169122527662798e-07,
+      "loss": 0.9134,
+      "step": 8453
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7923096002992829,
+      "learning_rate": 2.1640528690660302e-07,
+      "loss": 0.9626,
+      "step": 8454
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.703668965755326,
+      "learning_rate": 2.158989010742779e-07,
+      "loss": 0.8634,
+      "step": 8455
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.793890274039902,
+      "learning_rate": 2.1539309533070318e-07,
+      "loss": 0.9155,
+      "step": 8456
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6939145646838638,
+      "learning_rate": 2.1488786973721088e-07,
+      "loss": 0.8834,
+      "step": 8457
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7453927219303724,
+      "learning_rate": 2.1438322435506198e-07,
+      "loss": 0.9499,
+      "step": 8458
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9635421569272278,
+      "learning_rate": 2.138791592454448e-07,
+      "loss": 0.9039,
+      "step": 8459
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7649214325215867,
+      "learning_rate": 2.1337567446947982e-07,
+      "loss": 0.8305,
+      "step": 8460
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8188915626847422,
+      "learning_rate": 2.1287277008821662e-07,
+      "loss": 0.9119,
+      "step": 8461
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.747092580848208,
+      "learning_rate": 2.1237044616263413e-07,
+      "loss": 0.8697,
+      "step": 8462
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.675739433841438,
+      "learning_rate": 2.118687027536398e-07,
+      "loss": 0.8252,
+      "step": 8463
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8392976662828129,
+      "learning_rate": 2.113675399220727e-07,
+      "loss": 0.928,
+      "step": 8464
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7327069080525423,
+      "learning_rate": 2.1086695772870092e-07,
+      "loss": 0.8857,
+      "step": 8465
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8065015711432533,
+      "learning_rate": 2.1036695623422033e-07,
+      "loss": 1.0524,
+      "step": 8466
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6886803154044214,
+      "learning_rate": 2.0986753549925963e-07,
+      "loss": 0.8886,
+      "step": 8467
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7236494289751492,
+      "learning_rate": 2.0936869558437322e-07,
+      "loss": 0.8714,
+      "step": 8468
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8117039882009591,
+      "learning_rate": 2.0887043655004825e-07,
+      "loss": 0.9404,
+      "step": 8469
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6567330758464889,
+      "learning_rate": 2.0837275845670135e-07,
+      "loss": 0.8387,
+      "step": 8470
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.713941943336574,
+      "learning_rate": 2.0787566136467708e-07,
+      "loss": 0.8128,
+      "step": 8471
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7699394482941314,
+      "learning_rate": 2.0737914533424885e-07,
+      "loss": 1.0469,
+      "step": 8472
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7326594808579814,
+      "learning_rate": 2.0688321042562186e-07,
+      "loss": 0.844,
+      "step": 8473
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7034135984118404,
+      "learning_rate": 2.0638785669893026e-07,
+      "loss": 0.8652,
+      "step": 8474
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6692342093247456,
+      "learning_rate": 2.0589308421423704e-07,
+      "loss": 0.8674,
+      "step": 8475
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.710778118295827,
+      "learning_rate": 2.053988930315348e-07,
+      "loss": 0.8904,
+      "step": 8476
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7821696875223789,
+      "learning_rate": 2.0490528321074732e-07,
+      "loss": 1.0428,
+      "step": 8477
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6976011601590875,
+      "learning_rate": 2.0441225481172444e-07,
+      "loss": 0.9578,
+      "step": 8478
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7993345218138387,
+      "learning_rate": 2.0391980789424892e-07,
+      "loss": 1.0652,
+      "step": 8479
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.682202090907931,
+      "learning_rate": 2.034279425180319e-07,
+      "loss": 0.9084,
+      "step": 8480
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.753365719757043,
+      "learning_rate": 2.0293665874271228e-07,
+      "loss": 0.8736,
+      "step": 8481
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7986869587500376,
+      "learning_rate": 2.0244595662786138e-07,
+      "loss": 0.8739,
+      "step": 8482
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6918596011043319,
+      "learning_rate": 2.019558362329782e-07,
+      "loss": 0.8443,
+      "step": 8483
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.756852160939759,
+      "learning_rate": 2.014662976174908e-07,
+      "loss": 0.865,
+      "step": 8484
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7422059397927088,
+      "learning_rate": 2.0097734084075726e-07,
+      "loss": 0.8756,
+      "step": 8485
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7333984969005671,
+      "learning_rate": 2.004889659620668e-07,
+      "loss": 0.8575,
+      "step": 8486
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8960209357448137,
+      "learning_rate": 2.0000117304063483e-07,
+      "loss": 0.8923,
+      "step": 8487
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7626632707194032,
+      "learning_rate": 1.9951396213560904e-07,
+      "loss": 0.9281,
+      "step": 8488
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7132937473455472,
+      "learning_rate": 1.9902733330606605e-07,
+      "loss": 0.8907,
+      "step": 8489
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7014448067685382,
+      "learning_rate": 1.9854128661100924e-07,
+      "loss": 0.7954,
+      "step": 8490
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7639021809221131,
+      "learning_rate": 1.980558221093748e-07,
+      "loss": 0.9172,
+      "step": 8491
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.806218603040422,
+      "learning_rate": 1.9757093986002729e-07,
+      "loss": 0.8525,
+      "step": 8492
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6727564552144012,
+      "learning_rate": 1.9708663992175969e-07,
+      "loss": 0.881,
+      "step": 8493
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6952936396284455,
+      "learning_rate": 1.966029223532945e-07,
+      "loss": 0.8536,
+      "step": 8494
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6982785184908866,
+      "learning_rate": 1.9611978721328473e-07,
+      "loss": 0.8891,
+      "step": 8495
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7711968866501198,
+      "learning_rate": 1.9563723456031303e-07,
+      "loss": 0.8332,
+      "step": 8496
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6929181513787197,
+      "learning_rate": 1.9515526445288923e-07,
+      "loss": 0.9622,
+      "step": 8497
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.1043061295119343,
+      "learning_rate": 1.946738769494555e-07,
+      "loss": 0.8803,
+      "step": 8498
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6806616032077493,
+      "learning_rate": 1.9419307210837955e-07,
+      "loss": 0.8256,
+      "step": 8499
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.730238576405702,
+      "learning_rate": 1.9371284998796146e-07,
+      "loss": 0.8754,
+      "step": 8500
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.792395664226214,
+      "learning_rate": 1.9323321064643131e-07,
+      "loss": 0.7976,
+      "step": 8501
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7664670932858075,
+      "learning_rate": 1.9275415414194476e-07,
+      "loss": 1.0223,
+      "step": 8502
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7972478238874077,
+      "learning_rate": 1.9227568053259094e-07,
+      "loss": 1.0165,
+      "step": 8503
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7192192013580099,
+      "learning_rate": 1.9179778987638508e-07,
+      "loss": 0.8467,
+      "step": 8504
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7802727537977003,
+      "learning_rate": 1.913204822312742e-07,
+      "loss": 0.902,
+      "step": 8505
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7024995147397617,
+      "learning_rate": 1.9084375765513197e-07,
+      "loss": 0.9046,
+      "step": 8506
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7209341527423896,
+      "learning_rate": 1.9036761620576438e-07,
+      "loss": 0.8791,
+      "step": 8507
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8096865620255889,
+      "learning_rate": 1.898920579409047e-07,
+      "loss": 0.8976,
+      "step": 8508
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7725630223497932,
+      "learning_rate": 1.894170829182157e-07,
+      "loss": 0.8603,
+      "step": 8509
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6244322786712522,
+      "learning_rate": 1.8894269119528963e-07,
+      "loss": 0.9762,
+      "step": 8510
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6681300451454673,
+      "learning_rate": 1.8846888282964936e-07,
+      "loss": 0.9358,
+      "step": 8511
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6045761553831774,
+      "learning_rate": 1.8799565787874397e-07,
+      "loss": 0.8531,
+      "step": 8512
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7400974822927735,
+      "learning_rate": 1.8752301639995528e-07,
+      "loss": 1.0396,
+      "step": 8513
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.728215221601243,
+      "learning_rate": 1.8705095845059252e-07,
+      "loss": 0.8429,
+      "step": 8514
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7616294074481065,
+      "learning_rate": 1.8657948408789261e-07,
+      "loss": 0.9534,
+      "step": 8515
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7567945222315553,
+      "learning_rate": 1.8610859336902488e-07,
+      "loss": 0.9328,
+      "step": 8516
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.753965621669419,
+      "learning_rate": 1.8563828635108694e-07,
+      "loss": 0.8379,
+      "step": 8517
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7738314017867929,
+      "learning_rate": 1.8516856309110375e-07,
+      "loss": 1.0479,
+      "step": 8518
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7415406504982403,
+      "learning_rate": 1.8469942364603145e-07,
+      "loss": 0.9295,
+      "step": 8519
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.675114518750102,
+      "learning_rate": 1.8423086807275558e-07,
+      "loss": 0.9131,
+      "step": 8520
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7979009052966526,
+      "learning_rate": 1.8376289642808854e-07,
+      "loss": 0.8788,
+      "step": 8521
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6164700638480836,
+      "learning_rate": 1.8329550876877488e-07,
+      "loss": 0.8423,
+      "step": 8522
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.88115533334524,
+      "learning_rate": 1.8282870515148654e-07,
+      "loss": 0.8835,
+      "step": 8523
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7723357516652573,
+      "learning_rate": 1.8236248563282545e-07,
+      "loss": 1.0316,
+      "step": 8524
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7045642686426519,
+      "learning_rate": 1.8189685026932137e-07,
+      "loss": 0.9114,
+      "step": 8525
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7512391285362112,
+      "learning_rate": 1.8143179911743415e-07,
+      "loss": 0.9187,
+      "step": 8526
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7594527495060834,
+      "learning_rate": 1.8096733223355477e-07,
+      "loss": 1.0647,
+      "step": 8527
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6821180999355072,
+      "learning_rate": 1.805034496739988e-07,
+      "loss": 0.9178,
+      "step": 8528
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6517212399609746,
+      "learning_rate": 1.8004015149501564e-07,
+      "loss": 0.8735,
+      "step": 8529
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7777244297333116,
+      "learning_rate": 1.795774377527809e-07,
+      "loss": 1.072,
+      "step": 8530
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7602072672588146,
+      "learning_rate": 1.7911530850339976e-07,
+      "loss": 1.042,
+      "step": 8531
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8446555072881352,
+      "learning_rate": 1.7865376380290845e-07,
+      "loss": 0.8834,
+      "step": 8532
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7885163155853431,
+      "learning_rate": 1.7819280370726943e-07,
+      "loss": 1.028,
+      "step": 8533
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7696796241308165,
+      "learning_rate": 1.7773242827237635e-07,
+      "loss": 1.0395,
+      "step": 8534
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.665116675871728,
+      "learning_rate": 1.7727263755405178e-07,
+      "loss": 0.9253,
+      "step": 8535
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.018858750091648,
+      "learning_rate": 1.7681343160804608e-07,
+      "loss": 0.8293,
+      "step": 8536
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.812482035625714,
+      "learning_rate": 1.7635481049003979e-07,
+      "loss": 1.0238,
+      "step": 8537
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6641465448501882,
+      "learning_rate": 1.7589677425564222e-07,
+      "loss": 0.9254,
+      "step": 8538
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7543591172854383,
+      "learning_rate": 1.7543932296039234e-07,
+      "loss": 0.9038,
+      "step": 8539
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8754535017864549,
+      "learning_rate": 1.7498245665975689e-07,
+      "loss": 0.8137,
+      "step": 8540
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6411972501537255,
+      "learning_rate": 1.7452617540913265e-07,
+      "loss": 0.8819,
+      "step": 8541
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7660797706993099,
+      "learning_rate": 1.7407047926384646e-07,
+      "loss": 0.9021,
+      "step": 8542
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7731319376752432,
+      "learning_rate": 1.7361536827915138e-07,
+      "loss": 0.8874,
+      "step": 8543
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6959005306150499,
+      "learning_rate": 1.7316084251023214e-07,
+      "loss": 0.9626,
+      "step": 8544
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6749510753422585,
+      "learning_rate": 1.727069020122024e-07,
+      "loss": 0.9228,
+      "step": 8545
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7582386757157733,
+      "learning_rate": 1.722535468401021e-07,
+      "loss": 0.8895,
+      "step": 8546
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6423586996199162,
+      "learning_rate": 1.7180077704890274e-07,
+      "loss": 0.911,
+      "step": 8547
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8098696594569355,
+      "learning_rate": 1.7134859269350546e-07,
+      "loss": 0.9774,
+      "step": 8548
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.643070687851437,
+      "learning_rate": 1.7089699382873748e-07,
+      "loss": 0.8728,
+      "step": 8549
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7424715862674363,
+      "learning_rate": 1.7044598050935722e-07,
+      "loss": 0.8723,
+      "step": 8550
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7171669083927323,
+      "learning_rate": 1.6999555279005264e-07,
+      "loss": 0.8998,
+      "step": 8551
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7666861252827353,
+      "learning_rate": 1.695457107254378e-07,
+      "loss": 0.8664,
+      "step": 8552
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.803948347648774,
+      "learning_rate": 1.6909645437005906e-07,
+      "loss": 0.866,
+      "step": 8553
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7417987067852403,
+      "learning_rate": 1.6864778377839064e-07,
+      "loss": 0.8685,
+      "step": 8554
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7885228084323712,
+      "learning_rate": 1.6819969900483347e-07,
+      "loss": 0.9358,
+      "step": 8555
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.748405721495499,
+      "learning_rate": 1.6775220010372185e-07,
+      "loss": 0.8232,
+      "step": 8556
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6992602968009074,
+      "learning_rate": 1.6730528712931404e-07,
+      "loss": 0.9112,
+      "step": 8557
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.722812051708577,
+      "learning_rate": 1.6685896013580226e-07,
+      "loss": 0.8543,
+      "step": 8558
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7436185208111048,
+      "learning_rate": 1.6641321917730268e-07,
+      "loss": 0.8767,
+      "step": 8559
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6559671466711385,
+      "learning_rate": 1.6596806430786428e-07,
+      "loss": 0.8702,
+      "step": 8560
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8175757998681108,
+      "learning_rate": 1.6552349558146442e-07,
+      "loss": 0.8718,
+      "step": 8561
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9094451019551455,
+      "learning_rate": 1.650795130520072e-07,
+      "loss": 0.834,
+      "step": 8562
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6691151087680551,
+      "learning_rate": 1.6463611677332792e-07,
+      "loss": 0.8769,
+      "step": 8563
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8929159848740218,
+      "learning_rate": 1.6419330679918855e-07,
+      "loss": 0.9367,
+      "step": 8564
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7709278948730423,
+      "learning_rate": 1.6375108318328282e-07,
+      "loss": 0.9334,
+      "step": 8565
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9020729073538922,
+      "learning_rate": 1.633094459792317e-07,
+      "loss": 0.9192,
+      "step": 8566
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7258142148819962,
+      "learning_rate": 1.6286839524058462e-07,
+      "loss": 0.9369,
+      "step": 8567
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6046401197121125,
+      "learning_rate": 1.6242793102082043e-07,
+      "loss": 0.8619,
+      "step": 8568
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7651653168311088,
+      "learning_rate": 1.6198805337334755e-07,
+      "loss": 0.9917,
+      "step": 8569
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.606156060285631,
+      "learning_rate": 1.6154876235150273e-07,
+      "loss": 0.9296,
+      "step": 8570
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7712999735482209,
+      "learning_rate": 1.6111005800855063e-07,
+      "loss": 1.009,
+      "step": 8571
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.714640988178553,
+      "learning_rate": 1.6067194039768585e-07,
+      "loss": 0.8256,
+      "step": 8572
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8085848849985176,
+      "learning_rate": 1.6023440957203262e-07,
+      "loss": 0.9058,
+      "step": 8573
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.741224452877189,
+      "learning_rate": 1.5979746558464237e-07,
+      "loss": 0.7891,
+      "step": 8574
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.910881743105309,
+      "learning_rate": 1.593611084884955e-07,
+      "loss": 0.9369,
+      "step": 8575
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6899516118649593,
+      "learning_rate": 1.589253383365036e-07,
+      "loss": 0.822,
+      "step": 8576
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.742788046430878,
+      "learning_rate": 1.584901551815038e-07,
+      "loss": 0.8711,
+      "step": 8577
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6805741548736162,
+      "learning_rate": 1.5805555907626336e-07,
+      "loss": 0.8182,
+      "step": 8578
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6785250935364462,
+      "learning_rate": 1.5762155007347957e-07,
+      "loss": 0.881,
+      "step": 8579
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7208351744492294,
+      "learning_rate": 1.57188128225777e-07,
+      "loss": 0.8583,
+      "step": 8580
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.885055618111959,
+      "learning_rate": 1.5675529358570918e-07,
+      "loss": 0.9109,
+      "step": 8581
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6745257921811123,
+      "learning_rate": 1.5632304620575965e-07,
+      "loss": 0.8848,
+      "step": 8582
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5965811298179093,
+      "learning_rate": 1.5589138613833933e-07,
+      "loss": 0.8847,
+      "step": 8583
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6985480216095172,
+      "learning_rate": 1.5546031343578793e-07,
+      "loss": 0.7977,
+      "step": 8584
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7027249039381507,
+      "learning_rate": 1.5502982815037648e-07,
+      "loss": 0.9643,
+      "step": 8585
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7083201196376805,
+      "learning_rate": 1.5459993033430043e-07,
+      "loss": 0.8199,
+      "step": 8586
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6519003964403711,
+      "learning_rate": 1.5417062003968808e-07,
+      "loss": 0.9287,
+      "step": 8587
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6143944983311755,
+      "learning_rate": 1.537418973185939e-07,
+      "loss": 0.7871,
+      "step": 8588
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.1449883088908543,
+      "learning_rate": 1.5331376222300297e-07,
+      "loss": 0.9641,
+      "step": 8589
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7421187784891359,
+      "learning_rate": 1.5288621480482658e-07,
+      "loss": 0.9642,
+      "step": 8590
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8935650249451195,
+      "learning_rate": 1.524592551159071e-07,
+      "loss": 0.9316,
+      "step": 8591
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8363662934056069,
+      "learning_rate": 1.520328832080159e-07,
+      "loss": 0.8487,
+      "step": 8592
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7115228840756376,
+      "learning_rate": 1.5160709913285e-07,
+      "loss": 0.8388,
+      "step": 8593
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7042478061518613,
+      "learning_rate": 1.5118190294203915e-07,
+      "loss": 0.8898,
+      "step": 8594
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6773195577168234,
+      "learning_rate": 1.5075729468713829e-07,
+      "loss": 0.9223,
+      "step": 8595
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7626343854936184,
+      "learning_rate": 1.5033327441963398e-07,
+      "loss": 0.892,
+      "step": 8596
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.647664347513808,
+      "learning_rate": 1.4990984219093897e-07,
+      "loss": 0.882,
+      "step": 8597
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7576961355777236,
+      "learning_rate": 1.4948699805239718e-07,
+      "loss": 0.8661,
+      "step": 8598
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.843885358596449,
+      "learning_rate": 1.490647420552782e-07,
+      "loss": 0.9258,
+      "step": 8599
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7791036932392247,
+      "learning_rate": 1.486430742507833e-07,
+      "loss": 1.038,
+      "step": 8600
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.611416654882545,
+      "learning_rate": 1.4822199469004094e-07,
+      "loss": 0.8511,
+      "step": 8601
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7663359506786613,
+      "learning_rate": 1.4780150342410814e-07,
+      "loss": 0.8642,
+      "step": 8602
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.248257448631594,
+      "learning_rate": 1.473816005039713e-07,
+      "loss": 0.8756,
+      "step": 8603
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.664191794382842,
+      "learning_rate": 1.4696228598054473e-07,
+      "loss": 0.926,
+      "step": 8604
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.781773582720752,
+      "learning_rate": 1.4654355990467217e-07,
+      "loss": 1.0483,
+      "step": 8605
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.687709807486767,
+      "learning_rate": 1.4612542232712522e-07,
+      "loss": 0.9152,
+      "step": 8606
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.713970020227116,
+      "learning_rate": 1.4570787329860504e-07,
+      "loss": 0.8974,
+      "step": 8607
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8031453211671327,
+      "learning_rate": 1.4529091286973994e-07,
+      "loss": 0.9181,
+      "step": 8608
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7843435250796245,
+      "learning_rate": 1.44874541091089e-07,
+      "loss": 0.9556,
+      "step": 8609
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8064631469895245,
+      "learning_rate": 1.444587580131379e-07,
+      "loss": 0.8684,
+      "step": 8610
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7298369812512413,
+      "learning_rate": 1.4404356368630134e-07,
+      "loss": 0.8653,
+      "step": 8611
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7255770693568269,
+      "learning_rate": 1.4362895816092403e-07,
+      "loss": 0.8636,
+      "step": 8612
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8130630922670764,
+      "learning_rate": 1.4321494148727855e-07,
+      "loss": 1.0349,
+      "step": 8613
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7420409139949444,
+      "learning_rate": 1.4280151371556472e-07,
+      "loss": 1.0132,
+      "step": 8614
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.30230043707605,
+      "learning_rate": 1.4238867489591302e-07,
+      "loss": 0.8974,
+      "step": 8615
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6548265642707094,
+      "learning_rate": 1.4197642507838115e-07,
+      "loss": 0.9238,
+      "step": 8616
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9029855809286464,
+      "learning_rate": 1.4156476431295586e-07,
+      "loss": 0.856,
+      "step": 8617
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.0342822602750545,
+      "learning_rate": 1.4115369264955215e-07,
+      "loss": 0.8207,
+      "step": 8618
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7511513432428554,
+      "learning_rate": 1.4074321013801462e-07,
+      "loss": 0.8704,
+      "step": 8619
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7209689901016054,
+      "learning_rate": 1.4033331682811568e-07,
+      "loss": 0.818,
+      "step": 8620
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.651193755266146,
+      "learning_rate": 1.3992401276955502e-07,
+      "loss": 0.9065,
+      "step": 8621
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7734970710199283,
+      "learning_rate": 1.3951529801196295e-07,
+      "loss": 0.9759,
+      "step": 8622
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7350606916823506,
+      "learning_rate": 1.391071726048987e-07,
+      "loss": 1.0326,
+      "step": 8623
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7185308467968199,
+      "learning_rate": 1.3869963659784657e-07,
+      "loss": 0.8437,
+      "step": 8624
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8911353916896554,
+      "learning_rate": 1.3829269004022426e-07,
+      "loss": 0.9339,
+      "step": 8625
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8372614409965347,
+      "learning_rate": 1.3788633298137288e-07,
+      "loss": 0.8918,
+      "step": 8626
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8143927551793197,
+      "learning_rate": 1.3748056547056632e-07,
+      "loss": 0.9107,
+      "step": 8627
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6863219394322964,
+      "learning_rate": 1.3707538755700522e-07,
+      "loss": 0.8685,
+      "step": 8628
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.689296259782041,
+      "learning_rate": 1.3667079928981807e-07,
+      "loss": 0.9151,
+      "step": 8629
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.77867239537121,
+      "learning_rate": 1.3626680071806342e-07,
+      "loss": 0.8845,
+      "step": 8630
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.658972879340762,
+      "learning_rate": 1.358633918907265e-07,
+      "loss": 0.8578,
+      "step": 8631
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8144927322028328,
+      "learning_rate": 1.354605728567232e-07,
+      "loss": 0.8501,
+      "step": 8632
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6879652895161257,
+      "learning_rate": 1.3505834366489502e-07,
+      "loss": 0.9072,
+      "step": 8633
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.692956573005911,
+      "learning_rate": 1.3465670436401513e-07,
+      "loss": 0.8445,
+      "step": 8634
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7670710228349,
+      "learning_rate": 1.3425565500278348e-07,
+      "loss": 1.0488,
+      "step": 8635
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6600288723290713,
+      "learning_rate": 1.3385519562982842e-07,
+      "loss": 0.8662,
+      "step": 8636
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8537000796519754,
+      "learning_rate": 1.334553262937066e-07,
+      "loss": 0.9443,
+      "step": 8637
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.866380707334537,
+      "learning_rate": 1.3305604704290432e-07,
+      "loss": 0.9162,
+      "step": 8638
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6739359278297261,
+      "learning_rate": 1.32657357925835e-07,
+      "loss": 0.9714,
+      "step": 8639
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6035213794021221,
+      "learning_rate": 1.322592589908417e-07,
+      "loss": 0.8831,
+      "step": 8640
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.802876835066766,
+      "learning_rate": 1.318617502861952e-07,
+      "loss": 0.8485,
+      "step": 8641
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6553619531839114,
+      "learning_rate": 1.3146483186009418e-07,
+      "loss": 0.8322,
+      "step": 8642
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7674791427060506,
+      "learning_rate": 1.3106850376066626e-07,
+      "loss": 1.0342,
+      "step": 8643
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7942198810367045,
+      "learning_rate": 1.3067276603596856e-07,
+      "loss": 1.0256,
+      "step": 8644
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.664552276911771,
+      "learning_rate": 1.3027761873398437e-07,
+      "loss": 0.985,
+      "step": 8645
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7239029450737329,
+      "learning_rate": 1.2988306190262757e-07,
+      "loss": 0.9023,
+      "step": 8646
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7248444248076231,
+      "learning_rate": 1.2948909558974045e-07,
+      "loss": 0.8841,
+      "step": 8647
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.732531426524349,
+      "learning_rate": 1.2909571984309032e-07,
+      "loss": 0.8131,
+      "step": 8648
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6443824266504774,
+      "learning_rate": 1.2870293471037743e-07,
+      "loss": 0.8407,
+      "step": 8649
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6113504178037183,
+      "learning_rate": 1.2831074023922752e-07,
+      "loss": 0.839,
+      "step": 8650
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7367237237901707,
+      "learning_rate": 1.279191364771959e-07,
+      "loss": 0.8609,
+      "step": 8651
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6241966916615753,
+      "learning_rate": 1.2752812347176514e-07,
+      "loss": 0.8761,
+      "step": 8652
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.692614323211419,
+      "learning_rate": 1.271377012703473e-07,
+      "loss": 0.8808,
+      "step": 8653
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.707329389195086,
+      "learning_rate": 1.2674786992028287e-07,
+      "loss": 0.886,
+      "step": 8654
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.778124637957207,
+      "learning_rate": 1.263586294688396e-07,
+      "loss": 1.0452,
+      "step": 8655
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7941056500910548,
+      "learning_rate": 1.259699799632147e-07,
+      "loss": 0.8921,
+      "step": 8656
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6272474393238914,
+      "learning_rate": 1.2558192145053327e-07,
+      "loss": 0.9049,
+      "step": 8657
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6868110655488864,
+      "learning_rate": 1.251944539778477e-07,
+      "loss": 0.8849,
+      "step": 8658
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9194643456643747,
+      "learning_rate": 1.2480757759214145e-07,
+      "loss": 0.9094,
+      "step": 8659
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.696345155028626,
+      "learning_rate": 1.2442129234032373e-07,
+      "loss": 0.8693,
+      "step": 8660
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7626151921816402,
+      "learning_rate": 1.2403559826923205e-07,
+      "loss": 0.7961,
+      "step": 8661
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7404996254326943,
+      "learning_rate": 1.236504954256351e-07,
+      "loss": 0.9566,
+      "step": 8662
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.679352627921948,
+      "learning_rate": 1.2326598385622723e-07,
+      "loss": 0.833,
+      "step": 8663
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6771439780358839,
+      "learning_rate": 1.2288206360763056e-07,
+      "loss": 0.8738,
+      "step": 8664
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.701352170001282,
+      "learning_rate": 1.2249873472639783e-07,
+      "loss": 0.8534,
+      "step": 8665
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7715490483093246,
+      "learning_rate": 1.2211599725900914e-07,
+      "loss": 0.9436,
+      "step": 8666
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.823500016875931,
+      "learning_rate": 1.2173385125187177e-07,
+      "loss": 0.8881,
+      "step": 8667
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9388614534985507,
+      "learning_rate": 1.213522967513231e-07,
+      "loss": 0.8037,
+      "step": 8668
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.680898710735779,
+      "learning_rate": 1.2097133380362835e-07,
+      "loss": 0.8771,
+      "step": 8669
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.76648139891266,
+      "learning_rate": 1.2059096245497948e-07,
+      "loss": 0.8195,
+      "step": 8670
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6935585888768638,
+      "learning_rate": 1.202111827514979e-07,
+      "loss": 0.8861,
+      "step": 8671
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8098601469387469,
+      "learning_rate": 1.1983199473923457e-07,
+      "loss": 1.0908,
+      "step": 8672
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6684278826931422,
+      "learning_rate": 1.1945339846416603e-07,
+      "loss": 0.9437,
+      "step": 8673
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.754932193984157,
+      "learning_rate": 1.1907539397219837e-07,
+      "loss": 0.8732,
+      "step": 8674
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6832489565316293,
+      "learning_rate": 1.1869798130916654e-07,
+      "loss": 0.9166,
+      "step": 8675
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6736674346043008,
+      "learning_rate": 1.1832116052083232e-07,
+      "loss": 0.9056,
+      "step": 8676
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8126309096571516,
+      "learning_rate": 1.1794493165288745e-07,
+      "loss": 0.8342,
+      "step": 8677
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6860264305810373,
+      "learning_rate": 1.1756929475095103e-07,
+      "loss": 0.8266,
+      "step": 8678
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7989676083173267,
+      "learning_rate": 1.1719424986056938e-07,
+      "loss": 0.8745,
+      "step": 8679
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6671562433377647,
+      "learning_rate": 1.168197970272189e-07,
+      "loss": 0.838,
+      "step": 8680
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7190502965856658,
+      "learning_rate": 1.1644593629630274e-07,
+      "loss": 0.9228,
+      "step": 8681
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7165512355743622,
+      "learning_rate": 1.1607266771315295e-07,
+      "loss": 0.9108,
+      "step": 8682
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6235037306610982,
+      "learning_rate": 1.1569999132303001e-07,
+      "loss": 0.8189,
+      "step": 8683
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7018717058913932,
+      "learning_rate": 1.1532790717112164e-07,
+      "loss": 0.9212,
+      "step": 8684
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7679287715121683,
+      "learning_rate": 1.1495641530254454e-07,
+      "loss": 0.8869,
+      "step": 8685
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.677306390973604,
+      "learning_rate": 1.1458551576234322e-07,
+      "loss": 0.8992,
+      "step": 8686
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8237467326056576,
+      "learning_rate": 1.1421520859549063e-07,
+      "loss": 0.9093,
+      "step": 8687
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.4676196670502355,
+      "learning_rate": 1.1384549384688804e-07,
+      "loss": 0.8605,
+      "step": 8688
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7721265487395383,
+      "learning_rate": 1.134763715613646e-07,
+      "loss": 0.8752,
+      "step": 8689
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.746287931031475,
+      "learning_rate": 1.1310784178367728e-07,
+      "loss": 0.8731,
+      "step": 8690
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8021840596513031,
+      "learning_rate": 1.1273990455851203e-07,
+      "loss": 1.0133,
+      "step": 8691
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7972779039730296,
+      "learning_rate": 1.1237255993048202e-07,
+      "loss": 0.9204,
+      "step": 8692
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5821575651945894,
+      "learning_rate": 1.1200580794412997e-07,
+      "loss": 0.786,
+      "step": 8693
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.685144881516343,
+      "learning_rate": 1.1163964864392474e-07,
+      "loss": 0.8352,
+      "step": 8694
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.802961833034365,
+      "learning_rate": 1.1127408207426471e-07,
+      "loss": 0.8492,
+      "step": 8695
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7771086493368791,
+      "learning_rate": 1.109091082794761e-07,
+      "loss": 1.0999,
+      "step": 8696
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8864989562134697,
+      "learning_rate": 1.1054472730381405e-07,
+      "loss": 0.8683,
+      "step": 8697
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7794715471711953,
+      "learning_rate": 1.1018093919145989e-07,
+      "loss": 0.9312,
+      "step": 8698
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8782141428372994,
+      "learning_rate": 1.0981774398652446e-07,
+      "loss": 0.8952,
+      "step": 8699
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6887724005940208,
+      "learning_rate": 1.0945514173304695e-07,
+      "loss": 0.836,
+      "step": 8700
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6561020357661516,
+      "learning_rate": 1.0909313247499331e-07,
+      "loss": 0.9074,
+      "step": 8701
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.807447588981726,
+      "learning_rate": 1.08731716256259e-07,
+      "loss": 0.8806,
+      "step": 8702
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6915235198421559,
+      "learning_rate": 1.0837089312066729e-07,
+      "loss": 0.8927,
+      "step": 8703
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6864662471900609,
+      "learning_rate": 1.0801066311196872e-07,
+      "loss": 0.8595,
+      "step": 8704
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.910660121310086,
+      "learning_rate": 1.0765102627384227e-07,
+      "loss": 0.8439,
+      "step": 8705
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8040777317197338,
+      "learning_rate": 1.0729198264989582e-07,
+      "loss": 0.8684,
+      "step": 8706
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6611983708931068,
+      "learning_rate": 1.0693353228366344e-07,
+      "loss": 0.7713,
+      "step": 8707
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7055110134140223,
+      "learning_rate": 1.0657567521860978e-07,
+      "loss": 0.8861,
+      "step": 8708
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7751152070795508,
+      "learning_rate": 1.0621841149812573e-07,
+      "loss": 0.8178,
+      "step": 8709
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.734301326268237,
+      "learning_rate": 1.0586174116552994e-07,
+      "loss": 0.8762,
+      "step": 8710
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.625066302517425,
+      "learning_rate": 1.0550566426407117e-07,
+      "loss": 0.8754,
+      "step": 8711
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6829850258789252,
+      "learning_rate": 1.0515018083692486e-07,
+      "loss": 0.9457,
+      "step": 8712
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.785946456771606,
+      "learning_rate": 1.0479529092719376e-07,
+      "loss": 0.885,
+      "step": 8713
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8912625434400956,
+      "learning_rate": 1.0444099457791013e-07,
+      "loss": 0.8776,
+      "step": 8714
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7802010199443186,
+      "learning_rate": 1.0408729183203403e-07,
+      "loss": 0.9867,
+      "step": 8715
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7398046738498467,
+      "learning_rate": 1.0373418273245228e-07,
+      "loss": 0.9068,
+      "step": 8716
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6551546580853094,
+      "learning_rate": 1.0338166732198062e-07,
+      "loss": 0.8787,
+      "step": 8717
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7321368640287063,
+      "learning_rate": 1.0302974564336265e-07,
+      "loss": 0.9183,
+      "step": 8718
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.768267936714945,
+      "learning_rate": 1.0267841773927089e-07,
+      "loss": 1.0255,
+      "step": 8719
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6649649099514554,
+      "learning_rate": 1.0232768365230462e-07,
+      "loss": 0.7575,
+      "step": 8720
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.650647563833598,
+      "learning_rate": 1.01977543424992e-07,
+      "loss": 0.9331,
+      "step": 8721
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7096864216095242,
+      "learning_rate": 1.0162799709978743e-07,
+      "loss": 0.9114,
+      "step": 8722
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6977216704583267,
+      "learning_rate": 1.012790447190759e-07,
+      "loss": 0.8739,
+      "step": 8723
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.654371209262714,
+      "learning_rate": 1.0093068632516855e-07,
+      "loss": 0.8947,
+      "step": 8724
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6502943794088205,
+      "learning_rate": 1.005829219603055e-07,
+      "loss": 0.8863,
+      "step": 8725
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7266107456447235,
+      "learning_rate": 1.0023575166665356e-07,
+      "loss": 0.9146,
+      "step": 8726
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7191674171784719,
+      "learning_rate": 9.98891754863085e-08,
+      "loss": 0.9053,
+      "step": 8727
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6508448271613672,
+      "learning_rate": 9.954319346129504e-08,
+      "loss": 0.8483,
+      "step": 8728
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6806593406810597,
+      "learning_rate": 9.919780563356296e-08,
+      "loss": 0.8375,
+      "step": 8729
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.756659986191418,
+      "learning_rate": 9.885301204499321e-08,
+      "loss": 0.8698,
+      "step": 8730
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6574648906527947,
+      "learning_rate": 9.850881273739232e-08,
+      "loss": 0.9046,
+      "step": 8731
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7825988557562138,
+      "learning_rate": 9.816520775249583e-08,
+      "loss": 0.9176,
+      "step": 8732
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9055536481280362,
+      "learning_rate": 9.782219713196706e-08,
+      "loss": 0.8832,
+      "step": 8733
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.701554984357639,
+      "learning_rate": 9.747978091739774e-08,
+      "loss": 0.8438,
+      "step": 8734
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7732869975969348,
+      "learning_rate": 9.713795915030577e-08,
+      "loss": 0.8652,
+      "step": 8735
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6621618550626442,
+      "learning_rate": 9.679673187214023e-08,
+      "loss": 0.9237,
+      "step": 8736
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6623637146218884,
+      "learning_rate": 9.645609912427412e-08,
+      "loss": 0.8362,
+      "step": 8737
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.743433721770903,
+      "learning_rate": 9.611606094801052e-08,
+      "loss": 0.927,
+      "step": 8738
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6880934156260312,
+      "learning_rate": 9.577661738458144e-08,
+      "loss": 0.8818,
+      "step": 8739
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8033331168405395,
+      "learning_rate": 9.543776847514508e-08,
+      "loss": 0.8549,
+      "step": 8740
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8162257005797307,
+      "learning_rate": 9.509951426078745e-08,
+      "loss": 0.8896,
+      "step": 8741
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8397332680300942,
+      "learning_rate": 9.476185478252354e-08,
+      "loss": 0.8801,
+      "step": 8742
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7746170797697094,
+      "learning_rate": 9.442479008129557e-08,
+      "loss": 1.0676,
+      "step": 8743
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7582210475904232,
+      "learning_rate": 9.40883201979742e-08,
+      "loss": 1.008,
+      "step": 8744
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7809745981474663,
+      "learning_rate": 9.37524451733568e-08,
+      "loss": 0.9593,
+      "step": 8745
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7212151641640505,
+      "learning_rate": 9.341716504817022e-08,
+      "loss": 0.9188,
+      "step": 8746
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.712081039534405,
+      "learning_rate": 9.308247986306862e-08,
+      "loss": 0.9496,
+      "step": 8747
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7772235921773951,
+      "learning_rate": 9.274838965863175e-08,
+      "loss": 1.0282,
+      "step": 8748
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7538186775136482,
+      "learning_rate": 9.241489447537111e-08,
+      "loss": 0.8495,
+      "step": 8749
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7060304883884183,
+      "learning_rate": 9.208199435372379e-08,
+      "loss": 0.844,
+      "step": 8750
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7742980726500839,
+      "learning_rate": 9.174968933405415e-08,
+      "loss": 1.0129,
+      "step": 8751
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7265033516315342,
+      "learning_rate": 9.141797945665609e-08,
+      "loss": 0.9487,
+      "step": 8752
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9427486818306074,
+      "learning_rate": 9.108686476175132e-08,
+      "loss": 0.9347,
+      "step": 8753
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.729457791940581,
+      "learning_rate": 9.075634528948718e-08,
+      "loss": 0.8118,
+      "step": 8754
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7991561550423116,
+      "learning_rate": 9.042642107994103e-08,
+      "loss": 0.8532,
+      "step": 8755
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7364859575347862,
+      "learning_rate": 9.009709217311702e-08,
+      "loss": 0.8869,
+      "step": 8756
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7292755646672495,
+      "learning_rate": 8.976835860894762e-08,
+      "loss": 0.857,
+      "step": 8757
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6628914866427786,
+      "learning_rate": 8.944022042729317e-08,
+      "loss": 0.9134,
+      "step": 8758
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7078757929066355,
+      "learning_rate": 8.91126776679413e-08,
+      "loss": 0.8594,
+      "step": 8759
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7924786921108733,
+      "learning_rate": 8.878573037060689e-08,
+      "loss": 0.8872,
+      "step": 8760
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.777690336494681,
+      "learning_rate": 8.845937857493492e-08,
+      "loss": 0.9944,
+      "step": 8761
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7146020668537298,
+      "learning_rate": 8.813362232049594e-08,
+      "loss": 0.96,
+      "step": 8762
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7702086092082934,
+      "learning_rate": 8.780846164678836e-08,
+      "loss": 0.8883,
+      "step": 8763
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.793273497276005,
+      "learning_rate": 8.748389659324008e-08,
+      "loss": 0.9039,
+      "step": 8764
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8563030741518667,
+      "learning_rate": 8.715992719920575e-08,
+      "loss": 0.8821,
+      "step": 8765
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.692136041261812,
+      "learning_rate": 8.683655350396724e-08,
+      "loss": 0.8431,
+      "step": 8766
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8242016865772352,
+      "learning_rate": 8.651377554673435e-08,
+      "loss": 0.9337,
+      "step": 8767
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7128776837757196,
+      "learning_rate": 8.619159336664684e-08,
+      "loss": 0.8677,
+      "step": 8768
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8163318864051665,
+      "learning_rate": 8.587000700276793e-08,
+      "loss": 0.8186,
+      "step": 8769
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6389646007746306,
+      "learning_rate": 8.554901649409253e-08,
+      "loss": 0.8349,
+      "step": 8770
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7453384275113522,
+      "learning_rate": 8.522862187954173e-08,
+      "loss": 0.8911,
+      "step": 8771
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4937871094428208,
+      "learning_rate": 8.49088231979639e-08,
+      "loss": 0.8432,
+      "step": 8772
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7842683633016247,
+      "learning_rate": 8.458962048813634e-08,
+      "loss": 0.9421,
+      "step": 8773
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7105912446655767,
+      "learning_rate": 8.427101378876368e-08,
+      "loss": 0.8833,
+      "step": 8774
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7085382583201034,
+      "learning_rate": 8.39530031384772e-08,
+      "loss": 0.8722,
+      "step": 8775
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.713771971270094,
+      "learning_rate": 8.36355885758372e-08,
+      "loss": 0.9537,
+      "step": 8776
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7232713531532322,
+      "learning_rate": 8.331877013933176e-08,
+      "loss": 0.8804,
+      "step": 8777
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8043590291619638,
+      "learning_rate": 8.300254786737627e-08,
+      "loss": 0.8998,
+      "step": 8778
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.748231085629352,
+      "learning_rate": 8.268692179831228e-08,
+      "loss": 0.9492,
+      "step": 8779
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.944237886563713,
+      "learning_rate": 8.237189197041196e-08,
+      "loss": 1.0028,
+      "step": 8780
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.725507934137953,
+      "learning_rate": 8.205745842187363e-08,
+      "loss": 0.7632,
+      "step": 8781
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.742370024203653,
+      "learning_rate": 8.174362119082291e-08,
+      "loss": 0.8635,
+      "step": 8782
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8342276961445105,
+      "learning_rate": 8.143038031531381e-08,
+      "loss": 0.875,
+      "step": 8783
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8158857503321166,
+      "learning_rate": 8.111773583332871e-08,
+      "loss": 0.9179,
+      "step": 8784
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7771395373419845,
+      "learning_rate": 8.080568778277509e-08,
+      "loss": 0.9493,
+      "step": 8785
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.783695062681277,
+      "learning_rate": 8.049423620149155e-08,
+      "loss": 0.8192,
+      "step": 8786
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6812631544417553,
+      "learning_rate": 8.018338112724178e-08,
+      "loss": 0.8695,
+      "step": 8787
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6300141931628267,
+      "learning_rate": 7.98731225977184e-08,
+      "loss": 0.8343,
+      "step": 8788
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6797131687964026,
+      "learning_rate": 7.956346065054132e-08,
+      "loss": 0.8064,
+      "step": 8789
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6378605104665973,
+      "learning_rate": 7.925439532325774e-08,
+      "loss": 0.8713,
+      "step": 8790
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.861710388795536,
+      "learning_rate": 7.894592665334266e-08,
+      "loss": 0.8817,
+      "step": 8791
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7381407927136665,
+      "learning_rate": 7.863805467820007e-08,
+      "loss": 0.9509,
+      "step": 8792
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.728224629830989,
+      "learning_rate": 7.833077943515954e-08,
+      "loss": 0.9442,
+      "step": 8793
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7252641636734705,
+      "learning_rate": 7.802410096147961e-08,
+      "loss": 0.9085,
+      "step": 8794
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7766720271773432,
+      "learning_rate": 7.771801929434608e-08,
+      "loss": 1.0634,
+      "step": 8795
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9603008739156298,
+      "learning_rate": 7.74125344708726e-08,
+      "loss": 0.8883,
+      "step": 8796
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7685373426472952,
+      "learning_rate": 7.71076465281001e-08,
+      "loss": 0.8771,
+      "step": 8797
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.671431423901742,
+      "learning_rate": 7.680335550299678e-08,
+      "loss": 0.8536,
+      "step": 8798
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.800484202467439,
+      "learning_rate": 7.64996614324598e-08,
+      "loss": 0.9059,
+      "step": 8799
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7373861939408914,
+      "learning_rate": 7.619656435331302e-08,
+      "loss": 0.8655,
+      "step": 8800
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7105118362804141,
+      "learning_rate": 7.589406430230706e-08,
+      "loss": 0.9478,
+      "step": 8801
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7741530499471447,
+      "learning_rate": 7.559216131612256e-08,
+      "loss": 0.9314,
+      "step": 8802
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6790622925476417,
+      "learning_rate": 7.529085543136472e-08,
+      "loss": 0.8588,
+      "step": 8803
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.646597252695301,
+      "learning_rate": 7.499014668456872e-08,
+      "loss": 0.9009,
+      "step": 8804
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.819361423177791,
+      "learning_rate": 7.469003511219708e-08,
+      "loss": 0.9112,
+      "step": 8805
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.661352790042645,
+      "learning_rate": 7.439052075063847e-08,
+      "loss": 0.9193,
+      "step": 8806
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6938598123814248,
+      "learning_rate": 7.409160363621049e-08,
+      "loss": 0.8867,
+      "step": 8807
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.748930993359509,
+      "learning_rate": 7.379328380515805e-08,
+      "loss": 0.8952,
+      "step": 8808
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7017792350390337,
+      "learning_rate": 7.349556129365276e-08,
+      "loss": 0.8428,
+      "step": 8809
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7104673229158864,
+      "learning_rate": 7.31984361377952e-08,
+      "loss": 0.8835,
+      "step": 8810
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.721271285473092,
+      "learning_rate": 7.290190837361211e-08,
+      "loss": 0.8317,
+      "step": 8811
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6217396730571771,
+      "learning_rate": 7.260597803705971e-08,
+      "loss": 0.8727,
+      "step": 8812
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7369141221640836,
+      "learning_rate": 7.231064516401932e-08,
+      "loss": 0.8797,
+      "step": 8813
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8638751535000941,
+      "learning_rate": 7.201590979030116e-08,
+      "loss": 0.8864,
+      "step": 8814
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7853878754829022,
+      "learning_rate": 7.172177195164387e-08,
+      "loss": 1.044,
+      "step": 8815
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6186404965288494,
+      "learning_rate": 7.142823168371171e-08,
+      "loss": 0.7876,
+      "step": 8816
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.579585154508702,
+      "learning_rate": 7.113528902209787e-08,
+      "loss": 0.8693,
+      "step": 8817
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8001981285371624,
+      "learning_rate": 7.084294400232283e-08,
+      "loss": 0.8813,
+      "step": 8818
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.8180598032321683,
+      "learning_rate": 7.055119665983379e-08,
+      "loss": 1.0557,
+      "step": 8819
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7977290683950893,
+      "learning_rate": 7.026004703000689e-08,
+      "loss": 0.825,
+      "step": 8820
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7001088242551985,
+      "learning_rate": 6.9969495148145e-08,
+      "loss": 0.9354,
+      "step": 8821
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.797356298583357,
+      "learning_rate": 6.967954104947771e-08,
+      "loss": 0.8669,
+      "step": 8822
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.8009441176804548,
+      "learning_rate": 6.939018476916359e-08,
+      "loss": 1.0258,
+      "step": 8823
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9012258700201452,
+      "learning_rate": 6.91014263422879e-08,
+      "loss": 0.9468,
+      "step": 8824
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6712749695536122,
+      "learning_rate": 6.881326580386372e-08,
+      "loss": 0.8585,
+      "step": 8825
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6949411744251415,
+      "learning_rate": 6.852570318883145e-08,
+      "loss": 0.8451,
+      "step": 8826
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7847498028469047,
+      "learning_rate": 6.823873853205931e-08,
+      "loss": 0.9019,
+      "step": 8827
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8301198485257983,
+      "learning_rate": 6.79523718683417e-08,
+      "loss": 0.8909,
+      "step": 8828
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7997265839082333,
+      "learning_rate": 6.766660323240303e-08,
+      "loss": 0.9405,
+      "step": 8829
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7141982040544148,
+      "learning_rate": 6.738143265889285e-08,
+      "loss": 0.896,
+      "step": 8830
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6634732983490532,
+      "learning_rate": 6.709686018238958e-08,
+      "loss": 0.8502,
+      "step": 8831
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.803037141831488,
+      "learning_rate": 6.681288583739787e-08,
+      "loss": 0.9035,
+      "step": 8832
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6619339651063416,
+      "learning_rate": 6.652950965835181e-08,
+      "loss": 0.793,
+      "step": 8833
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6958177435297506,
+      "learning_rate": 6.624673167961004e-08,
+      "loss": 0.9443,
+      "step": 8834
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5378204699646796,
+      "learning_rate": 6.59645519354618e-08,
+      "loss": 0.8718,
+      "step": 8835
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.764638309614953,
+      "learning_rate": 6.568297046012196e-08,
+      "loss": 0.9329,
+      "step": 8836
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7290949591877172,
+      "learning_rate": 6.540198728773262e-08,
+      "loss": 0.9618,
+      "step": 8837
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.872213947790235,
+      "learning_rate": 6.512160245236432e-08,
+      "loss": 0.9711,
+      "step": 8838
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8483003544116654,
+      "learning_rate": 6.484181598801542e-08,
+      "loss": 0.9476,
+      "step": 8839
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8165979406763404,
+      "learning_rate": 6.456262792861046e-08,
+      "loss": 0.9298,
+      "step": 8840
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.743048501710946,
+      "learning_rate": 6.428403830800123e-08,
+      "loss": 0.9003,
+      "step": 8841
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7804193480048996,
+      "learning_rate": 6.400604715996905e-08,
+      "loss": 1.0569,
+      "step": 8842
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7242733565395683,
+      "learning_rate": 6.372865451822086e-08,
+      "loss": 0.8394,
+      "step": 8843
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6835111351209955,
+      "learning_rate": 6.345186041639029e-08,
+      "loss": 0.8691,
+      "step": 8844
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7500758629832467,
+      "learning_rate": 6.317566488804105e-08,
+      "loss": 0.8303,
+      "step": 8845
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.699550610875349,
+      "learning_rate": 6.290006796666249e-08,
+      "loss": 0.8955,
+      "step": 8846
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7506613634780688,
+      "learning_rate": 6.262506968567061e-08,
+      "loss": 0.9774,
+      "step": 8847
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7124983288795061,
+      "learning_rate": 6.235067007841155e-08,
+      "loss": 0.8778,
+      "step": 8848
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8466498050615583,
+      "learning_rate": 6.207686917815592e-08,
+      "loss": 0.899,
+      "step": 8849
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7591888241633542,
+      "learning_rate": 6.18036670181038e-08,
+      "loss": 0.8723,
+      "step": 8850
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.61232954830691,
+      "learning_rate": 6.153106363138151e-08,
+      "loss": 0.8892,
+      "step": 8851
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6548054021145893,
+      "learning_rate": 6.125905905104368e-08,
+      "loss": 0.8387,
+      "step": 8852
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5697542340565545,
+      "learning_rate": 6.098765331007061e-08,
+      "loss": 0.8567,
+      "step": 8853
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.797859432906638,
+      "learning_rate": 6.071684644137265e-08,
+      "loss": 0.9176,
+      "step": 8854
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.69822809057246,
+      "learning_rate": 6.044663847778521e-08,
+      "loss": 0.8485,
+      "step": 8855
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8179272236354664,
+      "learning_rate": 6.017702945207149e-08,
+      "loss": 0.913,
+      "step": 8856
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7251087436150738,
+      "learning_rate": 5.990801939692315e-08,
+      "loss": 0.892,
+      "step": 8857
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7816913788683594,
+      "learning_rate": 5.963960834495908e-08,
+      "loss": 1.0217,
+      "step": 8858
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.71616828150927,
+      "learning_rate": 5.937179632872436e-08,
+      "loss": 0.8829,
+      "step": 8859
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6681775415608753,
+      "learning_rate": 5.910458338069192e-08,
+      "loss": 0.8707,
+      "step": 8860
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6147134810100792,
+      "learning_rate": 5.8837969533263595e-08,
+      "loss": 0.9361,
+      "step": 8861
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7241914758552606,
+      "learning_rate": 5.857195481876521e-08,
+      "loss": 0.8721,
+      "step": 8862
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7194770741885168,
+      "learning_rate": 5.830653926945373e-08,
+      "loss": 0.9785,
+      "step": 8863
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6891964027238264,
+      "learning_rate": 5.804172291751064e-08,
+      "loss": 0.8501,
+      "step": 8864
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7327046697770807,
+      "learning_rate": 5.777750579504582e-08,
+      "loss": 0.8441,
+      "step": 8865
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.751938693775302,
+      "learning_rate": 5.751388793409696e-08,
+      "loss": 0.9844,
+      "step": 8866
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.8386238707148712,
+      "learning_rate": 5.725086936662905e-08,
+      "loss": 1.0297,
+      "step": 8867
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7462363264918382,
+      "learning_rate": 5.6988450124533244e-08,
+      "loss": 0.9933,
+      "step": 8868
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8599863604485445,
+      "learning_rate": 5.672663023962854e-08,
+      "loss": 1.0067,
+      "step": 8869
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8456713218337186,
+      "learning_rate": 5.646540974366288e-08,
+      "loss": 0.9604,
+      "step": 8870
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.738856334355234,
+      "learning_rate": 5.620478866830814e-08,
+      "loss": 0.8973,
+      "step": 8871
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8165274407170784,
+      "learning_rate": 5.594476704516738e-08,
+      "loss": 0.8524,
+      "step": 8872
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7069556373185137,
+      "learning_rate": 5.568534490576816e-08,
+      "loss": 0.8868,
+      "step": 8873
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.699828107224752,
+      "learning_rate": 5.542652228156698e-08,
+      "loss": 0.8865,
+      "step": 8874
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8733041992213415,
+      "learning_rate": 5.5168299203945953e-08,
+      "loss": 0.8818,
+      "step": 8875
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6901173063260295,
+      "learning_rate": 5.491067570421615e-08,
+      "loss": 0.8674,
+      "step": 8876
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6086175012147987,
+      "learning_rate": 5.465365181361537e-08,
+      "loss": 0.8415,
+      "step": 8877
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7622831247538544,
+      "learning_rate": 5.439722756330812e-08,
+      "loss": 0.9099,
+      "step": 8878
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.671792320404714,
+      "learning_rate": 5.4141402984387304e-08,
+      "loss": 0.8571,
+      "step": 8879
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7779374968832276,
+      "learning_rate": 5.388617810787256e-08,
+      "loss": 0.8924,
+      "step": 8880
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7545494700455597,
+      "learning_rate": 5.363155296471079e-08,
+      "loss": 0.8814,
+      "step": 8881
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7355268452741641,
+      "learning_rate": 5.3377527585775635e-08,
+      "loss": 0.8626,
+      "step": 8882
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7587017700937684,
+      "learning_rate": 5.3124102001868574e-08,
+      "loss": 0.8561,
+      "step": 8883
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6831612168029113,
+      "learning_rate": 5.287127624371946e-08,
+      "loss": 0.8679,
+      "step": 8884
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6889580196336205,
+      "learning_rate": 5.261905034198267e-08,
+      "loss": 0.8795,
+      "step": 8885
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.765930972216828,
+      "learning_rate": 5.236742432724262e-08,
+      "loss": 1.0508,
+      "step": 8886
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7923678839673987,
+      "learning_rate": 5.211639823000936e-08,
+      "loss": 0.9425,
+      "step": 8887
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.8011042401078649,
+      "learning_rate": 5.1865972080720195e-08,
+      "loss": 1.0891,
+      "step": 8888
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7539197783992857,
+      "learning_rate": 5.16161459097414e-08,
+      "loss": 0.9203,
+      "step": 8889
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.705251960380442,
+      "learning_rate": 5.13669197473643e-08,
+      "loss": 0.859,
+      "step": 8890
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6878493932151493,
+      "learning_rate": 5.11182936238086e-08,
+      "loss": 0.8167,
+      "step": 8891
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5971559035553664,
+      "learning_rate": 5.0870267569221864e-08,
+      "loss": 0.9114,
+      "step": 8892
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6730817709332049,
+      "learning_rate": 5.0622841613676695e-08,
+      "loss": 0.9578,
+      "step": 8893
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6941228875662697,
+      "learning_rate": 5.0376015787175215e-08,
+      "loss": 0.9023,
+      "step": 8894
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.848185508010583,
+      "learning_rate": 5.0129790119646804e-08,
+      "loss": 0.8739,
+      "step": 8895
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7554975845581249,
+      "learning_rate": 4.988416464094481e-08,
+      "loss": 0.8673,
+      "step": 8896
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.763305474072755,
+      "learning_rate": 4.963913938085374e-08,
+      "loss": 0.8286,
+      "step": 8897
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7380162147368385,
+      "learning_rate": 4.939471436908427e-08,
+      "loss": 0.9477,
+      "step": 8898
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6987738026288308,
+      "learning_rate": 4.9150889635272145e-08,
+      "loss": 0.9271,
+      "step": 8899
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7148690992722675,
+      "learning_rate": 4.890766520898316e-08,
+      "loss": 0.8764,
+      "step": 8900
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6698686989127507,
+      "learning_rate": 4.8665041119708714e-08,
+      "loss": 0.8878,
+      "step": 8901
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7169047369672084,
+      "learning_rate": 4.8423017396868054e-08,
+      "loss": 0.9491,
+      "step": 8902
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6538514911559579,
+      "learning_rate": 4.818159406980716e-08,
+      "loss": 0.8639,
+      "step": 8903
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9285773021157036,
+      "learning_rate": 4.7940771167799274e-08,
+      "loss": 0.8546,
+      "step": 8904
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6899697224651593,
+      "learning_rate": 4.770054872004548e-08,
+      "loss": 0.8791,
+      "step": 8905
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7001991090521789,
+      "learning_rate": 4.7460926755673595e-08,
+      "loss": 0.8987,
+      "step": 8906
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7738860319744307,
+      "learning_rate": 4.72219053037376e-08,
+      "loss": 1.0038,
+      "step": 8907
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7701036416453981,
+      "learning_rate": 4.698348439322098e-08,
+      "loss": 0.827,
+      "step": 8908
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7863045138553155,
+      "learning_rate": 4.674566405303227e-08,
+      "loss": 1.0146,
+      "step": 8909
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6208247051974858,
+      "learning_rate": 4.65084443120084e-08,
+      "loss": 0.8782,
+      "step": 8910
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7880261381925586,
+      "learning_rate": 4.627182519891305e-08,
+      "loss": 1.0752,
+      "step": 8911
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7063538310636899,
+      "learning_rate": 4.6035806742436575e-08,
+      "loss": 0.825,
+      "step": 8912
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6472008082219505,
+      "learning_rate": 4.5800388971197764e-08,
+      "loss": 0.9382,
+      "step": 8913
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.672508794677758,
+      "learning_rate": 4.5565571913741005e-08,
+      "loss": 0.8332,
+      "step": 8914
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6763003775437368,
+      "learning_rate": 4.533135559853963e-08,
+      "loss": 0.8696,
+      "step": 8915
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8043807929063636,
+      "learning_rate": 4.5097740053993144e-08,
+      "loss": 0.8836,
+      "step": 8916
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6326466386802474,
+      "learning_rate": 4.486472530842723e-08,
+      "loss": 0.8574,
+      "step": 8917
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7313987111186204,
+      "learning_rate": 4.463231139009649e-08,
+      "loss": 0.9043,
+      "step": 8918
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6927496784043867,
+      "learning_rate": 4.440049832718174e-08,
+      "loss": 0.9356,
+      "step": 8919
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.870356306661973,
+      "learning_rate": 4.416928614779104e-08,
+      "loss": 0.9193,
+      "step": 8920
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.705461592435337,
+      "learning_rate": 4.393867487995973e-08,
+      "loss": 0.924,
+      "step": 8921
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7317126631987951,
+      "learning_rate": 4.370866455165046e-08,
+      "loss": 0.9012,
+      "step": 8922
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7245441685321954,
+      "learning_rate": 4.347925519075258e-08,
+      "loss": 0.8882,
+      "step": 8923
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8179619861907725,
+      "learning_rate": 4.325044682508217e-08,
+      "loss": 0.8556,
+      "step": 8924
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.848362654376589,
+      "learning_rate": 4.3022239482384264e-08,
+      "loss": 0.903,
+      "step": 8925
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5910201405758972,
+      "learning_rate": 4.279463319032895e-08,
+      "loss": 0.8657,
+      "step": 8926
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6620280011002693,
+      "learning_rate": 4.256762797651415e-08,
+      "loss": 0.8268,
+      "step": 8927
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.751251381784036,
+      "learning_rate": 4.234122386846562e-08,
+      "loss": 0.9114,
+      "step": 8928
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7932472079022234,
+      "learning_rate": 4.2115420893635296e-08,
+      "loss": 1.0179,
+      "step": 8929
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6705358779688988,
+      "learning_rate": 4.1890219079402384e-08,
+      "loss": 0.9032,
+      "step": 8930
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7754452620224036,
+      "learning_rate": 4.166561845307393e-08,
+      "loss": 1.0661,
+      "step": 8931
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9250148332686925,
+      "learning_rate": 4.1441619041883155e-08,
+      "loss": 0.9392,
+      "step": 8932
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6643434823376908,
+      "learning_rate": 4.1218220872990546e-08,
+      "loss": 0.8892,
+      "step": 8933
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.730597087840161,
+      "learning_rate": 4.099542397348444e-08,
+      "loss": 0.8652,
+      "step": 8934
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6901388846520238,
+      "learning_rate": 4.0773228370379334e-08,
+      "loss": 0.8938,
+      "step": 8935
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6976235775855437,
+      "learning_rate": 4.055163409061758e-08,
+      "loss": 0.811,
+      "step": 8936
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6700276766233098,
+      "learning_rate": 4.033064116106766e-08,
+      "loss": 0.8798,
+      "step": 8937
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8001882379960406,
+      "learning_rate": 4.01102496085265e-08,
+      "loss": 0.9604,
+      "step": 8938
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8398321897652261,
+      "learning_rate": 3.989045945971659e-08,
+      "loss": 0.9396,
+      "step": 8939
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7432841475041494,
+      "learning_rate": 3.967127074128885e-08,
+      "loss": 0.8576,
+      "step": 8940
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.72804025308454,
+      "learning_rate": 3.9452683479819783e-08,
+      "loss": 0.9422,
+      "step": 8941
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7281732295649466,
+      "learning_rate": 3.923469770181543e-08,
+      "loss": 0.9144,
+      "step": 8942
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7344771298564283,
+      "learning_rate": 3.901731343370574e-08,
+      "loss": 0.8813,
+      "step": 8943
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8121729413924828,
+      "learning_rate": 3.8800530701849634e-08,
+      "loss": 0.9379,
+      "step": 8944
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.695819052500904,
+      "learning_rate": 3.858434953253332e-08,
+      "loss": 0.9707,
+      "step": 8945
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7703958874659151,
+      "learning_rate": 3.836876995196914e-08,
+      "loss": 0.8029,
+      "step": 8946
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.659591993290325,
+      "learning_rate": 3.8153791986297315e-08,
+      "loss": 0.9332,
+      "step": 8947
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7570856516879534,
+      "learning_rate": 3.793941566158421e-08,
+      "loss": 0.8263,
+      "step": 8948
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7582010964140862,
+      "learning_rate": 3.772564100382348e-08,
+      "loss": 0.8749,
+      "step": 8949
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.810719522043132,
+      "learning_rate": 3.75124680389366e-08,
+      "loss": 0.9397,
+      "step": 8950
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7139878045010424,
+      "learning_rate": 3.729989679277124e-08,
+      "loss": 0.9454,
+      "step": 8951
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8472551925009655,
+      "learning_rate": 3.7087927291102335e-08,
+      "loss": 0.9097,
+      "step": 8952
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.653718743968741,
+      "learning_rate": 3.687655955963154e-08,
+      "loss": 0.905,
+      "step": 8953
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6624409750389941,
+      "learning_rate": 3.666579362398892e-08,
+      "loss": 0.9247,
+      "step": 8954
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.78575587822623,
+      "learning_rate": 3.645562950973014e-08,
+      "loss": 1.0602,
+      "step": 8955
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6849520477787194,
+      "learning_rate": 3.624606724233759e-08,
+      "loss": 0.8711,
+      "step": 8956
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7236569883947854,
+      "learning_rate": 3.603710684722317e-08,
+      "loss": 0.8779,
+      "step": 8957
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.673275018096668,
+      "learning_rate": 3.5828748349722164e-08,
+      "loss": 0.8322,
+      "step": 8958
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.726566012505798,
+      "learning_rate": 3.562099177509937e-08,
+      "loss": 0.9682,
+      "step": 8959
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8254854572908077,
+      "learning_rate": 3.541383714854685e-08,
+      "loss": 0.8792,
+      "step": 8960
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7636441178940484,
+      "learning_rate": 3.520728449518118e-08,
+      "loss": 1.04,
+      "step": 8961
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7230886840477964,
+      "learning_rate": 3.5001333840049e-08,
+      "loss": 0.9147,
+      "step": 8962
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8304015594768872,
+      "learning_rate": 3.4795985208122e-08,
+      "loss": 0.903,
+      "step": 8963
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.918036581863025,
+      "learning_rate": 3.4591238624299696e-08,
+      "loss": 0.8819,
+      "step": 8964
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7421933983926192,
+      "learning_rate": 3.438709411340779e-08,
+      "loss": 1.0517,
+      "step": 8965
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6327324776391086,
+      "learning_rate": 3.418355170019982e-08,
+      "loss": 0.803,
+      "step": 8966
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7479503801981549,
+      "learning_rate": 3.3980611409356045e-08,
+      "loss": 0.8941,
+      "step": 8967
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8155400739378578,
+      "learning_rate": 3.377827326548344e-08,
+      "loss": 0.9456,
+      "step": 8968
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7506774851964663,
+      "learning_rate": 3.357653729311683e-08,
+      "loss": 1.0565,
+      "step": 8969
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6666653612872917,
+      "learning_rate": 3.337540351671664e-08,
+      "loss": 0.9143,
+      "step": 8970
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6774366935344318,
+      "learning_rate": 3.317487196067115e-08,
+      "loss": 0.8973,
+      "step": 8971
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7062203811175196,
+      "learning_rate": 3.29749426492959e-08,
+      "loss": 0.9158,
+      "step": 8972
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.732725148731328,
+      "learning_rate": 3.277561560683262e-08,
+      "loss": 0.9027,
+      "step": 8973
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8192702145432427,
+      "learning_rate": 3.257689085745086e-08,
+      "loss": 0.9689,
+      "step": 8974
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9127811301605193,
+      "learning_rate": 3.237876842524634e-08,
+      "loss": 0.8451,
+      "step": 8975
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9551643038628586,
+      "learning_rate": 3.2181248334242075e-08,
+      "loss": 0.8868,
+      "step": 8976
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6531556440667723,
+      "learning_rate": 3.198433060838779e-08,
+      "loss": 0.8514,
+      "step": 8977
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7617610696275698,
+      "learning_rate": 3.1788015271561054e-08,
+      "loss": 0.9401,
+      "step": 8978
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7385264135254228,
+      "learning_rate": 3.159230234756561e-08,
+      "loss": 0.8762,
+      "step": 8979
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6966012306652218,
+      "learning_rate": 3.139719186013246e-08,
+      "loss": 0.8954,
+      "step": 8980
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7788894648144027,
+      "learning_rate": 3.1202683832918804e-08,
+      "loss": 0.9353,
+      "step": 8981
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7243292209200727,
+      "learning_rate": 3.100877828950965e-08,
+      "loss": 0.8627,
+      "step": 8982
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.747644751973078,
+      "learning_rate": 3.0815475253417326e-08,
+      "loss": 0.8787,
+      "step": 8983
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7613750029026443,
+      "learning_rate": 3.062277474807917e-08,
+      "loss": 0.9184,
+      "step": 8984
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7696073657812026,
+      "learning_rate": 3.043067679686262e-08,
+      "loss": 0.9189,
+      "step": 8985
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6524156062781945,
+      "learning_rate": 3.0239181423058485e-08,
+      "loss": 0.9444,
+      "step": 8986
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.2027497377659473,
+      "learning_rate": 3.004828864988707e-08,
+      "loss": 0.8853,
+      "step": 8987
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8070469209104079,
+      "learning_rate": 2.985799850049487e-08,
+      "loss": 0.8894,
+      "step": 8988
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7739491741500097,
+      "learning_rate": 2.9668310997955085e-08,
+      "loss": 1.0391,
+      "step": 8989
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.627604044868102,
+      "learning_rate": 2.9479226165268216e-08,
+      "loss": 0.9493,
+      "step": 8990
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6394821025515003,
+      "learning_rate": 2.9290744025360916e-08,
+      "loss": 0.8999,
+      "step": 8991
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.753405174210761,
+      "learning_rate": 2.9102864601087133e-08,
+      "loss": 0.9064,
+      "step": 8992
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7882189206375205,
+      "learning_rate": 2.8915587915228638e-08,
+      "loss": 1.0211,
+      "step": 8993
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.710210630472524,
+      "learning_rate": 2.872891399049338e-08,
+      "loss": 0.8593,
+      "step": 8994
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8178562948556511,
+      "learning_rate": 2.8542842849515474e-08,
+      "loss": 0.906,
+      "step": 8995
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8271902199095564,
+      "learning_rate": 2.835737451485687e-08,
+      "loss": 0.8583,
+      "step": 8996
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.126638612163813,
+      "learning_rate": 2.8172509009006808e-08,
+      "loss": 0.857,
+      "step": 8997
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7935405921507497,
+      "learning_rate": 2.798824635438069e-08,
+      "loss": 0.9019,
+      "step": 8998
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.699933654604272,
+      "learning_rate": 2.780458657332008e-08,
+      "loss": 0.9168,
+      "step": 8999
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8058365640916076,
+      "learning_rate": 2.7621529688096062e-08,
+      "loss": 0.8817,
+      "step": 9000
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6488815471621763,
+      "learning_rate": 2.7439075720903653e-08,
+      "loss": 0.8765,
+      "step": 9001
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7473592009089598,
+      "learning_rate": 2.7257224693866268e-08,
+      "loss": 0.8877,
+      "step": 9002
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.227114455321381,
+      "learning_rate": 2.7075976629033496e-08,
+      "loss": 0.9331,
+      "step": 9003
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8472669422075099,
+      "learning_rate": 2.6895331548383864e-08,
+      "loss": 0.8964,
+      "step": 9004
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7208095557403624,
+      "learning_rate": 2.6715289473819296e-08,
+      "loss": 0.8377,
+      "step": 9005
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7639713919724314,
+      "learning_rate": 2.6535850427171774e-08,
+      "loss": 0.8762,
+      "step": 9006
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7659523203589909,
+      "learning_rate": 2.6357014430198892e-08,
+      "loss": 1.0693,
+      "step": 9007
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6515840890174034,
+      "learning_rate": 2.617878150458386e-08,
+      "loss": 0.9298,
+      "step": 9008
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.71372722419767,
+      "learning_rate": 2.6001151671939952e-08,
+      "loss": 0.9311,
+      "step": 9009
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6879951497225332,
+      "learning_rate": 2.582412495380382e-08,
+      "loss": 0.8909,
+      "step": 9010
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6746786627703782,
+      "learning_rate": 2.5647701371641075e-08,
+      "loss": 0.9172,
+      "step": 9011
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8568393935652887,
+      "learning_rate": 2.5471880946844053e-08,
+      "loss": 0.8532,
+      "step": 9012
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.697554459197382,
+      "learning_rate": 2.529666370073125e-08,
+      "loss": 0.9229,
+      "step": 9013
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7776368607503963,
+      "learning_rate": 2.5122049654547897e-08,
+      "loss": 0.9878,
+      "step": 9014
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7011011919752606,
+      "learning_rate": 2.4948038829467613e-08,
+      "loss": 0.8909,
+      "step": 9015
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6866618702751146,
+      "learning_rate": 2.4774631246589075e-08,
+      "loss": 0.8783,
+      "step": 9016
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.671798130577716,
+      "learning_rate": 2.4601826926938245e-08,
+      "loss": 0.8749,
+      "step": 9017
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7867919363608207,
+      "learning_rate": 2.4429625891468912e-08,
+      "loss": 1.0168,
+      "step": 9018
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6524951184308798,
+      "learning_rate": 2.4258028161061042e-08,
+      "loss": 0.8569,
+      "step": 9019
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.857310186828111,
+      "learning_rate": 2.408703375652133e-08,
+      "loss": 0.8681,
+      "step": 9020
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7614645044924335,
+      "learning_rate": 2.391664269858318e-08,
+      "loss": 0.8711,
+      "step": 9021
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7526430011833185,
+      "learning_rate": 2.3746855007907854e-08,
+      "loss": 0.8317,
+      "step": 9022
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8105588881337562,
+      "learning_rate": 2.3577670705081655e-08,
+      "loss": 0.9152,
+      "step": 9023
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8198491536312023,
+      "learning_rate": 2.3409089810618736e-08,
+      "loss": 1.0323,
+      "step": 9024
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6356789329134283,
+      "learning_rate": 2.324111234496107e-08,
+      "loss": 0.8006,
+      "step": 9025
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6402744950731307,
+      "learning_rate": 2.3073738328476258e-08,
+      "loss": 0.8656,
+      "step": 9026
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.712331613595779,
+      "learning_rate": 2.2906967781458066e-08,
+      "loss": 0.9188,
+      "step": 9027
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.695258374561053,
+      "learning_rate": 2.274080072412921e-08,
+      "loss": 0.9149,
+      "step": 9028
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.667315166218039,
+      "learning_rate": 2.2575237176637453e-08,
+      "loss": 0.8553,
+      "step": 9029
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7941765784662156,
+      "learning_rate": 2.2410277159057858e-08,
+      "loss": 0.954,
+      "step": 9030
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6972773660947091,
+      "learning_rate": 2.224592069139275e-08,
+      "loss": 0.8544,
+      "step": 9031
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7603201851553251,
+      "learning_rate": 2.2082167793570642e-08,
+      "loss": 1.0269,
+      "step": 9032
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7972780395656742,
+      "learning_rate": 2.1919018485446753e-08,
+      "loss": 0.8914,
+      "step": 9033
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.713669507627122,
+      "learning_rate": 2.175647278680415e-08,
+      "loss": 0.8558,
+      "step": 9034
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7353618726132312,
+      "learning_rate": 2.159453071735207e-08,
+      "loss": 0.8913,
+      "step": 9035
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7185103497506375,
+      "learning_rate": 2.1433192296725914e-08,
+      "loss": 0.8386,
+      "step": 9036
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.63880072788424,
+      "learning_rate": 2.1272457544488922e-08,
+      "loss": 0.8639,
+      "step": 9037
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7998498321645506,
+      "learning_rate": 2.1112326480131063e-08,
+      "loss": 0.9624,
+      "step": 9038
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7336335606768198,
+      "learning_rate": 2.095279912306847e-08,
+      "loss": 0.9058,
+      "step": 9039
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.857660517657548,
+      "learning_rate": 2.0793875492644e-08,
+      "loss": 0.9399,
+      "step": 9040
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7136757087458103,
+      "learning_rate": 2.0635555608128354e-08,
+      "loss": 0.9564,
+      "step": 9041
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.648849224453651,
+      "learning_rate": 2.0477839488718398e-08,
+      "loss": 0.832,
+      "step": 9042
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.714201623558537,
+      "learning_rate": 2.032072715353717e-08,
+      "loss": 0.9119,
+      "step": 9043
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7478967296912449,
+      "learning_rate": 2.016421862163498e-08,
+      "loss": 0.9167,
+      "step": 9044
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7008623821094497,
+      "learning_rate": 2.0008313911989985e-08,
+      "loss": 0.9065,
+      "step": 9045
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.717681618082401,
+      "learning_rate": 1.9853013043504843e-08,
+      "loss": 0.9012,
+      "step": 9046
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6808209079669376,
+      "learning_rate": 1.969831603501171e-08,
+      "loss": 0.8696,
+      "step": 9047
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7490170896943678,
+      "learning_rate": 1.954422290526725e-08,
+      "loss": 0.9231,
+      "step": 9048
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6329333937599875,
+      "learning_rate": 1.939073367295541e-08,
+      "loss": 0.9354,
+      "step": 9049
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8000361868559058,
+      "learning_rate": 1.9237848356688517e-08,
+      "loss": 0.9261,
+      "step": 9050
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7095769030442776,
+      "learning_rate": 1.9085566975003966e-08,
+      "loss": 0.9077,
+      "step": 9051
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8001885973775742,
+      "learning_rate": 1.893388954636588e-08,
+      "loss": 1.0631,
+      "step": 9052
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7488987883584868,
+      "learning_rate": 1.878281608916621e-08,
+      "loss": 0.8623,
+      "step": 9053
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.79115705751056,
+      "learning_rate": 1.8632346621723085e-08,
+      "loss": 0.7667,
+      "step": 9054
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6895989913953078,
+      "learning_rate": 1.8482481162280797e-08,
+      "loss": 0.8023,
+      "step": 9055
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6880258206829248,
+      "learning_rate": 1.833321972901203e-08,
+      "loss": 0.7568,
+      "step": 9056
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6998083083517264,
+      "learning_rate": 1.8184562340014533e-08,
+      "loss": 0.8463,
+      "step": 9057
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7082504943783157,
+      "learning_rate": 1.8036509013313886e-08,
+      "loss": 0.9124,
+      "step": 9058
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8395185715589875,
+      "learning_rate": 1.7889059766862392e-08,
+      "loss": 0.868,
+      "step": 9059
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7232114719760543,
+      "learning_rate": 1.7742214618537978e-08,
+      "loss": 0.893,
+      "step": 9060
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.732287602766679,
+      "learning_rate": 1.7595973586145842e-08,
+      "loss": 0.8948,
+      "step": 9061
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8373315503179548,
+      "learning_rate": 1.7450336687420132e-08,
+      "loss": 0.8846,
+      "step": 9062
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6988471940477583,
+      "learning_rate": 1.7305303940017838e-08,
+      "loss": 0.8713,
+      "step": 9063
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6154185306074311,
+      "learning_rate": 1.716087536152544e-08,
+      "loss": 0.9268,
+      "step": 9064
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6956034955472434,
+      "learning_rate": 1.701705096945505e-08,
+      "loss": 0.8376,
+      "step": 9065
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7905778417450919,
+      "learning_rate": 1.6873830781246603e-08,
+      "loss": 1.0475,
+      "step": 9066
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6930014888452642,
+      "learning_rate": 1.673121481426565e-08,
+      "loss": 0.8943,
+      "step": 9067
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7385071820795834,
+      "learning_rate": 1.6589203085804473e-08,
+      "loss": 0.8992,
+      "step": 9068
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.807488707910084,
+      "learning_rate": 1.6447795613083185e-08,
+      "loss": 0.9414,
+      "step": 9069
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8712453961363233,
+      "learning_rate": 1.6306992413247512e-08,
+      "loss": 0.9696,
+      "step": 9070
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6057756238716798,
+      "learning_rate": 1.616679350337047e-08,
+      "loss": 0.8964,
+      "step": 9071
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7403326619418713,
+      "learning_rate": 1.602719890045179e-08,
+      "loss": 0.9162,
+      "step": 9072
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.742054600595073,
+      "learning_rate": 1.5888208621417377e-08,
+      "loss": 0.9154,
+      "step": 9073
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7853893303960242,
+      "learning_rate": 1.574982268312042e-08,
+      "loss": 0.9125,
+      "step": 9074
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.83761020856176,
+      "learning_rate": 1.561204110234138e-08,
+      "loss": 0.9192,
+      "step": 9075
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7089640134701365,
+      "learning_rate": 1.547486389578523e-08,
+      "loss": 0.8141,
+      "step": 9076
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7827373941860172,
+      "learning_rate": 1.5338291080086444e-08,
+      "loss": 0.8751,
+      "step": 9077
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.861494262388883,
+      "learning_rate": 1.5202322671805104e-08,
+      "loss": 0.913,
+      "step": 9078
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7785025146223936,
+      "learning_rate": 1.5066958687426915e-08,
+      "loss": 0.9149,
+      "step": 9079
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8103344790437834,
+      "learning_rate": 1.493219914336541e-08,
+      "loss": 0.8377,
+      "step": 9080
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7982179672632662,
+      "learning_rate": 1.4798044055961414e-08,
+      "loss": 1.0349,
+      "step": 9081
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.9191032215125934,
+      "learning_rate": 1.4664493441480798e-08,
+      "loss": 0.9087,
+      "step": 9082
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8226221196107963,
+      "learning_rate": 1.4531547316117833e-08,
+      "loss": 0.8079,
+      "step": 9083
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7944971360217326,
+      "learning_rate": 1.4399205695991847e-08,
+      "loss": 0.9084,
+      "step": 9084
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.723446214042162,
+      "learning_rate": 1.4267468597150558e-08,
+      "loss": 0.902,
+      "step": 9085
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.758819129178057,
+      "learning_rate": 1.4136336035566744e-08,
+      "loss": 0.8155,
+      "step": 9086
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.690446515285022,
+      "learning_rate": 1.4005808027141576e-08,
+      "loss": 0.8442,
+      "step": 9087
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7178650777181188,
+      "learning_rate": 1.3875884587700727e-08,
+      "loss": 0.9156,
+      "step": 9088
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8440204890789689,
+      "learning_rate": 1.3746565732999372e-08,
+      "loss": 0.911,
+      "step": 9089
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8831473392886526,
+      "learning_rate": 1.3617851478716637e-08,
+      "loss": 0.88,
+      "step": 9090
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.738609579311347,
+      "learning_rate": 1.3489741840460591e-08,
+      "loss": 0.8092,
+      "step": 9091
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7699042220963042,
+      "learning_rate": 1.3362236833763809e-08,
+      "loss": 0.9627,
+      "step": 9092
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7119957305818998,
+      "learning_rate": 1.3235336474087812e-08,
+      "loss": 0.89,
+      "step": 9093
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7283917105327136,
+      "learning_rate": 1.3109040776819181e-08,
+      "loss": 0.9074,
+      "step": 9094
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7809673917823186,
+      "learning_rate": 1.2983349757271779e-08,
+      "loss": 1.0231,
+      "step": 9095
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9068419779083436,
+      "learning_rate": 1.285826343068619e-08,
+      "loss": 0.8768,
+      "step": 9096
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8644712209673977,
+      "learning_rate": 1.2733781812229728e-08,
+      "loss": 0.9349,
+      "step": 9097
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.825391201906082,
+      "learning_rate": 1.2609904916995319e-08,
+      "loss": 0.9274,
+      "step": 9098
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8177325574104135,
+      "learning_rate": 1.248663276000428e-08,
+      "loss": 0.8698,
+      "step": 9099
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5993057887215836,
+      "learning_rate": 1.23639653562041e-08,
+      "loss": 0.9245,
+      "step": 9100
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7717911843510411,
+      "learning_rate": 1.2241902720467325e-08,
+      "loss": 1.07,
+      "step": 9101
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6821630528082228,
+      "learning_rate": 1.2120444867596003e-08,
+      "loss": 0.8692,
+      "step": 9102
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7746723796187034,
+      "learning_rate": 1.1999591812316136e-08,
+      "loss": 0.9126,
+      "step": 9103
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8492099322470268,
+      "learning_rate": 1.1879343569282109e-08,
+      "loss": 0.9184,
+      "step": 9104
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7623791065358076,
+      "learning_rate": 1.1759700153073928e-08,
+      "loss": 0.9777,
+      "step": 9105
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7654244386832343,
+      "learning_rate": 1.1640661578199985e-08,
+      "loss": 0.8186,
+      "step": 9106
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7304830298380984,
+      "learning_rate": 1.1522227859092627e-08,
+      "loss": 0.9392,
+      "step": 9107
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.690768074159443,
+      "learning_rate": 1.1404399010113698e-08,
+      "loss": 0.8725,
+      "step": 9108
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7280855868652891,
+      "learning_rate": 1.1287175045548993e-08,
+      "loss": 0.881,
+      "step": 9109
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6950658892271546,
+      "learning_rate": 1.117055597961325e-08,
+      "loss": 0.8178,
+      "step": 9110
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7151001594030106,
+      "learning_rate": 1.105454182644683e-08,
+      "loss": 0.8519,
+      "step": 9111
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7004414943541661,
+      "learning_rate": 1.0939132600116808e-08,
+      "loss": 0.9615,
+      "step": 9112
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.606747081021208,
+      "learning_rate": 1.0824328314616995e-08,
+      "loss": 0.8498,
+      "step": 9113
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6274400127430584,
+      "learning_rate": 1.0710128983867364e-08,
+      "loss": 0.8736,
+      "step": 9114
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6272621288648101,
+      "learning_rate": 1.0596534621715171e-08,
+      "loss": 0.9128,
+      "step": 9115
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7189807665646388,
+      "learning_rate": 1.0483545241934401e-08,
+      "loss": 0.8904,
+      "step": 9116
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8173224840161402,
+      "learning_rate": 1.037116085822576e-08,
+      "loss": 0.9057,
+      "step": 9117
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7270116196563388,
+      "learning_rate": 1.0259381484215014e-08,
+      "loss": 0.8728,
+      "step": 9118
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7185204396194957,
+      "learning_rate": 1.0148207133456877e-08,
+      "loss": 0.8975,
+      "step": 9119
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6422698095934387,
+      "learning_rate": 1.0037637819431123e-08,
+      "loss": 0.9116,
+      "step": 9120
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7174578228824693,
+      "learning_rate": 9.927673555544804e-09,
+      "loss": 0.8597,
+      "step": 9121
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7745232725329894,
+      "learning_rate": 9.818314355131697e-09,
+      "loss": 0.8474,
+      "step": 9122
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.799376446663371,
+      "learning_rate": 9.709560231451198e-09,
+      "loss": 0.8533,
+      "step": 9123
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7464341523903155,
+      "learning_rate": 9.60141119769109e-09,
+      "loss": 0.9681,
+      "step": 9124
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9377358995752976,
+      "learning_rate": 9.493867266964219e-09,
+      "loss": 0.8759,
+      "step": 9125
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6861470075195868,
+      "learning_rate": 9.386928452310707e-09,
+      "loss": 0.8306,
+      "step": 9126
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.70984096912081,
+      "learning_rate": 9.280594766697959e-09,
+      "loss": 0.8176,
+      "step": 9127
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6742680743120448,
+      "learning_rate": 9.174866223018441e-09,
+      "loss": 0.7947,
+      "step": 9128
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9394316850888744,
+      "learning_rate": 9.069742834092454e-09,
+      "loss": 0.9579,
+      "step": 9129
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7993923890408288,
+      "learning_rate": 8.965224612665912e-09,
+      "loss": 0.866,
+      "step": 9130
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7447403404168589,
+      "learning_rate": 8.861311571413122e-09,
+      "loss": 0.8641,
+      "step": 9131
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7850976512753851,
+      "learning_rate": 8.758003722933451e-09,
+      "loss": 1.0539,
+      "step": 9132
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7658521117293198,
+      "learning_rate": 8.655301079752986e-09,
+      "loss": 0.8853,
+      "step": 9133
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6773223351139095,
+      "learning_rate": 8.553203654325104e-09,
+      "loss": 0.8556,
+      "step": 9134
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8132855294922835,
+      "learning_rate": 8.451711459029343e-09,
+      "loss": 1.04,
+      "step": 9135
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6863286563636097,
+      "learning_rate": 8.350824506172528e-09,
+      "loss": 0.8706,
+      "step": 9136
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.65650098637169,
+      "learning_rate": 8.250542807986538e-09,
+      "loss": 0.8251,
+      "step": 9137
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.741410215587374,
+      "learning_rate": 8.15086637663165e-09,
+      "loss": 0.8967,
+      "step": 9138
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.982185707419509,
+      "learning_rate": 8.051795224193748e-09,
+      "loss": 0.8117,
+      "step": 9139
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7479311987933726,
+      "learning_rate": 7.953329362685447e-09,
+      "loss": 0.8447,
+      "step": 9140
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7109635207171099,
+      "learning_rate": 7.855468804046085e-09,
+      "loss": 0.8834,
+      "step": 9141
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.682281990730583,
+      "learning_rate": 7.758213560141726e-09,
+      "loss": 0.8489,
+      "step": 9142
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.784424990257666,
+      "learning_rate": 7.661563642765158e-09,
+      "loss": 0.9257,
+      "step": 9143
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.75055616743191,
+      "learning_rate": 7.565519063634785e-09,
+      "loss": 0.9217,
+      "step": 9144
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7264691161134966,
+      "learning_rate": 7.470079834396849e-09,
+      "loss": 0.9823,
+      "step": 9145
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7127240808791075,
+      "learning_rate": 7.375245966623757e-09,
+      "loss": 0.8854,
+      "step": 9146
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7692033642720633,
+      "learning_rate": 7.281017471814089e-09,
+      "loss": 0.9256,
+      "step": 9147
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.654493168379517,
+      "learning_rate": 7.187394361393707e-09,
+      "loss": 0.9111,
+      "step": 9148
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.829072790970889,
+      "learning_rate": 7.09437664671464e-09,
+      "loss": 0.9651,
+      "step": 9149
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8457075067193203,
+      "learning_rate": 7.001964339055644e-09,
+      "loss": 0.8753,
+      "step": 9150
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.820544486890795,
+      "learning_rate": 6.910157449621646e-09,
+      "loss": 0.9111,
+      "step": 9151
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6908804980856922,
+      "learning_rate": 6.818955989545406e-09,
+      "loss": 0.8361,
+      "step": 9152
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7754779786135602,
+      "learning_rate": 6.728359969884191e-09,
+      "loss": 0.8911,
+      "step": 9153
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6739641787319068,
+      "learning_rate": 6.638369401624212e-09,
+      "loss": 0.914,
+      "step": 9154
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.578525368276803,
+      "learning_rate": 6.548984295676741e-09,
+      "loss": 0.8516,
+      "step": 9155
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6925381320760478,
+      "learning_rate": 6.4602046628797765e-09,
+      "loss": 0.9197,
+      "step": 9156
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.685470067717621,
+      "learning_rate": 6.37203051399804e-09,
+      "loss": 0.8734,
+      "step": 9157
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7825497932004423,
+      "learning_rate": 6.284461859723534e-09,
+      "loss": 1.0306,
+      "step": 9158
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6005891106138797,
+      "learning_rate": 6.197498710673877e-09,
+      "loss": 0.83,
+      "step": 9159
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6659898542888691,
+      "learning_rate": 6.111141077393967e-09,
+      "loss": 0.8889,
+      "step": 9160
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7843773793192037,
+      "learning_rate": 6.025388970354873e-09,
+      "loss": 0.9343,
+      "step": 9161
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.709312990622651,
+      "learning_rate": 5.940242399953833e-09,
+      "loss": 0.8766,
+      "step": 9162
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7987757016052872,
+      "learning_rate": 5.855701376515921e-09,
+      "loss": 0.8605,
+      "step": 9163
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7417187056297736,
+      "learning_rate": 5.771765910291272e-09,
+      "loss": 0.8628,
+      "step": 9164
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6695044749664891,
+      "learning_rate": 5.688436011457854e-09,
+      "loss": 0.8849,
+      "step": 9165
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7464852151295442,
+      "learning_rate": 5.605711690119808e-09,
+      "loss": 0.8689,
+      "step": 9166
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7433133950511028,
+      "learning_rate": 5.5235929563074446e-09,
+      "loss": 0.9274,
+      "step": 9167
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7687967092378002,
+      "learning_rate": 5.442079819977797e-09,
+      "loss": 1.0319,
+      "step": 9168
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7533186276697215,
+      "learning_rate": 5.361172291014627e-09,
+      "loss": 1.0757,
+      "step": 9169
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.803827272520431,
+      "learning_rate": 5.280870379228975e-09,
+      "loss": 0.9688,
+      "step": 9170
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8319383936184692,
+      "learning_rate": 5.201174094356942e-09,
+      "loss": 0.8795,
+      "step": 9171
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7917142422047312,
+      "learning_rate": 5.122083446062464e-09,
+      "loss": 1.0606,
+      "step": 9172
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7444879660423172,
+      "learning_rate": 5.043598443935094e-09,
+      "loss": 0.9081,
+      "step": 9173
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5490749042177736,
+      "learning_rate": 4.965719097491661e-09,
+      "loss": 0.8728,
+      "step": 9174
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6796411304409415,
+      "learning_rate": 4.888445416175725e-09,
+      "loss": 0.8342,
+      "step": 9175
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6459189189222507,
+      "learning_rate": 4.8117774093559e-09,
+      "loss": 0.9059,
+      "step": 9176
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8093140156129675,
+      "learning_rate": 4.7357150863291955e-09,
+      "loss": 0.863,
+      "step": 9177
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7868077251150232,
+      "learning_rate": 4.660258456318789e-09,
+      "loss": 0.8175,
+      "step": 9178
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7996668198164496,
+      "learning_rate": 4.5854075284729184e-09,
+      "loss": 0.9129,
+      "step": 9179
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5685924330054664,
+      "learning_rate": 4.5111623118687666e-09,
+      "loss": 0.8879,
+      "step": 9180
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7122473341977509,
+      "learning_rate": 4.437522815508577e-09,
+      "loss": 0.9346,
+      "step": 9181
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7153049102960125,
+      "learning_rate": 4.3644890483202086e-09,
+      "loss": 0.8689,
+      "step": 9182
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6551719665955429,
+      "learning_rate": 4.292061019160465e-09,
+      "loss": 0.8898,
+      "step": 9183
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6802026216778958,
+      "learning_rate": 4.2202387368112105e-09,
+      "loss": 0.8848,
+      "step": 9184
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7901308050121532,
+      "learning_rate": 4.149022209981035e-09,
+      "loss": 0.9608,
+      "step": 9185
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7099148401147537,
+      "learning_rate": 4.078411447305253e-09,
+      "loss": 0.8323,
+      "step": 9186
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6697862061336277,
+      "learning_rate": 4.008406457345904e-09,
+      "loss": 0.8961,
+      "step": 9187
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7393976808328753,
+      "learning_rate": 3.939007248590643e-09,
+      "loss": 0.8644,
+      "step": 9188
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8592704895006116,
+      "learning_rate": 3.870213829454961e-09,
+      "loss": 0.9447,
+      "step": 9189
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.752231416735354,
+      "learning_rate": 3.802026208279963e-09,
+      "loss": 0.9468,
+      "step": 9190
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7003641237682152,
+      "learning_rate": 3.734444393334036e-09,
+      "loss": 0.8536,
+      "step": 9191
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7829653479416725,
+      "learning_rate": 3.66746839281118e-09,
+      "loss": 0.8838,
+      "step": 9192
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7774107316254032,
+      "learning_rate": 3.601098214833232e-09,
+      "loss": 0.986,
+      "step": 9193
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.724201884433632,
+      "learning_rate": 3.5353338674470883e-09,
+      "loss": 0.9272,
+      "step": 9194
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.791735323561752,
+      "learning_rate": 3.4701753586269262e-09,
+      "loss": 0.8899,
+      "step": 9195
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.8046994858282555,
+      "learning_rate": 3.4056226962742024e-09,
+      "loss": 1.0684,
+      "step": 9196
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6199620216039676,
+      "learning_rate": 3.3416758882154344e-09,
+      "loss": 0.8862,
+      "step": 9197
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.694106765170152,
+      "learning_rate": 3.2783349422044197e-09,
+      "loss": 0.8702,
+      "step": 9198
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7415696415325952,
+      "learning_rate": 3.2155998659211265e-09,
+      "loss": 0.8989,
+      "step": 9199
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6724028420362715,
+      "learning_rate": 3.153470666973357e-09,
+      "loss": 0.8568,
+      "step": 9200
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.741978350469708,
+      "learning_rate": 3.091947352893976e-09,
+      "loss": 0.864,
+      "step": 9201
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8347308123330335,
+      "learning_rate": 3.031029931143126e-09,
+      "loss": 0.9434,
+      "step": 9202
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7361159595619056,
+      "learning_rate": 2.9707184091071206e-09,
+      "loss": 0.888,
+      "step": 9203
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6745277667295864,
+      "learning_rate": 2.9110127940984447e-09,
+      "loss": 0.7828,
+      "step": 9204
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8624210834481494,
+      "learning_rate": 2.8519130933574168e-09,
+      "loss": 0.8727,
+      "step": 9205
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9420242658487823,
+      "learning_rate": 2.7934193140499723e-09,
+      "loss": 0.9323,
+      "step": 9206
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.668156707492051,
+      "learning_rate": 2.73553146326766e-09,
+      "loss": 0.8295,
+      "step": 9207
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.75298747884214,
+      "learning_rate": 2.6782495480309757e-09,
+      "loss": 0.8856,
+      "step": 9208
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7328517413471247,
+      "learning_rate": 2.621573575284919e-09,
+      "loss": 0.8712,
+      "step": 9209
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7675900322948337,
+      "learning_rate": 2.565503551901216e-09,
+      "loss": 1.034,
+      "step": 9210
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7147983652449887,
+      "learning_rate": 2.5100394846794272e-09,
+      "loss": 0.8474,
+      "step": 9211
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8194909961322767,
+      "learning_rate": 2.455181380344174e-09,
+      "loss": 0.8328,
+      "step": 9212
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6677750753663276,
+      "learning_rate": 2.4009292455468013e-09,
+      "loss": 0.8632,
+      "step": 9213
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8015452141650738,
+      "learning_rate": 2.347283086867047e-09,
+      "loss": 0.884,
+      "step": 9214
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7053056673751479,
+      "learning_rate": 2.294242910808042e-09,
+      "loss": 0.8569,
+      "step": 9215
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7870679862855685,
+      "learning_rate": 2.2418087238018638e-09,
+      "loss": 1.0511,
+      "step": 9216
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7725984634506644,
+      "learning_rate": 2.189980532206759e-09,
+      "loss": 0.8602,
+      "step": 9217
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7162118703527607,
+      "learning_rate": 2.1387583423060354e-09,
+      "loss": 0.8651,
+      "step": 9218
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7869090857516297,
+      "learning_rate": 2.0881421603113904e-09,
+      "loss": 0.8646,
+      "step": 9219
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8112580336366757,
+      "learning_rate": 2.038131992359582e-09,
+      "loss": 0.84,
+      "step": 9220
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7552453134607935,
+      "learning_rate": 1.9887278445152037e-09,
+      "loss": 0.9001,
+      "step": 9221
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.770616304493373,
+      "learning_rate": 1.9399297227684634e-09,
+      "loss": 1.0231,
+      "step": 9222
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7410290288123016,
+      "learning_rate": 1.8917376330357395e-09,
+      "loss": 0.8115,
+      "step": 9223
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.783793368000215,
+      "learning_rate": 1.8441515811612465e-09,
+      "loss": 0.9071,
+      "step": 9224
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.718071172083732,
+      "learning_rate": 1.7971715729148131e-09,
+      "loss": 0.8739,
+      "step": 9225
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.689541569774579,
+      "learning_rate": 1.7507976139924388e-09,
+      "loss": 0.845,
+      "step": 9226
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5944375630462957,
+      "learning_rate": 1.7050297100174029e-09,
+      "loss": 0.8899,
+      "step": 9227
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8896529716412072,
+      "learning_rate": 1.6598678665397106e-09,
+      "loss": 0.9372,
+      "step": 9228
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7883531679192002,
+      "learning_rate": 1.6153120890344266e-09,
+      "loss": 0.8629,
+      "step": 9229
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.688139275768094,
+      "learning_rate": 1.5713623829050063e-09,
+      "loss": 0.8625,
+      "step": 9230
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.769601675892439,
+      "learning_rate": 1.52801875347941e-09,
+      "loss": 0.9351,
+      "step": 9231
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.686950453235554,
+      "learning_rate": 1.4852812060145438e-09,
+      "loss": 0.8068,
+      "step": 9232
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6115900019069742,
+      "learning_rate": 1.4431497456918186e-09,
+      "loss": 0.8802,
+      "step": 9233
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.685260222683556,
+      "learning_rate": 1.4016243776193705e-09,
+      "loss": 0.8976,
+      "step": 9234
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7676793964300495,
+      "learning_rate": 1.3607051068331712e-09,
+      "loss": 0.884,
+      "step": 9235
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6721572400314098,
+      "learning_rate": 1.320391938294252e-09,
+      "loss": 0.8607,
+      "step": 9236
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7288428051240505,
+      "learning_rate": 1.2806848768909252e-09,
+      "loss": 0.9181,
+      "step": 9237
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7248314516585734,
+      "learning_rate": 1.241583927437673e-09,
+      "loss": 0.7547,
+      "step": 9238
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8152562821450142,
+      "learning_rate": 1.2030890946757024e-09,
+      "loss": 0.9568,
+      "step": 9239
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7129744543587961,
+      "learning_rate": 1.1652003832729464e-09,
+      "loss": 0.862,
+      "step": 9240
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6826210880891392,
+      "learning_rate": 1.1279177978229527e-09,
+      "loss": 0.7989,
+      "step": 9241
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8414224128793462,
+      "learning_rate": 1.0912413428471046e-09,
+      "loss": 0.9147,
+      "step": 9242
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6660388927677798,
+      "learning_rate": 1.0551710227912904e-09,
+      "loss": 0.8029,
+      "step": 9243
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7712005816402376,
+      "learning_rate": 1.0197068420308986e-09,
+      "loss": 0.9259,
+      "step": 9244
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7562717522451998,
+      "learning_rate": 9.84848804864713e-10,
+      "loss": 0.8605,
+      "step": 9245
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7626233880234385,
+      "learning_rate": 9.505969155193528e-10,
+      "loss": 0.8751,
+      "step": 9246
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7657631848052613,
+      "learning_rate": 9.169511781492724e-10,
+      "loss": 1.0402,
+      "step": 9247
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6212134503198508,
+      "learning_rate": 8.83911596832876e-10,
+      "loss": 0.8869,
+      "step": 9248
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8614988222888011,
+      "learning_rate": 8.514781755769586e-10,
+      "loss": 0.8919,
+      "step": 9249
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.803076439215676,
+      "learning_rate": 8.196509183139301e-10,
+      "loss": 0.9004,
+      "step": 9250
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6903965231622837,
+      "learning_rate": 7.884298289029258e-10,
+      "loss": 0.8651,
+      "step": 9251
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.810782489416566,
+      "learning_rate": 7.578149111298061e-10,
+      "loss": 0.9026,
+      "step": 9252
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.114020211513277,
+      "learning_rate": 7.278061687066018e-10,
+      "loss": 0.9181,
+      "step": 9253
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7613803320428254,
+      "learning_rate": 6.984036052720688e-10,
+      "loss": 0.8922,
+      "step": 9254
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.691582899743085,
+      "learning_rate": 6.696072243911333e-10,
+      "loss": 0.9479,
+      "step": 9255
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.618488145780062,
+      "learning_rate": 6.414170295560019e-10,
+      "loss": 0.8667,
+      "step": 9256
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6461631390311573,
+      "learning_rate": 6.138330241839408e-10,
+      "loss": 0.8966,
+      "step": 9257
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6940069990940216,
+      "learning_rate": 5.868552116206072e-10,
+      "loss": 0.837,
+      "step": 9258
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.712907037219412,
+      "learning_rate": 5.604835951367183e-10,
+      "loss": 0.9584,
+      "step": 9259
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.80865466947985,
+      "learning_rate": 5.347181779297162e-10,
+      "loss": 0.8905,
+      "step": 9260
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6943266445550123,
+      "learning_rate": 5.095589631237686e-10,
+      "loss": 0.7743,
+      "step": 9261
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6516714029218185,
+      "learning_rate": 4.850059537697682e-10,
+      "loss": 0.9337,
+      "step": 9262
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6778305443568975,
+      "learning_rate": 4.610591528447783e-10,
+      "loss": 0.9002,
+      "step": 9263
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7855790963533185,
+      "learning_rate": 4.37718563252032e-10,
+      "loss": 0.9147,
+      "step": 9264
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7179218980857869,
+      "learning_rate": 4.14984187822598e-10,
+      "loss": 0.9487,
+      "step": 9265
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7832879426405934,
+      "learning_rate": 3.928560293120498e-10,
+      "loss": 0.8127,
+      "step": 9266
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6660065315784636,
+      "learning_rate": 3.7133409040435165e-10,
+      "loss": 0.8945,
+      "step": 9267
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.8032130897953331,
+      "learning_rate": 3.5041837370852763e-10,
+      "loss": 0.8849,
+      "step": 9268
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7439199649405395,
+      "learning_rate": 3.3010888176088217e-10,
+      "loss": 0.9488,
+      "step": 9269
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6366513913097684,
+      "learning_rate": 3.1040561702444517e-10,
+      "loss": 0.8299,
+      "step": 9270
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7554527824213058,
+      "learning_rate": 2.913085818873063e-10,
+      "loss": 0.9064,
+      "step": 9271
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7596989810085981,
+      "learning_rate": 2.7281777866594584e-10,
+      "loss": 0.9059,
+      "step": 9272
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.614690178679153,
+      "learning_rate": 2.5493320960190413e-10,
+      "loss": 0.8463,
+      "step": 9273
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7325704202541454,
+      "learning_rate": 2.376548768640019e-10,
+      "loss": 0.8454,
+      "step": 9274
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.71311509657776,
+      "learning_rate": 2.2098278254722993e-10,
+      "loss": 0.7916,
+      "step": 9275
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6898908948729612,
+      "learning_rate": 2.0491692867330438e-10,
+      "loss": 0.9123,
+      "step": 9276
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6649349260250577,
+      "learning_rate": 1.8945731719011152e-10,
+      "loss": 0.9142,
+      "step": 9277
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5917590118855425,
+      "learning_rate": 1.7460394997170783e-10,
+      "loss": 0.8329,
+      "step": 9278
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6164112088879237,
+      "learning_rate": 1.6035682881998526e-10,
+      "loss": 0.8622,
+      "step": 9279
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7845646856090713,
+      "learning_rate": 1.4671595546245087e-10,
+      "loss": 0.872,
+      "step": 9280
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.795501175954851,
+      "learning_rate": 1.3368133155222673e-10,
+      "loss": 0.8339,
+      "step": 9281
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7800859391995763,
+      "learning_rate": 1.2125295867027043e-10,
+      "loss": 0.9319,
+      "step": 9282
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7562846578273954,
+      "learning_rate": 1.0943083832370971e-10,
+      "loss": 0.8798,
+      "step": 9283
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.750762004855234,
+      "learning_rate": 9.821497194584251e-11,
+      "loss": 0.8722,
+      "step": 9284
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.737622442790977,
+      "learning_rate": 8.760536089724713e-11,
+      "loss": 0.8604,
+      "step": 9285
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6217067216408663,
+      "learning_rate": 7.760200646300675e-11,
+      "loss": 0.8714,
+      "step": 9286
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7702148068771453,
+      "learning_rate": 6.820490985715022e-11,
+      "loss": 0.8923,
+      "step": 9287
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7590042620966585,
+      "learning_rate": 5.941407221932149e-11,
+      "loss": 1.0117,
+      "step": 9288
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7517794323985483,
+      "learning_rate": 5.122949461422444e-11,
+      "loss": 1.0238,
+      "step": 9289
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.78198599112025,
+      "learning_rate": 4.365117803550867e-11,
+      "loss": 0.8927,
+      "step": 9290
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.8005681139348992,
+      "learning_rate": 3.667912340132862e-11,
+      "loss": 0.8684,
+      "step": 9291
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.8270837844191012,
+      "learning_rate": 3.031333155767424e-11,
+      "loss": 0.9145,
+      "step": 9292
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7498665486363356,
+      "learning_rate": 2.4553803275595422e-11,
+      "loss": 0.8724,
+      "step": 9293
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6695096817029753,
+      "learning_rate": 1.9400539253977558e-11,
+      "loss": 0.9088,
+      "step": 9294
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6716326178141856,
+      "learning_rate": 1.4853540117321098e-11,
+      "loss": 0.9037,
+      "step": 9295
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6850984595928384,
+      "learning_rate": 1.0912806417961997e-11,
+      "loss": 0.8323,
+      "step": 9296
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.734363032129179,
+      "learning_rate": 7.578338632741045e-12,
+      "loss": 0.8643,
+      "step": 9297
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7727923476558051,
+      "learning_rate": 4.850137165779423e-12,
+      "loss": 1.0634,
+      "step": 9298
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7799045450905264,
+      "learning_rate": 2.728202349033815e-12,
+      "loss": 0.8718,
+      "step": 9299
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7925661467446528,
+      "learning_rate": 1.2125344384106285e-12,
+      "loss": 0.9312,
+      "step": 9300
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7025758032701872,
+      "learning_rate": 3.031336187619971e-13,
+      "loss": 0.9755,
+      "step": 9301
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4419020241733236,
+      "learning_rate": 0.0,
+      "loss": 0.9002,
+      "step": 9302
+    },
+    {
+      "epoch": 1.0,
+      "step": 9302,
+      "total_flos": 4294892380667904.0,
+      "train_loss": 0.9256457089672958,
+      "train_runtime": 61794.9627,
+      "train_samples_per_second": 38.538,
+      "train_steps_per_second": 0.151
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 9302,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 1000,
+  "total_flos": 4294892380667904.0,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}