diff --git "a/checkpoint-2000/trainer_state.json" "b/checkpoint-2000/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-2000/trainer_state.json"
@@ -0,0 +1,14033 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.3731548232063164,
+  "eval_steps": 500,
+  "global_step": 2000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0006865774116031583,
+      "grad_norm": 2.40625,
+      "learning_rate": 1.1363636363636364e-07,
+      "loss": 0.3846,
+      "step": 1
+    },
+    {
+      "epoch": 0.0013731548232063166,
+      "grad_norm": 2.3125,
+      "learning_rate": 2.2727272727272729e-07,
+      "loss": 0.3841,
+      "step": 2
+    },
+    {
+      "epoch": 0.0020597322348094747,
+      "grad_norm": 2.40625,
+      "learning_rate": 3.409090909090909e-07,
+      "loss": 0.3764,
+      "step": 3
+    },
+    {
+      "epoch": 0.0027463096464126332,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.5454545454545457e-07,
+      "loss": 0.3733,
+      "step": 4
+    },
+    {
+      "epoch": 0.0034328870580157913,
+      "grad_norm": 2.765625,
+      "learning_rate": 5.681818181818182e-07,
+      "loss": 0.3723,
+      "step": 5
+    },
+    {
+      "epoch": 0.004119464469618949,
+      "grad_norm": 2.796875,
+      "learning_rate": 6.818181818181818e-07,
+      "loss": 0.3946,
+      "step": 6
+    },
+    {
+      "epoch": 0.004806041881222108,
+      "grad_norm": 2.8125,
+      "learning_rate": 7.954545454545455e-07,
+      "loss": 0.3755,
+      "step": 7
+    },
+    {
+      "epoch": 0.0054926192928252664,
+      "grad_norm": 2.859375,
+      "learning_rate": 9.090909090909091e-07,
+      "loss": 0.3977,
+      "step": 8
+    },
+    {
+      "epoch": 0.006179196704428424,
+      "grad_norm": 2.90625,
+      "learning_rate": 1.0227272727272729e-06,
+      "loss": 0.3898,
+      "step": 9
+    },
+    {
+      "epoch": 0.006865774116031583,
+      "grad_norm": 2.59375,
+      "learning_rate": 1.1363636363636364e-06,
+      "loss": 0.3905,
+      "step": 10
+    },
+    {
+      "epoch": 0.007552351527634741,
+      "grad_norm": 2.890625,
+      "learning_rate": 1.25e-06,
+      "loss": 0.3924,
+      "step": 11
+    },
+    {
+      "epoch": 0.008238928939237899,
+      "grad_norm": 2.53125,
+      "learning_rate": 1.3636363636363636e-06,
+      "loss": 0.3748,
+      "step": 12
+    },
+    {
+      "epoch": 0.008925506350841057,
+      "grad_norm": 2.59375,
+      "learning_rate": 1.4772727272727275e-06,
+      "loss": 0.3694,
+      "step": 13
+    },
+    {
+      "epoch": 0.009612083762444216,
+      "grad_norm": 2.90625,
+      "learning_rate": 1.590909090909091e-06,
+      "loss": 0.3876,
+      "step": 14
+    },
+    {
+      "epoch": 0.010298661174047374,
+      "grad_norm": 2.34375,
+      "learning_rate": 1.7045454545454546e-06,
+      "loss": 0.3751,
+      "step": 15
+    },
+    {
+      "epoch": 0.010985238585650533,
+      "grad_norm": 2.609375,
+      "learning_rate": 1.8181818181818183e-06,
+      "loss": 0.3687,
+      "step": 16
+    },
+    {
+      "epoch": 0.01167181599725369,
+      "grad_norm": 2.359375,
+      "learning_rate": 1.931818181818182e-06,
+      "loss": 0.3853,
+      "step": 17
+    },
+    {
+      "epoch": 0.012358393408856848,
+      "grad_norm": 2.3125,
+      "learning_rate": 2.0454545454545457e-06,
+      "loss": 0.361,
+      "step": 18
+    },
+    {
+      "epoch": 0.013044970820460007,
+      "grad_norm": 2.3125,
+      "learning_rate": 2.1590909090909092e-06,
+      "loss": 0.3646,
+      "step": 19
+    },
+    {
+      "epoch": 0.013731548232063165,
+      "grad_norm": 2.171875,
+      "learning_rate": 2.2727272727272728e-06,
+      "loss": 0.3737,
+      "step": 20
+    },
+    {
+      "epoch": 0.014418125643666324,
+      "grad_norm": 2.078125,
+      "learning_rate": 2.3863636363636367e-06,
+      "loss": 0.3679,
+      "step": 21
+    },
+    {
+      "epoch": 0.015104703055269482,
+      "grad_norm": 1.90625,
+      "learning_rate": 2.5e-06,
+      "loss": 0.3576,
+      "step": 22
+    },
+    {
+      "epoch": 0.01579128046687264,
+      "grad_norm": 2.25,
+      "learning_rate": 2.6136363636363637e-06,
+      "loss": 0.361,
+      "step": 23
+    },
+    {
+      "epoch": 0.016477857878475798,
+      "grad_norm": 2.171875,
+      "learning_rate": 2.7272727272727272e-06,
+      "loss": 0.36,
+      "step": 24
+    },
+    {
+      "epoch": 0.017164435290078956,
+      "grad_norm": 1.765625,
+      "learning_rate": 2.8409090909090916e-06,
+      "loss": 0.351,
+      "step": 25
+    },
+    {
+      "epoch": 0.017851012701682115,
+      "grad_norm": 1.6875,
+      "learning_rate": 2.954545454545455e-06,
+      "loss": 0.3357,
+      "step": 26
+    },
+    {
+      "epoch": 0.018537590113285273,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.0681818181818186e-06,
+      "loss": 0.3369,
+      "step": 27
+    },
+    {
+      "epoch": 0.01922416752488843,
+      "grad_norm": 1.8203125,
+      "learning_rate": 3.181818181818182e-06,
+      "loss": 0.3419,
+      "step": 28
+    },
+    {
+      "epoch": 0.01991074493649159,
+      "grad_norm": 1.78125,
+      "learning_rate": 3.2954545454545456e-06,
+      "loss": 0.3258,
+      "step": 29
+    },
+    {
+      "epoch": 0.02059732234809475,
+      "grad_norm": 1.7109375,
+      "learning_rate": 3.409090909090909e-06,
+      "loss": 0.3213,
+      "step": 30
+    },
+    {
+      "epoch": 0.021283899759697907,
+      "grad_norm": 1.5546875,
+      "learning_rate": 3.522727272727273e-06,
+      "loss": 0.3325,
+      "step": 31
+    },
+    {
+      "epoch": 0.021970477171301066,
+      "grad_norm": 1.4921875,
+      "learning_rate": 3.6363636363636366e-06,
+      "loss": 0.3298,
+      "step": 32
+    },
+    {
+      "epoch": 0.02265705458290422,
+      "grad_norm": 1.4296875,
+      "learning_rate": 3.7500000000000005e-06,
+      "loss": 0.3228,
+      "step": 33
+    },
+    {
+      "epoch": 0.02334363199450738,
+      "grad_norm": 1.515625,
+      "learning_rate": 3.863636363636364e-06,
+      "loss": 0.315,
+      "step": 34
+    },
+    {
+      "epoch": 0.024030209406110538,
+      "grad_norm": 1.3125,
+      "learning_rate": 3.9772727272727275e-06,
+      "loss": 0.3168,
+      "step": 35
+    },
+    {
+      "epoch": 0.024716786817713696,
+      "grad_norm": 1.296875,
+      "learning_rate": 4.0909090909090915e-06,
+      "loss": 0.3128,
+      "step": 36
+    },
+    {
+      "epoch": 0.025403364229316855,
+      "grad_norm": 1.125,
+      "learning_rate": 4.204545454545455e-06,
+      "loss": 0.3154,
+      "step": 37
+    },
+    {
+      "epoch": 0.026089941640920013,
+      "grad_norm": 1.09375,
+      "learning_rate": 4.3181818181818185e-06,
+      "loss": 0.3083,
+      "step": 38
+    },
+    {
+      "epoch": 0.026776519052523172,
+      "grad_norm": 0.9921875,
+      "learning_rate": 4.4318181818181824e-06,
+      "loss": 0.3043,
+      "step": 39
+    },
+    {
+      "epoch": 0.02746309646412633,
+      "grad_norm": 0.94140625,
+      "learning_rate": 4.5454545454545455e-06,
+      "loss": 0.2977,
+      "step": 40
+    },
+    {
+      "epoch": 0.02814967387572949,
+      "grad_norm": 0.8671875,
+      "learning_rate": 4.6590909090909095e-06,
+      "loss": 0.272,
+      "step": 41
+    },
+    {
+      "epoch": 0.028836251287332648,
+      "grad_norm": 0.84765625,
+      "learning_rate": 4.772727272727273e-06,
+      "loss": 0.2774,
+      "step": 42
+    },
+    {
+      "epoch": 0.029522828698935806,
+      "grad_norm": 0.80859375,
+      "learning_rate": 4.8863636363636365e-06,
+      "loss": 0.267,
+      "step": 43
+    },
+    {
+      "epoch": 0.030209406110538965,
+      "grad_norm": 0.828125,
+      "learning_rate": 5e-06,
+      "loss": 0.2646,
+      "step": 44
+    },
+    {
+      "epoch": 0.030895983522142123,
+      "grad_norm": 0.85546875,
+      "learning_rate": 5.113636363636364e-06,
+      "loss": 0.2592,
+      "step": 45
+    },
+    {
+      "epoch": 0.03158256093374528,
+      "grad_norm": 0.74609375,
+      "learning_rate": 5.2272727272727274e-06,
+      "loss": 0.2453,
+      "step": 46
+    },
+    {
+      "epoch": 0.03226913834534844,
+      "grad_norm": 0.78125,
+      "learning_rate": 5.340909090909091e-06,
+      "loss": 0.2393,
+      "step": 47
+    },
+    {
+      "epoch": 0.032955715756951595,
+      "grad_norm": 0.78125,
+      "learning_rate": 5.4545454545454545e-06,
+      "loss": 0.2339,
+      "step": 48
+    },
+    {
+      "epoch": 0.03364229316855476,
+      "grad_norm": 0.76171875,
+      "learning_rate": 5.568181818181818e-06,
+      "loss": 0.229,
+      "step": 49
+    },
+    {
+      "epoch": 0.03432887058015791,
+      "grad_norm": 0.7421875,
+      "learning_rate": 5.681818181818183e-06,
+      "loss": 0.2383,
+      "step": 50
+    },
+    {
+      "epoch": 0.035015447991761074,
+      "grad_norm": 0.77734375,
+      "learning_rate": 5.795454545454546e-06,
+      "loss": 0.2255,
+      "step": 51
+    },
+    {
+      "epoch": 0.03570202540336423,
+      "grad_norm": 0.7421875,
+      "learning_rate": 5.90909090909091e-06,
+      "loss": 0.2251,
+      "step": 52
+    },
+    {
+      "epoch": 0.036388602814967384,
+      "grad_norm": 0.8515625,
+      "learning_rate": 6.022727272727273e-06,
+      "loss": 0.2179,
+      "step": 53
+    },
+    {
+      "epoch": 0.037075180226570546,
+      "grad_norm": 0.7578125,
+      "learning_rate": 6.136363636363637e-06,
+      "loss": 0.1995,
+      "step": 54
+    },
+    {
+      "epoch": 0.0377617576381737,
+      "grad_norm": 0.6875,
+      "learning_rate": 6.25e-06,
+      "loss": 0.1978,
+      "step": 55
+    },
+    {
+      "epoch": 0.03844833504977686,
+      "grad_norm": 0.68359375,
+      "learning_rate": 6.363636363636364e-06,
+      "loss": 0.1784,
+      "step": 56
+    },
+    {
+      "epoch": 0.03913491246138002,
+      "grad_norm": 0.671875,
+      "learning_rate": 6.477272727272727e-06,
+      "loss": 0.1888,
+      "step": 57
+    },
+    {
+      "epoch": 0.03982148987298318,
+      "grad_norm": 0.640625,
+      "learning_rate": 6.590909090909091e-06,
+      "loss": 0.1702,
+      "step": 58
+    },
+    {
+      "epoch": 0.040508067284586335,
+      "grad_norm": 0.6171875,
+      "learning_rate": 6.704545454545454e-06,
+      "loss": 0.1694,
+      "step": 59
+    },
+    {
+      "epoch": 0.0411946446961895,
+      "grad_norm": 0.5625,
+      "learning_rate": 6.818181818181818e-06,
+      "loss": 0.1613,
+      "step": 60
+    },
+    {
+      "epoch": 0.04188122210779265,
+      "grad_norm": 0.52734375,
+      "learning_rate": 6.931818181818183e-06,
+      "loss": 0.1639,
+      "step": 61
+    },
+    {
+      "epoch": 0.042567799519395814,
+      "grad_norm": 0.50390625,
+      "learning_rate": 7.045454545454546e-06,
+      "loss": 0.1582,
+      "step": 62
+    },
+    {
+      "epoch": 0.04325437693099897,
+      "grad_norm": 0.470703125,
+      "learning_rate": 7.15909090909091e-06,
+      "loss": 0.1443,
+      "step": 63
+    },
+    {
+      "epoch": 0.04394095434260213,
+      "grad_norm": 0.43359375,
+      "learning_rate": 7.272727272727273e-06,
+      "loss": 0.1513,
+      "step": 64
+    },
+    {
+      "epoch": 0.04462753175420529,
+      "grad_norm": 0.421875,
+      "learning_rate": 7.386363636363637e-06,
+      "loss": 0.1329,
+      "step": 65
+    },
+    {
+      "epoch": 0.04531410916580844,
+      "grad_norm": 0.416015625,
+      "learning_rate": 7.500000000000001e-06,
+      "loss": 0.1387,
+      "step": 66
+    },
+    {
+      "epoch": 0.046000686577411604,
+      "grad_norm": 0.390625,
+      "learning_rate": 7.613636363636364e-06,
+      "loss": 0.1359,
+      "step": 67
+    },
+    {
+      "epoch": 0.04668726398901476,
+      "grad_norm": 0.37109375,
+      "learning_rate": 7.727272727272727e-06,
+      "loss": 0.1389,
+      "step": 68
+    },
+    {
+      "epoch": 0.04737384140061792,
+      "grad_norm": 0.373046875,
+      "learning_rate": 7.840909090909091e-06,
+      "loss": 0.1393,
+      "step": 69
+    },
+    {
+      "epoch": 0.048060418812221076,
+      "grad_norm": 0.36328125,
+      "learning_rate": 7.954545454545455e-06,
+      "loss": 0.1389,
+      "step": 70
+    },
+    {
+      "epoch": 0.04874699622382424,
+      "grad_norm": 0.361328125,
+      "learning_rate": 8.068181818181819e-06,
+      "loss": 0.1198,
+      "step": 71
+    },
+    {
+      "epoch": 0.04943357363542739,
+      "grad_norm": 0.337890625,
+      "learning_rate": 8.181818181818183e-06,
+      "loss": 0.1287,
+      "step": 72
+    },
+    {
+      "epoch": 0.050120151047030555,
+      "grad_norm": 0.33984375,
+      "learning_rate": 8.295454545454547e-06,
+      "loss": 0.1271,
+      "step": 73
+    },
+    {
+      "epoch": 0.05080672845863371,
+      "grad_norm": 0.330078125,
+      "learning_rate": 8.40909090909091e-06,
+      "loss": 0.1125,
+      "step": 74
+    },
+    {
+      "epoch": 0.05149330587023687,
+      "grad_norm": 0.318359375,
+      "learning_rate": 8.522727272727273e-06,
+      "loss": 0.123,
+      "step": 75
+    },
+    {
+      "epoch": 0.05217988328184003,
+      "grad_norm": 0.330078125,
+      "learning_rate": 8.636363636363637e-06,
+      "loss": 0.1208,
+      "step": 76
+    },
+    {
+      "epoch": 0.05286646069344319,
+      "grad_norm": 0.345703125,
+      "learning_rate": 8.750000000000001e-06,
+      "loss": 0.1148,
+      "step": 77
+    },
+    {
+      "epoch": 0.053553038105046344,
+      "grad_norm": 0.3203125,
+      "learning_rate": 8.863636363636365e-06,
+      "loss": 0.1234,
+      "step": 78
+    },
+    {
+      "epoch": 0.0542396155166495,
+      "grad_norm": 0.3125,
+      "learning_rate": 8.977272727272727e-06,
+      "loss": 0.1128,
+      "step": 79
+    },
+    {
+      "epoch": 0.05492619292825266,
+      "grad_norm": 0.30078125,
+      "learning_rate": 9.090909090909091e-06,
+      "loss": 0.1171,
+      "step": 80
+    },
+    {
+      "epoch": 0.055612770339855816,
+      "grad_norm": 0.25,
+      "learning_rate": 9.204545454545455e-06,
+      "loss": 0.1153,
+      "step": 81
+    },
+    {
+      "epoch": 0.05629934775145898,
+      "grad_norm": 0.287109375,
+      "learning_rate": 9.318181818181819e-06,
+      "loss": 0.1118,
+      "step": 82
+    },
+    {
+      "epoch": 0.05698592516306213,
+      "grad_norm": 0.283203125,
+      "learning_rate": 9.431818181818183e-06,
+      "loss": 0.1093,
+      "step": 83
+    },
+    {
+      "epoch": 0.057672502574665295,
+      "grad_norm": 0.310546875,
+      "learning_rate": 9.545454545454547e-06,
+      "loss": 0.1152,
+      "step": 84
+    },
+    {
+      "epoch": 0.05835907998626845,
+      "grad_norm": 0.275390625,
+      "learning_rate": 9.65909090909091e-06,
+      "loss": 0.099,
+      "step": 85
+    },
+    {
+      "epoch": 0.05904565739787161,
+      "grad_norm": 0.298828125,
+      "learning_rate": 9.772727272727273e-06,
+      "loss": 0.1053,
+      "step": 86
+    },
+    {
+      "epoch": 0.05973223480947477,
+      "grad_norm": 0.25390625,
+      "learning_rate": 9.886363636363637e-06,
+      "loss": 0.1124,
+      "step": 87
+    },
+    {
+      "epoch": 0.06041881222107793,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1e-05,
+      "loss": 0.1157,
+      "step": 88
+    },
+    {
+      "epoch": 0.061105389632681084,
+      "grad_norm": 0.296875,
+      "learning_rate": 9.999998787742986e-06,
+      "loss": 0.1174,
+      "step": 89
+    },
+    {
+      "epoch": 0.061791967044284246,
+      "grad_norm": 0.25390625,
+      "learning_rate": 9.999995150972593e-06,
+      "loss": 0.1121,
+      "step": 90
+    },
+    {
+      "epoch": 0.0624785444558874,
+      "grad_norm": 0.28125,
+      "learning_rate": 9.999989089690783e-06,
+      "loss": 0.1057,
+      "step": 91
+    },
+    {
+      "epoch": 0.06316512186749056,
+      "grad_norm": 0.255859375,
+      "learning_rate": 9.999980603900821e-06,
+      "loss": 0.0984,
+      "step": 92
+    },
+    {
+      "epoch": 0.06385169927909372,
+      "grad_norm": 0.26171875,
+      "learning_rate": 9.99996969360728e-06,
+      "loss": 0.1075,
+      "step": 93
+    },
+    {
+      "epoch": 0.06453827669069688,
+      "grad_norm": 0.23046875,
+      "learning_rate": 9.999956358816037e-06,
+      "loss": 0.1144,
+      "step": 94
+    },
+    {
+      "epoch": 0.06522485410230003,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 9.999940599534277e-06,
+      "loss": 0.1091,
+      "step": 95
+    },
+    {
+      "epoch": 0.06591143151390319,
+      "grad_norm": 0.263671875,
+      "learning_rate": 9.99992241577049e-06,
+      "loss": 0.1046,
+      "step": 96
+    },
+    {
+      "epoch": 0.06659800892550635,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 9.999901807534473e-06,
+      "loss": 0.1162,
+      "step": 97
+    },
+    {
+      "epoch": 0.06728458633710951,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 9.999878774837331e-06,
+      "loss": 0.1147,
+      "step": 98
+    },
+    {
+      "epoch": 0.06797116374871266,
+      "grad_norm": 0.25,
+      "learning_rate": 9.999853317691473e-06,
+      "loss": 0.1039,
+      "step": 99
+    },
+    {
+      "epoch": 0.06865774116031582,
+      "grad_norm": 0.251953125,
+      "learning_rate": 9.999825436110612e-06,
+      "loss": 0.1065,
+      "step": 100
+    },
+    {
+      "epoch": 0.06934431857191899,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 9.999795130109777e-06,
+      "loss": 0.1178,
+      "step": 101
+    },
+    {
+      "epoch": 0.07003089598352215,
+      "grad_norm": 0.248046875,
+      "learning_rate": 9.999762399705288e-06,
+      "loss": 0.1082,
+      "step": 102
+    },
+    {
+      "epoch": 0.0707174733951253,
+      "grad_norm": 0.25390625,
+      "learning_rate": 9.999727244914785e-06,
+      "loss": 0.1033,
+      "step": 103
+    },
+    {
+      "epoch": 0.07140405080672846,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 9.999689665757205e-06,
+      "loss": 0.1072,
+      "step": 104
+    },
+    {
+      "epoch": 0.07209062821833162,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 9.9996496622528e-06,
+      "loss": 0.1028,
+      "step": 105
+    },
+    {
+      "epoch": 0.07277720562993477,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 9.999607234423118e-06,
+      "loss": 0.1007,
+      "step": 106
+    },
+    {
+      "epoch": 0.07346378304153793,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 9.99956238229102e-06,
+      "loss": 0.0927,
+      "step": 107
+    },
+    {
+      "epoch": 0.07415036045314109,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 9.999515105880674e-06,
+      "loss": 0.1045,
+      "step": 108
+    },
+    {
+      "epoch": 0.07483693786474425,
+      "grad_norm": 0.232421875,
+      "learning_rate": 9.999465405217547e-06,
+      "loss": 0.0942,
+      "step": 109
+    },
+    {
+      "epoch": 0.0755235152763474,
+      "grad_norm": 0.25390625,
+      "learning_rate": 9.99941328032842e-06,
+      "loss": 0.1107,
+      "step": 110
+    },
+    {
+      "epoch": 0.07621009268795056,
+      "grad_norm": 0.236328125,
+      "learning_rate": 9.999358731241378e-06,
+      "loss": 0.0951,
+      "step": 111
+    },
+    {
+      "epoch": 0.07689667009955373,
+      "grad_norm": 0.232421875,
+      "learning_rate": 9.999301757985807e-06,
+      "loss": 0.1076,
+      "step": 112
+    },
+    {
+      "epoch": 0.07758324751115689,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 9.999242360592406e-06,
+      "loss": 0.1148,
+      "step": 113
+    },
+    {
+      "epoch": 0.07826982492276004,
+      "grad_norm": 0.24609375,
+      "learning_rate": 9.999180539093177e-06,
+      "loss": 0.0985,
+      "step": 114
+    },
+    {
+      "epoch": 0.0789564023343632,
+      "grad_norm": 0.255859375,
+      "learning_rate": 9.999116293521427e-06,
+      "loss": 0.1042,
+      "step": 115
+    },
+    {
+      "epoch": 0.07964297974596636,
+      "grad_norm": 0.2265625,
+      "learning_rate": 9.999049623911771e-06,
+      "loss": 0.1002,
+      "step": 116
+    },
+    {
+      "epoch": 0.08032955715756952,
+      "grad_norm": 0.220703125,
+      "learning_rate": 9.998980530300129e-06,
+      "loss": 0.1059,
+      "step": 117
+    },
+    {
+      "epoch": 0.08101613456917267,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 9.998909012723729e-06,
+      "loss": 0.1062,
+      "step": 118
+    },
+    {
+      "epoch": 0.08170271198077583,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 9.9988350712211e-06,
+      "loss": 0.0996,
+      "step": 119
+    },
+    {
+      "epoch": 0.082389289392379,
+      "grad_norm": 0.220703125,
+      "learning_rate": 9.998758705832084e-06,
+      "loss": 0.1009,
+      "step": 120
+    },
+    {
+      "epoch": 0.08307586680398214,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 9.998679916597822e-06,
+      "loss": 0.0994,
+      "step": 121
+    },
+    {
+      "epoch": 0.0837624442155853,
+      "grad_norm": 0.212890625,
+      "learning_rate": 9.998598703560766e-06,
+      "loss": 0.1044,
+      "step": 122
+    },
+    {
+      "epoch": 0.08444902162718847,
+      "grad_norm": 0.205078125,
+      "learning_rate": 9.998515066764672e-06,
+      "loss": 0.104,
+      "step": 123
+    },
+    {
+      "epoch": 0.08513559903879163,
+      "grad_norm": 0.20703125,
+      "learning_rate": 9.998429006254601e-06,
+      "loss": 0.1032,
+      "step": 124
+    },
+    {
+      "epoch": 0.08582217645039478,
+      "grad_norm": 0.21875,
+      "learning_rate": 9.99834052207692e-06,
+      "loss": 0.0962,
+      "step": 125
+    },
+    {
+      "epoch": 0.08650875386199794,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 9.998249614279306e-06,
+      "loss": 0.1067,
+      "step": 126
+    },
+    {
+      "epoch": 0.0871953312736011,
+      "grad_norm": 0.212890625,
+      "learning_rate": 9.998156282910736e-06,
+      "loss": 0.107,
+      "step": 127
+    },
+    {
+      "epoch": 0.08788190868520426,
+      "grad_norm": 0.23828125,
+      "learning_rate": 9.998060528021493e-06,
+      "loss": 0.1016,
+      "step": 128
+    },
+    {
+      "epoch": 0.08856848609680741,
+      "grad_norm": 0.224609375,
+      "learning_rate": 9.997962349663172e-06,
+      "loss": 0.1048,
+      "step": 129
+    },
+    {
+      "epoch": 0.08925506350841057,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 9.997861747888669e-06,
+      "loss": 0.1086,
+      "step": 130
+    },
+    {
+      "epoch": 0.08994164092001374,
+      "grad_norm": 0.234375,
+      "learning_rate": 9.997758722752182e-06,
+      "loss": 0.1057,
+      "step": 131
+    },
+    {
+      "epoch": 0.09062821833161688,
+      "grad_norm": 0.212890625,
+      "learning_rate": 9.997653274309225e-06,
+      "loss": 0.102,
+      "step": 132
+    },
+    {
+      "epoch": 0.09131479574322005,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 9.997545402616607e-06,
+      "loss": 0.0985,
+      "step": 133
+    },
+    {
+      "epoch": 0.09200137315482321,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 9.997435107732451e-06,
+      "loss": 0.1032,
+      "step": 134
+    },
+    {
+      "epoch": 0.09268795056642637,
+      "grad_norm": 0.21875,
+      "learning_rate": 9.997322389716178e-06,
+      "loss": 0.094,
+      "step": 135
+    },
+    {
+      "epoch": 0.09337452797802952,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 9.99720724862852e-06,
+      "loss": 0.1037,
+      "step": 136
+    },
+    {
+      "epoch": 0.09406110538963268,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 9.997089684531514e-06,
+      "loss": 0.0986,
+      "step": 137
+    },
+    {
+      "epoch": 0.09474768280123584,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 9.996969697488499e-06,
+      "loss": 0.1036,
+      "step": 138
+    },
+    {
+      "epoch": 0.095434260212839,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 9.996847287564124e-06,
+      "loss": 0.0916,
+      "step": 139
+    },
+    {
+      "epoch": 0.09612083762444215,
+      "grad_norm": 0.228515625,
+      "learning_rate": 9.99672245482434e-06,
+      "loss": 0.104,
+      "step": 140
+    },
+    {
+      "epoch": 0.09680741503604531,
+      "grad_norm": 0.21484375,
+      "learning_rate": 9.996595199336404e-06,
+      "loss": 0.0989,
+      "step": 141
+    },
+    {
+      "epoch": 0.09749399244764848,
+      "grad_norm": 0.25390625,
+      "learning_rate": 9.99646552116888e-06,
+      "loss": 0.1048,
+      "step": 142
+    },
+    {
+      "epoch": 0.09818056985925164,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 9.996333420391635e-06,
+      "loss": 0.1001,
+      "step": 143
+    },
+    {
+      "epoch": 0.09886714727085479,
+      "grad_norm": 0.23046875,
+      "learning_rate": 9.996198897075842e-06,
+      "loss": 0.1055,
+      "step": 144
+    },
+    {
+      "epoch": 0.09955372468245795,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 9.996061951293981e-06,
+      "loss": 0.1003,
+      "step": 145
+    },
+    {
+      "epoch": 0.10024030209406111,
+      "grad_norm": 0.240234375,
+      "learning_rate": 9.995922583119836e-06,
+      "loss": 0.0953,
+      "step": 146
+    },
+    {
+      "epoch": 0.10092687950566426,
+      "grad_norm": 0.22265625,
+      "learning_rate": 9.995780792628494e-06,
+      "loss": 0.106,
+      "step": 147
+    },
+    {
+      "epoch": 0.10161345691726742,
+      "grad_norm": 0.220703125,
+      "learning_rate": 9.99563657989635e-06,
+      "loss": 0.091,
+      "step": 148
+    },
+    {
+      "epoch": 0.10230003432887058,
+      "grad_norm": 0.2421875,
+      "learning_rate": 9.995489945001104e-06,
+      "loss": 0.1103,
+      "step": 149
+    },
+    {
+      "epoch": 0.10298661174047374,
+      "grad_norm": 0.22265625,
+      "learning_rate": 9.995340888021761e-06,
+      "loss": 0.0943,
+      "step": 150
+    },
+    {
+      "epoch": 0.10367318915207689,
+      "grad_norm": 0.232421875,
+      "learning_rate": 9.995189409038626e-06,
+      "loss": 0.0981,
+      "step": 151
+    },
+    {
+      "epoch": 0.10435976656368005,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 9.995035508133316e-06,
+      "loss": 0.1139,
+      "step": 152
+    },
+    {
+      "epoch": 0.10504634397528322,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 9.994879185388747e-06,
+      "loss": 0.1007,
+      "step": 153
+    },
+    {
+      "epoch": 0.10573292138688638,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 9.994720440889147e-06,
+      "loss": 0.1007,
+      "step": 154
+    },
+    {
+      "epoch": 0.10641949879848953,
+      "grad_norm": 0.234375,
+      "learning_rate": 9.994559274720043e-06,
+      "loss": 0.105,
+      "step": 155
+    },
+    {
+      "epoch": 0.10710607621009269,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 9.994395686968267e-06,
+      "loss": 0.1054,
+      "step": 156
+    },
+    {
+      "epoch": 0.10779265362169585,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 9.994229677721957e-06,
+      "loss": 0.0967,
+      "step": 157
+    },
+    {
+      "epoch": 0.108479231033299,
+      "grad_norm": 0.228515625,
+      "learning_rate": 9.994061247070557e-06,
+      "loss": 0.1028,
+      "step": 158
+    },
+    {
+      "epoch": 0.10916580844490216,
+      "grad_norm": 0.220703125,
+      "learning_rate": 9.993890395104812e-06,
+      "loss": 0.1025,
+      "step": 159
+    },
+    {
+      "epoch": 0.10985238585650532,
+      "grad_norm": 0.201171875,
+      "learning_rate": 9.993717121916778e-06,
+      "loss": 0.0982,
+      "step": 160
+    },
+    {
+      "epoch": 0.11053896326810848,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 9.993541427599805e-06,
+      "loss": 0.0981,
+      "step": 161
+    },
+    {
+      "epoch": 0.11122554067971163,
+      "grad_norm": 0.19921875,
+      "learning_rate": 9.993363312248559e-06,
+      "loss": 0.0906,
+      "step": 162
+    },
+    {
+      "epoch": 0.1119121180913148,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 9.993182775959004e-06,
+      "loss": 0.1043,
+      "step": 163
+    },
+    {
+      "epoch": 0.11259869550291796,
+      "grad_norm": 0.232421875,
+      "learning_rate": 9.992999818828408e-06,
+      "loss": 0.0994,
+      "step": 164
+    },
+    {
+      "epoch": 0.11328527291452112,
+      "grad_norm": 0.21875,
+      "learning_rate": 9.992814440955346e-06,
+      "loss": 0.0947,
+      "step": 165
+    },
+    {
+      "epoch": 0.11397185032612427,
+      "grad_norm": 0.224609375,
+      "learning_rate": 9.992626642439695e-06,
+      "loss": 0.0992,
+      "step": 166
+    },
+    {
+      "epoch": 0.11465842773772743,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 9.992436423382639e-06,
+      "loss": 0.096,
+      "step": 167
+    },
+    {
+      "epoch": 0.11534500514933059,
+      "grad_norm": 0.201171875,
+      "learning_rate": 9.992243783886663e-06,
+      "loss": 0.105,
+      "step": 168
+    },
+    {
+      "epoch": 0.11603158256093375,
+      "grad_norm": 0.22265625,
+      "learning_rate": 9.992048724055559e-06,
+      "loss": 0.1074,
+      "step": 169
+    },
+    {
+      "epoch": 0.1167181599725369,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 9.99185124399442e-06,
+      "loss": 0.1122,
+      "step": 170
+    },
+    {
+      "epoch": 0.11740473738414006,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 9.991651343809642e-06,
+      "loss": 0.1047,
+      "step": 171
+    },
+    {
+      "epoch": 0.11809131479574322,
+      "grad_norm": 0.208984375,
+      "learning_rate": 9.991449023608932e-06,
+      "loss": 0.094,
+      "step": 172
+    },
+    {
+      "epoch": 0.11877789220734637,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 9.991244283501294e-06,
+      "loss": 0.096,
+      "step": 173
+    },
+    {
+      "epoch": 0.11946446961894953,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 9.991037123597041e-06,
+      "loss": 0.0936,
+      "step": 174
+    },
+    {
+      "epoch": 0.1201510470305527,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 9.990827544007783e-06,
+      "loss": 0.0996,
+      "step": 175
+    },
+    {
+      "epoch": 0.12083762444215586,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 9.990615544846439e-06,
+      "loss": 0.1031,
+      "step": 176
+    },
+    {
+      "epoch": 0.121524201853759,
+      "grad_norm": 0.21484375,
+      "learning_rate": 9.99040112622723e-06,
+      "loss": 0.1034,
+      "step": 177
+    },
+    {
+      "epoch": 0.12221077926536217,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 9.990184288265679e-06,
+      "loss": 0.0939,
+      "step": 178
+    },
+    {
+      "epoch": 0.12289735667696533,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 9.989965031078616e-06,
+      "loss": 0.0897,
+      "step": 179
+    },
+    {
+      "epoch": 0.12358393408856849,
+      "grad_norm": 0.201171875,
+      "learning_rate": 9.989743354784174e-06,
+      "loss": 0.0891,
+      "step": 180
+    },
+    {
+      "epoch": 0.12427051150017164,
+      "grad_norm": 0.236328125,
+      "learning_rate": 9.989519259501786e-06,
+      "loss": 0.0971,
+      "step": 181
+    },
+    {
+      "epoch": 0.1249570889117748,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 9.989292745352191e-06,
+      "loss": 0.0916,
+      "step": 182
+    },
+    {
+      "epoch": 0.12564366632337795,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 9.98906381245743e-06,
+      "loss": 0.1044,
+      "step": 183
+    },
+    {
+      "epoch": 0.1263302437349811,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 9.988832460940846e-06,
+      "loss": 0.0914,
+      "step": 184
+    },
+    {
+      "epoch": 0.12701682114658427,
+      "grad_norm": 0.2421875,
+      "learning_rate": 9.98859869092709e-06,
+      "loss": 0.1019,
+      "step": 185
+    },
+    {
+      "epoch": 0.12770339855818744,
+      "grad_norm": 0.203125,
+      "learning_rate": 9.98836250254211e-06,
+      "loss": 0.091,
+      "step": 186
+    },
+    {
+      "epoch": 0.1283899759697906,
+      "grad_norm": 0.21875,
+      "learning_rate": 9.988123895913162e-06,
+      "loss": 0.1086,
+      "step": 187
+    },
+    {
+      "epoch": 0.12907655338139376,
+      "grad_norm": 0.322265625,
+      "learning_rate": 9.987882871168801e-06,
+      "loss": 0.1062,
+      "step": 188
+    },
+    {
+      "epoch": 0.12976313079299692,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 9.987639428438888e-06,
+      "loss": 0.0931,
+      "step": 189
+    },
+    {
+      "epoch": 0.13044970820460006,
+      "grad_norm": 0.208984375,
+      "learning_rate": 9.987393567854585e-06,
+      "loss": 0.0901,
+      "step": 190
+    },
+    {
+      "epoch": 0.13113628561620322,
+      "grad_norm": 0.21875,
+      "learning_rate": 9.987145289548356e-06,
+      "loss": 0.0905,
+      "step": 191
+    },
+    {
+      "epoch": 0.13182286302780638,
+      "grad_norm": 0.23046875,
+      "learning_rate": 9.986894593653969e-06,
+      "loss": 0.1013,
+      "step": 192
+    },
+    {
+      "epoch": 0.13250944043940954,
+      "grad_norm": 0.212890625,
+      "learning_rate": 9.986641480306495e-06,
+      "loss": 0.1104,
+      "step": 193
+    },
+    {
+      "epoch": 0.1331960178510127,
+      "grad_norm": 0.224609375,
+      "learning_rate": 9.986385949642307e-06,
+      "loss": 0.1112,
+      "step": 194
+    },
+    {
+      "epoch": 0.13388259526261587,
+      "grad_norm": 0.244140625,
+      "learning_rate": 9.986128001799077e-06,
+      "loss": 0.0941,
+      "step": 195
+    },
+    {
+      "epoch": 0.13456917267421903,
+      "grad_norm": 0.205078125,
+      "learning_rate": 9.985867636915784e-06,
+      "loss": 0.108,
+      "step": 196
+    },
+    {
+      "epoch": 0.13525575008582216,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 9.98560485513271e-06,
+      "loss": 0.1118,
+      "step": 197
+    },
+    {
+      "epoch": 0.13594232749742532,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 9.985339656591434e-06,
+      "loss": 0.1043,
+      "step": 198
+    },
+    {
+      "epoch": 0.1366289049090285,
+      "grad_norm": 0.2109375,
+      "learning_rate": 9.985072041434841e-06,
+      "loss": 0.1017,
+      "step": 199
+    },
+    {
+      "epoch": 0.13731548232063165,
+      "grad_norm": 0.193359375,
+      "learning_rate": 9.984802009807117e-06,
+      "loss": 0.0939,
+      "step": 200
+    },
+    {
+      "epoch": 0.1380020597322348,
+      "grad_norm": 0.203125,
+      "learning_rate": 9.984529561853749e-06,
+      "loss": 0.0927,
+      "step": 201
+    },
+    {
+      "epoch": 0.13868863714383797,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 9.984254697721528e-06,
+      "loss": 0.0954,
+      "step": 202
+    },
+    {
+      "epoch": 0.13937521455544113,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 9.983977417558544e-06,
+      "loss": 0.0941,
+      "step": 203
+    },
+    {
+      "epoch": 0.1400617919670443,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 9.98369772151419e-06,
+      "loss": 0.0933,
+      "step": 204
+    },
+    {
+      "epoch": 0.14074836937864743,
+      "grad_norm": 0.20703125,
+      "learning_rate": 9.983415609739165e-06,
+      "loss": 0.1003,
+      "step": 205
+    },
+    {
+      "epoch": 0.1414349467902506,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 9.98313108238546e-06,
+      "loss": 0.1117,
+      "step": 206
+    },
+    {
+      "epoch": 0.14212152420185376,
+      "grad_norm": 0.22265625,
+      "learning_rate": 9.982844139606375e-06,
+      "loss": 0.0929,
+      "step": 207
+    },
+    {
+      "epoch": 0.14280810161345692,
+      "grad_norm": 0.24609375,
+      "learning_rate": 9.982554781556512e-06,
+      "loss": 0.0968,
+      "step": 208
+    },
+    {
+      "epoch": 0.14349467902506008,
+      "grad_norm": 0.19921875,
+      "learning_rate": 9.982263008391769e-06,
+      "loss": 0.0987,
+      "step": 209
+    },
+    {
+      "epoch": 0.14418125643666324,
+      "grad_norm": 0.2890625,
+      "learning_rate": 9.981968820269347e-06,
+      "loss": 0.0868,
+      "step": 210
+    },
+    {
+      "epoch": 0.1448678338482664,
+      "grad_norm": 0.24609375,
+      "learning_rate": 9.98167221734775e-06,
+      "loss": 0.1,
+      "step": 211
+    },
+    {
+      "epoch": 0.14555441125986954,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 9.981373199786782e-06,
+      "loss": 0.0918,
+      "step": 212
+    },
+    {
+      "epoch": 0.1462409886714727,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 9.981071767747547e-06,
+      "loss": 0.0921,
+      "step": 213
+    },
+    {
+      "epoch": 0.14692756608307586,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 9.980767921392453e-06,
+      "loss": 0.099,
+      "step": 214
+    },
+    {
+      "epoch": 0.14761414349467902,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 9.980461660885204e-06,
+      "loss": 0.1032,
+      "step": 215
+    },
+    {
+      "epoch": 0.14830072090628219,
+      "grad_norm": 0.18359375,
+      "learning_rate": 9.98015298639081e-06,
+      "loss": 0.0977,
+      "step": 216
+    },
+    {
+      "epoch": 0.14898729831788535,
+      "grad_norm": 0.240234375,
+      "learning_rate": 9.979841898075577e-06,
+      "loss": 0.0908,
+      "step": 217
+    },
+    {
+      "epoch": 0.1496738757294885,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 9.979528396107114e-06,
+      "loss": 0.0983,
+      "step": 218
+    },
+    {
+      "epoch": 0.15036045314109167,
+      "grad_norm": 0.25,
+      "learning_rate": 9.97921248065433e-06,
+      "loss": 0.0964,
+      "step": 219
+    },
+    {
+      "epoch": 0.1510470305526948,
+      "grad_norm": 0.220703125,
+      "learning_rate": 9.978894151887435e-06,
+      "loss": 0.0945,
+      "step": 220
+    },
+    {
+      "epoch": 0.15173360796429797,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 9.978573409977937e-06,
+      "loss": 0.1029,
+      "step": 221
+    },
+    {
+      "epoch": 0.15242018537590113,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 9.978250255098645e-06,
+      "loss": 0.0971,
+      "step": 222
+    },
+    {
+      "epoch": 0.1531067627875043,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 9.977924687423672e-06,
+      "loss": 0.0919,
+      "step": 223
+    },
+    {
+      "epoch": 0.15379334019910745,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 9.977596707128424e-06,
+      "loss": 0.0905,
+      "step": 224
+    },
+    {
+      "epoch": 0.15447991761071062,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 9.977266314389611e-06,
+      "loss": 0.0938,
+      "step": 225
+    },
+    {
+      "epoch": 0.15516649502231378,
+      "grad_norm": 0.23828125,
+      "learning_rate": 9.976933509385245e-06,
+      "loss": 0.1117,
+      "step": 226
+    },
+    {
+      "epoch": 0.1558530724339169,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 9.976598292294632e-06,
+      "loss": 0.0955,
+      "step": 227
+    },
+    {
+      "epoch": 0.15653964984552007,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 9.976260663298384e-06,
+      "loss": 0.0898,
+      "step": 228
+    },
+    {
+      "epoch": 0.15722622725712324,
+      "grad_norm": 0.236328125,
+      "learning_rate": 9.975920622578403e-06,
+      "loss": 0.0928,
+      "step": 229
+    },
+    {
+      "epoch": 0.1579128046687264,
+      "grad_norm": 0.1953125,
+      "learning_rate": 9.975578170317905e-06,
+      "loss": 0.0937,
+      "step": 230
+    },
+    {
+      "epoch": 0.15859938208032956,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 9.97523330670139e-06,
+      "loss": 0.1031,
+      "step": 231
+    },
+    {
+      "epoch": 0.15928595949193272,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 9.974886031914665e-06,
+      "loss": 0.101,
+      "step": 232
+    },
+    {
+      "epoch": 0.15997253690353588,
+      "grad_norm": 0.228515625,
+      "learning_rate": 9.974536346144838e-06,
+      "loss": 0.0929,
+      "step": 233
+    },
+    {
+      "epoch": 0.16065911431513905,
+      "grad_norm": 0.212890625,
+      "learning_rate": 9.974184249580309e-06,
+      "loss": 0.1055,
+      "step": 234
+    },
+    {
+      "epoch": 0.16134569172674218,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 9.973829742410784e-06,
+      "loss": 0.0933,
+      "step": 235
+    },
+    {
+      "epoch": 0.16203226913834534,
+      "grad_norm": 0.23046875,
+      "learning_rate": 9.973472824827262e-06,
+      "loss": 0.0971,
+      "step": 236
+    },
+    {
+      "epoch": 0.1627188465499485,
+      "grad_norm": 0.19921875,
+      "learning_rate": 9.973113497022047e-06,
+      "loss": 0.0927,
+      "step": 237
+    },
+    {
+      "epoch": 0.16340542396155167,
+      "grad_norm": 0.1875,
+      "learning_rate": 9.972751759188736e-06,
+      "loss": 0.1003,
+      "step": 238
+    },
+    {
+      "epoch": 0.16409200137315483,
+      "grad_norm": 0.19921875,
+      "learning_rate": 9.972387611522227e-06,
+      "loss": 0.1083,
+      "step": 239
+    },
+    {
+      "epoch": 0.164778578784758,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 9.972021054218712e-06,
+      "loss": 0.0992,
+      "step": 240
+    },
+    {
+      "epoch": 0.16546515619636115,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 9.971652087475691e-06,
+      "loss": 0.0978,
+      "step": 241
+    },
+    {
+      "epoch": 0.16615173360796429,
+      "grad_norm": 0.232421875,
+      "learning_rate": 9.971280711491952e-06,
+      "loss": 0.1012,
+      "step": 242
+    },
+    {
+      "epoch": 0.16683831101956745,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 9.970906926467588e-06,
+      "loss": 0.1042,
+      "step": 243
+    },
+    {
+      "epoch": 0.1675248884311706,
+      "grad_norm": 0.216796875,
+      "learning_rate": 9.970530732603984e-06,
+      "loss": 0.0952,
+      "step": 244
+    },
+    {
+      "epoch": 0.16821146584277377,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 9.97015213010383e-06,
+      "loss": 0.1012,
+      "step": 245
+    },
+    {
+      "epoch": 0.16889804325437693,
+      "grad_norm": 0.2109375,
+      "learning_rate": 9.969771119171108e-06,
+      "loss": 0.0903,
+      "step": 246
+    },
+    {
+      "epoch": 0.1695846206659801,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 9.969387700011098e-06,
+      "loss": 0.0908,
+      "step": 247
+    },
+    {
+      "epoch": 0.17027119807758326,
+      "grad_norm": 0.259765625,
+      "learning_rate": 9.969001872830383e-06,
+      "loss": 0.1127,
+      "step": 248
+    },
+    {
+      "epoch": 0.1709577754891864,
+      "grad_norm": 0.21875,
+      "learning_rate": 9.968613637836833e-06,
+      "loss": 0.1004,
+      "step": 249
+    },
+    {
+      "epoch": 0.17164435290078955,
+      "grad_norm": 0.244140625,
+      "learning_rate": 9.968222995239628e-06,
+      "loss": 0.1032,
+      "step": 250
+    },
+    {
+      "epoch": 0.17233093031239272,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 9.967829945249234e-06,
+      "loss": 0.0922,
+      "step": 251
+    },
+    {
+      "epoch": 0.17301750772399588,
+      "grad_norm": 0.203125,
+      "learning_rate": 9.967434488077422e-06,
+      "loss": 0.1,
+      "step": 252
+    },
+    {
+      "epoch": 0.17370408513559904,
+      "grad_norm": 0.224609375,
+      "learning_rate": 9.967036623937252e-06,
+      "loss": 0.0892,
+      "step": 253
+    },
+    {
+      "epoch": 0.1743906625472022,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 9.966636353043092e-06,
+      "loss": 0.0886,
+      "step": 254
+    },
+    {
+      "epoch": 0.17507723995880536,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 9.966233675610599e-06,
+      "loss": 0.1015,
+      "step": 255
+    },
+    {
+      "epoch": 0.17576381737040853,
+      "grad_norm": 0.23046875,
+      "learning_rate": 9.965828591856725e-06,
+      "loss": 0.0894,
+      "step": 256
+    },
+    {
+      "epoch": 0.17645039478201166,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 9.965421101999721e-06,
+      "loss": 0.1061,
+      "step": 257
+    },
+    {
+      "epoch": 0.17713697219361482,
+      "grad_norm": 0.19921875,
+      "learning_rate": 9.965011206259138e-06,
+      "loss": 0.0962,
+      "step": 258
+    },
+    {
+      "epoch": 0.17782354960521798,
+      "grad_norm": 0.228515625,
+      "learning_rate": 9.964598904855818e-06,
+      "loss": 0.1039,
+      "step": 259
+    },
+    {
+      "epoch": 0.17851012701682115,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 9.964184198011903e-06,
+      "loss": 0.1045,
+      "step": 260
+    },
+    {
+      "epoch": 0.1791967044284243,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 9.963767085950824e-06,
+      "loss": 0.1036,
+      "step": 261
+    },
+    {
+      "epoch": 0.17988328184002747,
+      "grad_norm": 0.19921875,
+      "learning_rate": 9.96334756889732e-06,
+      "loss": 0.0984,
+      "step": 262
+    },
+    {
+      "epoch": 0.18056985925163063,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 9.962925647077414e-06,
+      "loss": 0.0972,
+      "step": 263
+    },
+    {
+      "epoch": 0.18125643666323377,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 9.962501320718432e-06,
+      "loss": 0.0923,
+      "step": 264
+    },
+    {
+      "epoch": 0.18194301407483693,
+      "grad_norm": 0.212890625,
+      "learning_rate": 9.96207459004899e-06,
+      "loss": 0.101,
+      "step": 265
+    },
+    {
+      "epoch": 0.1826295914864401,
+      "grad_norm": 0.21484375,
+      "learning_rate": 9.961645455299006e-06,
+      "loss": 0.1007,
+      "step": 266
+    },
+    {
+      "epoch": 0.18331616889804325,
+      "grad_norm": 0.189453125,
+      "learning_rate": 9.961213916699685e-06,
+      "loss": 0.1012,
+      "step": 267
+    },
+    {
+      "epoch": 0.18400274630964641,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 9.960779974483537e-06,
+      "loss": 0.1073,
+      "step": 268
+    },
+    {
+      "epoch": 0.18468932372124958,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 9.96034362888436e-06,
+      "loss": 0.0986,
+      "step": 269
+    },
+    {
+      "epoch": 0.18537590113285274,
+      "grad_norm": 0.234375,
+      "learning_rate": 9.959904880137246e-06,
+      "loss": 0.115,
+      "step": 270
+    },
+    {
+      "epoch": 0.1860624785444559,
+      "grad_norm": 0.19921875,
+      "learning_rate": 9.959463728478586e-06,
+      "loss": 0.0976,
+      "step": 271
+    },
+    {
+      "epoch": 0.18674905595605903,
+      "grad_norm": 0.2109375,
+      "learning_rate": 9.959020174146066e-06,
+      "loss": 0.0991,
+      "step": 272
+    },
+    {
+      "epoch": 0.1874356333676622,
+      "grad_norm": 0.259765625,
+      "learning_rate": 9.958574217378663e-06,
+      "loss": 0.094,
+      "step": 273
+    },
+    {
+      "epoch": 0.18812221077926536,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 9.95812585841665e-06,
+      "loss": 0.0899,
+      "step": 274
+    },
+    {
+      "epoch": 0.18880878819086852,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 9.957675097501594e-06,
+      "loss": 0.1032,
+      "step": 275
+    },
+    {
+      "epoch": 0.18949536560247168,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 9.957221934876355e-06,
+      "loss": 0.0978,
+      "step": 276
+    },
+    {
+      "epoch": 0.19018194301407484,
+      "grad_norm": 0.203125,
+      "learning_rate": 9.956766370785093e-06,
+      "loss": 0.0896,
+      "step": 277
+    },
+    {
+      "epoch": 0.190868520425678,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 9.956308405473252e-06,
+      "loss": 0.1042,
+      "step": 278
+    },
+    {
+      "epoch": 0.19155509783728114,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 9.95584803918758e-06,
+      "loss": 0.0983,
+      "step": 279
+    },
+    {
+      "epoch": 0.1922416752488843,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 9.955385272176108e-06,
+      "loss": 0.1007,
+      "step": 280
+    },
+    {
+      "epoch": 0.19292825266048746,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 9.95492010468817e-06,
+      "loss": 0.0934,
+      "step": 281
+    },
+    {
+      "epoch": 0.19361483007209063,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 9.954452536974387e-06,
+      "loss": 0.1007,
+      "step": 282
+    },
+    {
+      "epoch": 0.1943014074836938,
+      "grad_norm": 0.20703125,
+      "learning_rate": 9.953982569286679e-06,
+      "loss": 0.0822,
+      "step": 283
+    },
+    {
+      "epoch": 0.19498798489529695,
+      "grad_norm": 0.21875,
+      "learning_rate": 9.953510201878251e-06,
+      "loss": 0.0981,
+      "step": 284
+    },
+    {
+      "epoch": 0.1956745623069001,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 9.953035435003608e-06,
+      "loss": 0.0935,
+      "step": 285
+    },
+    {
+      "epoch": 0.19636113971850327,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 9.952558268918546e-06,
+      "loss": 0.099,
+      "step": 286
+    },
+    {
+      "epoch": 0.1970477171301064,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 9.952078703880153e-06,
+      "loss": 0.0895,
+      "step": 287
+    },
+    {
+      "epoch": 0.19773429454170957,
+      "grad_norm": 0.26171875,
+      "learning_rate": 9.951596740146809e-06,
+      "loss": 0.0967,
+      "step": 288
+    },
+    {
+      "epoch": 0.19842087195331273,
+      "grad_norm": 0.2265625,
+      "learning_rate": 9.951112377978185e-06,
+      "loss": 0.0887,
+      "step": 289
+    },
+    {
+      "epoch": 0.1991074493649159,
+      "grad_norm": 0.212890625,
+      "learning_rate": 9.950625617635247e-06,
+      "loss": 0.0994,
+      "step": 290
+    },
+    {
+      "epoch": 0.19979402677651906,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 9.950136459380253e-06,
+      "loss": 0.0947,
+      "step": 291
+    },
+    {
+      "epoch": 0.20048060418812222,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 9.949644903476752e-06,
+      "loss": 0.0983,
+      "step": 292
+    },
+    {
+      "epoch": 0.20116718159972538,
+      "grad_norm": 0.263671875,
+      "learning_rate": 9.949150950189586e-06,
+      "loss": 0.0938,
+      "step": 293
+    },
+    {
+      "epoch": 0.20185375901132852,
+      "grad_norm": 0.25,
+      "learning_rate": 9.948654599784886e-06,
+      "loss": 0.0973,
+      "step": 294
+    },
+    {
+      "epoch": 0.20254033642293168,
+      "grad_norm": 0.228515625,
+      "learning_rate": 9.948155852530075e-06,
+      "loss": 0.0858,
+      "step": 295
+    },
+    {
+      "epoch": 0.20322691383453484,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 9.947654708693872e-06,
+      "loss": 0.0964,
+      "step": 296
+    },
+    {
+      "epoch": 0.203913491246138,
+      "grad_norm": 0.20703125,
+      "learning_rate": 9.947151168546281e-06,
+      "loss": 0.0989,
+      "step": 297
+    },
+    {
+      "epoch": 0.20460006865774116,
+      "grad_norm": 0.20703125,
+      "learning_rate": 9.946645232358602e-06,
+      "loss": 0.0926,
+      "step": 298
+    },
+    {
+      "epoch": 0.20528664606934433,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 9.94613690040342e-06,
+      "loss": 0.1045,
+      "step": 299
+    },
+    {
+      "epoch": 0.2059732234809475,
+      "grad_norm": 0.203125,
+      "learning_rate": 9.945626172954617e-06,
+      "loss": 0.1054,
+      "step": 300
+    },
+    {
+      "epoch": 0.20665980089255062,
+      "grad_norm": 0.216796875,
+      "learning_rate": 9.945113050287363e-06,
+      "loss": 0.0928,
+      "step": 301
+    },
+    {
+      "epoch": 0.20734637830415378,
+      "grad_norm": 0.2421875,
+      "learning_rate": 9.94459753267812e-06,
+      "loss": 0.087,
+      "step": 302
+    },
+    {
+      "epoch": 0.20803295571575695,
+      "grad_norm": 0.255859375,
+      "learning_rate": 9.944079620404638e-06,
+      "loss": 0.0919,
+      "step": 303
+    },
+    {
+      "epoch": 0.2087195331273601,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 9.943559313745957e-06,
+      "loss": 0.0975,
+      "step": 304
+    },
+    {
+      "epoch": 0.20940611053896327,
+      "grad_norm": 0.193359375,
+      "learning_rate": 9.943036612982409e-06,
+      "loss": 0.0966,
+      "step": 305
+    },
+    {
+      "epoch": 0.21009268795056643,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 9.942511518395616e-06,
+      "loss": 0.0973,
+      "step": 306
+    },
+    {
+      "epoch": 0.2107792653621696,
+      "grad_norm": 0.23828125,
+      "learning_rate": 9.941984030268487e-06,
+      "loss": 0.0938,
+      "step": 307
+    },
+    {
+      "epoch": 0.21146584277377276,
+      "grad_norm": 0.25390625,
+      "learning_rate": 9.941454148885226e-06,
+      "loss": 0.0972,
+      "step": 308
+    },
+    {
+      "epoch": 0.2121524201853759,
+      "grad_norm": 0.232421875,
+      "learning_rate": 9.940921874531322e-06,
+      "loss": 0.0927,
+      "step": 309
+    },
+    {
+      "epoch": 0.21283899759697905,
+      "grad_norm": 0.25390625,
+      "learning_rate": 9.94038720749355e-06,
+      "loss": 0.0986,
+      "step": 310
+    },
+    {
+      "epoch": 0.2135255750085822,
+      "grad_norm": 0.220703125,
+      "learning_rate": 9.939850148059983e-06,
+      "loss": 0.0929,
+      "step": 311
+    },
+    {
+      "epoch": 0.21421215242018538,
+      "grad_norm": 0.201171875,
+      "learning_rate": 9.939310696519977e-06,
+      "loss": 0.0928,
+      "step": 312
+    },
+    {
+      "epoch": 0.21489872983178854,
+      "grad_norm": 0.271484375,
+      "learning_rate": 9.938768853164176e-06,
+      "loss": 0.1035,
+      "step": 313
+    },
+    {
+      "epoch": 0.2155853072433917,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 9.93822461828452e-06,
+      "loss": 0.0978,
+      "step": 314
+    },
+    {
+      "epoch": 0.21627188465499486,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 9.937677992174228e-06,
+      "loss": 0.0922,
+      "step": 315
+    },
+    {
+      "epoch": 0.216958462066598,
+      "grad_norm": 0.21875,
+      "learning_rate": 9.937128975127814e-06,
+      "loss": 0.1061,
+      "step": 316
+    },
+    {
+      "epoch": 0.21764503947820116,
+      "grad_norm": 0.275390625,
+      "learning_rate": 9.936577567441074e-06,
+      "loss": 0.0954,
+      "step": 317
+    },
+    {
+      "epoch": 0.21833161688980432,
+      "grad_norm": 0.208984375,
+      "learning_rate": 9.936023769411103e-06,
+      "loss": 0.0877,
+      "step": 318
+    },
+    {
+      "epoch": 0.21901819430140748,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 9.935467581336269e-06,
+      "loss": 0.0985,
+      "step": 319
+    },
+    {
+      "epoch": 0.21970477171301064,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 9.93490900351624e-06,
+      "loss": 0.085,
+      "step": 320
+    },
+    {
+      "epoch": 0.2203913491246138,
+      "grad_norm": 0.201171875,
+      "learning_rate": 9.934348036251969e-06,
+      "loss": 0.0949,
+      "step": 321
+    },
+    {
+      "epoch": 0.22107792653621697,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 9.933784679845687e-06,
+      "loss": 0.101,
+      "step": 322
+    },
+    {
+      "epoch": 0.22176450394782013,
+      "grad_norm": 0.193359375,
+      "learning_rate": 9.933218934600927e-06,
+      "loss": 0.1031,
+      "step": 323
+    },
+    {
+      "epoch": 0.22245108135942326,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 9.9326508008225e-06,
+      "loss": 0.104,
+      "step": 324
+    },
+    {
+      "epoch": 0.22313765877102643,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 9.932080278816503e-06,
+      "loss": 0.0875,
+      "step": 325
+    },
+    {
+      "epoch": 0.2238242361826296,
+      "grad_norm": 0.236328125,
+      "learning_rate": 9.931507368890323e-06,
+      "loss": 0.0937,
+      "step": 326
+    },
+    {
+      "epoch": 0.22451081359423275,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 9.930932071352635e-06,
+      "loss": 0.099,
+      "step": 327
+    },
+    {
+      "epoch": 0.2251973910058359,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 9.930354386513399e-06,
+      "loss": 0.0978,
+      "step": 328
+    },
+    {
+      "epoch": 0.22588396841743907,
+      "grad_norm": 0.2109375,
+      "learning_rate": 9.929774314683856e-06,
+      "loss": 0.095,
+      "step": 329
+    },
+    {
+      "epoch": 0.22657054582904224,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 9.929191856176543e-06,
+      "loss": 0.0905,
+      "step": 330
+    },
+    {
+      "epoch": 0.22725712324064537,
+      "grad_norm": 0.203125,
+      "learning_rate": 9.928607011305273e-06,
+      "loss": 0.0984,
+      "step": 331
+    },
+    {
+      "epoch": 0.22794370065224853,
+      "grad_norm": 0.2109375,
+      "learning_rate": 9.928019780385152e-06,
+      "loss": 0.1044,
+      "step": 332
+    },
+    {
+      "epoch": 0.2286302780638517,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 9.927430163732566e-06,
+      "loss": 0.095,
+      "step": 333
+    },
+    {
+      "epoch": 0.22931685547545486,
+      "grad_norm": 0.212890625,
+      "learning_rate": 9.926838161665195e-06,
+      "loss": 0.0926,
+      "step": 334
+    },
+    {
+      "epoch": 0.23000343288705802,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 9.926243774501993e-06,
+      "loss": 0.0958,
+      "step": 335
+    },
+    {
+      "epoch": 0.23069001029866118,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 9.925647002563205e-06,
+      "loss": 0.0873,
+      "step": 336
+    },
+    {
+      "epoch": 0.23137658771026434,
+      "grad_norm": 0.234375,
+      "learning_rate": 9.92504784617036e-06,
+      "loss": 0.103,
+      "step": 337
+    },
+    {
+      "epoch": 0.2320631651218675,
+      "grad_norm": 0.216796875,
+      "learning_rate": 9.924446305646278e-06,
+      "loss": 0.096,
+      "step": 338
+    },
+    {
+      "epoch": 0.23274974253347064,
+      "grad_norm": 0.244140625,
+      "learning_rate": 9.923842381315049e-06,
+      "loss": 0.097,
+      "step": 339
+    },
+    {
+      "epoch": 0.2334363199450738,
+      "grad_norm": 0.2109375,
+      "learning_rate": 9.92323607350206e-06,
+      "loss": 0.092,
+      "step": 340
+    },
+    {
+      "epoch": 0.23412289735667696,
+      "grad_norm": 0.255859375,
+      "learning_rate": 9.92262738253398e-06,
+      "loss": 0.0976,
+      "step": 341
+    },
+    {
+      "epoch": 0.23480947476828012,
+      "grad_norm": 0.19921875,
+      "learning_rate": 9.922016308738757e-06,
+      "loss": 0.095,
+      "step": 342
+    },
+    {
+      "epoch": 0.2354960521798833,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 9.921402852445627e-06,
+      "loss": 0.0933,
+      "step": 343
+    },
+    {
+      "epoch": 0.23618262959148645,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 9.920787013985106e-06,
+      "loss": 0.0818,
+      "step": 344
+    },
+    {
+      "epoch": 0.2368692070030896,
+      "grad_norm": 0.1953125,
+      "learning_rate": 9.920168793689e-06,
+      "loss": 0.0961,
+      "step": 345
+    },
+    {
+      "epoch": 0.23755578441469274,
+      "grad_norm": 0.212890625,
+      "learning_rate": 9.919548191890395e-06,
+      "loss": 0.0869,
+      "step": 346
+    },
+    {
+      "epoch": 0.2382423618262959,
+      "grad_norm": 0.203125,
+      "learning_rate": 9.918925208923654e-06,
+      "loss": 0.098,
+      "step": 347
+    },
+    {
+      "epoch": 0.23892893923789907,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 9.918299845124433e-06,
+      "loss": 0.102,
+      "step": 348
+    },
+    {
+      "epoch": 0.23961551664950223,
+      "grad_norm": 0.21484375,
+      "learning_rate": 9.917672100829664e-06,
+      "loss": 0.0854,
+      "step": 349
+    },
+    {
+      "epoch": 0.2403020940611054,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 9.917041976377564e-06,
+      "loss": 0.0958,
+      "step": 350
+    },
+    {
+      "epoch": 0.24098867147270855,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 9.916409472107632e-06,
+      "loss": 0.1014,
+      "step": 351
+    },
+    {
+      "epoch": 0.24167524888431172,
+      "grad_norm": 0.2109375,
+      "learning_rate": 9.915774588360649e-06,
+      "loss": 0.0944,
+      "step": 352
+    },
+    {
+      "epoch": 0.24236182629591485,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 9.915137325478677e-06,
+      "loss": 0.1103,
+      "step": 353
+    },
+    {
+      "epoch": 0.243048403707518,
+      "grad_norm": 0.216796875,
+      "learning_rate": 9.914497683805065e-06,
+      "loss": 0.0994,
+      "step": 354
+    },
+    {
+      "epoch": 0.24373498111912117,
+      "grad_norm": 0.228515625,
+      "learning_rate": 9.913855663684438e-06,
+      "loss": 0.0969,
+      "step": 355
+    },
+    {
+      "epoch": 0.24442155853072434,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 9.9132112654627e-06,
+      "loss": 0.0917,
+      "step": 356
+    },
+    {
+      "epoch": 0.2451081359423275,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 9.912564489487047e-06,
+      "loss": 0.1032,
+      "step": 357
+    },
+    {
+      "epoch": 0.24579471335393066,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 9.911915336105943e-06,
+      "loss": 0.0962,
+      "step": 358
+    },
+    {
+      "epoch": 0.24648129076553382,
+      "grad_norm": 0.228515625,
+      "learning_rate": 9.911263805669147e-06,
+      "loss": 0.1077,
+      "step": 359
+    },
+    {
+      "epoch": 0.24716786817713698,
+      "grad_norm": 0.21875,
+      "learning_rate": 9.910609898527686e-06,
+      "loss": 0.0919,
+      "step": 360
+    },
+    {
+      "epoch": 0.24785444558874012,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 9.909953615033872e-06,
+      "loss": 0.0982,
+      "step": 361
+    },
+    {
+      "epoch": 0.24854102300034328,
+      "grad_norm": 0.205078125,
+      "learning_rate": 9.9092949555413e-06,
+      "loss": 0.0987,
+      "step": 362
+    },
+    {
+      "epoch": 0.24922760041194644,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 9.908633920404844e-06,
+      "loss": 0.0919,
+      "step": 363
+    },
+    {
+      "epoch": 0.2499141778235496,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 9.907970509980657e-06,
+      "loss": 0.0949,
+      "step": 364
+    },
+    {
+      "epoch": 0.25060075523515274,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 9.90730472462617e-06,
+      "loss": 0.0891,
+      "step": 365
+    },
+    {
+      "epoch": 0.2512873326467559,
+      "grad_norm": 0.21875,
+      "learning_rate": 9.906636564700096e-06,
+      "loss": 0.0919,
+      "step": 366
+    },
+    {
+      "epoch": 0.25197391005835906,
+      "grad_norm": 0.251953125,
+      "learning_rate": 9.905966030562426e-06,
+      "loss": 0.1023,
+      "step": 367
+    },
+    {
+      "epoch": 0.2526604874699622,
+      "grad_norm": 0.216796875,
+      "learning_rate": 9.905293122574433e-06,
+      "loss": 0.0918,
+      "step": 368
+    },
+    {
+      "epoch": 0.2533470648815654,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 9.904617841098666e-06,
+      "loss": 0.0921,
+      "step": 369
+    },
+    {
+      "epoch": 0.25403364229316855,
+      "grad_norm": 0.26953125,
+      "learning_rate": 9.903940186498953e-06,
+      "loss": 0.0997,
+      "step": 370
+    },
+    {
+      "epoch": 0.2547202197047717,
+      "grad_norm": 0.19921875,
+      "learning_rate": 9.903260159140404e-06,
+      "loss": 0.0927,
+      "step": 371
+    },
+    {
+      "epoch": 0.2554067971163749,
+      "grad_norm": 0.275390625,
+      "learning_rate": 9.902577759389402e-06,
+      "loss": 0.098,
+      "step": 372
+    },
+    {
+      "epoch": 0.25609337452797803,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 9.901892987613612e-06,
+      "loss": 0.1043,
+      "step": 373
+    },
+    {
+      "epoch": 0.2567799519395812,
+      "grad_norm": 0.205078125,
+      "learning_rate": 9.901205844181976e-06,
+      "loss": 0.0867,
+      "step": 374
+    },
+    {
+      "epoch": 0.25746652935118436,
+      "grad_norm": 0.248046875,
+      "learning_rate": 9.900516329464713e-06,
+      "loss": 0.0958,
+      "step": 375
+    },
+    {
+      "epoch": 0.2581531067627875,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 9.89982444383332e-06,
+      "loss": 0.1,
+      "step": 376
+    },
+    {
+      "epoch": 0.2588396841743907,
+      "grad_norm": 0.20703125,
+      "learning_rate": 9.899130187660573e-06,
+      "loss": 0.0964,
+      "step": 377
+    },
+    {
+      "epoch": 0.25952626158599384,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 9.898433561320525e-06,
+      "loss": 0.0912,
+      "step": 378
+    },
+    {
+      "epoch": 0.26021283899759695,
+      "grad_norm": 0.2734375,
+      "learning_rate": 9.897734565188504e-06,
+      "loss": 0.1104,
+      "step": 379
+    },
+    {
+      "epoch": 0.2608994164092001,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 9.897033199641114e-06,
+      "loss": 0.0868,
+      "step": 380
+    },
+    {
+      "epoch": 0.2615859938208033,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 9.896329465056238e-06,
+      "loss": 0.1038,
+      "step": 381
+    },
+    {
+      "epoch": 0.26227257123240644,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 9.895623361813036e-06,
+      "loss": 0.0906,
+      "step": 382
+    },
+    {
+      "epoch": 0.2629591486440096,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 9.894914890291944e-06,
+      "loss": 0.0906,
+      "step": 383
+    },
+    {
+      "epoch": 0.26364572605561276,
+      "grad_norm": 0.201171875,
+      "learning_rate": 9.89420405087467e-06,
+      "loss": 0.1056,
+      "step": 384
+    },
+    {
+      "epoch": 0.2643323034672159,
+      "grad_norm": 0.216796875,
+      "learning_rate": 9.893490843944201e-06,
+      "loss": 0.0885,
+      "step": 385
+    },
+    {
+      "epoch": 0.2650188808788191,
+      "grad_norm": 0.234375,
+      "learning_rate": 9.892775269884802e-06,
+      "loss": 0.0868,
+      "step": 386
+    },
+    {
+      "epoch": 0.26570545829042225,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 9.892057329082009e-06,
+      "loss": 0.0858,
+      "step": 387
+    },
+    {
+      "epoch": 0.2663920357020254,
+      "grad_norm": 0.240234375,
+      "learning_rate": 9.891337021922633e-06,
+      "loss": 0.0905,
+      "step": 388
+    },
+    {
+      "epoch": 0.26707861311362857,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 9.890614348794764e-06,
+      "loss": 0.0923,
+      "step": 389
+    },
+    {
+      "epoch": 0.26776519052523173,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 9.889889310087766e-06,
+      "loss": 0.0966,
+      "step": 390
+    },
+    {
+      "epoch": 0.2684517679368349,
+      "grad_norm": 0.208984375,
+      "learning_rate": 9.889161906192271e-06,
+      "loss": 0.089,
+      "step": 391
+    },
+    {
+      "epoch": 0.26913834534843806,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 9.888432137500194e-06,
+      "loss": 0.0962,
+      "step": 392
+    },
+    {
+      "epoch": 0.2698249227600412,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 9.88770000440472e-06,
+      "loss": 0.0899,
+      "step": 393
+    },
+    {
+      "epoch": 0.2705115001716443,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 9.886965507300309e-06,
+      "loss": 0.0892,
+      "step": 394
+    },
+    {
+      "epoch": 0.2711980775832475,
+      "grad_norm": 0.23046875,
+      "learning_rate": 9.886228646582694e-06,
+      "loss": 0.0982,
+      "step": 395
+    },
+    {
+      "epoch": 0.27188465499485065,
+      "grad_norm": 0.259765625,
+      "learning_rate": 9.885489422648878e-06,
+      "loss": 0.1012,
+      "step": 396
+    },
+    {
+      "epoch": 0.2725712324064538,
+      "grad_norm": 0.234375,
+      "learning_rate": 9.884747835897145e-06,
+      "loss": 0.0991,
+      "step": 397
+    },
+    {
+      "epoch": 0.273257809818057,
+      "grad_norm": 0.205078125,
+      "learning_rate": 9.884003886727044e-06,
+      "loss": 0.0867,
+      "step": 398
+    },
+    {
+      "epoch": 0.27394438722966014,
+      "grad_norm": 0.201171875,
+      "learning_rate": 9.883257575539404e-06,
+      "loss": 0.088,
+      "step": 399
+    },
+    {
+      "epoch": 0.2746309646412633,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 9.88250890273632e-06,
+      "loss": 0.1028,
+      "step": 400
+    },
+    {
+      "epoch": 0.27531754205286646,
+      "grad_norm": 0.19921875,
+      "learning_rate": 9.881757868721166e-06,
+      "loss": 0.0931,
+      "step": 401
+    },
+    {
+      "epoch": 0.2760041194644696,
+      "grad_norm": 0.21484375,
+      "learning_rate": 9.881004473898585e-06,
+      "loss": 0.0885,
+      "step": 402
+    },
+    {
+      "epoch": 0.2766906968760728,
+      "grad_norm": 0.2734375,
+      "learning_rate": 9.880248718674486e-06,
+      "loss": 0.1049,
+      "step": 403
+    },
+    {
+      "epoch": 0.27737727428767595,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 9.879490603456062e-06,
+      "loss": 0.0886,
+      "step": 404
+    },
+    {
+      "epoch": 0.2780638516992791,
+      "grad_norm": 0.2109375,
+      "learning_rate": 9.878730128651768e-06,
+      "loss": 0.1003,
+      "step": 405
+    },
+    {
+      "epoch": 0.27875042911088227,
+      "grad_norm": 0.203125,
+      "learning_rate": 9.877967294671333e-06,
+      "loss": 0.102,
+      "step": 406
+    },
+    {
+      "epoch": 0.27943700652248543,
+      "grad_norm": 0.234375,
+      "learning_rate": 9.87720210192576e-06,
+      "loss": 0.0914,
+      "step": 407
+    },
+    {
+      "epoch": 0.2801235839340886,
+      "grad_norm": 0.2109375,
+      "learning_rate": 9.876434550827315e-06,
+      "loss": 0.0903,
+      "step": 408
+    },
+    {
+      "epoch": 0.2808101613456917,
+      "grad_norm": 0.2265625,
+      "learning_rate": 9.875664641789545e-06,
+      "loss": 0.1074,
+      "step": 409
+    },
+    {
+      "epoch": 0.28149673875729486,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 9.874892375227262e-06,
+      "loss": 0.099,
+      "step": 410
+    },
+    {
+      "epoch": 0.282183316168898,
+      "grad_norm": 0.23046875,
+      "learning_rate": 9.874117751556544e-06,
+      "loss": 0.0941,
+      "step": 411
+    },
+    {
+      "epoch": 0.2828698935805012,
+      "grad_norm": 0.203125,
+      "learning_rate": 9.873340771194749e-06,
+      "loss": 0.0936,
+      "step": 412
+    },
+    {
+      "epoch": 0.28355647099210435,
+      "grad_norm": 0.224609375,
+      "learning_rate": 9.872561434560493e-06,
+      "loss": 0.0957,
+      "step": 413
+    },
+    {
+      "epoch": 0.2842430484037075,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 9.871779742073675e-06,
+      "loss": 0.0835,
+      "step": 414
+    },
+    {
+      "epoch": 0.28492962581531067,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 9.870995694155449e-06,
+      "loss": 0.0935,
+      "step": 415
+    },
+    {
+      "epoch": 0.28561620322691383,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 9.87020929122825e-06,
+      "loss": 0.0938,
+      "step": 416
+    },
+    {
+      "epoch": 0.286302780638517,
+      "grad_norm": 0.2421875,
+      "learning_rate": 9.869420533715777e-06,
+      "loss": 0.0766,
+      "step": 417
+    },
+    {
+      "epoch": 0.28698935805012016,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 9.868629422042994e-06,
+      "loss": 0.1023,
+      "step": 418
+    },
+    {
+      "epoch": 0.2876759354617233,
+      "grad_norm": 0.251953125,
+      "learning_rate": 9.867835956636137e-06,
+      "loss": 0.0972,
+      "step": 419
+    },
+    {
+      "epoch": 0.2883625128733265,
+      "grad_norm": 0.2265625,
+      "learning_rate": 9.867040137922712e-06,
+      "loss": 0.1057,
+      "step": 420
+    },
+    {
+      "epoch": 0.28904909028492964,
+      "grad_norm": 0.212890625,
+      "learning_rate": 9.866241966331491e-06,
+      "loss": 0.1022,
+      "step": 421
+    },
+    {
+      "epoch": 0.2897356676965328,
+      "grad_norm": 0.197265625,
+      "learning_rate": 9.865441442292513e-06,
+      "loss": 0.09,
+      "step": 422
+    },
+    {
+      "epoch": 0.29042224510813597,
+      "grad_norm": 0.20703125,
+      "learning_rate": 9.864638566237084e-06,
+      "loss": 0.0921,
+      "step": 423
+    },
+    {
+      "epoch": 0.2911088225197391,
+      "grad_norm": 0.22265625,
+      "learning_rate": 9.86383333859778e-06,
+      "loss": 0.0901,
+      "step": 424
+    },
+    {
+      "epoch": 0.29179539993134224,
+      "grad_norm": 0.234375,
+      "learning_rate": 9.863025759808444e-06,
+      "loss": 0.0955,
+      "step": 425
+    },
+    {
+      "epoch": 0.2924819773429454,
+      "grad_norm": 0.232421875,
+      "learning_rate": 9.862215830304178e-06,
+      "loss": 0.0903,
+      "step": 426
+    },
+    {
+      "epoch": 0.29316855475454856,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 9.861403550521361e-06,
+      "loss": 0.1042,
+      "step": 427
+    },
+    {
+      "epoch": 0.2938551321661517,
+      "grad_norm": 0.21875,
+      "learning_rate": 9.860588920897633e-06,
+      "loss": 0.0906,
+      "step": 428
+    },
+    {
+      "epoch": 0.2945417095777549,
+      "grad_norm": 0.248046875,
+      "learning_rate": 9.859771941871903e-06,
+      "loss": 0.0827,
+      "step": 429
+    },
+    {
+      "epoch": 0.29522828698935805,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 9.858952613884339e-06,
+      "loss": 0.1057,
+      "step": 430
+    },
+    {
+      "epoch": 0.2959148644009612,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 9.858130937376384e-06,
+      "loss": 0.0979,
+      "step": 431
+    },
+    {
+      "epoch": 0.29660144181256437,
+      "grad_norm": 0.21875,
+      "learning_rate": 9.857306912790737e-06,
+      "loss": 0.1004,
+      "step": 432
+    },
+    {
+      "epoch": 0.29728801922416753,
+      "grad_norm": 0.234375,
+      "learning_rate": 9.85648054057137e-06,
+      "loss": 0.0941,
+      "step": 433
+    },
+    {
+      "epoch": 0.2979745966357707,
+      "grad_norm": 0.208984375,
+      "learning_rate": 9.855651821163516e-06,
+      "loss": 0.0807,
+      "step": 434
+    },
+    {
+      "epoch": 0.29866117404737386,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 9.854820755013672e-06,
+      "loss": 0.0885,
+      "step": 435
+    },
+    {
+      "epoch": 0.299347751458977,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 9.853987342569604e-06,
+      "loss": 0.0833,
+      "step": 436
+    },
+    {
+      "epoch": 0.3000343288705802,
+      "grad_norm": 0.201171875,
+      "learning_rate": 9.853151584280332e-06,
+      "loss": 0.0822,
+      "step": 437
+    },
+    {
+      "epoch": 0.30072090628218334,
+      "grad_norm": 0.224609375,
+      "learning_rate": 9.852313480596155e-06,
+      "loss": 0.0978,
+      "step": 438
+    },
+    {
+      "epoch": 0.30140748369378645,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 9.851473031968621e-06,
+      "loss": 0.0953,
+      "step": 439
+    },
+    {
+      "epoch": 0.3020940611053896,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 9.850630238850549e-06,
+      "loss": 0.1123,
+      "step": 440
+    },
+    {
+      "epoch": 0.3027806385169928,
+      "grad_norm": 0.208984375,
+      "learning_rate": 9.849785101696022e-06,
+      "loss": 0.0968,
+      "step": 441
+    },
+    {
+      "epoch": 0.30346721592859593,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 9.848937620960382e-06,
+      "loss": 0.0934,
+      "step": 442
+    },
+    {
+      "epoch": 0.3041537933401991,
+      "grad_norm": 0.25,
+      "learning_rate": 9.848087797100234e-06,
+      "loss": 0.0853,
+      "step": 443
+    },
+    {
+      "epoch": 0.30484037075180226,
+      "grad_norm": 0.228515625,
+      "learning_rate": 9.84723563057345e-06,
+      "loss": 0.09,
+      "step": 444
+    },
+    {
+      "epoch": 0.3055269481634054,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 9.84638112183916e-06,
+      "loss": 0.0927,
+      "step": 445
+    },
+    {
+      "epoch": 0.3062135255750086,
+      "grad_norm": 0.220703125,
+      "learning_rate": 9.845524271357757e-06,
+      "loss": 0.08,
+      "step": 446
+    },
+    {
+      "epoch": 0.30690010298661174,
+      "grad_norm": 0.212890625,
+      "learning_rate": 9.844665079590892e-06,
+      "loss": 0.0847,
+      "step": 447
+    },
+    {
+      "epoch": 0.3075866803982149,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 9.843803547001487e-06,
+      "loss": 0.0904,
+      "step": 448
+    },
+    {
+      "epoch": 0.30827325780981807,
+      "grad_norm": 0.21875,
+      "learning_rate": 9.842939674053715e-06,
+      "loss": 0.0883,
+      "step": 449
+    },
+    {
+      "epoch": 0.30895983522142123,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 9.842073461213017e-06,
+      "loss": 0.0949,
+      "step": 450
+    },
+    {
+      "epoch": 0.3096464126330244,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 9.841204908946091e-06,
+      "loss": 0.0871,
+      "step": 451
+    },
+    {
+      "epoch": 0.31033299004462755,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 9.840334017720896e-06,
+      "loss": 0.0929,
+      "step": 452
+    },
+    {
+      "epoch": 0.3110195674562307,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 9.839460788006652e-06,
+      "loss": 0.087,
+      "step": 453
+    },
+    {
+      "epoch": 0.3117061448678338,
+      "grad_norm": 0.232421875,
+      "learning_rate": 9.838585220273837e-06,
+      "loss": 0.0859,
+      "step": 454
+    },
+    {
+      "epoch": 0.312392722279437,
+      "grad_norm": 0.21875,
+      "learning_rate": 9.837707314994192e-06,
+      "loss": 0.0918,
+      "step": 455
+    },
+    {
+      "epoch": 0.31307929969104015,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 9.836827072640716e-06,
+      "loss": 0.0972,
+      "step": 456
+    },
+    {
+      "epoch": 0.3137658771026433,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 9.835944493687665e-06,
+      "loss": 0.0952,
+      "step": 457
+    },
+    {
+      "epoch": 0.31445245451424647,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 9.835059578610556e-06,
+      "loss": 0.1021,
+      "step": 458
+    },
+    {
+      "epoch": 0.31513903192584963,
+      "grad_norm": 0.232421875,
+      "learning_rate": 9.834172327886166e-06,
+      "loss": 0.0904,
+      "step": 459
+    },
+    {
+      "epoch": 0.3158256093374528,
+      "grad_norm": 0.259765625,
+      "learning_rate": 9.833282741992526e-06,
+      "loss": 0.1068,
+      "step": 460
+    },
+    {
+      "epoch": 0.31651218674905596,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 9.832390821408929e-06,
+      "loss": 0.0954,
+      "step": 461
+    },
+    {
+      "epoch": 0.3171987641606591,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 9.831496566615927e-06,
+      "loss": 0.1049,
+      "step": 462
+    },
+    {
+      "epoch": 0.3178853415722623,
+      "grad_norm": 0.220703125,
+      "learning_rate": 9.830599978095323e-06,
+      "loss": 0.093,
+      "step": 463
+    },
+    {
+      "epoch": 0.31857191898386544,
+      "grad_norm": 0.20703125,
+      "learning_rate": 9.829701056330188e-06,
+      "loss": 0.1035,
+      "step": 464
+    },
+    {
+      "epoch": 0.3192584963954686,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 9.828799801804837e-06,
+      "loss": 0.0903,
+      "step": 465
+    },
+    {
+      "epoch": 0.31994507380707177,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 9.827896215004853e-06,
+      "loss": 0.0996,
+      "step": 466
+    },
+    {
+      "epoch": 0.32063165121867493,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 9.826990296417071e-06,
+      "loss": 0.0985,
+      "step": 467
+    },
+    {
+      "epoch": 0.3213182286302781,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 9.826082046529581e-06,
+      "loss": 0.0991,
+      "step": 468
+    },
+    {
+      "epoch": 0.3220048060418812,
+      "grad_norm": 0.19921875,
+      "learning_rate": 9.825171465831732e-06,
+      "loss": 0.0898,
+      "step": 469
+    },
+    {
+      "epoch": 0.32269138345348436,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 9.824258554814126e-06,
+      "loss": 0.0929,
+      "step": 470
+    },
+    {
+      "epoch": 0.3233779608650875,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 9.823343313968624e-06,
+      "loss": 0.0918,
+      "step": 471
+    },
+    {
+      "epoch": 0.3240645382766907,
+      "grad_norm": 0.21484375,
+      "learning_rate": 9.82242574378834e-06,
+      "loss": 0.0935,
+      "step": 472
+    },
+    {
+      "epoch": 0.32475111568829385,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 9.821505844767642e-06,
+      "loss": 0.0906,
+      "step": 473
+    },
+    {
+      "epoch": 0.325437693099897,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 9.820583617402153e-06,
+      "loss": 0.0886,
+      "step": 474
+    },
+    {
+      "epoch": 0.32612427051150017,
+      "grad_norm": 0.224609375,
+      "learning_rate": 9.819659062188754e-06,
+      "loss": 0.0859,
+      "step": 475
+    },
+    {
+      "epoch": 0.32681084792310333,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 9.818732179625578e-06,
+      "loss": 0.0906,
+      "step": 476
+    },
+    {
+      "epoch": 0.3274974253347065,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 9.817802970212009e-06,
+      "loss": 0.091,
+      "step": 477
+    },
+    {
+      "epoch": 0.32818400274630966,
+      "grad_norm": 0.267578125,
+      "learning_rate": 9.81687143444869e-06,
+      "loss": 0.097,
+      "step": 478
+    },
+    {
+      "epoch": 0.3288705801579128,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 9.815937572837511e-06,
+      "loss": 0.0868,
+      "step": 479
+    },
+    {
+      "epoch": 0.329557157569516,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 9.815001385881624e-06,
+      "loss": 0.0853,
+      "step": 480
+    },
+    {
+      "epoch": 0.33024373498111914,
+      "grad_norm": 0.25,
+      "learning_rate": 9.814062874085424e-06,
+      "loss": 0.0875,
+      "step": 481
+    },
+    {
+      "epoch": 0.3309303123927223,
+      "grad_norm": 0.234375,
+      "learning_rate": 9.813122037954567e-06,
+      "loss": 0.1021,
+      "step": 482
+    },
+    {
+      "epoch": 0.33161688980432547,
+      "grad_norm": 0.212890625,
+      "learning_rate": 9.812178877995954e-06,
+      "loss": 0.0872,
+      "step": 483
+    },
+    {
+      "epoch": 0.33230346721592857,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 9.811233394717742e-06,
+      "loss": 0.0862,
+      "step": 484
+    },
+    {
+      "epoch": 0.33299004462753173,
+      "grad_norm": 0.22265625,
+      "learning_rate": 9.810285588629342e-06,
+      "loss": 0.1053,
+      "step": 485
+    },
+    {
+      "epoch": 0.3336766220391349,
+      "grad_norm": 0.21875,
+      "learning_rate": 9.809335460241412e-06,
+      "loss": 0.0936,
+      "step": 486
+    },
+    {
+      "epoch": 0.33436319945073806,
+      "grad_norm": 0.20703125,
+      "learning_rate": 9.808383010065863e-06,
+      "loss": 0.0917,
+      "step": 487
+    },
+    {
+      "epoch": 0.3350497768623412,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 9.807428238615858e-06,
+      "loss": 0.0797,
+      "step": 488
+    },
+    {
+      "epoch": 0.3357363542739444,
+      "grad_norm": 0.22265625,
+      "learning_rate": 9.806471146405809e-06,
+      "loss": 0.0943,
+      "step": 489
+    },
+    {
+      "epoch": 0.33642293168554754,
+      "grad_norm": 0.189453125,
+      "learning_rate": 9.805511733951379e-06,
+      "loss": 0.0846,
+      "step": 490
+    },
+    {
+      "epoch": 0.3371095090971507,
+      "grad_norm": 0.234375,
+      "learning_rate": 9.804550001769478e-06,
+      "loss": 0.0947,
+      "step": 491
+    },
+    {
+      "epoch": 0.33779608650875387,
+      "grad_norm": 0.224609375,
+      "learning_rate": 9.803585950378274e-06,
+      "loss": 0.0914,
+      "step": 492
+    },
+    {
+      "epoch": 0.33848266392035703,
+      "grad_norm": 0.234375,
+      "learning_rate": 9.802619580297178e-06,
+      "loss": 0.0881,
+      "step": 493
+    },
+    {
+      "epoch": 0.3391692413319602,
+      "grad_norm": 0.248046875,
+      "learning_rate": 9.801650892046851e-06,
+      "loss": 0.0989,
+      "step": 494
+    },
+    {
+      "epoch": 0.33985581874356335,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 9.800679886149203e-06,
+      "loss": 0.0987,
+      "step": 495
+    },
+    {
+      "epoch": 0.3405423961551665,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 9.799706563127395e-06,
+      "loss": 0.094,
+      "step": 496
+    },
+    {
+      "epoch": 0.3412289735667697,
+      "grad_norm": 0.255859375,
+      "learning_rate": 9.798730923505833e-06,
+      "loss": 0.0847,
+      "step": 497
+    },
+    {
+      "epoch": 0.3419155509783728,
+      "grad_norm": 0.224609375,
+      "learning_rate": 9.797752967810176e-06,
+      "loss": 0.0962,
+      "step": 498
+    },
+    {
+      "epoch": 0.34260212838997595,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 9.796772696567323e-06,
+      "loss": 0.087,
+      "step": 499
+    },
+    {
+      "epoch": 0.3432887058015791,
+      "grad_norm": 0.28125,
+      "learning_rate": 9.795790110305431e-06,
+      "loss": 0.0883,
+      "step": 500
+    },
+    {
+      "epoch": 0.34397528321318227,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 9.794805209553896e-06,
+      "loss": 0.0878,
+      "step": 501
+    },
+    {
+      "epoch": 0.34466186062478543,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 9.793817994843362e-06,
+      "loss": 0.0869,
+      "step": 502
+    },
+    {
+      "epoch": 0.3453484380363886,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 9.792828466705725e-06,
+      "loss": 0.1071,
+      "step": 503
+    },
+    {
+      "epoch": 0.34603501544799176,
+      "grad_norm": 0.236328125,
+      "learning_rate": 9.79183662567412e-06,
+      "loss": 0.0876,
+      "step": 504
+    },
+    {
+      "epoch": 0.3467215928595949,
+      "grad_norm": 0.248046875,
+      "learning_rate": 9.790842472282935e-06,
+      "loss": 0.0966,
+      "step": 505
+    },
+    {
+      "epoch": 0.3474081702711981,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 9.789846007067802e-06,
+      "loss": 0.0864,
+      "step": 506
+    },
+    {
+      "epoch": 0.34809474768280124,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 9.788847230565592e-06,
+      "loss": 0.0874,
+      "step": 507
+    },
+    {
+      "epoch": 0.3487813250944044,
+      "grad_norm": 0.203125,
+      "learning_rate": 9.787846143314433e-06,
+      "loss": 0.095,
+      "step": 508
+    },
+    {
+      "epoch": 0.34946790250600757,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 9.786842745853685e-06,
+      "loss": 0.0874,
+      "step": 509
+    },
+    {
+      "epoch": 0.35015447991761073,
+      "grad_norm": 0.197265625,
+      "learning_rate": 9.785837038723966e-06,
+      "loss": 0.094,
+      "step": 510
+    },
+    {
+      "epoch": 0.3508410573292139,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 9.784829022467128e-06,
+      "loss": 0.0997,
+      "step": 511
+    },
+    {
+      "epoch": 0.35152763474081705,
+      "grad_norm": 0.255859375,
+      "learning_rate": 9.783818697626273e-06,
+      "loss": 0.0945,
+      "step": 512
+    },
+    {
+      "epoch": 0.35221421215242016,
+      "grad_norm": 0.189453125,
+      "learning_rate": 9.782806064745742e-06,
+      "loss": 0.0823,
+      "step": 513
+    },
+    {
+      "epoch": 0.3529007895640233,
+      "grad_norm": 0.21484375,
+      "learning_rate": 9.781791124371124e-06,
+      "loss": 0.0941,
+      "step": 514
+    },
+    {
+      "epoch": 0.3535873669756265,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 9.78077387704925e-06,
+      "loss": 0.0885,
+      "step": 515
+    },
+    {
+      "epoch": 0.35427394438722964,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 9.779754323328192e-06,
+      "loss": 0.0902,
+      "step": 516
+    },
+    {
+      "epoch": 0.3549605217988328,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 9.778732463757267e-06,
+      "loss": 0.0936,
+      "step": 517
+    },
+    {
+      "epoch": 0.35564709921043597,
+      "grad_norm": 0.263671875,
+      "learning_rate": 9.777708298887034e-06,
+      "loss": 0.0913,
+      "step": 518
+    },
+    {
+      "epoch": 0.35633367662203913,
+      "grad_norm": 0.26953125,
+      "learning_rate": 9.776681829269291e-06,
+      "loss": 0.0778,
+      "step": 519
+    },
+    {
+      "epoch": 0.3570202540336423,
+      "grad_norm": 0.21875,
+      "learning_rate": 9.775653055457082e-06,
+      "loss": 0.0873,
+      "step": 520
+    },
+    {
+      "epoch": 0.35770683144524545,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 9.774621978004692e-06,
+      "loss": 0.0895,
+      "step": 521
+    },
+    {
+      "epoch": 0.3583934088568486,
+      "grad_norm": 0.2578125,
+      "learning_rate": 9.773588597467642e-06,
+      "loss": 0.089,
+      "step": 522
+    },
+    {
+      "epoch": 0.3590799862684518,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 9.772552914402701e-06,
+      "loss": 0.0893,
+      "step": 523
+    },
+    {
+      "epoch": 0.35976656368005494,
+      "grad_norm": 0.2734375,
+      "learning_rate": 9.771514929367875e-06,
+      "loss": 0.0975,
+      "step": 524
+    },
+    {
+      "epoch": 0.3604531410916581,
+      "grad_norm": 0.25,
+      "learning_rate": 9.77047464292241e-06,
+      "loss": 0.0911,
+      "step": 525
+    },
+    {
+      "epoch": 0.36113971850326126,
+      "grad_norm": 0.23828125,
+      "learning_rate": 9.76943205562679e-06,
+      "loss": 0.0896,
+      "step": 526
+    },
+    {
+      "epoch": 0.3618262959148644,
+      "grad_norm": 0.205078125,
+      "learning_rate": 9.768387168042745e-06,
+      "loss": 0.0857,
+      "step": 527
+    },
+    {
+      "epoch": 0.36251287332646753,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 9.76733998073324e-06,
+      "loss": 0.0957,
+      "step": 528
+    },
+    {
+      "epoch": 0.3631994507380707,
+      "grad_norm": 0.189453125,
+      "learning_rate": 9.766290494262477e-06,
+      "loss": 0.0909,
+      "step": 529
+    },
+    {
+      "epoch": 0.36388602814967386,
+      "grad_norm": 0.2734375,
+      "learning_rate": 9.7652387091959e-06,
+      "loss": 0.1133,
+      "step": 530
+    },
+    {
+      "epoch": 0.364572605561277,
+      "grad_norm": 0.2421875,
+      "learning_rate": 9.764184626100193e-06,
+      "loss": 0.088,
+      "step": 531
+    },
+    {
+      "epoch": 0.3652591829728802,
+      "grad_norm": 0.240234375,
+      "learning_rate": 9.763128245543272e-06,
+      "loss": 0.0921,
+      "step": 532
+    },
+    {
+      "epoch": 0.36594576038448334,
+      "grad_norm": 0.197265625,
+      "learning_rate": 9.7620695680943e-06,
+      "loss": 0.093,
+      "step": 533
+    },
+    {
+      "epoch": 0.3666323377960865,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 9.761008594323666e-06,
+      "loss": 0.0937,
+      "step": 534
+    },
+    {
+      "epoch": 0.36731891520768967,
+      "grad_norm": 0.251953125,
+      "learning_rate": 9.759945324803006e-06,
+      "loss": 0.0908,
+      "step": 535
+    },
+    {
+      "epoch": 0.36800549261929283,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 9.75887976010519e-06,
+      "loss": 0.0848,
+      "step": 536
+    },
+    {
+      "epoch": 0.368692070030896,
+      "grad_norm": 0.228515625,
+      "learning_rate": 9.75781190080432e-06,
+      "loss": 0.0993,
+      "step": 537
+    },
+    {
+      "epoch": 0.36937864744249915,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 9.756741747475744e-06,
+      "loss": 0.0925,
+      "step": 538
+    },
+    {
+      "epoch": 0.3700652248541023,
+      "grad_norm": 0.21484375,
+      "learning_rate": 9.755669300696035e-06,
+      "loss": 0.0982,
+      "step": 539
+    },
+    {
+      "epoch": 0.3707518022657055,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 9.75459456104301e-06,
+      "loss": 0.088,
+      "step": 540
+    },
+    {
+      "epoch": 0.37143837967730864,
+      "grad_norm": 0.1953125,
+      "learning_rate": 9.753517529095716e-06,
+      "loss": 0.0958,
+      "step": 541
+    },
+    {
+      "epoch": 0.3721249570889118,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 9.752438205434439e-06,
+      "loss": 0.0841,
+      "step": 542
+    },
+    {
+      "epoch": 0.3728115345005149,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 9.751356590640696e-06,
+      "loss": 0.089,
+      "step": 543
+    },
+    {
+      "epoch": 0.37349811191211807,
+      "grad_norm": 0.20703125,
+      "learning_rate": 9.750272685297241e-06,
+      "loss": 0.094,
+      "step": 544
+    },
+    {
+      "epoch": 0.37418468932372123,
+      "grad_norm": 0.212890625,
+      "learning_rate": 9.749186489988065e-06,
+      "loss": 0.092,
+      "step": 545
+    },
+    {
+      "epoch": 0.3748712667353244,
+      "grad_norm": 0.208984375,
+      "learning_rate": 9.748098005298384e-06,
+      "loss": 0.0963,
+      "step": 546
+    },
+    {
+      "epoch": 0.37555784414692756,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 9.747007231814656e-06,
+      "loss": 0.0936,
+      "step": 547
+    },
+    {
+      "epoch": 0.3762444215585307,
+      "grad_norm": 0.255859375,
+      "learning_rate": 9.745914170124568e-06,
+      "loss": 0.0988,
+      "step": 548
+    },
+    {
+      "epoch": 0.3769309989701339,
+      "grad_norm": 0.1875,
+      "learning_rate": 9.74481882081704e-06,
+      "loss": 0.0847,
+      "step": 549
+    },
+    {
+      "epoch": 0.37761757638173704,
+      "grad_norm": 0.26171875,
+      "learning_rate": 9.743721184482226e-06,
+      "loss": 0.0951,
+      "step": 550
+    },
+    {
+      "epoch": 0.3783041537933402,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 9.742621261711512e-06,
+      "loss": 0.0889,
+      "step": 551
+    },
+    {
+      "epoch": 0.37899073120494337,
+      "grad_norm": 0.20703125,
+      "learning_rate": 9.741519053097516e-06,
+      "loss": 0.0962,
+      "step": 552
+    },
+    {
+      "epoch": 0.3796773086165465,
+      "grad_norm": 0.2109375,
+      "learning_rate": 9.740414559234085e-06,
+      "loss": 0.0914,
+      "step": 553
+    },
+    {
+      "epoch": 0.3803638860281497,
+      "grad_norm": 0.22265625,
+      "learning_rate": 9.739307780716301e-06,
+      "loss": 0.102,
+      "step": 554
+    },
+    {
+      "epoch": 0.38105046343975285,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 9.738198718140471e-06,
+      "loss": 0.0872,
+      "step": 555
+    },
+    {
+      "epoch": 0.381737040851356,
+      "grad_norm": 0.255859375,
+      "learning_rate": 9.737087372104143e-06,
+      "loss": 0.0892,
+      "step": 556
+    },
+    {
+      "epoch": 0.3824236182629592,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 9.735973743206085e-06,
+      "loss": 0.0966,
+      "step": 557
+    },
+    {
+      "epoch": 0.3831101956745623,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 9.7348578320463e-06,
+      "loss": 0.0916,
+      "step": 558
+    },
+    {
+      "epoch": 0.38379677308616544,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 9.73373963922602e-06,
+      "loss": 0.0992,
+      "step": 559
+    },
+    {
+      "epoch": 0.3844833504977686,
+      "grad_norm": 0.244140625,
+      "learning_rate": 9.732619165347705e-06,
+      "loss": 0.0834,
+      "step": 560
+    },
+    {
+      "epoch": 0.38516992790937177,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 9.731496411015046e-06,
+      "loss": 0.0895,
+      "step": 561
+    },
+    {
+      "epoch": 0.38585650532097493,
+      "grad_norm": 0.25390625,
+      "learning_rate": 9.73037137683296e-06,
+      "loss": 0.0962,
+      "step": 562
+    },
+    {
+      "epoch": 0.3865430827325781,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 9.729244063407594e-06,
+      "loss": 0.0891,
+      "step": 563
+    },
+    {
+      "epoch": 0.38722966014418125,
+      "grad_norm": 0.220703125,
+      "learning_rate": 9.728114471346324e-06,
+      "loss": 0.0876,
+      "step": 564
+    },
+    {
+      "epoch": 0.3879162375557844,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 9.726982601257755e-06,
+      "loss": 0.104,
+      "step": 565
+    },
+    {
+      "epoch": 0.3886028149673876,
+      "grad_norm": 0.251953125,
+      "learning_rate": 9.725848453751712e-06,
+      "loss": 0.0966,
+      "step": 566
+    },
+    {
+      "epoch": 0.38928939237899074,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 9.724712029439255e-06,
+      "loss": 0.0911,
+      "step": 567
+    },
+    {
+      "epoch": 0.3899759697905939,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 9.723573328932669e-06,
+      "loss": 0.0952,
+      "step": 568
+    },
+    {
+      "epoch": 0.39066254720219706,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 9.722432352845458e-06,
+      "loss": 0.092,
+      "step": 569
+    },
+    {
+      "epoch": 0.3913491246138002,
+      "grad_norm": 0.193359375,
+      "learning_rate": 9.721289101792367e-06,
+      "loss": 0.091,
+      "step": 570
+    },
+    {
+      "epoch": 0.3920357020254034,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 9.72014357638935e-06,
+      "loss": 0.0921,
+      "step": 571
+    },
+    {
+      "epoch": 0.39272227943700655,
+      "grad_norm": 0.220703125,
+      "learning_rate": 9.718995777253598e-06,
+      "loss": 0.0881,
+      "step": 572
+    },
+    {
+      "epoch": 0.39340885684860966,
+      "grad_norm": 0.173828125,
+      "learning_rate": 9.717845705003523e-06,
+      "loss": 0.0906,
+      "step": 573
+    },
+    {
+      "epoch": 0.3940954342602128,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 9.716693360258761e-06,
+      "loss": 0.0937,
+      "step": 574
+    },
+    {
+      "epoch": 0.394782011671816,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 9.715538743640177e-06,
+      "loss": 0.093,
+      "step": 575
+    },
+    {
+      "epoch": 0.39546858908341914,
+      "grad_norm": 0.21875,
+      "learning_rate": 9.71438185576985e-06,
+      "loss": 0.0853,
+      "step": 576
+    },
+    {
+      "epoch": 0.3961551664950223,
+      "grad_norm": 0.208984375,
+      "learning_rate": 9.71322269727109e-06,
+      "loss": 0.1017,
+      "step": 577
+    },
+    {
+      "epoch": 0.39684174390662547,
+      "grad_norm": 0.255859375,
+      "learning_rate": 9.712061268768436e-06,
+      "loss": 0.0893,
+      "step": 578
+    },
+    {
+      "epoch": 0.39752832131822863,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 9.710897570887639e-06,
+      "loss": 0.0971,
+      "step": 579
+    },
+    {
+      "epoch": 0.3982148987298318,
+      "grad_norm": 0.216796875,
+      "learning_rate": 9.709731604255675e-06,
+      "loss": 0.0851,
+      "step": 580
+    },
+    {
+      "epoch": 0.39890147614143495,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 9.70856336950075e-06,
+      "loss": 0.0945,
+      "step": 581
+    },
+    {
+      "epoch": 0.3995880535530381,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 9.707392867252282e-06,
+      "loss": 0.0988,
+      "step": 582
+    },
+    {
+      "epoch": 0.4002746309646413,
+      "grad_norm": 0.212890625,
+      "learning_rate": 9.706220098140917e-06,
+      "loss": 0.0804,
+      "step": 583
+    },
+    {
+      "epoch": 0.40096120837624444,
+      "grad_norm": 0.201171875,
+      "learning_rate": 9.705045062798519e-06,
+      "loss": 0.1108,
+      "step": 584
+    },
+    {
+      "epoch": 0.4016477857878476,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 9.703867761858177e-06,
+      "loss": 0.1004,
+      "step": 585
+    },
+    {
+      "epoch": 0.40233436319945076,
+      "grad_norm": 0.244140625,
+      "learning_rate": 9.702688195954198e-06,
+      "loss": 0.0885,
+      "step": 586
+    },
+    {
+      "epoch": 0.4030209406110539,
+      "grad_norm": 0.2421875,
+      "learning_rate": 9.70150636572211e-06,
+      "loss": 0.0875,
+      "step": 587
+    },
+    {
+      "epoch": 0.40370751802265703,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 9.700322271798657e-06,
+      "loss": 0.1066,
+      "step": 588
+    },
+    {
+      "epoch": 0.4043940954342602,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 9.69913591482181e-06,
+      "loss": 0.0932,
+      "step": 589
+    },
+    {
+      "epoch": 0.40508067284586335,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 9.697947295430754e-06,
+      "loss": 0.0834,
+      "step": 590
+    },
+    {
+      "epoch": 0.4057672502574665,
+      "grad_norm": 0.234375,
+      "learning_rate": 9.696756414265894e-06,
+      "loss": 0.0933,
+      "step": 591
+    },
+    {
+      "epoch": 0.4064538276690697,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 9.695563271968853e-06,
+      "loss": 0.1106,
+      "step": 592
+    },
+    {
+      "epoch": 0.40714040508067284,
+      "grad_norm": 0.197265625,
+      "learning_rate": 9.694367869182479e-06,
+      "loss": 0.088,
+      "step": 593
+    },
+    {
+      "epoch": 0.407826982492276,
+      "grad_norm": 0.251953125,
+      "learning_rate": 9.693170206550824e-06,
+      "loss": 0.0869,
+      "step": 594
+    },
+    {
+      "epoch": 0.40851355990387916,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 9.69197028471917e-06,
+      "loss": 0.0909,
+      "step": 595
+    },
+    {
+      "epoch": 0.4092001373154823,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 9.690768104334015e-06,
+      "loss": 0.0981,
+      "step": 596
+    },
+    {
+      "epoch": 0.4098867147270855,
+      "grad_norm": 0.23046875,
+      "learning_rate": 9.689563666043065e-06,
+      "loss": 0.0959,
+      "step": 597
+    },
+    {
+      "epoch": 0.41057329213868865,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 9.688356970495252e-06,
+      "loss": 0.0975,
+      "step": 598
+    },
+    {
+      "epoch": 0.4112598695502918,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 9.687148018340717e-06,
+      "loss": 0.0953,
+      "step": 599
+    },
+    {
+      "epoch": 0.411946446961895,
+      "grad_norm": 0.20703125,
+      "learning_rate": 9.685936810230824e-06,
+      "loss": 0.0924,
+      "step": 600
+    },
+    {
+      "epoch": 0.41263302437349814,
+      "grad_norm": 0.2109375,
+      "learning_rate": 9.684723346818149e-06,
+      "loss": 0.1025,
+      "step": 601
+    },
+    {
+      "epoch": 0.41331960178510124,
+      "grad_norm": 0.220703125,
+      "learning_rate": 9.683507628756477e-06,
+      "loss": 0.0843,
+      "step": 602
+    },
+    {
+      "epoch": 0.4140061791967044,
+      "grad_norm": 0.21875,
+      "learning_rate": 9.682289656700823e-06,
+      "loss": 0.0928,
+      "step": 603
+    },
+    {
+      "epoch": 0.41469275660830757,
+      "grad_norm": 0.1953125,
+      "learning_rate": 9.6810694313074e-06,
+      "loss": 0.0798,
+      "step": 604
+    },
+    {
+      "epoch": 0.41537933401991073,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 9.679846953233644e-06,
+      "loss": 0.0853,
+      "step": 605
+    },
+    {
+      "epoch": 0.4160659114315139,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 9.678622223138203e-06,
+      "loss": 0.1036,
+      "step": 606
+    },
+    {
+      "epoch": 0.41675248884311705,
+      "grad_norm": 0.212890625,
+      "learning_rate": 9.677395241680939e-06,
+      "loss": 0.0824,
+      "step": 607
+    },
+    {
+      "epoch": 0.4174390662547202,
+      "grad_norm": 0.283203125,
+      "learning_rate": 9.676166009522925e-06,
+      "loss": 0.0829,
+      "step": 608
+    },
+    {
+      "epoch": 0.4181256436663234,
+      "grad_norm": 0.2109375,
+      "learning_rate": 9.674934527326447e-06,
+      "loss": 0.0921,
+      "step": 609
+    },
+    {
+      "epoch": 0.41881222107792654,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 9.673700795755008e-06,
+      "loss": 0.0963,
+      "step": 610
+    },
+    {
+      "epoch": 0.4194987984895297,
+      "grad_norm": 0.2265625,
+      "learning_rate": 9.672464815473315e-06,
+      "loss": 0.0973,
+      "step": 611
+    },
+    {
+      "epoch": 0.42018537590113286,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 9.67122658714729e-06,
+      "loss": 0.1012,
+      "step": 612
+    },
+    {
+      "epoch": 0.420871953312736,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 9.669986111444071e-06,
+      "loss": 0.0861,
+      "step": 613
+    },
+    {
+      "epoch": 0.4215585307243392,
+      "grad_norm": 0.26953125,
+      "learning_rate": 9.668743389032001e-06,
+      "loss": 0.0952,
+      "step": 614
+    },
+    {
+      "epoch": 0.42224510813594235,
+      "grad_norm": 0.23046875,
+      "learning_rate": 9.667498420580632e-06,
+      "loss": 0.0837,
+      "step": 615
+    },
+    {
+      "epoch": 0.4229316855475455,
+      "grad_norm": 0.18359375,
+      "learning_rate": 9.666251206760732e-06,
+      "loss": 0.0875,
+      "step": 616
+    },
+    {
+      "epoch": 0.4236182629591486,
+      "grad_norm": 0.212890625,
+      "learning_rate": 9.665001748244276e-06,
+      "loss": 0.097,
+      "step": 617
+    },
+    {
+      "epoch": 0.4243048403707518,
+      "grad_norm": 0.23046875,
+      "learning_rate": 9.663750045704448e-06,
+      "loss": 0.1012,
+      "step": 618
+    },
+    {
+      "epoch": 0.42499141778235494,
+      "grad_norm": 0.232421875,
+      "learning_rate": 9.66249609981564e-06,
+      "loss": 0.0981,
+      "step": 619
+    },
+    {
+      "epoch": 0.4256779951939581,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 9.661239911253457e-06,
+      "loss": 0.0875,
+      "step": 620
+    },
+    {
+      "epoch": 0.42636457260556127,
+      "grad_norm": 0.189453125,
+      "learning_rate": 9.659981480694708e-06,
+      "loss": 0.0878,
+      "step": 621
+    },
+    {
+      "epoch": 0.4270511500171644,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 9.65872080881741e-06,
+      "loss": 0.1059,
+      "step": 622
+    },
+    {
+      "epoch": 0.4277377274287676,
+      "grad_norm": 0.26953125,
+      "learning_rate": 9.65745789630079e-06,
+      "loss": 0.0881,
+      "step": 623
+    },
+    {
+      "epoch": 0.42842430484037075,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 9.656192743825283e-06,
+      "loss": 0.0803,
+      "step": 624
+    },
+    {
+      "epoch": 0.4291108822519739,
+      "grad_norm": 0.2578125,
+      "learning_rate": 9.654925352072526e-06,
+      "loss": 0.1024,
+      "step": 625
+    },
+    {
+      "epoch": 0.4297974596635771,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 9.653655721725367e-06,
+      "loss": 0.0908,
+      "step": 626
+    },
+    {
+      "epoch": 0.43048403707518024,
+      "grad_norm": 0.234375,
+      "learning_rate": 9.652383853467858e-06,
+      "loss": 0.098,
+      "step": 627
+    },
+    {
+      "epoch": 0.4311706144867834,
+      "grad_norm": 0.28125,
+      "learning_rate": 9.651109747985257e-06,
+      "loss": 0.0886,
+      "step": 628
+    },
+    {
+      "epoch": 0.43185719189838656,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 9.649833405964029e-06,
+      "loss": 0.0918,
+      "step": 629
+    },
+    {
+      "epoch": 0.4325437693099897,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 9.64855482809184e-06,
+      "loss": 0.0918,
+      "step": 630
+    },
+    {
+      "epoch": 0.4332303467215929,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 9.647274015057568e-06,
+      "loss": 0.0898,
+      "step": 631
+    },
+    {
+      "epoch": 0.433916924133196,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 9.645990967551287e-06,
+      "loss": 0.091,
+      "step": 632
+    },
+    {
+      "epoch": 0.43460350154479915,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 9.64470568626428e-06,
+      "loss": 0.0992,
+      "step": 633
+    },
+    {
+      "epoch": 0.4352900789564023,
+      "grad_norm": 0.50390625,
+      "learning_rate": 9.64341817188903e-06,
+      "loss": 0.0993,
+      "step": 634
+    },
+    {
+      "epoch": 0.4359766563680055,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 9.642128425119226e-06,
+      "loss": 0.0911,
+      "step": 635
+    },
+    {
+      "epoch": 0.43666323377960864,
+      "grad_norm": 0.244140625,
+      "learning_rate": 9.640836446649761e-06,
+      "loss": 0.0856,
+      "step": 636
+    },
+    {
+      "epoch": 0.4373498111912118,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 9.639542237176726e-06,
+      "loss": 0.0947,
+      "step": 637
+    },
+    {
+      "epoch": 0.43803638860281496,
+      "grad_norm": 0.25390625,
+      "learning_rate": 9.638245797397418e-06,
+      "loss": 0.0994,
+      "step": 638
+    },
+    {
+      "epoch": 0.4387229660144181,
+      "grad_norm": 0.244140625,
+      "learning_rate": 9.636947128010332e-06,
+      "loss": 0.0913,
+      "step": 639
+    },
+    {
+      "epoch": 0.4394095434260213,
+      "grad_norm": 0.251953125,
+      "learning_rate": 9.635646229715168e-06,
+      "loss": 0.0931,
+      "step": 640
+    },
+    {
+      "epoch": 0.44009612083762445,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 9.634343103212824e-06,
+      "loss": 0.0836,
+      "step": 641
+    },
+    {
+      "epoch": 0.4407826982492276,
+      "grad_norm": 0.263671875,
+      "learning_rate": 9.6330377492054e-06,
+      "loss": 0.0949,
+      "step": 642
+    },
+    {
+      "epoch": 0.4414692756608308,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 9.631730168396196e-06,
+      "loss": 0.0928,
+      "step": 643
+    },
+    {
+      "epoch": 0.44215585307243394,
+      "grad_norm": 0.26171875,
+      "learning_rate": 9.630420361489711e-06,
+      "loss": 0.0959,
+      "step": 644
+    },
+    {
+      "epoch": 0.4428424304840371,
+      "grad_norm": 0.244140625,
+      "learning_rate": 9.629108329191646e-06,
+      "loss": 0.0878,
+      "step": 645
+    },
+    {
+      "epoch": 0.44352900789564026,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 9.627794072208896e-06,
+      "loss": 0.0937,
+      "step": 646
+    },
+    {
+      "epoch": 0.44421558530724337,
+      "grad_norm": 0.244140625,
+      "learning_rate": 9.62647759124956e-06,
+      "loss": 0.0934,
+      "step": 647
+    },
+    {
+      "epoch": 0.44490216271884653,
+      "grad_norm": 0.212890625,
+      "learning_rate": 9.62515888702293e-06,
+      "loss": 0.0859,
+      "step": 648
+    },
+    {
+      "epoch": 0.4455887401304497,
+      "grad_norm": 0.24609375,
+      "learning_rate": 9.6238379602395e-06,
+      "loss": 0.0907,
+      "step": 649
+    },
+    {
+      "epoch": 0.44627531754205285,
+      "grad_norm": 0.275390625,
+      "learning_rate": 9.622514811610964e-06,
+      "loss": 0.0988,
+      "step": 650
+    },
+    {
+      "epoch": 0.446961894953656,
+      "grad_norm": 0.2421875,
+      "learning_rate": 9.621189441850206e-06,
+      "loss": 0.0888,
+      "step": 651
+    },
+    {
+      "epoch": 0.4476484723652592,
+      "grad_norm": 0.22265625,
+      "learning_rate": 9.619861851671306e-06,
+      "loss": 0.0824,
+      "step": 652
+    },
+    {
+      "epoch": 0.44833504977686234,
+      "grad_norm": 0.23828125,
+      "learning_rate": 9.618532041789548e-06,
+      "loss": 0.105,
+      "step": 653
+    },
+    {
+      "epoch": 0.4490216271884655,
+      "grad_norm": 0.21484375,
+      "learning_rate": 9.61720001292141e-06,
+      "loss": 0.0842,
+      "step": 654
+    },
+    {
+      "epoch": 0.44970820460006866,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 9.61586576578456e-06,
+      "loss": 0.1037,
+      "step": 655
+    },
+    {
+      "epoch": 0.4503947820116718,
+      "grad_norm": 0.216796875,
+      "learning_rate": 9.614529301097867e-06,
+      "loss": 0.0926,
+      "step": 656
+    },
+    {
+      "epoch": 0.451081359423275,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 9.613190619581393e-06,
+      "loss": 0.0871,
+      "step": 657
+    },
+    {
+      "epoch": 0.45176793683487815,
+      "grad_norm": 0.232421875,
+      "learning_rate": 9.611849721956392e-06,
+      "loss": 0.0875,
+      "step": 658
+    },
+    {
+      "epoch": 0.4524545142464813,
+      "grad_norm": 0.22265625,
+      "learning_rate": 9.610506608945315e-06,
+      "loss": 0.0892,
+      "step": 659
+    },
+    {
+      "epoch": 0.45314109165808447,
+      "grad_norm": 0.220703125,
+      "learning_rate": 9.609161281271808e-06,
+      "loss": 0.1082,
+      "step": 660
+    },
+    {
+      "epoch": 0.45382766906968763,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 9.607813739660705e-06,
+      "loss": 0.0917,
+      "step": 661
+    },
+    {
+      "epoch": 0.45451424648129074,
+      "grad_norm": 0.2109375,
+      "learning_rate": 9.606463984838034e-06,
+      "loss": 0.081,
+      "step": 662
+    },
+    {
+      "epoch": 0.4552008238928939,
+      "grad_norm": 0.203125,
+      "learning_rate": 9.605112017531022e-06,
+      "loss": 0.0854,
+      "step": 663
+    },
+    {
+      "epoch": 0.45588740130449706,
+      "grad_norm": 0.236328125,
+      "learning_rate": 9.603757838468079e-06,
+      "loss": 0.09,
+      "step": 664
+    },
+    {
+      "epoch": 0.4565739787161002,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 9.602401448378816e-06,
+      "loss": 0.0966,
+      "step": 665
+    },
+    {
+      "epoch": 0.4572605561277034,
+      "grad_norm": 0.263671875,
+      "learning_rate": 9.601042847994021e-06,
+      "loss": 0.0938,
+      "step": 666
+    },
+    {
+      "epoch": 0.45794713353930655,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 9.59968203804569e-06,
+      "loss": 0.0846,
+      "step": 667
+    },
+    {
+      "epoch": 0.4586337109509097,
+      "grad_norm": 0.26171875,
+      "learning_rate": 9.598319019267e-06,
+      "loss": 0.084,
+      "step": 668
+    },
+    {
+      "epoch": 0.4593202883625129,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 9.596953792392318e-06,
+      "loss": 0.1041,
+      "step": 669
+    },
+    {
+      "epoch": 0.46000686577411604,
+      "grad_norm": 0.232421875,
+      "learning_rate": 9.595586358157202e-06,
+      "loss": 0.099,
+      "step": 670
+    },
+    {
+      "epoch": 0.4606934431857192,
+      "grad_norm": 0.259765625,
+      "learning_rate": 9.5942167172984e-06,
+      "loss": 0.0913,
+      "step": 671
+    },
+    {
+      "epoch": 0.46138002059732236,
+      "grad_norm": 0.2265625,
+      "learning_rate": 9.592844870553849e-06,
+      "loss": 0.0859,
+      "step": 672
+    },
+    {
+      "epoch": 0.4620665980089255,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 9.591470818662672e-06,
+      "loss": 0.0881,
+      "step": 673
+    },
+    {
+      "epoch": 0.4627531754205287,
+      "grad_norm": 0.224609375,
+      "learning_rate": 9.590094562365184e-06,
+      "loss": 0.0875,
+      "step": 674
+    },
+    {
+      "epoch": 0.46343975283213185,
+      "grad_norm": 0.22265625,
+      "learning_rate": 9.588716102402882e-06,
+      "loss": 0.0943,
+      "step": 675
+    },
+    {
+      "epoch": 0.464126330243735,
+      "grad_norm": 0.22265625,
+      "learning_rate": 9.58733543951846e-06,
+      "loss": 0.0885,
+      "step": 676
+    },
+    {
+      "epoch": 0.4648129076553381,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 9.58595257445579e-06,
+      "loss": 0.0931,
+      "step": 677
+    },
+    {
+      "epoch": 0.4654994850669413,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 9.584567507959929e-06,
+      "loss": 0.0895,
+      "step": 678
+    },
+    {
+      "epoch": 0.46618606247854444,
+      "grad_norm": 0.2265625,
+      "learning_rate": 9.583180240777128e-06,
+      "loss": 0.0886,
+      "step": 679
+    },
+    {
+      "epoch": 0.4668726398901476,
+      "grad_norm": 0.236328125,
+      "learning_rate": 9.581790773654821e-06,
+      "loss": 0.0919,
+      "step": 680
+    },
+    {
+      "epoch": 0.46755921730175076,
+      "grad_norm": 0.203125,
+      "learning_rate": 9.580399107341627e-06,
+      "loss": 0.0925,
+      "step": 681
+    },
+    {
+      "epoch": 0.4682457947133539,
+      "grad_norm": 0.1953125,
+      "learning_rate": 9.579005242587344e-06,
+      "loss": 0.0885,
+      "step": 682
+    },
+    {
+      "epoch": 0.4689323721249571,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 9.577609180142967e-06,
+      "loss": 0.0856,
+      "step": 683
+    },
+    {
+      "epoch": 0.46961894953656025,
+      "grad_norm": 0.224609375,
+      "learning_rate": 9.576210920760662e-06,
+      "loss": 0.0814,
+      "step": 684
+    },
+    {
+      "epoch": 0.4703055269481634,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 9.574810465193787e-06,
+      "loss": 0.0827,
+      "step": 685
+    },
+    {
+      "epoch": 0.4709921043597666,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 9.57340781419688e-06,
+      "loss": 0.0833,
+      "step": 686
+    },
+    {
+      "epoch": 0.47167868177136973,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 9.572002968525662e-06,
+      "loss": 0.0804,
+      "step": 687
+    },
+    {
+      "epoch": 0.4723652591829729,
+      "grad_norm": 0.220703125,
+      "learning_rate": 9.57059592893704e-06,
+      "loss": 0.0803,
+      "step": 688
+    },
+    {
+      "epoch": 0.47305183659457606,
+      "grad_norm": 0.2421875,
+      "learning_rate": 9.569186696189095e-06,
+      "loss": 0.1013,
+      "step": 689
+    },
+    {
+      "epoch": 0.4737384140061792,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 9.567775271041099e-06,
+      "loss": 0.0967,
+      "step": 690
+    },
+    {
+      "epoch": 0.4744249914177824,
+      "grad_norm": 0.2421875,
+      "learning_rate": 9.566361654253499e-06,
+      "loss": 0.0883,
+      "step": 691
+    },
+    {
+      "epoch": 0.4751115688293855,
+      "grad_norm": 0.240234375,
+      "learning_rate": 9.564945846587925e-06,
+      "loss": 0.0926,
+      "step": 692
+    },
+    {
+      "epoch": 0.47579814624098865,
+      "grad_norm": 0.208984375,
+      "learning_rate": 9.563527848807186e-06,
+      "loss": 0.0891,
+      "step": 693
+    },
+    {
+      "epoch": 0.4764847236525918,
+      "grad_norm": 0.205078125,
+      "learning_rate": 9.562107661675276e-06,
+      "loss": 0.0887,
+      "step": 694
+    },
+    {
+      "epoch": 0.477171301064195,
+      "grad_norm": 0.201171875,
+      "learning_rate": 9.560685285957361e-06,
+      "loss": 0.0904,
+      "step": 695
+    },
+    {
+      "epoch": 0.47785787847579814,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 9.55926072241979e-06,
+      "loss": 0.0873,
+      "step": 696
+    },
+    {
+      "epoch": 0.4785444558874013,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 9.557833971830093e-06,
+      "loss": 0.0858,
+      "step": 697
+    },
+    {
+      "epoch": 0.47923103329900446,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 9.556405034956974e-06,
+      "loss": 0.0855,
+      "step": 698
+    },
+    {
+      "epoch": 0.4799176107106076,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 9.554973912570316e-06,
+      "loss": 0.0942,
+      "step": 699
+    },
+    {
+      "epoch": 0.4806041881222108,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 9.553540605441182e-06,
+      "loss": 0.0919,
+      "step": 700
+    },
+    {
+      "epoch": 0.48129076553381395,
+      "grad_norm": 0.20703125,
+      "learning_rate": 9.552105114341811e-06,
+      "loss": 0.0748,
+      "step": 701
+    },
+    {
+      "epoch": 0.4819773429454171,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 9.550667440045618e-06,
+      "loss": 0.0837,
+      "step": 702
+    },
+    {
+      "epoch": 0.48266392035702027,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 9.549227583327193e-06,
+      "loss": 0.0874,
+      "step": 703
+    },
+    {
+      "epoch": 0.48335049776862343,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 9.547785544962303e-06,
+      "loss": 0.0939,
+      "step": 704
+    },
+    {
+      "epoch": 0.4840370751802266,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 9.546341325727893e-06,
+      "loss": 0.087,
+      "step": 705
+    },
+    {
+      "epoch": 0.4847236525918297,
+      "grad_norm": 0.208984375,
+      "learning_rate": 9.54489492640208e-06,
+      "loss": 0.0801,
+      "step": 706
+    },
+    {
+      "epoch": 0.48541023000343286,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 9.543446347764159e-06,
+      "loss": 0.0827,
+      "step": 707
+    },
+    {
+      "epoch": 0.486096807415036,
+      "grad_norm": 0.2109375,
+      "learning_rate": 9.541995590594589e-06,
+      "loss": 0.087,
+      "step": 708
+    },
+    {
+      "epoch": 0.4867833848266392,
+      "grad_norm": 0.294921875,
+      "learning_rate": 9.540542655675014e-06,
+      "loss": 0.0983,
+      "step": 709
+    },
+    {
+      "epoch": 0.48746996223824235,
+      "grad_norm": 0.234375,
+      "learning_rate": 9.539087543788251e-06,
+      "loss": 0.0954,
+      "step": 710
+    },
+    {
+      "epoch": 0.4881565396498455,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 9.537630255718285e-06,
+      "loss": 0.0854,
+      "step": 711
+    },
+    {
+      "epoch": 0.4888431170614487,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 9.53617079225027e-06,
+      "loss": 0.0974,
+      "step": 712
+    },
+    {
+      "epoch": 0.48952969447305184,
+      "grad_norm": 0.20703125,
+      "learning_rate": 9.534709154170542e-06,
+      "loss": 0.0904,
+      "step": 713
+    },
+    {
+      "epoch": 0.490216271884655,
+      "grad_norm": 0.19921875,
+      "learning_rate": 9.533245342266604e-06,
+      "loss": 0.089,
+      "step": 714
+    },
+    {
+      "epoch": 0.49090284929625816,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 9.531779357327125e-06,
+      "loss": 0.0885,
+      "step": 715
+    },
+    {
+      "epoch": 0.4915894267078613,
+      "grad_norm": 0.2578125,
+      "learning_rate": 9.530311200141957e-06,
+      "loss": 0.0882,
+      "step": 716
+    },
+    {
+      "epoch": 0.4922760041194645,
+      "grad_norm": 0.240234375,
+      "learning_rate": 9.528840871502108e-06,
+      "loss": 0.0928,
+      "step": 717
+    },
+    {
+      "epoch": 0.49296258153106765,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 9.527368372199767e-06,
+      "loss": 0.0887,
+      "step": 718
+    },
+    {
+      "epoch": 0.4936491589426708,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 9.525893703028289e-06,
+      "loss": 0.0862,
+      "step": 719
+    },
+    {
+      "epoch": 0.49433573635427397,
+      "grad_norm": 0.236328125,
+      "learning_rate": 9.524416864782196e-06,
+      "loss": 0.0853,
+      "step": 720
+    },
+    {
+      "epoch": 0.4950223137658771,
+      "grad_norm": 0.203125,
+      "learning_rate": 9.522937858257177e-06,
+      "loss": 0.079,
+      "step": 721
+    },
+    {
+      "epoch": 0.49570889117748024,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 9.5214566842501e-06,
+      "loss": 0.0954,
+      "step": 722
+    },
+    {
+      "epoch": 0.4963954685890834,
+      "grad_norm": 0.21875,
+      "learning_rate": 9.519973343558984e-06,
+      "loss": 0.083,
+      "step": 723
+    },
+    {
+      "epoch": 0.49708204600068656,
+      "grad_norm": 0.24609375,
+      "learning_rate": 9.518487836983035e-06,
+      "loss": 0.0983,
+      "step": 724
+    },
+    {
+      "epoch": 0.4977686234122897,
+      "grad_norm": 0.26171875,
+      "learning_rate": 9.517000165322607e-06,
+      "loss": 0.0814,
+      "step": 725
+    },
+    {
+      "epoch": 0.4984552008238929,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 9.515510329379234e-06,
+      "loss": 0.0906,
+      "step": 726
+    },
+    {
+      "epoch": 0.49914177823549605,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 9.514018329955608e-06,
+      "loss": 0.0901,
+      "step": 727
+    },
+    {
+      "epoch": 0.4998283556470992,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 9.51252416785559e-06,
+      "loss": 0.0884,
+      "step": 728
+    },
+    {
+      "epoch": 0.5005149330587023,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 9.51102784388421e-06,
+      "loss": 0.098,
+      "step": 729
+    },
+    {
+      "epoch": 0.5012015104703055,
+      "grad_norm": 0.23046875,
+      "learning_rate": 9.509529358847655e-06,
+      "loss": 0.0966,
+      "step": 730
+    },
+    {
+      "epoch": 0.5018880878819086,
+      "grad_norm": 0.22265625,
+      "learning_rate": 9.508028713553282e-06,
+      "loss": 0.0978,
+      "step": 731
+    },
+    {
+      "epoch": 0.5025746652935118,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 9.50652590880961e-06,
+      "loss": 0.0869,
+      "step": 732
+    },
+    {
+      "epoch": 0.503261242705115,
+      "grad_norm": 0.1875,
+      "learning_rate": 9.505020945426318e-06,
+      "loss": 0.0828,
+      "step": 733
+    },
+    {
+      "epoch": 0.5039478201167181,
+      "grad_norm": 0.19140625,
+      "learning_rate": 9.503513824214254e-06,
+      "loss": 0.0817,
+      "step": 734
+    },
+    {
+      "epoch": 0.5046343975283213,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 9.502004545985428e-06,
+      "loss": 0.0916,
+      "step": 735
+    },
+    {
+      "epoch": 0.5053209749399244,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 9.500493111553007e-06,
+      "loss": 0.0991,
+      "step": 736
+    },
+    {
+      "epoch": 0.5060075523515276,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 9.498979521731327e-06,
+      "loss": 0.0897,
+      "step": 737
+    },
+    {
+      "epoch": 0.5066941297631308,
+      "grad_norm": 0.259765625,
+      "learning_rate": 9.497463777335875e-06,
+      "loss": 0.0942,
+      "step": 738
+    },
+    {
+      "epoch": 0.5073807071747339,
+      "grad_norm": 0.302734375,
+      "learning_rate": 9.495945879183312e-06,
+      "loss": 0.09,
+      "step": 739
+    },
+    {
+      "epoch": 0.5080672845863371,
+      "grad_norm": 0.1953125,
+      "learning_rate": 9.49442582809145e-06,
+      "loss": 0.0879,
+      "step": 740
+    },
+    {
+      "epoch": 0.5087538619979403,
+      "grad_norm": 0.32421875,
+      "learning_rate": 9.49290362487926e-06,
+      "loss": 0.0956,
+      "step": 741
+    },
+    {
+      "epoch": 0.5094404394095434,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 9.49137927036688e-06,
+      "loss": 0.0989,
+      "step": 742
+    },
+    {
+      "epoch": 0.5101270168211466,
+      "grad_norm": 0.26953125,
+      "learning_rate": 9.489852765375602e-06,
+      "loss": 0.0936,
+      "step": 743
+    },
+    {
+      "epoch": 0.5108135942327497,
+      "grad_norm": 0.216796875,
+      "learning_rate": 9.488324110727878e-06,
+      "loss": 0.0779,
+      "step": 744
+    },
+    {
+      "epoch": 0.5115001716443529,
+      "grad_norm": 0.294921875,
+      "learning_rate": 9.486793307247318e-06,
+      "loss": 0.091,
+      "step": 745
+    },
+    {
+      "epoch": 0.5121867490559561,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 9.48526035575869e-06,
+      "loss": 0.0828,
+      "step": 746
+    },
+    {
+      "epoch": 0.5128733264675592,
+      "grad_norm": 0.2421875,
+      "learning_rate": 9.483725257087919e-06,
+      "loss": 0.0881,
+      "step": 747
+    },
+    {
+      "epoch": 0.5135599038791624,
+      "grad_norm": 0.208984375,
+      "learning_rate": 9.482188012062084e-06,
+      "loss": 0.0798,
+      "step": 748
+    },
+    {
+      "epoch": 0.5142464812907656,
+      "grad_norm": 0.259765625,
+      "learning_rate": 9.480648621509426e-06,
+      "loss": 0.0896,
+      "step": 749
+    },
+    {
+      "epoch": 0.5149330587023687,
+      "grad_norm": 0.208984375,
+      "learning_rate": 9.47910708625934e-06,
+      "loss": 0.0931,
+      "step": 750
+    },
+    {
+      "epoch": 0.5156196361139719,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 9.477563407142372e-06,
+      "loss": 0.0936,
+      "step": 751
+    },
+    {
+      "epoch": 0.516306213525575,
+      "grad_norm": 0.19921875,
+      "learning_rate": 9.476017584990229e-06,
+      "loss": 0.0826,
+      "step": 752
+    },
+    {
+      "epoch": 0.5169927909371782,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 9.474469620635773e-06,
+      "loss": 0.0883,
+      "step": 753
+    },
+    {
+      "epoch": 0.5176793683487814,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 9.472919514913013e-06,
+      "loss": 0.0854,
+      "step": 754
+    },
+    {
+      "epoch": 0.5183659457603845,
+      "grad_norm": 0.203125,
+      "learning_rate": 9.471367268657121e-06,
+      "loss": 0.0927,
+      "step": 755
+    },
+    {
+      "epoch": 0.5190525231719877,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 9.469812882704413e-06,
+      "loss": 0.0849,
+      "step": 756
+    },
+    {
+      "epoch": 0.5197391005835909,
+      "grad_norm": 0.2421875,
+      "learning_rate": 9.468256357892367e-06,
+      "loss": 0.0984,
+      "step": 757
+    },
+    {
+      "epoch": 0.5204256779951939,
+      "grad_norm": 0.220703125,
+      "learning_rate": 9.466697695059604e-06,
+      "loss": 0.0906,
+      "step": 758
+    },
+    {
+      "epoch": 0.5211122554067971,
+      "grad_norm": 0.181640625,
+      "learning_rate": 9.465136895045907e-06,
+      "loss": 0.0925,
+      "step": 759
+    },
+    {
+      "epoch": 0.5217988328184002,
+      "grad_norm": 0.333984375,
+      "learning_rate": 9.4635739586922e-06,
+      "loss": 0.0933,
+      "step": 760
+    },
+    {
+      "epoch": 0.5224854102300034,
+      "grad_norm": 0.240234375,
+      "learning_rate": 9.462008886840567e-06,
+      "loss": 0.0806,
+      "step": 761
+    },
+    {
+      "epoch": 0.5231719876416066,
+      "grad_norm": 0.205078125,
+      "learning_rate": 9.460441680334236e-06,
+      "loss": 0.0898,
+      "step": 762
+    },
+    {
+      "epoch": 0.5238585650532097,
+      "grad_norm": 0.23046875,
+      "learning_rate": 9.458872340017592e-06,
+      "loss": 0.0845,
+      "step": 763
+    },
+    {
+      "epoch": 0.5245451424648129,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 9.45730086673616e-06,
+      "loss": 0.1001,
+      "step": 764
+    },
+    {
+      "epoch": 0.525231719876416,
+      "grad_norm": 0.232421875,
+      "learning_rate": 9.455727261336626e-06,
+      "loss": 0.0888,
+      "step": 765
+    },
+    {
+      "epoch": 0.5259182972880192,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 9.454151524666815e-06,
+      "loss": 0.0954,
+      "step": 766
+    },
+    {
+      "epoch": 0.5266048746996224,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 9.452573657575705e-06,
+      "loss": 0.09,
+      "step": 767
+    },
+    {
+      "epoch": 0.5272914521112255,
+      "grad_norm": 0.216796875,
+      "learning_rate": 9.450993660913418e-06,
+      "loss": 0.0806,
+      "step": 768
+    },
+    {
+      "epoch": 0.5279780295228287,
+      "grad_norm": 0.224609375,
+      "learning_rate": 9.449411535531227e-06,
+      "loss": 0.0865,
+      "step": 769
+    },
+    {
+      "epoch": 0.5286646069344318,
+      "grad_norm": 0.244140625,
+      "learning_rate": 9.447827282281551e-06,
+      "loss": 0.0971,
+      "step": 770
+    },
+    {
+      "epoch": 0.529351184346035,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 9.44624090201796e-06,
+      "loss": 0.0911,
+      "step": 771
+    },
+    {
+      "epoch": 0.5300377617576382,
+      "grad_norm": 0.302734375,
+      "learning_rate": 9.444652395595159e-06,
+      "loss": 0.1002,
+      "step": 772
+    },
+    {
+      "epoch": 0.5307243391692413,
+      "grad_norm": 0.20703125,
+      "learning_rate": 9.443061763869007e-06,
+      "loss": 0.0907,
+      "step": 773
+    },
+    {
+      "epoch": 0.5314109165808445,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 9.44146900769651e-06,
+      "loss": 0.0953,
+      "step": 774
+    },
+    {
+      "epoch": 0.5320974939924477,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 9.439874127935807e-06,
+      "loss": 0.0919,
+      "step": 775
+    },
+    {
+      "epoch": 0.5327840714040508,
+      "grad_norm": 0.208984375,
+      "learning_rate": 9.438277125446194e-06,
+      "loss": 0.0909,
+      "step": 776
+    },
+    {
+      "epoch": 0.533470648815654,
+      "grad_norm": 0.20703125,
+      "learning_rate": 9.436678001088106e-06,
+      "loss": 0.0924,
+      "step": 777
+    },
+    {
+      "epoch": 0.5341572262272571,
+      "grad_norm": 0.21875,
+      "learning_rate": 9.435076755723119e-06,
+      "loss": 0.0819,
+      "step": 778
+    },
+    {
+      "epoch": 0.5348438036388603,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 9.43347339021395e-06,
+      "loss": 0.0939,
+      "step": 779
+    },
+    {
+      "epoch": 0.5355303810504635,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 9.431867905424466e-06,
+      "loss": 0.0871,
+      "step": 780
+    },
+    {
+      "epoch": 0.5362169584620666,
+      "grad_norm": 0.232421875,
+      "learning_rate": 9.430260302219672e-06,
+      "loss": 0.093,
+      "step": 781
+    },
+    {
+      "epoch": 0.5369035358736698,
+      "grad_norm": 0.2109375,
+      "learning_rate": 9.428650581465713e-06,
+      "loss": 0.0909,
+      "step": 782
+    },
+    {
+      "epoch": 0.537590113285273,
+      "grad_norm": 0.216796875,
+      "learning_rate": 9.427038744029872e-06,
+      "loss": 0.0985,
+      "step": 783
+    },
+    {
+      "epoch": 0.5382766906968761,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 9.425424790780581e-06,
+      "loss": 0.0802,
+      "step": 784
+    },
+    {
+      "epoch": 0.5389632681084793,
+      "grad_norm": 0.181640625,
+      "learning_rate": 9.423808722587407e-06,
+      "loss": 0.0829,
+      "step": 785
+    },
+    {
+      "epoch": 0.5396498455200824,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 9.422190540321055e-06,
+      "loss": 0.0898,
+      "step": 786
+    },
+    {
+      "epoch": 0.5403364229316856,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 9.42057024485337e-06,
+      "loss": 0.0875,
+      "step": 787
+    },
+    {
+      "epoch": 0.5410230003432887,
+      "grad_norm": 0.220703125,
+      "learning_rate": 9.418947837057338e-06,
+      "loss": 0.0919,
+      "step": 788
+    },
+    {
+      "epoch": 0.5417095777548918,
+      "grad_norm": 0.21484375,
+      "learning_rate": 9.41732331780708e-06,
+      "loss": 0.0909,
+      "step": 789
+    },
+    {
+      "epoch": 0.542396155166495,
+      "grad_norm": 0.234375,
+      "learning_rate": 9.415696687977857e-06,
+      "loss": 0.0841,
+      "step": 790
+    },
+    {
+      "epoch": 0.5430827325780981,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 9.414067948446064e-06,
+      "loss": 0.0854,
+      "step": 791
+    },
+    {
+      "epoch": 0.5437693099897013,
+      "grad_norm": 0.2421875,
+      "learning_rate": 9.412437100089236e-06,
+      "loss": 0.0892,
+      "step": 792
+    },
+    {
+      "epoch": 0.5444558874013045,
+      "grad_norm": 0.2578125,
+      "learning_rate": 9.410804143786046e-06,
+      "loss": 0.0893,
+      "step": 793
+    },
+    {
+      "epoch": 0.5451424648129076,
+      "grad_norm": 0.2265625,
+      "learning_rate": 9.409169080416296e-06,
+      "loss": 0.0839,
+      "step": 794
+    },
+    {
+      "epoch": 0.5458290422245108,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 9.407531910860928e-06,
+      "loss": 0.0854,
+      "step": 795
+    },
+    {
+      "epoch": 0.546515619636114,
+      "grad_norm": 0.265625,
+      "learning_rate": 9.405892636002016e-06,
+      "loss": 0.0882,
+      "step": 796
+    },
+    {
+      "epoch": 0.5472021970477171,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 9.404251256722772e-06,
+      "loss": 0.0893,
+      "step": 797
+    },
+    {
+      "epoch": 0.5478887744593203,
+      "grad_norm": 0.193359375,
+      "learning_rate": 9.402607773907539e-06,
+      "loss": 0.0863,
+      "step": 798
+    },
+    {
+      "epoch": 0.5485753518709234,
+      "grad_norm": 0.21484375,
+      "learning_rate": 9.400962188441795e-06,
+      "loss": 0.0925,
+      "step": 799
+    },
+    {
+      "epoch": 0.5492619292825266,
+      "grad_norm": 0.240234375,
+      "learning_rate": 9.39931450121215e-06,
+      "loss": 0.0791,
+      "step": 800
+    },
+    {
+      "epoch": 0.5499485066941298,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 9.397664713106345e-06,
+      "loss": 0.0975,
+      "step": 801
+    },
+    {
+      "epoch": 0.5506350841057329,
+      "grad_norm": 0.2109375,
+      "learning_rate": 9.396012825013256e-06,
+      "loss": 0.0974,
+      "step": 802
+    },
+    {
+      "epoch": 0.5513216615173361,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 9.394358837822886e-06,
+      "loss": 0.0994,
+      "step": 803
+    },
+    {
+      "epoch": 0.5520082389289392,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 9.392702752426377e-06,
+      "loss": 0.0914,
+      "step": 804
+    },
+    {
+      "epoch": 0.5526948163405424,
+      "grad_norm": 0.208984375,
+      "learning_rate": 9.391044569715987e-06,
+      "loss": 0.0945,
+      "step": 805
+    },
+    {
+      "epoch": 0.5533813937521456,
+      "grad_norm": 0.236328125,
+      "learning_rate": 9.389384290585123e-06,
+      "loss": 0.0945,
+      "step": 806
+    },
+    {
+      "epoch": 0.5540679711637487,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 9.387721915928309e-06,
+      "loss": 0.0841,
+      "step": 807
+    },
+    {
+      "epoch": 0.5547545485753519,
+      "grad_norm": 0.2421875,
+      "learning_rate": 9.386057446641195e-06,
+      "loss": 0.0953,
+      "step": 808
+    },
+    {
+      "epoch": 0.555441125986955,
+      "grad_norm": 0.220703125,
+      "learning_rate": 9.384390883620573e-06,
+      "loss": 0.0887,
+      "step": 809
+    },
+    {
+      "epoch": 0.5561277033985582,
+      "grad_norm": 0.2265625,
+      "learning_rate": 9.38272222776435e-06,
+      "loss": 0.0971,
+      "step": 810
+    },
+    {
+      "epoch": 0.5568142808101614,
+      "grad_norm": 0.259765625,
+      "learning_rate": 9.381051479971569e-06,
+      "loss": 0.0854,
+      "step": 811
+    },
+    {
+      "epoch": 0.5575008582217645,
+      "grad_norm": 0.22265625,
+      "learning_rate": 9.379378641142394e-06,
+      "loss": 0.0904,
+      "step": 812
+    },
+    {
+      "epoch": 0.5581874356333677,
+      "grad_norm": 0.208984375,
+      "learning_rate": 9.377703712178122e-06,
+      "loss": 0.0966,
+      "step": 813
+    },
+    {
+      "epoch": 0.5588740130449709,
+      "grad_norm": 0.234375,
+      "learning_rate": 9.37602669398117e-06,
+      "loss": 0.0889,
+      "step": 814
+    },
+    {
+      "epoch": 0.559560590456574,
+      "grad_norm": 0.259765625,
+      "learning_rate": 9.374347587455087e-06,
+      "loss": 0.089,
+      "step": 815
+    },
+    {
+      "epoch": 0.5602471678681772,
+      "grad_norm": 0.205078125,
+      "learning_rate": 9.372666393504537e-06,
+      "loss": 0.0852,
+      "step": 816
+    },
+    {
+      "epoch": 0.5609337452797803,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 9.370983113035323e-06,
+      "loss": 0.1005,
+      "step": 817
+    },
+    {
+      "epoch": 0.5616203226913834,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 9.369297746954358e-06,
+      "loss": 0.0817,
+      "step": 818
+    },
+    {
+      "epoch": 0.5623069001029866,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 9.367610296169689e-06,
+      "loss": 0.0821,
+      "step": 819
+    },
+    {
+      "epoch": 0.5629934775145897,
+      "grad_norm": 0.193359375,
+      "learning_rate": 9.365920761590478e-06,
+      "loss": 0.1006,
+      "step": 820
+    },
+    {
+      "epoch": 0.5636800549261929,
+      "grad_norm": 0.23046875,
+      "learning_rate": 9.36422914412702e-06,
+      "loss": 0.0936,
+      "step": 821
+    },
+    {
+      "epoch": 0.564366632337796,
+      "grad_norm": 0.333984375,
+      "learning_rate": 9.362535444690721e-06,
+      "loss": 0.0789,
+      "step": 822
+    },
+    {
+      "epoch": 0.5650532097493992,
+      "grad_norm": 0.2109375,
+      "learning_rate": 9.360839664194116e-06,
+      "loss": 0.0921,
+      "step": 823
+    },
+    {
+      "epoch": 0.5657397871610024,
+      "grad_norm": 0.21484375,
+      "learning_rate": 9.35914180355086e-06,
+      "loss": 0.0935,
+      "step": 824
+    },
+    {
+      "epoch": 0.5664263645726055,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 9.357441863675727e-06,
+      "loss": 0.0996,
+      "step": 825
+    },
+    {
+      "epoch": 0.5671129419842087,
+      "grad_norm": 0.25390625,
+      "learning_rate": 9.355739845484611e-06,
+      "loss": 0.0833,
+      "step": 826
+    },
+    {
+      "epoch": 0.5677995193958119,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 9.354035749894527e-06,
+      "loss": 0.0955,
+      "step": 827
+    },
+    {
+      "epoch": 0.568486096807415,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 9.352329577823613e-06,
+      "loss": 0.0887,
+      "step": 828
+    },
+    {
+      "epoch": 0.5691726742190182,
+      "grad_norm": 0.208984375,
+      "learning_rate": 9.350621330191116e-06,
+      "loss": 0.0924,
+      "step": 829
+    },
+    {
+      "epoch": 0.5698592516306213,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 9.348911007917411e-06,
+      "loss": 0.0899,
+      "step": 830
+    },
+    {
+      "epoch": 0.5705458290422245,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 9.347198611923986e-06,
+      "loss": 0.0949,
+      "step": 831
+    },
+    {
+      "epoch": 0.5712324064538277,
+      "grad_norm": 0.2265625,
+      "learning_rate": 9.345484143133447e-06,
+      "loss": 0.0807,
+      "step": 832
+    },
+    {
+      "epoch": 0.5719189838654308,
+      "grad_norm": 0.23046875,
+      "learning_rate": 9.343767602469519e-06,
+      "loss": 0.0875,
+      "step": 833
+    },
+    {
+      "epoch": 0.572605561277034,
+      "grad_norm": 0.234375,
+      "learning_rate": 9.342048990857037e-06,
+      "loss": 0.0927,
+      "step": 834
+    },
+    {
+      "epoch": 0.5732921386886372,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 9.340328309221962e-06,
+      "loss": 0.0823,
+      "step": 835
+    },
+    {
+      "epoch": 0.5739787161002403,
+      "grad_norm": 0.19140625,
+      "learning_rate": 9.33860555849136e-06,
+      "loss": 0.0861,
+      "step": 836
+    },
+    {
+      "epoch": 0.5746652935118435,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 9.336880739593415e-06,
+      "loss": 0.098,
+      "step": 837
+    },
+    {
+      "epoch": 0.5753518709234466,
+      "grad_norm": 0.259765625,
+      "learning_rate": 9.335153853457431e-06,
+      "loss": 0.0935,
+      "step": 838
+    },
+    {
+      "epoch": 0.5760384483350498,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 9.333424901013818e-06,
+      "loss": 0.0903,
+      "step": 839
+    },
+    {
+      "epoch": 0.576725025746653,
+      "grad_norm": 0.201171875,
+      "learning_rate": 9.331693883194105e-06,
+      "loss": 0.0881,
+      "step": 840
+    },
+    {
+      "epoch": 0.5774116031582561,
+      "grad_norm": 0.19140625,
+      "learning_rate": 9.329960800930929e-06,
+      "loss": 0.088,
+      "step": 841
+    },
+    {
+      "epoch": 0.5780981805698593,
+      "grad_norm": 0.2109375,
+      "learning_rate": 9.328225655158045e-06,
+      "loss": 0.0897,
+      "step": 842
+    },
+    {
+      "epoch": 0.5787847579814624,
+      "grad_norm": 0.216796875,
+      "learning_rate": 9.32648844681031e-06,
+      "loss": 0.0833,
+      "step": 843
+    },
+    {
+      "epoch": 0.5794713353930656,
+      "grad_norm": 0.228515625,
+      "learning_rate": 9.324749176823704e-06,
+      "loss": 0.0827,
+      "step": 844
+    },
+    {
+      "epoch": 0.5801579128046688,
+      "grad_norm": 0.25390625,
+      "learning_rate": 9.323007846135312e-06,
+      "loss": 0.1096,
+      "step": 845
+    },
+    {
+      "epoch": 0.5808444902162719,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 9.321264455683327e-06,
+      "loss": 0.0973,
+      "step": 846
+    },
+    {
+      "epoch": 0.5815310676278751,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 9.31951900640706e-06,
+      "loss": 0.0812,
+      "step": 847
+    },
+    {
+      "epoch": 0.5822176450394781,
+      "grad_norm": 0.2265625,
+      "learning_rate": 9.317771499246918e-06,
+      "loss": 0.0804,
+      "step": 848
+    },
+    {
+      "epoch": 0.5829042224510813,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 9.316021935144431e-06,
+      "loss": 0.0876,
+      "step": 849
+    },
+    {
+      "epoch": 0.5835907998626845,
+      "grad_norm": 0.212890625,
+      "learning_rate": 9.314270315042225e-06,
+      "loss": 0.1036,
+      "step": 850
+    },
+    {
+      "epoch": 0.5842773772742876,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 9.312516639884047e-06,
+      "loss": 0.0865,
+      "step": 851
+    },
+    {
+      "epoch": 0.5849639546858908,
+      "grad_norm": 0.28515625,
+      "learning_rate": 9.310760910614736e-06,
+      "loss": 0.0805,
+      "step": 852
+    },
+    {
+      "epoch": 0.585650532097494,
+      "grad_norm": 0.220703125,
+      "learning_rate": 9.309003128180249e-06,
+      "loss": 0.0887,
+      "step": 853
+    },
+    {
+      "epoch": 0.5863371095090971,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 9.307243293527645e-06,
+      "loss": 0.0957,
+      "step": 854
+    },
+    {
+      "epoch": 0.5870236869207003,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 9.30548140760509e-06,
+      "loss": 0.0871,
+      "step": 855
+    },
+    {
+      "epoch": 0.5877102643323034,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 9.303717471361855e-06,
+      "loss": 0.0943,
+      "step": 856
+    },
+    {
+      "epoch": 0.5883968417439066,
+      "grad_norm": 0.3515625,
+      "learning_rate": 9.301951485748314e-06,
+      "loss": 0.0915,
+      "step": 857
+    },
+    {
+      "epoch": 0.5890834191555098,
+      "grad_norm": 0.19921875,
+      "learning_rate": 9.300183451715945e-06,
+      "loss": 0.0843,
+      "step": 858
+    },
+    {
+      "epoch": 0.5897699965671129,
+      "grad_norm": 0.21875,
+      "learning_rate": 9.298413370217333e-06,
+      "loss": 0.0784,
+      "step": 859
+    },
+    {
+      "epoch": 0.5904565739787161,
+      "grad_norm": 0.216796875,
+      "learning_rate": 9.296641242206165e-06,
+      "loss": 0.0832,
+      "step": 860
+    },
+    {
+      "epoch": 0.5911431513903193,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 9.294867068637227e-06,
+      "loss": 0.0913,
+      "step": 861
+    },
+    {
+      "epoch": 0.5918297288019224,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 9.29309085046641e-06,
+      "loss": 0.0988,
+      "step": 862
+    },
+    {
+      "epoch": 0.5925163062135256,
+      "grad_norm": 0.216796875,
+      "learning_rate": 9.29131258865071e-06,
+      "loss": 0.0884,
+      "step": 863
+    },
+    {
+      "epoch": 0.5932028836251287,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 9.289532284148218e-06,
+      "loss": 0.0831,
+      "step": 864
+    },
+    {
+      "epoch": 0.5938894610367319,
+      "grad_norm": 0.255859375,
+      "learning_rate": 9.287749937918125e-06,
+      "loss": 0.0959,
+      "step": 865
+    },
+    {
+      "epoch": 0.5945760384483351,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 9.285965550920732e-06,
+      "loss": 0.0992,
+      "step": 866
+    },
+    {
+      "epoch": 0.5952626158599382,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 9.284179124117426e-06,
+      "loss": 0.0922,
+      "step": 867
+    },
+    {
+      "epoch": 0.5959491932715414,
+      "grad_norm": 0.251953125,
+      "learning_rate": 9.282390658470703e-06,
+      "loss": 0.101,
+      "step": 868
+    },
+    {
+      "epoch": 0.5966357706831446,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 9.280600154944153e-06,
+      "loss": 0.0898,
+      "step": 869
+    },
+    {
+      "epoch": 0.5973223480947477,
+      "grad_norm": 0.240234375,
+      "learning_rate": 9.278807614502467e-06,
+      "loss": 0.0886,
+      "step": 870
+    },
+    {
+      "epoch": 0.5980089255063509,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 9.27701303811143e-06,
+      "loss": 0.0858,
+      "step": 871
+    },
+    {
+      "epoch": 0.598695502917954,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 9.275216426737924e-06,
+      "loss": 0.0838,
+      "step": 872
+    },
+    {
+      "epoch": 0.5993820803295572,
+      "grad_norm": 0.2109375,
+      "learning_rate": 9.273417781349933e-06,
+      "loss": 0.0836,
+      "step": 873
+    },
+    {
+      "epoch": 0.6000686577411604,
+      "grad_norm": 0.220703125,
+      "learning_rate": 9.271617102916528e-06,
+      "loss": 0.0918,
+      "step": 874
+    },
+    {
+      "epoch": 0.6007552351527635,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 9.269814392407883e-06,
+      "loss": 0.0831,
+      "step": 875
+    },
+    {
+      "epoch": 0.6014418125643667,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 9.268009650795264e-06,
+      "loss": 0.0977,
+      "step": 876
+    },
+    {
+      "epoch": 0.6021283899759697,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 9.26620287905103e-06,
+      "loss": 0.0859,
+      "step": 877
+    },
+    {
+      "epoch": 0.6028149673875729,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 9.264394078148636e-06,
+      "loss": 0.0827,
+      "step": 878
+    },
+    {
+      "epoch": 0.6035015447991761,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 9.26258324906263e-06,
+      "loss": 0.0829,
+      "step": 879
+    },
+    {
+      "epoch": 0.6041881222107792,
+      "grad_norm": 0.193359375,
+      "learning_rate": 9.260770392768652e-06,
+      "loss": 0.0744,
+      "step": 880
+    },
+    {
+      "epoch": 0.6048746996223824,
+      "grad_norm": 0.224609375,
+      "learning_rate": 9.258955510243431e-06,
+      "loss": 0.0882,
+      "step": 881
+    },
+    {
+      "epoch": 0.6055612770339855,
+      "grad_norm": 0.232421875,
+      "learning_rate": 9.257138602464795e-06,
+      "loss": 0.0882,
+      "step": 882
+    },
+    {
+      "epoch": 0.6062478544455887,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 9.255319670411658e-06,
+      "loss": 0.0842,
+      "step": 883
+    },
+    {
+      "epoch": 0.6069344318571919,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 9.253498715064025e-06,
+      "loss": 0.086,
+      "step": 884
+    },
+    {
+      "epoch": 0.607621009268795,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 9.251675737402992e-06,
+      "loss": 0.0915,
+      "step": 885
+    },
+    {
+      "epoch": 0.6083075866803982,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 9.249850738410749e-06,
+      "loss": 0.0926,
+      "step": 886
+    },
+    {
+      "epoch": 0.6089941640920014,
+      "grad_norm": 0.212890625,
+      "learning_rate": 9.248023719070563e-06,
+      "loss": 0.0854,
+      "step": 887
+    },
+    {
+      "epoch": 0.6096807415036045,
+      "grad_norm": 0.2734375,
+      "learning_rate": 9.246194680366802e-06,
+      "loss": 0.0915,
+      "step": 888
+    },
+    {
+      "epoch": 0.6103673189152077,
+      "grad_norm": 0.189453125,
+      "learning_rate": 9.244363623284916e-06,
+      "loss": 0.0858,
+      "step": 889
+    },
+    {
+      "epoch": 0.6110538963268108,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 9.242530548811444e-06,
+      "loss": 0.0899,
+      "step": 890
+    },
+    {
+      "epoch": 0.611740473738414,
+      "grad_norm": 0.26171875,
+      "learning_rate": 9.240695457934012e-06,
+      "loss": 0.0988,
+      "step": 891
+    },
+    {
+      "epoch": 0.6124270511500172,
+      "grad_norm": 0.216796875,
+      "learning_rate": 9.23885835164133e-06,
+      "loss": 0.0814,
+      "step": 892
+    },
+    {
+      "epoch": 0.6131136285616203,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 9.237019230923196e-06,
+      "loss": 0.0888,
+      "step": 893
+    },
+    {
+      "epoch": 0.6138002059732235,
+      "grad_norm": 0.20703125,
+      "learning_rate": 9.235178096770494e-06,
+      "loss": 0.0976,
+      "step": 894
+    },
+    {
+      "epoch": 0.6144867833848267,
+      "grad_norm": 0.20703125,
+      "learning_rate": 9.233334950175194e-06,
+      "loss": 0.0887,
+      "step": 895
+    },
+    {
+      "epoch": 0.6151733607964298,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 9.231489792130343e-06,
+      "loss": 0.0803,
+      "step": 896
+    },
+    {
+      "epoch": 0.615859938208033,
+      "grad_norm": 0.265625,
+      "learning_rate": 9.229642623630081e-06,
+      "loss": 0.0805,
+      "step": 897
+    },
+    {
+      "epoch": 0.6165465156196361,
+      "grad_norm": 0.220703125,
+      "learning_rate": 9.227793445669627e-06,
+      "loss": 0.0833,
+      "step": 898
+    },
+    {
+      "epoch": 0.6172330930312393,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 9.225942259245281e-06,
+      "loss": 0.0898,
+      "step": 899
+    },
+    {
+      "epoch": 0.6179196704428425,
+      "grad_norm": 0.2734375,
+      "learning_rate": 9.224089065354428e-06,
+      "loss": 0.0982,
+      "step": 900
+    },
+    {
+      "epoch": 0.6186062478544456,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 9.222233864995533e-06,
+      "loss": 0.0932,
+      "step": 901
+    },
+    {
+      "epoch": 0.6192928252660488,
+      "grad_norm": 0.240234375,
+      "learning_rate": 9.220376659168141e-06,
+      "loss": 0.1014,
+      "step": 902
+    },
+    {
+      "epoch": 0.619979402677652,
+      "grad_norm": 0.263671875,
+      "learning_rate": 9.21851744887288e-06,
+      "loss": 0.0886,
+      "step": 903
+    },
+    {
+      "epoch": 0.6206659800892551,
+      "grad_norm": 0.216796875,
+      "learning_rate": 9.216656235111463e-06,
+      "loss": 0.0871,
+      "step": 904
+    },
+    {
+      "epoch": 0.6213525575008583,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 9.214793018886666e-06,
+      "loss": 0.0915,
+      "step": 905
+    },
+    {
+      "epoch": 0.6220391349124614,
+      "grad_norm": 0.20703125,
+      "learning_rate": 9.212927801202361e-06,
+      "loss": 0.0902,
+      "step": 906
+    },
+    {
+      "epoch": 0.6227257123240645,
+      "grad_norm": 0.19140625,
+      "learning_rate": 9.211060583063489e-06,
+      "loss": 0.0735,
+      "step": 907
+    },
+    {
+      "epoch": 0.6234122897356676,
+      "grad_norm": 0.259765625,
+      "learning_rate": 9.209191365476074e-06,
+      "loss": 0.0956,
+      "step": 908
+    },
+    {
+      "epoch": 0.6240988671472708,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 9.207320149447212e-06,
+      "loss": 0.0876,
+      "step": 909
+    },
+    {
+      "epoch": 0.624785444558874,
+      "grad_norm": 0.20703125,
+      "learning_rate": 9.20544693598508e-06,
+      "loss": 0.0861,
+      "step": 910
+    },
+    {
+      "epoch": 0.6254720219704771,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 9.20357172609893e-06,
+      "loss": 0.085,
+      "step": 911
+    },
+    {
+      "epoch": 0.6261585993820803,
+      "grad_norm": 0.220703125,
+      "learning_rate": 9.201694520799086e-06,
+      "loss": 0.0918,
+      "step": 912
+    },
+    {
+      "epoch": 0.6268451767936835,
+      "grad_norm": 0.244140625,
+      "learning_rate": 9.199815321096953e-06,
+      "loss": 0.0878,
+      "step": 913
+    },
+    {
+      "epoch": 0.6275317542052866,
+      "grad_norm": 0.265625,
+      "learning_rate": 9.19793412800501e-06,
+      "loss": 0.0878,
+      "step": 914
+    },
+    {
+      "epoch": 0.6282183316168898,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 9.196050942536806e-06,
+      "loss": 0.0824,
+      "step": 915
+    },
+    {
+      "epoch": 0.6289049090284929,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 9.194165765706963e-06,
+      "loss": 0.0964,
+      "step": 916
+    },
+    {
+      "epoch": 0.6295914864400961,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 9.192278598531182e-06,
+      "loss": 0.092,
+      "step": 917
+    },
+    {
+      "epoch": 0.6302780638516993,
+      "grad_norm": 0.251953125,
+      "learning_rate": 9.19038944202623e-06,
+      "loss": 0.0863,
+      "step": 918
+    },
+    {
+      "epoch": 0.6309646412633024,
+      "grad_norm": 0.224609375,
+      "learning_rate": 9.18849829720995e-06,
+      "loss": 0.0859,
+      "step": 919
+    },
+    {
+      "epoch": 0.6316512186749056,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 9.186605165101253e-06,
+      "loss": 0.0972,
+      "step": 920
+    },
+    {
+      "epoch": 0.6323377960865088,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 9.184710046720123e-06,
+      "loss": 0.0879,
+      "step": 921
+    },
+    {
+      "epoch": 0.6330243734981119,
+      "grad_norm": 0.201171875,
+      "learning_rate": 9.182812943087614e-06,
+      "loss": 0.0803,
+      "step": 922
+    },
+    {
+      "epoch": 0.6337109509097151,
+      "grad_norm": 0.2109375,
+      "learning_rate": 9.18091385522585e-06,
+      "loss": 0.0815,
+      "step": 923
+    },
+    {
+      "epoch": 0.6343975283213182,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 9.179012784158023e-06,
+      "loss": 0.0862,
+      "step": 924
+    },
+    {
+      "epoch": 0.6350841057329214,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 9.177109730908393e-06,
+      "loss": 0.0813,
+      "step": 925
+    },
+    {
+      "epoch": 0.6357706831445246,
+      "grad_norm": 0.197265625,
+      "learning_rate": 9.175204696502288e-06,
+      "loss": 0.0993,
+      "step": 926
+    },
+    {
+      "epoch": 0.6364572605561277,
+      "grad_norm": 0.224609375,
+      "learning_rate": 9.173297681966105e-06,
+      "loss": 0.0888,
+      "step": 927
+    },
+    {
+      "epoch": 0.6371438379677309,
+      "grad_norm": 0.20703125,
+      "learning_rate": 9.171388688327307e-06,
+      "loss": 0.0875,
+      "step": 928
+    },
+    {
+      "epoch": 0.637830415379334,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 9.169477716614425e-06,
+      "loss": 0.0851,
+      "step": 929
+    },
+    {
+      "epoch": 0.6385169927909372,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 9.167564767857052e-06,
+      "loss": 0.0768,
+      "step": 930
+    },
+    {
+      "epoch": 0.6392035702025404,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 9.165649843085848e-06,
+      "loss": 0.0901,
+      "step": 931
+    },
+    {
+      "epoch": 0.6398901476141435,
+      "grad_norm": 0.1953125,
+      "learning_rate": 9.163732943332539e-06,
+      "loss": 0.0845,
+      "step": 932
+    },
+    {
+      "epoch": 0.6405767250257467,
+      "grad_norm": 0.263671875,
+      "learning_rate": 9.161814069629914e-06,
+      "loss": 0.0786,
+      "step": 933
+    },
+    {
+      "epoch": 0.6412633024373499,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 9.159893223011824e-06,
+      "loss": 0.0942,
+      "step": 934
+    },
+    {
+      "epoch": 0.641949879848953,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 9.157970404513185e-06,
+      "loss": 0.0897,
+      "step": 935
+    },
+    {
+      "epoch": 0.6426364572605562,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 9.156045615169978e-06,
+      "loss": 0.0896,
+      "step": 936
+    },
+    {
+      "epoch": 0.6433230346721592,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 9.154118856019239e-06,
+      "loss": 0.0875,
+      "step": 937
+    },
+    {
+      "epoch": 0.6440096120837624,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 9.15219012809907e-06,
+      "loss": 0.0805,
+      "step": 938
+    },
+    {
+      "epoch": 0.6446961894953656,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 9.150259432448632e-06,
+      "loss": 0.092,
+      "step": 939
+    },
+    {
+      "epoch": 0.6453827669069687,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 9.148326770108147e-06,
+      "loss": 0.0862,
+      "step": 940
+    },
+    {
+      "epoch": 0.6460693443185719,
+      "grad_norm": 0.216796875,
+      "learning_rate": 9.146392142118899e-06,
+      "loss": 0.0869,
+      "step": 941
+    },
+    {
+      "epoch": 0.646755921730175,
+      "grad_norm": 0.330078125,
+      "learning_rate": 9.144455549523227e-06,
+      "loss": 0.0805,
+      "step": 942
+    },
+    {
+      "epoch": 0.6474424991417782,
+      "grad_norm": 0.228515625,
+      "learning_rate": 9.14251699336453e-06,
+      "loss": 0.0933,
+      "step": 943
+    },
+    {
+      "epoch": 0.6481290765533814,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 9.140576474687263e-06,
+      "loss": 0.0838,
+      "step": 944
+    },
+    {
+      "epoch": 0.6488156539649845,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 9.138633994536945e-06,
+      "loss": 0.0862,
+      "step": 945
+    },
+    {
+      "epoch": 0.6495022313765877,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 9.136689553960144e-06,
+      "loss": 0.0874,
+      "step": 946
+    },
+    {
+      "epoch": 0.6501888087881909,
+      "grad_norm": 0.2578125,
+      "learning_rate": 9.134743154004488e-06,
+      "loss": 0.0862,
+      "step": 947
+    },
+    {
+      "epoch": 0.650875386199794,
+      "grad_norm": 0.275390625,
+      "learning_rate": 9.132794795718662e-06,
+      "loss": 0.0922,
+      "step": 948
+    },
+    {
+      "epoch": 0.6515619636113972,
+      "grad_norm": 0.2734375,
+      "learning_rate": 9.1308444801524e-06,
+      "loss": 0.0965,
+      "step": 949
+    },
+    {
+      "epoch": 0.6522485410230003,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 9.128892208356496e-06,
+      "loss": 0.0861,
+      "step": 950
+    },
+    {
+      "epoch": 0.6529351184346035,
+      "grad_norm": 0.201171875,
+      "learning_rate": 9.126937981382802e-06,
+      "loss": 0.0776,
+      "step": 951
+    },
+    {
+      "epoch": 0.6536216958462067,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 9.12498180028421e-06,
+      "loss": 0.0888,
+      "step": 952
+    },
+    {
+      "epoch": 0.6543082732578098,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 9.12302366611468e-06,
+      "loss": 0.094,
+      "step": 953
+    },
+    {
+      "epoch": 0.654994850669413,
+      "grad_norm": 0.205078125,
+      "learning_rate": 9.121063579929214e-06,
+      "loss": 0.0887,
+      "step": 954
+    },
+    {
+      "epoch": 0.6556814280810161,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 9.119101542783868e-06,
+      "loss": 0.0937,
+      "step": 955
+    },
+    {
+      "epoch": 0.6563680054926193,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 9.117137555735753e-06,
+      "loss": 0.0921,
+      "step": 956
+    },
+    {
+      "epoch": 0.6570545829042225,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 9.115171619843025e-06,
+      "loss": 0.0943,
+      "step": 957
+    },
+    {
+      "epoch": 0.6577411603158256,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 9.113203736164894e-06,
+      "loss": 0.0875,
+      "step": 958
+    },
+    {
+      "epoch": 0.6584277377274288,
+      "grad_norm": 0.26171875,
+      "learning_rate": 9.111233905761618e-06,
+      "loss": 0.0907,
+      "step": 959
+    },
+    {
+      "epoch": 0.659114315139032,
+      "grad_norm": 0.2421875,
+      "learning_rate": 9.109262129694506e-06,
+      "loss": 0.0853,
+      "step": 960
+    },
+    {
+      "epoch": 0.6598008925506351,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 9.107288409025909e-06,
+      "loss": 0.1071,
+      "step": 961
+    },
+    {
+      "epoch": 0.6604874699622383,
+      "grad_norm": 0.240234375,
+      "learning_rate": 9.105312744819232e-06,
+      "loss": 0.08,
+      "step": 962
+    },
+    {
+      "epoch": 0.6611740473738414,
+      "grad_norm": 0.21484375,
+      "learning_rate": 9.103335138138926e-06,
+      "loss": 0.1027,
+      "step": 963
+    },
+    {
+      "epoch": 0.6618606247854446,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 9.101355590050489e-06,
+      "loss": 0.0876,
+      "step": 964
+    },
+    {
+      "epoch": 0.6625472021970478,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 9.09937410162046e-06,
+      "loss": 0.0934,
+      "step": 965
+    },
+    {
+      "epoch": 0.6632337796086509,
+      "grad_norm": 0.25390625,
+      "learning_rate": 9.09739067391643e-06,
+      "loss": 0.0899,
+      "step": 966
+    },
+    {
+      "epoch": 0.663920357020254,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 9.095405308007029e-06,
+      "loss": 0.0808,
+      "step": 967
+    },
+    {
+      "epoch": 0.6646069344318571,
+      "grad_norm": 0.20703125,
+      "learning_rate": 9.093418004961939e-06,
+      "loss": 0.0847,
+      "step": 968
+    },
+    {
+      "epoch": 0.6652935118434603,
+      "grad_norm": 0.208984375,
+      "learning_rate": 9.091428765851877e-06,
+      "loss": 0.085,
+      "step": 969
+    },
+    {
+      "epoch": 0.6659800892550635,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 9.089437591748607e-06,
+      "loss": 0.0868,
+      "step": 970
+    },
+    {
+      "epoch": 0.6666666666666666,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 9.087444483724937e-06,
+      "loss": 0.0836,
+      "step": 971
+    },
+    {
+      "epoch": 0.6673532440782698,
+      "grad_norm": 0.2109375,
+      "learning_rate": 9.085449442854715e-06,
+      "loss": 0.0834,
+      "step": 972
+    },
+    {
+      "epoch": 0.668039821489873,
+      "grad_norm": 0.193359375,
+      "learning_rate": 9.08345247021283e-06,
+      "loss": 0.0814,
+      "step": 973
+    },
+    {
+      "epoch": 0.6687263989014761,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 9.081453566875215e-06,
+      "loss": 0.0997,
+      "step": 974
+    },
+    {
+      "epoch": 0.6694129763130793,
+      "grad_norm": 0.248046875,
+      "learning_rate": 9.079452733918841e-06,
+      "loss": 0.0944,
+      "step": 975
+    },
+    {
+      "epoch": 0.6700995537246824,
+      "grad_norm": 0.205078125,
+      "learning_rate": 9.077449972421716e-06,
+      "loss": 0.0903,
+      "step": 976
+    },
+    {
+      "epoch": 0.6707861311362856,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 9.07544528346289e-06,
+      "loss": 0.0991,
+      "step": 977
+    },
+    {
+      "epoch": 0.6714727085478888,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 9.073438668122454e-06,
+      "loss": 0.0749,
+      "step": 978
+    },
+    {
+      "epoch": 0.6721592859594919,
+      "grad_norm": 0.205078125,
+      "learning_rate": 9.07143012748153e-06,
+      "loss": 0.0952,
+      "step": 979
+    },
+    {
+      "epoch": 0.6728458633710951,
+      "grad_norm": 0.216796875,
+      "learning_rate": 9.069419662622284e-06,
+      "loss": 0.0932,
+      "step": 980
+    },
+    {
+      "epoch": 0.6735324407826982,
+      "grad_norm": 0.263671875,
+      "learning_rate": 9.067407274627917e-06,
+      "loss": 0.091,
+      "step": 981
+    },
+    {
+      "epoch": 0.6742190181943014,
+      "grad_norm": 0.220703125,
+      "learning_rate": 9.06539296458266e-06,
+      "loss": 0.0909,
+      "step": 982
+    },
+    {
+      "epoch": 0.6749055956059046,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 9.063376733571791e-06,
+      "loss": 0.0866,
+      "step": 983
+    },
+    {
+      "epoch": 0.6755921730175077,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 9.061358582681614e-06,
+      "loss": 0.0875,
+      "step": 984
+    },
+    {
+      "epoch": 0.6762787504291109,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 9.05933851299947e-06,
+      "loss": 0.095,
+      "step": 985
+    },
+    {
+      "epoch": 0.6769653278407141,
+      "grad_norm": 0.251953125,
+      "learning_rate": 9.057316525613735e-06,
+      "loss": 0.0827,
+      "step": 986
+    },
+    {
+      "epoch": 0.6776519052523172,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 9.055292621613815e-06,
+      "loss": 0.1015,
+      "step": 987
+    },
+    {
+      "epoch": 0.6783384826639204,
+      "grad_norm": 0.2265625,
+      "learning_rate": 9.053266802090152e-06,
+      "loss": 0.0923,
+      "step": 988
+    },
+    {
+      "epoch": 0.6790250600755235,
+      "grad_norm": 0.189453125,
+      "learning_rate": 9.051239068134221e-06,
+      "loss": 0.0868,
+      "step": 989
+    },
+    {
+      "epoch": 0.6797116374871267,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 9.049209420838522e-06,
+      "loss": 0.0831,
+      "step": 990
+    },
+    {
+      "epoch": 0.6803982148987299,
+      "grad_norm": 0.216796875,
+      "learning_rate": 9.047177861296595e-06,
+      "loss": 0.0849,
+      "step": 991
+    },
+    {
+      "epoch": 0.681084792310333,
+      "grad_norm": 0.220703125,
+      "learning_rate": 9.045144390603e-06,
+      "loss": 0.0842,
+      "step": 992
+    },
+    {
+      "epoch": 0.6817713697219362,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 9.043109009853337e-06,
+      "loss": 0.0832,
+      "step": 993
+    },
+    {
+      "epoch": 0.6824579471335394,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 9.041071720144232e-06,
+      "loss": 0.0819,
+      "step": 994
+    },
+    {
+      "epoch": 0.6831445245451425,
+      "grad_norm": 0.228515625,
+      "learning_rate": 9.039032522573328e-06,
+      "loss": 0.0964,
+      "step": 995
+    },
+    {
+      "epoch": 0.6838311019567456,
+      "grad_norm": 0.25390625,
+      "learning_rate": 9.036991418239316e-06,
+      "loss": 0.0771,
+      "step": 996
+    },
+    {
+      "epoch": 0.6845176793683487,
+      "grad_norm": 0.228515625,
+      "learning_rate": 9.034948408241898e-06,
+      "loss": 0.0841,
+      "step": 997
+    },
+    {
+      "epoch": 0.6852042567799519,
+      "grad_norm": 0.259765625,
+      "learning_rate": 9.032903493681811e-06,
+      "loss": 0.0818,
+      "step": 998
+    },
+    {
+      "epoch": 0.685890834191555,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 9.030856675660816e-06,
+      "loss": 0.087,
+      "step": 999
+    },
+    {
+      "epoch": 0.6865774116031582,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 9.028807955281701e-06,
+      "loss": 0.0878,
+      "step": 1000
+    },
+    {
+      "epoch": 0.6872639890147614,
+      "grad_norm": 0.220703125,
+      "learning_rate": 9.026757333648272e-06,
+      "loss": 0.0864,
+      "step": 1001
+    },
+    {
+      "epoch": 0.6879505664263645,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 9.02470481186537e-06,
+      "loss": 0.0869,
+      "step": 1002
+    },
+    {
+      "epoch": 0.6886371438379677,
+      "grad_norm": 0.203125,
+      "learning_rate": 9.022650391038851e-06,
+      "loss": 0.085,
+      "step": 1003
+    },
+    {
+      "epoch": 0.6893237212495709,
+      "grad_norm": 0.25390625,
+      "learning_rate": 9.020594072275602e-06,
+      "loss": 0.0868,
+      "step": 1004
+    },
+    {
+      "epoch": 0.690010298661174,
+      "grad_norm": 0.208984375,
+      "learning_rate": 9.018535856683525e-06,
+      "loss": 0.0926,
+      "step": 1005
+    },
+    {
+      "epoch": 0.6906968760727772,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 9.016475745371547e-06,
+      "loss": 0.0865,
+      "step": 1006
+    },
+    {
+      "epoch": 0.6913834534843804,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 9.014413739449619e-06,
+      "loss": 0.0937,
+      "step": 1007
+    },
+    {
+      "epoch": 0.6920700308959835,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 9.012349840028707e-06,
+      "loss": 0.0965,
+      "step": 1008
+    },
+    {
+      "epoch": 0.6927566083075867,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 9.010284048220805e-06,
+      "loss": 0.0825,
+      "step": 1009
+    },
+    {
+      "epoch": 0.6934431857191898,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 9.008216365138917e-06,
+      "loss": 0.0952,
+      "step": 1010
+    },
+    {
+      "epoch": 0.694129763130793,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 9.006146791897077e-06,
+      "loss": 0.0814,
+      "step": 1011
+    },
+    {
+      "epoch": 0.6948163405423962,
+      "grad_norm": 0.21484375,
+      "learning_rate": 9.004075329610324e-06,
+      "loss": 0.0818,
+      "step": 1012
+    },
+    {
+      "epoch": 0.6955029179539993,
+      "grad_norm": 0.20703125,
+      "learning_rate": 9.00200197939473e-06,
+      "loss": 0.0724,
+      "step": 1013
+    },
+    {
+      "epoch": 0.6961894953656025,
+      "grad_norm": 0.60546875,
+      "learning_rate": 8.999926742367372e-06,
+      "loss": 0.0806,
+      "step": 1014
+    },
+    {
+      "epoch": 0.6968760727772056,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 8.997849619646349e-06,
+      "loss": 0.0844,
+      "step": 1015
+    },
+    {
+      "epoch": 0.6975626501888088,
+      "grad_norm": 0.2421875,
+      "learning_rate": 8.995770612350778e-06,
+      "loss": 0.0863,
+      "step": 1016
+    },
+    {
+      "epoch": 0.698249227600412,
+      "grad_norm": 0.23046875,
+      "learning_rate": 8.993689721600781e-06,
+      "loss": 0.0907,
+      "step": 1017
+    },
+    {
+      "epoch": 0.6989358050120151,
+      "grad_norm": 0.19921875,
+      "learning_rate": 8.99160694851751e-06,
+      "loss": 0.0827,
+      "step": 1018
+    },
+    {
+      "epoch": 0.6996223824236183,
+      "grad_norm": 0.2265625,
+      "learning_rate": 8.98952229422312e-06,
+      "loss": 0.0865,
+      "step": 1019
+    },
+    {
+      "epoch": 0.7003089598352215,
+      "grad_norm": 0.23828125,
+      "learning_rate": 8.987435759840782e-06,
+      "loss": 0.0926,
+      "step": 1020
+    },
+    {
+      "epoch": 0.7009955372468246,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 8.985347346494682e-06,
+      "loss": 0.0905,
+      "step": 1021
+    },
+    {
+      "epoch": 0.7016821146584278,
+      "grad_norm": 0.216796875,
+      "learning_rate": 8.983257055310018e-06,
+      "loss": 0.0832,
+      "step": 1022
+    },
+    {
+      "epoch": 0.7023686920700309,
+      "grad_norm": 0.24609375,
+      "learning_rate": 8.981164887412998e-06,
+      "loss": 0.0906,
+      "step": 1023
+    },
+    {
+      "epoch": 0.7030552694816341,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 8.979070843930841e-06,
+      "loss": 0.0805,
+      "step": 1024
+    },
+    {
+      "epoch": 0.7037418468932373,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 8.97697492599178e-06,
+      "loss": 0.0755,
+      "step": 1025
+    },
+    {
+      "epoch": 0.7044284243048403,
+      "grad_norm": 0.208984375,
+      "learning_rate": 8.974877134725054e-06,
+      "loss": 0.0901,
+      "step": 1026
+    },
+    {
+      "epoch": 0.7051150017164435,
+      "grad_norm": 0.248046875,
+      "learning_rate": 8.972777471260914e-06,
+      "loss": 0.0792,
+      "step": 1027
+    },
+    {
+      "epoch": 0.7058015791280466,
+      "grad_norm": 0.20703125,
+      "learning_rate": 8.970675936730618e-06,
+      "loss": 0.0873,
+      "step": 1028
+    },
+    {
+      "epoch": 0.7064881565396498,
+      "grad_norm": 0.193359375,
+      "learning_rate": 8.968572532266432e-06,
+      "loss": 0.0859,
+      "step": 1029
+    },
+    {
+      "epoch": 0.707174733951253,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 8.96646725900163e-06,
+      "loss": 0.0938,
+      "step": 1030
+    },
+    {
+      "epoch": 0.7078613113628561,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 8.964360118070495e-06,
+      "loss": 0.0828,
+      "step": 1031
+    },
+    {
+      "epoch": 0.7085478887744593,
+      "grad_norm": 0.205078125,
+      "learning_rate": 8.96225111060831e-06,
+      "loss": 0.0836,
+      "step": 1032
+    },
+    {
+      "epoch": 0.7092344661860625,
+      "grad_norm": 0.228515625,
+      "learning_rate": 8.960140237751374e-06,
+      "loss": 0.0843,
+      "step": 1033
+    },
+    {
+      "epoch": 0.7099210435976656,
+      "grad_norm": 0.21484375,
+      "learning_rate": 8.95802750063698e-06,
+      "loss": 0.0863,
+      "step": 1034
+    },
+    {
+      "epoch": 0.7106076210092688,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 8.955912900403432e-06,
+      "loss": 0.0877,
+      "step": 1035
+    },
+    {
+      "epoch": 0.7112941984208719,
+      "grad_norm": 0.21484375,
+      "learning_rate": 8.953796438190038e-06,
+      "loss": 0.0862,
+      "step": 1036
+    },
+    {
+      "epoch": 0.7119807758324751,
+      "grad_norm": 0.20703125,
+      "learning_rate": 8.951678115137103e-06,
+      "loss": 0.0933,
+      "step": 1037
+    },
+    {
+      "epoch": 0.7126673532440783,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 8.949557932385946e-06,
+      "loss": 0.0864,
+      "step": 1038
+    },
+    {
+      "epoch": 0.7133539306556814,
+      "grad_norm": 0.228515625,
+      "learning_rate": 8.947435891078873e-06,
+      "loss": 0.0856,
+      "step": 1039
+    },
+    {
+      "epoch": 0.7140405080672846,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 8.945311992359206e-06,
+      "loss": 0.0854,
+      "step": 1040
+    },
+    {
+      "epoch": 0.7147270854788877,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 8.943186237371256e-06,
+      "loss": 0.089,
+      "step": 1041
+    },
+    {
+      "epoch": 0.7154136628904909,
+      "grad_norm": 0.21875,
+      "learning_rate": 8.941058627260342e-06,
+      "loss": 0.0806,
+      "step": 1042
+    },
+    {
+      "epoch": 0.7161002403020941,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 8.938929163172779e-06,
+      "loss": 0.0937,
+      "step": 1043
+    },
+    {
+      "epoch": 0.7167868177136972,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 8.936797846255882e-06,
+      "loss": 0.0878,
+      "step": 1044
+    },
+    {
+      "epoch": 0.7174733951253004,
+      "grad_norm": 0.23046875,
+      "learning_rate": 8.934664677657963e-06,
+      "loss": 0.0908,
+      "step": 1045
+    },
+    {
+      "epoch": 0.7181599725369036,
+      "grad_norm": 0.220703125,
+      "learning_rate": 8.932529658528332e-06,
+      "loss": 0.0854,
+      "step": 1046
+    },
+    {
+      "epoch": 0.7188465499485067,
+      "grad_norm": 0.265625,
+      "learning_rate": 8.9303927900173e-06,
+      "loss": 0.0899,
+      "step": 1047
+    },
+    {
+      "epoch": 0.7195331273601099,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 8.928254073276166e-06,
+      "loss": 0.0921,
+      "step": 1048
+    },
+    {
+      "epoch": 0.720219704771713,
+      "grad_norm": 0.265625,
+      "learning_rate": 8.926113509457232e-06,
+      "loss": 0.0919,
+      "step": 1049
+    },
+    {
+      "epoch": 0.7209062821833162,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 8.923971099713795e-06,
+      "loss": 0.0906,
+      "step": 1050
+    },
+    {
+      "epoch": 0.7215928595949194,
+      "grad_norm": 0.341796875,
+      "learning_rate": 8.92182684520014e-06,
+      "loss": 0.0916,
+      "step": 1051
+    },
+    {
+      "epoch": 0.7222794370065225,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 8.919680747071554e-06,
+      "loss": 0.076,
+      "step": 1052
+    },
+    {
+      "epoch": 0.7229660144181257,
+      "grad_norm": 0.27734375,
+      "learning_rate": 8.917532806484313e-06,
+      "loss": 0.0873,
+      "step": 1053
+    },
+    {
+      "epoch": 0.7236525918297289,
+      "grad_norm": 0.25,
+      "learning_rate": 8.915383024595683e-06,
+      "loss": 0.0899,
+      "step": 1054
+    },
+    {
+      "epoch": 0.724339169241332,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 8.91323140256393e-06,
+      "loss": 0.0906,
+      "step": 1055
+    },
+    {
+      "epoch": 0.7250257466529351,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 8.911077941548306e-06,
+      "loss": 0.0802,
+      "step": 1056
+    },
+    {
+      "epoch": 0.7257123240645382,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 8.908922642709053e-06,
+      "loss": 0.0889,
+      "step": 1057
+    },
+    {
+      "epoch": 0.7263989014761414,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 8.906765507207405e-06,
+      "loss": 0.0867,
+      "step": 1058
+    },
+    {
+      "epoch": 0.7270854788877446,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 8.904606536205587e-06,
+      "loss": 0.084,
+      "step": 1059
+    },
+    {
+      "epoch": 0.7277720562993477,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 8.90244573086681e-06,
+      "loss": 0.0872,
+      "step": 1060
+    },
+    {
+      "epoch": 0.7284586337109509,
+      "grad_norm": 0.2265625,
+      "learning_rate": 8.900283092355274e-06,
+      "loss": 0.0837,
+      "step": 1061
+    },
+    {
+      "epoch": 0.729145211122554,
+      "grad_norm": 0.21875,
+      "learning_rate": 8.89811862183617e-06,
+      "loss": 0.0881,
+      "step": 1062
+    },
+    {
+      "epoch": 0.7298317885341572,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 8.895952320475669e-06,
+      "loss": 0.0867,
+      "step": 1063
+    },
+    {
+      "epoch": 0.7305183659457604,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 8.893784189440937e-06,
+      "loss": 0.0952,
+      "step": 1064
+    },
+    {
+      "epoch": 0.7312049433573635,
+      "grad_norm": 0.251953125,
+      "learning_rate": 8.89161422990012e-06,
+      "loss": 0.0841,
+      "step": 1065
+    },
+    {
+      "epoch": 0.7318915207689667,
+      "grad_norm": 0.251953125,
+      "learning_rate": 8.88944244302235e-06,
+      "loss": 0.1001,
+      "step": 1066
+    },
+    {
+      "epoch": 0.7325780981805698,
+      "grad_norm": 0.21484375,
+      "learning_rate": 8.887268829977745e-06,
+      "loss": 0.0929,
+      "step": 1067
+    },
+    {
+      "epoch": 0.733264675592173,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 8.885093391937406e-06,
+      "loss": 0.0933,
+      "step": 1068
+    },
+    {
+      "epoch": 0.7339512530037762,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 8.882916130073418e-06,
+      "loss": 0.0812,
+      "step": 1069
+    },
+    {
+      "epoch": 0.7346378304153793,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 8.880737045558849e-06,
+      "loss": 0.0835,
+      "step": 1070
+    },
+    {
+      "epoch": 0.7353244078269825,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 8.878556139567744e-06,
+      "loss": 0.0829,
+      "step": 1071
+    },
+    {
+      "epoch": 0.7360109852385857,
+      "grad_norm": 0.21875,
+      "learning_rate": 8.876373413275139e-06,
+      "loss": 0.0863,
+      "step": 1072
+    },
+    {
+      "epoch": 0.7366975626501888,
+      "grad_norm": 0.216796875,
+      "learning_rate": 8.87418886785704e-06,
+      "loss": 0.0848,
+      "step": 1073
+    },
+    {
+      "epoch": 0.737384140061792,
+      "grad_norm": 0.2578125,
+      "learning_rate": 8.872002504490443e-06,
+      "loss": 0.0959,
+      "step": 1074
+    },
+    {
+      "epoch": 0.7380707174733951,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 8.869814324353316e-06,
+      "loss": 0.0851,
+      "step": 1075
+    },
+    {
+      "epoch": 0.7387572948849983,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 8.86762432862461e-06,
+      "loss": 0.0911,
+      "step": 1076
+    },
+    {
+      "epoch": 0.7394438722966015,
+      "grad_norm": 0.2734375,
+      "learning_rate": 8.86543251848425e-06,
+      "loss": 0.0842,
+      "step": 1077
+    },
+    {
+      "epoch": 0.7401304497082046,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 8.863238895113145e-06,
+      "loss": 0.0846,
+      "step": 1078
+    },
+    {
+      "epoch": 0.7408170271198078,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 8.861043459693174e-06,
+      "loss": 0.0837,
+      "step": 1079
+    },
+    {
+      "epoch": 0.741503604531411,
+      "grad_norm": 0.205078125,
+      "learning_rate": 8.858846213407201e-06,
+      "loss": 0.0921,
+      "step": 1080
+    },
+    {
+      "epoch": 0.7421901819430141,
+      "grad_norm": 0.298828125,
+      "learning_rate": 8.856647157439055e-06,
+      "loss": 0.0771,
+      "step": 1081
+    },
+    {
+      "epoch": 0.7428767593546173,
+      "grad_norm": 0.310546875,
+      "learning_rate": 8.854446292973548e-06,
+      "loss": 0.0853,
+      "step": 1082
+    },
+    {
+      "epoch": 0.7435633367662204,
+      "grad_norm": 0.25,
+      "learning_rate": 8.852243621196463e-06,
+      "loss": 0.0928,
+      "step": 1083
+    },
+    {
+      "epoch": 0.7442499141778236,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 8.85003914329456e-06,
+      "loss": 0.083,
+      "step": 1084
+    },
+    {
+      "epoch": 0.7449364915894267,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 8.847832860455563e-06,
+      "loss": 0.0872,
+      "step": 1085
+    },
+    {
+      "epoch": 0.7456230690010298,
+      "grad_norm": 0.25390625,
+      "learning_rate": 8.84562477386818e-06,
+      "loss": 0.0971,
+      "step": 1086
+    },
+    {
+      "epoch": 0.746309646412633,
+      "grad_norm": 0.240234375,
+      "learning_rate": 8.843414884722086e-06,
+      "loss": 0.0806,
+      "step": 1087
+    },
+    {
+      "epoch": 0.7469962238242361,
+      "grad_norm": 0.2734375,
+      "learning_rate": 8.841203194207925e-06,
+      "loss": 0.0854,
+      "step": 1088
+    },
+    {
+      "epoch": 0.7476828012358393,
+      "grad_norm": 0.279296875,
+      "learning_rate": 8.838989703517315e-06,
+      "loss": 0.0825,
+      "step": 1089
+    },
+    {
+      "epoch": 0.7483693786474425,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 8.836774413842843e-06,
+      "loss": 0.0824,
+      "step": 1090
+    },
+    {
+      "epoch": 0.7490559560590456,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 8.834557326378062e-06,
+      "loss": 0.085,
+      "step": 1091
+    },
+    {
+      "epoch": 0.7497425334706488,
+      "grad_norm": 0.22265625,
+      "learning_rate": 8.832338442317499e-06,
+      "loss": 0.0916,
+      "step": 1092
+    },
+    {
+      "epoch": 0.750429110882252,
+      "grad_norm": 0.2109375,
+      "learning_rate": 8.830117762856648e-06,
+      "loss": 0.0938,
+      "step": 1093
+    },
+    {
+      "epoch": 0.7511156882938551,
+      "grad_norm": 0.2265625,
+      "learning_rate": 8.827895289191965e-06,
+      "loss": 0.0772,
+      "step": 1094
+    },
+    {
+      "epoch": 0.7518022657054583,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 8.825671022520876e-06,
+      "loss": 0.0983,
+      "step": 1095
+    },
+    {
+      "epoch": 0.7524888431170614,
+      "grad_norm": 0.228515625,
+      "learning_rate": 8.823444964041777e-06,
+      "loss": 0.085,
+      "step": 1096
+    },
+    {
+      "epoch": 0.7531754205286646,
+      "grad_norm": 0.224609375,
+      "learning_rate": 8.821217114954022e-06,
+      "loss": 0.0937,
+      "step": 1097
+    },
+    {
+      "epoch": 0.7538619979402678,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 8.818987476457936e-06,
+      "loss": 0.088,
+      "step": 1098
+    },
+    {
+      "epoch": 0.7545485753518709,
+      "grad_norm": 0.251953125,
+      "learning_rate": 8.816756049754805e-06,
+      "loss": 0.0794,
+      "step": 1099
+    },
+    {
+      "epoch": 0.7552351527634741,
+      "grad_norm": 0.2578125,
+      "learning_rate": 8.81452283604688e-06,
+      "loss": 0.0947,
+      "step": 1100
+    },
+    {
+      "epoch": 0.7559217301750772,
+      "grad_norm": 0.21484375,
+      "learning_rate": 8.81228783653737e-06,
+      "loss": 0.0869,
+      "step": 1101
+    },
+    {
+      "epoch": 0.7566083075866804,
+      "grad_norm": 0.23828125,
+      "learning_rate": 8.810051052430457e-06,
+      "loss": 0.0932,
+      "step": 1102
+    },
+    {
+      "epoch": 0.7572948849982836,
+      "grad_norm": 0.2265625,
+      "learning_rate": 8.80781248493127e-06,
+      "loss": 0.0799,
+      "step": 1103
+    },
+    {
+      "epoch": 0.7579814624098867,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 8.805572135245911e-06,
+      "loss": 0.0899,
+      "step": 1104
+    },
+    {
+      "epoch": 0.7586680398214899,
+      "grad_norm": 0.201171875,
+      "learning_rate": 8.803330004581438e-06,
+      "loss": 0.0856,
+      "step": 1105
+    },
+    {
+      "epoch": 0.759354617233093,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 8.801086094145867e-06,
+      "loss": 0.0878,
+      "step": 1106
+    },
+    {
+      "epoch": 0.7600411946446962,
+      "grad_norm": 0.24609375,
+      "learning_rate": 8.79884040514817e-06,
+      "loss": 0.1001,
+      "step": 1107
+    },
+    {
+      "epoch": 0.7607277720562994,
+      "grad_norm": 0.234375,
+      "learning_rate": 8.796592938798287e-06,
+      "loss": 0.0834,
+      "step": 1108
+    },
+    {
+      "epoch": 0.7614143494679025,
+      "grad_norm": 0.21484375,
+      "learning_rate": 8.79434369630711e-06,
+      "loss": 0.0826,
+      "step": 1109
+    },
+    {
+      "epoch": 0.7621009268795057,
+      "grad_norm": 0.22265625,
+      "learning_rate": 8.792092678886483e-06,
+      "loss": 0.0963,
+      "step": 1110
+    },
+    {
+      "epoch": 0.7627875042911089,
+      "grad_norm": 0.220703125,
+      "learning_rate": 8.789839887749215e-06,
+      "loss": 0.0921,
+      "step": 1111
+    },
+    {
+      "epoch": 0.763474081702712,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 8.787585324109067e-06,
+      "loss": 0.0798,
+      "step": 1112
+    },
+    {
+      "epoch": 0.7641606591143152,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 8.785328989180753e-06,
+      "loss": 0.0914,
+      "step": 1113
+    },
+    {
+      "epoch": 0.7648472365259184,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 8.783070884179943e-06,
+      "loss": 0.1003,
+      "step": 1114
+    },
+    {
+      "epoch": 0.7655338139375214,
+      "grad_norm": 0.21875,
+      "learning_rate": 8.780811010323264e-06,
+      "loss": 0.0813,
+      "step": 1115
+    },
+    {
+      "epoch": 0.7662203913491246,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 8.77854936882829e-06,
+      "loss": 0.0881,
+      "step": 1116
+    },
+    {
+      "epoch": 0.7669069687607277,
+      "grad_norm": 0.201171875,
+      "learning_rate": 8.776285960913549e-06,
+      "loss": 0.0903,
+      "step": 1117
+    },
+    {
+      "epoch": 0.7675935461723309,
+      "grad_norm": 0.2109375,
+      "learning_rate": 8.774020787798525e-06,
+      "loss": 0.0858,
+      "step": 1118
+    },
+    {
+      "epoch": 0.768280123583934,
+      "grad_norm": 0.248046875,
+      "learning_rate": 8.771753850703649e-06,
+      "loss": 0.0832,
+      "step": 1119
+    },
+    {
+      "epoch": 0.7689667009955372,
+      "grad_norm": 0.224609375,
+      "learning_rate": 8.7694851508503e-06,
+      "loss": 0.0922,
+      "step": 1120
+    },
+    {
+      "epoch": 0.7696532784071404,
+      "grad_norm": 0.19140625,
+      "learning_rate": 8.767214689460815e-06,
+      "loss": 0.0787,
+      "step": 1121
+    },
+    {
+      "epoch": 0.7703398558187435,
+      "grad_norm": 0.220703125,
+      "learning_rate": 8.764942467758476e-06,
+      "loss": 0.0866,
+      "step": 1122
+    },
+    {
+      "epoch": 0.7710264332303467,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 8.762668486967508e-06,
+      "loss": 0.0868,
+      "step": 1123
+    },
+    {
+      "epoch": 0.7717130106419499,
+      "grad_norm": 0.244140625,
+      "learning_rate": 8.760392748313092e-06,
+      "loss": 0.09,
+      "step": 1124
+    },
+    {
+      "epoch": 0.772399588053553,
+      "grad_norm": 0.220703125,
+      "learning_rate": 8.758115253021348e-06,
+      "loss": 0.0844,
+      "step": 1125
+    },
+    {
+      "epoch": 0.7730861654651562,
+      "grad_norm": 0.21875,
+      "learning_rate": 8.755836002319353e-06,
+      "loss": 0.0851,
+      "step": 1126
+    },
+    {
+      "epoch": 0.7737727428767593,
+      "grad_norm": 0.2109375,
+      "learning_rate": 8.75355499743512e-06,
+      "loss": 0.0906,
+      "step": 1127
+    },
+    {
+      "epoch": 0.7744593202883625,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 8.751272239597612e-06,
+      "loss": 0.0998,
+      "step": 1128
+    },
+    {
+      "epoch": 0.7751458976999657,
+      "grad_norm": 0.205078125,
+      "learning_rate": 8.748987730036733e-06,
+      "loss": 0.0811,
+      "step": 1129
+    },
+    {
+      "epoch": 0.7758324751115688,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 8.746701469983336e-06,
+      "loss": 0.0786,
+      "step": 1130
+    },
+    {
+      "epoch": 0.776519052523172,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 8.744413460669213e-06,
+      "loss": 0.0848,
+      "step": 1131
+    },
+    {
+      "epoch": 0.7772056299347752,
+      "grad_norm": 0.201171875,
+      "learning_rate": 8.7421237033271e-06,
+      "loss": 0.079,
+      "step": 1132
+    },
+    {
+      "epoch": 0.7778922073463783,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 8.739832199190677e-06,
+      "loss": 0.0886,
+      "step": 1133
+    },
+    {
+      "epoch": 0.7785787847579815,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 8.737538949494557e-06,
+      "loss": 0.0877,
+      "step": 1134
+    },
+    {
+      "epoch": 0.7792653621695846,
+      "grad_norm": 0.19921875,
+      "learning_rate": 8.735243955474305e-06,
+      "loss": 0.0901,
+      "step": 1135
+    },
+    {
+      "epoch": 0.7799519395811878,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 8.732947218366414e-06,
+      "loss": 0.0806,
+      "step": 1136
+    },
+    {
+      "epoch": 0.780638516992791,
+      "grad_norm": 0.251953125,
+      "learning_rate": 8.730648739408326e-06,
+      "loss": 0.093,
+      "step": 1137
+    },
+    {
+      "epoch": 0.7813250944043941,
+      "grad_norm": 0.3125,
+      "learning_rate": 8.728348519838418e-06,
+      "loss": 0.0909,
+      "step": 1138
+    },
+    {
+      "epoch": 0.7820116718159973,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 8.726046560896002e-06,
+      "loss": 0.0932,
+      "step": 1139
+    },
+    {
+      "epoch": 0.7826982492276005,
+      "grad_norm": 0.236328125,
+      "learning_rate": 8.723742863821332e-06,
+      "loss": 0.0893,
+      "step": 1140
+    },
+    {
+      "epoch": 0.7833848266392036,
+      "grad_norm": 0.205078125,
+      "learning_rate": 8.721437429855594e-06,
+      "loss": 0.0857,
+      "step": 1141
+    },
+    {
+      "epoch": 0.7840714040508068,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 8.719130260240913e-06,
+      "loss": 0.0963,
+      "step": 1142
+    },
+    {
+      "epoch": 0.7847579814624099,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 8.716821356220348e-06,
+      "loss": 0.098,
+      "step": 1143
+    },
+    {
+      "epoch": 0.7854445588740131,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 8.71451071903789e-06,
+      "loss": 0.0974,
+      "step": 1144
+    },
+    {
+      "epoch": 0.7861311362856161,
+      "grad_norm": 0.23046875,
+      "learning_rate": 8.712198349938468e-06,
+      "loss": 0.0932,
+      "step": 1145
+    },
+    {
+      "epoch": 0.7868177136972193,
+      "grad_norm": 0.21875,
+      "learning_rate": 8.709884250167944e-06,
+      "loss": 0.0944,
+      "step": 1146
+    },
+    {
+      "epoch": 0.7875042911088225,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 8.707568420973107e-06,
+      "loss": 0.0898,
+      "step": 1147
+    },
+    {
+      "epoch": 0.7881908685204256,
+      "grad_norm": 0.236328125,
+      "learning_rate": 8.705250863601684e-06,
+      "loss": 0.0866,
+      "step": 1148
+    },
+    {
+      "epoch": 0.7888774459320288,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 8.70293157930233e-06,
+      "loss": 0.0852,
+      "step": 1149
+    },
+    {
+      "epoch": 0.789564023343632,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 8.700610569324632e-06,
+      "loss": 0.0762,
+      "step": 1150
+    },
+    {
+      "epoch": 0.7902506007552351,
+      "grad_norm": 0.203125,
+      "learning_rate": 8.698287834919103e-06,
+      "loss": 0.0851,
+      "step": 1151
+    },
+    {
+      "epoch": 0.7909371781668383,
+      "grad_norm": 0.23046875,
+      "learning_rate": 8.695963377337191e-06,
+      "loss": 0.0926,
+      "step": 1152
+    },
+    {
+      "epoch": 0.7916237555784414,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 8.693637197831269e-06,
+      "loss": 0.0924,
+      "step": 1153
+    },
+    {
+      "epoch": 0.7923103329900446,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 8.691309297654637e-06,
+      "loss": 0.0786,
+      "step": 1154
+    },
+    {
+      "epoch": 0.7929969104016478,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 8.688979678061523e-06,
+      "loss": 0.0853,
+      "step": 1155
+    },
+    {
+      "epoch": 0.7936834878132509,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 8.686648340307081e-06,
+      "loss": 0.081,
+      "step": 1156
+    },
+    {
+      "epoch": 0.7943700652248541,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 8.684315285647394e-06,
+      "loss": 0.0794,
+      "step": 1157
+    },
+    {
+      "epoch": 0.7950566426364573,
+      "grad_norm": 0.25,
+      "learning_rate": 8.681980515339464e-06,
+      "loss": 0.0835,
+      "step": 1158
+    },
+    {
+      "epoch": 0.7957432200480604,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 8.679644030641223e-06,
+      "loss": 0.0765,
+      "step": 1159
+    },
+    {
+      "epoch": 0.7964297974596636,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 8.677305832811524e-06,
+      "loss": 0.0821,
+      "step": 1160
+    },
+    {
+      "epoch": 0.7971163748712667,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 8.674965923110143e-06,
+      "loss": 0.0872,
+      "step": 1161
+    },
+    {
+      "epoch": 0.7978029522828699,
+      "grad_norm": 0.201171875,
+      "learning_rate": 8.672624302797778e-06,
+      "loss": 0.0831,
+      "step": 1162
+    },
+    {
+      "epoch": 0.7984895296944731,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 8.670280973136049e-06,
+      "loss": 0.0774,
+      "step": 1163
+    },
+    {
+      "epoch": 0.7991761071060762,
+      "grad_norm": 0.220703125,
+      "learning_rate": 8.667935935387497e-06,
+      "loss": 0.0823,
+      "step": 1164
+    },
+    {
+      "epoch": 0.7998626845176794,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 8.665589190815588e-06,
+      "loss": 0.0877,
+      "step": 1165
+    },
+    {
+      "epoch": 0.8005492619292826,
+      "grad_norm": 0.205078125,
+      "learning_rate": 8.663240740684698e-06,
+      "loss": 0.0922,
+      "step": 1166
+    },
+    {
+      "epoch": 0.8012358393408857,
+      "grad_norm": 0.193359375,
+      "learning_rate": 8.660890586260127e-06,
+      "loss": 0.0987,
+      "step": 1167
+    },
+    {
+      "epoch": 0.8019224167524889,
+      "grad_norm": 0.189453125,
+      "learning_rate": 8.658538728808097e-06,
+      "loss": 0.0974,
+      "step": 1168
+    },
+    {
+      "epoch": 0.802608994164092,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 8.656185169595743e-06,
+      "loss": 0.0737,
+      "step": 1169
+    },
+    {
+      "epoch": 0.8032955715756952,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 8.653829909891114e-06,
+      "loss": 0.0912,
+      "step": 1170
+    },
+    {
+      "epoch": 0.8039821489872984,
+      "grad_norm": 0.224609375,
+      "learning_rate": 8.651472950963184e-06,
+      "loss": 0.085,
+      "step": 1171
+    },
+    {
+      "epoch": 0.8046687263989015,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 8.649114294081833e-06,
+      "loss": 0.0859,
+      "step": 1172
+    },
+    {
+      "epoch": 0.8053553038105047,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 8.646753940517863e-06,
+      "loss": 0.079,
+      "step": 1173
+    },
+    {
+      "epoch": 0.8060418812221078,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 8.644391891542987e-06,
+      "loss": 0.0956,
+      "step": 1174
+    },
+    {
+      "epoch": 0.8067284586337109,
+      "grad_norm": 0.1953125,
+      "learning_rate": 8.64202814842983e-06,
+      "loss": 0.0794,
+      "step": 1175
+    },
+    {
+      "epoch": 0.8074150360453141,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 8.639662712451935e-06,
+      "loss": 0.0854,
+      "step": 1176
+    },
+    {
+      "epoch": 0.8081016134569172,
+      "grad_norm": 0.205078125,
+      "learning_rate": 8.637295584883752e-06,
+      "loss": 0.0766,
+      "step": 1177
+    },
+    {
+      "epoch": 0.8087881908685204,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 8.634926767000642e-06,
+      "loss": 0.0768,
+      "step": 1178
+    },
+    {
+      "epoch": 0.8094747682801235,
+      "grad_norm": 0.224609375,
+      "learning_rate": 8.632556260078882e-06,
+      "loss": 0.0916,
+      "step": 1179
+    },
+    {
+      "epoch": 0.8101613456917267,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 8.630184065395655e-06,
+      "loss": 0.0817,
+      "step": 1180
+    },
+    {
+      "epoch": 0.8108479231033299,
+      "grad_norm": 0.22265625,
+      "learning_rate": 8.627810184229054e-06,
+      "loss": 0.0804,
+      "step": 1181
+    },
+    {
+      "epoch": 0.811534500514933,
+      "grad_norm": 0.201171875,
+      "learning_rate": 8.62543461785808e-06,
+      "loss": 0.0836,
+      "step": 1182
+    },
+    {
+      "epoch": 0.8122210779265362,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 8.623057367562646e-06,
+      "loss": 0.0819,
+      "step": 1183
+    },
+    {
+      "epoch": 0.8129076553381394,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 8.620678434623563e-06,
+      "loss": 0.0911,
+      "step": 1184
+    },
+    {
+      "epoch": 0.8135942327497425,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 8.618297820322558e-06,
+      "loss": 0.088,
+      "step": 1185
+    },
+    {
+      "epoch": 0.8142808101613457,
+      "grad_norm": 0.21484375,
+      "learning_rate": 8.615915525942263e-06,
+      "loss": 0.0855,
+      "step": 1186
+    },
+    {
+      "epoch": 0.8149673875729488,
+      "grad_norm": 0.1953125,
+      "learning_rate": 8.61353155276621e-06,
+      "loss": 0.0911,
+      "step": 1187
+    },
+    {
+      "epoch": 0.815653964984552,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 8.611145902078838e-06,
+      "loss": 0.0819,
+      "step": 1188
+    },
+    {
+      "epoch": 0.8163405423961552,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 8.608758575165491e-06,
+      "loss": 0.0844,
+      "step": 1189
+    },
+    {
+      "epoch": 0.8170271198077583,
+      "grad_norm": 0.25,
+      "learning_rate": 8.606369573312414e-06,
+      "loss": 0.0965,
+      "step": 1190
+    },
+    {
+      "epoch": 0.8177136972193615,
+      "grad_norm": 0.197265625,
+      "learning_rate": 8.603978897806756e-06,
+      "loss": 0.082,
+      "step": 1191
+    },
+    {
+      "epoch": 0.8184002746309647,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 8.601586549936567e-06,
+      "loss": 0.0912,
+      "step": 1192
+    },
+    {
+      "epoch": 0.8190868520425678,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 8.5991925309908e-06,
+      "loss": 0.0802,
+      "step": 1193
+    },
+    {
+      "epoch": 0.819773429454171,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 8.596796842259306e-06,
+      "loss": 0.0891,
+      "step": 1194
+    },
+    {
+      "epoch": 0.8204600068657741,
+      "grad_norm": 0.240234375,
+      "learning_rate": 8.594399485032834e-06,
+      "loss": 0.094,
+      "step": 1195
+    },
+    {
+      "epoch": 0.8211465842773773,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 8.592000460603039e-06,
+      "loss": 0.0885,
+      "step": 1196
+    },
+    {
+      "epoch": 0.8218331616889805,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 8.589599770262464e-06,
+      "loss": 0.0911,
+      "step": 1197
+    },
+    {
+      "epoch": 0.8225197391005836,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 8.587197415304558e-06,
+      "loss": 0.0861,
+      "step": 1198
+    },
+    {
+      "epoch": 0.8232063165121868,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 8.584793397023664e-06,
+      "loss": 0.0881,
+      "step": 1199
+    },
+    {
+      "epoch": 0.82389289392379,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 8.582387716715021e-06,
+      "loss": 0.0896,
+      "step": 1200
+    },
+    {
+      "epoch": 0.8245794713353931,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 8.579980375674763e-06,
+      "loss": 0.097,
+      "step": 1201
+    },
+    {
+      "epoch": 0.8252660487469963,
+      "grad_norm": 0.228515625,
+      "learning_rate": 8.57757137519992e-06,
+      "loss": 0.0891,
+      "step": 1202
+    },
+    {
+      "epoch": 0.8259526261585994,
+      "grad_norm": 0.228515625,
+      "learning_rate": 8.575160716588416e-06,
+      "loss": 0.0873,
+      "step": 1203
+    },
+    {
+      "epoch": 0.8266392035702025,
+      "grad_norm": 0.208984375,
+      "learning_rate": 8.572748401139066e-06,
+      "loss": 0.085,
+      "step": 1204
+    },
+    {
+      "epoch": 0.8273257809818056,
+      "grad_norm": 0.2421875,
+      "learning_rate": 8.57033443015158e-06,
+      "loss": 0.0954,
+      "step": 1205
+    },
+    {
+      "epoch": 0.8280123583934088,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 8.567918804926562e-06,
+      "loss": 0.0856,
+      "step": 1206
+    },
+    {
+      "epoch": 0.828698935805012,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 8.565501526765501e-06,
+      "loss": 0.0943,
+      "step": 1207
+    },
+    {
+      "epoch": 0.8293855132166151,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 8.563082596970785e-06,
+      "loss": 0.0903,
+      "step": 1208
+    },
+    {
+      "epoch": 0.8300720906282183,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 8.56066201684568e-06,
+      "loss": 0.0828,
+      "step": 1209
+    },
+    {
+      "epoch": 0.8307586680398215,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 8.558239787694353e-06,
+      "loss": 0.091,
+      "step": 1210
+    },
+    {
+      "epoch": 0.8314452454514246,
+      "grad_norm": 0.275390625,
+      "learning_rate": 8.555815910821856e-06,
+      "loss": 0.0885,
+      "step": 1211
+    },
+    {
+      "epoch": 0.8321318228630278,
+      "grad_norm": 0.2265625,
+      "learning_rate": 8.553390387534126e-06,
+      "loss": 0.0855,
+      "step": 1212
+    },
+    {
+      "epoch": 0.8328184002746309,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 8.550963219137986e-06,
+      "loss": 0.0944,
+      "step": 1213
+    },
+    {
+      "epoch": 0.8335049776862341,
+      "grad_norm": 0.263671875,
+      "learning_rate": 8.548534406941153e-06,
+      "loss": 0.1078,
+      "step": 1214
+    },
+    {
+      "epoch": 0.8341915550978373,
+      "grad_norm": 0.220703125,
+      "learning_rate": 8.54610395225222e-06,
+      "loss": 0.0991,
+      "step": 1215
+    },
+    {
+      "epoch": 0.8348781325094404,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 8.543671856380672e-06,
+      "loss": 0.0888,
+      "step": 1216
+    },
+    {
+      "epoch": 0.8355647099210436,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 8.541238120636876e-06,
+      "loss": 0.0818,
+      "step": 1217
+    },
+    {
+      "epoch": 0.8362512873326468,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 8.53880274633208e-06,
+      "loss": 0.0792,
+      "step": 1218
+    },
+    {
+      "epoch": 0.8369378647442499,
+      "grad_norm": 0.251953125,
+      "learning_rate": 8.53636573477842e-06,
+      "loss": 0.0923,
+      "step": 1219
+    },
+    {
+      "epoch": 0.8376244421558531,
+      "grad_norm": 0.23828125,
+      "learning_rate": 8.53392708728891e-06,
+      "loss": 0.0921,
+      "step": 1220
+    },
+    {
+      "epoch": 0.8383110195674562,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 8.531486805177446e-06,
+      "loss": 0.0929,
+      "step": 1221
+    },
+    {
+      "epoch": 0.8389975969790594,
+      "grad_norm": 0.228515625,
+      "learning_rate": 8.529044889758808e-06,
+      "loss": 0.0776,
+      "step": 1222
+    },
+    {
+      "epoch": 0.8396841743906626,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 8.526601342348649e-06,
+      "loss": 0.088,
+      "step": 1223
+    },
+    {
+      "epoch": 0.8403707518022657,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 8.524156164263509e-06,
+      "loss": 0.0817,
+      "step": 1224
+    },
+    {
+      "epoch": 0.8410573292138689,
+      "grad_norm": 0.22265625,
+      "learning_rate": 8.521709356820803e-06,
+      "loss": 0.0796,
+      "step": 1225
+    },
+    {
+      "epoch": 0.841743906625472,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 8.519260921338823e-06,
+      "loss": 0.093,
+      "step": 1226
+    },
+    {
+      "epoch": 0.8424304840370752,
+      "grad_norm": 0.21875,
+      "learning_rate": 8.51681085913674e-06,
+      "loss": 0.0872,
+      "step": 1227
+    },
+    {
+      "epoch": 0.8431170614486784,
+      "grad_norm": 0.244140625,
+      "learning_rate": 8.514359171534603e-06,
+      "loss": 0.0947,
+      "step": 1228
+    },
+    {
+      "epoch": 0.8438036388602815,
+      "grad_norm": 0.220703125,
+      "learning_rate": 8.511905859853329e-06,
+      "loss": 0.0876,
+      "step": 1229
+    },
+    {
+      "epoch": 0.8444902162718847,
+      "grad_norm": 0.26171875,
+      "learning_rate": 8.509450925414717e-06,
+      "loss": 0.0796,
+      "step": 1230
+    },
+    {
+      "epoch": 0.8451767936834879,
+      "grad_norm": 0.232421875,
+      "learning_rate": 8.506994369541442e-06,
+      "loss": 0.0896,
+      "step": 1231
+    },
+    {
+      "epoch": 0.845863371095091,
+      "grad_norm": 0.375,
+      "learning_rate": 8.504536193557049e-06,
+      "loss": 0.0962,
+      "step": 1232
+    },
+    {
+      "epoch": 0.8465499485066942,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 8.502076398785953e-06,
+      "loss": 0.0783,
+      "step": 1233
+    },
+    {
+      "epoch": 0.8472365259182972,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 8.499614986553445e-06,
+      "loss": 0.0686,
+      "step": 1234
+    },
+    {
+      "epoch": 0.8479231033299004,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 8.49715195818569e-06,
+      "loss": 0.0763,
+      "step": 1235
+    },
+    {
+      "epoch": 0.8486096807415036,
+      "grad_norm": 0.2734375,
+      "learning_rate": 8.494687315009718e-06,
+      "loss": 0.0925,
+      "step": 1236
+    },
+    {
+      "epoch": 0.8492962581531067,
+      "grad_norm": 0.23828125,
+      "learning_rate": 8.492221058353432e-06,
+      "loss": 0.0831,
+      "step": 1237
+    },
+    {
+      "epoch": 0.8499828355647099,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 8.489753189545605e-06,
+      "loss": 0.0905,
+      "step": 1238
+    },
+    {
+      "epoch": 0.850669412976313,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 8.487283709915877e-06,
+      "loss": 0.0917,
+      "step": 1239
+    },
+    {
+      "epoch": 0.8513559903879162,
+      "grad_norm": 0.2109375,
+      "learning_rate": 8.484812620794757e-06,
+      "loss": 0.0836,
+      "step": 1240
+    },
+    {
+      "epoch": 0.8520425677995194,
+      "grad_norm": 0.259765625,
+      "learning_rate": 8.482339923513619e-06,
+      "loss": 0.0815,
+      "step": 1241
+    },
+    {
+      "epoch": 0.8527291452111225,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 8.479865619404705e-06,
+      "loss": 0.0972,
+      "step": 1242
+    },
+    {
+      "epoch": 0.8534157226227257,
+      "grad_norm": 0.263671875,
+      "learning_rate": 8.477389709801126e-06,
+      "loss": 0.0976,
+      "step": 1243
+    },
+    {
+      "epoch": 0.8541023000343289,
+      "grad_norm": 0.2265625,
+      "learning_rate": 8.474912196036852e-06,
+      "loss": 0.1028,
+      "step": 1244
+    },
+    {
+      "epoch": 0.854788877445932,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 8.472433079446719e-06,
+      "loss": 0.0971,
+      "step": 1245
+    },
+    {
+      "epoch": 0.8554754548575352,
+      "grad_norm": 0.291015625,
+      "learning_rate": 8.469952361366431e-06,
+      "loss": 0.0868,
+      "step": 1246
+    },
+    {
+      "epoch": 0.8561620322691383,
+      "grad_norm": 0.26953125,
+      "learning_rate": 8.46747004313255e-06,
+      "loss": 0.0848,
+      "step": 1247
+    },
+    {
+      "epoch": 0.8568486096807415,
+      "grad_norm": 0.201171875,
+      "learning_rate": 8.4649861260825e-06,
+      "loss": 0.0914,
+      "step": 1248
+    },
+    {
+      "epoch": 0.8575351870923447,
+      "grad_norm": 0.26171875,
+      "learning_rate": 8.462500611554572e-06,
+      "loss": 0.0839,
+      "step": 1249
+    },
+    {
+      "epoch": 0.8582217645039478,
+      "grad_norm": 0.251953125,
+      "learning_rate": 8.46001350088791e-06,
+      "loss": 0.0799,
+      "step": 1250
+    },
+    {
+      "epoch": 0.858908341915551,
+      "grad_norm": 0.1953125,
+      "learning_rate": 8.457524795422522e-06,
+      "loss": 0.0926,
+      "step": 1251
+    },
+    {
+      "epoch": 0.8595949193271541,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 8.455034496499276e-06,
+      "loss": 0.0863,
+      "step": 1252
+    },
+    {
+      "epoch": 0.8602814967387573,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 8.452542605459898e-06,
+      "loss": 0.0863,
+      "step": 1253
+    },
+    {
+      "epoch": 0.8609680741503605,
+      "grad_norm": 0.244140625,
+      "learning_rate": 8.450049123646972e-06,
+      "loss": 0.0909,
+      "step": 1254
+    },
+    {
+      "epoch": 0.8616546515619636,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 8.447554052403937e-06,
+      "loss": 0.0741,
+      "step": 1255
+    },
+    {
+      "epoch": 0.8623412289735668,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 8.445057393075088e-06,
+      "loss": 0.0835,
+      "step": 1256
+    },
+    {
+      "epoch": 0.86302780638517,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 8.44255914700558e-06,
+      "loss": 0.0779,
+      "step": 1257
+    },
+    {
+      "epoch": 0.8637143837967731,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 8.440059315541418e-06,
+      "loss": 0.0789,
+      "step": 1258
+    },
+    {
+      "epoch": 0.8644009612083763,
+      "grad_norm": 0.205078125,
+      "learning_rate": 8.437557900029467e-06,
+      "loss": 0.0867,
+      "step": 1259
+    },
+    {
+      "epoch": 0.8650875386199794,
+      "grad_norm": 0.19140625,
+      "learning_rate": 8.435054901817438e-06,
+      "loss": 0.079,
+      "step": 1260
+    },
+    {
+      "epoch": 0.8657741160315826,
+      "grad_norm": 0.20703125,
+      "learning_rate": 8.432550322253901e-06,
+      "loss": 0.0923,
+      "step": 1261
+    },
+    {
+      "epoch": 0.8664606934431858,
+      "grad_norm": 0.1953125,
+      "learning_rate": 8.430044162688277e-06,
+      "loss": 0.085,
+      "step": 1262
+    },
+    {
+      "epoch": 0.8671472708547889,
+      "grad_norm": 0.20703125,
+      "learning_rate": 8.427536424470833e-06,
+      "loss": 0.0876,
+      "step": 1263
+    },
+    {
+      "epoch": 0.867833848266392,
+      "grad_norm": 0.248046875,
+      "learning_rate": 8.425027108952693e-06,
+      "loss": 0.0863,
+      "step": 1264
+    },
+    {
+      "epoch": 0.8685204256779951,
+      "grad_norm": 0.2109375,
+      "learning_rate": 8.422516217485826e-06,
+      "loss": 0.0877,
+      "step": 1265
+    },
+    {
+      "epoch": 0.8692070030895983,
+      "grad_norm": 0.2578125,
+      "learning_rate": 8.420003751423055e-06,
+      "loss": 0.0899,
+      "step": 1266
+    },
+    {
+      "epoch": 0.8698935805012015,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 8.41748971211805e-06,
+      "loss": 0.0826,
+      "step": 1267
+    },
+    {
+      "epoch": 0.8705801579128046,
+      "grad_norm": 0.201171875,
+      "learning_rate": 8.414974100925321e-06,
+      "loss": 0.0849,
+      "step": 1268
+    },
+    {
+      "epoch": 0.8712667353244078,
+      "grad_norm": 0.251953125,
+      "learning_rate": 8.41245691920024e-06,
+      "loss": 0.094,
+      "step": 1269
+    },
+    {
+      "epoch": 0.871953312736011,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 8.409938168299007e-06,
+      "loss": 0.0919,
+      "step": 1270
+    },
+    {
+      "epoch": 0.8726398901476141,
+      "grad_norm": 0.2109375,
+      "learning_rate": 8.407417849578683e-06,
+      "loss": 0.0793,
+      "step": 1271
+    },
+    {
+      "epoch": 0.8733264675592173,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 8.404895964397166e-06,
+      "loss": 0.0948,
+      "step": 1272
+    },
+    {
+      "epoch": 0.8740130449708204,
+      "grad_norm": 0.240234375,
+      "learning_rate": 8.402372514113201e-06,
+      "loss": 0.0842,
+      "step": 1273
+    },
+    {
+      "epoch": 0.8746996223824236,
+      "grad_norm": 0.216796875,
+      "learning_rate": 8.39984750008637e-06,
+      "loss": 0.0898,
+      "step": 1274
+    },
+    {
+      "epoch": 0.8753861997940268,
+      "grad_norm": 0.2421875,
+      "learning_rate": 8.39732092367711e-06,
+      "loss": 0.0929,
+      "step": 1275
+    },
+    {
+      "epoch": 0.8760727772056299,
+      "grad_norm": 0.240234375,
+      "learning_rate": 8.394792786246686e-06,
+      "loss": 0.084,
+      "step": 1276
+    },
+    {
+      "epoch": 0.8767593546172331,
+      "grad_norm": 0.25,
+      "learning_rate": 8.392263089157212e-06,
+      "loss": 0.0832,
+      "step": 1277
+    },
+    {
+      "epoch": 0.8774459320288363,
+      "grad_norm": 0.234375,
+      "learning_rate": 8.389731833771642e-06,
+      "loss": 0.0805,
+      "step": 1278
+    },
+    {
+      "epoch": 0.8781325094404394,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 8.387199021453765e-06,
+      "loss": 0.0859,
+      "step": 1279
+    },
+    {
+      "epoch": 0.8788190868520426,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 8.384664653568213e-06,
+      "loss": 0.0963,
+      "step": 1280
+    },
+    {
+      "epoch": 0.8795056642636457,
+      "grad_norm": 0.23828125,
+      "learning_rate": 8.382128731480456e-06,
+      "loss": 0.0729,
+      "step": 1281
+    },
+    {
+      "epoch": 0.8801922416752489,
+      "grad_norm": 0.212890625,
+      "learning_rate": 8.379591256556802e-06,
+      "loss": 0.101,
+      "step": 1282
+    },
+    {
+      "epoch": 0.8808788190868521,
+      "grad_norm": 0.234375,
+      "learning_rate": 8.377052230164388e-06,
+      "loss": 0.0912,
+      "step": 1283
+    },
+    {
+      "epoch": 0.8815653964984552,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 8.3745116536712e-06,
+      "loss": 0.0836,
+      "step": 1284
+    },
+    {
+      "epoch": 0.8822519739100584,
+      "grad_norm": 0.25,
+      "learning_rate": 8.371969528446047e-06,
+      "loss": 0.0926,
+      "step": 1285
+    },
+    {
+      "epoch": 0.8829385513216615,
+      "grad_norm": 0.2578125,
+      "learning_rate": 8.36942585585858e-06,
+      "loss": 0.0964,
+      "step": 1286
+    },
+    {
+      "epoch": 0.8836251287332647,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 8.366880637279281e-06,
+      "loss": 0.0835,
+      "step": 1287
+    },
+    {
+      "epoch": 0.8843117061448679,
+      "grad_norm": 0.232421875,
+      "learning_rate": 8.364333874079462e-06,
+      "loss": 0.0936,
+      "step": 1288
+    },
+    {
+      "epoch": 0.884998283556471,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 8.361785567631275e-06,
+      "loss": 0.0805,
+      "step": 1289
+    },
+    {
+      "epoch": 0.8856848609680742,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 8.359235719307697e-06,
+      "loss": 0.0894,
+      "step": 1290
+    },
+    {
+      "epoch": 0.8863714383796774,
+      "grad_norm": 0.21484375,
+      "learning_rate": 8.356684330482537e-06,
+      "loss": 0.086,
+      "step": 1291
+    },
+    {
+      "epoch": 0.8870580157912805,
+      "grad_norm": 0.263671875,
+      "learning_rate": 8.354131402530436e-06,
+      "loss": 0.0891,
+      "step": 1292
+    },
+    {
+      "epoch": 0.8877445932028837,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 8.351576936826862e-06,
+      "loss": 0.0845,
+      "step": 1293
+    },
+    {
+      "epoch": 0.8884311706144867,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 8.34902093474811e-06,
+      "loss": 0.0733,
+      "step": 1294
+    },
+    {
+      "epoch": 0.8891177480260899,
+      "grad_norm": 0.228515625,
+      "learning_rate": 8.346463397671309e-06,
+      "loss": 0.0919,
+      "step": 1295
+    },
+    {
+      "epoch": 0.8898043254376931,
+      "grad_norm": 0.21484375,
+      "learning_rate": 8.343904326974409e-06,
+      "loss": 0.1012,
+      "step": 1296
+    },
+    {
+      "epoch": 0.8904909028492962,
+      "grad_norm": 0.21484375,
+      "learning_rate": 8.341343724036188e-06,
+      "loss": 0.0751,
+      "step": 1297
+    },
+    {
+      "epoch": 0.8911774802608994,
+      "grad_norm": 0.208984375,
+      "learning_rate": 8.33878159023625e-06,
+      "loss": 0.0799,
+      "step": 1298
+    },
+    {
+      "epoch": 0.8918640576725025,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 8.336217926955027e-06,
+      "loss": 0.083,
+      "step": 1299
+    },
+    {
+      "epoch": 0.8925506350841057,
+      "grad_norm": 0.201171875,
+      "learning_rate": 8.333652735573767e-06,
+      "loss": 0.0912,
+      "step": 1300
+    },
+    {
+      "epoch": 0.8932372124957089,
+      "grad_norm": 0.248046875,
+      "learning_rate": 8.33108601747455e-06,
+      "loss": 0.0945,
+      "step": 1301
+    },
+    {
+      "epoch": 0.893923789907312,
+      "grad_norm": 0.236328125,
+      "learning_rate": 8.32851777404027e-06,
+      "loss": 0.0884,
+      "step": 1302
+    },
+    {
+      "epoch": 0.8946103673189152,
+      "grad_norm": 0.224609375,
+      "learning_rate": 8.325948006654652e-06,
+      "loss": 0.0912,
+      "step": 1303
+    },
+    {
+      "epoch": 0.8952969447305184,
+      "grad_norm": 0.212890625,
+      "learning_rate": 8.323376716702236e-06,
+      "loss": 0.0815,
+      "step": 1304
+    },
+    {
+      "epoch": 0.8959835221421215,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 8.320803905568385e-06,
+      "loss": 0.0856,
+      "step": 1305
+    },
+    {
+      "epoch": 0.8966700995537247,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 8.31822957463928e-06,
+      "loss": 0.0749,
+      "step": 1306
+    },
+    {
+      "epoch": 0.8973566769653278,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 8.31565372530192e-06,
+      "loss": 0.0828,
+      "step": 1307
+    },
+    {
+      "epoch": 0.898043254376931,
+      "grad_norm": 0.265625,
+      "learning_rate": 8.313076358944122e-06,
+      "loss": 0.0831,
+      "step": 1308
+    },
+    {
+      "epoch": 0.8987298317885342,
+      "grad_norm": 0.25390625,
+      "learning_rate": 8.310497476954527e-06,
+      "loss": 0.0975,
+      "step": 1309
+    },
+    {
+      "epoch": 0.8994164092001373,
+      "grad_norm": 0.212890625,
+      "learning_rate": 8.307917080722582e-06,
+      "loss": 0.0881,
+      "step": 1310
+    },
+    {
+      "epoch": 0.9001029866117405,
+      "grad_norm": 0.21875,
+      "learning_rate": 8.30533517163856e-06,
+      "loss": 0.0917,
+      "step": 1311
+    },
+    {
+      "epoch": 0.9007895640233436,
+      "grad_norm": 0.197265625,
+      "learning_rate": 8.302751751093539e-06,
+      "loss": 0.0849,
+      "step": 1312
+    },
+    {
+      "epoch": 0.9014761414349468,
+      "grad_norm": 0.21484375,
+      "learning_rate": 8.30016682047942e-06,
+      "loss": 0.0869,
+      "step": 1313
+    },
+    {
+      "epoch": 0.90216271884655,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 8.297580381188916e-06,
+      "loss": 0.0776,
+      "step": 1314
+    },
+    {
+      "epoch": 0.9028492962581531,
+      "grad_norm": 0.248046875,
+      "learning_rate": 8.294992434615547e-06,
+      "loss": 0.0856,
+      "step": 1315
+    },
+    {
+      "epoch": 0.9035358736697563,
+      "grad_norm": 0.21875,
+      "learning_rate": 8.29240298215365e-06,
+      "loss": 0.0882,
+      "step": 1316
+    },
+    {
+      "epoch": 0.9042224510813595,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 8.289812025198376e-06,
+      "loss": 0.078,
+      "step": 1317
+    },
+    {
+      "epoch": 0.9049090284929626,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 8.287219565145676e-06,
+      "loss": 0.0817,
+      "step": 1318
+    },
+    {
+      "epoch": 0.9055956059045658,
+      "grad_norm": 0.201171875,
+      "learning_rate": 8.284625603392326e-06,
+      "loss": 0.0931,
+      "step": 1319
+    },
+    {
+      "epoch": 0.9062821833161689,
+      "grad_norm": 0.28515625,
+      "learning_rate": 8.282030141335899e-06,
+      "loss": 0.0794,
+      "step": 1320
+    },
+    {
+      "epoch": 0.9069687607277721,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 8.279433180374781e-06,
+      "loss": 0.0836,
+      "step": 1321
+    },
+    {
+      "epoch": 0.9076553381393753,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 8.276834721908164e-06,
+      "loss": 0.0848,
+      "step": 1322
+    },
+    {
+      "epoch": 0.9083419155509783,
+      "grad_norm": 0.255859375,
+      "learning_rate": 8.274234767336048e-06,
+      "loss": 0.0888,
+      "step": 1323
+    },
+    {
+      "epoch": 0.9090284929625815,
+      "grad_norm": 0.30078125,
+      "learning_rate": 8.271633318059241e-06,
+      "loss": 0.0898,
+      "step": 1324
+    },
+    {
+      "epoch": 0.9097150703741846,
+      "grad_norm": 0.205078125,
+      "learning_rate": 8.269030375479352e-06,
+      "loss": 0.0869,
+      "step": 1325
+    },
+    {
+      "epoch": 0.9104016477857878,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 8.266425940998797e-06,
+      "loss": 0.0872,
+      "step": 1326
+    },
+    {
+      "epoch": 0.911088225197391,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 8.263820016020798e-06,
+      "loss": 0.0893,
+      "step": 1327
+    },
+    {
+      "epoch": 0.9117748026089941,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 8.261212601949374e-06,
+      "loss": 0.0878,
+      "step": 1328
+    },
+    {
+      "epoch": 0.9124613800205973,
+      "grad_norm": 0.21875,
+      "learning_rate": 8.258603700189351e-06,
+      "loss": 0.0835,
+      "step": 1329
+    },
+    {
+      "epoch": 0.9131479574322005,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 8.255993312146355e-06,
+      "loss": 0.0942,
+      "step": 1330
+    },
+    {
+      "epoch": 0.9138345348438036,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 8.253381439226815e-06,
+      "loss": 0.0823,
+      "step": 1331
+    },
+    {
+      "epoch": 0.9145211122554068,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 8.250768082837957e-06,
+      "loss": 0.0941,
+      "step": 1332
+    },
+    {
+      "epoch": 0.9152076896670099,
+      "grad_norm": 0.236328125,
+      "learning_rate": 8.24815324438781e-06,
+      "loss": 0.0885,
+      "step": 1333
+    },
+    {
+      "epoch": 0.9158942670786131,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 8.245536925285193e-06,
+      "loss": 0.0933,
+      "step": 1334
+    },
+    {
+      "epoch": 0.9165808444902163,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 8.242919126939734e-06,
+      "loss": 0.0949,
+      "step": 1335
+    },
+    {
+      "epoch": 0.9172674219018194,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 8.240299850761851e-06,
+      "loss": 0.0913,
+      "step": 1336
+    },
+    {
+      "epoch": 0.9179539993134226,
+      "grad_norm": 0.26953125,
+      "learning_rate": 8.23767909816276e-06,
+      "loss": 0.0825,
+      "step": 1337
+    },
+    {
+      "epoch": 0.9186405767250257,
+      "grad_norm": 0.224609375,
+      "learning_rate": 8.235056870554476e-06,
+      "loss": 0.085,
+      "step": 1338
+    },
+    {
+      "epoch": 0.9193271541366289,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 8.232433169349799e-06,
+      "loss": 0.083,
+      "step": 1339
+    },
+    {
+      "epoch": 0.9200137315482321,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 8.229807995962332e-06,
+      "loss": 0.0843,
+      "step": 1340
+    },
+    {
+      "epoch": 0.9207003089598352,
+      "grad_norm": 0.314453125,
+      "learning_rate": 8.227181351806467e-06,
+      "loss": 0.0923,
+      "step": 1341
+    },
+    {
+      "epoch": 0.9213868863714384,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 8.224553238297392e-06,
+      "loss": 0.0853,
+      "step": 1342
+    },
+    {
+      "epoch": 0.9220734637830416,
+      "grad_norm": 0.236328125,
+      "learning_rate": 8.221923656851084e-06,
+      "loss": 0.081,
+      "step": 1343
+    },
+    {
+      "epoch": 0.9227600411946447,
+      "grad_norm": 0.3046875,
+      "learning_rate": 8.219292608884309e-06,
+      "loss": 0.0847,
+      "step": 1344
+    },
+    {
+      "epoch": 0.9234466186062479,
+      "grad_norm": 0.21875,
+      "learning_rate": 8.216660095814627e-06,
+      "loss": 0.0892,
+      "step": 1345
+    },
+    {
+      "epoch": 0.924133196017851,
+      "grad_norm": 0.25390625,
+      "learning_rate": 8.214026119060385e-06,
+      "loss": 0.0821,
+      "step": 1346
+    },
+    {
+      "epoch": 0.9248197734294542,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 8.211390680040722e-06,
+      "loss": 0.0885,
+      "step": 1347
+    },
+    {
+      "epoch": 0.9255063508410574,
+      "grad_norm": 0.283203125,
+      "learning_rate": 8.208753780175557e-06,
+      "loss": 0.0937,
+      "step": 1348
+    },
+    {
+      "epoch": 0.9261929282526605,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 8.206115420885608e-06,
+      "loss": 0.083,
+      "step": 1349
+    },
+    {
+      "epoch": 0.9268795056642637,
+      "grad_norm": 0.23046875,
+      "learning_rate": 8.203475603592365e-06,
+      "loss": 0.0743,
+      "step": 1350
+    },
+    {
+      "epoch": 0.9275660830758669,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 8.200834329718118e-06,
+      "loss": 0.0808,
+      "step": 1351
+    },
+    {
+      "epoch": 0.92825266048747,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 8.198191600685931e-06,
+      "loss": 0.0837,
+      "step": 1352
+    },
+    {
+      "epoch": 0.9289392378990731,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 8.195547417919656e-06,
+      "loss": 0.0826,
+      "step": 1353
+    },
+    {
+      "epoch": 0.9296258153106762,
+      "grad_norm": 0.203125,
+      "learning_rate": 8.19290178284393e-06,
+      "loss": 0.082,
+      "step": 1354
+    },
+    {
+      "epoch": 0.9303123927222794,
+      "grad_norm": 0.24609375,
+      "learning_rate": 8.190254696884171e-06,
+      "loss": 0.0868,
+      "step": 1355
+    },
+    {
+      "epoch": 0.9309989701338826,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 8.187606161466575e-06,
+      "loss": 0.0799,
+      "step": 1356
+    },
+    {
+      "epoch": 0.9316855475454857,
+      "grad_norm": 0.220703125,
+      "learning_rate": 8.184956178018124e-06,
+      "loss": 0.0796,
+      "step": 1357
+    },
+    {
+      "epoch": 0.9323721249570889,
+      "grad_norm": 0.216796875,
+      "learning_rate": 8.182304747966581e-06,
+      "loss": 0.0803,
+      "step": 1358
+    },
+    {
+      "epoch": 0.933058702368692,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 8.179651872740484e-06,
+      "loss": 0.0881,
+      "step": 1359
+    },
+    {
+      "epoch": 0.9337452797802952,
+      "grad_norm": 0.19921875,
+      "learning_rate": 8.176997553769146e-06,
+      "loss": 0.085,
+      "step": 1360
+    },
+    {
+      "epoch": 0.9344318571918984,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 8.174341792482673e-06,
+      "loss": 0.0747,
+      "step": 1361
+    },
+    {
+      "epoch": 0.9351184346035015,
+      "grad_norm": 0.341796875,
+      "learning_rate": 8.171684590311933e-06,
+      "loss": 0.0925,
+      "step": 1362
+    },
+    {
+      "epoch": 0.9358050120151047,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 8.169025948688576e-06,
+      "loss": 0.0881,
+      "step": 1363
+    },
+    {
+      "epoch": 0.9364915894267078,
+      "grad_norm": 0.236328125,
+      "learning_rate": 8.166365869045029e-06,
+      "loss": 0.0829,
+      "step": 1364
+    },
+    {
+      "epoch": 0.937178166838311,
+      "grad_norm": 0.25390625,
+      "learning_rate": 8.16370435281449e-06,
+      "loss": 0.085,
+      "step": 1365
+    },
+    {
+      "epoch": 0.9378647442499142,
+      "grad_norm": 0.251953125,
+      "learning_rate": 8.161041401430934e-06,
+      "loss": 0.1009,
+      "step": 1366
+    },
+    {
+      "epoch": 0.9385513216615173,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 8.158377016329111e-06,
+      "loss": 0.0864,
+      "step": 1367
+    },
+    {
+      "epoch": 0.9392378990731205,
+      "grad_norm": 0.203125,
+      "learning_rate": 8.155711198944536e-06,
+      "loss": 0.0798,
+      "step": 1368
+    },
+    {
+      "epoch": 0.9399244764847237,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 8.153043950713504e-06,
+      "loss": 0.0901,
+      "step": 1369
+    },
+    {
+      "epoch": 0.9406110538963268,
+      "grad_norm": 0.21875,
+      "learning_rate": 8.150375273073074e-06,
+      "loss": 0.0887,
+      "step": 1370
+    },
+    {
+      "epoch": 0.94129763130793,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 8.147705167461084e-06,
+      "loss": 0.0809,
+      "step": 1371
+    },
+    {
+      "epoch": 0.9419842087195331,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 8.14503363531613e-06,
+      "loss": 0.0786,
+      "step": 1372
+    },
+    {
+      "epoch": 0.9426707861311363,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 8.142360678077585e-06,
+      "loss": 0.0934,
+      "step": 1373
+    },
+    {
+      "epoch": 0.9433573635427395,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 8.139686297185588e-06,
+      "loss": 0.0809,
+      "step": 1374
+    },
+    {
+      "epoch": 0.9440439409543426,
+      "grad_norm": 0.23046875,
+      "learning_rate": 8.137010494081044e-06,
+      "loss": 0.0931,
+      "step": 1375
+    },
+    {
+      "epoch": 0.9447305183659458,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 8.134333270205624e-06,
+      "loss": 0.0831,
+      "step": 1376
+    },
+    {
+      "epoch": 0.945417095777549,
+      "grad_norm": 0.1875,
+      "learning_rate": 8.131654627001766e-06,
+      "loss": 0.0865,
+      "step": 1377
+    },
+    {
+      "epoch": 0.9461036731891521,
+      "grad_norm": 0.267578125,
+      "learning_rate": 8.128974565912672e-06,
+      "loss": 0.0909,
+      "step": 1378
+    },
+    {
+      "epoch": 0.9467902506007553,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 8.126293088382306e-06,
+      "loss": 0.0805,
+      "step": 1379
+    },
+    {
+      "epoch": 0.9474768280123584,
+      "grad_norm": 0.255859375,
+      "learning_rate": 8.123610195855395e-06,
+      "loss": 0.0761,
+      "step": 1380
+    },
+    {
+      "epoch": 0.9481634054239616,
+      "grad_norm": 0.208984375,
+      "learning_rate": 8.120925889777437e-06,
+      "loss": 0.0899,
+      "step": 1381
+    },
+    {
+      "epoch": 0.9488499828355648,
+      "grad_norm": 0.25390625,
+      "learning_rate": 8.118240171594678e-06,
+      "loss": 0.0794,
+      "step": 1382
+    },
+    {
+      "epoch": 0.9495365602471678,
+      "grad_norm": 0.185546875,
+      "learning_rate": 8.115553042754136e-06,
+      "loss": 0.0864,
+      "step": 1383
+    },
+    {
+      "epoch": 0.950223137658771,
+      "grad_norm": 0.21484375,
+      "learning_rate": 8.112864504703582e-06,
+      "loss": 0.0894,
+      "step": 1384
+    },
+    {
+      "epoch": 0.9509097150703741,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 8.110174558891552e-06,
+      "loss": 0.0976,
+      "step": 1385
+    },
+    {
+      "epoch": 0.9515962924819773,
+      "grad_norm": 0.203125,
+      "learning_rate": 8.107483206767334e-06,
+      "loss": 0.0905,
+      "step": 1386
+    },
+    {
+      "epoch": 0.9522828698935805,
+      "grad_norm": 0.259765625,
+      "learning_rate": 8.10479044978098e-06,
+      "loss": 0.0882,
+      "step": 1387
+    },
+    {
+      "epoch": 0.9529694473051836,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 8.102096289383293e-06,
+      "loss": 0.0816,
+      "step": 1388
+    },
+    {
+      "epoch": 0.9536560247167868,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 8.099400727025837e-06,
+      "loss": 0.0911,
+      "step": 1389
+    },
+    {
+      "epoch": 0.95434260212839,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 8.096703764160929e-06,
+      "loss": 0.0931,
+      "step": 1390
+    },
+    {
+      "epoch": 0.9550291795399931,
+      "grad_norm": 0.212890625,
+      "learning_rate": 8.094005402241643e-06,
+      "loss": 0.0832,
+      "step": 1391
+    },
+    {
+      "epoch": 0.9557157569515963,
+      "grad_norm": 0.205078125,
+      "learning_rate": 8.0913056427218e-06,
+      "loss": 0.0822,
+      "step": 1392
+    },
+    {
+      "epoch": 0.9564023343631994,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 8.088604487055986e-06,
+      "loss": 0.0884,
+      "step": 1393
+    },
+    {
+      "epoch": 0.9570889117748026,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 8.085901936699526e-06,
+      "loss": 0.0832,
+      "step": 1394
+    },
+    {
+      "epoch": 0.9577754891864058,
+      "grad_norm": 0.25390625,
+      "learning_rate": 8.083197993108505e-06,
+      "loss": 0.095,
+      "step": 1395
+    },
+    {
+      "epoch": 0.9584620665980089,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 8.080492657739754e-06,
+      "loss": 0.0856,
+      "step": 1396
+    },
+    {
+      "epoch": 0.9591486440096121,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 8.07778593205086e-06,
+      "loss": 0.0861,
+      "step": 1397
+    },
+    {
+      "epoch": 0.9598352214212152,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 8.075077817500151e-06,
+      "loss": 0.0936,
+      "step": 1398
+    },
+    {
+      "epoch": 0.9605217988328184,
+      "grad_norm": 0.240234375,
+      "learning_rate": 8.072368315546712e-06,
+      "loss": 0.0866,
+      "step": 1399
+    },
+    {
+      "epoch": 0.9612083762444216,
+      "grad_norm": 0.2421875,
+      "learning_rate": 8.069657427650364e-06,
+      "loss": 0.082,
+      "step": 1400
+    },
+    {
+      "epoch": 0.9618949536560247,
+      "grad_norm": 0.1953125,
+      "learning_rate": 8.06694515527169e-06,
+      "loss": 0.0855,
+      "step": 1401
+    },
+    {
+      "epoch": 0.9625815310676279,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 8.064231499872004e-06,
+      "loss": 0.0789,
+      "step": 1402
+    },
+    {
+      "epoch": 0.9632681084792311,
+      "grad_norm": 0.203125,
+      "learning_rate": 8.061516462913376e-06,
+      "loss": 0.0833,
+      "step": 1403
+    },
+    {
+      "epoch": 0.9639546858908342,
+      "grad_norm": 0.21484375,
+      "learning_rate": 8.058800045858613e-06,
+      "loss": 0.0872,
+      "step": 1404
+    },
+    {
+      "epoch": 0.9646412633024374,
+      "grad_norm": 0.21484375,
+      "learning_rate": 8.056082250171267e-06,
+      "loss": 0.0829,
+      "step": 1405
+    },
+    {
+      "epoch": 0.9653278407140405,
+      "grad_norm": 0.2578125,
+      "learning_rate": 8.053363077315641e-06,
+      "loss": 0.0822,
+      "step": 1406
+    },
+    {
+      "epoch": 0.9660144181256437,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 8.050642528756769e-06,
+      "loss": 0.0896,
+      "step": 1407
+    },
+    {
+      "epoch": 0.9667009955372469,
+      "grad_norm": 0.212890625,
+      "learning_rate": 8.047920605960428e-06,
+      "loss": 0.0902,
+      "step": 1408
+    },
+    {
+      "epoch": 0.96738757294885,
+      "grad_norm": 0.220703125,
+      "learning_rate": 8.045197310393142e-06,
+      "loss": 0.0937,
+      "step": 1409
+    },
+    {
+      "epoch": 0.9680741503604532,
+      "grad_norm": 0.244140625,
+      "learning_rate": 8.042472643522172e-06,
+      "loss": 0.0883,
+      "step": 1410
+    },
+    {
+      "epoch": 0.9687607277720564,
+      "grad_norm": 0.267578125,
+      "learning_rate": 8.03974660681551e-06,
+      "loss": 0.0891,
+      "step": 1411
+    },
+    {
+      "epoch": 0.9694473051836594,
+      "grad_norm": 0.240234375,
+      "learning_rate": 8.037019201741901e-06,
+      "loss": 0.0954,
+      "step": 1412
+    },
+    {
+      "epoch": 0.9701338825952626,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 8.034290429770812e-06,
+      "loss": 0.0892,
+      "step": 1413
+    },
+    {
+      "epoch": 0.9708204600068657,
+      "grad_norm": 0.228515625,
+      "learning_rate": 8.031560292372455e-06,
+      "loss": 0.0799,
+      "step": 1414
+    },
+    {
+      "epoch": 0.9715070374184689,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 8.028828791017778e-06,
+      "loss": 0.0816,
+      "step": 1415
+    },
+    {
+      "epoch": 0.972193614830072,
+      "grad_norm": 0.21484375,
+      "learning_rate": 8.026095927178458e-06,
+      "loss": 0.0836,
+      "step": 1416
+    },
+    {
+      "epoch": 0.9728801922416752,
+      "grad_norm": 0.205078125,
+      "learning_rate": 8.023361702326914e-06,
+      "loss": 0.0846,
+      "step": 1417
+    },
+    {
+      "epoch": 0.9735667696532784,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 8.02062611793629e-06,
+      "loss": 0.0868,
+      "step": 1418
+    },
+    {
+      "epoch": 0.9742533470648815,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 8.017889175480473e-06,
+      "loss": 0.0917,
+      "step": 1419
+    },
+    {
+      "epoch": 0.9749399244764847,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 8.015150876434066e-06,
+      "loss": 0.0809,
+      "step": 1420
+    },
+    {
+      "epoch": 0.9756265018880879,
+      "grad_norm": 0.23046875,
+      "learning_rate": 8.012411222272421e-06,
+      "loss": 0.0935,
+      "step": 1421
+    },
+    {
+      "epoch": 0.976313079299691,
+      "grad_norm": 0.234375,
+      "learning_rate": 8.009670214471605e-06,
+      "loss": 0.08,
+      "step": 1422
+    },
+    {
+      "epoch": 0.9769996567112942,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 8.006927854508425e-06,
+      "loss": 0.0747,
+      "step": 1423
+    },
+    {
+      "epoch": 0.9776862341228973,
+      "grad_norm": 0.21484375,
+      "learning_rate": 8.004184143860408e-06,
+      "loss": 0.0826,
+      "step": 1424
+    },
+    {
+      "epoch": 0.9783728115345005,
+      "grad_norm": 0.25390625,
+      "learning_rate": 8.001439084005817e-06,
+      "loss": 0.0861,
+      "step": 1425
+    },
+    {
+      "epoch": 0.9790593889461037,
+      "grad_norm": 0.22265625,
+      "learning_rate": 7.998692676423637e-06,
+      "loss": 0.076,
+      "step": 1426
+    },
+    {
+      "epoch": 0.9797459663577068,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 7.995944922593578e-06,
+      "loss": 0.0958,
+      "step": 1427
+    },
+    {
+      "epoch": 0.98043254376931,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 7.993195823996077e-06,
+      "loss": 0.0851,
+      "step": 1428
+    },
+    {
+      "epoch": 0.9811191211809132,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 7.990445382112299e-06,
+      "loss": 0.0878,
+      "step": 1429
+    },
+    {
+      "epoch": 0.9818056985925163,
+      "grad_norm": 0.21484375,
+      "learning_rate": 7.987693598424126e-06,
+      "loss": 0.0941,
+      "step": 1430
+    },
+    {
+      "epoch": 0.9824922760041195,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 7.984940474414168e-06,
+      "loss": 0.0838,
+      "step": 1431
+    },
+    {
+      "epoch": 0.9831788534157226,
+      "grad_norm": 0.22265625,
+      "learning_rate": 7.982186011565755e-06,
+      "loss": 0.0779,
+      "step": 1432
+    },
+    {
+      "epoch": 0.9838654308273258,
+      "grad_norm": 0.220703125,
+      "learning_rate": 7.979430211362942e-06,
+      "loss": 0.0899,
+      "step": 1433
+    },
+    {
+      "epoch": 0.984552008238929,
+      "grad_norm": 0.201171875,
+      "learning_rate": 7.9766730752905e-06,
+      "loss": 0.0848,
+      "step": 1434
+    },
+    {
+      "epoch": 0.9852385856505321,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 7.973914604833917e-06,
+      "loss": 0.0867,
+      "step": 1435
+    },
+    {
+      "epoch": 0.9859251630621353,
+      "grad_norm": 0.34765625,
+      "learning_rate": 7.971154801479409e-06,
+      "loss": 0.1027,
+      "step": 1436
+    },
+    {
+      "epoch": 0.9866117404737385,
+      "grad_norm": 0.310546875,
+      "learning_rate": 7.968393666713904e-06,
+      "loss": 0.0847,
+      "step": 1437
+    },
+    {
+      "epoch": 0.9872983178853416,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 7.965631202025048e-06,
+      "loss": 0.0777,
+      "step": 1438
+    },
+    {
+      "epoch": 0.9879848952969448,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 7.962867408901206e-06,
+      "loss": 0.0854,
+      "step": 1439
+    },
+    {
+      "epoch": 0.9886714727085479,
+      "grad_norm": 0.251953125,
+      "learning_rate": 7.960102288831454e-06,
+      "loss": 0.0912,
+      "step": 1440
+    },
+    {
+      "epoch": 0.9893580501201511,
+      "grad_norm": 0.1875,
+      "learning_rate": 7.957335843305589e-06,
+      "loss": 0.0747,
+      "step": 1441
+    },
+    {
+      "epoch": 0.9900446275317542,
+      "grad_norm": 0.2265625,
+      "learning_rate": 7.954568073814118e-06,
+      "loss": 0.0909,
+      "step": 1442
+    },
+    {
+      "epoch": 0.9907312049433573,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 7.951798981848261e-06,
+      "loss": 0.0807,
+      "step": 1443
+    },
+    {
+      "epoch": 0.9914177823549605,
+      "grad_norm": 0.22265625,
+      "learning_rate": 7.949028568899951e-06,
+      "loss": 0.0823,
+      "step": 1444
+    },
+    {
+      "epoch": 0.9921043597665636,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 7.946256836461837e-06,
+      "loss": 0.1013,
+      "step": 1445
+    },
+    {
+      "epoch": 0.9927909371781668,
+      "grad_norm": 0.28125,
+      "learning_rate": 7.943483786027276e-06,
+      "loss": 0.0912,
+      "step": 1446
+    },
+    {
+      "epoch": 0.99347751458977,
+      "grad_norm": 0.2421875,
+      "learning_rate": 7.940709419090328e-06,
+      "loss": 0.0827,
+      "step": 1447
+    },
+    {
+      "epoch": 0.9941640920013731,
+      "grad_norm": 0.24609375,
+      "learning_rate": 7.937933737145777e-06,
+      "loss": 0.0799,
+      "step": 1448
+    },
+    {
+      "epoch": 0.9948506694129763,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 7.935156741689102e-06,
+      "loss": 0.0897,
+      "step": 1449
+    },
+    {
+      "epoch": 0.9955372468245794,
+      "grad_norm": 0.228515625,
+      "learning_rate": 7.932378434216497e-06,
+      "loss": 0.0927,
+      "step": 1450
+    },
+    {
+      "epoch": 0.9962238242361826,
+      "grad_norm": 0.2265625,
+      "learning_rate": 7.92959881622486e-06,
+      "loss": 0.0832,
+      "step": 1451
+    },
+    {
+      "epoch": 0.9969104016477858,
+      "grad_norm": 0.19921875,
+      "learning_rate": 7.926817889211798e-06,
+      "loss": 0.0822,
+      "step": 1452
+    },
+    {
+      "epoch": 0.9975969790593889,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 7.924035654675619e-06,
+      "loss": 0.0772,
+      "step": 1453
+    },
+    {
+      "epoch": 0.9982835564709921,
+      "grad_norm": 0.20703125,
+      "learning_rate": 7.92125211411534e-06,
+      "loss": 0.0755,
+      "step": 1454
+    },
+    {
+      "epoch": 0.9989701338825953,
+      "grad_norm": 0.28515625,
+      "learning_rate": 7.918467269030678e-06,
+      "loss": 0.088,
+      "step": 1455
+    },
+    {
+      "epoch": 0.9996567112941984,
+      "grad_norm": 0.20703125,
+      "learning_rate": 7.915681120922055e-06,
+      "loss": 0.081,
+      "step": 1456
+    },
+    {
+      "epoch": 1.0003432887058015,
+      "grad_norm": 0.275390625,
+      "learning_rate": 7.912893671290595e-06,
+      "loss": 0.1102,
+      "step": 1457
+    },
+    {
+      "epoch": 1.0010298661174046,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 7.91010492163812e-06,
+      "loss": 0.0832,
+      "step": 1458
+    },
+    {
+      "epoch": 1.0017164435290078,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 7.907314873467158e-06,
+      "loss": 0.093,
+      "step": 1459
+    },
+    {
+      "epoch": 1.002403020940611,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 7.904523528280932e-06,
+      "loss": 0.0997,
+      "step": 1460
+    },
+    {
+      "epoch": 1.0030895983522141,
+      "grad_norm": 0.158203125,
+      "learning_rate": 7.901730887583367e-06,
+      "loss": 0.0825,
+      "step": 1461
+    },
+    {
+      "epoch": 1.0037761757638173,
+      "grad_norm": 0.17578125,
+      "learning_rate": 7.898936952879082e-06,
+      "loss": 0.0868,
+      "step": 1462
+    },
+    {
+      "epoch": 1.0044627531754204,
+      "grad_norm": 0.1640625,
+      "learning_rate": 7.896141725673398e-06,
+      "loss": 0.0849,
+      "step": 1463
+    },
+    {
+      "epoch": 1.0051493305870236,
+      "grad_norm": 0.189453125,
+      "learning_rate": 7.893345207472329e-06,
+      "loss": 0.0811,
+      "step": 1464
+    },
+    {
+      "epoch": 1.0058359079986268,
+      "grad_norm": 0.177734375,
+      "learning_rate": 7.890547399782584e-06,
+      "loss": 0.0933,
+      "step": 1465
+    },
+    {
+      "epoch": 1.00652248541023,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 7.887748304111573e-06,
+      "loss": 0.0811,
+      "step": 1466
+    },
+    {
+      "epoch": 1.007209062821833,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 7.88494792196739e-06,
+      "loss": 0.0888,
+      "step": 1467
+    },
+    {
+      "epoch": 1.0078956402334363,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 7.882146254858832e-06,
+      "loss": 0.0859,
+      "step": 1468
+    },
+    {
+      "epoch": 1.0085822176450394,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 7.879343304295379e-06,
+      "loss": 0.0823,
+      "step": 1469
+    },
+    {
+      "epoch": 1.0092687950566426,
+      "grad_norm": 0.181640625,
+      "learning_rate": 7.87653907178721e-06,
+      "loss": 0.0847,
+      "step": 1470
+    },
+    {
+      "epoch": 1.0099553724682457,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 7.87373355884519e-06,
+      "loss": 0.0864,
+      "step": 1471
+    },
+    {
+      "epoch": 1.010641949879849,
+      "grad_norm": 0.18359375,
+      "learning_rate": 7.870926766980879e-06,
+      "loss": 0.0819,
+      "step": 1472
+    },
+    {
+      "epoch": 1.011328527291452,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 7.86811869770652e-06,
+      "loss": 0.0785,
+      "step": 1473
+    },
+    {
+      "epoch": 1.0120151047030552,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 7.86530935253505e-06,
+      "loss": 0.0792,
+      "step": 1474
+    },
+    {
+      "epoch": 1.0127016821146584,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 7.862498732980089e-06,
+      "loss": 0.0866,
+      "step": 1475
+    },
+    {
+      "epoch": 1.0133882595262615,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 7.859686840555944e-06,
+      "loss": 0.0827,
+      "step": 1476
+    },
+    {
+      "epoch": 1.0140748369378647,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 7.856873676777607e-06,
+      "loss": 0.0852,
+      "step": 1477
+    },
+    {
+      "epoch": 1.0147614143494679,
+      "grad_norm": 0.1640625,
+      "learning_rate": 7.854059243160763e-06,
+      "loss": 0.077,
+      "step": 1478
+    },
+    {
+      "epoch": 1.015447991761071,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 7.85124354122177e-06,
+      "loss": 0.0958,
+      "step": 1479
+    },
+    {
+      "epoch": 1.0161345691726742,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 7.848426572477677e-06,
+      "loss": 0.0831,
+      "step": 1480
+    },
+    {
+      "epoch": 1.0168211465842774,
+      "grad_norm": 0.1796875,
+      "learning_rate": 7.84560833844621e-06,
+      "loss": 0.0822,
+      "step": 1481
+    },
+    {
+      "epoch": 1.0175077239958805,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 7.842788840645785e-06,
+      "loss": 0.0863,
+      "step": 1482
+    },
+    {
+      "epoch": 1.0181943014074837,
+      "grad_norm": 0.181640625,
+      "learning_rate": 7.839968080595488e-06,
+      "loss": 0.0854,
+      "step": 1483
+    },
+    {
+      "epoch": 1.0188808788190868,
+      "grad_norm": 0.16796875,
+      "learning_rate": 7.837146059815096e-06,
+      "loss": 0.0882,
+      "step": 1484
+    },
+    {
+      "epoch": 1.01956745623069,
+      "grad_norm": 0.171875,
+      "learning_rate": 7.834322779825053e-06,
+      "loss": 0.0915,
+      "step": 1485
+    },
+    {
+      "epoch": 1.0202540336422932,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 7.831498242146495e-06,
+      "loss": 0.0914,
+      "step": 1486
+    },
+    {
+      "epoch": 1.0209406110538963,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 7.828672448301224e-06,
+      "loss": 0.09,
+      "step": 1487
+    },
+    {
+      "epoch": 1.0216271884654995,
+      "grad_norm": 0.17578125,
+      "learning_rate": 7.825845399811723e-06,
+      "loss": 0.0849,
+      "step": 1488
+    },
+    {
+      "epoch": 1.0223137658771027,
+      "grad_norm": 0.1640625,
+      "learning_rate": 7.823017098201159e-06,
+      "loss": 0.0821,
+      "step": 1489
+    },
+    {
+      "epoch": 1.0230003432887058,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 7.820187544993361e-06,
+      "loss": 0.0866,
+      "step": 1490
+    },
+    {
+      "epoch": 1.023686920700309,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 7.817356741712839e-06,
+      "loss": 0.0782,
+      "step": 1491
+    },
+    {
+      "epoch": 1.0243734981119121,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 7.814524689884776e-06,
+      "loss": 0.0777,
+      "step": 1492
+    },
+    {
+      "epoch": 1.0250600755235153,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 7.811691391035029e-06,
+      "loss": 0.0801,
+      "step": 1493
+    },
+    {
+      "epoch": 1.0257466529351185,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 7.808856846690125e-06,
+      "loss": 0.0832,
+      "step": 1494
+    },
+    {
+      "epoch": 1.0264332303467216,
+      "grad_norm": 0.1953125,
+      "learning_rate": 7.806021058377259e-06,
+      "loss": 0.0889,
+      "step": 1495
+    },
+    {
+      "epoch": 1.0271198077583248,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 7.8031840276243e-06,
+      "loss": 0.0811,
+      "step": 1496
+    },
+    {
+      "epoch": 1.027806385169928,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 7.800345755959792e-06,
+      "loss": 0.0913,
+      "step": 1497
+    },
+    {
+      "epoch": 1.0284929625815311,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 7.797506244912937e-06,
+      "loss": 0.0795,
+      "step": 1498
+    },
+    {
+      "epoch": 1.0291795399931343,
+      "grad_norm": 0.16796875,
+      "learning_rate": 7.794665496013608e-06,
+      "loss": 0.0811,
+      "step": 1499
+    },
+    {
+      "epoch": 1.0298661174047374,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 7.791823510792348e-06,
+      "loss": 0.0963,
+      "step": 1500
+    },
+    {
+      "epoch": 1.0305526948163406,
+      "grad_norm": 0.1796875,
+      "learning_rate": 7.788980290780364e-06,
+      "loss": 0.0801,
+      "step": 1501
+    },
+    {
+      "epoch": 1.0312392722279438,
+      "grad_norm": 0.17578125,
+      "learning_rate": 7.786135837509531e-06,
+      "loss": 0.08,
+      "step": 1502
+    },
+    {
+      "epoch": 1.031925849639547,
+      "grad_norm": 0.162109375,
+      "learning_rate": 7.783290152512382e-06,
+      "loss": 0.0818,
+      "step": 1503
+    },
+    {
+      "epoch": 1.03261242705115,
+      "grad_norm": 0.162109375,
+      "learning_rate": 7.78044323732212e-06,
+      "loss": 0.0773,
+      "step": 1504
+    },
+    {
+      "epoch": 1.0332990044627532,
+      "grad_norm": 0.169921875,
+      "learning_rate": 7.77759509347261e-06,
+      "loss": 0.0947,
+      "step": 1505
+    },
+    {
+      "epoch": 1.0339855818743564,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 7.774745722498374e-06,
+      "loss": 0.0809,
+      "step": 1506
+    },
+    {
+      "epoch": 1.0346721592859596,
+      "grad_norm": 0.177734375,
+      "learning_rate": 7.771895125934601e-06,
+      "loss": 0.0797,
+      "step": 1507
+    },
+    {
+      "epoch": 1.0353587366975627,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 7.769043305317138e-06,
+      "loss": 0.0875,
+      "step": 1508
+    },
+    {
+      "epoch": 1.036045314109166,
+      "grad_norm": 0.15625,
+      "learning_rate": 7.766190262182487e-06,
+      "loss": 0.0824,
+      "step": 1509
+    },
+    {
+      "epoch": 1.036731891520769,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 7.763335998067822e-06,
+      "loss": 0.0952,
+      "step": 1510
+    },
+    {
+      "epoch": 1.0374184689323722,
+      "grad_norm": 0.177734375,
+      "learning_rate": 7.76048051451096e-06,
+      "loss": 0.0883,
+      "step": 1511
+    },
+    {
+      "epoch": 1.0381050463439754,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 7.75762381305038e-06,
+      "loss": 0.0933,
+      "step": 1512
+    },
+    {
+      "epoch": 1.0387916237555785,
+      "grad_norm": 0.185546875,
+      "learning_rate": 7.75476589522522e-06,
+      "loss": 0.0946,
+      "step": 1513
+    },
+    {
+      "epoch": 1.0394782011671817,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 7.751906762575273e-06,
+      "loss": 0.083,
+      "step": 1514
+    },
+    {
+      "epoch": 1.0401647785787849,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 7.749046416640983e-06,
+      "loss": 0.0756,
+      "step": 1515
+    },
+    {
+      "epoch": 1.0408513559903878,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 7.74618485896345e-06,
+      "loss": 0.0935,
+      "step": 1516
+    },
+    {
+      "epoch": 1.041537933401991,
+      "grad_norm": 0.16015625,
+      "learning_rate": 7.743322091084427e-06,
+      "loss": 0.0803,
+      "step": 1517
+    },
+    {
+      "epoch": 1.0422245108135941,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 7.740458114546319e-06,
+      "loss": 0.0722,
+      "step": 1518
+    },
+    {
+      "epoch": 1.0429110882251973,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 7.737592930892182e-06,
+      "loss": 0.0959,
+      "step": 1519
+    },
+    {
+      "epoch": 1.0435976656368005,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 7.734726541665722e-06,
+      "loss": 0.0879,
+      "step": 1520
+    },
+    {
+      "epoch": 1.0442842430484036,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 7.731858948411293e-06,
+      "loss": 0.0819,
+      "step": 1521
+    },
+    {
+      "epoch": 1.0449708204600068,
+      "grad_norm": 0.169921875,
+      "learning_rate": 7.7289901526739e-06,
+      "loss": 0.0915,
+      "step": 1522
+    },
+    {
+      "epoch": 1.04565739787161,
+      "grad_norm": 0.16015625,
+      "learning_rate": 7.7261201559992e-06,
+      "loss": 0.088,
+      "step": 1523
+    },
+    {
+      "epoch": 1.046343975283213,
+      "grad_norm": 0.1875,
+      "learning_rate": 7.723248959933486e-06,
+      "loss": 0.0818,
+      "step": 1524
+    },
+    {
+      "epoch": 1.0470305526948163,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 7.720376566023708e-06,
+      "loss": 0.0824,
+      "step": 1525
+    },
+    {
+      "epoch": 1.0477171301064194,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 7.717502975817452e-06,
+      "loss": 0.077,
+      "step": 1526
+    },
+    {
+      "epoch": 1.0484037075180226,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 7.71462819086296e-06,
+      "loss": 0.0907,
+      "step": 1527
+    },
+    {
+      "epoch": 1.0490902849296257,
+      "grad_norm": 0.197265625,
+      "learning_rate": 7.711752212709106e-06,
+      "loss": 0.0816,
+      "step": 1528
+    },
+    {
+      "epoch": 1.049776862341229,
+      "grad_norm": 0.162109375,
+      "learning_rate": 7.708875042905414e-06,
+      "loss": 0.0845,
+      "step": 1529
+    },
+    {
+      "epoch": 1.050463439752832,
+      "grad_norm": 0.158203125,
+      "learning_rate": 7.70599668300205e-06,
+      "loss": 0.0875,
+      "step": 1530
+    },
+    {
+      "epoch": 1.0511500171644352,
+      "grad_norm": 0.162109375,
+      "learning_rate": 7.703117134549814e-06,
+      "loss": 0.0844,
+      "step": 1531
+    },
+    {
+      "epoch": 1.0518365945760384,
+      "grad_norm": 0.16796875,
+      "learning_rate": 7.700236399100159e-06,
+      "loss": 0.0874,
+      "step": 1532
+    },
+    {
+      "epoch": 1.0525231719876416,
+      "grad_norm": 0.185546875,
+      "learning_rate": 7.697354478205165e-06,
+      "loss": 0.0978,
+      "step": 1533
+    },
+    {
+      "epoch": 1.0532097493992447,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 7.694471373417556e-06,
+      "loss": 0.0823,
+      "step": 1534
+    },
+    {
+      "epoch": 1.0538963268108479,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 7.691587086290696e-06,
+      "loss": 0.0914,
+      "step": 1535
+    },
+    {
+      "epoch": 1.054582904222451,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 7.688701618378583e-06,
+      "loss": 0.0882,
+      "step": 1536
+    },
+    {
+      "epoch": 1.0552694816340542,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 7.685814971235852e-06,
+      "loss": 0.08,
+      "step": 1537
+    },
+    {
+      "epoch": 1.0559560590456574,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 7.682927146417772e-06,
+      "loss": 0.0839,
+      "step": 1538
+    },
+    {
+      "epoch": 1.0566426364572605,
+      "grad_norm": 0.16796875,
+      "learning_rate": 7.680038145480248e-06,
+      "loss": 0.0747,
+      "step": 1539
+    },
+    {
+      "epoch": 1.0573292138688637,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 7.677147969979824e-06,
+      "loss": 0.0889,
+      "step": 1540
+    },
+    {
+      "epoch": 1.0580157912804669,
+      "grad_norm": 0.181640625,
+      "learning_rate": 7.674256621473664e-06,
+      "loss": 0.0914,
+      "step": 1541
+    },
+    {
+      "epoch": 1.05870236869207,
+      "grad_norm": 0.1640625,
+      "learning_rate": 7.671364101519574e-06,
+      "loss": 0.0863,
+      "step": 1542
+    },
+    {
+      "epoch": 1.0593889461036732,
+      "grad_norm": 0.1875,
+      "learning_rate": 7.668470411675987e-06,
+      "loss": 0.0925,
+      "step": 1543
+    },
+    {
+      "epoch": 1.0600755235152763,
+      "grad_norm": 0.1640625,
+      "learning_rate": 7.665575553501973e-06,
+      "loss": 0.0963,
+      "step": 1544
+    },
+    {
+      "epoch": 1.0607621009268795,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 7.662679528557224e-06,
+      "loss": 0.0845,
+      "step": 1545
+    },
+    {
+      "epoch": 1.0614486783384827,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 7.659782338402058e-06,
+      "loss": 0.0856,
+      "step": 1546
+    },
+    {
+      "epoch": 1.0621352557500858,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 7.65688398459743e-06,
+      "loss": 0.0866,
+      "step": 1547
+    },
+    {
+      "epoch": 1.062821833161689,
+      "grad_norm": 0.1796875,
+      "learning_rate": 7.653984468704918e-06,
+      "loss": 0.0809,
+      "step": 1548
+    },
+    {
+      "epoch": 1.0635084105732922,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 7.651083792286725e-06,
+      "loss": 0.0989,
+      "step": 1549
+    },
+    {
+      "epoch": 1.0641949879848953,
+      "grad_norm": 0.185546875,
+      "learning_rate": 7.648181956905677e-06,
+      "loss": 0.0893,
+      "step": 1550
+    },
+    {
+      "epoch": 1.0648815653964985,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 7.645278964125234e-06,
+      "loss": 0.0857,
+      "step": 1551
+    },
+    {
+      "epoch": 1.0655681428081016,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 7.64237481550947e-06,
+      "loss": 0.0854,
+      "step": 1552
+    },
+    {
+      "epoch": 1.0662547202197048,
+      "grad_norm": 0.162109375,
+      "learning_rate": 7.639469512623081e-06,
+      "loss": 0.079,
+      "step": 1553
+    },
+    {
+      "epoch": 1.066941297631308,
+      "grad_norm": 0.1796875,
+      "learning_rate": 7.636563057031394e-06,
+      "loss": 0.0817,
+      "step": 1554
+    },
+    {
+      "epoch": 1.0676278750429111,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 7.63365545030035e-06,
+      "loss": 0.0904,
+      "step": 1555
+    },
+    {
+      "epoch": 1.0683144524545143,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 7.63074669399651e-06,
+      "loss": 0.0829,
+      "step": 1556
+    },
+    {
+      "epoch": 1.0690010298661174,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 7.627836789687058e-06,
+      "loss": 0.0897,
+      "step": 1557
+    },
+    {
+      "epoch": 1.0696876072777206,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 7.624925738939794e-06,
+      "loss": 0.0746,
+      "step": 1558
+    },
+    {
+      "epoch": 1.0703741846893238,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 7.622013543323138e-06,
+      "loss": 0.0938,
+      "step": 1559
+    },
+    {
+      "epoch": 1.071060762100927,
+      "grad_norm": 0.171875,
+      "learning_rate": 7.619100204406127e-06,
+      "loss": 0.0821,
+      "step": 1560
+    },
+    {
+      "epoch": 1.07174733951253,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 7.616185723758407e-06,
+      "loss": 0.0905,
+      "step": 1561
+    },
+    {
+      "epoch": 1.0724339169241333,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 7.613270102950249e-06,
+      "loss": 0.0845,
+      "step": 1562
+    },
+    {
+      "epoch": 1.0731204943357364,
+      "grad_norm": 0.1640625,
+      "learning_rate": 7.610353343552532e-06,
+      "loss": 0.0777,
+      "step": 1563
+    },
+    {
+      "epoch": 1.0738070717473396,
+      "grad_norm": 0.185546875,
+      "learning_rate": 7.60743544713675e-06,
+      "loss": 0.0815,
+      "step": 1564
+    },
+    {
+      "epoch": 1.0744936491589427,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 7.604516415275011e-06,
+      "loss": 0.0822,
+      "step": 1565
+    },
+    {
+      "epoch": 1.075180226570546,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 7.601596249540033e-06,
+      "loss": 0.0891,
+      "step": 1566
+    },
+    {
+      "epoch": 1.075866803982149,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 7.598674951505145e-06,
+      "loss": 0.0976,
+      "step": 1567
+    },
+    {
+      "epoch": 1.0765533813937522,
+      "grad_norm": 0.18359375,
+      "learning_rate": 7.595752522744287e-06,
+      "loss": 0.0845,
+      "step": 1568
+    },
+    {
+      "epoch": 1.0772399588053554,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 7.5928289648320074e-06,
+      "loss": 0.0843,
+      "step": 1569
+    },
+    {
+      "epoch": 1.0779265362169586,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 7.589904279343465e-06,
+      "loss": 0.0776,
+      "step": 1570
+    },
+    {
+      "epoch": 1.0786131136285617,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 7.586978467854422e-06,
+      "loss": 0.0785,
+      "step": 1571
+    },
+    {
+      "epoch": 1.0792996910401649,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 7.584051531941252e-06,
+      "loss": 0.0792,
+      "step": 1572
+    },
+    {
+      "epoch": 1.079986268451768,
+      "grad_norm": 0.15625,
+      "learning_rate": 7.581123473180932e-06,
+      "loss": 0.0806,
+      "step": 1573
+    },
+    {
+      "epoch": 1.0806728458633712,
+      "grad_norm": 0.1875,
+      "learning_rate": 7.578194293151042e-06,
+      "loss": 0.0938,
+      "step": 1574
+    },
+    {
+      "epoch": 1.0813594232749741,
+      "grad_norm": 0.166015625,
+      "learning_rate": 7.575263993429769e-06,
+      "loss": 0.0986,
+      "step": 1575
+    },
+    {
+      "epoch": 1.0820460006865775,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 7.572332575595904e-06,
+      "loss": 0.0922,
+      "step": 1576
+    },
+    {
+      "epoch": 1.0827325780981805,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 7.569400041228838e-06,
+      "loss": 0.0775,
+      "step": 1577
+    },
+    {
+      "epoch": 1.0834191555097836,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 7.566466391908563e-06,
+      "loss": 0.0796,
+      "step": 1578
+    },
+    {
+      "epoch": 1.0841057329213868,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 7.5635316292156766e-06,
+      "loss": 0.0799,
+      "step": 1579
+    },
+    {
+      "epoch": 1.08479231033299,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 7.5605957547313695e-06,
+      "loss": 0.0906,
+      "step": 1580
+    },
+    {
+      "epoch": 1.0854788877445931,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 7.557658770037437e-06,
+      "loss": 0.0974,
+      "step": 1581
+    },
+    {
+      "epoch": 1.0861654651561963,
+      "grad_norm": 0.166015625,
+      "learning_rate": 7.55472067671627e-06,
+      "loss": 0.0831,
+      "step": 1582
+    },
+    {
+      "epoch": 1.0868520425677994,
+      "grad_norm": 0.15234375,
+      "learning_rate": 7.551781476350854e-06,
+      "loss": 0.0822,
+      "step": 1583
+    },
+    {
+      "epoch": 1.0875386199794026,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 7.548841170524779e-06,
+      "loss": 0.0889,
+      "step": 1584
+    },
+    {
+      "epoch": 1.0882251973910058,
+      "grad_norm": 0.16796875,
+      "learning_rate": 7.545899760822224e-06,
+      "loss": 0.0851,
+      "step": 1585
+    },
+    {
+      "epoch": 1.088911774802609,
+      "grad_norm": 0.17578125,
+      "learning_rate": 7.5429572488279615e-06,
+      "loss": 0.0929,
+      "step": 1586
+    },
+    {
+      "epoch": 1.089598352214212,
+      "grad_norm": 0.166015625,
+      "learning_rate": 7.540013636127364e-06,
+      "loss": 0.0846,
+      "step": 1587
+    },
+    {
+      "epoch": 1.0902849296258152,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 7.537068924306392e-06,
+      "loss": 0.0776,
+      "step": 1588
+    },
+    {
+      "epoch": 1.0909715070374184,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 7.5341231149516e-06,
+      "loss": 0.0828,
+      "step": 1589
+    },
+    {
+      "epoch": 1.0916580844490216,
+      "grad_norm": 0.162109375,
+      "learning_rate": 7.5311762096501355e-06,
+      "loss": 0.0733,
+      "step": 1590
+    },
+    {
+      "epoch": 1.0923446618606247,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 7.528228209989733e-06,
+      "loss": 0.0906,
+      "step": 1591
+    },
+    {
+      "epoch": 1.093031239272228,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 7.525279117558719e-06,
+      "loss": 0.0833,
+      "step": 1592
+    },
+    {
+      "epoch": 1.093717816683831,
+      "grad_norm": 0.17578125,
+      "learning_rate": 7.5223289339460085e-06,
+      "loss": 0.0888,
+      "step": 1593
+    },
+    {
+      "epoch": 1.0944043940954342,
+      "grad_norm": 0.18359375,
+      "learning_rate": 7.519377660741103e-06,
+      "loss": 0.0962,
+      "step": 1594
+    },
+    {
+      "epoch": 1.0950909715070374,
+      "grad_norm": 0.158203125,
+      "learning_rate": 7.516425299534093e-06,
+      "loss": 0.0804,
+      "step": 1595
+    },
+    {
+      "epoch": 1.0957775489186405,
+      "grad_norm": 0.177734375,
+      "learning_rate": 7.513471851915657e-06,
+      "loss": 0.082,
+      "step": 1596
+    },
+    {
+      "epoch": 1.0964641263302437,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 7.510517319477048e-06,
+      "loss": 0.0818,
+      "step": 1597
+    },
+    {
+      "epoch": 1.0971507037418469,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 7.507561703810122e-06,
+      "loss": 0.0794,
+      "step": 1598
+    },
+    {
+      "epoch": 1.09783728115345,
+      "grad_norm": 0.171875,
+      "learning_rate": 7.5046050065073014e-06,
+      "loss": 0.0915,
+      "step": 1599
+    },
+    {
+      "epoch": 1.0985238585650532,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 7.501647229161599e-06,
+      "loss": 0.0775,
+      "step": 1600
+    },
+    {
+      "epoch": 1.0992104359766564,
+      "grad_norm": 0.171875,
+      "learning_rate": 7.49868837336661e-06,
+      "loss": 0.0776,
+      "step": 1601
+    },
+    {
+      "epoch": 1.0998970133882595,
+      "grad_norm": 0.1875,
+      "learning_rate": 7.495728440716511e-06,
+      "loss": 0.0894,
+      "step": 1602
+    },
+    {
+      "epoch": 1.1005835907998627,
+      "grad_norm": 0.15625,
+      "learning_rate": 7.492767432806055e-06,
+      "loss": 0.0893,
+      "step": 1603
+    },
+    {
+      "epoch": 1.1012701682114658,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 7.489805351230575e-06,
+      "loss": 0.0872,
+      "step": 1604
+    },
+    {
+      "epoch": 1.101956745623069,
+      "grad_norm": 0.16796875,
+      "learning_rate": 7.486842197585988e-06,
+      "loss": 0.087,
+      "step": 1605
+    },
+    {
+      "epoch": 1.1026433230346722,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 7.483877973468782e-06,
+      "loss": 0.0903,
+      "step": 1606
+    },
+    {
+      "epoch": 1.1033299004462753,
+      "grad_norm": 0.181640625,
+      "learning_rate": 7.480912680476025e-06,
+      "loss": 0.0863,
+      "step": 1607
+    },
+    {
+      "epoch": 1.1040164778578785,
+      "grad_norm": 0.177734375,
+      "learning_rate": 7.477946320205358e-06,
+      "loss": 0.0873,
+      "step": 1608
+    },
+    {
+      "epoch": 1.1047030552694816,
+      "grad_norm": 0.162109375,
+      "learning_rate": 7.474978894255001e-06,
+      "loss": 0.0782,
+      "step": 1609
+    },
+    {
+      "epoch": 1.1053896326810848,
+      "grad_norm": 0.166015625,
+      "learning_rate": 7.472010404223749e-06,
+      "loss": 0.0778,
+      "step": 1610
+    },
+    {
+      "epoch": 1.106076210092688,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 7.469040851710965e-06,
+      "loss": 0.0801,
+      "step": 1611
+    },
+    {
+      "epoch": 1.1067627875042911,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 7.466070238316584e-06,
+      "loss": 0.0801,
+      "step": 1612
+    },
+    {
+      "epoch": 1.1074493649158943,
+      "grad_norm": 0.177734375,
+      "learning_rate": 7.463098565641121e-06,
+      "loss": 0.0868,
+      "step": 1613
+    },
+    {
+      "epoch": 1.1081359423274975,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 7.4601258352856565e-06,
+      "loss": 0.0859,
+      "step": 1614
+    },
+    {
+      "epoch": 1.1088225197391006,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 7.4571520488518365e-06,
+      "loss": 0.0911,
+      "step": 1615
+    },
+    {
+      "epoch": 1.1095090971507038,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 7.454177207941884e-06,
+      "loss": 0.0784,
+      "step": 1616
+    },
+    {
+      "epoch": 1.110195674562307,
+      "grad_norm": 0.166015625,
+      "learning_rate": 7.451201314158586e-06,
+      "loss": 0.0884,
+      "step": 1617
+    },
+    {
+      "epoch": 1.11088225197391,
+      "grad_norm": 0.1875,
+      "learning_rate": 7.448224369105297e-06,
+      "loss": 0.0937,
+      "step": 1618
+    },
+    {
+      "epoch": 1.1115688293855133,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 7.445246374385935e-06,
+      "loss": 0.075,
+      "step": 1619
+    },
+    {
+      "epoch": 1.1122554067971164,
+      "grad_norm": 0.1875,
+      "learning_rate": 7.44226733160499e-06,
+      "loss": 0.094,
+      "step": 1620
+    },
+    {
+      "epoch": 1.1129419842087196,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 7.439287242367514e-06,
+      "loss": 0.0778,
+      "step": 1621
+    },
+    {
+      "epoch": 1.1136285616203228,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 7.43630610827912e-06,
+      "loss": 0.0891,
+      "step": 1622
+    },
+    {
+      "epoch": 1.114315139031926,
+      "grad_norm": 0.166015625,
+      "learning_rate": 7.433323930945986e-06,
+      "loss": 0.0822,
+      "step": 1623
+    },
+    {
+      "epoch": 1.115001716443529,
+      "grad_norm": 0.181640625,
+      "learning_rate": 7.430340711974855e-06,
+      "loss": 0.0889,
+      "step": 1624
+    },
+    {
+      "epoch": 1.1156882938551322,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 7.427356452973024e-06,
+      "loss": 0.0903,
+      "step": 1625
+    },
+    {
+      "epoch": 1.1163748712667354,
+      "grad_norm": 0.1953125,
+      "learning_rate": 7.424371155548359e-06,
+      "loss": 0.0847,
+      "step": 1626
+    },
+    {
+      "epoch": 1.1170614486783386,
+      "grad_norm": 0.181640625,
+      "learning_rate": 7.421384821309278e-06,
+      "loss": 0.0832,
+      "step": 1627
+    },
+    {
+      "epoch": 1.1177480260899417,
+      "grad_norm": 0.166015625,
+      "learning_rate": 7.418397451864763e-06,
+      "loss": 0.092,
+      "step": 1628
+    },
+    {
+      "epoch": 1.1184346035015449,
+      "grad_norm": 0.181640625,
+      "learning_rate": 7.41540904882435e-06,
+      "loss": 0.09,
+      "step": 1629
+    },
+    {
+      "epoch": 1.119121180913148,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 7.412419613798135e-06,
+      "loss": 0.0882,
+      "step": 1630
+    },
+    {
+      "epoch": 1.1198077583247512,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 7.409429148396767e-06,
+      "loss": 0.084,
+      "step": 1631
+    },
+    {
+      "epoch": 1.1204943357363544,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 7.406437654231453e-06,
+      "loss": 0.0814,
+      "step": 1632
+    },
+    {
+      "epoch": 1.1211809131479575,
+      "grad_norm": 0.162109375,
+      "learning_rate": 7.403445132913951e-06,
+      "loss": 0.0797,
+      "step": 1633
+    },
+    {
+      "epoch": 1.1218674905595605,
+      "grad_norm": 0.166015625,
+      "learning_rate": 7.400451586056578e-06,
+      "loss": 0.0895,
+      "step": 1634
+    },
+    {
+      "epoch": 1.1225540679711639,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 7.397457015272194e-06,
+      "loss": 0.0938,
+      "step": 1635
+    },
+    {
+      "epoch": 1.1232406453827668,
+      "grad_norm": 0.18359375,
+      "learning_rate": 7.394461422174224e-06,
+      "loss": 0.0783,
+      "step": 1636
+    },
+    {
+      "epoch": 1.12392722279437,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 7.391464808376629e-06,
+      "loss": 0.0889,
+      "step": 1637
+    },
+    {
+      "epoch": 1.1246138002059731,
+      "grad_norm": 0.173828125,
+      "learning_rate": 7.388467175493932e-06,
+      "loss": 0.0771,
+      "step": 1638
+    },
+    {
+      "epoch": 1.1253003776175763,
+      "grad_norm": 0.162109375,
+      "learning_rate": 7.385468525141199e-06,
+      "loss": 0.0838,
+      "step": 1639
+    },
+    {
+      "epoch": 1.1259869550291794,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 7.382468858934046e-06,
+      "loss": 0.0843,
+      "step": 1640
+    },
+    {
+      "epoch": 1.1266735324407826,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 7.379468178488633e-06,
+      "loss": 0.0793,
+      "step": 1641
+    },
+    {
+      "epoch": 1.1273601098523858,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 7.376466485421672e-06,
+      "loss": 0.0967,
+      "step": 1642
+    },
+    {
+      "epoch": 1.128046687263989,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 7.373463781350419e-06,
+      "loss": 0.0918,
+      "step": 1643
+    },
+    {
+      "epoch": 1.128733264675592,
+      "grad_norm": 0.193359375,
+      "learning_rate": 7.370460067892672e-06,
+      "loss": 0.0921,
+      "step": 1644
+    },
+    {
+      "epoch": 1.1294198420871953,
+      "grad_norm": 0.15625,
+      "learning_rate": 7.367455346666774e-06,
+      "loss": 0.0854,
+      "step": 1645
+    },
+    {
+      "epoch": 1.1301064194987984,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 7.36444961929161e-06,
+      "loss": 0.0726,
+      "step": 1646
+    },
+    {
+      "epoch": 1.1307929969104016,
+      "grad_norm": 0.177734375,
+      "learning_rate": 7.361442887386613e-06,
+      "loss": 0.0868,
+      "step": 1647
+    },
+    {
+      "epoch": 1.1314795743220047,
+      "grad_norm": 0.171875,
+      "learning_rate": 7.358435152571749e-06,
+      "loss": 0.0818,
+      "step": 1648
+    },
+    {
+      "epoch": 1.132166151733608,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 7.355426416467529e-06,
+      "loss": 0.0851,
+      "step": 1649
+    },
+    {
+      "epoch": 1.132852729145211,
+      "grad_norm": 0.158203125,
+      "learning_rate": 7.352416680695001e-06,
+      "loss": 0.0839,
+      "step": 1650
+    },
+    {
+      "epoch": 1.1335393065568142,
+      "grad_norm": 0.177734375,
+      "learning_rate": 7.349405946875758e-06,
+      "loss": 0.0957,
+      "step": 1651
+    },
+    {
+      "epoch": 1.1342258839684174,
+      "grad_norm": 0.173828125,
+      "learning_rate": 7.34639421663192e-06,
+      "loss": 0.0863,
+      "step": 1652
+    },
+    {
+      "epoch": 1.1349124613800206,
+      "grad_norm": 0.177734375,
+      "learning_rate": 7.343381491586153e-06,
+      "loss": 0.0916,
+      "step": 1653
+    },
+    {
+      "epoch": 1.1355990387916237,
+      "grad_norm": 0.181640625,
+      "learning_rate": 7.340367773361656e-06,
+      "loss": 0.0943,
+      "step": 1654
+    },
+    {
+      "epoch": 1.1362856162032269,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 7.3373530635821614e-06,
+      "loss": 0.0858,
+      "step": 1655
+    },
+    {
+      "epoch": 1.13697219361483,
+      "grad_norm": 0.15625,
+      "learning_rate": 7.334337363871936e-06,
+      "loss": 0.0841,
+      "step": 1656
+    },
+    {
+      "epoch": 1.1376587710264332,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 7.3313206758557855e-06,
+      "loss": 0.0882,
+      "step": 1657
+    },
+    {
+      "epoch": 1.1383453484380364,
+      "grad_norm": 0.166015625,
+      "learning_rate": 7.32830300115904e-06,
+      "loss": 0.0838,
+      "step": 1658
+    },
+    {
+      "epoch": 1.1390319258496395,
+      "grad_norm": 0.16796875,
+      "learning_rate": 7.325284341407566e-06,
+      "loss": 0.078,
+      "step": 1659
+    },
+    {
+      "epoch": 1.1397185032612427,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 7.322264698227759e-06,
+      "loss": 0.0886,
+      "step": 1660
+    },
+    {
+      "epoch": 1.1404050806728459,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 7.319244073246546e-06,
+      "loss": 0.0956,
+      "step": 1661
+    },
+    {
+      "epoch": 1.141091658084449,
+      "grad_norm": 0.171875,
+      "learning_rate": 7.316222468091384e-06,
+      "loss": 0.0997,
+      "step": 1662
+    },
+    {
+      "epoch": 1.1417782354960522,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 7.3131998843902515e-06,
+      "loss": 0.0807,
+      "step": 1663
+    },
+    {
+      "epoch": 1.1424648129076553,
+      "grad_norm": 0.16796875,
+      "learning_rate": 7.310176323771663e-06,
+      "loss": 0.0921,
+      "step": 1664
+    },
+    {
+      "epoch": 1.1431513903192585,
+      "grad_norm": 0.16015625,
+      "learning_rate": 7.307151787864654e-06,
+      "loss": 0.0815,
+      "step": 1665
+    },
+    {
+      "epoch": 1.1438379677308617,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 7.304126278298784e-06,
+      "loss": 0.0841,
+      "step": 1666
+    },
+    {
+      "epoch": 1.1445245451424648,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 7.301099796704144e-06,
+      "loss": 0.0813,
+      "step": 1667
+    },
+    {
+      "epoch": 1.145211122554068,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 7.298072344711342e-06,
+      "loss": 0.0797,
+      "step": 1668
+    },
+    {
+      "epoch": 1.1458976999656711,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 7.295043923951512e-06,
+      "loss": 0.0849,
+      "step": 1669
+    },
+    {
+      "epoch": 1.1465842773772743,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 7.292014536056309e-06,
+      "loss": 0.0856,
+      "step": 1670
+    },
+    {
+      "epoch": 1.1472708547888775,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 7.288984182657908e-06,
+      "loss": 0.0882,
+      "step": 1671
+    },
+    {
+      "epoch": 1.1479574322004806,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 7.285952865389007e-06,
+      "loss": 0.0926,
+      "step": 1672
+    },
+    {
+      "epoch": 1.1486440096120838,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 7.282920585882822e-06,
+      "loss": 0.0858,
+      "step": 1673
+    },
+    {
+      "epoch": 1.149330587023687,
+      "grad_norm": 0.140625,
+      "learning_rate": 7.2798873457730875e-06,
+      "loss": 0.0797,
+      "step": 1674
+    },
+    {
+      "epoch": 1.1500171644352901,
+      "grad_norm": 0.158203125,
+      "learning_rate": 7.276853146694056e-06,
+      "loss": 0.0829,
+      "step": 1675
+    },
+    {
+      "epoch": 1.1507037418468933,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 7.273817990280494e-06,
+      "loss": 0.0867,
+      "step": 1676
+    },
+    {
+      "epoch": 1.1513903192584964,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 7.270781878167684e-06,
+      "loss": 0.0886,
+      "step": 1677
+    },
+    {
+      "epoch": 1.1520768966700996,
+      "grad_norm": 0.185546875,
+      "learning_rate": 7.26774481199143e-06,
+      "loss": 0.0865,
+      "step": 1678
+    },
+    {
+      "epoch": 1.1527634740817028,
+      "grad_norm": 0.177734375,
+      "learning_rate": 7.264706793388043e-06,
+      "loss": 0.0861,
+      "step": 1679
+    },
+    {
+      "epoch": 1.153450051493306,
+      "grad_norm": 0.193359375,
+      "learning_rate": 7.261667823994351e-06,
+      "loss": 0.0909,
+      "step": 1680
+    },
+    {
+      "epoch": 1.154136628904909,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 7.258627905447687e-06,
+      "loss": 0.0902,
+      "step": 1681
+    },
+    {
+      "epoch": 1.1548232063165123,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 7.255587039385909e-06,
+      "loss": 0.0868,
+      "step": 1682
+    },
+    {
+      "epoch": 1.1555097837281154,
+      "grad_norm": 0.189453125,
+      "learning_rate": 7.252545227447371e-06,
+      "loss": 0.0842,
+      "step": 1683
+    },
+    {
+      "epoch": 1.1561963611397186,
+      "grad_norm": 0.18359375,
+      "learning_rate": 7.249502471270947e-06,
+      "loss": 0.0826,
+      "step": 1684
+    },
+    {
+      "epoch": 1.1568829385513217,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 7.246458772496015e-06,
+      "loss": 0.0919,
+      "step": 1685
+    },
+    {
+      "epoch": 1.157569515962925,
+      "grad_norm": 0.1796875,
+      "learning_rate": 7.243414132762461e-06,
+      "loss": 0.0858,
+      "step": 1686
+    },
+    {
+      "epoch": 1.158256093374528,
+      "grad_norm": 0.16796875,
+      "learning_rate": 7.240368553710679e-06,
+      "loss": 0.0935,
+      "step": 1687
+    },
+    {
+      "epoch": 1.1589426707861312,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 7.237322036981568e-06,
+      "loss": 0.0782,
+      "step": 1688
+    },
+    {
+      "epoch": 1.1596292481977344,
+      "grad_norm": 0.181640625,
+      "learning_rate": 7.234274584216534e-06,
+      "loss": 0.0877,
+      "step": 1689
+    },
+    {
+      "epoch": 1.1603158256093375,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 7.2312261970574885e-06,
+      "loss": 0.093,
+      "step": 1690
+    },
+    {
+      "epoch": 1.1610024030209405,
+      "grad_norm": 0.171875,
+      "learning_rate": 7.228176877146841e-06,
+      "loss": 0.0851,
+      "step": 1691
+    },
+    {
+      "epoch": 1.1616889804325439,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 7.225126626127507e-06,
+      "loss": 0.0787,
+      "step": 1692
+    },
+    {
+      "epoch": 1.1623755578441468,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 7.222075445642904e-06,
+      "loss": 0.0934,
+      "step": 1693
+    },
+    {
+      "epoch": 1.1630621352557502,
+      "grad_norm": 0.177734375,
+      "learning_rate": 7.219023337336952e-06,
+      "loss": 0.0743,
+      "step": 1694
+    },
+    {
+      "epoch": 1.1637487126673531,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 7.215970302854065e-06,
+      "loss": 0.0807,
+      "step": 1695
+    },
+    {
+      "epoch": 1.1644352900789565,
+      "grad_norm": 0.169921875,
+      "learning_rate": 7.212916343839163e-06,
+      "loss": 0.0796,
+      "step": 1696
+    },
+    {
+      "epoch": 1.1651218674905595,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 7.209861461937659e-06,
+      "loss": 0.086,
+      "step": 1697
+    },
+    {
+      "epoch": 1.1658084449021628,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 7.206805658795464e-06,
+      "loss": 0.0773,
+      "step": 1698
+    },
+    {
+      "epoch": 1.1664950223137658,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 7.203748936058987e-06,
+      "loss": 0.0787,
+      "step": 1699
+    },
+    {
+      "epoch": 1.167181599725369,
+      "grad_norm": 0.177734375,
+      "learning_rate": 7.200691295375134e-06,
+      "loss": 0.0904,
+      "step": 1700
+    },
+    {
+      "epoch": 1.167868177136972,
+      "grad_norm": 0.1640625,
+      "learning_rate": 7.1976327383913016e-06,
+      "loss": 0.0826,
+      "step": 1701
+    },
+    {
+      "epoch": 1.1685547545485753,
+      "grad_norm": 0.18359375,
+      "learning_rate": 7.194573266755381e-06,
+      "loss": 0.0888,
+      "step": 1702
+    },
+    {
+      "epoch": 1.1692413319601784,
+      "grad_norm": 0.162109375,
+      "learning_rate": 7.1915128821157585e-06,
+      "loss": 0.0764,
+      "step": 1703
+    },
+    {
+      "epoch": 1.1699279093717816,
+      "grad_norm": 0.166015625,
+      "learning_rate": 7.188451586121312e-06,
+      "loss": 0.0858,
+      "step": 1704
+    },
+    {
+      "epoch": 1.1706144867833848,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 7.1853893804214055e-06,
+      "loss": 0.0829,
+      "step": 1705
+    },
+    {
+      "epoch": 1.171301064194988,
+      "grad_norm": 0.177734375,
+      "learning_rate": 7.182326266665901e-06,
+      "loss": 0.0871,
+      "step": 1706
+    },
+    {
+      "epoch": 1.171987641606591,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 7.179262246505142e-06,
+      "loss": 0.0891,
+      "step": 1707
+    },
+    {
+      "epoch": 1.1726742190181942,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 7.176197321589967e-06,
+      "loss": 0.1015,
+      "step": 1708
+    },
+    {
+      "epoch": 1.1733607964297974,
+      "grad_norm": 0.16796875,
+      "learning_rate": 7.173131493571699e-06,
+      "loss": 0.078,
+      "step": 1709
+    },
+    {
+      "epoch": 1.1740473738414006,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 7.1700647641021435e-06,
+      "loss": 0.083,
+      "step": 1710
+    },
+    {
+      "epoch": 1.1747339512530037,
+      "grad_norm": 0.158203125,
+      "learning_rate": 7.166997134833601e-06,
+      "loss": 0.0784,
+      "step": 1711
+    },
+    {
+      "epoch": 1.175420528664607,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 7.163928607418849e-06,
+      "loss": 0.079,
+      "step": 1712
+    },
+    {
+      "epoch": 1.17610710607621,
+      "grad_norm": 0.166015625,
+      "learning_rate": 7.1608591835111515e-06,
+      "loss": 0.08,
+      "step": 1713
+    },
+    {
+      "epoch": 1.1767936834878132,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 7.157788864764255e-06,
+      "loss": 0.0862,
+      "step": 1714
+    },
+    {
+      "epoch": 1.1774802608994164,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 7.154717652832389e-06,
+      "loss": 0.0777,
+      "step": 1715
+    },
+    {
+      "epoch": 1.1781668383110195,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 7.1516455493702654e-06,
+      "loss": 0.0874,
+      "step": 1716
+    },
+    {
+      "epoch": 1.1788534157226227,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 7.148572556033071e-06,
+      "loss": 0.0866,
+      "step": 1717
+    },
+    {
+      "epoch": 1.1795399931342259,
+      "grad_norm": 0.20703125,
+      "learning_rate": 7.145498674476479e-06,
+      "loss": 0.0894,
+      "step": 1718
+    },
+    {
+      "epoch": 1.180226570545829,
+      "grad_norm": 0.16796875,
+      "learning_rate": 7.142423906356641e-06,
+      "loss": 0.0847,
+      "step": 1719
+    },
+    {
+      "epoch": 1.1809131479574322,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 7.139348253330177e-06,
+      "loss": 0.0776,
+      "step": 1720
+    },
+    {
+      "epoch": 1.1815997253690353,
+      "grad_norm": 0.158203125,
+      "learning_rate": 7.136271717054193e-06,
+      "loss": 0.0853,
+      "step": 1721
+    },
+    {
+      "epoch": 1.1822863027806385,
+      "grad_norm": 0.169921875,
+      "learning_rate": 7.133194299186268e-06,
+      "loss": 0.085,
+      "step": 1722
+    },
+    {
+      "epoch": 1.1829728801922417,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 7.130116001384456e-06,
+      "loss": 0.0801,
+      "step": 1723
+    },
+    {
+      "epoch": 1.1836594576038448,
+      "grad_norm": 0.1796875,
+      "learning_rate": 7.1270368253072855e-06,
+      "loss": 0.0856,
+      "step": 1724
+    },
+    {
+      "epoch": 1.184346035015448,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 7.123956772613756e-06,
+      "loss": 0.0956,
+      "step": 1725
+    },
+    {
+      "epoch": 1.1850326124270512,
+      "grad_norm": 0.18359375,
+      "learning_rate": 7.120875844963343e-06,
+      "loss": 0.0903,
+      "step": 1726
+    },
+    {
+      "epoch": 1.1857191898386543,
+      "grad_norm": 0.193359375,
+      "learning_rate": 7.11779404401599e-06,
+      "loss": 0.0885,
+      "step": 1727
+    },
+    {
+      "epoch": 1.1864057672502575,
+      "grad_norm": 0.16015625,
+      "learning_rate": 7.114711371432113e-06,
+      "loss": 0.0781,
+      "step": 1728
+    },
+    {
+      "epoch": 1.1870923446618606,
+      "grad_norm": 0.18359375,
+      "learning_rate": 7.111627828872597e-06,
+      "loss": 0.0911,
+      "step": 1729
+    },
+    {
+      "epoch": 1.1877789220734638,
+      "grad_norm": 0.1796875,
+      "learning_rate": 7.108543417998796e-06,
+      "loss": 0.0828,
+      "step": 1730
+    },
+    {
+      "epoch": 1.188465499485067,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 7.105458140472533e-06,
+      "loss": 0.0859,
+      "step": 1731
+    },
+    {
+      "epoch": 1.1891520768966701,
+      "grad_norm": 0.18359375,
+      "learning_rate": 7.1023719979560934e-06,
+      "loss": 0.0913,
+      "step": 1732
+    },
+    {
+      "epoch": 1.1898386543082733,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 7.099284992112234e-06,
+      "loss": 0.087,
+      "step": 1733
+    },
+    {
+      "epoch": 1.1905252317198765,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 7.096197124604175e-06,
+      "loss": 0.0801,
+      "step": 1734
+    },
+    {
+      "epoch": 1.1912118091314796,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 7.0931083970955984e-06,
+      "loss": 0.0797,
+      "step": 1735
+    },
+    {
+      "epoch": 1.1918983865430828,
+      "grad_norm": 0.193359375,
+      "learning_rate": 7.090018811250653e-06,
+      "loss": 0.1028,
+      "step": 1736
+    },
+    {
+      "epoch": 1.192584963954686,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 7.086928368733949e-06,
+      "loss": 0.0799,
+      "step": 1737
+    },
+    {
+      "epoch": 1.193271541366289,
+      "grad_norm": 0.169921875,
+      "learning_rate": 7.083837071210557e-06,
+      "loss": 0.086,
+      "step": 1738
+    },
+    {
+      "epoch": 1.1939581187778923,
+      "grad_norm": 0.169921875,
+      "learning_rate": 7.08074492034601e-06,
+      "loss": 0.0815,
+      "step": 1739
+    },
+    {
+      "epoch": 1.1946446961894954,
+      "grad_norm": 0.166015625,
+      "learning_rate": 7.0776519178063e-06,
+      "loss": 0.0875,
+      "step": 1740
+    },
+    {
+      "epoch": 1.1953312736010986,
+      "grad_norm": 0.169921875,
+      "learning_rate": 7.0745580652578765e-06,
+      "loss": 0.0891,
+      "step": 1741
+    },
+    {
+      "epoch": 1.1960178510127017,
+      "grad_norm": 0.1640625,
+      "learning_rate": 7.071463364367647e-06,
+      "loss": 0.084,
+      "step": 1742
+    },
+    {
+      "epoch": 1.196704428424305,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 7.068367816802978e-06,
+      "loss": 0.0814,
+      "step": 1743
+    },
+    {
+      "epoch": 1.197391005835908,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 7.065271424231694e-06,
+      "loss": 0.0872,
+      "step": 1744
+    },
+    {
+      "epoch": 1.1980775832475112,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 7.06217418832207e-06,
+      "loss": 0.0785,
+      "step": 1745
+    },
+    {
+      "epoch": 1.1987641606591144,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 7.059076110742836e-06,
+      "loss": 0.0796,
+      "step": 1746
+    },
+    {
+      "epoch": 1.1994507380707176,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 7.0559771931631805e-06,
+      "loss": 0.09,
+      "step": 1747
+    },
+    {
+      "epoch": 1.2001373154823207,
+      "grad_norm": 0.16796875,
+      "learning_rate": 7.052877437252739e-06,
+      "loss": 0.0859,
+      "step": 1748
+    },
+    {
+      "epoch": 1.2008238928939239,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 7.0497768446816e-06,
+      "loss": 0.0819,
+      "step": 1749
+    },
+    {
+      "epoch": 1.201510470305527,
+      "grad_norm": 0.185546875,
+      "learning_rate": 7.046675417120306e-06,
+      "loss": 0.0896,
+      "step": 1750
+    },
+    {
+      "epoch": 1.2021970477171302,
+      "grad_norm": 0.162109375,
+      "learning_rate": 7.043573156239844e-06,
+      "loss": 0.0838,
+      "step": 1751
+    },
+    {
+      "epoch": 1.2028836251287331,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 7.040470063711655e-06,
+      "loss": 0.0812,
+      "step": 1752
+    },
+    {
+      "epoch": 1.2035702025403365,
+      "grad_norm": 0.16015625,
+      "learning_rate": 7.037366141207627e-06,
+      "loss": 0.0854,
+      "step": 1753
+    },
+    {
+      "epoch": 1.2042567799519395,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 7.034261390400091e-06,
+      "loss": 0.0796,
+      "step": 1754
+    },
+    {
+      "epoch": 1.2049433573635429,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 7.0311558129618295e-06,
+      "loss": 0.0859,
+      "step": 1755
+    },
+    {
+      "epoch": 1.2056299347751458,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 7.028049410566068e-06,
+      "loss": 0.0886,
+      "step": 1756
+    },
+    {
+      "epoch": 1.2063165121867492,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 7.024942184886474e-06,
+      "loss": 0.0929,
+      "step": 1757
+    },
+    {
+      "epoch": 1.2070030895983521,
+      "grad_norm": 0.18359375,
+      "learning_rate": 7.0218341375971674e-06,
+      "loss": 0.089,
+      "step": 1758
+    },
+    {
+      "epoch": 1.2076896670099553,
+      "grad_norm": 0.1640625,
+      "learning_rate": 7.0187252703727e-06,
+      "loss": 0.0757,
+      "step": 1759
+    },
+    {
+      "epoch": 1.2083762444215584,
+      "grad_norm": 0.166015625,
+      "learning_rate": 7.015615584888072e-06,
+      "loss": 0.0771,
+      "step": 1760
+    },
+    {
+      "epoch": 1.2090628218331616,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 7.01250508281872e-06,
+      "loss": 0.0903,
+      "step": 1761
+    },
+    {
+      "epoch": 1.2097493992447648,
+      "grad_norm": 0.171875,
+      "learning_rate": 7.009393765840526e-06,
+      "loss": 0.0796,
+      "step": 1762
+    },
+    {
+      "epoch": 1.210435976656368,
+      "grad_norm": 0.16796875,
+      "learning_rate": 7.006281635629808e-06,
+      "loss": 0.0897,
+      "step": 1763
+    },
+    {
+      "epoch": 1.211122554067971,
+      "grad_norm": 0.171875,
+      "learning_rate": 7.003168693863319e-06,
+      "loss": 0.0817,
+      "step": 1764
+    },
+    {
+      "epoch": 1.2118091314795743,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 7.000054942218256e-06,
+      "loss": 0.083,
+      "step": 1765
+    },
+    {
+      "epoch": 1.2124957088911774,
+      "grad_norm": 0.185546875,
+      "learning_rate": 6.996940382372249e-06,
+      "loss": 0.0866,
+      "step": 1766
+    },
+    {
+      "epoch": 1.2131822863027806,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 6.993825016003362e-06,
+      "loss": 0.0914,
+      "step": 1767
+    },
+    {
+      "epoch": 1.2138688637143837,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 6.990708844790093e-06,
+      "loss": 0.086,
+      "step": 1768
+    },
+    {
+      "epoch": 1.214555441125987,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 6.987591870411378e-06,
+      "loss": 0.0855,
+      "step": 1769
+    },
+    {
+      "epoch": 1.21524201853759,
+      "grad_norm": 0.19140625,
+      "learning_rate": 6.984474094546583e-06,
+      "loss": 0.0955,
+      "step": 1770
+    },
+    {
+      "epoch": 1.2159285959491932,
+      "grad_norm": 0.177734375,
+      "learning_rate": 6.981355518875508e-06,
+      "loss": 0.0941,
+      "step": 1771
+    },
+    {
+      "epoch": 1.2166151733607964,
+      "grad_norm": 0.177734375,
+      "learning_rate": 6.9782361450783765e-06,
+      "loss": 0.0825,
+      "step": 1772
+    },
+    {
+      "epoch": 1.2173017507723995,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 6.975115974835852e-06,
+      "loss": 0.0991,
+      "step": 1773
+    },
+    {
+      "epoch": 1.2179883281840027,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 6.971995009829022e-06,
+      "loss": 0.0829,
+      "step": 1774
+    },
+    {
+      "epoch": 1.2186749055956059,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 6.968873251739402e-06,
+      "loss": 0.0843,
+      "step": 1775
+    },
+    {
+      "epoch": 1.219361483007209,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 6.965750702248936e-06,
+      "loss": 0.0826,
+      "step": 1776
+    },
+    {
+      "epoch": 1.2200480604188122,
+      "grad_norm": 0.177734375,
+      "learning_rate": 6.962627363039995e-06,
+      "loss": 0.1041,
+      "step": 1777
+    },
+    {
+      "epoch": 1.2207346378304154,
+      "grad_norm": 0.18359375,
+      "learning_rate": 6.9595032357953716e-06,
+      "loss": 0.0821,
+      "step": 1778
+    },
+    {
+      "epoch": 1.2214212152420185,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 6.956378322198287e-06,
+      "loss": 0.0781,
+      "step": 1779
+    },
+    {
+      "epoch": 1.2221077926536217,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 6.953252623932388e-06,
+      "loss": 0.0839,
+      "step": 1780
+    },
+    {
+      "epoch": 1.2227943700652248,
+      "grad_norm": 0.1796875,
+      "learning_rate": 6.950126142681738e-06,
+      "loss": 0.0862,
+      "step": 1781
+    },
+    {
+      "epoch": 1.223480947476828,
+      "grad_norm": 0.177734375,
+      "learning_rate": 6.946998880130827e-06,
+      "loss": 0.0809,
+      "step": 1782
+    },
+    {
+      "epoch": 1.2241675248884312,
+      "grad_norm": 0.1796875,
+      "learning_rate": 6.943870837964561e-06,
+      "loss": 0.09,
+      "step": 1783
+    },
+    {
+      "epoch": 1.2248541023000343,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 6.940742017868274e-06,
+      "loss": 0.0753,
+      "step": 1784
+    },
+    {
+      "epoch": 1.2255406797116375,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 6.937612421527712e-06,
+      "loss": 0.082,
+      "step": 1785
+    },
+    {
+      "epoch": 1.2262272571232407,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 6.93448205062904e-06,
+      "loss": 0.0838,
+      "step": 1786
+    },
+    {
+      "epoch": 1.2269138345348438,
+      "grad_norm": 0.20703125,
+      "learning_rate": 6.9313509068588455e-06,
+      "loss": 0.0875,
+      "step": 1787
+    },
+    {
+      "epoch": 1.227600411946447,
+      "grad_norm": 0.1796875,
+      "learning_rate": 6.928218991904128e-06,
+      "loss": 0.0813,
+      "step": 1788
+    },
+    {
+      "epoch": 1.2282869893580501,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 6.9250863074523024e-06,
+      "loss": 0.0908,
+      "step": 1789
+    },
+    {
+      "epoch": 1.2289735667696533,
+      "grad_norm": 0.17578125,
+      "learning_rate": 6.921952855191199e-06,
+      "loss": 0.0976,
+      "step": 1790
+    },
+    {
+      "epoch": 1.2296601441812565,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 6.918818636809064e-06,
+      "loss": 0.0804,
+      "step": 1791
+    },
+    {
+      "epoch": 1.2303467215928596,
+      "grad_norm": 0.185546875,
+      "learning_rate": 6.9156836539945535e-06,
+      "loss": 0.0973,
+      "step": 1792
+    },
+    {
+      "epoch": 1.2310332990044628,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 6.912547908436735e-06,
+      "loss": 0.082,
+      "step": 1793
+    },
+    {
+      "epoch": 1.231719876416066,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 6.909411401825088e-06,
+      "loss": 0.0789,
+      "step": 1794
+    },
+    {
+      "epoch": 1.2324064538276691,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 6.906274135849506e-06,
+      "loss": 0.0905,
+      "step": 1795
+    },
+    {
+      "epoch": 1.2330930312392723,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 6.903136112200285e-06,
+      "loss": 0.0833,
+      "step": 1796
+    },
+    {
+      "epoch": 1.2337796086508754,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 6.899997332568132e-06,
+      "loss": 0.0887,
+      "step": 1797
+    },
+    {
+      "epoch": 1.2344661860624786,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 6.896857798644163e-06,
+      "loss": 0.0809,
+      "step": 1798
+    },
+    {
+      "epoch": 1.2351527634740818,
+      "grad_norm": 0.1875,
+      "learning_rate": 6.893717512119898e-06,
+      "loss": 0.0769,
+      "step": 1799
+    },
+    {
+      "epoch": 1.235839340885685,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 6.890576474687264e-06,
+      "loss": 0.0952,
+      "step": 1800
+    },
+    {
+      "epoch": 1.236525918297288,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 6.88743468803859e-06,
+      "loss": 0.0876,
+      "step": 1801
+    },
+    {
+      "epoch": 1.2372124957088912,
+      "grad_norm": 0.173828125,
+      "learning_rate": 6.884292153866616e-06,
+      "loss": 0.0901,
+      "step": 1802
+    },
+    {
+      "epoch": 1.2378990731204944,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 6.881148873864472e-06,
+      "loss": 0.0808,
+      "step": 1803
+    },
+    {
+      "epoch": 1.2385856505320976,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 6.878004849725702e-06,
+      "loss": 0.0805,
+      "step": 1804
+    },
+    {
+      "epoch": 1.2392722279437007,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 6.874860083144245e-06,
+      "loss": 0.0827,
+      "step": 1805
+    },
+    {
+      "epoch": 1.239958805355304,
+      "grad_norm": 0.17578125,
+      "learning_rate": 6.871714575814441e-06,
+      "loss": 0.0823,
+      "step": 1806
+    },
+    {
+      "epoch": 1.240645382766907,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 6.868568329431029e-06,
+      "loss": 0.0906,
+      "step": 1807
+    },
+    {
+      "epoch": 1.2413319601785102,
+      "grad_norm": 0.169921875,
+      "learning_rate": 6.865421345689147e-06,
+      "loss": 0.0767,
+      "step": 1808
+    },
+    {
+      "epoch": 1.2420185375901134,
+      "grad_norm": 0.16796875,
+      "learning_rate": 6.86227362628433e-06,
+      "loss": 0.0845,
+      "step": 1809
+    },
+    {
+      "epoch": 1.2427051150017165,
+      "grad_norm": 0.166015625,
+      "learning_rate": 6.859125172912507e-06,
+      "loss": 0.083,
+      "step": 1810
+    },
+    {
+      "epoch": 1.2433916924133195,
+      "grad_norm": 0.177734375,
+      "learning_rate": 6.855975987270007e-06,
+      "loss": 0.0823,
+      "step": 1811
+    },
+    {
+      "epoch": 1.2440782698249229,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 6.8528260710535495e-06,
+      "loss": 0.0897,
+      "step": 1812
+    },
+    {
+      "epoch": 1.2447648472365258,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 6.8496754259602515e-06,
+      "loss": 0.0917,
+      "step": 1813
+    },
+    {
+      "epoch": 1.2454514246481292,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 6.846524053687619e-06,
+      "loss": 0.0872,
+      "step": 1814
+    },
+    {
+      "epoch": 1.2461380020597321,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 6.84337195593355e-06,
+      "loss": 0.0873,
+      "step": 1815
+    },
+    {
+      "epoch": 1.2468245794713355,
+      "grad_norm": 0.19140625,
+      "learning_rate": 6.840219134396334e-06,
+      "loss": 0.083,
+      "step": 1816
+    },
+    {
+      "epoch": 1.2475111568829385,
+      "grad_norm": 0.189453125,
+      "learning_rate": 6.837065590774654e-06,
+      "loss": 0.0944,
+      "step": 1817
+    },
+    {
+      "epoch": 1.2481977342945416,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 6.833911326767578e-06,
+      "loss": 0.0837,
+      "step": 1818
+    },
+    {
+      "epoch": 1.2488843117061448,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 6.830756344074562e-06,
+      "loss": 0.0795,
+      "step": 1819
+    },
+    {
+      "epoch": 1.249570889117748,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 6.82760064439545e-06,
+      "loss": 0.0911,
+      "step": 1820
+    },
+    {
+      "epoch": 1.250257466529351,
+      "grad_norm": 0.177734375,
+      "learning_rate": 6.8244442294304745e-06,
+      "loss": 0.088,
+      "step": 1821
+    },
+    {
+      "epoch": 1.2509440439409543,
+      "grad_norm": 0.177734375,
+      "learning_rate": 6.8212871008802515e-06,
+      "loss": 0.0903,
+      "step": 1822
+    },
+    {
+      "epoch": 1.2516306213525574,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 6.818129260445781e-06,
+      "loss": 0.0767,
+      "step": 1823
+    },
+    {
+      "epoch": 1.2523171987641606,
+      "grad_norm": 0.173828125,
+      "learning_rate": 6.814970709828448e-06,
+      "loss": 0.0857,
+      "step": 1824
+    },
+    {
+      "epoch": 1.2530037761757638,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 6.8118114507300174e-06,
+      "loss": 0.0814,
+      "step": 1825
+    },
+    {
+      "epoch": 1.253690353587367,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 6.808651484852637e-06,
+      "loss": 0.0846,
+      "step": 1826
+    },
+    {
+      "epoch": 1.25437693099897,
+      "grad_norm": 0.19140625,
+      "learning_rate": 6.80549081389884e-06,
+      "loss": 0.0867,
+      "step": 1827
+    },
+    {
+      "epoch": 1.2550635084105732,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 6.802329439571531e-06,
+      "loss": 0.0809,
+      "step": 1828
+    },
+    {
+      "epoch": 1.2557500858221764,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 6.799167363574001e-06,
+      "loss": 0.075,
+      "step": 1829
+    },
+    {
+      "epoch": 1.2564366632337796,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 6.796004587609914e-06,
+      "loss": 0.0762,
+      "step": 1830
+    },
+    {
+      "epoch": 1.2571232406453827,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 6.792841113383316e-06,
+      "loss": 0.0738,
+      "step": 1831
+    },
+    {
+      "epoch": 1.2578098180569859,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 6.789676942598626e-06,
+      "loss": 0.0801,
+      "step": 1832
+    },
+    {
+      "epoch": 1.258496395468589,
+      "grad_norm": 0.15234375,
+      "learning_rate": 6.786512076960637e-06,
+      "loss": 0.0795,
+      "step": 1833
+    },
+    {
+      "epoch": 1.2591829728801922,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 6.783346518174519e-06,
+      "loss": 0.0761,
+      "step": 1834
+    },
+    {
+      "epoch": 1.2598695502917954,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 6.780180267945814e-06,
+      "loss": 0.085,
+      "step": 1835
+    },
+    {
+      "epoch": 1.2605561277033985,
+      "grad_norm": 0.17578125,
+      "learning_rate": 6.777013327980438e-06,
+      "loss": 0.0854,
+      "step": 1836
+    },
+    {
+      "epoch": 1.2612427051150017,
+      "grad_norm": 0.193359375,
+      "learning_rate": 6.773845699984677e-06,
+      "loss": 0.0829,
+      "step": 1837
+    },
+    {
+      "epoch": 1.2619292825266049,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 6.770677385665188e-06,
+      "loss": 0.0969,
+      "step": 1838
+    },
+    {
+      "epoch": 1.262615859938208,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 6.767508386729003e-06,
+      "loss": 0.0813,
+      "step": 1839
+    },
+    {
+      "epoch": 1.2633024373498112,
+      "grad_norm": 0.15625,
+      "learning_rate": 6.764338704883511e-06,
+      "loss": 0.0792,
+      "step": 1840
+    },
+    {
+      "epoch": 1.2639890147614143,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 6.761168341836479e-06,
+      "loss": 0.0868,
+      "step": 1841
+    },
+    {
+      "epoch": 1.2646755921730175,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 6.7579972992960405e-06,
+      "loss": 0.0884,
+      "step": 1842
+    },
+    {
+      "epoch": 1.2653621695846207,
+      "grad_norm": 0.169921875,
+      "learning_rate": 6.75482557897069e-06,
+      "loss": 0.0804,
+      "step": 1843
+    },
+    {
+      "epoch": 1.2660487469962238,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 6.751653182569289e-06,
+      "loss": 0.0865,
+      "step": 1844
+    },
+    {
+      "epoch": 1.266735324407827,
+      "grad_norm": 0.17578125,
+      "learning_rate": 6.7484801118010655e-06,
+      "loss": 0.0934,
+      "step": 1845
+    },
+    {
+      "epoch": 1.2674219018194302,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 6.745306368375611e-06,
+      "loss": 0.0866,
+      "step": 1846
+    },
+    {
+      "epoch": 1.2681084792310333,
+      "grad_norm": 0.1640625,
+      "learning_rate": 6.742131954002874e-06,
+      "loss": 0.0896,
+      "step": 1847
+    },
+    {
+      "epoch": 1.2687950566426365,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 6.73895687039317e-06,
+      "loss": 0.0914,
+      "step": 1848
+    },
+    {
+      "epoch": 1.2694816340542396,
+      "grad_norm": 0.166015625,
+      "learning_rate": 6.735781119257174e-06,
+      "loss": 0.0878,
+      "step": 1849
+    },
+    {
+      "epoch": 1.2701682114658428,
+      "grad_norm": 0.16796875,
+      "learning_rate": 6.732604702305921e-06,
+      "loss": 0.0842,
+      "step": 1850
+    },
+    {
+      "epoch": 1.270854788877446,
+      "grad_norm": 0.169921875,
+      "learning_rate": 6.729427621250803e-06,
+      "loss": 0.0964,
+      "step": 1851
+    },
+    {
+      "epoch": 1.2715413662890491,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 6.726249877803568e-06,
+      "loss": 0.0902,
+      "step": 1852
+    },
+    {
+      "epoch": 1.2722279437006523,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 6.723071473676327e-06,
+      "loss": 0.0952,
+      "step": 1853
+    },
+    {
+      "epoch": 1.2729145211122554,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 6.719892410581542e-06,
+      "loss": 0.0821,
+      "step": 1854
+    },
+    {
+      "epoch": 1.2736010985238586,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 6.7167126902320325e-06,
+      "loss": 0.0747,
+      "step": 1855
+    },
+    {
+      "epoch": 1.2742876759354618,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 6.713532314340968e-06,
+      "loss": 0.0816,
+      "step": 1856
+    },
+    {
+      "epoch": 1.274974253347065,
+      "grad_norm": 0.169921875,
+      "learning_rate": 6.710351284621876e-06,
+      "loss": 0.0878,
+      "step": 1857
+    },
+    {
+      "epoch": 1.275660830758668,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 6.707169602788638e-06,
+      "loss": 0.0751,
+      "step": 1858
+    },
+    {
+      "epoch": 1.2763474081702713,
+      "grad_norm": 0.1953125,
+      "learning_rate": 6.703987270555477e-06,
+      "loss": 0.0942,
+      "step": 1859
+    },
+    {
+      "epoch": 1.2770339855818744,
+      "grad_norm": 0.173828125,
+      "learning_rate": 6.700804289636977e-06,
+      "loss": 0.0758,
+      "step": 1860
+    },
+    {
+      "epoch": 1.2777205629934776,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 6.697620661748068e-06,
+      "loss": 0.0849,
+      "step": 1861
+    },
+    {
+      "epoch": 1.2784071404050807,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 6.694436388604025e-06,
+      "loss": 0.0808,
+      "step": 1862
+    },
+    {
+      "epoch": 1.279093717816684,
+      "grad_norm": 0.173828125,
+      "learning_rate": 6.691251471920475e-06,
+      "loss": 0.0846,
+      "step": 1863
+    },
+    {
+      "epoch": 1.279780295228287,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 6.688065913413391e-06,
+      "loss": 0.0876,
+      "step": 1864
+    },
+    {
+      "epoch": 1.2804668726398902,
+      "grad_norm": 0.158203125,
+      "learning_rate": 6.684879714799088e-06,
+      "loss": 0.0769,
+      "step": 1865
+    },
+    {
+      "epoch": 1.2811534500514932,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 6.6816928777942325e-06,
+      "loss": 0.0823,
+      "step": 1866
+    },
+    {
+      "epoch": 1.2818400274630966,
+      "grad_norm": 0.162109375,
+      "learning_rate": 6.67850540411583e-06,
+      "loss": 0.0737,
+      "step": 1867
+    },
+    {
+      "epoch": 1.2825266048746995,
+      "grad_norm": 0.173828125,
+      "learning_rate": 6.67531729548123e-06,
+      "loss": 0.0725,
+      "step": 1868
+    },
+    {
+      "epoch": 1.2832131822863029,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 6.6721285536081245e-06,
+      "loss": 0.0786,
+      "step": 1869
+    },
+    {
+      "epoch": 1.2838997596979058,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 6.6689391802145476e-06,
+      "loss": 0.0863,
+      "step": 1870
+    },
+    {
+      "epoch": 1.2845863371095092,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 6.665749177018872e-06,
+      "loss": 0.0836,
+      "step": 1871
+    },
+    {
+      "epoch": 1.2852729145211121,
+      "grad_norm": 0.189453125,
+      "learning_rate": 6.662558545739812e-06,
+      "loss": 0.0824,
+      "step": 1872
+    },
+    {
+      "epoch": 1.2859594919327155,
+      "grad_norm": 0.171875,
+      "learning_rate": 6.659367288096417e-06,
+      "loss": 0.0881,
+      "step": 1873
+    },
+    {
+      "epoch": 1.2866460693443185,
+      "grad_norm": 0.171875,
+      "learning_rate": 6.656175405808075e-06,
+      "loss": 0.0838,
+      "step": 1874
+    },
+    {
+      "epoch": 1.2873326467559219,
+      "grad_norm": 0.177734375,
+      "learning_rate": 6.652982900594514e-06,
+      "loss": 0.0898,
+      "step": 1875
+    },
+    {
+      "epoch": 1.2880192241675248,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 6.649789774175792e-06,
+      "loss": 0.0884,
+      "step": 1876
+    },
+    {
+      "epoch": 1.2887058015791282,
+      "grad_norm": 0.16015625,
+      "learning_rate": 6.646596028272306e-06,
+      "loss": 0.0806,
+      "step": 1877
+    },
+    {
+      "epoch": 1.2893923789907311,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 6.643401664604785e-06,
+      "loss": 0.0932,
+      "step": 1878
+    },
+    {
+      "epoch": 1.2900789564023345,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 6.640206684894291e-06,
+      "loss": 0.0818,
+      "step": 1879
+    },
+    {
+      "epoch": 1.2907655338139374,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 6.637011090862219e-06,
+      "loss": 0.0824,
+      "step": 1880
+    },
+    {
+      "epoch": 1.2914521112255406,
+      "grad_norm": 0.17578125,
+      "learning_rate": 6.6338148842302914e-06,
+      "loss": 0.1001,
+      "step": 1881
+    },
+    {
+      "epoch": 1.2921386886371438,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 6.630618066720568e-06,
+      "loss": 0.0759,
+      "step": 1882
+    },
+    {
+      "epoch": 1.292825266048747,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 6.627420640055428e-06,
+      "loss": 0.0905,
+      "step": 1883
+    },
+    {
+      "epoch": 1.29351184346035,
+      "grad_norm": 0.177734375,
+      "learning_rate": 6.624222605957586e-06,
+      "loss": 0.079,
+      "step": 1884
+    },
+    {
+      "epoch": 1.2941984208719532,
+      "grad_norm": 0.171875,
+      "learning_rate": 6.62102396615008e-06,
+      "loss": 0.0796,
+      "step": 1885
+    },
+    {
+      "epoch": 1.2948849982835564,
+      "grad_norm": 0.17578125,
+      "learning_rate": 6.617824722356281e-06,
+      "loss": 0.0865,
+      "step": 1886
+    },
+    {
+      "epoch": 1.2955715756951596,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 6.614624876299874e-06,
+      "loss": 0.0912,
+      "step": 1887
+    },
+    {
+      "epoch": 1.2962581531067627,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 6.611424429704879e-06,
+      "loss": 0.0945,
+      "step": 1888
+    },
+    {
+      "epoch": 1.296944730518366,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 6.608223384295635e-06,
+      "loss": 0.0961,
+      "step": 1889
+    },
+    {
+      "epoch": 1.297631307929969,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 6.6050217417968025e-06,
+      "loss": 0.0914,
+      "step": 1890
+    },
+    {
+      "epoch": 1.2983178853415722,
+      "grad_norm": 0.169921875,
+      "learning_rate": 6.601819503933367e-06,
+      "loss": 0.0817,
+      "step": 1891
+    },
+    {
+      "epoch": 1.2990044627531754,
+      "grad_norm": 0.1640625,
+      "learning_rate": 6.598616672430631e-06,
+      "loss": 0.0874,
+      "step": 1892
+    },
+    {
+      "epoch": 1.2996910401647785,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 6.595413249014222e-06,
+      "loss": 0.077,
+      "step": 1893
+    },
+    {
+      "epoch": 1.3003776175763817,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 6.592209235410081e-06,
+      "loss": 0.0865,
+      "step": 1894
+    },
+    {
+      "epoch": 1.3010641949879849,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 6.589004633344469e-06,
+      "loss": 0.0812,
+      "step": 1895
+    },
+    {
+      "epoch": 1.301750772399588,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 6.585799444543967e-06,
+      "loss": 0.0827,
+      "step": 1896
+    },
+    {
+      "epoch": 1.3024373498111912,
+      "grad_norm": 0.177734375,
+      "learning_rate": 6.582593670735466e-06,
+      "loss": 0.0991,
+      "step": 1897
+    },
+    {
+      "epoch": 1.3031239272227944,
+      "grad_norm": 0.173828125,
+      "learning_rate": 6.57938731364618e-06,
+      "loss": 0.0942,
+      "step": 1898
+    },
+    {
+      "epoch": 1.3038105046343975,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 6.576180375003628e-06,
+      "loss": 0.0771,
+      "step": 1899
+    },
+    {
+      "epoch": 1.3044970820460007,
+      "grad_norm": 0.162109375,
+      "learning_rate": 6.5729728565356495e-06,
+      "loss": 0.08,
+      "step": 1900
+    },
+    {
+      "epoch": 1.3051836594576038,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 6.569764759970397e-06,
+      "loss": 0.0904,
+      "step": 1901
+    },
+    {
+      "epoch": 1.305870236869207,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 6.566556087036329e-06,
+      "loss": 0.0856,
+      "step": 1902
+    },
+    {
+      "epoch": 1.3065568142808102,
+      "grad_norm": 0.166015625,
+      "learning_rate": 6.5633468394622145e-06,
+      "loss": 0.0854,
+      "step": 1903
+    },
+    {
+      "epoch": 1.3072433916924133,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 6.560137018977139e-06,
+      "loss": 0.0848,
+      "step": 1904
+    },
+    {
+      "epoch": 1.3079299691040165,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 6.556926627310492e-06,
+      "loss": 0.0785,
+      "step": 1905
+    },
+    {
+      "epoch": 1.3086165465156196,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 6.5537156661919685e-06,
+      "loss": 0.0927,
+      "step": 1906
+    },
+    {
+      "epoch": 1.3093031239272228,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 6.550504137351576e-06,
+      "loss": 0.0745,
+      "step": 1907
+    },
+    {
+      "epoch": 1.309989701338826,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 6.547292042519623e-06,
+      "loss": 0.0811,
+      "step": 1908
+    },
+    {
+      "epoch": 1.3106762787504291,
+      "grad_norm": 0.1875,
+      "learning_rate": 6.544079383426725e-06,
+      "loss": 0.0929,
+      "step": 1909
+    },
+    {
+      "epoch": 1.3113628561620323,
+      "grad_norm": 0.17578125,
+      "learning_rate": 6.5408661618038e-06,
+      "loss": 0.0777,
+      "step": 1910
+    },
+    {
+      "epoch": 1.3120494335736355,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 6.537652379382076e-06,
+      "loss": 0.0793,
+      "step": 1911
+    },
+    {
+      "epoch": 1.3127360109852386,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 6.53443803789307e-06,
+      "loss": 0.0794,
+      "step": 1912
+    },
+    {
+      "epoch": 1.3134225883968418,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 6.5312231390686135e-06,
+      "loss": 0.0815,
+      "step": 1913
+    },
+    {
+      "epoch": 1.314109165808445,
+      "grad_norm": 0.189453125,
+      "learning_rate": 6.52800768464083e-06,
+      "loss": 0.0968,
+      "step": 1914
+    },
+    {
+      "epoch": 1.314795743220048,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 6.524791676342146e-06,
+      "loss": 0.0816,
+      "step": 1915
+    },
+    {
+      "epoch": 1.3154823206316513,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 6.521575115905284e-06,
+      "loss": 0.0855,
+      "step": 1916
+    },
+    {
+      "epoch": 1.3161688980432544,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 6.518358005063268e-06,
+      "loss": 0.0866,
+      "step": 1917
+    },
+    {
+      "epoch": 1.3168554754548576,
+      "grad_norm": 0.171875,
+      "learning_rate": 6.515140345549411e-06,
+      "loss": 0.0895,
+      "step": 1918
+    },
+    {
+      "epoch": 1.3175420528664608,
+      "grad_norm": 0.181640625,
+      "learning_rate": 6.511922139097332e-06,
+      "loss": 0.0886,
+      "step": 1919
+    },
+    {
+      "epoch": 1.318228630278064,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 6.5087033874409354e-06,
+      "loss": 0.094,
+      "step": 1920
+    },
+    {
+      "epoch": 1.318915207689667,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 6.505484092314425e-06,
+      "loss": 0.0765,
+      "step": 1921
+    },
+    {
+      "epoch": 1.3196017851012702,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 6.502264255452295e-06,
+      "loss": 0.0837,
+      "step": 1922
+    },
+    {
+      "epoch": 1.3202883625128734,
+      "grad_norm": 0.17578125,
+      "learning_rate": 6.499043878589332e-06,
+      "loss": 0.0845,
+      "step": 1923
+    },
+    {
+      "epoch": 1.3209749399244766,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 6.4958229634606115e-06,
+      "loss": 0.0883,
+      "step": 1924
+    },
+    {
+      "epoch": 1.3216615173360795,
+      "grad_norm": 0.1640625,
+      "learning_rate": 6.492601511801504e-06,
+      "loss": 0.086,
+      "step": 1925
+    },
+    {
+      "epoch": 1.322348094747683,
+      "grad_norm": 0.17578125,
+      "learning_rate": 6.489379525347666e-06,
+      "loss": 0.0703,
+      "step": 1926
+    },
+    {
+      "epoch": 1.3230346721592858,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 6.4861570058350394e-06,
+      "loss": 0.0793,
+      "step": 1927
+    },
+    {
+      "epoch": 1.3237212495708892,
+      "grad_norm": 0.166015625,
+      "learning_rate": 6.482933954999858e-06,
+      "loss": 0.0777,
+      "step": 1928
+    },
+    {
+      "epoch": 1.3244078269824922,
+      "grad_norm": 0.189453125,
+      "learning_rate": 6.479710374578642e-06,
+      "loss": 0.0901,
+      "step": 1929
+    },
+    {
+      "epoch": 1.3250944043940955,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 6.476486266308192e-06,
+      "loss": 0.1017,
+      "step": 1930
+    },
+    {
+      "epoch": 1.3257809818056985,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 6.473261631925597e-06,
+      "loss": 0.0902,
+      "step": 1931
+    },
+    {
+      "epoch": 1.3264675592173019,
+      "grad_norm": 0.169921875,
+      "learning_rate": 6.470036473168228e-06,
+      "loss": 0.0905,
+      "step": 1932
+    },
+    {
+      "epoch": 1.3271541366289048,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 6.466810791773739e-06,
+      "loss": 0.0745,
+      "step": 1933
+    },
+    {
+      "epoch": 1.3278407140405082,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 6.463584589480066e-06,
+      "loss": 0.0914,
+      "step": 1934
+    },
+    {
+      "epoch": 1.3285272914521111,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 6.4603578680254255e-06,
+      "loss": 0.0875,
+      "step": 1935
+    },
+    {
+      "epoch": 1.3292138688637145,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 6.457130629148312e-06,
+      "loss": 0.0842,
+      "step": 1936
+    },
+    {
+      "epoch": 1.3299004462753174,
+      "grad_norm": 0.1875,
+      "learning_rate": 6.4539028745875006e-06,
+      "loss": 0.0876,
+      "step": 1937
+    },
+    {
+      "epoch": 1.3305870236869208,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 6.450674606082046e-06,
+      "loss": 0.1012,
+      "step": 1938
+    },
+    {
+      "epoch": 1.3312736010985238,
+      "grad_norm": 0.16796875,
+      "learning_rate": 6.447445825371272e-06,
+      "loss": 0.0747,
+      "step": 1939
+    },
+    {
+      "epoch": 1.331960178510127,
+      "grad_norm": 0.1796875,
+      "learning_rate": 6.444216534194791e-06,
+      "loss": 0.0896,
+      "step": 1940
+    },
+    {
+      "epoch": 1.33264675592173,
+      "grad_norm": 0.1875,
+      "learning_rate": 6.440986734292479e-06,
+      "loss": 0.0882,
+      "step": 1941
+    },
+    {
+      "epoch": 1.3333333333333333,
+      "grad_norm": 0.185546875,
+      "learning_rate": 6.437756427404491e-06,
+      "loss": 0.0886,
+      "step": 1942
+    },
+    {
+      "epoch": 1.3340199107449364,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 6.434525615271257e-06,
+      "loss": 0.0846,
+      "step": 1943
+    },
+    {
+      "epoch": 1.3347064881565396,
+      "grad_norm": 0.19921875,
+      "learning_rate": 6.431294299633473e-06,
+      "loss": 0.0831,
+      "step": 1944
+    },
+    {
+      "epoch": 1.3353930655681427,
+      "grad_norm": 0.17578125,
+      "learning_rate": 6.428062482232114e-06,
+      "loss": 0.0881,
+      "step": 1945
+    },
+    {
+      "epoch": 1.336079642979746,
+      "grad_norm": 0.177734375,
+      "learning_rate": 6.424830164808418e-06,
+      "loss": 0.0833,
+      "step": 1946
+    },
+    {
+      "epoch": 1.336766220391349,
+      "grad_norm": 0.169921875,
+      "learning_rate": 6.421597349103899e-06,
+      "loss": 0.0932,
+      "step": 1947
+    },
+    {
+      "epoch": 1.3374527978029522,
+      "grad_norm": 0.166015625,
+      "learning_rate": 6.418364036860334e-06,
+      "loss": 0.0811,
+      "step": 1948
+    },
+    {
+      "epoch": 1.3381393752145554,
+      "grad_norm": 0.1640625,
+      "learning_rate": 6.41513022981977e-06,
+      "loss": 0.0777,
+      "step": 1949
+    },
+    {
+      "epoch": 1.3388259526261586,
+      "grad_norm": 0.17578125,
+      "learning_rate": 6.4118959297245235e-06,
+      "loss": 0.0826,
+      "step": 1950
+    },
+    {
+      "epoch": 1.3395125300377617,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 6.4086611383171685e-06,
+      "loss": 0.0897,
+      "step": 1951
+    },
+    {
+      "epoch": 1.3401991074493649,
+      "grad_norm": 0.189453125,
+      "learning_rate": 6.405425857340554e-06,
+      "loss": 0.0876,
+      "step": 1952
+    },
+    {
+      "epoch": 1.340885684860968,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 6.4021900885377845e-06,
+      "loss": 0.0818,
+      "step": 1953
+    },
+    {
+      "epoch": 1.3415722622725712,
+      "grad_norm": 0.1953125,
+      "learning_rate": 6.3989538336522295e-06,
+      "loss": 0.0926,
+      "step": 1954
+    },
+    {
+      "epoch": 1.3422588396841744,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 6.395717094427524e-06,
+      "loss": 0.0836,
+      "step": 1955
+    },
+    {
+      "epoch": 1.3429454170957775,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 6.39247987260756e-06,
+      "loss": 0.0882,
+      "step": 1956
+    },
+    {
+      "epoch": 1.3436319945073807,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 6.389242169936491e-06,
+      "loss": 0.0817,
+      "step": 1957
+    },
+    {
+      "epoch": 1.3443185719189839,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 6.386003988158729e-06,
+      "loss": 0.0735,
+      "step": 1958
+    },
+    {
+      "epoch": 1.345005149330587,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 6.382765329018945e-06,
+      "loss": 0.0849,
+      "step": 1959
+    },
+    {
+      "epoch": 1.3456917267421902,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 6.3795261942620665e-06,
+      "loss": 0.0791,
+      "step": 1960
+    },
+    {
+      "epoch": 1.3463783041537933,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 6.376286585633279e-06,
+      "loss": 0.0916,
+      "step": 1961
+    },
+    {
+      "epoch": 1.3470648815653965,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 6.373046504878019e-06,
+      "loss": 0.0809,
+      "step": 1962
+    },
+    {
+      "epoch": 1.3477514589769997,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 6.369805953741981e-06,
+      "loss": 0.08,
+      "step": 1963
+    },
+    {
+      "epoch": 1.3484380363886028,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 6.366564933971111e-06,
+      "loss": 0.0763,
+      "step": 1964
+    },
+    {
+      "epoch": 1.349124613800206,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 6.363323447311611e-06,
+      "loss": 0.0796,
+      "step": 1965
+    },
+    {
+      "epoch": 1.3498111912118091,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 6.36008149550993e-06,
+      "loss": 0.0879,
+      "step": 1966
+    },
+    {
+      "epoch": 1.3504977686234123,
+      "grad_norm": 0.1875,
+      "learning_rate": 6.356839080312772e-06,
+      "loss": 0.0842,
+      "step": 1967
+    },
+    {
+      "epoch": 1.3511843460350155,
+      "grad_norm": 0.177734375,
+      "learning_rate": 6.353596203467085e-06,
+      "loss": 0.0864,
+      "step": 1968
+    },
+    {
+      "epoch": 1.3518709234466186,
+      "grad_norm": 0.1796875,
+      "learning_rate": 6.350352866720073e-06,
+      "loss": 0.083,
+      "step": 1969
+    },
+    {
+      "epoch": 1.3525575008582218,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 6.347109071819181e-06,
+      "loss": 0.0877,
+      "step": 1970
+    },
+    {
+      "epoch": 1.353244078269825,
+      "grad_norm": 0.16796875,
+      "learning_rate": 6.3438648205121045e-06,
+      "loss": 0.0829,
+      "step": 1971
+    },
+    {
+      "epoch": 1.3539306556814281,
+      "grad_norm": 0.171875,
+      "learning_rate": 6.340620114546784e-06,
+      "loss": 0.0731,
+      "step": 1972
+    },
+    {
+      "epoch": 1.3546172330930313,
+      "grad_norm": 0.17578125,
+      "learning_rate": 6.337374955671406e-06,
+      "loss": 0.0735,
+      "step": 1973
+    },
+    {
+      "epoch": 1.3553038105046344,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 6.3341293456344e-06,
+      "loss": 0.0836,
+      "step": 1974
+    },
+    {
+      "epoch": 1.3559903879162376,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 6.3308832861844376e-06,
+      "loss": 0.0796,
+      "step": 1975
+    },
+    {
+      "epoch": 1.3566769653278408,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 6.3276367790704315e-06,
+      "loss": 0.0849,
+      "step": 1976
+    },
+    {
+      "epoch": 1.357363542739444,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 6.324389826041544e-06,
+      "loss": 0.0759,
+      "step": 1977
+    },
+    {
+      "epoch": 1.358050120151047,
+      "grad_norm": 0.1796875,
+      "learning_rate": 6.321142428847166e-06,
+      "loss": 0.081,
+      "step": 1978
+    },
+    {
+      "epoch": 1.3587366975626503,
+      "grad_norm": 0.181640625,
+      "learning_rate": 6.317894589236936e-06,
+      "loss": 0.086,
+      "step": 1979
+    },
+    {
+      "epoch": 1.3594232749742534,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 6.314646308960725e-06,
+      "loss": 0.0781,
+      "step": 1980
+    },
+    {
+      "epoch": 1.3601098523858566,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 6.3113975897686475e-06,
+      "loss": 0.0761,
+      "step": 1981
+    },
+    {
+      "epoch": 1.3607964297974597,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 6.3081484334110495e-06,
+      "loss": 0.0778,
+      "step": 1982
+    },
+    {
+      "epoch": 1.361483007209063,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 6.304898841638514e-06,
+      "loss": 0.0935,
+      "step": 1983
+    },
+    {
+      "epoch": 1.362169584620666,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 6.30164881620186e-06,
+      "loss": 0.0949,
+      "step": 1984
+    },
+    {
+      "epoch": 1.3628561620322692,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 6.298398358852139e-06,
+      "loss": 0.0837,
+      "step": 1985
+    },
+    {
+      "epoch": 1.3635427394438722,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 6.295147471340634e-06,
+      "loss": 0.0836,
+      "step": 1986
+    },
+    {
+      "epoch": 1.3642293168554755,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 6.291896155418865e-06,
+      "loss": 0.0868,
+      "step": 1987
+    },
+    {
+      "epoch": 1.3649158942670785,
+      "grad_norm": 0.185546875,
+      "learning_rate": 6.2886444128385745e-06,
+      "loss": 0.0838,
+      "step": 1988
+    },
+    {
+      "epoch": 1.3656024716786819,
+      "grad_norm": 0.162109375,
+      "learning_rate": 6.285392245351741e-06,
+      "loss": 0.0765,
+      "step": 1989
+    },
+    {
+      "epoch": 1.3662890490902848,
+      "grad_norm": 0.181640625,
+      "learning_rate": 6.2821396547105705e-06,
+      "loss": 0.0843,
+      "step": 1990
+    },
+    {
+      "epoch": 1.3669756265018882,
+      "grad_norm": 0.16796875,
+      "learning_rate": 6.2788866426674964e-06,
+      "loss": 0.083,
+      "step": 1991
+    },
+    {
+      "epoch": 1.3676622039134911,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 6.275633210975179e-06,
+      "loss": 0.1028,
+      "step": 1992
+    },
+    {
+      "epoch": 1.3683487813250945,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 6.272379361386507e-06,
+      "loss": 0.0862,
+      "step": 1993
+    },
+    {
+      "epoch": 1.3690353587366975,
+      "grad_norm": 0.169921875,
+      "learning_rate": 6.269125095654588e-06,
+      "loss": 0.0757,
+      "step": 1994
+    },
+    {
+      "epoch": 1.3697219361483008,
+      "grad_norm": 0.193359375,
+      "learning_rate": 6.265870415532763e-06,
+      "loss": 0.0858,
+      "step": 1995
+    },
+    {
+      "epoch": 1.3704085135599038,
+      "grad_norm": 0.177734375,
+      "learning_rate": 6.26261532277459e-06,
+      "loss": 0.0983,
+      "step": 1996
+    },
+    {
+      "epoch": 1.3710950909715072,
+      "grad_norm": 0.171875,
+      "learning_rate": 6.259359819133849e-06,
+      "loss": 0.0873,
+      "step": 1997
+    },
+    {
+      "epoch": 1.37178166838311,
+      "grad_norm": 0.244140625,
+      "learning_rate": 6.256103906364545e-06,
+      "loss": 0.0859,
+      "step": 1998
+    },
+    {
+      "epoch": 1.3724682457947133,
+      "grad_norm": 0.169921875,
+      "learning_rate": 6.252847586220901e-06,
+      "loss": 0.0891,
+      "step": 1999
+    },
+    {
+      "epoch": 1.3731548232063164,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 6.249590860457362e-06,
+      "loss": 0.0823,
+      "step": 2000
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 4368,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 1000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 2.060987570061312e+18,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}