diff --git "a/checkpoint-9250/trainer_state.json" "b/checkpoint-9250/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-9250/trainer_state.json"
@@ -0,0 +1,64783 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 7.388411588411588,
+  "eval_steps": 500,
+  "global_step": 9250,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0007992007992007992,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.0002,
+      "loss": 3.2116,
+      "step": 1
+    },
+    {
+      "epoch": 0.0015984015984015984,
+      "grad_norm": 3.046875,
+      "learning_rate": 0.0002,
+      "loss": 2.9686,
+      "step": 2
+    },
+    {
+      "epoch": 0.0023976023976023976,
+      "grad_norm": 2.703125,
+      "learning_rate": 0.0002,
+      "loss": 2.734,
+      "step": 3
+    },
+    {
+      "epoch": 0.003196803196803197,
+      "grad_norm": 4.5625,
+      "learning_rate": 0.0002,
+      "loss": 2.4344,
+      "step": 4
+    },
+    {
+      "epoch": 0.003996003996003996,
+      "grad_norm": 3.5,
+      "learning_rate": 0.0002,
+      "loss": 2.1558,
+      "step": 5
+    },
+    {
+      "epoch": 0.004795204795204795,
+      "grad_norm": 4.5625,
+      "learning_rate": 0.0002,
+      "loss": 2.0167,
+      "step": 6
+    },
+    {
+      "epoch": 0.005594405594405594,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.0002,
+      "loss": 1.9392,
+      "step": 7
+    },
+    {
+      "epoch": 0.006393606393606394,
+      "grad_norm": 5.90625,
+      "learning_rate": 0.0002,
+      "loss": 1.9313,
+      "step": 8
+    },
+    {
+      "epoch": 0.007192807192807193,
+      "grad_norm": 2.96875,
+      "learning_rate": 0.0002,
+      "loss": 1.9121,
+      "step": 9
+    },
+    {
+      "epoch": 0.007992007992007992,
+      "grad_norm": 8.125,
+      "learning_rate": 0.0002,
+      "loss": 1.9119,
+      "step": 10
+    },
+    {
+      "epoch": 0.008791208791208791,
+      "grad_norm": 10.0625,
+      "learning_rate": 0.0002,
+      "loss": 1.9204,
+      "step": 11
+    },
+    {
+      "epoch": 0.00959040959040959,
+      "grad_norm": 4.65625,
+      "learning_rate": 0.0002,
+      "loss": 1.8705,
+      "step": 12
+    },
+    {
+      "epoch": 0.01038961038961039,
+      "grad_norm": 7.5625,
+      "learning_rate": 0.0002,
+      "loss": 1.8554,
+      "step": 13
+    },
+    {
+      "epoch": 0.011188811188811189,
+      "grad_norm": 10.875,
+      "learning_rate": 0.0002,
+      "loss": 1.851,
+      "step": 14
+    },
+    {
+      "epoch": 0.011988011988011988,
+      "grad_norm": 8.5625,
+      "learning_rate": 0.0002,
+      "loss": 1.8158,
+      "step": 15
+    },
+    {
+      "epoch": 0.012787212787212787,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.0002,
+      "loss": 1.7605,
+      "step": 16
+    },
+    {
+      "epoch": 0.013586413586413586,
+      "grad_norm": 10.25,
+      "learning_rate": 0.0002,
+      "loss": 1.7539,
+      "step": 17
+    },
+    {
+      "epoch": 0.014385614385614386,
+      "grad_norm": 12.3125,
+      "learning_rate": 0.0002,
+      "loss": 1.7519,
+      "step": 18
+    },
+    {
+      "epoch": 0.015184815184815185,
+      "grad_norm": 10.625,
+      "learning_rate": 0.0002,
+      "loss": 1.7278,
+      "step": 19
+    },
+    {
+      "epoch": 0.015984015984015984,
+      "grad_norm": 6.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.6722,
+      "step": 20
+    },
+    {
+      "epoch": 0.016783216783216783,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.0002,
+      "loss": 1.6398,
+      "step": 21
+    },
+    {
+      "epoch": 0.017582417582417582,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.0002,
+      "loss": 1.6283,
+      "step": 22
+    },
+    {
+      "epoch": 0.01838161838161838,
+      "grad_norm": 0.8515625,
+      "learning_rate": 0.0002,
+      "loss": 1.6145,
+      "step": 23
+    },
+    {
+      "epoch": 0.01918081918081918,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.0002,
+      "loss": 1.59,
+      "step": 24
+    },
+    {
+      "epoch": 0.01998001998001998,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0002,
+      "loss": 1.566,
+      "step": 25
+    },
+    {
+      "epoch": 0.02077922077922078,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.0002,
+      "loss": 1.566,
+      "step": 26
+    },
+    {
+      "epoch": 0.02157842157842158,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0002,
+      "loss": 1.5317,
+      "step": 27
+    },
+    {
+      "epoch": 0.022377622377622378,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0002,
+      "loss": 1.5039,
+      "step": 28
+    },
+    {
+      "epoch": 0.023176823176823177,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0002,
+      "loss": 1.4993,
+      "step": 29
+    },
+    {
+      "epoch": 0.023976023976023976,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.0002,
+      "loss": 1.4725,
+      "step": 30
+    },
+    {
+      "epoch": 0.024775224775224775,
+      "grad_norm": 0.859375,
+      "learning_rate": 0.0002,
+      "loss": 1.4626,
+      "step": 31
+    },
+    {
+      "epoch": 0.025574425574425574,
+      "grad_norm": 0.96484375,
+      "learning_rate": 0.0002,
+      "loss": 1.4509,
+      "step": 32
+    },
+    {
+      "epoch": 0.026373626373626374,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.0002,
+      "loss": 1.4543,
+      "step": 33
+    },
+    {
+      "epoch": 0.027172827172827173,
+      "grad_norm": 0.85546875,
+      "learning_rate": 0.0002,
+      "loss": 1.4315,
+      "step": 34
+    },
+    {
+      "epoch": 0.027972027972027972,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.0002,
+      "loss": 1.4186,
+      "step": 35
+    },
+    {
+      "epoch": 0.02877122877122877,
+      "grad_norm": 0.75,
+      "learning_rate": 0.0002,
+      "loss": 1.404,
+      "step": 36
+    },
+    {
+      "epoch": 0.02957042957042957,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.0002,
+      "loss": 1.4039,
+      "step": 37
+    },
+    {
+      "epoch": 0.03036963036963037,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.0002,
+      "loss": 1.3757,
+      "step": 38
+    },
+    {
+      "epoch": 0.03116883116883117,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.0002,
+      "loss": 1.365,
+      "step": 39
+    },
+    {
+      "epoch": 0.03196803196803197,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.0002,
+      "loss": 1.3511,
+      "step": 40
+    },
+    {
+      "epoch": 0.03276723276723277,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.0002,
+      "loss": 1.3416,
+      "step": 41
+    },
+    {
+      "epoch": 0.033566433566433566,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.0002,
+      "loss": 1.3243,
+      "step": 42
+    },
+    {
+      "epoch": 0.03436563436563437,
+      "grad_norm": 0.890625,
+      "learning_rate": 0.0002,
+      "loss": 1.3228,
+      "step": 43
+    },
+    {
+      "epoch": 0.035164835164835165,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.0002,
+      "loss": 1.3174,
+      "step": 44
+    },
+    {
+      "epoch": 0.03596403596403597,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.0002,
+      "loss": 1.2993,
+      "step": 45
+    },
+    {
+      "epoch": 0.03676323676323676,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.0002,
+      "loss": 1.2973,
+      "step": 46
+    },
+    {
+      "epoch": 0.037562437562437566,
+      "grad_norm": 0.90625,
+      "learning_rate": 0.0002,
+      "loss": 1.2756,
+      "step": 47
+    },
+    {
+      "epoch": 0.03836163836163836,
+      "grad_norm": 0.84375,
+      "learning_rate": 0.0002,
+      "loss": 1.2762,
+      "step": 48
+    },
+    {
+      "epoch": 0.039160839160839164,
+      "grad_norm": 0.828125,
+      "learning_rate": 0.0002,
+      "loss": 1.2659,
+      "step": 49
+    },
+    {
+      "epoch": 0.03996003996003996,
+      "grad_norm": 0.796875,
+      "learning_rate": 0.0002,
+      "loss": 1.2436,
+      "step": 50
+    },
+    {
+      "epoch": 0.04075924075924076,
+      "grad_norm": 0.80078125,
+      "learning_rate": 0.0002,
+      "loss": 1.2321,
+      "step": 51
+    },
+    {
+      "epoch": 0.04155844155844156,
+      "grad_norm": 0.828125,
+      "learning_rate": 0.0002,
+      "loss": 1.2447,
+      "step": 52
+    },
+    {
+      "epoch": 0.04235764235764236,
+      "grad_norm": 0.85546875,
+      "learning_rate": 0.0002,
+      "loss": 1.2188,
+      "step": 53
+    },
+    {
+      "epoch": 0.04315684315684316,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.0002,
+      "loss": 1.2265,
+      "step": 54
+    },
+    {
+      "epoch": 0.04395604395604396,
+      "grad_norm": 1.25,
+      "learning_rate": 0.0002,
+      "loss": 1.2319,
+      "step": 55
+    },
+    {
+      "epoch": 0.044755244755244755,
+      "grad_norm": 0.84375,
+      "learning_rate": 0.0002,
+      "loss": 1.2058,
+      "step": 56
+    },
+    {
+      "epoch": 0.04555444555444556,
+      "grad_norm": 1.125,
+      "learning_rate": 0.0002,
+      "loss": 1.2101,
+      "step": 57
+    },
+    {
+      "epoch": 0.046353646353646354,
+      "grad_norm": 0.82421875,
+      "learning_rate": 0.0002,
+      "loss": 1.1909,
+      "step": 58
+    },
+    {
+      "epoch": 0.047152847152847156,
+      "grad_norm": 0.95703125,
+      "learning_rate": 0.0002,
+      "loss": 1.2001,
+      "step": 59
+    },
+    {
+      "epoch": 0.04795204795204795,
+      "grad_norm": 0.9609375,
+      "learning_rate": 0.0002,
+      "loss": 1.1865,
+      "step": 60
+    },
+    {
+      "epoch": 0.048751248751248755,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.0002,
+      "loss": 1.1945,
+      "step": 61
+    },
+    {
+      "epoch": 0.04955044955044955,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.0002,
+      "loss": 1.1705,
+      "step": 62
+    },
+    {
+      "epoch": 0.05034965034965035,
+      "grad_norm": 1.484375,
+      "learning_rate": 0.0002,
+      "loss": 1.1735,
+      "step": 63
+    },
+    {
+      "epoch": 0.05114885114885115,
+      "grad_norm": 0.8359375,
+      "learning_rate": 0.0002,
+      "loss": 1.1707,
+      "step": 64
+    },
+    {
+      "epoch": 0.05194805194805195,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.184,
+      "step": 65
+    },
+    {
+      "epoch": 0.05274725274725275,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.0002,
+      "loss": 1.1475,
+      "step": 66
+    },
+    {
+      "epoch": 0.05354645354645355,
+      "grad_norm": 0.953125,
+      "learning_rate": 0.0002,
+      "loss": 1.1537,
+      "step": 67
+    },
+    {
+      "epoch": 0.054345654345654346,
+      "grad_norm": 0.921875,
+      "learning_rate": 0.0002,
+      "loss": 1.151,
+      "step": 68
+    },
+    {
+      "epoch": 0.05514485514485515,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.0002,
+      "loss": 1.1522,
+      "step": 69
+    },
+    {
+      "epoch": 0.055944055944055944,
+      "grad_norm": 0.98828125,
+      "learning_rate": 0.0002,
+      "loss": 1.1533,
+      "step": 70
+    },
+    {
+      "epoch": 0.05674325674325675,
+      "grad_norm": 0.86328125,
+      "learning_rate": 0.0002,
+      "loss": 1.1393,
+      "step": 71
+    },
+    {
+      "epoch": 0.05754245754245754,
+      "grad_norm": 0.9140625,
+      "learning_rate": 0.0002,
+      "loss": 1.1469,
+      "step": 72
+    },
+    {
+      "epoch": 0.058341658341658345,
+      "grad_norm": 0.85546875,
+      "learning_rate": 0.0002,
+      "loss": 1.1306,
+      "step": 73
+    },
+    {
+      "epoch": 0.05914085914085914,
+      "grad_norm": 0.953125,
+      "learning_rate": 0.0002,
+      "loss": 1.1305,
+      "step": 74
+    },
+    {
+      "epoch": 0.059940059940059943,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.0002,
+      "loss": 1.1233,
+      "step": 75
+    },
+    {
+      "epoch": 0.06073926073926074,
+      "grad_norm": 0.80078125,
+      "learning_rate": 0.0002,
+      "loss": 1.1207,
+      "step": 76
+    },
+    {
+      "epoch": 0.06153846153846154,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.1107,
+      "step": 77
+    },
+    {
+      "epoch": 0.06233766233766234,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.0002,
+      "loss": 1.1134,
+      "step": 78
+    },
+    {
+      "epoch": 0.06313686313686313,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.0002,
+      "loss": 1.1057,
+      "step": 79
+    },
+    {
+      "epoch": 0.06393606393606394,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.0002,
+      "loss": 1.1146,
+      "step": 80
+    },
+    {
+      "epoch": 0.06473526473526474,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.0002,
+      "loss": 1.1159,
+      "step": 81
+    },
+    {
+      "epoch": 0.06553446553446554,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.0002,
+      "loss": 1.0952,
+      "step": 82
+    },
+    {
+      "epoch": 0.06633366633366633,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.0002,
+      "loss": 1.0873,
+      "step": 83
+    },
+    {
+      "epoch": 0.06713286713286713,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0002,
+      "loss": 1.0875,
+      "step": 84
+    },
+    {
+      "epoch": 0.06793206793206794,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.0002,
+      "loss": 1.0839,
+      "step": 85
+    },
+    {
+      "epoch": 0.06873126873126874,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.0002,
+      "loss": 1.081,
+      "step": 86
+    },
+    {
+      "epoch": 0.06953046953046953,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.0002,
+      "loss": 1.0859,
+      "step": 87
+    },
+    {
+      "epoch": 0.07032967032967033,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.0002,
+      "loss": 1.0902,
+      "step": 88
+    },
+    {
+      "epoch": 0.07112887112887113,
+      "grad_norm": 1.6328125,
+      "learning_rate": 0.0002,
+      "loss": 1.0967,
+      "step": 89
+    },
+    {
+      "epoch": 0.07192807192807193,
+      "grad_norm": 0.89453125,
+      "learning_rate": 0.0002,
+      "loss": 1.0817,
+      "step": 90
+    },
+    {
+      "epoch": 0.07272727272727272,
+      "grad_norm": 1.828125,
+      "learning_rate": 0.0002,
+      "loss": 1.088,
+      "step": 91
+    },
+    {
+      "epoch": 0.07352647352647353,
+      "grad_norm": 0.9609375,
+      "learning_rate": 0.0002,
+      "loss": 1.0768,
+      "step": 92
+    },
+    {
+      "epoch": 0.07432567432567433,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.0002,
+      "loss": 1.0927,
+      "step": 93
+    },
+    {
+      "epoch": 0.07512487512487513,
+      "grad_norm": 1.375,
+      "learning_rate": 0.0002,
+      "loss": 1.0891,
+      "step": 94
+    },
+    {
+      "epoch": 0.07592407592407592,
+      "grad_norm": 1.703125,
+      "learning_rate": 0.0002,
+      "loss": 1.0823,
+      "step": 95
+    },
+    {
+      "epoch": 0.07672327672327672,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.0002,
+      "loss": 1.0898,
+      "step": 96
+    },
+    {
+      "epoch": 0.07752247752247753,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.0002,
+      "loss": 1.0714,
+      "step": 97
+    },
+    {
+      "epoch": 0.07832167832167833,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.0757,
+      "step": 98
+    },
+    {
+      "epoch": 0.07912087912087912,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.0002,
+      "loss": 1.0757,
+      "step": 99
+    },
+    {
+      "epoch": 0.07992007992007992,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.0604,
+      "step": 100
+    },
+    {
+      "epoch": 0.08071928071928072,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.0002,
+      "loss": 1.0717,
+      "step": 101
+    },
+    {
+      "epoch": 0.08151848151848153,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.0002,
+      "loss": 1.0629,
+      "step": 102
+    },
+    {
+      "epoch": 0.08231768231768231,
+      "grad_norm": 0.94140625,
+      "learning_rate": 0.0002,
+      "loss": 1.0499,
+      "step": 103
+    },
+    {
+      "epoch": 0.08311688311688312,
+      "grad_norm": 0.9765625,
+      "learning_rate": 0.0002,
+      "loss": 1.0726,
+      "step": 104
+    },
+    {
+      "epoch": 0.08391608391608392,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.0002,
+      "loss": 1.0493,
+      "step": 105
+    },
+    {
+      "epoch": 0.08471528471528472,
+      "grad_norm": 0.890625,
+      "learning_rate": 0.0002,
+      "loss": 1.0698,
+      "step": 106
+    },
+    {
+      "epoch": 0.08551448551448551,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.0002,
+      "loss": 1.0598,
+      "step": 107
+    },
+    {
+      "epoch": 0.08631368631368631,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.0002,
+      "loss": 1.06,
+      "step": 108
+    },
+    {
+      "epoch": 0.08711288711288712,
+      "grad_norm": 0.625,
+      "learning_rate": 0.0002,
+      "loss": 1.046,
+      "step": 109
+    },
+    {
+      "epoch": 0.08791208791208792,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.0002,
+      "loss": 1.066,
+      "step": 110
+    },
+    {
+      "epoch": 0.08871128871128871,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.0002,
+      "loss": 1.0511,
+      "step": 111
+    },
+    {
+      "epoch": 0.08951048951048951,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.0002,
+      "loss": 1.0463,
+      "step": 112
+    },
+    {
+      "epoch": 0.09030969030969031,
+      "grad_norm": 0.8359375,
+      "learning_rate": 0.0002,
+      "loss": 1.0591,
+      "step": 113
+    },
+    {
+      "epoch": 0.09110889110889112,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.0002,
+      "loss": 1.0349,
+      "step": 114
+    },
+    {
+      "epoch": 0.0919080919080919,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0002,
+      "loss": 1.0605,
+      "step": 115
+    },
+    {
+      "epoch": 0.09270729270729271,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.0002,
+      "loss": 1.0461,
+      "step": 116
+    },
+    {
+      "epoch": 0.09350649350649351,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0002,
+      "loss": 1.0437,
+      "step": 117
+    },
+    {
+      "epoch": 0.09430569430569431,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0002,
+      "loss": 1.0402,
+      "step": 118
+    },
+    {
+      "epoch": 0.0951048951048951,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0002,
+      "loss": 1.0477,
+      "step": 119
+    },
+    {
+      "epoch": 0.0959040959040959,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0002,
+      "loss": 1.0404,
+      "step": 120
+    },
+    {
+      "epoch": 0.0967032967032967,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0002,
+      "loss": 1.0353,
+      "step": 121
+    },
+    {
+      "epoch": 0.09750249750249751,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0002,
+      "loss": 1.0339,
+      "step": 122
+    },
+    {
+      "epoch": 0.0983016983016983,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0002,
+      "loss": 1.0462,
+      "step": 123
+    },
+    {
+      "epoch": 0.0991008991008991,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0002,
+      "loss": 1.0377,
+      "step": 124
+    },
+    {
+      "epoch": 0.0999000999000999,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0002,
+      "loss": 1.0252,
+      "step": 125
+    },
+    {
+      "epoch": 0.1006993006993007,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0002,
+      "loss": 1.0261,
+      "step": 126
+    },
+    {
+      "epoch": 0.1014985014985015,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.0002,
+      "loss": 1.0389,
+      "step": 127
+    },
+    {
+      "epoch": 0.1022977022977023,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0002,
+      "loss": 1.027,
+      "step": 128
+    },
+    {
+      "epoch": 0.1030969030969031,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.0002,
+      "loss": 1.0256,
+      "step": 129
+    },
+    {
+      "epoch": 0.1038961038961039,
+      "grad_norm": 0.88671875,
+      "learning_rate": 0.0002,
+      "loss": 1.0353,
+      "step": 130
+    },
+    {
+      "epoch": 0.10469530469530469,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.0371,
+      "step": 131
+    },
+    {
+      "epoch": 0.1054945054945055,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.0002,
+      "loss": 1.0318,
+      "step": 132
+    },
+    {
+      "epoch": 0.1062937062937063,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.0002,
+      "loss": 1.0355,
+      "step": 133
+    },
+    {
+      "epoch": 0.1070929070929071,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.0002,
+      "loss": 1.0378,
+      "step": 134
+    },
+    {
+      "epoch": 0.10789210789210789,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.0002,
+      "loss": 1.0231,
+      "step": 135
+    },
+    {
+      "epoch": 0.10869130869130869,
+      "grad_norm": 0.80078125,
+      "learning_rate": 0.0002,
+      "loss": 1.0222,
+      "step": 136
+    },
+    {
+      "epoch": 0.1094905094905095,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.0002,
+      "loss": 1.0841,
+      "step": 137
+    },
+    {
+      "epoch": 0.1102897102897103,
+      "grad_norm": 1.640625,
+      "learning_rate": 0.0002,
+      "loss": 1.0375,
+      "step": 138
+    },
+    {
+      "epoch": 0.11108891108891109,
+      "grad_norm": 0.90234375,
+      "learning_rate": 0.0002,
+      "loss": 1.0293,
+      "step": 139
+    },
+    {
+      "epoch": 0.11188811188811189,
+      "grad_norm": 1.4140625,
+      "learning_rate": 0.0002,
+      "loss": 1.0401,
+      "step": 140
+    },
+    {
+      "epoch": 0.11268731268731269,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.0002,
+      "loss": 1.0388,
+      "step": 141
+    },
+    {
+      "epoch": 0.1134865134865135,
+      "grad_norm": 1.5234375,
+      "learning_rate": 0.0002,
+      "loss": 1.0298,
+      "step": 142
+    },
+    {
+      "epoch": 0.11428571428571428,
+      "grad_norm": 0.859375,
+      "learning_rate": 0.0002,
+      "loss": 1.0251,
+      "step": 143
+    },
+    {
+      "epoch": 0.11508491508491508,
+      "grad_norm": 1.4921875,
+      "learning_rate": 0.0002,
+      "loss": 1.0393,
+      "step": 144
+    },
+    {
+      "epoch": 0.11588411588411589,
+      "grad_norm": 0.859375,
+      "learning_rate": 0.0002,
+      "loss": 1.017,
+      "step": 145
+    },
+    {
+      "epoch": 0.11668331668331669,
+      "grad_norm": 1.421875,
+      "learning_rate": 0.0002,
+      "loss": 1.0405,
+      "step": 146
+    },
+    {
+      "epoch": 0.11748251748251748,
+      "grad_norm": 0.8671875,
+      "learning_rate": 0.0002,
+      "loss": 1.0274,
+      "step": 147
+    },
+    {
+      "epoch": 0.11828171828171828,
+      "grad_norm": 1.4921875,
+      "learning_rate": 0.0002,
+      "loss": 1.0285,
+      "step": 148
+    },
+    {
+      "epoch": 0.11908091908091908,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.0002,
+      "loss": 1.0318,
+      "step": 149
+    },
+    {
+      "epoch": 0.11988011988011989,
+      "grad_norm": 1.640625,
+      "learning_rate": 0.0002,
+      "loss": 1.0365,
+      "step": 150
+    },
+    {
+      "epoch": 0.12067932067932068,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.0002,
+      "loss": 1.0278,
+      "step": 151
+    },
+    {
+      "epoch": 0.12147852147852148,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.0002,
+      "loss": 1.0288,
+      "step": 152
+    },
+    {
+      "epoch": 0.12227772227772228,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.0002,
+      "loss": 1.0322,
+      "step": 153
+    },
+    {
+      "epoch": 0.12307692307692308,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.0002,
+      "loss": 1.019,
+      "step": 154
+    },
+    {
+      "epoch": 0.12387612387612387,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.0002,
+      "loss": 1.0248,
+      "step": 155
+    },
+    {
+      "epoch": 0.12467532467532468,
+      "grad_norm": 0.96484375,
+      "learning_rate": 0.0002,
+      "loss": 1.0256,
+      "step": 156
+    },
+    {
+      "epoch": 0.12547452547452548,
+      "grad_norm": 0.84375,
+      "learning_rate": 0.0002,
+      "loss": 1.0221,
+      "step": 157
+    },
+    {
+      "epoch": 0.12627372627372627,
+      "grad_norm": 0.875,
+      "learning_rate": 0.0002,
+      "loss": 1.0258,
+      "step": 158
+    },
+    {
+      "epoch": 0.12707292707292708,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.0002,
+      "loss": 1.0205,
+      "step": 159
+    },
+    {
+      "epoch": 0.12787212787212787,
+      "grad_norm": 0.75,
+      "learning_rate": 0.0002,
+      "loss": 1.0135,
+      "step": 160
+    },
+    {
+      "epoch": 0.12867132867132866,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.0002,
+      "loss": 1.0225,
+      "step": 161
+    },
+    {
+      "epoch": 0.12947052947052948,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.0002,
+      "loss": 1.0249,
+      "step": 162
+    },
+    {
+      "epoch": 0.13026973026973027,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0002,
+      "loss": 1.0221,
+      "step": 163
+    },
+    {
+      "epoch": 0.13106893106893108,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.0002,
+      "loss": 1.0263,
+      "step": 164
+    },
+    {
+      "epoch": 0.13186813186813187,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0002,
+      "loss": 1.0179,
+      "step": 165
+    },
+    {
+      "epoch": 0.13266733266733266,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0002,
+      "loss": 1.0019,
+      "step": 166
+    },
+    {
+      "epoch": 0.13346653346653348,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0002,
+      "loss": 1.0215,
+      "step": 167
+    },
+    {
+      "epoch": 0.13426573426573427,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.0002,
+      "loss": 1.0191,
+      "step": 168
+    },
+    {
+      "epoch": 0.13506493506493505,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.0002,
+      "loss": 1.0083,
+      "step": 169
+    },
+    {
+      "epoch": 0.13586413586413587,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0002,
+      "loss": 1.0227,
+      "step": 170
+    },
+    {
+      "epoch": 0.13666333666333666,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0002,
+      "loss": 1.0147,
+      "step": 171
+    },
+    {
+      "epoch": 0.13746253746253748,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0002,
+      "loss": 1.021,
+      "step": 172
+    },
+    {
+      "epoch": 0.13826173826173827,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0002,
+      "loss": 1.0158,
+      "step": 173
+    },
+    {
+      "epoch": 0.13906093906093905,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.0002,
+      "loss": 1.01,
+      "step": 174
+    },
+    {
+      "epoch": 0.13986013986013987,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0002,
+      "loss": 1.0202,
+      "step": 175
+    },
+    {
+      "epoch": 0.14065934065934066,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.0002,
+      "loss": 1.0105,
+      "step": 176
+    },
+    {
+      "epoch": 0.14145854145854145,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0002,
+      "loss": 1.0117,
+      "step": 177
+    },
+    {
+      "epoch": 0.14225774225774226,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0002,
+      "loss": 1.0192,
+      "step": 178
+    },
+    {
+      "epoch": 0.14305694305694305,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.0002,
+      "loss": 1.0107,
+      "step": 179
+    },
+    {
+      "epoch": 0.14385614385614387,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.0002,
+      "loss": 1.0049,
+      "step": 180
+    },
+    {
+      "epoch": 0.14465534465534466,
+      "grad_norm": 0.375,
+      "learning_rate": 0.0002,
+      "loss": 1.0037,
+      "step": 181
+    },
+    {
+      "epoch": 0.14545454545454545,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0002,
+      "loss": 0.9951,
+      "step": 182
+    },
+    {
+      "epoch": 0.14625374625374626,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.0002,
+      "loss": 1.0104,
+      "step": 183
+    },
+    {
+      "epoch": 0.14705294705294705,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0002,
+      "loss": 1.0101,
+      "step": 184
+    },
+    {
+      "epoch": 0.14785214785214784,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0002,
+      "loss": 1.0091,
+      "step": 185
+    },
+    {
+      "epoch": 0.14865134865134866,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0002,
+      "loss": 1.0074,
+      "step": 186
+    },
+    {
+      "epoch": 0.14945054945054945,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.0002,
+      "loss": 1.0123,
+      "step": 187
+    },
+    {
+      "epoch": 0.15024975024975026,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.0002,
+      "loss": 1.0396,
+      "step": 188
+    },
+    {
+      "epoch": 0.15104895104895105,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.0002,
+      "loss": 1.0196,
+      "step": 189
+    },
+    {
+      "epoch": 0.15184815184815184,
+      "grad_norm": 0.796875,
+      "learning_rate": 0.0002,
+      "loss": 1.0079,
+      "step": 190
+    },
+    {
+      "epoch": 0.15264735264735266,
+      "grad_norm": 3.484375,
+      "learning_rate": 0.0002,
+      "loss": 1.0459,
+      "step": 191
+    },
+    {
+      "epoch": 0.15344655344655345,
+      "grad_norm": 0.83984375,
+      "learning_rate": 0.0002,
+      "loss": 1.0136,
+      "step": 192
+    },
+    {
+      "epoch": 0.15424575424575424,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0002,
+      "loss": 1.0378,
+      "step": 193
+    },
+    {
+      "epoch": 0.15504495504495505,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.0002,
+      "loss": 1.0416,
+      "step": 194
+    },
+    {
+      "epoch": 0.15584415584415584,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.0002,
+      "loss": 1.0514,
+      "step": 195
+    },
+    {
+      "epoch": 0.15664335664335666,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.0002,
+      "loss": 1.0412,
+      "step": 196
+    },
+    {
+      "epoch": 0.15744255744255745,
+      "grad_norm": 3.171875,
+      "learning_rate": 0.0002,
+      "loss": 1.0968,
+      "step": 197
+    },
+    {
+      "epoch": 0.15824175824175823,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.0002,
+      "loss": 1.0774,
+      "step": 198
+    },
+    {
+      "epoch": 0.15904095904095905,
+      "grad_norm": 1.7890625,
+      "learning_rate": 0.0002,
+      "loss": 1.0554,
+      "step": 199
+    },
+    {
+      "epoch": 0.15984015984015984,
+      "grad_norm": 1.46875,
+      "learning_rate": 0.0002,
+      "loss": 1.064,
+      "step": 200
+    },
+    {
+      "epoch": 0.16063936063936063,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.0002,
+      "loss": 1.0419,
+      "step": 201
+    },
+    {
+      "epoch": 0.16143856143856145,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.0002,
+      "loss": 1.0409,
+      "step": 202
+    },
+    {
+      "epoch": 0.16223776223776223,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.0002,
+      "loss": 1.0371,
+      "step": 203
+    },
+    {
+      "epoch": 0.16303696303696305,
+      "grad_norm": 0.96875,
+      "learning_rate": 0.0002,
+      "loss": 1.0238,
+      "step": 204
+    },
+    {
+      "epoch": 0.16383616383616384,
+      "grad_norm": 0.97265625,
+      "learning_rate": 0.0002,
+      "loss": 1.0403,
+      "step": 205
+    },
+    {
+      "epoch": 0.16463536463536463,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.0002,
+      "loss": 1.0403,
+      "step": 206
+    },
+    {
+      "epoch": 0.16543456543456544,
+      "grad_norm": 0.95703125,
+      "learning_rate": 0.0002,
+      "loss": 1.0254,
+      "step": 207
+    },
+    {
+      "epoch": 0.16623376623376623,
+      "grad_norm": 0.8984375,
+      "learning_rate": 0.0002,
+      "loss": 1.033,
+      "step": 208
+    },
+    {
+      "epoch": 0.16703296703296702,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.0002,
+      "loss": 1.0272,
+      "step": 209
+    },
+    {
+      "epoch": 0.16783216783216784,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.0002,
+      "loss": 1.0248,
+      "step": 210
+    },
+    {
+      "epoch": 0.16863136863136863,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.0002,
+      "loss": 1.0287,
+      "step": 211
+    },
+    {
+      "epoch": 0.16943056943056944,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.0002,
+      "loss": 1.0129,
+      "step": 212
+    },
+    {
+      "epoch": 0.17022977022977023,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.0002,
+      "loss": 1.0276,
+      "step": 213
+    },
+    {
+      "epoch": 0.17102897102897102,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0002,
+      "loss": 1.0164,
+      "step": 214
+    },
+    {
+      "epoch": 0.17182817182817184,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0002,
+      "loss": 1.0174,
+      "step": 215
+    },
+    {
+      "epoch": 0.17262737262737263,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0002,
+      "loss": 1.0293,
+      "step": 216
+    },
+    {
+      "epoch": 0.17342657342657342,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0002,
+      "loss": 1.017,
+      "step": 217
+    },
+    {
+      "epoch": 0.17422577422577423,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0002,
+      "loss": 1.0159,
+      "step": 218
+    },
+    {
+      "epoch": 0.17502497502497502,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0002,
+      "loss": 1.0051,
+      "step": 219
+    },
+    {
+      "epoch": 0.17582417582417584,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0002,
+      "loss": 1.0038,
+      "step": 220
+    },
+    {
+      "epoch": 0.17662337662337663,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0002,
+      "loss": 1.0147,
+      "step": 221
+    },
+    {
+      "epoch": 0.17742257742257742,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0002,
+      "loss": 1.0116,
+      "step": 222
+    },
+    {
+      "epoch": 0.17822177822177823,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0002,
+      "loss": 1.0103,
+      "step": 223
+    },
+    {
+      "epoch": 0.17902097902097902,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0002,
+      "loss": 1.0087,
+      "step": 224
+    },
+    {
+      "epoch": 0.1798201798201798,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0002,
+      "loss": 1.0171,
+      "step": 225
+    },
+    {
+      "epoch": 0.18061938061938063,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9997,
+      "step": 226
+    },
+    {
+      "epoch": 0.18141858141858141,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0002,
+      "loss": 1.0111,
+      "step": 227
+    },
+    {
+      "epoch": 0.18221778221778223,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0002,
+      "loss": 1.0054,
+      "step": 228
+    },
+    {
+      "epoch": 0.18301698301698302,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9988,
+      "step": 229
+    },
+    {
+      "epoch": 0.1838161838161838,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0002,
+      "loss": 1.0008,
+      "step": 230
+    },
+    {
+      "epoch": 0.18461538461538463,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0002,
+      "loss": 1.0037,
+      "step": 231
+    },
+    {
+      "epoch": 0.18541458541458541,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.0002,
+      "loss": 1.0063,
+      "step": 232
+    },
+    {
+      "epoch": 0.1862137862137862,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9908,
+      "step": 233
+    },
+    {
+      "epoch": 0.18701298701298702,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.0002,
+      "loss": 1.0117,
+      "step": 234
+    },
+    {
+      "epoch": 0.1878121878121878,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.0002,
+      "loss": 1.0002,
+      "step": 235
+    },
+    {
+      "epoch": 0.18861138861138863,
+      "grad_norm": 1.0,
+      "learning_rate": 0.0002,
+      "loss": 0.9998,
+      "step": 236
+    },
+    {
+      "epoch": 0.18941058941058941,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.0002,
+      "loss": 1.0021,
+      "step": 237
+    },
+    {
+      "epoch": 0.1902097902097902,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.0002,
+      "loss": 0.9999,
+      "step": 238
+    },
+    {
+      "epoch": 0.19100899100899102,
+      "grad_norm": 0.625,
+      "learning_rate": 0.0002,
+      "loss": 0.9985,
+      "step": 239
+    },
+    {
+      "epoch": 0.1918081918081918,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.0002,
+      "loss": 1.0006,
+      "step": 240
+    },
+    {
+      "epoch": 0.1926073926073926,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0002,
+      "loss": 0.9983,
+      "step": 241
+    },
+    {
+      "epoch": 0.1934065934065934,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0002,
+      "loss": 1.0014,
+      "step": 242
+    },
+    {
+      "epoch": 0.1942057942057942,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.0002,
+      "loss": 1.001,
+      "step": 243
+    },
+    {
+      "epoch": 0.19500499500499502,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0002,
+      "loss": 0.9903,
+      "step": 244
+    },
+    {
+      "epoch": 0.1958041958041958,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.0002,
+      "loss": 0.995,
+      "step": 245
+    },
+    {
+      "epoch": 0.1966033966033966,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0002,
+      "loss": 0.9974,
+      "step": 246
+    },
+    {
+      "epoch": 0.1974025974025974,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.0002,
+      "loss": 0.9987,
+      "step": 247
+    },
+    {
+      "epoch": 0.1982017982017982,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0002,
+      "loss": 1.0066,
+      "step": 248
+    },
+    {
+      "epoch": 0.199000999000999,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.0002,
+      "loss": 0.99,
+      "step": 249
+    },
+    {
+      "epoch": 0.1998001998001998,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9969,
+      "step": 250
+    },
+    {
+      "epoch": 0.2005994005994006,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0002,
+      "loss": 0.9983,
+      "step": 251
+    },
+    {
+      "epoch": 0.2013986013986014,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0002,
+      "loss": 1.0044,
+      "step": 252
+    },
+    {
+      "epoch": 0.2021978021978022,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0002,
+      "loss": 0.9928,
+      "step": 253
+    },
+    {
+      "epoch": 0.202997002997003,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0002,
+      "loss": 0.9865,
+      "step": 254
+    },
+    {
+      "epoch": 0.2037962037962038,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9919,
+      "step": 255
+    },
+    {
+      "epoch": 0.2045954045954046,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.0002,
+      "loss": 0.998,
+      "step": 256
+    },
+    {
+      "epoch": 0.20539460539460538,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.0002,
+      "loss": 0.9927,
+      "step": 257
+    },
+    {
+      "epoch": 0.2061938061938062,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9888,
+      "step": 258
+    },
+    {
+      "epoch": 0.206993006993007,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0002,
+      "loss": 0.9876,
+      "step": 259
+    },
+    {
+      "epoch": 0.2077922077922078,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9935,
+      "step": 260
+    },
+    {
+      "epoch": 0.2085914085914086,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0002,
+      "loss": 0.9843,
+      "step": 261
+    },
+    {
+      "epoch": 0.20939060939060938,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.0002,
+      "loss": 0.9917,
+      "step": 262
+    },
+    {
+      "epoch": 0.2101898101898102,
+      "grad_norm": 0.93359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9974,
+      "step": 263
+    },
+    {
+      "epoch": 0.210989010989011,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.0002,
+      "loss": 1.008,
+      "step": 264
+    },
+    {
+      "epoch": 0.21178821178821178,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0002,
+      "loss": 0.9963,
+      "step": 265
+    },
+    {
+      "epoch": 0.2125874125874126,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.0002,
+      "loss": 1.0049,
+      "step": 266
+    },
+    {
+      "epoch": 0.21338661338661338,
+      "grad_norm": 1.6875,
+      "learning_rate": 0.0002,
+      "loss": 0.9916,
+      "step": 267
+    },
+    {
+      "epoch": 0.2141858141858142,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.0002,
+      "loss": 0.9931,
+      "step": 268
+    },
+    {
+      "epoch": 0.214985014985015,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0002,
+      "loss": 1.0043,
+      "step": 269
+    },
+    {
+      "epoch": 0.21578421578421578,
+      "grad_norm": 0.99609375,
+      "learning_rate": 0.0002,
+      "loss": 0.9899,
+      "step": 270
+    },
+    {
+      "epoch": 0.2165834165834166,
+      "grad_norm": 5.3125,
+      "learning_rate": 0.0002,
+      "loss": 1.0724,
+      "step": 271
+    },
+    {
+      "epoch": 0.21738261738261738,
+      "grad_norm": 5.0625,
+      "learning_rate": 0.0002,
+      "loss": 1.0816,
+      "step": 272
+    },
+    {
+      "epoch": 0.21818181818181817,
+      "grad_norm": 1.734375,
+      "learning_rate": 0.0002,
+      "loss": 1.0166,
+      "step": 273
+    },
+    {
+      "epoch": 0.218981018981019,
+      "grad_norm": 3.703125,
+      "learning_rate": 0.0002,
+      "loss": 1.0355,
+      "step": 274
+    },
+    {
+      "epoch": 0.21978021978021978,
+      "grad_norm": 3.1875,
+      "learning_rate": 0.0002,
+      "loss": 1.0681,
+      "step": 275
+    },
+    {
+      "epoch": 0.2205794205794206,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.0401,
+      "step": 276
+    },
+    {
+      "epoch": 0.22137862137862138,
+      "grad_norm": 1.6796875,
+      "learning_rate": 0.0002,
+      "loss": 1.0299,
+      "step": 277
+    },
+    {
+      "epoch": 0.22217782217782217,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.0002,
+      "loss": 1.0256,
+      "step": 278
+    },
+    {
+      "epoch": 0.222977022977023,
+      "grad_norm": 1.125,
+      "learning_rate": 0.0002,
+      "loss": 1.023,
+      "step": 279
+    },
+    {
+      "epoch": 0.22377622377622378,
+      "grad_norm": 1.734375,
+      "learning_rate": 0.0002,
+      "loss": 1.0184,
+      "step": 280
+    },
+    {
+      "epoch": 0.22457542457542456,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.0154,
+      "step": 281
+    },
+    {
+      "epoch": 0.22537462537462538,
+      "grad_norm": 1.5390625,
+      "learning_rate": 0.0002,
+      "loss": 1.0054,
+      "step": 282
+    },
+    {
+      "epoch": 0.22617382617382617,
+      "grad_norm": 1.125,
+      "learning_rate": 0.0002,
+      "loss": 1.0123,
+      "step": 283
+    },
+    {
+      "epoch": 0.226973026973027,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.0002,
+      "loss": 1.0089,
+      "step": 284
+    },
+    {
+      "epoch": 0.22777222777222778,
+      "grad_norm": 1.546875,
+      "learning_rate": 0.0002,
+      "loss": 1.0159,
+      "step": 285
+    },
+    {
+      "epoch": 0.22857142857142856,
+      "grad_norm": 1.5078125,
+      "learning_rate": 0.0002,
+      "loss": 1.0028,
+      "step": 286
+    },
+    {
+      "epoch": 0.22937062937062938,
+      "grad_norm": 1.3515625,
+      "learning_rate": 0.0002,
+      "loss": 1.0018,
+      "step": 287
+    },
+    {
+      "epoch": 0.23016983016983017,
+      "grad_norm": 1.3984375,
+      "learning_rate": 0.0002,
+      "loss": 1.0481,
+      "step": 288
+    },
+    {
+      "epoch": 0.23096903096903096,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.0002,
+      "loss": 1.0038,
+      "step": 289
+    },
+    {
+      "epoch": 0.23176823176823177,
+      "grad_norm": 1.375,
+      "learning_rate": 0.0002,
+      "loss": 1.0024,
+      "step": 290
+    },
+    {
+      "epoch": 0.23256743256743256,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9914,
+      "step": 291
+    },
+    {
+      "epoch": 0.23336663336663338,
+      "grad_norm": 1.4296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9959,
+      "step": 292
+    },
+    {
+      "epoch": 0.23416583416583417,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.0002,
+      "loss": 1.0094,
+      "step": 293
+    },
+    {
+      "epoch": 0.23496503496503496,
+      "grad_norm": 1.125,
+      "learning_rate": 0.0002,
+      "loss": 0.9992,
+      "step": 294
+    },
+    {
+      "epoch": 0.23576423576423577,
+      "grad_norm": 0.984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9999,
+      "step": 295
+    },
+    {
+      "epoch": 0.23656343656343656,
+      "grad_norm": 0.96484375,
+      "learning_rate": 0.0002,
+      "loss": 0.9901,
+      "step": 296
+    },
+    {
+      "epoch": 0.23736263736263735,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.0002,
+      "loss": 0.9891,
+      "step": 297
+    },
+    {
+      "epoch": 0.23816183816183817,
+      "grad_norm": 0.796875,
+      "learning_rate": 0.0002,
+      "loss": 1.003,
+      "step": 298
+    },
+    {
+      "epoch": 0.23896103896103896,
+      "grad_norm": 0.625,
+      "learning_rate": 0.0002,
+      "loss": 1.0007,
+      "step": 299
+    },
+    {
+      "epoch": 0.23976023976023977,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9954,
+      "step": 300
+    },
+    {
+      "epoch": 0.24055944055944056,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.0002,
+      "loss": 0.9901,
+      "step": 301
+    },
+    {
+      "epoch": 0.24135864135864135,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.0002,
+      "loss": 0.9953,
+      "step": 302
+    },
+    {
+      "epoch": 0.24215784215784217,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9948,
+      "step": 303
+    },
+    {
+      "epoch": 0.24295704295704296,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0002,
+      "loss": 0.9864,
+      "step": 304
+    },
+    {
+      "epoch": 0.24375624375624375,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0002,
+      "loss": 0.9957,
+      "step": 305
+    },
+    {
+      "epoch": 0.24455544455544456,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9828,
+      "step": 306
+    },
+    {
+      "epoch": 0.24535464535464535,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9981,
+      "step": 307
+    },
+    {
+      "epoch": 0.24615384615384617,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9915,
+      "step": 308
+    },
+    {
+      "epoch": 0.24695304695304696,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0002,
+      "loss": 0.9876,
+      "step": 309
+    },
+    {
+      "epoch": 0.24775224775224775,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0002,
+      "loss": 0.994,
+      "step": 310
+    },
+    {
+      "epoch": 0.24855144855144856,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0002,
+      "loss": 0.9861,
+      "step": 311
+    },
+    {
+      "epoch": 0.24935064935064935,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.0002,
+      "loss": 0.9949,
+      "step": 312
+    },
+    {
+      "epoch": 0.25014985014985014,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0002,
+      "loss": 0.987,
+      "step": 313
+    },
+    {
+      "epoch": 0.25094905094905096,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9863,
+      "step": 314
+    },
+    {
+      "epoch": 0.2517482517482518,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0002,
+      "loss": 0.9894,
+      "step": 315
+    },
+    {
+      "epoch": 0.25254745254745253,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.0002,
+      "loss": 0.9915,
+      "step": 316
+    },
+    {
+      "epoch": 0.25334665334665335,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.0002,
+      "loss": 0.9934,
+      "step": 317
+    },
+    {
+      "epoch": 0.25414585414585417,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.0002,
+      "loss": 0.996,
+      "step": 318
+    },
+    {
+      "epoch": 0.2549450549450549,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9822,
+      "step": 319
+    },
+    {
+      "epoch": 0.25574425574425574,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.0002,
+      "loss": 1.0072,
+      "step": 320
+    },
+    {
+      "epoch": 0.25654345654345656,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9891,
+      "step": 321
+    },
+    {
+      "epoch": 0.2573426573426573,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0002,
+      "loss": 0.9901,
+      "step": 322
+    },
+    {
+      "epoch": 0.25814185814185814,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0002,
+      "loss": 0.9991,
+      "step": 323
+    },
+    {
+      "epoch": 0.25894105894105895,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0002,
+      "loss": 0.9899,
+      "step": 324
+    },
+    {
+      "epoch": 0.2597402597402597,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.0002,
+      "loss": 0.9921,
+      "step": 325
+    },
+    {
+      "epoch": 0.26053946053946053,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9796,
+      "step": 326
+    },
+    {
+      "epoch": 0.26133866133866135,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9879,
+      "step": 327
+    },
+    {
+      "epoch": 0.26213786213786217,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9948,
+      "step": 328
+    },
+    {
+      "epoch": 0.2629370629370629,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9917,
+      "step": 329
+    },
+    {
+      "epoch": 0.26373626373626374,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9928,
+      "step": 330
+    },
+    {
+      "epoch": 0.26453546453546456,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0002,
+      "loss": 0.9894,
+      "step": 331
+    },
+    {
+      "epoch": 0.2653346653346653,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0002,
+      "loss": 0.9864,
+      "step": 332
+    },
+    {
+      "epoch": 0.26613386613386614,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0002,
+      "loss": 0.9796,
+      "step": 333
+    },
+    {
+      "epoch": 0.26693306693306695,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.0002,
+      "loss": 0.9782,
+      "step": 334
+    },
+    {
+      "epoch": 0.2677322677322677,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9848,
+      "step": 335
+    },
+    {
+      "epoch": 0.26853146853146853,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9848,
+      "step": 336
+    },
+    {
+      "epoch": 0.26933066933066935,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9832,
+      "step": 337
+    },
+    {
+      "epoch": 0.2701298701298701,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9817,
+      "step": 338
+    },
+    {
+      "epoch": 0.2709290709290709,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0002,
+      "loss": 0.9831,
+      "step": 339
+    },
+    {
+      "epoch": 0.27172827172827174,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9785,
+      "step": 340
+    },
+    {
+      "epoch": 0.2725274725274725,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.0002,
+      "loss": 0.989,
+      "step": 341
+    },
+    {
+      "epoch": 0.2733266733266733,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.0002,
+      "loss": 0.9831,
+      "step": 342
+    },
+    {
+      "epoch": 0.27412587412587414,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.0002,
+      "loss": 0.9801,
+      "step": 343
+    },
+    {
+      "epoch": 0.27492507492507495,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9868,
+      "step": 344
+    },
+    {
+      "epoch": 0.2757242757242757,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0002,
+      "loss": 0.989,
+      "step": 345
+    },
+    {
+      "epoch": 0.27652347652347653,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9765,
+      "step": 346
+    },
+    {
+      "epoch": 0.27732267732267735,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9844,
+      "step": 347
+    },
+    {
+      "epoch": 0.2781218781218781,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9885,
+      "step": 348
+    },
+    {
+      "epoch": 0.2789210789210789,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0002,
+      "loss": 0.9754,
+      "step": 349
+    },
+    {
+      "epoch": 0.27972027972027974,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0002,
+      "loss": 0.9797,
+      "step": 350
+    },
+    {
+      "epoch": 0.2805194805194805,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0002,
+      "loss": 0.9778,
+      "step": 351
+    },
+    {
+      "epoch": 0.2813186813186813,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0002,
+      "loss": 0.9793,
+      "step": 352
+    },
+    {
+      "epoch": 0.28211788211788213,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.0002,
+      "loss": 0.9795,
+      "step": 353
+    },
+    {
+      "epoch": 0.2829170829170829,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0002,
+      "loss": 0.983,
+      "step": 354
+    },
+    {
+      "epoch": 0.2837162837162837,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0002,
+      "loss": 0.976,
+      "step": 355
+    },
+    {
+      "epoch": 0.28451548451548453,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0002,
+      "loss": 0.9719,
+      "step": 356
+    },
+    {
+      "epoch": 0.2853146853146853,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0002,
+      "loss": 0.9774,
+      "step": 357
+    },
+    {
+      "epoch": 0.2861138861138861,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0002,
+      "loss": 0.9754,
+      "step": 358
+    },
+    {
+      "epoch": 0.2869130869130869,
+      "grad_norm": 0.90625,
+      "learning_rate": 0.0002,
+      "loss": 1.0023,
+      "step": 359
+    },
+    {
+      "epoch": 0.28771228771228774,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9793,
+      "step": 360
+    },
+    {
+      "epoch": 0.2885114885114885,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0002,
+      "loss": 0.982,
+      "step": 361
+    },
+    {
+      "epoch": 0.2893106893106893,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0002,
+      "loss": 0.9803,
+      "step": 362
+    },
+    {
+      "epoch": 0.29010989010989013,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0002,
+      "loss": 0.977,
+      "step": 363
+    },
+    {
+      "epoch": 0.2909090909090909,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0002,
+      "loss": 0.985,
+      "step": 364
+    },
+    {
+      "epoch": 0.2917082917082917,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9818,
+      "step": 365
+    },
+    {
+      "epoch": 0.29250749250749253,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9947,
+      "step": 366
+    },
+    {
+      "epoch": 0.2933066933066933,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0002,
+      "loss": 0.981,
+      "step": 367
+    },
+    {
+      "epoch": 0.2941058941058941,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9836,
+      "step": 368
+    },
+    {
+      "epoch": 0.2949050949050949,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9741,
+      "step": 369
+    },
+    {
+      "epoch": 0.2957042957042957,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0002,
+      "loss": 0.976,
+      "step": 370
+    },
+    {
+      "epoch": 0.2965034965034965,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0002,
+      "loss": 0.9789,
+      "step": 371
+    },
+    {
+      "epoch": 0.2973026973026973,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9917,
+      "step": 372
+    },
+    {
+      "epoch": 0.2981018981018981,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.0002,
+      "loss": 0.9778,
+      "step": 373
+    },
+    {
+      "epoch": 0.2989010989010989,
+      "grad_norm": 0.90625,
+      "learning_rate": 0.0002,
+      "loss": 0.9772,
+      "step": 374
+    },
+    {
+      "epoch": 0.2997002997002997,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.0002,
+      "loss": 0.9921,
+      "step": 375
+    },
+    {
+      "epoch": 0.3004995004995005,
+      "grad_norm": 0.75,
+      "learning_rate": 0.0002,
+      "loss": 0.9764,
+      "step": 376
+    },
+    {
+      "epoch": 0.3012987012987013,
+      "grad_norm": 1.640625,
+      "learning_rate": 0.0002,
+      "loss": 0.9932,
+      "step": 377
+    },
+    {
+      "epoch": 0.3020979020979021,
+      "grad_norm": 1.0,
+      "learning_rate": 0.0002,
+      "loss": 0.9853,
+      "step": 378
+    },
+    {
+      "epoch": 0.3028971028971029,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.0002,
+      "loss": 0.9954,
+      "step": 379
+    },
+    {
+      "epoch": 0.3036963036963037,
+      "grad_norm": 0.93359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9834,
+      "step": 380
+    },
+    {
+      "epoch": 0.3044955044955045,
+      "grad_norm": 2.78125,
+      "learning_rate": 0.0002,
+      "loss": 1.025,
+      "step": 381
+    },
+    {
+      "epoch": 0.3052947052947053,
+      "grad_norm": 1.8046875,
+      "learning_rate": 0.0002,
+      "loss": 1.0111,
+      "step": 382
+    },
+    {
+      "epoch": 0.3060939060939061,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9947,
+      "step": 383
+    },
+    {
+      "epoch": 0.3068931068931069,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.0002,
+      "loss": 0.9982,
+      "step": 384
+    },
+    {
+      "epoch": 0.3076923076923077,
+      "grad_norm": 1.53125,
+      "learning_rate": 0.0002,
+      "loss": 1.0028,
+      "step": 385
+    },
+    {
+      "epoch": 0.30849150849150847,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.0002,
+      "loss": 0.9915,
+      "step": 386
+    },
+    {
+      "epoch": 0.3092907092907093,
+      "grad_norm": 1.34375,
+      "learning_rate": 0.0002,
+      "loss": 0.9934,
+      "step": 387
+    },
+    {
+      "epoch": 0.3100899100899101,
+      "grad_norm": 0.96875,
+      "learning_rate": 0.0002,
+      "loss": 0.9919,
+      "step": 388
+    },
+    {
+      "epoch": 0.31088911088911086,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.0002,
+      "loss": 0.9982,
+      "step": 389
+    },
+    {
+      "epoch": 0.3116883116883117,
+      "grad_norm": 0.90234375,
+      "learning_rate": 0.0002,
+      "loss": 0.9804,
+      "step": 390
+    },
+    {
+      "epoch": 0.3124875124875125,
+      "grad_norm": 0.94140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9811,
+      "step": 391
+    },
+    {
+      "epoch": 0.3132867132867133,
+      "grad_norm": 1.4765625,
+      "learning_rate": 0.0002,
+      "loss": 0.9871,
+      "step": 392
+    },
+    {
+      "epoch": 0.3140859140859141,
+      "grad_norm": 0.9453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9841,
+      "step": 393
+    },
+    {
+      "epoch": 0.3148851148851149,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.0002,
+      "loss": 0.989,
+      "step": 394
+    },
+    {
+      "epoch": 0.3156843156843157,
+      "grad_norm": 0.9140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9861,
+      "step": 395
+    },
+    {
+      "epoch": 0.31648351648351647,
+      "grad_norm": 0.94921875,
+      "learning_rate": 0.0002,
+      "loss": 0.9875,
+      "step": 396
+    },
+    {
+      "epoch": 0.3172827172827173,
+      "grad_norm": 0.89453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9877,
+      "step": 397
+    },
+    {
+      "epoch": 0.3180819180819181,
+      "grad_norm": 0.8203125,
+      "learning_rate": 0.0002,
+      "loss": 0.9823,
+      "step": 398
+    },
+    {
+      "epoch": 0.31888111888111886,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.0002,
+      "loss": 0.9777,
+      "step": 399
+    },
+    {
+      "epoch": 0.3196803196803197,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.0002,
+      "loss": 0.9886,
+      "step": 400
+    },
+    {
+      "epoch": 0.3204795204795205,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.0002,
+      "loss": 0.9856,
+      "step": 401
+    },
+    {
+      "epoch": 0.32127872127872126,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0002,
+      "loss": 0.9786,
+      "step": 402
+    },
+    {
+      "epoch": 0.3220779220779221,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0002,
+      "loss": 0.9778,
+      "step": 403
+    },
+    {
+      "epoch": 0.3228771228771229,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9796,
+      "step": 404
+    },
+    {
+      "epoch": 0.32367632367632365,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9739,
+      "step": 405
+    },
+    {
+      "epoch": 0.32447552447552447,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.0002,
+      "loss": 0.9844,
+      "step": 406
+    },
+    {
+      "epoch": 0.3252747252747253,
+      "grad_norm": 0.87109375,
+      "learning_rate": 0.0002,
+      "loss": 0.9847,
+      "step": 407
+    },
+    {
+      "epoch": 0.3260739260739261,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.0002,
+      "loss": 0.9812,
+      "step": 408
+    },
+    {
+      "epoch": 0.32687312687312686,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0002,
+      "loss": 0.9782,
+      "step": 409
+    },
+    {
+      "epoch": 0.3276723276723277,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9722,
+      "step": 410
+    },
+    {
+      "epoch": 0.3284715284715285,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.0002,
+      "loss": 0.9673,
+      "step": 411
+    },
+    {
+      "epoch": 0.32927072927072926,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0002,
+      "loss": 0.9775,
+      "step": 412
+    },
+    {
+      "epoch": 0.3300699300699301,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9737,
+      "step": 413
+    },
+    {
+      "epoch": 0.3308691308691309,
+      "grad_norm": 0.83203125,
+      "learning_rate": 0.0002,
+      "loss": 0.9722,
+      "step": 414
+    },
+    {
+      "epoch": 0.33166833166833165,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.0002,
+      "loss": 0.975,
+      "step": 415
+    },
+    {
+      "epoch": 0.33246753246753247,
+      "grad_norm": 0.9375,
+      "learning_rate": 0.0002,
+      "loss": 0.9811,
+      "step": 416
+    },
+    {
+      "epoch": 0.3332667332667333,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9751,
+      "step": 417
+    },
+    {
+      "epoch": 0.33406593406593404,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9806,
+      "step": 418
+    },
+    {
+      "epoch": 0.33486513486513486,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9752,
+      "step": 419
+    },
+    {
+      "epoch": 0.3356643356643357,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0002,
+      "loss": 0.966,
+      "step": 420
+    },
+    {
+      "epoch": 0.33646353646353644,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0002,
+      "loss": 0.9753,
+      "step": 421
+    },
+    {
+      "epoch": 0.33726273726273726,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9634,
+      "step": 422
+    },
+    {
+      "epoch": 0.33806193806193807,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9671,
+      "step": 423
+    },
+    {
+      "epoch": 0.3388611388611389,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9711,
+      "step": 424
+    },
+    {
+      "epoch": 0.33966033966033965,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0002,
+      "loss": 1.013,
+      "step": 425
+    },
+    {
+      "epoch": 0.34045954045954047,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0002,
+      "loss": 0.9675,
+      "step": 426
+    },
+    {
+      "epoch": 0.3412587412587413,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0002,
+      "loss": 0.9732,
+      "step": 427
+    },
+    {
+      "epoch": 0.34205794205794204,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0002,
+      "loss": 0.9699,
+      "step": 428
+    },
+    {
+      "epoch": 0.34285714285714286,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9758,
+      "step": 429
+    },
+    {
+      "epoch": 0.3436563436563437,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9761,
+      "step": 430
+    },
+    {
+      "epoch": 0.34445554445554444,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0002,
+      "loss": 0.9673,
+      "step": 431
+    },
+    {
+      "epoch": 0.34525474525474525,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0002,
+      "loss": 0.9802,
+      "step": 432
+    },
+    {
+      "epoch": 0.34605394605394607,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0002,
+      "loss": 0.9703,
+      "step": 433
+    },
+    {
+      "epoch": 0.34685314685314683,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0002,
+      "loss": 0.9706,
+      "step": 434
+    },
+    {
+      "epoch": 0.34765234765234765,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0002,
+      "loss": 0.981,
+      "step": 435
+    },
+    {
+      "epoch": 0.34845154845154847,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9759,
+      "step": 436
+    },
+    {
+      "epoch": 0.3492507492507492,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0002,
+      "loss": 0.9593,
+      "step": 437
+    },
+    {
+      "epoch": 0.35004995004995004,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0002,
+      "loss": 0.9729,
+      "step": 438
+    },
+    {
+      "epoch": 0.35084915084915086,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9667,
+      "step": 439
+    },
+    {
+      "epoch": 0.3516483516483517,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0002,
+      "loss": 0.9724,
+      "step": 440
+    },
+    {
+      "epoch": 0.35244755244755244,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.0002,
+      "loss": 0.9696,
+      "step": 441
+    },
+    {
+      "epoch": 0.35324675324675325,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0002,
+      "loss": 0.973,
+      "step": 442
+    },
+    {
+      "epoch": 0.35404595404595407,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0002,
+      "loss": 0.9721,
+      "step": 443
+    },
+    {
+      "epoch": 0.35484515484515483,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0002,
+      "loss": 0.9729,
+      "step": 444
+    },
+    {
+      "epoch": 0.35564435564435565,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0002,
+      "loss": 0.9707,
+      "step": 445
+    },
+    {
+      "epoch": 0.35644355644355646,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.0002,
+      "loss": 0.9698,
+      "step": 446
+    },
+    {
+      "epoch": 0.3572427572427572,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9627,
+      "step": 447
+    },
+    {
+      "epoch": 0.35804195804195804,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9746,
+      "step": 448
+    },
+    {
+      "epoch": 0.35884115884115886,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9672,
+      "step": 449
+    },
+    {
+      "epoch": 0.3596403596403596,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0002,
+      "loss": 0.9622,
+      "step": 450
+    },
+    {
+      "epoch": 0.36043956043956044,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0002,
+      "loss": 0.9717,
+      "step": 451
+    },
+    {
+      "epoch": 0.36123876123876125,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9732,
+      "step": 452
+    },
+    {
+      "epoch": 0.362037962037962,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.0002,
+      "loss": 0.988,
+      "step": 453
+    },
+    {
+      "epoch": 0.36283716283716283,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0002,
+      "loss": 0.9639,
+      "step": 454
+    },
+    {
+      "epoch": 0.36363636363636365,
+      "grad_norm": 0.8125,
+      "learning_rate": 0.0002,
+      "loss": 0.9725,
+      "step": 455
+    },
+    {
+      "epoch": 0.36443556443556446,
+      "grad_norm": 1.484375,
+      "learning_rate": 0.0002,
+      "loss": 0.9809,
+      "step": 456
+    },
+    {
+      "epoch": 0.3652347652347652,
+      "grad_norm": 0.96875,
+      "learning_rate": 0.0002,
+      "loss": 0.9707,
+      "step": 457
+    },
+    {
+      "epoch": 0.36603396603396604,
+      "grad_norm": 0.94921875,
+      "learning_rate": 0.0002,
+      "loss": 0.9667,
+      "step": 458
+    },
+    {
+      "epoch": 0.36683316683316686,
+      "grad_norm": 0.97265625,
+      "learning_rate": 0.0002,
+      "loss": 0.9771,
+      "step": 459
+    },
+    {
+      "epoch": 0.3676323676323676,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9768,
+      "step": 460
+    },
+    {
+      "epoch": 0.36843156843156843,
+      "grad_norm": 0.96875,
+      "learning_rate": 0.0002,
+      "loss": 0.973,
+      "step": 461
+    },
+    {
+      "epoch": 0.36923076923076925,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.0002,
+      "loss": 0.9686,
+      "step": 462
+    },
+    {
+      "epoch": 0.37002997002997,
+      "grad_norm": 0.98828125,
+      "learning_rate": 0.0002,
+      "loss": 0.9802,
+      "step": 463
+    },
+    {
+      "epoch": 0.37082917082917083,
+      "grad_norm": 1.453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9825,
+      "step": 464
+    },
+    {
+      "epoch": 0.37162837162837165,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9736,
+      "step": 465
+    },
+    {
+      "epoch": 0.3724275724275724,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.0002,
+      "loss": 0.9777,
+      "step": 466
+    },
+    {
+      "epoch": 0.3732267732267732,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.0002,
+      "loss": 1.0053,
+      "step": 467
+    },
+    {
+      "epoch": 0.37402597402597404,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.0002,
+      "loss": 0.9675,
+      "step": 468
+    },
+    {
+      "epoch": 0.3748251748251748,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9852,
+      "step": 469
+    },
+    {
+      "epoch": 0.3756243756243756,
+      "grad_norm": 0.9609375,
+      "learning_rate": 0.0002,
+      "loss": 0.9758,
+      "step": 470
+    },
+    {
+      "epoch": 0.37642357642357643,
+      "grad_norm": 4.25,
+      "learning_rate": 0.0002,
+      "loss": 1.0281,
+      "step": 471
+    },
+    {
+      "epoch": 0.37722277722277725,
+      "grad_norm": 3.8125,
+      "learning_rate": 0.0002,
+      "loss": 1.0331,
+      "step": 472
+    },
+    {
+      "epoch": 0.378021978021978,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.0002,
+      "loss": 0.9964,
+      "step": 473
+    },
+    {
+      "epoch": 0.37882117882117883,
+      "grad_norm": 1.7890625,
+      "learning_rate": 0.0002,
+      "loss": 0.9965,
+      "step": 474
+    },
+    {
+      "epoch": 0.37962037962037964,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9866,
+      "step": 475
+    },
+    {
+      "epoch": 0.3804195804195804,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.0002,
+      "loss": 0.9816,
+      "step": 476
+    },
+    {
+      "epoch": 0.3812187812187812,
+      "grad_norm": 1.484375,
+      "learning_rate": 0.0002,
+      "loss": 0.9933,
+      "step": 477
+    },
+    {
+      "epoch": 0.38201798201798204,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.0002,
+      "loss": 0.9881,
+      "step": 478
+    },
+    {
+      "epoch": 0.3828171828171828,
+      "grad_norm": 1.6328125,
+      "learning_rate": 0.0002,
+      "loss": 0.9979,
+      "step": 479
+    },
+    {
+      "epoch": 0.3836163836163836,
+      "grad_norm": 0.91015625,
+      "learning_rate": 0.0002,
+      "loss": 0.9754,
+      "step": 480
+    },
+    {
+      "epoch": 0.38441558441558443,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.0002,
+      "loss": 0.9872,
+      "step": 481
+    },
+    {
+      "epoch": 0.3852147852147852,
+      "grad_norm": 1.53125,
+      "learning_rate": 0.0002,
+      "loss": 0.9851,
+      "step": 482
+    },
+    {
+      "epoch": 0.386013986013986,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9769,
+      "step": 483
+    },
+    {
+      "epoch": 0.3868131868131868,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.0002,
+      "loss": 0.982,
+      "step": 484
+    },
+    {
+      "epoch": 0.3876123876123876,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.0002,
+      "loss": 0.9722,
+      "step": 485
+    },
+    {
+      "epoch": 0.3884115884115884,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.0002,
+      "loss": 0.9848,
+      "step": 486
+    },
+    {
+      "epoch": 0.3892107892107892,
+      "grad_norm": 0.83984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9681,
+      "step": 487
+    },
+    {
+      "epoch": 0.39000999000999004,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.0002,
+      "loss": 0.9778,
+      "step": 488
+    },
+    {
+      "epoch": 0.3908091908091908,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.0002,
+      "loss": 0.9768,
+      "step": 489
+    },
+    {
+      "epoch": 0.3916083916083916,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.0002,
+      "loss": 0.9671,
+      "step": 490
+    },
+    {
+      "epoch": 0.39240759240759243,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9813,
+      "step": 491
+    },
+    {
+      "epoch": 0.3932067932067932,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.0002,
+      "loss": 0.9819,
+      "step": 492
+    },
+    {
+      "epoch": 0.394005994005994,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9845,
+      "step": 493
+    },
+    {
+      "epoch": 0.3948051948051948,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0002,
+      "loss": 0.9789,
+      "step": 494
+    },
+    {
+      "epoch": 0.3956043956043956,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9736,
+      "step": 495
+    },
+    {
+      "epoch": 0.3964035964035964,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.0002,
+      "loss": 0.9737,
+      "step": 496
+    },
+    {
+      "epoch": 0.3972027972027972,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9673,
+      "step": 497
+    },
+    {
+      "epoch": 0.398001998001998,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9686,
+      "step": 498
+    },
+    {
+      "epoch": 0.3988011988011988,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0002,
+      "loss": 0.9774,
+      "step": 499
+    },
+    {
+      "epoch": 0.3996003996003996,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9708,
+      "step": 500
+    },
+    {
+      "epoch": 0.4003996003996004,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0002,
+      "loss": 0.9645,
+      "step": 501
+    },
+    {
+      "epoch": 0.4011988011988012,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9748,
+      "step": 502
+    },
+    {
+      "epoch": 0.401998001998002,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.0002,
+      "loss": 0.9789,
+      "step": 503
+    },
+    {
+      "epoch": 0.4027972027972028,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.0002,
+      "loss": 0.9741,
+      "step": 504
+    },
+    {
+      "epoch": 0.4035964035964036,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.0002,
+      "loss": 0.9733,
+      "step": 505
+    },
+    {
+      "epoch": 0.4043956043956044,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0002,
+      "loss": 0.971,
+      "step": 506
+    },
+    {
+      "epoch": 0.4051948051948052,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9747,
+      "step": 507
+    },
+    {
+      "epoch": 0.405994005994006,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.0002,
+      "loss": 0.9571,
+      "step": 508
+    },
+    {
+      "epoch": 0.4067932067932068,
+      "grad_norm": 4.375,
+      "learning_rate": 0.0002,
+      "loss": 0.9966,
+      "step": 509
+    },
+    {
+      "epoch": 0.4075924075924076,
+      "grad_norm": 3.71875,
+      "learning_rate": 0.0002,
+      "loss": 0.9856,
+      "step": 510
+    },
+    {
+      "epoch": 0.4083916083916084,
+      "grad_norm": 1.0,
+      "learning_rate": 0.0002,
+      "loss": 0.9844,
+      "step": 511
+    },
+    {
+      "epoch": 0.4091908091908092,
+      "grad_norm": 4.1875,
+      "learning_rate": 0.0002,
+      "loss": 1.0065,
+      "step": 512
+    },
+    {
+      "epoch": 0.40999000999001,
+      "grad_norm": 3.828125,
+      "learning_rate": 0.0002,
+      "loss": 1.0032,
+      "step": 513
+    },
+    {
+      "epoch": 0.41078921078921077,
+      "grad_norm": 1.609375,
+      "learning_rate": 0.0002,
+      "loss": 0.9871,
+      "step": 514
+    },
+    {
+      "epoch": 0.4115884115884116,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.0002,
+      "loss": 1.0058,
+      "step": 515
+    },
+    {
+      "epoch": 0.4123876123876124,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.0002,
+      "loss": 1.003,
+      "step": 516
+    },
+    {
+      "epoch": 0.41318681318681316,
+      "grad_norm": 1.6328125,
+      "learning_rate": 0.0002,
+      "loss": 0.9935,
+      "step": 517
+    },
+    {
+      "epoch": 0.413986013986014,
+      "grad_norm": 1.5625,
+      "learning_rate": 0.0002,
+      "loss": 0.9933,
+      "step": 518
+    },
+    {
+      "epoch": 0.4147852147852148,
+      "grad_norm": 0.9921875,
+      "learning_rate": 0.0002,
+      "loss": 0.9821,
+      "step": 519
+    },
+    {
+      "epoch": 0.4155844155844156,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.0002,
+      "loss": 0.981,
+      "step": 520
+    },
+    {
+      "epoch": 0.4163836163836164,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.0002,
+      "loss": 0.9768,
+      "step": 521
+    },
+    {
+      "epoch": 0.4171828171828172,
+      "grad_norm": 0.8515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9851,
+      "step": 522
+    },
+    {
+      "epoch": 0.417982017982018,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9805,
+      "step": 523
+    },
+    {
+      "epoch": 0.41878121878121877,
+      "grad_norm": 0.8515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9721,
+      "step": 524
+    },
+    {
+      "epoch": 0.4195804195804196,
+      "grad_norm": 0.91796875,
+      "learning_rate": 0.0002,
+      "loss": 0.9758,
+      "step": 525
+    },
+    {
+      "epoch": 0.4203796203796204,
+      "grad_norm": 0.75,
+      "learning_rate": 0.0002,
+      "loss": 0.9759,
+      "step": 526
+    },
+    {
+      "epoch": 0.42117882117882116,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9673,
+      "step": 527
+    },
+    {
+      "epoch": 0.421978021978022,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.0002,
+      "loss": 0.9705,
+      "step": 528
+    },
+    {
+      "epoch": 0.4227772227772228,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.0002,
+      "loss": 0.9699,
+      "step": 529
+    },
+    {
+      "epoch": 0.42357642357642356,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0002,
+      "loss": 0.977,
+      "step": 530
+    },
+    {
+      "epoch": 0.42437562437562437,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.0002,
+      "loss": 0.9698,
+      "step": 531
+    },
+    {
+      "epoch": 0.4251748251748252,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0002,
+      "loss": 0.9645,
+      "step": 532
+    },
+    {
+      "epoch": 0.42597402597402595,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0002,
+      "loss": 0.9701,
+      "step": 533
+    },
+    {
+      "epoch": 0.42677322677322677,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0002,
+      "loss": 0.9823,
+      "step": 534
+    },
+    {
+      "epoch": 0.4275724275724276,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9652,
+      "step": 535
+    },
+    {
+      "epoch": 0.4283716283716284,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.0002,
+      "loss": 1.0032,
+      "step": 536
+    },
+    {
+      "epoch": 0.42917082917082916,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9627,
+      "step": 537
+    },
+    {
+      "epoch": 0.42997002997003,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0002,
+      "loss": 0.9722,
+      "step": 538
+    },
+    {
+      "epoch": 0.4307692307692308,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9705,
+      "step": 539
+    },
+    {
+      "epoch": 0.43156843156843155,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.0002,
+      "loss": 0.9534,
+      "step": 540
+    },
+    {
+      "epoch": 0.43236763236763237,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0002,
+      "loss": 0.9605,
+      "step": 541
+    },
+    {
+      "epoch": 0.4331668331668332,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.0002,
+      "loss": 0.9728,
+      "step": 542
+    },
+    {
+      "epoch": 0.43396603396603395,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0002,
+      "loss": 0.9666,
+      "step": 543
+    },
+    {
+      "epoch": 0.43476523476523476,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0002,
+      "loss": 0.9846,
+      "step": 544
+    },
+    {
+      "epoch": 0.4355644355644356,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0002,
+      "loss": 0.9661,
+      "step": 545
+    },
+    {
+      "epoch": 0.43636363636363634,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9567,
+      "step": 546
+    },
+    {
+      "epoch": 0.43716283716283716,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0002,
+      "loss": 0.965,
+      "step": 547
+    },
+    {
+      "epoch": 0.437962037962038,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9692,
+      "step": 548
+    },
+    {
+      "epoch": 0.43876123876123874,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0002,
+      "loss": 0.9624,
+      "step": 549
+    },
+    {
+      "epoch": 0.43956043956043955,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0002,
+      "loss": 0.9629,
+      "step": 550
+    },
+    {
+      "epoch": 0.44035964035964037,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9585,
+      "step": 551
+    },
+    {
+      "epoch": 0.4411588411588412,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0002,
+      "loss": 0.9641,
+      "step": 552
+    },
+    {
+      "epoch": 0.44195804195804195,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9625,
+      "step": 553
+    },
+    {
+      "epoch": 0.44275724275724276,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.0002,
+      "loss": 0.9725,
+      "step": 554
+    },
+    {
+      "epoch": 0.4435564435564436,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9572,
+      "step": 555
+    },
+    {
+      "epoch": 0.44435564435564434,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9528,
+      "step": 556
+    },
+    {
+      "epoch": 0.44515484515484516,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9578,
+      "step": 557
+    },
+    {
+      "epoch": 0.445954045954046,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.0002,
+      "loss": 0.9924,
+      "step": 558
+    },
+    {
+      "epoch": 0.44675324675324674,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0002,
+      "loss": 0.9672,
+      "step": 559
+    },
+    {
+      "epoch": 0.44755244755244755,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0002,
+      "loss": 0.961,
+      "step": 560
+    },
+    {
+      "epoch": 0.44835164835164837,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9637,
+      "step": 561
+    },
+    {
+      "epoch": 0.44915084915084913,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0002,
+      "loss": 0.9612,
+      "step": 562
+    },
+    {
+      "epoch": 0.44995004995004995,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0002,
+      "loss": 0.9778,
+      "step": 563
+    },
+    {
+      "epoch": 0.45074925074925076,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0002,
+      "loss": 0.9764,
+      "step": 564
+    },
+    {
+      "epoch": 0.4515484515484515,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9674,
+      "step": 565
+    },
+    {
+      "epoch": 0.45234765234765234,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0002,
+      "loss": 0.975,
+      "step": 566
+    },
+    {
+      "epoch": 0.45314685314685316,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0002,
+      "loss": 0.9706,
+      "step": 567
+    },
+    {
+      "epoch": 0.453946053946054,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0002,
+      "loss": 0.9595,
+      "step": 568
+    },
+    {
+      "epoch": 0.45474525474525473,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0002,
+      "loss": 0.9616,
+      "step": 569
+    },
+    {
+      "epoch": 0.45554445554445555,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0002,
+      "loss": 0.9772,
+      "step": 570
+    },
+    {
+      "epoch": 0.45634365634365637,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0002,
+      "loss": 0.972,
+      "step": 571
+    },
+    {
+      "epoch": 0.45714285714285713,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9671,
+      "step": 572
+    },
+    {
+      "epoch": 0.45794205794205795,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9545,
+      "step": 573
+    },
+    {
+      "epoch": 0.45874125874125876,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0002,
+      "loss": 0.9589,
+      "step": 574
+    },
+    {
+      "epoch": 0.4595404595404595,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0002,
+      "loss": 0.9543,
+      "step": 575
+    },
+    {
+      "epoch": 0.46033966033966034,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0002,
+      "loss": 0.9687,
+      "step": 576
+    },
+    {
+      "epoch": 0.46113886113886116,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0002,
+      "loss": 0.9599,
+      "step": 577
+    },
+    {
+      "epoch": 0.4619380619380619,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9639,
+      "step": 578
+    },
+    {
+      "epoch": 0.46273726273726273,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0002,
+      "loss": 0.9622,
+      "step": 579
+    },
+    {
+      "epoch": 0.46353646353646355,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0002,
+      "loss": 0.9585,
+      "step": 580
+    },
+    {
+      "epoch": 0.4643356643356643,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0002,
+      "loss": 0.9655,
+      "step": 581
+    },
+    {
+      "epoch": 0.4651348651348651,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0002,
+      "loss": 0.9657,
+      "step": 582
+    },
+    {
+      "epoch": 0.46593406593406594,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0002,
+      "loss": 0.9736,
+      "step": 583
+    },
+    {
+      "epoch": 0.46673326673326676,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0002,
+      "loss": 0.9554,
+      "step": 584
+    },
+    {
+      "epoch": 0.4675324675324675,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0002,
+      "loss": 0.9578,
+      "step": 585
+    },
+    {
+      "epoch": 0.46833166833166834,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0002,
+      "loss": 0.9584,
+      "step": 586
+    },
+    {
+      "epoch": 0.46913086913086915,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9568,
+      "step": 587
+    },
+    {
+      "epoch": 0.4699300699300699,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9654,
+      "step": 588
+    },
+    {
+      "epoch": 0.47072927072927073,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0002,
+      "loss": 0.9606,
+      "step": 589
+    },
+    {
+      "epoch": 0.47152847152847155,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0002,
+      "loss": 0.9664,
+      "step": 590
+    },
+    {
+      "epoch": 0.4723276723276723,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9556,
+      "step": 591
+    },
+    {
+      "epoch": 0.4731268731268731,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9619,
+      "step": 592
+    },
+    {
+      "epoch": 0.47392607392607394,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0002,
+      "loss": 0.9612,
+      "step": 593
+    },
+    {
+      "epoch": 0.4747252747252747,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9556,
+      "step": 594
+    },
+    {
+      "epoch": 0.4755244755244755,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9562,
+      "step": 595
+    },
+    {
+      "epoch": 0.47632367632367634,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0002,
+      "loss": 0.9704,
+      "step": 596
+    },
+    {
+      "epoch": 0.4771228771228771,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0002,
+      "loss": 0.9603,
+      "step": 597
+    },
+    {
+      "epoch": 0.4779220779220779,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0002,
+      "loss": 0.9578,
+      "step": 598
+    },
+    {
+      "epoch": 0.47872127872127873,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0002,
+      "loss": 0.9675,
+      "step": 599
+    },
+    {
+      "epoch": 0.47952047952047955,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0002,
+      "loss": 0.9496,
+      "step": 600
+    },
+    {
+      "epoch": 0.4803196803196803,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9625,
+      "step": 601
+    },
+    {
+      "epoch": 0.4811188811188811,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9678,
+      "step": 602
+    },
+    {
+      "epoch": 0.48191808191808194,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0002,
+      "loss": 0.9612,
+      "step": 603
+    },
+    {
+      "epoch": 0.4827172827172827,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0002,
+      "loss": 0.954,
+      "step": 604
+    },
+    {
+      "epoch": 0.4835164835164835,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0002,
+      "loss": 0.9645,
+      "step": 605
+    },
+    {
+      "epoch": 0.48431568431568434,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0002,
+      "loss": 0.9512,
+      "step": 606
+    },
+    {
+      "epoch": 0.4851148851148851,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0002,
+      "loss": 0.95,
+      "step": 607
+    },
+    {
+      "epoch": 0.4859140859140859,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9608,
+      "step": 608
+    },
+    {
+      "epoch": 0.48671328671328673,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.0002,
+      "loss": 0.9713,
+      "step": 609
+    },
+    {
+      "epoch": 0.4875124875124875,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0002,
+      "loss": 0.9625,
+      "step": 610
+    },
+    {
+      "epoch": 0.4883116883116883,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9605,
+      "step": 611
+    },
+    {
+      "epoch": 0.4891108891108891,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0002,
+      "loss": 0.9594,
+      "step": 612
+    },
+    {
+      "epoch": 0.4899100899100899,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0002,
+      "loss": 0.953,
+      "step": 613
+    },
+    {
+      "epoch": 0.4907092907092907,
+      "grad_norm": 0.87109375,
+      "learning_rate": 0.0002,
+      "loss": 0.9565,
+      "step": 614
+    },
+    {
+      "epoch": 0.4915084915084915,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.0002,
+      "loss": 0.9683,
+      "step": 615
+    },
+    {
+      "epoch": 0.49230769230769234,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0002,
+      "loss": 0.9607,
+      "step": 616
+    },
+    {
+      "epoch": 0.4931068931068931,
+      "grad_norm": 3.09375,
+      "learning_rate": 0.0002,
+      "loss": 0.9799,
+      "step": 617
+    },
+    {
+      "epoch": 0.4939060939060939,
+      "grad_norm": 1.71875,
+      "learning_rate": 0.0002,
+      "loss": 0.9576,
+      "step": 618
+    },
+    {
+      "epoch": 0.49470529470529473,
+      "grad_norm": 3.3125,
+      "learning_rate": 0.0002,
+      "loss": 0.9817,
+      "step": 619
+    },
+    {
+      "epoch": 0.4955044955044955,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.0002,
+      "loss": 0.9602,
+      "step": 620
+    },
+    {
+      "epoch": 0.4963036963036963,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.0002,
+      "loss": 0.9746,
+      "step": 621
+    },
+    {
+      "epoch": 0.4971028971028971,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.0002,
+      "loss": 0.9646,
+      "step": 622
+    },
+    {
+      "epoch": 0.4979020979020979,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.0002,
+      "loss": 0.977,
+      "step": 623
+    },
+    {
+      "epoch": 0.4987012987012987,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9694,
+      "step": 624
+    },
+    {
+      "epoch": 0.4995004995004995,
+      "grad_norm": 0.89453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9741,
+      "step": 625
+    },
+    {
+      "epoch": 0.5002997002997003,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.0002,
+      "loss": 0.9771,
+      "step": 626
+    },
+    {
+      "epoch": 0.5010989010989011,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.0002,
+      "loss": 0.9678,
+      "step": 627
+    },
+    {
+      "epoch": 0.5018981018981019,
+      "grad_norm": 1.71875,
+      "learning_rate": 0.0002,
+      "loss": 0.985,
+      "step": 628
+    },
+    {
+      "epoch": 0.5026973026973027,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9604,
+      "step": 629
+    },
+    {
+      "epoch": 0.5034965034965035,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.0002,
+      "loss": 0.9795,
+      "step": 630
+    },
+    {
+      "epoch": 0.5042957042957043,
+      "grad_norm": 1.484375,
+      "learning_rate": 0.0002,
+      "loss": 0.9803,
+      "step": 631
+    },
+    {
+      "epoch": 0.5050949050949051,
+      "grad_norm": 4.125,
+      "learning_rate": 0.0002,
+      "loss": 1.0163,
+      "step": 632
+    },
+    {
+      "epoch": 0.5058941058941059,
+      "grad_norm": 3.828125,
+      "learning_rate": 0.0002,
+      "loss": 1.0087,
+      "step": 633
+    },
+    {
+      "epoch": 0.5066933066933067,
+      "grad_norm": 1.734375,
+      "learning_rate": 0.0002,
+      "loss": 0.985,
+      "step": 634
+    },
+    {
+      "epoch": 0.5074925074925075,
+      "grad_norm": 1.7109375,
+      "learning_rate": 0.0002,
+      "loss": 0.9874,
+      "step": 635
+    },
+    {
+      "epoch": 0.5082917082917083,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0002,
+      "loss": 0.9766,
+      "step": 636
+    },
+    {
+      "epoch": 0.509090909090909,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.0002,
+      "loss": 0.9683,
+      "step": 637
+    },
+    {
+      "epoch": 0.5098901098901099,
+      "grad_norm": 2.8125,
+      "learning_rate": 0.0002,
+      "loss": 0.9666,
+      "step": 638
+    },
+    {
+      "epoch": 0.5106893106893107,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.0002,
+      "loss": 0.9873,
+      "step": 639
+    },
+    {
+      "epoch": 0.5114885114885115,
+      "grad_norm": 3.484375,
+      "learning_rate": 0.0002,
+      "loss": 0.9802,
+      "step": 640
+    },
+    {
+      "epoch": 0.5122877122877123,
+      "grad_norm": 3.109375,
+      "learning_rate": 0.0002,
+      "loss": 0.9881,
+      "step": 641
+    },
+    {
+      "epoch": 0.5130869130869131,
+      "grad_norm": 3.15625,
+      "learning_rate": 0.0002,
+      "loss": 0.9783,
+      "step": 642
+    },
+    {
+      "epoch": 0.5138861138861139,
+      "grad_norm": 3.03125,
+      "learning_rate": 0.0002,
+      "loss": 0.9822,
+      "step": 643
+    },
+    {
+      "epoch": 0.5146853146853146,
+      "grad_norm": 1.7890625,
+      "learning_rate": 0.0002,
+      "loss": 0.9745,
+      "step": 644
+    },
+    {
+      "epoch": 0.5154845154845155,
+      "grad_norm": 1.6796875,
+      "learning_rate": 0.0002,
+      "loss": 0.9732,
+      "step": 645
+    },
+    {
+      "epoch": 0.5162837162837163,
+      "grad_norm": 2.640625,
+      "learning_rate": 0.0002,
+      "loss": 0.9754,
+      "step": 646
+    },
+    {
+      "epoch": 0.5170829170829171,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9666,
+      "step": 647
+    },
+    {
+      "epoch": 0.5178821178821179,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.0002,
+      "loss": 0.9754,
+      "step": 648
+    },
+    {
+      "epoch": 0.5186813186813187,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.0002,
+      "loss": 0.9681,
+      "step": 649
+    },
+    {
+      "epoch": 0.5194805194805194,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.0002,
+      "loss": 0.9702,
+      "step": 650
+    },
+    {
+      "epoch": 0.5202797202797202,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.0002,
+      "loss": 0.9748,
+      "step": 651
+    },
+    {
+      "epoch": 0.5210789210789211,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9633,
+      "step": 652
+    },
+    {
+      "epoch": 0.5218781218781219,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9639,
+      "step": 653
+    },
+    {
+      "epoch": 0.5226773226773227,
+      "grad_norm": 1.65625,
+      "learning_rate": 0.0002,
+      "loss": 0.9638,
+      "step": 654
+    },
+    {
+      "epoch": 0.5234765234765235,
+      "grad_norm": 1.59375,
+      "learning_rate": 0.0002,
+      "loss": 0.9577,
+      "step": 655
+    },
+    {
+      "epoch": 0.5242757242757243,
+      "grad_norm": 1.75,
+      "learning_rate": 0.0002,
+      "loss": 0.9527,
+      "step": 656
+    },
+    {
+      "epoch": 0.525074925074925,
+      "grad_norm": 1.6015625,
+      "learning_rate": 0.0002,
+      "loss": 0.9651,
+      "step": 657
+    },
+    {
+      "epoch": 0.5258741258741259,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9684,
+      "step": 658
+    },
+    {
+      "epoch": 0.5266733266733267,
+      "grad_norm": 1.828125,
+      "learning_rate": 0.0002,
+      "loss": 0.9685,
+      "step": 659
+    },
+    {
+      "epoch": 0.5274725274725275,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9737,
+      "step": 660
+    },
+    {
+      "epoch": 0.5282717282717283,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.0002,
+      "loss": 0.9637,
+      "step": 661
+    },
+    {
+      "epoch": 0.5290709290709291,
+      "grad_norm": 1.71875,
+      "learning_rate": 0.0002,
+      "loss": 0.9757,
+      "step": 662
+    },
+    {
+      "epoch": 0.5298701298701298,
+      "grad_norm": 1.59375,
+      "learning_rate": 0.0002,
+      "loss": 0.9603,
+      "step": 663
+    },
+    {
+      "epoch": 0.5306693306693306,
+      "grad_norm": 1.4453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9639,
+      "step": 664
+    },
+    {
+      "epoch": 0.5314685314685315,
+      "grad_norm": 1.40625,
+      "learning_rate": 0.0002,
+      "loss": 0.9722,
+      "step": 665
+    },
+    {
+      "epoch": 0.5322677322677323,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.0002,
+      "loss": 0.9585,
+      "step": 666
+    },
+    {
+      "epoch": 0.5330669330669331,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.0002,
+      "loss": 0.957,
+      "step": 667
+    },
+    {
+      "epoch": 0.5338661338661339,
+      "grad_norm": 1.5703125,
+      "learning_rate": 0.0002,
+      "loss": 0.957,
+      "step": 668
+    },
+    {
+      "epoch": 0.5346653346653346,
+      "grad_norm": 1.4921875,
+      "learning_rate": 0.0002,
+      "loss": 0.9568,
+      "step": 669
+    },
+    {
+      "epoch": 0.5354645354645354,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.0002,
+      "loss": 0.9592,
+      "step": 670
+    },
+    {
+      "epoch": 0.5362637362637362,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9539,
+      "step": 671
+    },
+    {
+      "epoch": 0.5370629370629371,
+      "grad_norm": 1.421875,
+      "learning_rate": 0.0002,
+      "loss": 0.9659,
+      "step": 672
+    },
+    {
+      "epoch": 0.5378621378621379,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.0002,
+      "loss": 0.9546,
+      "step": 673
+    },
+    {
+      "epoch": 0.5386613386613387,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.0002,
+      "loss": 0.9541,
+      "step": 674
+    },
+    {
+      "epoch": 0.5394605394605395,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.0002,
+      "loss": 0.9562,
+      "step": 675
+    },
+    {
+      "epoch": 0.5402597402597402,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.0002,
+      "loss": 0.9588,
+      "step": 676
+    },
+    {
+      "epoch": 0.541058941058941,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.0002,
+      "loss": 0.9748,
+      "step": 677
+    },
+    {
+      "epoch": 0.5418581418581419,
+      "grad_norm": 1.25,
+      "learning_rate": 0.0002,
+      "loss": 0.9585,
+      "step": 678
+    },
+    {
+      "epoch": 0.5426573426573427,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.0002,
+      "loss": 0.9598,
+      "step": 679
+    },
+    {
+      "epoch": 0.5434565434565435,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.0002,
+      "loss": 0.9593,
+      "step": 680
+    },
+    {
+      "epoch": 0.5442557442557443,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9593,
+      "step": 681
+    },
+    {
+      "epoch": 0.545054945054945,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.0002,
+      "loss": 0.9662,
+      "step": 682
+    },
+    {
+      "epoch": 0.5458541458541458,
+      "grad_norm": 0.9921875,
+      "learning_rate": 0.0002,
+      "loss": 0.9631,
+      "step": 683
+    },
+    {
+      "epoch": 0.5466533466533466,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.0002,
+      "loss": 0.9629,
+      "step": 684
+    },
+    {
+      "epoch": 0.5474525474525475,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9607,
+      "step": 685
+    },
+    {
+      "epoch": 0.5482517482517483,
+      "grad_norm": 1.3515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9548,
+      "step": 686
+    },
+    {
+      "epoch": 0.5490509490509491,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.0002,
+      "loss": 0.9686,
+      "step": 687
+    },
+    {
+      "epoch": 0.5498501498501499,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9729,
+      "step": 688
+    },
+    {
+      "epoch": 0.5506493506493506,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.0002,
+      "loss": 0.9578,
+      "step": 689
+    },
+    {
+      "epoch": 0.5514485514485514,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.0002,
+      "loss": 0.9595,
+      "step": 690
+    },
+    {
+      "epoch": 0.5522477522477522,
+      "grad_norm": 1.4375,
+      "learning_rate": 0.0002,
+      "loss": 0.9652,
+      "step": 691
+    },
+    {
+      "epoch": 0.5530469530469531,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0002,
+      "loss": 0.9667,
+      "step": 692
+    },
+    {
+      "epoch": 0.5538461538461539,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9716,
+      "step": 693
+    },
+    {
+      "epoch": 0.5546453546453547,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.0002,
+      "loss": 0.9626,
+      "step": 694
+    },
+    {
+      "epoch": 0.5554445554445554,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.0002,
+      "loss": 0.9607,
+      "step": 695
+    },
+    {
+      "epoch": 0.5562437562437562,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.0002,
+      "loss": 0.9571,
+      "step": 696
+    },
+    {
+      "epoch": 0.557042957042957,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.0002,
+      "loss": 0.9725,
+      "step": 697
+    },
+    {
+      "epoch": 0.5578421578421578,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9573,
+      "step": 698
+    },
+    {
+      "epoch": 0.5586413586413587,
+      "grad_norm": 0.93359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9757,
+      "step": 699
+    },
+    {
+      "epoch": 0.5594405594405595,
+      "grad_norm": 0.93359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9547,
+      "step": 700
+    },
+    {
+      "epoch": 0.5602397602397602,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.0002,
+      "loss": 0.9639,
+      "step": 701
+    },
+    {
+      "epoch": 0.561038961038961,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9574,
+      "step": 702
+    },
+    {
+      "epoch": 0.5618381618381618,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.0002,
+      "loss": 0.9612,
+      "step": 703
+    },
+    {
+      "epoch": 0.5626373626373626,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9603,
+      "step": 704
+    },
+    {
+      "epoch": 0.5634365634365635,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.0002,
+      "loss": 0.9559,
+      "step": 705
+    },
+    {
+      "epoch": 0.5642357642357643,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9654,
+      "step": 706
+    },
+    {
+      "epoch": 0.5650349650349651,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0002,
+      "loss": 0.9596,
+      "step": 707
+    },
+    {
+      "epoch": 0.5658341658341658,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0002,
+      "loss": 0.9577,
+      "step": 708
+    },
+    {
+      "epoch": 0.5666333666333666,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0002,
+      "loss": 0.9477,
+      "step": 709
+    },
+    {
+      "epoch": 0.5674325674325674,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9544,
+      "step": 710
+    },
+    {
+      "epoch": 0.5682317682317682,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0002,
+      "loss": 0.9678,
+      "step": 711
+    },
+    {
+      "epoch": 0.5690309690309691,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9547,
+      "step": 712
+    },
+    {
+      "epoch": 0.5698301698301699,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0002,
+      "loss": 0.952,
+      "step": 713
+    },
+    {
+      "epoch": 0.5706293706293706,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9449,
+      "step": 714
+    },
+    {
+      "epoch": 0.5714285714285714,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9617,
+      "step": 715
+    },
+    {
+      "epoch": 0.5722277722277722,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0002,
+      "loss": 0.9499,
+      "step": 716
+    },
+    {
+      "epoch": 0.573026973026973,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9448,
+      "step": 717
+    },
+    {
+      "epoch": 0.5738261738261738,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0002,
+      "loss": 0.9566,
+      "step": 718
+    },
+    {
+      "epoch": 0.5746253746253747,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.0002,
+      "loss": 0.9474,
+      "step": 719
+    },
+    {
+      "epoch": 0.5754245754245755,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0002,
+      "loss": 0.9512,
+      "step": 720
+    },
+    {
+      "epoch": 0.5762237762237762,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9597,
+      "step": 721
+    },
+    {
+      "epoch": 0.577022977022977,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9592,
+      "step": 722
+    },
+    {
+      "epoch": 0.5778221778221778,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0002,
+      "loss": 0.9454,
+      "step": 723
+    },
+    {
+      "epoch": 0.5786213786213786,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0002,
+      "loss": 0.9474,
+      "step": 724
+    },
+    {
+      "epoch": 0.5794205794205795,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9498,
+      "step": 725
+    },
+    {
+      "epoch": 0.5802197802197803,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0002,
+      "loss": 0.9513,
+      "step": 726
+    },
+    {
+      "epoch": 0.581018981018981,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0002,
+      "loss": 0.9618,
+      "step": 727
+    },
+    {
+      "epoch": 0.5818181818181818,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0002,
+      "loss": 0.948,
+      "step": 728
+    },
+    {
+      "epoch": 0.5826173826173826,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9427,
+      "step": 729
+    },
+    {
+      "epoch": 0.5834165834165834,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0002,
+      "loss": 0.9506,
+      "step": 730
+    },
+    {
+      "epoch": 0.5842157842157842,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9627,
+      "step": 731
+    },
+    {
+      "epoch": 0.5850149850149851,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9563,
+      "step": 732
+    },
+    {
+      "epoch": 0.5858141858141858,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0002,
+      "loss": 0.9508,
+      "step": 733
+    },
+    {
+      "epoch": 0.5866133866133866,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0002,
+      "loss": 0.9634,
+      "step": 734
+    },
+    {
+      "epoch": 0.5874125874125874,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9412,
+      "step": 735
+    },
+    {
+      "epoch": 0.5882117882117882,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0002,
+      "loss": 0.9478,
+      "step": 736
+    },
+    {
+      "epoch": 0.589010989010989,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0002,
+      "loss": 0.9461,
+      "step": 737
+    },
+    {
+      "epoch": 0.5898101898101898,
+      "grad_norm": 0.375,
+      "learning_rate": 0.0002,
+      "loss": 0.9528,
+      "step": 738
+    },
+    {
+      "epoch": 0.5906093906093907,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0002,
+      "loss": 0.9563,
+      "step": 739
+    },
+    {
+      "epoch": 0.5914085914085914,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.0002,
+      "loss": 0.9505,
+      "step": 740
+    },
+    {
+      "epoch": 0.5922077922077922,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.0002,
+      "loss": 0.9631,
+      "step": 741
+    },
+    {
+      "epoch": 0.593006993006993,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.0002,
+      "loss": 0.9508,
+      "step": 742
+    },
+    {
+      "epoch": 0.5938061938061938,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0002,
+      "loss": 0.9393,
+      "step": 743
+    },
+    {
+      "epoch": 0.5946053946053946,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.0002,
+      "loss": 0.9639,
+      "step": 744
+    },
+    {
+      "epoch": 0.5954045954045954,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.0002,
+      "loss": 0.9493,
+      "step": 745
+    },
+    {
+      "epoch": 0.5962037962037962,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0002,
+      "loss": 0.9542,
+      "step": 746
+    },
+    {
+      "epoch": 0.597002997002997,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.0002,
+      "loss": 0.9497,
+      "step": 747
+    },
+    {
+      "epoch": 0.5978021978021978,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0002,
+      "loss": 0.9544,
+      "step": 748
+    },
+    {
+      "epoch": 0.5986013986013986,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9514,
+      "step": 749
+    },
+    {
+      "epoch": 0.5994005994005994,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.0002,
+      "loss": 0.9518,
+      "step": 750
+    },
+    {
+      "epoch": 0.6001998001998002,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0002,
+      "loss": 0.9461,
+      "step": 751
+    },
+    {
+      "epoch": 0.600999000999001,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.0002,
+      "loss": 0.9509,
+      "step": 752
+    },
+    {
+      "epoch": 0.6017982017982018,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9563,
+      "step": 753
+    },
+    {
+      "epoch": 0.6025974025974026,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.0002,
+      "loss": 0.9607,
+      "step": 754
+    },
+    {
+      "epoch": 0.6033966033966034,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9561,
+      "step": 755
+    },
+    {
+      "epoch": 0.6041958041958042,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9481,
+      "step": 756
+    },
+    {
+      "epoch": 0.604995004995005,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9566,
+      "step": 757
+    },
+    {
+      "epoch": 0.6057942057942058,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9807,
+      "step": 758
+    },
+    {
+      "epoch": 0.6065934065934065,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9499,
+      "step": 759
+    },
+    {
+      "epoch": 0.6073926073926074,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0002,
+      "loss": 0.9561,
+      "step": 760
+    },
+    {
+      "epoch": 0.6081918081918082,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9444,
+      "step": 761
+    },
+    {
+      "epoch": 0.608991008991009,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9606,
+      "step": 762
+    },
+    {
+      "epoch": 0.6097902097902098,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.0002,
+      "loss": 0.9558,
+      "step": 763
+    },
+    {
+      "epoch": 0.6105894105894106,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9578,
+      "step": 764
+    },
+    {
+      "epoch": 0.6113886113886113,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0002,
+      "loss": 0.9636,
+      "step": 765
+    },
+    {
+      "epoch": 0.6121878121878122,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0002,
+      "loss": 0.9478,
+      "step": 766
+    },
+    {
+      "epoch": 0.612987012987013,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9505,
+      "step": 767
+    },
+    {
+      "epoch": 0.6137862137862138,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0002,
+      "loss": 0.949,
+      "step": 768
+    },
+    {
+      "epoch": 0.6145854145854146,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9581,
+      "step": 769
+    },
+    {
+      "epoch": 0.6153846153846154,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9566,
+      "step": 770
+    },
+    {
+      "epoch": 0.6161838161838162,
+      "grad_norm": 0.9375,
+      "learning_rate": 0.0002,
+      "loss": 0.9776,
+      "step": 771
+    },
+    {
+      "epoch": 0.6169830169830169,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.0002,
+      "loss": 0.9523,
+      "step": 772
+    },
+    {
+      "epoch": 0.6177822177822178,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0002,
+      "loss": 0.9474,
+      "step": 773
+    },
+    {
+      "epoch": 0.6185814185814186,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.0002,
+      "loss": 0.9549,
+      "step": 774
+    },
+    {
+      "epoch": 0.6193806193806194,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.0002,
+      "loss": 0.9621,
+      "step": 775
+    },
+    {
+      "epoch": 0.6201798201798202,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9582,
+      "step": 776
+    },
+    {
+      "epoch": 0.620979020979021,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.0002,
+      "loss": 0.9553,
+      "step": 777
+    },
+    {
+      "epoch": 0.6217782217782217,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0002,
+      "loss": 0.9599,
+      "step": 778
+    },
+    {
+      "epoch": 0.6225774225774225,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.0002,
+      "loss": 0.9587,
+      "step": 779
+    },
+    {
+      "epoch": 0.6233766233766234,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0002,
+      "loss": 0.954,
+      "step": 780
+    },
+    {
+      "epoch": 0.6241758241758242,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9522,
+      "step": 781
+    },
+    {
+      "epoch": 0.624975024975025,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.0002,
+      "loss": 0.9682,
+      "step": 782
+    },
+    {
+      "epoch": 0.6257742257742258,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0002,
+      "loss": 0.9605,
+      "step": 783
+    },
+    {
+      "epoch": 0.6265734265734266,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0002,
+      "loss": 0.965,
+      "step": 784
+    },
+    {
+      "epoch": 0.6273726273726273,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0002,
+      "loss": 0.9615,
+      "step": 785
+    },
+    {
+      "epoch": 0.6281718281718282,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0002,
+      "loss": 0.9594,
+      "step": 786
+    },
+    {
+      "epoch": 0.628971028971029,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0002,
+      "loss": 0.9503,
+      "step": 787
+    },
+    {
+      "epoch": 0.6297702297702298,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0002,
+      "loss": 0.9591,
+      "step": 788
+    },
+    {
+      "epoch": 0.6305694305694306,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.0002,
+      "loss": 0.9525,
+      "step": 789
+    },
+    {
+      "epoch": 0.6313686313686314,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0002,
+      "loss": 0.947,
+      "step": 790
+    },
+    {
+      "epoch": 0.6321678321678321,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.0002,
+      "loss": 0.9504,
+      "step": 791
+    },
+    {
+      "epoch": 0.6329670329670329,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.0002,
+      "loss": 0.9487,
+      "step": 792
+    },
+    {
+      "epoch": 0.6337662337662338,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0002,
+      "loss": 0.9546,
+      "step": 793
+    },
+    {
+      "epoch": 0.6345654345654346,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.0002,
+      "loss": 0.9521,
+      "step": 794
+    },
+    {
+      "epoch": 0.6353646353646354,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.0002,
+      "loss": 0.9547,
+      "step": 795
+    },
+    {
+      "epoch": 0.6361638361638362,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.0002,
+      "loss": 0.9538,
+      "step": 796
+    },
+    {
+      "epoch": 0.6369630369630369,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9424,
+      "step": 797
+    },
+    {
+      "epoch": 0.6377622377622377,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9457,
+      "step": 798
+    },
+    {
+      "epoch": 0.6385614385614385,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0002,
+      "loss": 0.9586,
+      "step": 799
+    },
+    {
+      "epoch": 0.6393606393606394,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9511,
+      "step": 800
+    },
+    {
+      "epoch": 0.6401598401598402,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9519,
+      "step": 801
+    },
+    {
+      "epoch": 0.640959040959041,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0002,
+      "loss": 0.9498,
+      "step": 802
+    },
+    {
+      "epoch": 0.6417582417582418,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0002,
+      "loss": 0.9484,
+      "step": 803
+    },
+    {
+      "epoch": 0.6425574425574425,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9541,
+      "step": 804
+    },
+    {
+      "epoch": 0.6433566433566433,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9462,
+      "step": 805
+    },
+    {
+      "epoch": 0.6441558441558441,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9506,
+      "step": 806
+    },
+    {
+      "epoch": 0.644955044955045,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0002,
+      "loss": 0.9571,
+      "step": 807
+    },
+    {
+      "epoch": 0.6457542457542458,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.0002,
+      "loss": 0.9455,
+      "step": 808
+    },
+    {
+      "epoch": 0.6465534465534466,
+      "grad_norm": 0.96875,
+      "learning_rate": 0.0002,
+      "loss": 0.9469,
+      "step": 809
+    },
+    {
+      "epoch": 0.6473526473526473,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9467,
+      "step": 810
+    },
+    {
+      "epoch": 0.6481518481518481,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.0002,
+      "loss": 0.953,
+      "step": 811
+    },
+    {
+      "epoch": 0.6489510489510489,
+      "grad_norm": 0.97265625,
+      "learning_rate": 0.0002,
+      "loss": 0.9639,
+      "step": 812
+    },
+    {
+      "epoch": 0.6497502497502498,
+      "grad_norm": 0.84375,
+      "learning_rate": 0.0002,
+      "loss": 0.9529,
+      "step": 813
+    },
+    {
+      "epoch": 0.6505494505494506,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0002,
+      "loss": 0.9507,
+      "step": 814
+    },
+    {
+      "epoch": 0.6513486513486514,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0002,
+      "loss": 0.9523,
+      "step": 815
+    },
+    {
+      "epoch": 0.6521478521478522,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0002,
+      "loss": 0.9541,
+      "step": 816
+    },
+    {
+      "epoch": 0.6529470529470529,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9484,
+      "step": 817
+    },
+    {
+      "epoch": 0.6537462537462537,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0002,
+      "loss": 0.9383,
+      "step": 818
+    },
+    {
+      "epoch": 0.6545454545454545,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9606,
+      "step": 819
+    },
+    {
+      "epoch": 0.6553446553446554,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.0002,
+      "loss": 0.9556,
+      "step": 820
+    },
+    {
+      "epoch": 0.6561438561438562,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9489,
+      "step": 821
+    },
+    {
+      "epoch": 0.656943056943057,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9447,
+      "step": 822
+    },
+    {
+      "epoch": 0.6577422577422577,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9572,
+      "step": 823
+    },
+    {
+      "epoch": 0.6585414585414585,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0002,
+      "loss": 0.9479,
+      "step": 824
+    },
+    {
+      "epoch": 0.6593406593406593,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9465,
+      "step": 825
+    },
+    {
+      "epoch": 0.6601398601398601,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9495,
+      "step": 826
+    },
+    {
+      "epoch": 0.660939060939061,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9516,
+      "step": 827
+    },
+    {
+      "epoch": 0.6617382617382618,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.0002,
+      "loss": 0.9553,
+      "step": 828
+    },
+    {
+      "epoch": 0.6625374625374625,
+      "grad_norm": 1.7265625,
+      "learning_rate": 0.0002,
+      "loss": 0.9652,
+      "step": 829
+    },
+    {
+      "epoch": 0.6633366633366633,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9591,
+      "step": 830
+    },
+    {
+      "epoch": 0.6641358641358641,
+      "grad_norm": 1.7265625,
+      "learning_rate": 0.0002,
+      "loss": 0.9634,
+      "step": 831
+    },
+    {
+      "epoch": 0.6649350649350649,
+      "grad_norm": 0.98828125,
+      "learning_rate": 0.0002,
+      "loss": 0.9629,
+      "step": 832
+    },
+    {
+      "epoch": 0.6657342657342658,
+      "grad_norm": 2.734375,
+      "learning_rate": 0.0002,
+      "loss": 0.996,
+      "step": 833
+    },
+    {
+      "epoch": 0.6665334665334666,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0002,
+      "loss": 0.9801,
+      "step": 834
+    },
+    {
+      "epoch": 0.6673326673326674,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0002,
+      "loss": 0.9717,
+      "step": 835
+    },
+    {
+      "epoch": 0.6681318681318681,
+      "grad_norm": 1.5546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9562,
+      "step": 836
+    },
+    {
+      "epoch": 0.6689310689310689,
+      "grad_norm": 2.90625,
+      "learning_rate": 0.0002,
+      "loss": 0.9736,
+      "step": 837
+    },
+    {
+      "epoch": 0.6697302697302697,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.0002,
+      "loss": 0.978,
+      "step": 838
+    },
+    {
+      "epoch": 0.6705294705294705,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.0002,
+      "loss": 0.9626,
+      "step": 839
+    },
+    {
+      "epoch": 0.6713286713286714,
+      "grad_norm": 1.703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9681,
+      "step": 840
+    },
+    {
+      "epoch": 0.6721278721278722,
+      "grad_norm": 1.7421875,
+      "learning_rate": 0.0002,
+      "loss": 0.9609,
+      "step": 841
+    },
+    {
+      "epoch": 0.6729270729270729,
+      "grad_norm": 1.5390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9685,
+      "step": 842
+    },
+    {
+      "epoch": 0.6737262737262737,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.0002,
+      "loss": 0.963,
+      "step": 843
+    },
+    {
+      "epoch": 0.6745254745254745,
+      "grad_norm": 1.5078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9596,
+      "step": 844
+    },
+    {
+      "epoch": 0.6753246753246753,
+      "grad_norm": 1.7421875,
+      "learning_rate": 0.0002,
+      "loss": 0.9441,
+      "step": 845
+    },
+    {
+      "epoch": 0.6761238761238761,
+      "grad_norm": 1.6328125,
+      "learning_rate": 0.0002,
+      "loss": 0.9487,
+      "step": 846
+    },
+    {
+      "epoch": 0.676923076923077,
+      "grad_norm": 1.6875,
+      "learning_rate": 0.0002,
+      "loss": 0.9604,
+      "step": 847
+    },
+    {
+      "epoch": 0.6777222777222778,
+      "grad_norm": 1.4921875,
+      "learning_rate": 0.0002,
+      "loss": 0.953,
+      "step": 848
+    },
+    {
+      "epoch": 0.6785214785214785,
+      "grad_norm": 1.515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9443,
+      "step": 849
+    },
+    {
+      "epoch": 0.6793206793206793,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9379,
+      "step": 850
+    },
+    {
+      "epoch": 0.6801198801198801,
+      "grad_norm": 1.5234375,
+      "learning_rate": 0.0002,
+      "loss": 0.9687,
+      "step": 851
+    },
+    {
+      "epoch": 0.6809190809190809,
+      "grad_norm": 1.3984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9617,
+      "step": 852
+    },
+    {
+      "epoch": 0.6817182817182817,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9533,
+      "step": 853
+    },
+    {
+      "epoch": 0.6825174825174826,
+      "grad_norm": 1.59375,
+      "learning_rate": 0.0002,
+      "loss": 0.9519,
+      "step": 854
+    },
+    {
+      "epoch": 0.6833166833166833,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.0002,
+      "loss": 0.957,
+      "step": 855
+    },
+    {
+      "epoch": 0.6841158841158841,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9442,
+      "step": 856
+    },
+    {
+      "epoch": 0.6849150849150849,
+      "grad_norm": 1.609375,
+      "learning_rate": 0.0002,
+      "loss": 0.9594,
+      "step": 857
+    },
+    {
+      "epoch": 0.6857142857142857,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.0002,
+      "loss": 0.947,
+      "step": 858
+    },
+    {
+      "epoch": 0.6865134865134865,
+      "grad_norm": 1.5390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9541,
+      "step": 859
+    },
+    {
+      "epoch": 0.6873126873126874,
+      "grad_norm": 1.375,
+      "learning_rate": 0.0002,
+      "loss": 0.9641,
+      "step": 860
+    },
+    {
+      "epoch": 0.6881118881118881,
+      "grad_norm": 0.953125,
+      "learning_rate": 0.0002,
+      "loss": 0.9531,
+      "step": 861
+    },
+    {
+      "epoch": 0.6889110889110889,
+      "grad_norm": 0.90234375,
+      "learning_rate": 0.0002,
+      "loss": 0.9482,
+      "step": 862
+    },
+    {
+      "epoch": 0.6897102897102897,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.0002,
+      "loss": 0.9451,
+      "step": 863
+    },
+    {
+      "epoch": 0.6905094905094905,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9553,
+      "step": 864
+    },
+    {
+      "epoch": 0.6913086913086913,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.0002,
+      "loss": 0.9599,
+      "step": 865
+    },
+    {
+      "epoch": 0.6921078921078921,
+      "grad_norm": 0.9375,
+      "learning_rate": 0.0002,
+      "loss": 0.9502,
+      "step": 866
+    },
+    {
+      "epoch": 0.692907092907093,
+      "grad_norm": 1.421875,
+      "learning_rate": 0.0002,
+      "loss": 0.9555,
+      "step": 867
+    },
+    {
+      "epoch": 0.6937062937062937,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.0002,
+      "loss": 0.958,
+      "step": 868
+    },
+    {
+      "epoch": 0.6945054945054945,
+      "grad_norm": 1.4296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9568,
+      "step": 869
+    },
+    {
+      "epoch": 0.6953046953046953,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.0002,
+      "loss": 0.9478,
+      "step": 870
+    },
+    {
+      "epoch": 0.6961038961038961,
+      "grad_norm": 1.3984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9406,
+      "step": 871
+    },
+    {
+      "epoch": 0.6969030969030969,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.0002,
+      "loss": 0.9509,
+      "step": 872
+    },
+    {
+      "epoch": 0.6977022977022977,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9464,
+      "step": 873
+    },
+    {
+      "epoch": 0.6985014985014985,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.0002,
+      "loss": 0.949,
+      "step": 874
+    },
+    {
+      "epoch": 0.6993006993006993,
+      "grad_norm": 1.640625,
+      "learning_rate": 0.0002,
+      "loss": 0.9852,
+      "step": 875
+    },
+    {
+      "epoch": 0.7000999000999001,
+      "grad_norm": 1.375,
+      "learning_rate": 0.0002,
+      "loss": 0.9498,
+      "step": 876
+    },
+    {
+      "epoch": 0.7008991008991009,
+      "grad_norm": 0.890625,
+      "learning_rate": 0.0002,
+      "loss": 0.9519,
+      "step": 877
+    },
+    {
+      "epoch": 0.7016983016983017,
+      "grad_norm": 0.8125,
+      "learning_rate": 0.0002,
+      "loss": 0.9516,
+      "step": 878
+    },
+    {
+      "epoch": 0.7024975024975025,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9485,
+      "step": 879
+    },
+    {
+      "epoch": 0.7032967032967034,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.0002,
+      "loss": 0.9507,
+      "step": 880
+    },
+    {
+      "epoch": 0.7040959040959041,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9412,
+      "step": 881
+    },
+    {
+      "epoch": 0.7048951048951049,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.0002,
+      "loss": 0.964,
+      "step": 882
+    },
+    {
+      "epoch": 0.7056943056943057,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0002,
+      "loss": 0.955,
+      "step": 883
+    },
+    {
+      "epoch": 0.7064935064935065,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0002,
+      "loss": 0.9485,
+      "step": 884
+    },
+    {
+      "epoch": 0.7072927072927073,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9502,
+      "step": 885
+    },
+    {
+      "epoch": 0.7080919080919081,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9631,
+      "step": 886
+    },
+    {
+      "epoch": 0.7088911088911088,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9518,
+      "step": 887
+    },
+    {
+      "epoch": 0.7096903096903097,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9429,
+      "step": 888
+    },
+    {
+      "epoch": 0.7104895104895105,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9486,
+      "step": 889
+    },
+    {
+      "epoch": 0.7112887112887113,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0002,
+      "loss": 0.9525,
+      "step": 890
+    },
+    {
+      "epoch": 0.7120879120879121,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0002,
+      "loss": 0.9462,
+      "step": 891
+    },
+    {
+      "epoch": 0.7128871128871129,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0002,
+      "loss": 0.9491,
+      "step": 892
+    },
+    {
+      "epoch": 0.7136863136863136,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.0002,
+      "loss": 0.9531,
+      "step": 893
+    },
+    {
+      "epoch": 0.7144855144855145,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0002,
+      "loss": 0.956,
+      "step": 894
+    },
+    {
+      "epoch": 0.7152847152847153,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.0002,
+      "loss": 0.9547,
+      "step": 895
+    },
+    {
+      "epoch": 0.7160839160839161,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0002,
+      "loss": 0.9549,
+      "step": 896
+    },
+    {
+      "epoch": 0.7168831168831169,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0002,
+      "loss": 0.9576,
+      "step": 897
+    },
+    {
+      "epoch": 0.7176823176823177,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9402,
+      "step": 898
+    },
+    {
+      "epoch": 0.7184815184815185,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0002,
+      "loss": 0.9507,
+      "step": 899
+    },
+    {
+      "epoch": 0.7192807192807192,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9507,
+      "step": 900
+    },
+    {
+      "epoch": 0.72007992007992,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0002,
+      "loss": 0.9494,
+      "step": 901
+    },
+    {
+      "epoch": 0.7208791208791209,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0002,
+      "loss": 0.9487,
+      "step": 902
+    },
+    {
+      "epoch": 0.7216783216783217,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0002,
+      "loss": 0.9491,
+      "step": 903
+    },
+    {
+      "epoch": 0.7224775224775225,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0002,
+      "loss": 0.957,
+      "step": 904
+    },
+    {
+      "epoch": 0.7232767232767233,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9517,
+      "step": 905
+    },
+    {
+      "epoch": 0.724075924075924,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9421,
+      "step": 906
+    },
+    {
+      "epoch": 0.7248751248751248,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0002,
+      "loss": 0.947,
+      "step": 907
+    },
+    {
+      "epoch": 0.7256743256743257,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.0002,
+      "loss": 0.9521,
+      "step": 908
+    },
+    {
+      "epoch": 0.7264735264735265,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.0002,
+      "loss": 0.9582,
+      "step": 909
+    },
+    {
+      "epoch": 0.7272727272727273,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.0002,
+      "loss": 0.9558,
+      "step": 910
+    },
+    {
+      "epoch": 0.7280719280719281,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.0002,
+      "loss": 0.9522,
+      "step": 911
+    },
+    {
+      "epoch": 0.7288711288711289,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0002,
+      "loss": 0.952,
+      "step": 912
+    },
+    {
+      "epoch": 0.7296703296703296,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0002,
+      "loss": 0.9445,
+      "step": 913
+    },
+    {
+      "epoch": 0.7304695304695304,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.0002,
+      "loss": 0.9388,
+      "step": 914
+    },
+    {
+      "epoch": 0.7312687312687313,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0002,
+      "loss": 0.9478,
+      "step": 915
+    },
+    {
+      "epoch": 0.7320679320679321,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9517,
+      "step": 916
+    },
+    {
+      "epoch": 0.7328671328671329,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.0002,
+      "loss": 0.9487,
+      "step": 917
+    },
+    {
+      "epoch": 0.7336663336663337,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0002,
+      "loss": 0.9378,
+      "step": 918
+    },
+    {
+      "epoch": 0.7344655344655344,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0002,
+      "loss": 0.946,
+      "step": 919
+    },
+    {
+      "epoch": 0.7352647352647352,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9529,
+      "step": 920
+    },
+    {
+      "epoch": 0.736063936063936,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0002,
+      "loss": 0.9515,
+      "step": 921
+    },
+    {
+      "epoch": 0.7368631368631369,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0002,
+      "loss": 0.9501,
+      "step": 922
+    },
+    {
+      "epoch": 0.7376623376623377,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0002,
+      "loss": 0.9452,
+      "step": 923
+    },
+    {
+      "epoch": 0.7384615384615385,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0002,
+      "loss": 0.9516,
+      "step": 924
+    },
+    {
+      "epoch": 0.7392607392607392,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0002,
+      "loss": 0.9302,
+      "step": 925
+    },
+    {
+      "epoch": 0.74005994005994,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0002,
+      "loss": 0.9502,
+      "step": 926
+    },
+    {
+      "epoch": 0.7408591408591408,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9408,
+      "step": 927
+    },
+    {
+      "epoch": 0.7416583416583417,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0002,
+      "loss": 0.9509,
+      "step": 928
+    },
+    {
+      "epoch": 0.7424575424575425,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.0002,
+      "loss": 0.9541,
+      "step": 929
+    },
+    {
+      "epoch": 0.7432567432567433,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9463,
+      "step": 930
+    },
+    {
+      "epoch": 0.7440559440559441,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9545,
+      "step": 931
+    },
+    {
+      "epoch": 0.7448551448551448,
+      "grad_norm": 0.875,
+      "learning_rate": 0.0002,
+      "loss": 0.9542,
+      "step": 932
+    },
+    {
+      "epoch": 0.7456543456543456,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0002,
+      "loss": 0.9428,
+      "step": 933
+    },
+    {
+      "epoch": 0.7464535464535464,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9433,
+      "step": 934
+    },
+    {
+      "epoch": 0.7472527472527473,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.0002,
+      "loss": 0.94,
+      "step": 935
+    },
+    {
+      "epoch": 0.7480519480519481,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0002,
+      "loss": 0.9431,
+      "step": 936
+    },
+    {
+      "epoch": 0.7488511488511489,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0002,
+      "loss": 0.9539,
+      "step": 937
+    },
+    {
+      "epoch": 0.7496503496503496,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.0002,
+      "loss": 0.9458,
+      "step": 938
+    },
+    {
+      "epoch": 0.7504495504495504,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0002,
+      "loss": 0.9423,
+      "step": 939
+    },
+    {
+      "epoch": 0.7512487512487512,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9488,
+      "step": 940
+    },
+    {
+      "epoch": 0.752047952047952,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0002,
+      "loss": 0.9415,
+      "step": 941
+    },
+    {
+      "epoch": 0.7528471528471529,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.0002,
+      "loss": 0.9544,
+      "step": 942
+    },
+    {
+      "epoch": 0.7536463536463537,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9465,
+      "step": 943
+    },
+    {
+      "epoch": 0.7544455544455545,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0002,
+      "loss": 0.9419,
+      "step": 944
+    },
+    {
+      "epoch": 0.7552447552447552,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9473,
+      "step": 945
+    },
+    {
+      "epoch": 0.756043956043956,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9431,
+      "step": 946
+    },
+    {
+      "epoch": 0.7568431568431568,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.0002,
+      "loss": 0.9423,
+      "step": 947
+    },
+    {
+      "epoch": 0.7576423576423577,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.0002,
+      "loss": 0.9454,
+      "step": 948
+    },
+    {
+      "epoch": 0.7584415584415585,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.0002,
+      "loss": 0.9414,
+      "step": 949
+    },
+    {
+      "epoch": 0.7592407592407593,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0002,
+      "loss": 0.9487,
+      "step": 950
+    },
+    {
+      "epoch": 0.76003996003996,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.0002,
+      "loss": 0.9491,
+      "step": 951
+    },
+    {
+      "epoch": 0.7608391608391608,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9462,
+      "step": 952
+    },
+    {
+      "epoch": 0.7616383616383616,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9394,
+      "step": 953
+    },
+    {
+      "epoch": 0.7624375624375624,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9415,
+      "step": 954
+    },
+    {
+      "epoch": 0.7632367632367633,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9429,
+      "step": 955
+    },
+    {
+      "epoch": 0.7640359640359641,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0002,
+      "loss": 0.9425,
+      "step": 956
+    },
+    {
+      "epoch": 0.7648351648351648,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0002,
+      "loss": 0.9479,
+      "step": 957
+    },
+    {
+      "epoch": 0.7656343656343656,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0002,
+      "loss": 0.9508,
+      "step": 958
+    },
+    {
+      "epoch": 0.7664335664335664,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0002,
+      "loss": 0.9442,
+      "step": 959
+    },
+    {
+      "epoch": 0.7672327672327672,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.0002,
+      "loss": 0.9442,
+      "step": 960
+    },
+    {
+      "epoch": 0.768031968031968,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9427,
+      "step": 961
+    },
+    {
+      "epoch": 0.7688311688311689,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.0002,
+      "loss": 0.9517,
+      "step": 962
+    },
+    {
+      "epoch": 0.7696303696303697,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0002,
+      "loss": 0.936,
+      "step": 963
+    },
+    {
+      "epoch": 0.7704295704295704,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.0002,
+      "loss": 0.9431,
+      "step": 964
+    },
+    {
+      "epoch": 0.7712287712287712,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9467,
+      "step": 965
+    },
+    {
+      "epoch": 0.772027972027972,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9394,
+      "step": 966
+    },
+    {
+      "epoch": 0.7728271728271728,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9481,
+      "step": 967
+    },
+    {
+      "epoch": 0.7736263736263737,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.0002,
+      "loss": 0.9515,
+      "step": 968
+    },
+    {
+      "epoch": 0.7744255744255745,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0002,
+      "loss": 0.9535,
+      "step": 969
+    },
+    {
+      "epoch": 0.7752247752247752,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9392,
+      "step": 970
+    },
+    {
+      "epoch": 0.776023976023976,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.0002,
+      "loss": 0.9418,
+      "step": 971
+    },
+    {
+      "epoch": 0.7768231768231768,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.0002,
+      "loss": 0.936,
+      "step": 972
+    },
+    {
+      "epoch": 0.7776223776223776,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0002,
+      "loss": 0.9343,
+      "step": 973
+    },
+    {
+      "epoch": 0.7784215784215784,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0002,
+      "loss": 0.9445,
+      "step": 974
+    },
+    {
+      "epoch": 0.7792207792207793,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0002,
+      "loss": 0.931,
+      "step": 975
+    },
+    {
+      "epoch": 0.7800199800199801,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.0002,
+      "loss": 0.9308,
+      "step": 976
+    },
+    {
+      "epoch": 0.7808191808191808,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.0002,
+      "loss": 0.9491,
+      "step": 977
+    },
+    {
+      "epoch": 0.7816183816183816,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0002,
+      "loss": 0.9509,
+      "step": 978
+    },
+    {
+      "epoch": 0.7824175824175824,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.0002,
+      "loss": 0.9433,
+      "step": 979
+    },
+    {
+      "epoch": 0.7832167832167832,
+      "grad_norm": 1.765625,
+      "learning_rate": 0.0002,
+      "loss": 0.9743,
+      "step": 980
+    },
+    {
+      "epoch": 0.784015984015984,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0002,
+      "loss": 0.9467,
+      "step": 981
+    },
+    {
+      "epoch": 0.7848151848151849,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.0002,
+      "loss": 0.9535,
+      "step": 982
+    },
+    {
+      "epoch": 0.7856143856143856,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0002,
+      "loss": 0.9484,
+      "step": 983
+    },
+    {
+      "epoch": 0.7864135864135864,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9344,
+      "step": 984
+    },
+    {
+      "epoch": 0.7872127872127872,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0002,
+      "loss": 0.9354,
+      "step": 985
+    },
+    {
+      "epoch": 0.788011988011988,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9421,
+      "step": 986
+    },
+    {
+      "epoch": 0.7888111888111888,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9457,
+      "step": 987
+    },
+    {
+      "epoch": 0.7896103896103897,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0002,
+      "loss": 0.9405,
+      "step": 988
+    },
+    {
+      "epoch": 0.7904095904095904,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0002,
+      "loss": 0.9322,
+      "step": 989
+    },
+    {
+      "epoch": 0.7912087912087912,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9431,
+      "step": 990
+    },
+    {
+      "epoch": 0.792007992007992,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0002,
+      "loss": 0.9385,
+      "step": 991
+    },
+    {
+      "epoch": 0.7928071928071928,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0002,
+      "loss": 0.9457,
+      "step": 992
+    },
+    {
+      "epoch": 0.7936063936063936,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0002,
+      "loss": 0.948,
+      "step": 993
+    },
+    {
+      "epoch": 0.7944055944055944,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9517,
+      "step": 994
+    },
+    {
+      "epoch": 0.7952047952047953,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0002,
+      "loss": 0.937,
+      "step": 995
+    },
+    {
+      "epoch": 0.796003996003996,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9394,
+      "step": 996
+    },
+    {
+      "epoch": 0.7968031968031968,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0002,
+      "loss": 0.9411,
+      "step": 997
+    },
+    {
+      "epoch": 0.7976023976023976,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0002,
+      "loss": 0.9384,
+      "step": 998
+    },
+    {
+      "epoch": 0.7984015984015984,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9469,
+      "step": 999
+    },
+    {
+      "epoch": 0.7992007992007992,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9444,
+      "step": 1000
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9444,
+      "step": 1001
+    },
+    {
+      "epoch": 0.8007992007992007,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9358,
+      "step": 1002
+    },
+    {
+      "epoch": 0.8015984015984016,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.0002,
+      "loss": 0.9524,
+      "step": 1003
+    },
+    {
+      "epoch": 0.8023976023976024,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.0002,
+      "loss": 0.9425,
+      "step": 1004
+    },
+    {
+      "epoch": 0.8031968031968032,
+      "grad_norm": 0.625,
+      "learning_rate": 0.0002,
+      "loss": 0.9398,
+      "step": 1005
+    },
+    {
+      "epoch": 0.803996003996004,
+      "grad_norm": 0.859375,
+      "learning_rate": 0.0002,
+      "loss": 0.9393,
+      "step": 1006
+    },
+    {
+      "epoch": 0.8047952047952048,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9391,
+      "step": 1007
+    },
+    {
+      "epoch": 0.8055944055944056,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.0002,
+      "loss": 0.9534,
+      "step": 1008
+    },
+    {
+      "epoch": 0.8063936063936064,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.0002,
+      "loss": 0.939,
+      "step": 1009
+    },
+    {
+      "epoch": 0.8071928071928072,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0002,
+      "loss": 0.9419,
+      "step": 1010
+    },
+    {
+      "epoch": 0.807992007992008,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.0002,
+      "loss": 0.9404,
+      "step": 1011
+    },
+    {
+      "epoch": 0.8087912087912088,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.0002,
+      "loss": 0.9411,
+      "step": 1012
+    },
+    {
+      "epoch": 0.8095904095904096,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0002,
+      "loss": 0.946,
+      "step": 1013
+    },
+    {
+      "epoch": 0.8103896103896104,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.0002,
+      "loss": 0.9412,
+      "step": 1014
+    },
+    {
+      "epoch": 0.8111888111888111,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9439,
+      "step": 1015
+    },
+    {
+      "epoch": 0.811988011988012,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9497,
+      "step": 1016
+    },
+    {
+      "epoch": 0.8127872127872128,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9415,
+      "step": 1017
+    },
+    {
+      "epoch": 0.8135864135864136,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0002,
+      "loss": 0.9309,
+      "step": 1018
+    },
+    {
+      "epoch": 0.8143856143856144,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.0002,
+      "loss": 0.9479,
+      "step": 1019
+    },
+    {
+      "epoch": 0.8151848151848152,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9433,
+      "step": 1020
+    },
+    {
+      "epoch": 0.8159840159840159,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9438,
+      "step": 1021
+    },
+    {
+      "epoch": 0.8167832167832167,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.0002,
+      "loss": 0.9352,
+      "step": 1022
+    },
+    {
+      "epoch": 0.8175824175824176,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0002,
+      "loss": 0.9438,
+      "step": 1023
+    },
+    {
+      "epoch": 0.8183816183816184,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.0002,
+      "loss": 0.9396,
+      "step": 1024
+    },
+    {
+      "epoch": 0.8191808191808192,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.0002,
+      "loss": 0.9516,
+      "step": 1025
+    },
+    {
+      "epoch": 0.81998001998002,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9346,
+      "step": 1026
+    },
+    {
+      "epoch": 0.8207792207792208,
+      "grad_norm": 0.94921875,
+      "learning_rate": 0.0002,
+      "loss": 0.9367,
+      "step": 1027
+    },
+    {
+      "epoch": 0.8215784215784215,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.0002,
+      "loss": 0.9428,
+      "step": 1028
+    },
+    {
+      "epoch": 0.8223776223776224,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.0002,
+      "loss": 0.9515,
+      "step": 1029
+    },
+    {
+      "epoch": 0.8231768231768232,
+      "grad_norm": 1.421875,
+      "learning_rate": 0.0002,
+      "loss": 0.9436,
+      "step": 1030
+    },
+    {
+      "epoch": 0.823976023976024,
+      "grad_norm": 0.88671875,
+      "learning_rate": 0.0002,
+      "loss": 0.9404,
+      "step": 1031
+    },
+    {
+      "epoch": 0.8247752247752248,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.0002,
+      "loss": 0.9489,
+      "step": 1032
+    },
+    {
+      "epoch": 0.8255744255744256,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9523,
+      "step": 1033
+    },
+    {
+      "epoch": 0.8263736263736263,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.0002,
+      "loss": 0.9348,
+      "step": 1034
+    },
+    {
+      "epoch": 0.8271728271728271,
+      "grad_norm": 0.828125,
+      "learning_rate": 0.0002,
+      "loss": 0.9511,
+      "step": 1035
+    },
+    {
+      "epoch": 0.827972027972028,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.0002,
+      "loss": 0.9498,
+      "step": 1036
+    },
+    {
+      "epoch": 0.8287712287712288,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.0002,
+      "loss": 0.9548,
+      "step": 1037
+    },
+    {
+      "epoch": 0.8295704295704296,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.0002,
+      "loss": 0.9519,
+      "step": 1038
+    },
+    {
+      "epoch": 0.8303696303696304,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.0002,
+      "loss": 0.9453,
+      "step": 1039
+    },
+    {
+      "epoch": 0.8311688311688312,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9471,
+      "step": 1040
+    },
+    {
+      "epoch": 0.8319680319680319,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.0002,
+      "loss": 0.9433,
+      "step": 1041
+    },
+    {
+      "epoch": 0.8327672327672327,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.0002,
+      "loss": 0.9504,
+      "step": 1042
+    },
+    {
+      "epoch": 0.8335664335664336,
+      "grad_norm": 0.98046875,
+      "learning_rate": 0.0002,
+      "loss": 0.9379,
+      "step": 1043
+    },
+    {
+      "epoch": 0.8343656343656344,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9442,
+      "step": 1044
+    },
+    {
+      "epoch": 0.8351648351648352,
+      "grad_norm": 0.83203125,
+      "learning_rate": 0.0002,
+      "loss": 0.9548,
+      "step": 1045
+    },
+    {
+      "epoch": 0.835964035964036,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.0002,
+      "loss": 0.9465,
+      "step": 1046
+    },
+    {
+      "epoch": 0.8367632367632367,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.0002,
+      "loss": 0.945,
+      "step": 1047
+    },
+    {
+      "epoch": 0.8375624375624375,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.0002,
+      "loss": 0.939,
+      "step": 1048
+    },
+    {
+      "epoch": 0.8383616383616384,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.0002,
+      "loss": 0.9467,
+      "step": 1049
+    },
+    {
+      "epoch": 0.8391608391608392,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.0002,
+      "loss": 0.9439,
+      "step": 1050
+    },
+    {
+      "epoch": 0.83996003996004,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0002,
+      "loss": 0.9456,
+      "step": 1051
+    },
+    {
+      "epoch": 0.8407592407592408,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9359,
+      "step": 1052
+    },
+    {
+      "epoch": 0.8415584415584415,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0002,
+      "loss": 0.9449,
+      "step": 1053
+    },
+    {
+      "epoch": 0.8423576423576423,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9441,
+      "step": 1054
+    },
+    {
+      "epoch": 0.8431568431568431,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0002,
+      "loss": 0.9414,
+      "step": 1055
+    },
+    {
+      "epoch": 0.843956043956044,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0002,
+      "loss": 0.944,
+      "step": 1056
+    },
+    {
+      "epoch": 0.8447552447552448,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.0002,
+      "loss": 0.9407,
+      "step": 1057
+    },
+    {
+      "epoch": 0.8455544455544456,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0002,
+      "loss": 0.9339,
+      "step": 1058
+    },
+    {
+      "epoch": 0.8463536463536464,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.0002,
+      "loss": 0.949,
+      "step": 1059
+    },
+    {
+      "epoch": 0.8471528471528471,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0002,
+      "loss": 0.9826,
+      "step": 1060
+    },
+    {
+      "epoch": 0.8479520479520479,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0002,
+      "loss": 0.9419,
+      "step": 1061
+    },
+    {
+      "epoch": 0.8487512487512487,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0002,
+      "loss": 0.9446,
+      "step": 1062
+    },
+    {
+      "epoch": 0.8495504495504496,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0002,
+      "loss": 0.9304,
+      "step": 1063
+    },
+    {
+      "epoch": 0.8503496503496504,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.0002,
+      "loss": 0.9671,
+      "step": 1064
+    },
+    {
+      "epoch": 0.8511488511488512,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0002,
+      "loss": 0.9449,
+      "step": 1065
+    },
+    {
+      "epoch": 0.8519480519480519,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0002,
+      "loss": 0.9448,
+      "step": 1066
+    },
+    {
+      "epoch": 0.8527472527472527,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9388,
+      "step": 1067
+    },
+    {
+      "epoch": 0.8535464535464535,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0002,
+      "loss": 0.9416,
+      "step": 1068
+    },
+    {
+      "epoch": 0.8543456543456543,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0002,
+      "loss": 0.9331,
+      "step": 1069
+    },
+    {
+      "epoch": 0.8551448551448552,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.0002,
+      "loss": 0.9439,
+      "step": 1070
+    },
+    {
+      "epoch": 0.855944055944056,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.0002,
+      "loss": 0.9393,
+      "step": 1071
+    },
+    {
+      "epoch": 0.8567432567432568,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.0002,
+      "loss": 0.946,
+      "step": 1072
+    },
+    {
+      "epoch": 0.8575424575424575,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0002,
+      "loss": 0.9413,
+      "step": 1073
+    },
+    {
+      "epoch": 0.8583416583416583,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.0002,
+      "loss": 0.9354,
+      "step": 1074
+    },
+    {
+      "epoch": 0.8591408591408591,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0002,
+      "loss": 0.9365,
+      "step": 1075
+    },
+    {
+      "epoch": 0.85994005994006,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.0002,
+      "loss": 0.9671,
+      "step": 1076
+    },
+    {
+      "epoch": 0.8607392607392608,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9513,
+      "step": 1077
+    },
+    {
+      "epoch": 0.8615384615384616,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0002,
+      "loss": 0.9385,
+      "step": 1078
+    },
+    {
+      "epoch": 0.8623376623376623,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0002,
+      "loss": 0.9421,
+      "step": 1079
+    },
+    {
+      "epoch": 0.8631368631368631,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0002,
+      "loss": 0.9312,
+      "step": 1080
+    },
+    {
+      "epoch": 0.8639360639360639,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9371,
+      "step": 1081
+    },
+    {
+      "epoch": 0.8647352647352647,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.0002,
+      "loss": 0.9436,
+      "step": 1082
+    },
+    {
+      "epoch": 0.8655344655344656,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0002,
+      "loss": 0.9632,
+      "step": 1083
+    },
+    {
+      "epoch": 0.8663336663336664,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9661,
+      "step": 1084
+    },
+    {
+      "epoch": 0.8671328671328671,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9284,
+      "step": 1085
+    },
+    {
+      "epoch": 0.8679320679320679,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9526,
+      "step": 1086
+    },
+    {
+      "epoch": 0.8687312687312687,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0002,
+      "loss": 0.943,
+      "step": 1087
+    },
+    {
+      "epoch": 0.8695304695304695,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0002,
+      "loss": 0.9404,
+      "step": 1088
+    },
+    {
+      "epoch": 0.8703296703296703,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9402,
+      "step": 1089
+    },
+    {
+      "epoch": 0.8711288711288712,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.0002,
+      "loss": 0.9269,
+      "step": 1090
+    },
+    {
+      "epoch": 0.871928071928072,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.0002,
+      "loss": 0.9325,
+      "step": 1091
+    },
+    {
+      "epoch": 0.8727272727272727,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0002,
+      "loss": 0.9427,
+      "step": 1092
+    },
+    {
+      "epoch": 0.8735264735264735,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.0002,
+      "loss": 0.9423,
+      "step": 1093
+    },
+    {
+      "epoch": 0.8743256743256743,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0002,
+      "loss": 0.94,
+      "step": 1094
+    },
+    {
+      "epoch": 0.8751248751248751,
+      "grad_norm": 0.375,
+      "learning_rate": 0.0002,
+      "loss": 0.9444,
+      "step": 1095
+    },
+    {
+      "epoch": 0.875924075924076,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9447,
+      "step": 1096
+    },
+    {
+      "epoch": 0.8767232767232768,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0002,
+      "loss": 0.939,
+      "step": 1097
+    },
+    {
+      "epoch": 0.8775224775224775,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9329,
+      "step": 1098
+    },
+    {
+      "epoch": 0.8783216783216783,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0002,
+      "loss": 0.9462,
+      "step": 1099
+    },
+    {
+      "epoch": 0.8791208791208791,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9465,
+      "step": 1100
+    },
+    {
+      "epoch": 0.8799200799200799,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.0002,
+      "loss": 0.9393,
+      "step": 1101
+    },
+    {
+      "epoch": 0.8807192807192807,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.0002,
+      "loss": 0.9396,
+      "step": 1102
+    },
+    {
+      "epoch": 0.8815184815184816,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9427,
+      "step": 1103
+    },
+    {
+      "epoch": 0.8823176823176824,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0002,
+      "loss": 0.9457,
+      "step": 1104
+    },
+    {
+      "epoch": 0.8831168831168831,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0002,
+      "loss": 0.9512,
+      "step": 1105
+    },
+    {
+      "epoch": 0.8839160839160839,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9462,
+      "step": 1106
+    },
+    {
+      "epoch": 0.8847152847152847,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0002,
+      "loss": 0.9364,
+      "step": 1107
+    },
+    {
+      "epoch": 0.8855144855144855,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9409,
+      "step": 1108
+    },
+    {
+      "epoch": 0.8863136863136863,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9398,
+      "step": 1109
+    },
+    {
+      "epoch": 0.8871128871128872,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0002,
+      "loss": 0.9413,
+      "step": 1110
+    },
+    {
+      "epoch": 0.8879120879120879,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0002,
+      "loss": 0.9394,
+      "step": 1111
+    },
+    {
+      "epoch": 0.8887112887112887,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0002,
+      "loss": 0.932,
+      "step": 1112
+    },
+    {
+      "epoch": 0.8895104895104895,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.0002,
+      "loss": 0.9468,
+      "step": 1113
+    },
+    {
+      "epoch": 0.8903096903096903,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0002,
+      "loss": 0.9449,
+      "step": 1114
+    },
+    {
+      "epoch": 0.8911088911088911,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0002,
+      "loss": 0.9555,
+      "step": 1115
+    },
+    {
+      "epoch": 0.891908091908092,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.0002,
+      "loss": 0.9366,
+      "step": 1116
+    },
+    {
+      "epoch": 0.8927072927072927,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9405,
+      "step": 1117
+    },
+    {
+      "epoch": 0.8935064935064935,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0002,
+      "loss": 0.9425,
+      "step": 1118
+    },
+    {
+      "epoch": 0.8943056943056943,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.0002,
+      "loss": 0.9437,
+      "step": 1119
+    },
+    {
+      "epoch": 0.8951048951048951,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0002,
+      "loss": 0.9488,
+      "step": 1120
+    },
+    {
+      "epoch": 0.8959040959040959,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9472,
+      "step": 1121
+    },
+    {
+      "epoch": 0.8967032967032967,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.0002,
+      "loss": 0.9304,
+      "step": 1122
+    },
+    {
+      "epoch": 0.8975024975024976,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.0002,
+      "loss": 0.9383,
+      "step": 1123
+    },
+    {
+      "epoch": 0.8983016983016983,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0002,
+      "loss": 0.9412,
+      "step": 1124
+    },
+    {
+      "epoch": 0.8991008991008991,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9414,
+      "step": 1125
+    },
+    {
+      "epoch": 0.8999000999000999,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.0002,
+      "loss": 0.9443,
+      "step": 1126
+    },
+    {
+      "epoch": 0.9006993006993007,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0002,
+      "loss": 0.9346,
+      "step": 1127
+    },
+    {
+      "epoch": 0.9014985014985015,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0002,
+      "loss": 0.944,
+      "step": 1128
+    },
+    {
+      "epoch": 0.9022977022977023,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9511,
+      "step": 1129
+    },
+    {
+      "epoch": 0.903096903096903,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9372,
+      "step": 1130
+    },
+    {
+      "epoch": 0.9038961038961039,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0002,
+      "loss": 0.9308,
+      "step": 1131
+    },
+    {
+      "epoch": 0.9046953046953047,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.0002,
+      "loss": 0.9475,
+      "step": 1132
+    },
+    {
+      "epoch": 0.9054945054945055,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9335,
+      "step": 1133
+    },
+    {
+      "epoch": 0.9062937062937063,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9362,
+      "step": 1134
+    },
+    {
+      "epoch": 0.9070929070929071,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9465,
+      "step": 1135
+    },
+    {
+      "epoch": 0.907892107892108,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0002,
+      "loss": 0.936,
+      "step": 1136
+    },
+    {
+      "epoch": 0.9086913086913087,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0002,
+      "loss": 0.933,
+      "step": 1137
+    },
+    {
+      "epoch": 0.9094905094905095,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0002,
+      "loss": 0.9399,
+      "step": 1138
+    },
+    {
+      "epoch": 0.9102897102897103,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0002,
+      "loss": 0.9511,
+      "step": 1139
+    },
+    {
+      "epoch": 0.9110889110889111,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0002,
+      "loss": 0.9392,
+      "step": 1140
+    },
+    {
+      "epoch": 0.9118881118881119,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0002,
+      "loss": 0.9357,
+      "step": 1141
+    },
+    {
+      "epoch": 0.9126873126873127,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0002,
+      "loss": 0.9328,
+      "step": 1142
+    },
+    {
+      "epoch": 0.9134865134865134,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9438,
+      "step": 1143
+    },
+    {
+      "epoch": 0.9142857142857143,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9275,
+      "step": 1144
+    },
+    {
+      "epoch": 0.9150849150849151,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9262,
+      "step": 1145
+    },
+    {
+      "epoch": 0.9158841158841159,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0002,
+      "loss": 0.9362,
+      "step": 1146
+    },
+    {
+      "epoch": 0.9166833166833167,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9398,
+      "step": 1147
+    },
+    {
+      "epoch": 0.9174825174825175,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0002,
+      "loss": 0.9338,
+      "step": 1148
+    },
+    {
+      "epoch": 0.9182817182817182,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0002,
+      "loss": 0.9387,
+      "step": 1149
+    },
+    {
+      "epoch": 0.919080919080919,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0002,
+      "loss": 0.9438,
+      "step": 1150
+    },
+    {
+      "epoch": 0.9198801198801199,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.0002,
+      "loss": 0.9406,
+      "step": 1151
+    },
+    {
+      "epoch": 0.9206793206793207,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0002,
+      "loss": 0.9468,
+      "step": 1152
+    },
+    {
+      "epoch": 0.9214785214785215,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0002,
+      "loss": 0.9333,
+      "step": 1153
+    },
+    {
+      "epoch": 0.9222777222777223,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.0002,
+      "loss": 0.9339,
+      "step": 1154
+    },
+    {
+      "epoch": 0.9230769230769231,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0002,
+      "loss": 0.9395,
+      "step": 1155
+    },
+    {
+      "epoch": 0.9238761238761238,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.0002,
+      "loss": 0.9363,
+      "step": 1156
+    },
+    {
+      "epoch": 0.9246753246753247,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9442,
+      "step": 1157
+    },
+    {
+      "epoch": 0.9254745254745255,
+      "grad_norm": 2.375,
+      "learning_rate": 0.0002,
+      "loss": 0.9493,
+      "step": 1158
+    },
+    {
+      "epoch": 0.9262737262737263,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9541,
+      "step": 1159
+    },
+    {
+      "epoch": 0.9270729270729271,
+      "grad_norm": 3.546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9484,
+      "step": 1160
+    },
+    {
+      "epoch": 0.9278721278721279,
+      "grad_norm": 2.90625,
+      "learning_rate": 0.0002,
+      "loss": 0.9599,
+      "step": 1161
+    },
+    {
+      "epoch": 0.9286713286713286,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.0002,
+      "loss": 0.9439,
+      "step": 1162
+    },
+    {
+      "epoch": 0.9294705294705294,
+      "grad_norm": 0.94140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9559,
+      "step": 1163
+    },
+    {
+      "epoch": 0.9302697302697303,
+      "grad_norm": 0.87109375,
+      "learning_rate": 0.0002,
+      "loss": 0.9517,
+      "step": 1164
+    },
+    {
+      "epoch": 0.9310689310689311,
+      "grad_norm": 0.98828125,
+      "learning_rate": 0.0002,
+      "loss": 0.9494,
+      "step": 1165
+    },
+    {
+      "epoch": 0.9318681318681319,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.0002,
+      "loss": 0.9535,
+      "step": 1166
+    },
+    {
+      "epoch": 0.9326673326673327,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.0002,
+      "loss": 0.9412,
+      "step": 1167
+    },
+    {
+      "epoch": 0.9334665334665335,
+      "grad_norm": 1.375,
+      "learning_rate": 0.0002,
+      "loss": 0.9478,
+      "step": 1168
+    },
+    {
+      "epoch": 0.9342657342657342,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9432,
+      "step": 1169
+    },
+    {
+      "epoch": 0.935064935064935,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9486,
+      "step": 1170
+    },
+    {
+      "epoch": 0.9358641358641359,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9442,
+      "step": 1171
+    },
+    {
+      "epoch": 0.9366633366633367,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9525,
+      "step": 1172
+    },
+    {
+      "epoch": 0.9374625374625375,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.0002,
+      "loss": 0.9447,
+      "step": 1173
+    },
+    {
+      "epoch": 0.9382617382617383,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.0002,
+      "loss": 0.9462,
+      "step": 1174
+    },
+    {
+      "epoch": 0.939060939060939,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.0002,
+      "loss": 0.9585,
+      "step": 1175
+    },
+    {
+      "epoch": 0.9398601398601398,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.0002,
+      "loss": 0.95,
+      "step": 1176
+    },
+    {
+      "epoch": 0.9406593406593406,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.0002,
+      "loss": 0.9418,
+      "step": 1177
+    },
+    {
+      "epoch": 0.9414585414585415,
+      "grad_norm": 1.0,
+      "learning_rate": 0.0002,
+      "loss": 0.938,
+      "step": 1178
+    },
+    {
+      "epoch": 0.9422577422577423,
+      "grad_norm": 0.84375,
+      "learning_rate": 0.0002,
+      "loss": 0.9535,
+      "step": 1179
+    },
+    {
+      "epoch": 0.9430569430569431,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.0002,
+      "loss": 0.941,
+      "step": 1180
+    },
+    {
+      "epoch": 0.9438561438561438,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.0002,
+      "loss": 0.9446,
+      "step": 1181
+    },
+    {
+      "epoch": 0.9446553446553446,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9406,
+      "step": 1182
+    },
+    {
+      "epoch": 0.9454545454545454,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.0002,
+      "loss": 0.9593,
+      "step": 1183
+    },
+    {
+      "epoch": 0.9462537462537463,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.0002,
+      "loss": 0.9334,
+      "step": 1184
+    },
+    {
+      "epoch": 0.9470529470529471,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0002,
+      "loss": 0.9429,
+      "step": 1185
+    },
+    {
+      "epoch": 0.9478521478521479,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.0002,
+      "loss": 0.9419,
+      "step": 1186
+    },
+    {
+      "epoch": 0.9486513486513487,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0002,
+      "loss": 0.9314,
+      "step": 1187
+    },
+    {
+      "epoch": 0.9494505494505494,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0002,
+      "loss": 0.936,
+      "step": 1188
+    },
+    {
+      "epoch": 0.9502497502497502,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0002,
+      "loss": 0.9393,
+      "step": 1189
+    },
+    {
+      "epoch": 0.951048951048951,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.0002,
+      "loss": 0.932,
+      "step": 1190
+    },
+    {
+      "epoch": 0.9518481518481519,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9316,
+      "step": 1191
+    },
+    {
+      "epoch": 0.9526473526473527,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0002,
+      "loss": 0.9404,
+      "step": 1192
+    },
+    {
+      "epoch": 0.9534465534465535,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0002,
+      "loss": 0.9301,
+      "step": 1193
+    },
+    {
+      "epoch": 0.9542457542457542,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9331,
+      "step": 1194
+    },
+    {
+      "epoch": 0.955044955044955,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9403,
+      "step": 1195
+    },
+    {
+      "epoch": 0.9558441558441558,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.0002,
+      "loss": 0.9766,
+      "step": 1196
+    },
+    {
+      "epoch": 0.9566433566433566,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0002,
+      "loss": 0.94,
+      "step": 1197
+    },
+    {
+      "epoch": 0.9574425574425575,
+      "grad_norm": 0.94140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9464,
+      "step": 1198
+    },
+    {
+      "epoch": 0.9582417582417583,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0002,
+      "loss": 0.944,
+      "step": 1199
+    },
+    {
+      "epoch": 0.9590409590409591,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0002,
+      "loss": 0.9389,
+      "step": 1200
+    },
+    {
+      "epoch": 0.9598401598401598,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.0002,
+      "loss": 0.947,
+      "step": 1201
+    },
+    {
+      "epoch": 0.9606393606393606,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0002,
+      "loss": 0.9404,
+      "step": 1202
+    },
+    {
+      "epoch": 0.9614385614385614,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.0002,
+      "loss": 0.9371,
+      "step": 1203
+    },
+    {
+      "epoch": 0.9622377622377623,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0002,
+      "loss": 0.9316,
+      "step": 1204
+    },
+    {
+      "epoch": 0.9630369630369631,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.0002,
+      "loss": 0.9416,
+      "step": 1205
+    },
+    {
+      "epoch": 0.9638361638361639,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0002,
+      "loss": 0.9433,
+      "step": 1206
+    },
+    {
+      "epoch": 0.9646353646353646,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0002,
+      "loss": 0.9397,
+      "step": 1207
+    },
+    {
+      "epoch": 0.9654345654345654,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0002,
+      "loss": 0.9428,
+      "step": 1208
+    },
+    {
+      "epoch": 0.9662337662337662,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9506,
+      "step": 1209
+    },
+    {
+      "epoch": 0.967032967032967,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0002,
+      "loss": 0.9429,
+      "step": 1210
+    },
+    {
+      "epoch": 0.9678321678321679,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9355,
+      "step": 1211
+    },
+    {
+      "epoch": 0.9686313686313687,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0002,
+      "loss": 0.943,
+      "step": 1212
+    },
+    {
+      "epoch": 0.9694305694305694,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.0002,
+      "loss": 0.9498,
+      "step": 1213
+    },
+    {
+      "epoch": 0.9702297702297702,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.0002,
+      "loss": 0.9406,
+      "step": 1214
+    },
+    {
+      "epoch": 0.971028971028971,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0002,
+      "loss": 0.9395,
+      "step": 1215
+    },
+    {
+      "epoch": 0.9718281718281718,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0002,
+      "loss": 0.9436,
+      "step": 1216
+    },
+    {
+      "epoch": 0.9726273726273726,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9375,
+      "step": 1217
+    },
+    {
+      "epoch": 0.9734265734265735,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9407,
+      "step": 1218
+    },
+    {
+      "epoch": 0.9742257742257743,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.0002,
+      "loss": 0.9298,
+      "step": 1219
+    },
+    {
+      "epoch": 0.975024975024975,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0002,
+      "loss": 0.9448,
+      "step": 1220
+    },
+    {
+      "epoch": 0.9758241758241758,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.0002,
+      "loss": 0.94,
+      "step": 1221
+    },
+    {
+      "epoch": 0.9766233766233766,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.0002,
+      "loss": 0.9365,
+      "step": 1222
+    },
+    {
+      "epoch": 0.9774225774225774,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0002,
+      "loss": 0.9435,
+      "step": 1223
+    },
+    {
+      "epoch": 0.9782217782217782,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9354,
+      "step": 1224
+    },
+    {
+      "epoch": 0.9790209790209791,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9338,
+      "step": 1225
+    },
+    {
+      "epoch": 0.9798201798201798,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0002,
+      "loss": 0.9355,
+      "step": 1226
+    },
+    {
+      "epoch": 0.9806193806193806,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0002,
+      "loss": 0.938,
+      "step": 1227
+    },
+    {
+      "epoch": 0.9814185814185814,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.0002,
+      "loss": 0.9358,
+      "step": 1228
+    },
+    {
+      "epoch": 0.9822177822177822,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9282,
+      "step": 1229
+    },
+    {
+      "epoch": 0.983016983016983,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.0002,
+      "loss": 0.936,
+      "step": 1230
+    },
+    {
+      "epoch": 0.9838161838161839,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.0002,
+      "loss": 0.9365,
+      "step": 1231
+    },
+    {
+      "epoch": 0.9846153846153847,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0002,
+      "loss": 0.9384,
+      "step": 1232
+    },
+    {
+      "epoch": 0.9854145854145854,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9337,
+      "step": 1233
+    },
+    {
+      "epoch": 0.9862137862137862,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0002,
+      "loss": 0.9407,
+      "step": 1234
+    },
+    {
+      "epoch": 0.987012987012987,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0002,
+      "loss": 0.9377,
+      "step": 1235
+    },
+    {
+      "epoch": 0.9878121878121878,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0002,
+      "loss": 0.9336,
+      "step": 1236
+    },
+    {
+      "epoch": 0.9886113886113886,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9384,
+      "step": 1237
+    },
+    {
+      "epoch": 0.9894105894105895,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0002,
+      "loss": 0.9239,
+      "step": 1238
+    },
+    {
+      "epoch": 0.9902097902097902,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9316,
+      "step": 1239
+    },
+    {
+      "epoch": 0.991008991008991,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9388,
+      "step": 1240
+    },
+    {
+      "epoch": 0.9918081918081918,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9426,
+      "step": 1241
+    },
+    {
+      "epoch": 0.9926073926073926,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0002,
+      "loss": 0.9359,
+      "step": 1242
+    },
+    {
+      "epoch": 0.9934065934065934,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0002,
+      "loss": 0.944,
+      "step": 1243
+    },
+    {
+      "epoch": 0.9942057942057942,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0002,
+      "loss": 0.9356,
+      "step": 1244
+    },
+    {
+      "epoch": 0.995004995004995,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9404,
+      "step": 1245
+    },
+    {
+      "epoch": 0.9958041958041958,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0002,
+      "loss": 0.9422,
+      "step": 1246
+    },
+    {
+      "epoch": 0.9966033966033966,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0002,
+      "loss": 0.9377,
+      "step": 1247
+    },
+    {
+      "epoch": 0.9974025974025974,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9373,
+      "step": 1248
+    },
+    {
+      "epoch": 0.9982017982017982,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.0002,
+      "loss": 0.9334,
+      "step": 1249
+    },
+    {
+      "epoch": 0.999000999000999,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9355,
+      "step": 1250
+    },
+    {
+      "epoch": 0.9998001998001999,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9339,
+      "step": 1251
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0002,
+      "loss": 0.2358,
+      "step": 1252
+    },
+    {
+      "epoch": 1.0007992007992008,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.0002,
+      "loss": 0.9372,
+      "step": 1253
+    },
+    {
+      "epoch": 1.0015984015984016,
+      "grad_norm": 0.75,
+      "learning_rate": 0.0002,
+      "loss": 0.9368,
+      "step": 1254
+    },
+    {
+      "epoch": 1.0023976023976024,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9396,
+      "step": 1255
+    },
+    {
+      "epoch": 1.0031968031968033,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9409,
+      "step": 1256
+    },
+    {
+      "epoch": 1.003996003996004,
+      "grad_norm": 0.625,
+      "learning_rate": 0.0002,
+      "loss": 0.932,
+      "step": 1257
+    },
+    {
+      "epoch": 1.004795204795205,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0002,
+      "loss": 0.9329,
+      "step": 1258
+    },
+    {
+      "epoch": 1.0055944055944055,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.0002,
+      "loss": 0.936,
+      "step": 1259
+    },
+    {
+      "epoch": 1.0063936063936063,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0002,
+      "loss": 0.9437,
+      "step": 1260
+    },
+    {
+      "epoch": 1.0071928071928071,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.0002,
+      "loss": 0.9371,
+      "step": 1261
+    },
+    {
+      "epoch": 1.007992007992008,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0002,
+      "loss": 0.9371,
+      "step": 1262
+    },
+    {
+      "epoch": 1.0087912087912088,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0002,
+      "loss": 0.9259,
+      "step": 1263
+    },
+    {
+      "epoch": 1.0095904095904096,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9412,
+      "step": 1264
+    },
+    {
+      "epoch": 1.0103896103896104,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0002,
+      "loss": 0.933,
+      "step": 1265
+    },
+    {
+      "epoch": 1.0111888111888112,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9364,
+      "step": 1266
+    },
+    {
+      "epoch": 1.011988011988012,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9288,
+      "step": 1267
+    },
+    {
+      "epoch": 1.0127872127872128,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0002,
+      "loss": 0.9485,
+      "step": 1268
+    },
+    {
+      "epoch": 1.0135864135864137,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9283,
+      "step": 1269
+    },
+    {
+      "epoch": 1.0143856143856145,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0002,
+      "loss": 0.9278,
+      "step": 1270
+    },
+    {
+      "epoch": 1.0151848151848153,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0002,
+      "loss": 0.93,
+      "step": 1271
+    },
+    {
+      "epoch": 1.0159840159840159,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9295,
+      "step": 1272
+    },
+    {
+      "epoch": 1.0167832167832167,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0002,
+      "loss": 0.939,
+      "step": 1273
+    },
+    {
+      "epoch": 1.0175824175824175,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0002,
+      "loss": 0.9304,
+      "step": 1274
+    },
+    {
+      "epoch": 1.0183816183816183,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0002,
+      "loss": 0.9326,
+      "step": 1275
+    },
+    {
+      "epoch": 1.0191808191808192,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0002,
+      "loss": 0.9334,
+      "step": 1276
+    },
+    {
+      "epoch": 1.01998001998002,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.0002,
+      "loss": 0.9164,
+      "step": 1277
+    },
+    {
+      "epoch": 1.0207792207792208,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.0002,
+      "loss": 0.9679,
+      "step": 1278
+    },
+    {
+      "epoch": 1.0215784215784216,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9393,
+      "step": 1279
+    },
+    {
+      "epoch": 1.0223776223776224,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9296,
+      "step": 1280
+    },
+    {
+      "epoch": 1.0231768231768232,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.0002,
+      "loss": 0.9391,
+      "step": 1281
+    },
+    {
+      "epoch": 1.023976023976024,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0002,
+      "loss": 0.939,
+      "step": 1282
+    },
+    {
+      "epoch": 1.0247752247752249,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.0002,
+      "loss": 0.9336,
+      "step": 1283
+    },
+    {
+      "epoch": 1.0255744255744257,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9453,
+      "step": 1284
+    },
+    {
+      "epoch": 1.0263736263736263,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9352,
+      "step": 1285
+    },
+    {
+      "epoch": 1.027172827172827,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0002,
+      "loss": 0.9384,
+      "step": 1286
+    },
+    {
+      "epoch": 1.027972027972028,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0002,
+      "loss": 0.9434,
+      "step": 1287
+    },
+    {
+      "epoch": 1.0287712287712287,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0002,
+      "loss": 0.9381,
+      "step": 1288
+    },
+    {
+      "epoch": 1.0295704295704295,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.0002,
+      "loss": 0.9455,
+      "step": 1289
+    },
+    {
+      "epoch": 1.0303696303696304,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.0002,
+      "loss": 0.9323,
+      "step": 1290
+    },
+    {
+      "epoch": 1.0311688311688312,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0002,
+      "loss": 0.9405,
+      "step": 1291
+    },
+    {
+      "epoch": 1.031968031968032,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0002,
+      "loss": 0.9367,
+      "step": 1292
+    },
+    {
+      "epoch": 1.0327672327672328,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9288,
+      "step": 1293
+    },
+    {
+      "epoch": 1.0335664335664336,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9355,
+      "step": 1294
+    },
+    {
+      "epoch": 1.0343656343656344,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9339,
+      "step": 1295
+    },
+    {
+      "epoch": 1.0351648351648353,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0002,
+      "loss": 0.9437,
+      "step": 1296
+    },
+    {
+      "epoch": 1.035964035964036,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9289,
+      "step": 1297
+    },
+    {
+      "epoch": 1.0367632367632367,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0002,
+      "loss": 0.9286,
+      "step": 1298
+    },
+    {
+      "epoch": 1.0375624375624375,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.0002,
+      "loss": 0.9325,
+      "step": 1299
+    },
+    {
+      "epoch": 1.0383616383616383,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0002,
+      "loss": 0.9316,
+      "step": 1300
+    },
+    {
+      "epoch": 1.0391608391608391,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.0002,
+      "loss": 0.9319,
+      "step": 1301
+    },
+    {
+      "epoch": 1.03996003996004,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0002,
+      "loss": 0.933,
+      "step": 1302
+    },
+    {
+      "epoch": 1.0407592407592408,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0002,
+      "loss": 0.9247,
+      "step": 1303
+    },
+    {
+      "epoch": 1.0415584415584416,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9378,
+      "step": 1304
+    },
+    {
+      "epoch": 1.0423576423576424,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.0002,
+      "loss": 0.9313,
+      "step": 1305
+    },
+    {
+      "epoch": 1.0431568431568432,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0002,
+      "loss": 0.93,
+      "step": 1306
+    },
+    {
+      "epoch": 1.043956043956044,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0002,
+      "loss": 0.9392,
+      "step": 1307
+    },
+    {
+      "epoch": 1.0447552447552448,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0002,
+      "loss": 0.9397,
+      "step": 1308
+    },
+    {
+      "epoch": 1.0455544455544457,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0002,
+      "loss": 0.9317,
+      "step": 1309
+    },
+    {
+      "epoch": 1.0463536463536462,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9344,
+      "step": 1310
+    },
+    {
+      "epoch": 1.047152847152847,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.0002,
+      "loss": 0.9298,
+      "step": 1311
+    },
+    {
+      "epoch": 1.0479520479520479,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0002,
+      "loss": 0.9352,
+      "step": 1312
+    },
+    {
+      "epoch": 1.0487512487512487,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.0002,
+      "loss": 0.9258,
+      "step": 1313
+    },
+    {
+      "epoch": 1.0495504495504495,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9268,
+      "step": 1314
+    },
+    {
+      "epoch": 1.0503496503496503,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0002,
+      "loss": 0.9337,
+      "step": 1315
+    },
+    {
+      "epoch": 1.0511488511488511,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0002,
+      "loss": 0.9359,
+      "step": 1316
+    },
+    {
+      "epoch": 1.051948051948052,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0002,
+      "loss": 0.9341,
+      "step": 1317
+    },
+    {
+      "epoch": 1.0527472527472528,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.0002,
+      "loss": 0.9385,
+      "step": 1318
+    },
+    {
+      "epoch": 1.0535464535464536,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0002,
+      "loss": 0.9307,
+      "step": 1319
+    },
+    {
+      "epoch": 1.0543456543456544,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0002,
+      "loss": 0.9357,
+      "step": 1320
+    },
+    {
+      "epoch": 1.0551448551448552,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9395,
+      "step": 1321
+    },
+    {
+      "epoch": 1.055944055944056,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0002,
+      "loss": 0.9275,
+      "step": 1322
+    },
+    {
+      "epoch": 1.0567432567432566,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0002,
+      "loss": 0.9222,
+      "step": 1323
+    },
+    {
+      "epoch": 1.0575424575424575,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0002,
+      "loss": 0.9293,
+      "step": 1324
+    },
+    {
+      "epoch": 1.0583416583416583,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.0002,
+      "loss": 0.9581,
+      "step": 1325
+    },
+    {
+      "epoch": 1.059140859140859,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0002,
+      "loss": 0.9259,
+      "step": 1326
+    },
+    {
+      "epoch": 1.05994005994006,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0002,
+      "loss": 0.9327,
+      "step": 1327
+    },
+    {
+      "epoch": 1.0607392607392607,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9344,
+      "step": 1328
+    },
+    {
+      "epoch": 1.0615384615384615,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9431,
+      "step": 1329
+    },
+    {
+      "epoch": 1.0623376623376624,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0002,
+      "loss": 0.9363,
+      "step": 1330
+    },
+    {
+      "epoch": 1.0631368631368632,
+      "grad_norm": 0.375,
+      "learning_rate": 0.0002,
+      "loss": 0.9342,
+      "step": 1331
+    },
+    {
+      "epoch": 1.063936063936064,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9242,
+      "step": 1332
+    },
+    {
+      "epoch": 1.0647352647352648,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0002,
+      "loss": 0.9385,
+      "step": 1333
+    },
+    {
+      "epoch": 1.0655344655344656,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0002,
+      "loss": 0.9363,
+      "step": 1334
+    },
+    {
+      "epoch": 1.0663336663336662,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9367,
+      "step": 1335
+    },
+    {
+      "epoch": 1.067132867132867,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9381,
+      "step": 1336
+    },
+    {
+      "epoch": 1.0679320679320679,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.0002,
+      "loss": 0.9399,
+      "step": 1337
+    },
+    {
+      "epoch": 1.0687312687312687,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0002,
+      "loss": 0.9397,
+      "step": 1338
+    },
+    {
+      "epoch": 1.0695304695304695,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.0002,
+      "loss": 0.9396,
+      "step": 1339
+    },
+    {
+      "epoch": 1.0703296703296703,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9306,
+      "step": 1340
+    },
+    {
+      "epoch": 1.0711288711288711,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.0002,
+      "loss": 0.929,
+      "step": 1341
+    },
+    {
+      "epoch": 1.071928071928072,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9429,
+      "step": 1342
+    },
+    {
+      "epoch": 1.0727272727272728,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0002,
+      "loss": 0.9499,
+      "step": 1343
+    },
+    {
+      "epoch": 1.0735264735264736,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0002,
+      "loss": 0.9336,
+      "step": 1344
+    },
+    {
+      "epoch": 1.0743256743256744,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0002,
+      "loss": 0.9191,
+      "step": 1345
+    },
+    {
+      "epoch": 1.0751248751248752,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0002,
+      "loss": 0.9243,
+      "step": 1346
+    },
+    {
+      "epoch": 1.075924075924076,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0002,
+      "loss": 0.9325,
+      "step": 1347
+    },
+    {
+      "epoch": 1.0767232767232766,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9362,
+      "step": 1348
+    },
+    {
+      "epoch": 1.0775224775224774,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0002,
+      "loss": 0.9354,
+      "step": 1349
+    },
+    {
+      "epoch": 1.0783216783216782,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9351,
+      "step": 1350
+    },
+    {
+      "epoch": 1.079120879120879,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.0002,
+      "loss": 0.9324,
+      "step": 1351
+    },
+    {
+      "epoch": 1.0799200799200799,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.0002,
+      "loss": 0.9229,
+      "step": 1352
+    },
+    {
+      "epoch": 1.0807192807192807,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.0002,
+      "loss": 0.9406,
+      "step": 1353
+    },
+    {
+      "epoch": 1.0815184815184815,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.0002,
+      "loss": 0.9278,
+      "step": 1354
+    },
+    {
+      "epoch": 1.0823176823176823,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9294,
+      "step": 1355
+    },
+    {
+      "epoch": 1.0831168831168831,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0002,
+      "loss": 0.9318,
+      "step": 1356
+    },
+    {
+      "epoch": 1.083916083916084,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0002,
+      "loss": 0.9327,
+      "step": 1357
+    },
+    {
+      "epoch": 1.0847152847152848,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.0002,
+      "loss": 0.9339,
+      "step": 1358
+    },
+    {
+      "epoch": 1.0855144855144856,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.0002,
+      "loss": 0.9343,
+      "step": 1359
+    },
+    {
+      "epoch": 1.0863136863136864,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9433,
+      "step": 1360
+    },
+    {
+      "epoch": 1.087112887112887,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.0002,
+      "loss": 0.9369,
+      "step": 1361
+    },
+    {
+      "epoch": 1.0879120879120878,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.0002,
+      "loss": 0.9335,
+      "step": 1362
+    },
+    {
+      "epoch": 1.0887112887112886,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0002,
+      "loss": 0.9289,
+      "step": 1363
+    },
+    {
+      "epoch": 1.0895104895104895,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0002,
+      "loss": 0.9274,
+      "step": 1364
+    },
+    {
+      "epoch": 1.0903096903096903,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0002,
+      "loss": 0.9346,
+      "step": 1365
+    },
+    {
+      "epoch": 1.091108891108891,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9329,
+      "step": 1366
+    },
+    {
+      "epoch": 1.091908091908092,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0002,
+      "loss": 0.9304,
+      "step": 1367
+    },
+    {
+      "epoch": 1.0927072927072927,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.0002,
+      "loss": 0.9315,
+      "step": 1368
+    },
+    {
+      "epoch": 1.0935064935064935,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0002,
+      "loss": 0.9376,
+      "step": 1369
+    },
+    {
+      "epoch": 1.0943056943056944,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0002,
+      "loss": 0.9363,
+      "step": 1370
+    },
+    {
+      "epoch": 1.0951048951048952,
+      "grad_norm": 0.75,
+      "learning_rate": 0.0002,
+      "loss": 0.9328,
+      "step": 1371
+    },
+    {
+      "epoch": 1.095904095904096,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.0002,
+      "loss": 0.9425,
+      "step": 1372
+    },
+    {
+      "epoch": 1.0967032967032968,
+      "grad_norm": 1.4609375,
+      "learning_rate": 0.0002,
+      "loss": 0.9346,
+      "step": 1373
+    },
+    {
+      "epoch": 1.0975024975024974,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0002,
+      "loss": 0.9282,
+      "step": 1374
+    },
+    {
+      "epoch": 1.0983016983016982,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0002,
+      "loss": 0.9444,
+      "step": 1375
+    },
+    {
+      "epoch": 1.099100899100899,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.0002,
+      "loss": 0.932,
+      "step": 1376
+    },
+    {
+      "epoch": 1.0999000999000998,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.0002,
+      "loss": 0.9439,
+      "step": 1377
+    },
+    {
+      "epoch": 1.1006993006993007,
+      "grad_norm": 1.6953125,
+      "learning_rate": 0.0002,
+      "loss": 0.9412,
+      "step": 1378
+    },
+    {
+      "epoch": 1.1014985014985015,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.0002,
+      "loss": 0.9368,
+      "step": 1379
+    },
+    {
+      "epoch": 1.1022977022977023,
+      "grad_norm": 0.89453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9445,
+      "step": 1380
+    },
+    {
+      "epoch": 1.1030969030969031,
+      "grad_norm": 1.5703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9492,
+      "step": 1381
+    },
+    {
+      "epoch": 1.103896103896104,
+      "grad_norm": 0.8125,
+      "learning_rate": 0.0002,
+      "loss": 0.943,
+      "step": 1382
+    },
+    {
+      "epoch": 1.1046953046953047,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9568,
+      "step": 1383
+    },
+    {
+      "epoch": 1.1054945054945056,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9467,
+      "step": 1384
+    },
+    {
+      "epoch": 1.1062937062937064,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.0002,
+      "loss": 0.971,
+      "step": 1385
+    },
+    {
+      "epoch": 1.1070929070929072,
+      "grad_norm": 1.828125,
+      "learning_rate": 0.0002,
+      "loss": 0.9598,
+      "step": 1386
+    },
+    {
+      "epoch": 1.1078921078921078,
+      "grad_norm": 1.734375,
+      "learning_rate": 0.0002,
+      "loss": 0.9659,
+      "step": 1387
+    },
+    {
+      "epoch": 1.1086913086913086,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.0002,
+      "loss": 0.9437,
+      "step": 1388
+    },
+    {
+      "epoch": 1.1094905094905094,
+      "grad_norm": 1.703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9568,
+      "step": 1389
+    },
+    {
+      "epoch": 1.1102897102897102,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.0002,
+      "loss": 0.9534,
+      "step": 1390
+    },
+    {
+      "epoch": 1.111088911088911,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.0002,
+      "loss": 0.9514,
+      "step": 1391
+    },
+    {
+      "epoch": 1.1118881118881119,
+      "grad_norm": 1.0,
+      "learning_rate": 0.0002,
+      "loss": 0.9509,
+      "step": 1392
+    },
+    {
+      "epoch": 1.1126873126873127,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9426,
+      "step": 1393
+    },
+    {
+      "epoch": 1.1134865134865135,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.0002,
+      "loss": 0.9469,
+      "step": 1394
+    },
+    {
+      "epoch": 1.1142857142857143,
+      "grad_norm": 0.8125,
+      "learning_rate": 0.0002,
+      "loss": 0.9535,
+      "step": 1395
+    },
+    {
+      "epoch": 1.1150849150849151,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.0002,
+      "loss": 0.9415,
+      "step": 1396
+    },
+    {
+      "epoch": 1.115884115884116,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9453,
+      "step": 1397
+    },
+    {
+      "epoch": 1.1166833166833168,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.0002,
+      "loss": 0.9409,
+      "step": 1398
+    },
+    {
+      "epoch": 1.1174825174825176,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.0002,
+      "loss": 0.9422,
+      "step": 1399
+    },
+    {
+      "epoch": 1.1182817182817182,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.0002,
+      "loss": 0.935,
+      "step": 1400
+    },
+    {
+      "epoch": 1.119080919080919,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.0002,
+      "loss": 0.9464,
+      "step": 1401
+    },
+    {
+      "epoch": 1.1198801198801198,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9373,
+      "step": 1402
+    },
+    {
+      "epoch": 1.1206793206793206,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0002,
+      "loss": 0.9476,
+      "step": 1403
+    },
+    {
+      "epoch": 1.1214785214785215,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9477,
+      "step": 1404
+    },
+    {
+      "epoch": 1.1222777222777223,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9312,
+      "step": 1405
+    },
+    {
+      "epoch": 1.123076923076923,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9305,
+      "step": 1406
+    },
+    {
+      "epoch": 1.123876123876124,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0002,
+      "loss": 0.9405,
+      "step": 1407
+    },
+    {
+      "epoch": 1.1246753246753247,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9276,
+      "step": 1408
+    },
+    {
+      "epoch": 1.1254745254745255,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.0002,
+      "loss": 0.9332,
+      "step": 1409
+    },
+    {
+      "epoch": 1.1262737262737263,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.0002,
+      "loss": 0.9283,
+      "step": 1410
+    },
+    {
+      "epoch": 1.1270729270729272,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.0002,
+      "loss": 0.9315,
+      "step": 1411
+    },
+    {
+      "epoch": 1.127872127872128,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.0002,
+      "loss": 0.9263,
+      "step": 1412
+    },
+    {
+      "epoch": 1.1286713286713286,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0002,
+      "loss": 0.9246,
+      "step": 1413
+    },
+    {
+      "epoch": 1.1294705294705294,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9373,
+      "step": 1414
+    },
+    {
+      "epoch": 1.1302697302697302,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0002,
+      "loss": 0.9307,
+      "step": 1415
+    },
+    {
+      "epoch": 1.131068931068931,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0002,
+      "loss": 0.9426,
+      "step": 1416
+    },
+    {
+      "epoch": 1.1318681318681318,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0002,
+      "loss": 0.9335,
+      "step": 1417
+    },
+    {
+      "epoch": 1.1326673326673327,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0002,
+      "loss": 0.929,
+      "step": 1418
+    },
+    {
+      "epoch": 1.1334665334665335,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0002,
+      "loss": 0.9286,
+      "step": 1419
+    },
+    {
+      "epoch": 1.1342657342657343,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0002,
+      "loss": 0.9314,
+      "step": 1420
+    },
+    {
+      "epoch": 1.135064935064935,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9299,
+      "step": 1421
+    },
+    {
+      "epoch": 1.135864135864136,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9255,
+      "step": 1422
+    },
+    {
+      "epoch": 1.1366633366633367,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0002,
+      "loss": 0.9374,
+      "step": 1423
+    },
+    {
+      "epoch": 1.1374625374625376,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0002,
+      "loss": 0.9337,
+      "step": 1424
+    },
+    {
+      "epoch": 1.1382617382617384,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0002,
+      "loss": 0.9261,
+      "step": 1425
+    },
+    {
+      "epoch": 1.139060939060939,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9305,
+      "step": 1426
+    },
+    {
+      "epoch": 1.1398601398601398,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0002,
+      "loss": 0.923,
+      "step": 1427
+    },
+    {
+      "epoch": 1.1406593406593406,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0002,
+      "loss": 0.928,
+      "step": 1428
+    },
+    {
+      "epoch": 1.1414585414585414,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0002,
+      "loss": 0.9353,
+      "step": 1429
+    },
+    {
+      "epoch": 1.1422577422577422,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0002,
+      "loss": 0.9333,
+      "step": 1430
+    },
+    {
+      "epoch": 1.143056943056943,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0002,
+      "loss": 0.9382,
+      "step": 1431
+    },
+    {
+      "epoch": 1.1438561438561439,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0002,
+      "loss": 0.9257,
+      "step": 1432
+    },
+    {
+      "epoch": 1.1446553446553447,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9349,
+      "step": 1433
+    },
+    {
+      "epoch": 1.1454545454545455,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9372,
+      "step": 1434
+    },
+    {
+      "epoch": 1.1462537462537463,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0002,
+      "loss": 0.924,
+      "step": 1435
+    },
+    {
+      "epoch": 1.1470529470529471,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0002,
+      "loss": 0.9228,
+      "step": 1436
+    },
+    {
+      "epoch": 1.1478521478521477,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.0002,
+      "loss": 0.9315,
+      "step": 1437
+    },
+    {
+      "epoch": 1.1486513486513488,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0002,
+      "loss": 0.937,
+      "step": 1438
+    },
+    {
+      "epoch": 1.1494505494505494,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9237,
+      "step": 1439
+    },
+    {
+      "epoch": 1.1502497502497502,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0002,
+      "loss": 0.9387,
+      "step": 1440
+    },
+    {
+      "epoch": 1.151048951048951,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0002,
+      "loss": 0.9177,
+      "step": 1441
+    },
+    {
+      "epoch": 1.1518481518481518,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0002,
+      "loss": 0.9349,
+      "step": 1442
+    },
+    {
+      "epoch": 1.1526473526473526,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0002,
+      "loss": 0.9463,
+      "step": 1443
+    },
+    {
+      "epoch": 1.1534465534465534,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9357,
+      "step": 1444
+    },
+    {
+      "epoch": 1.1542457542457543,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9301,
+      "step": 1445
+    },
+    {
+      "epoch": 1.155044955044955,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.0002,
+      "loss": 0.9399,
+      "step": 1446
+    },
+    {
+      "epoch": 1.155844155844156,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.0002,
+      "loss": 0.9418,
+      "step": 1447
+    },
+    {
+      "epoch": 1.1566433566433567,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0002,
+      "loss": 0.9425,
+      "step": 1448
+    },
+    {
+      "epoch": 1.1574425574425575,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0002,
+      "loss": 0.9285,
+      "step": 1449
+    },
+    {
+      "epoch": 1.1582417582417581,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0002,
+      "loss": 0.9284,
+      "step": 1450
+    },
+    {
+      "epoch": 1.1590409590409592,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0002,
+      "loss": 0.9379,
+      "step": 1451
+    },
+    {
+      "epoch": 1.1598401598401598,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0002,
+      "loss": 0.929,
+      "step": 1452
+    },
+    {
+      "epoch": 1.1606393606393606,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0002,
+      "loss": 0.9315,
+      "step": 1453
+    },
+    {
+      "epoch": 1.1614385614385614,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0002,
+      "loss": 0.9228,
+      "step": 1454
+    },
+    {
+      "epoch": 1.1622377622377622,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0002,
+      "loss": 0.9363,
+      "step": 1455
+    },
+    {
+      "epoch": 1.163036963036963,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.0002,
+      "loss": 0.9247,
+      "step": 1456
+    },
+    {
+      "epoch": 1.1638361638361638,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9242,
+      "step": 1457
+    },
+    {
+      "epoch": 1.1646353646353647,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0002,
+      "loss": 0.9292,
+      "step": 1458
+    },
+    {
+      "epoch": 1.1654345654345655,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0002,
+      "loss": 0.9325,
+      "step": 1459
+    },
+    {
+      "epoch": 1.1662337662337663,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9402,
+      "step": 1460
+    },
+    {
+      "epoch": 1.167032967032967,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0002,
+      "loss": 0.9421,
+      "step": 1461
+    },
+    {
+      "epoch": 1.167832167832168,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9228,
+      "step": 1462
+    },
+    {
+      "epoch": 1.1686313686313685,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0002,
+      "loss": 0.9273,
+      "step": 1463
+    },
+    {
+      "epoch": 1.1694305694305696,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0002,
+      "loss": 0.9319,
+      "step": 1464
+    },
+    {
+      "epoch": 1.1702297702297701,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9371,
+      "step": 1465
+    },
+    {
+      "epoch": 1.171028971028971,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0002,
+      "loss": 0.9285,
+      "step": 1466
+    },
+    {
+      "epoch": 1.1718281718281718,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0002,
+      "loss": 0.9268,
+      "step": 1467
+    },
+    {
+      "epoch": 1.1726273726273726,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0002,
+      "loss": 0.9269,
+      "step": 1468
+    },
+    {
+      "epoch": 1.1734265734265734,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0002,
+      "loss": 0.93,
+      "step": 1469
+    },
+    {
+      "epoch": 1.1742257742257742,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.0002,
+      "loss": 0.9493,
+      "step": 1470
+    },
+    {
+      "epoch": 1.175024975024975,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0002,
+      "loss": 0.925,
+      "step": 1471
+    },
+    {
+      "epoch": 1.1758241758241759,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0002,
+      "loss": 0.9332,
+      "step": 1472
+    },
+    {
+      "epoch": 1.1766233766233767,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0002,
+      "loss": 0.9347,
+      "step": 1473
+    },
+    {
+      "epoch": 1.1774225774225775,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9228,
+      "step": 1474
+    },
+    {
+      "epoch": 1.1782217782217783,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9329,
+      "step": 1475
+    },
+    {
+      "epoch": 1.179020979020979,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0002,
+      "loss": 0.9335,
+      "step": 1476
+    },
+    {
+      "epoch": 1.1798201798201797,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0002,
+      "loss": 0.9274,
+      "step": 1477
+    },
+    {
+      "epoch": 1.1806193806193805,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0002,
+      "loss": 0.9248,
+      "step": 1478
+    },
+    {
+      "epoch": 1.1814185814185814,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9293,
+      "step": 1479
+    },
+    {
+      "epoch": 1.1822177822177822,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9324,
+      "step": 1480
+    },
+    {
+      "epoch": 1.183016983016983,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9221,
+      "step": 1481
+    },
+    {
+      "epoch": 1.1838161838161838,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0002,
+      "loss": 0.9188,
+      "step": 1482
+    },
+    {
+      "epoch": 1.1846153846153846,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.0002,
+      "loss": 0.9314,
+      "step": 1483
+    },
+    {
+      "epoch": 1.1854145854145854,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.0002,
+      "loss": 0.9312,
+      "step": 1484
+    },
+    {
+      "epoch": 1.1862137862137863,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.0002,
+      "loss": 0.9301,
+      "step": 1485
+    },
+    {
+      "epoch": 1.187012987012987,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9215,
+      "step": 1486
+    },
+    {
+      "epoch": 1.187812187812188,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.0002,
+      "loss": 0.9228,
+      "step": 1487
+    },
+    {
+      "epoch": 1.1886113886113887,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9322,
+      "step": 1488
+    },
+    {
+      "epoch": 1.1894105894105893,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0002,
+      "loss": 0.9336,
+      "step": 1489
+    },
+    {
+      "epoch": 1.1902097902097901,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0002,
+      "loss": 0.9269,
+      "step": 1490
+    },
+    {
+      "epoch": 1.191008991008991,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0002,
+      "loss": 0.933,
+      "step": 1491
+    },
+    {
+      "epoch": 1.1918081918081918,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.0002,
+      "loss": 0.928,
+      "step": 1492
+    },
+    {
+      "epoch": 1.1926073926073926,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9324,
+      "step": 1493
+    },
+    {
+      "epoch": 1.1934065934065934,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.0002,
+      "loss": 0.9287,
+      "step": 1494
+    },
+    {
+      "epoch": 1.1942057942057942,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0002,
+      "loss": 0.9317,
+      "step": 1495
+    },
+    {
+      "epoch": 1.195004995004995,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9205,
+      "step": 1496
+    },
+    {
+      "epoch": 1.1958041958041958,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.0002,
+      "loss": 0.9344,
+      "step": 1497
+    },
+    {
+      "epoch": 1.1966033966033967,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0002,
+      "loss": 0.9241,
+      "step": 1498
+    },
+    {
+      "epoch": 1.1974025974025975,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0002,
+      "loss": 0.9211,
+      "step": 1499
+    },
+    {
+      "epoch": 1.1982017982017983,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.0002,
+      "loss": 0.936,
+      "step": 1500
+    },
+    {
+      "epoch": 1.199000999000999,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.0002,
+      "loss": 0.932,
+      "step": 1501
+    },
+    {
+      "epoch": 1.1998001998001997,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9256,
+      "step": 1502
+    },
+    {
+      "epoch": 1.2005994005994005,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0002,
+      "loss": 0.9361,
+      "step": 1503
+    },
+    {
+      "epoch": 1.2013986013986013,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0002,
+      "loss": 0.9209,
+      "step": 1504
+    },
+    {
+      "epoch": 1.2021978021978021,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0002,
+      "loss": 0.9202,
+      "step": 1505
+    },
+    {
+      "epoch": 1.202997002997003,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0002,
+      "loss": 0.931,
+      "step": 1506
+    },
+    {
+      "epoch": 1.2037962037962038,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0002,
+      "loss": 0.9325,
+      "step": 1507
+    },
+    {
+      "epoch": 1.2045954045954046,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0002,
+      "loss": 0.9267,
+      "step": 1508
+    },
+    {
+      "epoch": 1.2053946053946054,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0002,
+      "loss": 0.9287,
+      "step": 1509
+    },
+    {
+      "epoch": 1.2061938061938062,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.0002,
+      "loss": 0.9264,
+      "step": 1510
+    },
+    {
+      "epoch": 1.206993006993007,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9231,
+      "step": 1511
+    },
+    {
+      "epoch": 1.2077922077922079,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9228,
+      "step": 1512
+    },
+    {
+      "epoch": 1.2085914085914087,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.0002,
+      "loss": 0.9444,
+      "step": 1513
+    },
+    {
+      "epoch": 1.2093906093906095,
+      "grad_norm": 0.375,
+      "learning_rate": 0.0002,
+      "loss": 0.9187,
+      "step": 1514
+    },
+    {
+      "epoch": 1.21018981018981,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0002,
+      "loss": 0.934,
+      "step": 1515
+    },
+    {
+      "epoch": 1.210989010989011,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0002,
+      "loss": 0.9253,
+      "step": 1516
+    },
+    {
+      "epoch": 1.2117882117882117,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0002,
+      "loss": 0.9292,
+      "step": 1517
+    },
+    {
+      "epoch": 1.2125874125874125,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9241,
+      "step": 1518
+    },
+    {
+      "epoch": 1.2133866133866134,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9272,
+      "step": 1519
+    },
+    {
+      "epoch": 1.2141858141858142,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.0002,
+      "loss": 0.9193,
+      "step": 1520
+    },
+    {
+      "epoch": 1.214985014985015,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0002,
+      "loss": 0.9303,
+      "step": 1521
+    },
+    {
+      "epoch": 1.2157842157842158,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0002,
+      "loss": 0.9325,
+      "step": 1522
+    },
+    {
+      "epoch": 1.2165834165834166,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.0002,
+      "loss": 0.9206,
+      "step": 1523
+    },
+    {
+      "epoch": 1.2173826173826174,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0002,
+      "loss": 0.9235,
+      "step": 1524
+    },
+    {
+      "epoch": 1.2181818181818183,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9229,
+      "step": 1525
+    },
+    {
+      "epoch": 1.218981018981019,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9315,
+      "step": 1526
+    },
+    {
+      "epoch": 1.2197802197802199,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9313,
+      "step": 1527
+    },
+    {
+      "epoch": 1.2205794205794205,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0002,
+      "loss": 0.9285,
+      "step": 1528
+    },
+    {
+      "epoch": 1.2213786213786213,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0002,
+      "loss": 0.9376,
+      "step": 1529
+    },
+    {
+      "epoch": 1.2221778221778221,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0002,
+      "loss": 0.9241,
+      "step": 1530
+    },
+    {
+      "epoch": 1.222977022977023,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.0002,
+      "loss": 0.9164,
+      "step": 1531
+    },
+    {
+      "epoch": 1.2237762237762237,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0002,
+      "loss": 0.9304,
+      "step": 1532
+    },
+    {
+      "epoch": 1.2245754245754246,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9306,
+      "step": 1533
+    },
+    {
+      "epoch": 1.2253746253746254,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9263,
+      "step": 1534
+    },
+    {
+      "epoch": 1.2261738261738262,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.0002,
+      "loss": 0.9212,
+      "step": 1535
+    },
+    {
+      "epoch": 1.226973026973027,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.0002,
+      "loss": 0.942,
+      "step": 1536
+    },
+    {
+      "epoch": 1.2277722277722278,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0002,
+      "loss": 0.9325,
+      "step": 1537
+    },
+    {
+      "epoch": 1.2285714285714286,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0002,
+      "loss": 0.9222,
+      "step": 1538
+    },
+    {
+      "epoch": 1.2293706293706295,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0002,
+      "loss": 0.9396,
+      "step": 1539
+    },
+    {
+      "epoch": 1.2301698301698303,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.0002,
+      "loss": 0.9319,
+      "step": 1540
+    },
+    {
+      "epoch": 1.2309690309690309,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0002,
+      "loss": 0.9122,
+      "step": 1541
+    },
+    {
+      "epoch": 1.2317682317682317,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.0002,
+      "loss": 0.9232,
+      "step": 1542
+    },
+    {
+      "epoch": 1.2325674325674325,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.0002,
+      "loss": 0.929,
+      "step": 1543
+    },
+    {
+      "epoch": 1.2333666333666333,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9239,
+      "step": 1544
+    },
+    {
+      "epoch": 1.2341658341658341,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0002,
+      "loss": 0.9208,
+      "step": 1545
+    },
+    {
+      "epoch": 1.234965034965035,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9317,
+      "step": 1546
+    },
+    {
+      "epoch": 1.2357642357642358,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9209,
+      "step": 1547
+    },
+    {
+      "epoch": 1.2365634365634366,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9321,
+      "step": 1548
+    },
+    {
+      "epoch": 1.2373626373626374,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0002,
+      "loss": 0.9222,
+      "step": 1549
+    },
+    {
+      "epoch": 1.2381618381618382,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.0002,
+      "loss": 0.9267,
+      "step": 1550
+    },
+    {
+      "epoch": 1.238961038961039,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9408,
+      "step": 1551
+    },
+    {
+      "epoch": 1.2397602397602399,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0002,
+      "loss": 0.9242,
+      "step": 1552
+    },
+    {
+      "epoch": 1.2405594405594407,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.0002,
+      "loss": 0.9335,
+      "step": 1553
+    },
+    {
+      "epoch": 1.2413586413586413,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9319,
+      "step": 1554
+    },
+    {
+      "epoch": 1.242157842157842,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0002,
+      "loss": 0.925,
+      "step": 1555
+    },
+    {
+      "epoch": 1.242957042957043,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.0002,
+      "loss": 0.9257,
+      "step": 1556
+    },
+    {
+      "epoch": 1.2437562437562437,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9287,
+      "step": 1557
+    },
+    {
+      "epoch": 1.2445554445554445,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.0002,
+      "loss": 0.9273,
+      "step": 1558
+    },
+    {
+      "epoch": 1.2453546453546454,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0002,
+      "loss": 0.9313,
+      "step": 1559
+    },
+    {
+      "epoch": 1.2461538461538462,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9304,
+      "step": 1560
+    },
+    {
+      "epoch": 1.246953046953047,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0002,
+      "loss": 0.9261,
+      "step": 1561
+    },
+    {
+      "epoch": 1.2477522477522478,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0002,
+      "loss": 0.9206,
+      "step": 1562
+    },
+    {
+      "epoch": 1.2485514485514486,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0002,
+      "loss": 0.9277,
+      "step": 1563
+    },
+    {
+      "epoch": 1.2493506493506494,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0002,
+      "loss": 0.9321,
+      "step": 1564
+    },
+    {
+      "epoch": 1.25014985014985,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9266,
+      "step": 1565
+    },
+    {
+      "epoch": 1.250949050949051,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0002,
+      "loss": 0.9342,
+      "step": 1566
+    },
+    {
+      "epoch": 1.2517482517482517,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0002,
+      "loss": 0.9341,
+      "step": 1567
+    },
+    {
+      "epoch": 1.2525474525474525,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.0002,
+      "loss": 0.9124,
+      "step": 1568
+    },
+    {
+      "epoch": 1.2533466533466533,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0002,
+      "loss": 0.9268,
+      "step": 1569
+    },
+    {
+      "epoch": 1.254145854145854,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0002,
+      "loss": 0.924,
+      "step": 1570
+    },
+    {
+      "epoch": 1.254945054945055,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0002,
+      "loss": 0.9302,
+      "step": 1571
+    },
+    {
+      "epoch": 1.2557442557442557,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0002,
+      "loss": 0.929,
+      "step": 1572
+    },
+    {
+      "epoch": 1.2565434565434566,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0002,
+      "loss": 0.9282,
+      "step": 1573
+    },
+    {
+      "epoch": 1.2573426573426574,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9323,
+      "step": 1574
+    },
+    {
+      "epoch": 1.2581418581418582,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9326,
+      "step": 1575
+    },
+    {
+      "epoch": 1.258941058941059,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9202,
+      "step": 1576
+    },
+    {
+      "epoch": 1.2597402597402598,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9274,
+      "step": 1577
+    },
+    {
+      "epoch": 1.2605394605394604,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0002,
+      "loss": 0.9222,
+      "step": 1578
+    },
+    {
+      "epoch": 1.2613386613386615,
+      "grad_norm": 0.375,
+      "learning_rate": 0.0002,
+      "loss": 0.9265,
+      "step": 1579
+    },
+    {
+      "epoch": 1.262137862137862,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0002,
+      "loss": 0.9264,
+      "step": 1580
+    },
+    {
+      "epoch": 1.2629370629370629,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0002,
+      "loss": 0.9276,
+      "step": 1581
+    },
+    {
+      "epoch": 1.2637362637362637,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.0002,
+      "loss": 0.9321,
+      "step": 1582
+    },
+    {
+      "epoch": 1.2645354645354645,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.0002,
+      "loss": 0.9237,
+      "step": 1583
+    },
+    {
+      "epoch": 1.2653346653346653,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.0002,
+      "loss": 0.9285,
+      "step": 1584
+    },
+    {
+      "epoch": 1.2661338661338661,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.0002,
+      "loss": 0.9293,
+      "step": 1585
+    },
+    {
+      "epoch": 1.266933066933067,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9332,
+      "step": 1586
+    },
+    {
+      "epoch": 1.2677322677322678,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9345,
+      "step": 1587
+    },
+    {
+      "epoch": 1.2685314685314686,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9267,
+      "step": 1588
+    },
+    {
+      "epoch": 1.2693306693306694,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0002,
+      "loss": 0.9298,
+      "step": 1589
+    },
+    {
+      "epoch": 1.2701298701298702,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0002,
+      "loss": 0.9333,
+      "step": 1590
+    },
+    {
+      "epoch": 1.2709290709290708,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0002,
+      "loss": 0.9271,
+      "step": 1591
+    },
+    {
+      "epoch": 1.2717282717282719,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0002,
+      "loss": 0.9279,
+      "step": 1592
+    },
+    {
+      "epoch": 1.2725274725274724,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0002,
+      "loss": 0.9368,
+      "step": 1593
+    },
+    {
+      "epoch": 1.2733266733266733,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9282,
+      "step": 1594
+    },
+    {
+      "epoch": 1.274125874125874,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.0002,
+      "loss": 0.9386,
+      "step": 1595
+    },
+    {
+      "epoch": 1.274925074925075,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.0002,
+      "loss": 0.929,
+      "step": 1596
+    },
+    {
+      "epoch": 1.2757242757242757,
+      "grad_norm": 0.80078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9206,
+      "step": 1597
+    },
+    {
+      "epoch": 1.2765234765234765,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0002,
+      "loss": 0.9226,
+      "step": 1598
+    },
+    {
+      "epoch": 1.2773226773226773,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0002,
+      "loss": 0.9343,
+      "step": 1599
+    },
+    {
+      "epoch": 1.2781218781218782,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0002,
+      "loss": 0.9262,
+      "step": 1600
+    },
+    {
+      "epoch": 1.278921078921079,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0002,
+      "loss": 0.9256,
+      "step": 1601
+    },
+    {
+      "epoch": 1.2797202797202798,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.0002,
+      "loss": 0.9228,
+      "step": 1602
+    },
+    {
+      "epoch": 1.2805194805194806,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0002,
+      "loss": 0.936,
+      "step": 1603
+    },
+    {
+      "epoch": 1.2813186813186812,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0002,
+      "loss": 0.9354,
+      "step": 1604
+    },
+    {
+      "epoch": 1.2821178821178822,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0002,
+      "loss": 0.9355,
+      "step": 1605
+    },
+    {
+      "epoch": 1.2829170829170828,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.0002,
+      "loss": 0.922,
+      "step": 1606
+    },
+    {
+      "epoch": 1.2837162837162837,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0002,
+      "loss": 0.9245,
+      "step": 1607
+    },
+    {
+      "epoch": 1.2845154845154845,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9319,
+      "step": 1608
+    },
+    {
+      "epoch": 1.2853146853146853,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0002,
+      "loss": 0.9237,
+      "step": 1609
+    },
+    {
+      "epoch": 1.286113886113886,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0002,
+      "loss": 0.9212,
+      "step": 1610
+    },
+    {
+      "epoch": 1.286913086913087,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.0002,
+      "loss": 0.9242,
+      "step": 1611
+    },
+    {
+      "epoch": 1.2877122877122877,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0002,
+      "loss": 0.9184,
+      "step": 1612
+    },
+    {
+      "epoch": 1.2885114885114886,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.0002,
+      "loss": 0.9297,
+      "step": 1613
+    },
+    {
+      "epoch": 1.2893106893106894,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9242,
+      "step": 1614
+    },
+    {
+      "epoch": 1.2901098901098902,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0002,
+      "loss": 0.9254,
+      "step": 1615
+    },
+    {
+      "epoch": 1.290909090909091,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0002,
+      "loss": 0.9175,
+      "step": 1616
+    },
+    {
+      "epoch": 1.2917082917082916,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9292,
+      "step": 1617
+    },
+    {
+      "epoch": 1.2925074925074926,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.0002,
+      "loss": 0.9254,
+      "step": 1618
+    },
+    {
+      "epoch": 1.2933066933066932,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0002,
+      "loss": 0.9724,
+      "step": 1619
+    },
+    {
+      "epoch": 1.294105894105894,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0002,
+      "loss": 0.9317,
+      "step": 1620
+    },
+    {
+      "epoch": 1.2949050949050949,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0002,
+      "loss": 0.9367,
+      "step": 1621
+    },
+    {
+      "epoch": 1.2957042957042957,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9181,
+      "step": 1622
+    },
+    {
+      "epoch": 1.2965034965034965,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9243,
+      "step": 1623
+    },
+    {
+      "epoch": 1.2973026973026973,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0002,
+      "loss": 0.9253,
+      "step": 1624
+    },
+    {
+      "epoch": 1.2981018981018981,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0002,
+      "loss": 0.9198,
+      "step": 1625
+    },
+    {
+      "epoch": 1.298901098901099,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9219,
+      "step": 1626
+    },
+    {
+      "epoch": 1.2997002997002998,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0002,
+      "loss": 0.9237,
+      "step": 1627
+    },
+    {
+      "epoch": 1.3004995004995006,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.0002,
+      "loss": 0.9169,
+      "step": 1628
+    },
+    {
+      "epoch": 1.3012987012987014,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0002,
+      "loss": 0.9269,
+      "step": 1629
+    },
+    {
+      "epoch": 1.302097902097902,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9216,
+      "step": 1630
+    },
+    {
+      "epoch": 1.302897102897103,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9266,
+      "step": 1631
+    },
+    {
+      "epoch": 1.3036963036963036,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0002,
+      "loss": 0.9306,
+      "step": 1632
+    },
+    {
+      "epoch": 1.3044955044955044,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0002,
+      "loss": 0.9239,
+      "step": 1633
+    },
+    {
+      "epoch": 1.3052947052947053,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0002,
+      "loss": 0.9313,
+      "step": 1634
+    },
+    {
+      "epoch": 1.306093906093906,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9267,
+      "step": 1635
+    },
+    {
+      "epoch": 1.306893106893107,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0002,
+      "loss": 0.92,
+      "step": 1636
+    },
+    {
+      "epoch": 1.3076923076923077,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9299,
+      "step": 1637
+    },
+    {
+      "epoch": 1.3084915084915085,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9255,
+      "step": 1638
+    },
+    {
+      "epoch": 1.3092907092907093,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9259,
+      "step": 1639
+    },
+    {
+      "epoch": 1.3100899100899102,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0002,
+      "loss": 0.9192,
+      "step": 1640
+    },
+    {
+      "epoch": 1.3108891108891108,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0002,
+      "loss": 0.9328,
+      "step": 1641
+    },
+    {
+      "epoch": 1.3116883116883118,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0002,
+      "loss": 0.924,
+      "step": 1642
+    },
+    {
+      "epoch": 1.3124875124875124,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9203,
+      "step": 1643
+    },
+    {
+      "epoch": 1.3132867132867134,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9187,
+      "step": 1644
+    },
+    {
+      "epoch": 1.314085914085914,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0002,
+      "loss": 0.9188,
+      "step": 1645
+    },
+    {
+      "epoch": 1.3148851148851148,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0002,
+      "loss": 0.9228,
+      "step": 1646
+    },
+    {
+      "epoch": 1.3156843156843157,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9285,
+      "step": 1647
+    },
+    {
+      "epoch": 1.3164835164835165,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0002,
+      "loss": 0.9266,
+      "step": 1648
+    },
+    {
+      "epoch": 1.3172827172827173,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.0002,
+      "loss": 0.9343,
+      "step": 1649
+    },
+    {
+      "epoch": 1.318081918081918,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0002,
+      "loss": 0.9264,
+      "step": 1650
+    },
+    {
+      "epoch": 1.318881118881119,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0002,
+      "loss": 0.9331,
+      "step": 1651
+    },
+    {
+      "epoch": 1.3196803196803197,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0002,
+      "loss": 0.9195,
+      "step": 1652
+    },
+    {
+      "epoch": 1.3204795204795206,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.0002,
+      "loss": 0.9265,
+      "step": 1653
+    },
+    {
+      "epoch": 1.3212787212787211,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9261,
+      "step": 1654
+    },
+    {
+      "epoch": 1.3220779220779222,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0002,
+      "loss": 0.9263,
+      "step": 1655
+    },
+    {
+      "epoch": 1.3228771228771228,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0002,
+      "loss": 0.9294,
+      "step": 1656
+    },
+    {
+      "epoch": 1.3236763236763236,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0002,
+      "loss": 0.9233,
+      "step": 1657
+    },
+    {
+      "epoch": 1.3244755244755244,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9224,
+      "step": 1658
+    },
+    {
+      "epoch": 1.3252747252747252,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9194,
+      "step": 1659
+    },
+    {
+      "epoch": 1.326073926073926,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0002,
+      "loss": 0.9215,
+      "step": 1660
+    },
+    {
+      "epoch": 1.3268731268731269,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0002,
+      "loss": 0.9208,
+      "step": 1661
+    },
+    {
+      "epoch": 1.3276723276723277,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0002,
+      "loss": 0.931,
+      "step": 1662
+    },
+    {
+      "epoch": 1.3284715284715285,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0002,
+      "loss": 0.923,
+      "step": 1663
+    },
+    {
+      "epoch": 1.3292707292707293,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9227,
+      "step": 1664
+    },
+    {
+      "epoch": 1.3300699300699301,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0002,
+      "loss": 0.9281,
+      "step": 1665
+    },
+    {
+      "epoch": 1.330869130869131,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9226,
+      "step": 1666
+    },
+    {
+      "epoch": 1.3316683316683315,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.0002,
+      "loss": 0.9262,
+      "step": 1667
+    },
+    {
+      "epoch": 1.3324675324675326,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.0002,
+      "loss": 0.9305,
+      "step": 1668
+    },
+    {
+      "epoch": 1.3332667332667332,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9284,
+      "step": 1669
+    },
+    {
+      "epoch": 1.334065934065934,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0002,
+      "loss": 0.9264,
+      "step": 1670
+    },
+    {
+      "epoch": 1.3348651348651348,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9275,
+      "step": 1671
+    },
+    {
+      "epoch": 1.3356643356643356,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0002,
+      "loss": 0.9347,
+      "step": 1672
+    },
+    {
+      "epoch": 1.3364635364635364,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0002,
+      "loss": 0.9211,
+      "step": 1673
+    },
+    {
+      "epoch": 1.3372627372627373,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9169,
+      "step": 1674
+    },
+    {
+      "epoch": 1.338061938061938,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9223,
+      "step": 1675
+    },
+    {
+      "epoch": 1.3388611388611389,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0002,
+      "loss": 0.9265,
+      "step": 1676
+    },
+    {
+      "epoch": 1.3396603396603397,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9197,
+      "step": 1677
+    },
+    {
+      "epoch": 1.3404595404595405,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9263,
+      "step": 1678
+    },
+    {
+      "epoch": 1.3412587412587413,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9178,
+      "step": 1679
+    },
+    {
+      "epoch": 1.342057942057942,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0002,
+      "loss": 0.9305,
+      "step": 1680
+    },
+    {
+      "epoch": 1.342857142857143,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0002,
+      "loss": 0.9141,
+      "step": 1681
+    },
+    {
+      "epoch": 1.3436563436563436,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0002,
+      "loss": 0.9244,
+      "step": 1682
+    },
+    {
+      "epoch": 1.3444555444555444,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9247,
+      "step": 1683
+    },
+    {
+      "epoch": 1.3452547452547452,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0002,
+      "loss": 0.9266,
+      "step": 1684
+    },
+    {
+      "epoch": 1.346053946053946,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0002,
+      "loss": 0.9315,
+      "step": 1685
+    },
+    {
+      "epoch": 1.3468531468531468,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0002,
+      "loss": 0.9232,
+      "step": 1686
+    },
+    {
+      "epoch": 1.3476523476523476,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0002,
+      "loss": 0.9296,
+      "step": 1687
+    },
+    {
+      "epoch": 1.3484515484515485,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0002,
+      "loss": 0.9356,
+      "step": 1688
+    },
+    {
+      "epoch": 1.3492507492507493,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9159,
+      "step": 1689
+    },
+    {
+      "epoch": 1.35004995004995,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0002,
+      "loss": 0.9257,
+      "step": 1690
+    },
+    {
+      "epoch": 1.350849150849151,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9244,
+      "step": 1691
+    },
+    {
+      "epoch": 1.3516483516483517,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0002,
+      "loss": 0.9273,
+      "step": 1692
+    },
+    {
+      "epoch": 1.3524475524475523,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0002,
+      "loss": 0.9156,
+      "step": 1693
+    },
+    {
+      "epoch": 1.3532467532467534,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0002,
+      "loss": 0.9237,
+      "step": 1694
+    },
+    {
+      "epoch": 1.354045954045954,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0002,
+      "loss": 0.9216,
+      "step": 1695
+    },
+    {
+      "epoch": 1.3548451548451548,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9229,
+      "step": 1696
+    },
+    {
+      "epoch": 1.3556443556443556,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9251,
+      "step": 1697
+    },
+    {
+      "epoch": 1.3564435564435564,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0002,
+      "loss": 0.9184,
+      "step": 1698
+    },
+    {
+      "epoch": 1.3572427572427572,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9259,
+      "step": 1699
+    },
+    {
+      "epoch": 1.358041958041958,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0002,
+      "loss": 0.9303,
+      "step": 1700
+    },
+    {
+      "epoch": 1.3588411588411589,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0002,
+      "loss": 0.9186,
+      "step": 1701
+    },
+    {
+      "epoch": 1.3596403596403597,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0002,
+      "loss": 0.9302,
+      "step": 1702
+    },
+    {
+      "epoch": 1.3604395604395605,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0002,
+      "loss": 0.9309,
+      "step": 1703
+    },
+    {
+      "epoch": 1.3612387612387613,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9328,
+      "step": 1704
+    },
+    {
+      "epoch": 1.3620379620379621,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0002,
+      "loss": 0.9226,
+      "step": 1705
+    },
+    {
+      "epoch": 1.3628371628371627,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9268,
+      "step": 1706
+    },
+    {
+      "epoch": 1.3636363636363638,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0002,
+      "loss": 0.9216,
+      "step": 1707
+    },
+    {
+      "epoch": 1.3644355644355644,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0002,
+      "loss": 0.9203,
+      "step": 1708
+    },
+    {
+      "epoch": 1.3652347652347652,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.0002,
+      "loss": 0.9292,
+      "step": 1709
+    },
+    {
+      "epoch": 1.366033966033966,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0002,
+      "loss": 0.9193,
+      "step": 1710
+    },
+    {
+      "epoch": 1.3668331668331668,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0002,
+      "loss": 0.9306,
+      "step": 1711
+    },
+    {
+      "epoch": 1.3676323676323676,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9298,
+      "step": 1712
+    },
+    {
+      "epoch": 1.3684315684315684,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0002,
+      "loss": 0.9246,
+      "step": 1713
+    },
+    {
+      "epoch": 1.3692307692307693,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0002,
+      "loss": 0.9145,
+      "step": 1714
+    },
+    {
+      "epoch": 1.37002997002997,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0002,
+      "loss": 0.9244,
+      "step": 1715
+    },
+    {
+      "epoch": 1.3708291708291709,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9226,
+      "step": 1716
+    },
+    {
+      "epoch": 1.3716283716283717,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0002,
+      "loss": 0.9249,
+      "step": 1717
+    },
+    {
+      "epoch": 1.3724275724275725,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9278,
+      "step": 1718
+    },
+    {
+      "epoch": 1.3732267732267731,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9107,
+      "step": 1719
+    },
+    {
+      "epoch": 1.3740259740259742,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0002,
+      "loss": 0.9182,
+      "step": 1720
+    },
+    {
+      "epoch": 1.3748251748251747,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9304,
+      "step": 1721
+    },
+    {
+      "epoch": 1.3756243756243756,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.0002,
+      "loss": 0.9255,
+      "step": 1722
+    },
+    {
+      "epoch": 1.3764235764235764,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9191,
+      "step": 1723
+    },
+    {
+      "epoch": 1.3772227772227772,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0002,
+      "loss": 0.9187,
+      "step": 1724
+    },
+    {
+      "epoch": 1.378021978021978,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9254,
+      "step": 1725
+    },
+    {
+      "epoch": 1.3788211788211788,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0002,
+      "loss": 0.9283,
+      "step": 1726
+    },
+    {
+      "epoch": 1.3796203796203796,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9352,
+      "step": 1727
+    },
+    {
+      "epoch": 1.3804195804195805,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0002,
+      "loss": 0.9309,
+      "step": 1728
+    },
+    {
+      "epoch": 1.3812187812187813,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0002,
+      "loss": 0.9187,
+      "step": 1729
+    },
+    {
+      "epoch": 1.382017982017982,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0002,
+      "loss": 0.924,
+      "step": 1730
+    },
+    {
+      "epoch": 1.382817182817183,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0002,
+      "loss": 0.9196,
+      "step": 1731
+    },
+    {
+      "epoch": 1.3836163836163835,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0002,
+      "loss": 0.9252,
+      "step": 1732
+    },
+    {
+      "epoch": 1.3844155844155845,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9367,
+      "step": 1733
+    },
+    {
+      "epoch": 1.3852147852147851,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9195,
+      "step": 1734
+    },
+    {
+      "epoch": 1.386013986013986,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0002,
+      "loss": 0.914,
+      "step": 1735
+    },
+    {
+      "epoch": 1.3868131868131868,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0002,
+      "loss": 0.9247,
+      "step": 1736
+    },
+    {
+      "epoch": 1.3876123876123876,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0002,
+      "loss": 0.9209,
+      "step": 1737
+    },
+    {
+      "epoch": 1.3884115884115884,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0002,
+      "loss": 0.9294,
+      "step": 1738
+    },
+    {
+      "epoch": 1.3892107892107892,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9253,
+      "step": 1739
+    },
+    {
+      "epoch": 1.39000999000999,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9262,
+      "step": 1740
+    },
+    {
+      "epoch": 1.3908091908091909,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9206,
+      "step": 1741
+    },
+    {
+      "epoch": 1.3916083916083917,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9234,
+      "step": 1742
+    },
+    {
+      "epoch": 1.3924075924075925,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.0002,
+      "loss": 0.9262,
+      "step": 1743
+    },
+    {
+      "epoch": 1.3932067932067933,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9226,
+      "step": 1744
+    },
+    {
+      "epoch": 1.394005994005994,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9744,
+      "step": 1745
+    },
+    {
+      "epoch": 1.394805194805195,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9275,
+      "step": 1746
+    },
+    {
+      "epoch": 1.3956043956043955,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0002,
+      "loss": 0.9224,
+      "step": 1747
+    },
+    {
+      "epoch": 1.3964035964035963,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9326,
+      "step": 1748
+    },
+    {
+      "epoch": 1.3972027972027972,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0002,
+      "loss": 0.9184,
+      "step": 1749
+    },
+    {
+      "epoch": 1.398001998001998,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0002,
+      "loss": 0.9233,
+      "step": 1750
+    },
+    {
+      "epoch": 1.3988011988011988,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9207,
+      "step": 1751
+    },
+    {
+      "epoch": 1.3996003996003996,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9208,
+      "step": 1752
+    },
+    {
+      "epoch": 1.4003996003996004,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9173,
+      "step": 1753
+    },
+    {
+      "epoch": 1.4011988011988012,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0002,
+      "loss": 0.9237,
+      "step": 1754
+    },
+    {
+      "epoch": 1.401998001998002,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9148,
+      "step": 1755
+    },
+    {
+      "epoch": 1.4027972027972029,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9189,
+      "step": 1756
+    },
+    {
+      "epoch": 1.4035964035964037,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0002,
+      "loss": 0.9215,
+      "step": 1757
+    },
+    {
+      "epoch": 1.4043956043956043,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9248,
+      "step": 1758
+    },
+    {
+      "epoch": 1.4051948051948053,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.0002,
+      "loss": 0.9229,
+      "step": 1759
+    },
+    {
+      "epoch": 1.405994005994006,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9262,
+      "step": 1760
+    },
+    {
+      "epoch": 1.4067932067932067,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9287,
+      "step": 1761
+    },
+    {
+      "epoch": 1.4075924075924076,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9216,
+      "step": 1762
+    },
+    {
+      "epoch": 1.4083916083916084,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.0002,
+      "loss": 0.916,
+      "step": 1763
+    },
+    {
+      "epoch": 1.4091908091908092,
+      "grad_norm": 0.828125,
+      "learning_rate": 0.0002,
+      "loss": 0.9273,
+      "step": 1764
+    },
+    {
+      "epoch": 1.40999000999001,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.0002,
+      "loss": 0.9288,
+      "step": 1765
+    },
+    {
+      "epoch": 1.4107892107892108,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.0002,
+      "loss": 0.9143,
+      "step": 1766
+    },
+    {
+      "epoch": 1.4115884115884116,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9218,
+      "step": 1767
+    },
+    {
+      "epoch": 1.4123876123876125,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.0002,
+      "loss": 0.9271,
+      "step": 1768
+    },
+    {
+      "epoch": 1.413186813186813,
+      "grad_norm": 3.40625,
+      "learning_rate": 0.0002,
+      "loss": 0.9477,
+      "step": 1769
+    },
+    {
+      "epoch": 1.413986013986014,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.0002,
+      "loss": 0.9314,
+      "step": 1770
+    },
+    {
+      "epoch": 1.4147852147852147,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.0002,
+      "loss": 0.9195,
+      "step": 1771
+    },
+    {
+      "epoch": 1.4155844155844157,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9232,
+      "step": 1772
+    },
+    {
+      "epoch": 1.4163836163836163,
+      "grad_norm": 3.65625,
+      "learning_rate": 0.0002,
+      "loss": 0.9367,
+      "step": 1773
+    },
+    {
+      "epoch": 1.4171828171828171,
+      "grad_norm": 3.109375,
+      "learning_rate": 0.0002,
+      "loss": 0.9286,
+      "step": 1774
+    },
+    {
+      "epoch": 1.417982017982018,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.0002,
+      "loss": 0.9226,
+      "step": 1775
+    },
+    {
+      "epoch": 1.4187812187812188,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9232,
+      "step": 1776
+    },
+    {
+      "epoch": 1.4195804195804196,
+      "grad_norm": 0.83984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9251,
+      "step": 1777
+    },
+    {
+      "epoch": 1.4203796203796204,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0002,
+      "loss": 0.9291,
+      "step": 1778
+    },
+    {
+      "epoch": 1.4211788211788212,
+      "grad_norm": 1.515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9371,
+      "step": 1779
+    },
+    {
+      "epoch": 1.421978021978022,
+      "grad_norm": 1.5,
+      "learning_rate": 0.0002,
+      "loss": 0.9387,
+      "step": 1780
+    },
+    {
+      "epoch": 1.4227772227772228,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9314,
+      "step": 1781
+    },
+    {
+      "epoch": 1.4235764235764234,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0002,
+      "loss": 0.9338,
+      "step": 1782
+    },
+    {
+      "epoch": 1.4243756243756245,
+      "grad_norm": 1.421875,
+      "learning_rate": 0.0002,
+      "loss": 0.9263,
+      "step": 1783
+    },
+    {
+      "epoch": 1.425174825174825,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.0002,
+      "loss": 0.9263,
+      "step": 1784
+    },
+    {
+      "epoch": 1.425974025974026,
+      "grad_norm": 1.7265625,
+      "learning_rate": 0.0002,
+      "loss": 0.9312,
+      "step": 1785
+    },
+    {
+      "epoch": 1.4267732267732267,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9337,
+      "step": 1786
+    },
+    {
+      "epoch": 1.4275724275724275,
+      "grad_norm": 0.8671875,
+      "learning_rate": 0.0002,
+      "loss": 0.9338,
+      "step": 1787
+    },
+    {
+      "epoch": 1.4283716283716283,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0002,
+      "loss": 0.9272,
+      "step": 1788
+    },
+    {
+      "epoch": 1.4291708291708292,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.0002,
+      "loss": 0.9194,
+      "step": 1789
+    },
+    {
+      "epoch": 1.42997002997003,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9239,
+      "step": 1790
+    },
+    {
+      "epoch": 1.4307692307692308,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0002,
+      "loss": 0.9361,
+      "step": 1791
+    },
+    {
+      "epoch": 1.4315684315684316,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0002,
+      "loss": 0.9194,
+      "step": 1792
+    },
+    {
+      "epoch": 1.4323676323676324,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0002,
+      "loss": 0.925,
+      "step": 1793
+    },
+    {
+      "epoch": 1.4331668331668332,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0002,
+      "loss": 0.935,
+      "step": 1794
+    },
+    {
+      "epoch": 1.4339660339660338,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9438,
+      "step": 1795
+    },
+    {
+      "epoch": 1.4347652347652349,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9326,
+      "step": 1796
+    },
+    {
+      "epoch": 1.4355644355644355,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0002,
+      "loss": 0.9221,
+      "step": 1797
+    },
+    {
+      "epoch": 1.4363636363636363,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.0002,
+      "loss": 0.9208,
+      "step": 1798
+    },
+    {
+      "epoch": 1.437162837162837,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.0002,
+      "loss": 0.9262,
+      "step": 1799
+    },
+    {
+      "epoch": 1.437962037962038,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0002,
+      "loss": 0.9325,
+      "step": 1800
+    },
+    {
+      "epoch": 1.4387612387612387,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9227,
+      "step": 1801
+    },
+    {
+      "epoch": 1.4395604395604396,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.0002,
+      "loss": 0.9219,
+      "step": 1802
+    },
+    {
+      "epoch": 1.4403596403596404,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9357,
+      "step": 1803
+    },
+    {
+      "epoch": 1.4411588411588412,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0002,
+      "loss": 0.9292,
+      "step": 1804
+    },
+    {
+      "epoch": 1.441958041958042,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.0002,
+      "loss": 0.9234,
+      "step": 1805
+    },
+    {
+      "epoch": 1.4427572427572428,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9259,
+      "step": 1806
+    },
+    {
+      "epoch": 1.4435564435564436,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0002,
+      "loss": 0.917,
+      "step": 1807
+    },
+    {
+      "epoch": 1.4443556443556442,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9285,
+      "step": 1808
+    },
+    {
+      "epoch": 1.4451548451548453,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9231,
+      "step": 1809
+    },
+    {
+      "epoch": 1.4459540459540459,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0002,
+      "loss": 0.9246,
+      "step": 1810
+    },
+    {
+      "epoch": 1.4467532467532467,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0002,
+      "loss": 0.9267,
+      "step": 1811
+    },
+    {
+      "epoch": 1.4475524475524475,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9104,
+      "step": 1812
+    },
+    {
+      "epoch": 1.4483516483516483,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0002,
+      "loss": 0.9295,
+      "step": 1813
+    },
+    {
+      "epoch": 1.4491508491508491,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0002,
+      "loss": 0.9242,
+      "step": 1814
+    },
+    {
+      "epoch": 1.44995004995005,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0002,
+      "loss": 0.9231,
+      "step": 1815
+    },
+    {
+      "epoch": 1.4507492507492508,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0002,
+      "loss": 0.9211,
+      "step": 1816
+    },
+    {
+      "epoch": 1.4515484515484516,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.0002,
+      "loss": 0.9206,
+      "step": 1817
+    },
+    {
+      "epoch": 1.4523476523476524,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0002,
+      "loss": 0.928,
+      "step": 1818
+    },
+    {
+      "epoch": 1.4531468531468532,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0002,
+      "loss": 0.9277,
+      "step": 1819
+    },
+    {
+      "epoch": 1.453946053946054,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9217,
+      "step": 1820
+    },
+    {
+      "epoch": 1.4547452547452546,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9277,
+      "step": 1821
+    },
+    {
+      "epoch": 1.4555444555444557,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9227,
+      "step": 1822
+    },
+    {
+      "epoch": 1.4563436563436563,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0002,
+      "loss": 0.9212,
+      "step": 1823
+    },
+    {
+      "epoch": 1.457142857142857,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0002,
+      "loss": 0.9127,
+      "step": 1824
+    },
+    {
+      "epoch": 1.457942057942058,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9256,
+      "step": 1825
+    },
+    {
+      "epoch": 1.4587412587412587,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.0002,
+      "loss": 0.9224,
+      "step": 1826
+    },
+    {
+      "epoch": 1.4595404595404595,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9263,
+      "step": 1827
+    },
+    {
+      "epoch": 1.4603396603396603,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9102,
+      "step": 1828
+    },
+    {
+      "epoch": 1.4611388611388612,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9187,
+      "step": 1829
+    },
+    {
+      "epoch": 1.461938061938062,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0002,
+      "loss": 0.9257,
+      "step": 1830
+    },
+    {
+      "epoch": 1.4627372627372628,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9298,
+      "step": 1831
+    },
+    {
+      "epoch": 1.4635364635364636,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0002,
+      "loss": 0.965,
+      "step": 1832
+    },
+    {
+      "epoch": 1.4643356643356644,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0002,
+      "loss": 0.9263,
+      "step": 1833
+    },
+    {
+      "epoch": 1.465134865134865,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0002,
+      "loss": 0.9196,
+      "step": 1834
+    },
+    {
+      "epoch": 1.465934065934066,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0002,
+      "loss": 0.9159,
+      "step": 1835
+    },
+    {
+      "epoch": 1.4667332667332666,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0002,
+      "loss": 0.9178,
+      "step": 1836
+    },
+    {
+      "epoch": 1.4675324675324675,
+      "grad_norm": 1.40625,
+      "learning_rate": 0.0002,
+      "loss": 0.933,
+      "step": 1837
+    },
+    {
+      "epoch": 1.4683316683316683,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9259,
+      "step": 1838
+    },
+    {
+      "epoch": 1.469130869130869,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9405,
+      "step": 1839
+    },
+    {
+      "epoch": 1.46993006993007,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.0002,
+      "loss": 0.9384,
+      "step": 1840
+    },
+    {
+      "epoch": 1.4707292707292707,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9138,
+      "step": 1841
+    },
+    {
+      "epoch": 1.4715284715284715,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0002,
+      "loss": 0.9343,
+      "step": 1842
+    },
+    {
+      "epoch": 1.4723276723276724,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0002,
+      "loss": 0.9213,
+      "step": 1843
+    },
+    {
+      "epoch": 1.4731268731268732,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0002,
+      "loss": 0.9213,
+      "step": 1844
+    },
+    {
+      "epoch": 1.473926073926074,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0002,
+      "loss": 0.9189,
+      "step": 1845
+    },
+    {
+      "epoch": 1.4747252747252748,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9239,
+      "step": 1846
+    },
+    {
+      "epoch": 1.4755244755244754,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0002,
+      "loss": 0.926,
+      "step": 1847
+    },
+    {
+      "epoch": 1.4763236763236764,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0002,
+      "loss": 0.9244,
+      "step": 1848
+    },
+    {
+      "epoch": 1.477122877122877,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9802,
+      "step": 1849
+    },
+    {
+      "epoch": 1.4779220779220779,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0002,
+      "loss": 0.9271,
+      "step": 1850
+    },
+    {
+      "epoch": 1.4787212787212787,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9238,
+      "step": 1851
+    },
+    {
+      "epoch": 1.4795204795204795,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9284,
+      "step": 1852
+    },
+    {
+      "epoch": 1.4803196803196803,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9142,
+      "step": 1853
+    },
+    {
+      "epoch": 1.4811188811188811,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0002,
+      "loss": 0.934,
+      "step": 1854
+    },
+    {
+      "epoch": 1.481918081918082,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9172,
+      "step": 1855
+    },
+    {
+      "epoch": 1.4827172827172828,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0002,
+      "loss": 0.9146,
+      "step": 1856
+    },
+    {
+      "epoch": 1.4835164835164836,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0002,
+      "loss": 0.9128,
+      "step": 1857
+    },
+    {
+      "epoch": 1.4843156843156844,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0002,
+      "loss": 0.9241,
+      "step": 1858
+    },
+    {
+      "epoch": 1.4851148851148852,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0002,
+      "loss": 0.9271,
+      "step": 1859
+    },
+    {
+      "epoch": 1.4859140859140858,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9217,
+      "step": 1860
+    },
+    {
+      "epoch": 1.4867132867132868,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0002,
+      "loss": 0.926,
+      "step": 1861
+    },
+    {
+      "epoch": 1.4875124875124874,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.0002,
+      "loss": 0.9276,
+      "step": 1862
+    },
+    {
+      "epoch": 1.4883116883116883,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9231,
+      "step": 1863
+    },
+    {
+      "epoch": 1.489110889110889,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.0002,
+      "loss": 0.9495,
+      "step": 1864
+    },
+    {
+      "epoch": 1.4899100899100899,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9296,
+      "step": 1865
+    },
+    {
+      "epoch": 1.4907092907092907,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0002,
+      "loss": 0.9198,
+      "step": 1866
+    },
+    {
+      "epoch": 1.4915084915084915,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.0002,
+      "loss": 0.9199,
+      "step": 1867
+    },
+    {
+      "epoch": 1.4923076923076923,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0002,
+      "loss": 0.9312,
+      "step": 1868
+    },
+    {
+      "epoch": 1.4931068931068932,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.0002,
+      "loss": 0.9228,
+      "step": 1869
+    },
+    {
+      "epoch": 1.493906093906094,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9199,
+      "step": 1870
+    },
+    {
+      "epoch": 1.4947052947052948,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0002,
+      "loss": 0.9256,
+      "step": 1871
+    },
+    {
+      "epoch": 1.4955044955044956,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0002,
+      "loss": 0.9222,
+      "step": 1872
+    },
+    {
+      "epoch": 1.4963036963036962,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0002,
+      "loss": 0.92,
+      "step": 1873
+    },
+    {
+      "epoch": 1.4971028971028972,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9204,
+      "step": 1874
+    },
+    {
+      "epoch": 1.4979020979020978,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0002,
+      "loss": 0.9293,
+      "step": 1875
+    },
+    {
+      "epoch": 1.4987012987012986,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0002,
+      "loss": 0.9285,
+      "step": 1876
+    },
+    {
+      "epoch": 1.4995004995004995,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0002,
+      "loss": 0.921,
+      "step": 1877
+    },
+    {
+      "epoch": 1.5002997002997003,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0002,
+      "loss": 0.9254,
+      "step": 1878
+    },
+    {
+      "epoch": 1.501098901098901,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0002,
+      "loss": 0.9233,
+      "step": 1879
+    },
+    {
+      "epoch": 1.501898101898102,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9268,
+      "step": 1880
+    },
+    {
+      "epoch": 1.5026973026973027,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0002,
+      "loss": 0.9226,
+      "step": 1881
+    },
+    {
+      "epoch": 1.5034965034965035,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9211,
+      "step": 1882
+    },
+    {
+      "epoch": 1.5042957042957044,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9236,
+      "step": 1883
+    },
+    {
+      "epoch": 1.505094905094905,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.0002,
+      "loss": 0.9127,
+      "step": 1884
+    },
+    {
+      "epoch": 1.505894105894106,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0002,
+      "loss": 0.9237,
+      "step": 1885
+    },
+    {
+      "epoch": 1.5066933066933066,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9351,
+      "step": 1886
+    },
+    {
+      "epoch": 1.5074925074925076,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.0002,
+      "loss": 0.9205,
+      "step": 1887
+    },
+    {
+      "epoch": 1.5082917082917082,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9243,
+      "step": 1888
+    },
+    {
+      "epoch": 1.509090909090909,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.0002,
+      "loss": 0.9157,
+      "step": 1889
+    },
+    {
+      "epoch": 1.5098901098901099,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0002,
+      "loss": 0.9148,
+      "step": 1890
+    },
+    {
+      "epoch": 1.5106893106893107,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9229,
+      "step": 1891
+    },
+    {
+      "epoch": 1.5114885114885115,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.0002,
+      "loss": 0.9153,
+      "step": 1892
+    },
+    {
+      "epoch": 1.5122877122877123,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0002,
+      "loss": 0.9229,
+      "step": 1893
+    },
+    {
+      "epoch": 1.5130869130869131,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.0002,
+      "loss": 0.9139,
+      "step": 1894
+    },
+    {
+      "epoch": 1.513886113886114,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0002,
+      "loss": 0.9211,
+      "step": 1895
+    },
+    {
+      "epoch": 1.5146853146853148,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.0002,
+      "loss": 0.9247,
+      "step": 1896
+    },
+    {
+      "epoch": 1.5154845154845153,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0002,
+      "loss": 0.9173,
+      "step": 1897
+    },
+    {
+      "epoch": 1.5162837162837164,
+      "grad_norm": 2.703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9576,
+      "step": 1898
+    },
+    {
+      "epoch": 1.517082917082917,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0002,
+      "loss": 0.9182,
+      "step": 1899
+    },
+    {
+      "epoch": 1.517882117882118,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0002,
+      "loss": 0.9245,
+      "step": 1900
+    },
+    {
+      "epoch": 1.5186813186813186,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.0002,
+      "loss": 0.9179,
+      "step": 1901
+    },
+    {
+      "epoch": 1.5194805194805194,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.0002,
+      "loss": 0.9286,
+      "step": 1902
+    },
+    {
+      "epoch": 1.5202797202797202,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9309,
+      "step": 1903
+    },
+    {
+      "epoch": 1.521078921078921,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.0002,
+      "loss": 0.9206,
+      "step": 1904
+    },
+    {
+      "epoch": 1.5218781218781219,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9251,
+      "step": 1905
+    },
+    {
+      "epoch": 1.5226773226773227,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9211,
+      "step": 1906
+    },
+    {
+      "epoch": 1.5234765234765235,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0002,
+      "loss": 0.9158,
+      "step": 1907
+    },
+    {
+      "epoch": 1.5242757242757243,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9149,
+      "step": 1908
+    },
+    {
+      "epoch": 1.5250749250749251,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9223,
+      "step": 1909
+    },
+    {
+      "epoch": 1.5258741258741257,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0002,
+      "loss": 0.9214,
+      "step": 1910
+    },
+    {
+      "epoch": 1.5266733266733268,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0002,
+      "loss": 0.9312,
+      "step": 1911
+    },
+    {
+      "epoch": 1.5274725274725274,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9359,
+      "step": 1912
+    },
+    {
+      "epoch": 1.5282717282717284,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0002,
+      "loss": 0.9163,
+      "step": 1913
+    },
+    {
+      "epoch": 1.529070929070929,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.0002,
+      "loss": 0.9192,
+      "step": 1914
+    },
+    {
+      "epoch": 1.5298701298701298,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0002,
+      "loss": 0.9258,
+      "step": 1915
+    },
+    {
+      "epoch": 1.5306693306693306,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0002,
+      "loss": 0.9193,
+      "step": 1916
+    },
+    {
+      "epoch": 1.5314685314685315,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9118,
+      "step": 1917
+    },
+    {
+      "epoch": 1.5322677322677323,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0002,
+      "loss": 0.9128,
+      "step": 1918
+    },
+    {
+      "epoch": 1.533066933066933,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9221,
+      "step": 1919
+    },
+    {
+      "epoch": 1.533866133866134,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.0002,
+      "loss": 0.9177,
+      "step": 1920
+    },
+    {
+      "epoch": 1.5346653346653345,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0002,
+      "loss": 0.975,
+      "step": 1921
+    },
+    {
+      "epoch": 1.5354645354645355,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0002,
+      "loss": 0.9204,
+      "step": 1922
+    },
+    {
+      "epoch": 1.5362637362637361,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0002,
+      "loss": 0.9138,
+      "step": 1923
+    },
+    {
+      "epoch": 1.5370629370629372,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9237,
+      "step": 1924
+    },
+    {
+      "epoch": 1.5378621378621378,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9233,
+      "step": 1925
+    },
+    {
+      "epoch": 1.5386613386613388,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0002,
+      "loss": 0.9283,
+      "step": 1926
+    },
+    {
+      "epoch": 1.5394605394605394,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9235,
+      "step": 1927
+    },
+    {
+      "epoch": 1.5402597402597402,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.0002,
+      "loss": 0.916,
+      "step": 1928
+    },
+    {
+      "epoch": 1.541058941058941,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.0002,
+      "loss": 0.923,
+      "step": 1929
+    },
+    {
+      "epoch": 1.5418581418581419,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0002,
+      "loss": 0.923,
+      "step": 1930
+    },
+    {
+      "epoch": 1.5426573426573427,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9125,
+      "step": 1931
+    },
+    {
+      "epoch": 1.5434565434565435,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0002,
+      "loss": 0.9243,
+      "step": 1932
+    },
+    {
+      "epoch": 1.5442557442557443,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9218,
+      "step": 1933
+    },
+    {
+      "epoch": 1.545054945054945,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0002,
+      "loss": 0.918,
+      "step": 1934
+    },
+    {
+      "epoch": 1.545854145854146,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0002,
+      "loss": 0.928,
+      "step": 1935
+    },
+    {
+      "epoch": 1.5466533466533465,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0002,
+      "loss": 0.9232,
+      "step": 1936
+    },
+    {
+      "epoch": 1.5474525474525476,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0002,
+      "loss": 0.9294,
+      "step": 1937
+    },
+    {
+      "epoch": 1.5482517482517482,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9232,
+      "step": 1938
+    },
+    {
+      "epoch": 1.5490509490509492,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9121,
+      "step": 1939
+    },
+    {
+      "epoch": 1.5498501498501498,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0002,
+      "loss": 0.9302,
+      "step": 1940
+    },
+    {
+      "epoch": 1.5506493506493506,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9285,
+      "step": 1941
+    },
+    {
+      "epoch": 1.5514485514485514,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0002,
+      "loss": 0.9259,
+      "step": 1942
+    },
+    {
+      "epoch": 1.5522477522477522,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9254,
+      "step": 1943
+    },
+    {
+      "epoch": 1.553046953046953,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0002,
+      "loss": 0.922,
+      "step": 1944
+    },
+    {
+      "epoch": 1.5538461538461539,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0002,
+      "loss": 0.9268,
+      "step": 1945
+    },
+    {
+      "epoch": 1.5546453546453547,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0002,
+      "loss": 0.9206,
+      "step": 1946
+    },
+    {
+      "epoch": 1.5554445554445553,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9234,
+      "step": 1947
+    },
+    {
+      "epoch": 1.5562437562437563,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0002,
+      "loss": 0.9289,
+      "step": 1948
+    },
+    {
+      "epoch": 1.557042957042957,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9165,
+      "step": 1949
+    },
+    {
+      "epoch": 1.557842157842158,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0002,
+      "loss": 0.9198,
+      "step": 1950
+    },
+    {
+      "epoch": 1.5586413586413586,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0002,
+      "loss": 0.9199,
+      "step": 1951
+    },
+    {
+      "epoch": 1.5594405594405596,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9158,
+      "step": 1952
+    },
+    {
+      "epoch": 1.5602397602397602,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0002,
+      "loss": 0.9258,
+      "step": 1953
+    },
+    {
+      "epoch": 1.561038961038961,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0002,
+      "loss": 0.9217,
+      "step": 1954
+    },
+    {
+      "epoch": 1.5618381618381618,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.0002,
+      "loss": 0.9165,
+      "step": 1955
+    },
+    {
+      "epoch": 1.5626373626373626,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0002,
+      "loss": 0.9206,
+      "step": 1956
+    },
+    {
+      "epoch": 1.5634365634365635,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.0002,
+      "loss": 0.9163,
+      "step": 1957
+    },
+    {
+      "epoch": 1.5642357642357643,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9213,
+      "step": 1958
+    },
+    {
+      "epoch": 1.565034965034965,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0002,
+      "loss": 0.9227,
+      "step": 1959
+    },
+    {
+      "epoch": 1.5658341658341657,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0002,
+      "loss": 0.9156,
+      "step": 1960
+    },
+    {
+      "epoch": 1.5666333666333667,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0002,
+      "loss": 0.9204,
+      "step": 1961
+    },
+    {
+      "epoch": 1.5674325674325673,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0002,
+      "loss": 0.9161,
+      "step": 1962
+    },
+    {
+      "epoch": 1.5682317682317684,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9226,
+      "step": 1963
+    },
+    {
+      "epoch": 1.569030969030969,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0002,
+      "loss": 0.9131,
+      "step": 1964
+    },
+    {
+      "epoch": 1.56983016983017,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9197,
+      "step": 1965
+    },
+    {
+      "epoch": 1.5706293706293706,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0002,
+      "loss": 0.9256,
+      "step": 1966
+    },
+    {
+      "epoch": 1.5714285714285714,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9268,
+      "step": 1967
+    },
+    {
+      "epoch": 1.5722277722277722,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9152,
+      "step": 1968
+    },
+    {
+      "epoch": 1.573026973026973,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9224,
+      "step": 1969
+    },
+    {
+      "epoch": 1.5738261738261738,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0002,
+      "loss": 0.9133,
+      "step": 1970
+    },
+    {
+      "epoch": 1.5746253746253747,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0002,
+      "loss": 0.922,
+      "step": 1971
+    },
+    {
+      "epoch": 1.5754245754245755,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0002,
+      "loss": 0.9315,
+      "step": 1972
+    },
+    {
+      "epoch": 1.576223776223776,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9208,
+      "step": 1973
+    },
+    {
+      "epoch": 1.5770229770229771,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9181,
+      "step": 1974
+    },
+    {
+      "epoch": 1.5778221778221777,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.0002,
+      "loss": 0.9235,
+      "step": 1975
+    },
+    {
+      "epoch": 1.5786213786213787,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0002,
+      "loss": 0.918,
+      "step": 1976
+    },
+    {
+      "epoch": 1.5794205794205793,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.0002,
+      "loss": 0.9173,
+      "step": 1977
+    },
+    {
+      "epoch": 1.5802197802197804,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0002,
+      "loss": 0.9213,
+      "step": 1978
+    },
+    {
+      "epoch": 1.581018981018981,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0002,
+      "loss": 0.9172,
+      "step": 1979
+    },
+    {
+      "epoch": 1.5818181818181818,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9162,
+      "step": 1980
+    },
+    {
+      "epoch": 1.5826173826173826,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.0002,
+      "loss": 0.9217,
+      "step": 1981
+    },
+    {
+      "epoch": 1.5834165834165834,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0002,
+      "loss": 0.92,
+      "step": 1982
+    },
+    {
+      "epoch": 1.5842157842157842,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0002,
+      "loss": 0.9096,
+      "step": 1983
+    },
+    {
+      "epoch": 1.585014985014985,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.0002,
+      "loss": 0.9194,
+      "step": 1984
+    },
+    {
+      "epoch": 1.5858141858141859,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.0002,
+      "loss": 0.9164,
+      "step": 1985
+    },
+    {
+      "epoch": 1.5866133866133865,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0002,
+      "loss": 0.9202,
+      "step": 1986
+    },
+    {
+      "epoch": 1.5874125874125875,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9187,
+      "step": 1987
+    },
+    {
+      "epoch": 1.588211788211788,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.0002,
+      "loss": 0.9282,
+      "step": 1988
+    },
+    {
+      "epoch": 1.5890109890109891,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0002,
+      "loss": 0.921,
+      "step": 1989
+    },
+    {
+      "epoch": 1.5898101898101897,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0002,
+      "loss": 0.9149,
+      "step": 1990
+    },
+    {
+      "epoch": 1.5906093906093908,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0002,
+      "loss": 0.9214,
+      "step": 1991
+    },
+    {
+      "epoch": 1.5914085914085914,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0002,
+      "loss": 0.9211,
+      "step": 1992
+    },
+    {
+      "epoch": 1.5922077922077922,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0002,
+      "loss": 0.9127,
+      "step": 1993
+    },
+    {
+      "epoch": 1.593006993006993,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9209,
+      "step": 1994
+    },
+    {
+      "epoch": 1.5938061938061938,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9268,
+      "step": 1995
+    },
+    {
+      "epoch": 1.5946053946053946,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0002,
+      "loss": 0.9212,
+      "step": 1996
+    },
+    {
+      "epoch": 1.5954045954045954,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0002,
+      "loss": 0.9136,
+      "step": 1997
+    },
+    {
+      "epoch": 1.5962037962037963,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0002,
+      "loss": 0.9294,
+      "step": 1998
+    },
+    {
+      "epoch": 1.5970029970029969,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0002,
+      "loss": 0.932,
+      "step": 1999
+    },
+    {
+      "epoch": 1.597802197802198,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0002,
+      "loss": 0.9216,
+      "step": 2000
+    },
+    {
+      "epoch": 1.5986013986013985,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0002,
+      "loss": 0.9167,
+      "step": 2001
+    },
+    {
+      "epoch": 1.5994005994005995,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9213,
+      "step": 2002
+    },
+    {
+      "epoch": 1.6001998001998001,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9136,
+      "step": 2003
+    },
+    {
+      "epoch": 1.6009990009990012,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0002,
+      "loss": 0.9263,
+      "step": 2004
+    },
+    {
+      "epoch": 1.6017982017982018,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9144,
+      "step": 2005
+    },
+    {
+      "epoch": 1.6025974025974026,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9223,
+      "step": 2006
+    },
+    {
+      "epoch": 1.6033966033966034,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0002,
+      "loss": 0.9288,
+      "step": 2007
+    },
+    {
+      "epoch": 1.6041958041958042,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0002,
+      "loss": 0.9185,
+      "step": 2008
+    },
+    {
+      "epoch": 1.604995004995005,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9255,
+      "step": 2009
+    },
+    {
+      "epoch": 1.6057942057942058,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0002,
+      "loss": 0.9198,
+      "step": 2010
+    },
+    {
+      "epoch": 1.6065934065934067,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.0002,
+      "loss": 0.9575,
+      "step": 2011
+    },
+    {
+      "epoch": 1.6073926073926073,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0002,
+      "loss": 0.9151,
+      "step": 2012
+    },
+    {
+      "epoch": 1.6081918081918083,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0002,
+      "loss": 0.9247,
+      "step": 2013
+    },
+    {
+      "epoch": 1.6089910089910089,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9137,
+      "step": 2014
+    },
+    {
+      "epoch": 1.60979020979021,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9193,
+      "step": 2015
+    },
+    {
+      "epoch": 1.6105894105894105,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0002,
+      "loss": 0.918,
+      "step": 2016
+    },
+    {
+      "epoch": 1.6113886113886113,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0002,
+      "loss": 0.9239,
+      "step": 2017
+    },
+    {
+      "epoch": 1.6121878121878122,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9244,
+      "step": 2018
+    },
+    {
+      "epoch": 1.612987012987013,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0002,
+      "loss": 0.9155,
+      "step": 2019
+    },
+    {
+      "epoch": 1.6137862137862138,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0002,
+      "loss": 0.925,
+      "step": 2020
+    },
+    {
+      "epoch": 1.6145854145854146,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0002,
+      "loss": 0.9297,
+      "step": 2021
+    },
+    {
+      "epoch": 1.6153846153846154,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9247,
+      "step": 2022
+    },
+    {
+      "epoch": 1.6161838161838162,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0002,
+      "loss": 0.9253,
+      "step": 2023
+    },
+    {
+      "epoch": 1.616983016983017,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0002,
+      "loss": 0.9177,
+      "step": 2024
+    },
+    {
+      "epoch": 1.6177822177822176,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0002,
+      "loss": 0.9248,
+      "step": 2025
+    },
+    {
+      "epoch": 1.6185814185814187,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0002,
+      "loss": 0.9202,
+      "step": 2026
+    },
+    {
+      "epoch": 1.6193806193806193,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9278,
+      "step": 2027
+    },
+    {
+      "epoch": 1.6201798201798203,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0002,
+      "loss": 0.9149,
+      "step": 2028
+    },
+    {
+      "epoch": 1.620979020979021,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9128,
+      "step": 2029
+    },
+    {
+      "epoch": 1.6217782217782217,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0002,
+      "loss": 0.9186,
+      "step": 2030
+    },
+    {
+      "epoch": 1.6225774225774225,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0002,
+      "loss": 0.909,
+      "step": 2031
+    },
+    {
+      "epoch": 1.6233766233766234,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.0002,
+      "loss": 0.9128,
+      "step": 2032
+    },
+    {
+      "epoch": 1.6241758241758242,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9209,
+      "step": 2033
+    },
+    {
+      "epoch": 1.624975024975025,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9093,
+      "step": 2034
+    },
+    {
+      "epoch": 1.6257742257742258,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.0002,
+      "loss": 0.9174,
+      "step": 2035
+    },
+    {
+      "epoch": 1.6265734265734266,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0002,
+      "loss": 0.9185,
+      "step": 2036
+    },
+    {
+      "epoch": 1.6273726273726274,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.0002,
+      "loss": 0.9119,
+      "step": 2037
+    },
+    {
+      "epoch": 1.628171828171828,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9291,
+      "step": 2038
+    },
+    {
+      "epoch": 1.628971028971029,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9189,
+      "step": 2039
+    },
+    {
+      "epoch": 1.6297702297702297,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9301,
+      "step": 2040
+    },
+    {
+      "epoch": 1.6305694305694307,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9111,
+      "step": 2041
+    },
+    {
+      "epoch": 1.6313686313686313,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0002,
+      "loss": 0.9079,
+      "step": 2042
+    },
+    {
+      "epoch": 1.6321678321678321,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0002,
+      "loss": 0.9065,
+      "step": 2043
+    },
+    {
+      "epoch": 1.632967032967033,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.0002,
+      "loss": 0.9173,
+      "step": 2044
+    },
+    {
+      "epoch": 1.6337662337662338,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9174,
+      "step": 2045
+    },
+    {
+      "epoch": 1.6345654345654346,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0002,
+      "loss": 0.9206,
+      "step": 2046
+    },
+    {
+      "epoch": 1.6353646353646354,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.0002,
+      "loss": 0.9367,
+      "step": 2047
+    },
+    {
+      "epoch": 1.6361638361638362,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0002,
+      "loss": 0.9154,
+      "step": 2048
+    },
+    {
+      "epoch": 1.6369630369630368,
+      "grad_norm": 0.796875,
+      "learning_rate": 0.0002,
+      "loss": 0.9629,
+      "step": 2049
+    },
+    {
+      "epoch": 1.6377622377622378,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0002,
+      "loss": 0.9183,
+      "step": 2050
+    },
+    {
+      "epoch": 1.6385614385614384,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0002,
+      "loss": 0.9139,
+      "step": 2051
+    },
+    {
+      "epoch": 1.6393606393606395,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9218,
+      "step": 2052
+    },
+    {
+      "epoch": 1.64015984015984,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9215,
+      "step": 2053
+    },
+    {
+      "epoch": 1.640959040959041,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0002,
+      "loss": 0.9125,
+      "step": 2054
+    },
+    {
+      "epoch": 1.6417582417582417,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0002,
+      "loss": 0.9304,
+      "step": 2055
+    },
+    {
+      "epoch": 1.6425574425574425,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9159,
+      "step": 2056
+    },
+    {
+      "epoch": 1.6433566433566433,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0002,
+      "loss": 0.9231,
+      "step": 2057
+    },
+    {
+      "epoch": 1.6441558441558441,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0002,
+      "loss": 0.9292,
+      "step": 2058
+    },
+    {
+      "epoch": 1.644955044955045,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9275,
+      "step": 2059
+    },
+    {
+      "epoch": 1.6457542457542458,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.0002,
+      "loss": 0.9215,
+      "step": 2060
+    },
+    {
+      "epoch": 1.6465534465534466,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0002,
+      "loss": 0.9192,
+      "step": 2061
+    },
+    {
+      "epoch": 1.6473526473526472,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9185,
+      "step": 2062
+    },
+    {
+      "epoch": 1.6481518481518482,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9234,
+      "step": 2063
+    },
+    {
+      "epoch": 1.6489510489510488,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0002,
+      "loss": 0.915,
+      "step": 2064
+    },
+    {
+      "epoch": 1.6497502497502499,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0002,
+      "loss": 0.9186,
+      "step": 2065
+    },
+    {
+      "epoch": 1.6505494505494505,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0002,
+      "loss": 0.9167,
+      "step": 2066
+    },
+    {
+      "epoch": 1.6513486513486515,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9107,
+      "step": 2067
+    },
+    {
+      "epoch": 1.652147852147852,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.0002,
+      "loss": 0.9171,
+      "step": 2068
+    },
+    {
+      "epoch": 1.652947052947053,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.0002,
+      "loss": 0.9296,
+      "step": 2069
+    },
+    {
+      "epoch": 1.6537462537462537,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9096,
+      "step": 2070
+    },
+    {
+      "epoch": 1.6545454545454545,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0002,
+      "loss": 0.9253,
+      "step": 2071
+    },
+    {
+      "epoch": 1.6553446553446554,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0002,
+      "loss": 0.9077,
+      "step": 2072
+    },
+    {
+      "epoch": 1.6561438561438562,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.0002,
+      "loss": 0.9592,
+      "step": 2073
+    },
+    {
+      "epoch": 1.656943056943057,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0002,
+      "loss": 0.926,
+      "step": 2074
+    },
+    {
+      "epoch": 1.6577422577422576,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9142,
+      "step": 2075
+    },
+    {
+      "epoch": 1.6585414585414586,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9164,
+      "step": 2076
+    },
+    {
+      "epoch": 1.6593406593406592,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9171,
+      "step": 2077
+    },
+    {
+      "epoch": 1.6601398601398603,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9246,
+      "step": 2078
+    },
+    {
+      "epoch": 1.6609390609390609,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9234,
+      "step": 2079
+    },
+    {
+      "epoch": 1.661738261738262,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.0002,
+      "loss": 0.9187,
+      "step": 2080
+    },
+    {
+      "epoch": 1.6625374625374625,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0002,
+      "loss": 0.9178,
+      "step": 2081
+    },
+    {
+      "epoch": 1.6633366633366633,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9239,
+      "step": 2082
+    },
+    {
+      "epoch": 1.6641358641358641,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9217,
+      "step": 2083
+    },
+    {
+      "epoch": 1.664935064935065,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0002,
+      "loss": 0.9149,
+      "step": 2084
+    },
+    {
+      "epoch": 1.6657342657342658,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0002,
+      "loss": 0.9173,
+      "step": 2085
+    },
+    {
+      "epoch": 1.6665334665334666,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9154,
+      "step": 2086
+    },
+    {
+      "epoch": 1.6673326673326674,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0002,
+      "loss": 0.9106,
+      "step": 2087
+    },
+    {
+      "epoch": 1.668131868131868,
+      "grad_norm": 0.91015625,
+      "learning_rate": 0.0002,
+      "loss": 0.9596,
+      "step": 2088
+    },
+    {
+      "epoch": 1.668931068931069,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9086,
+      "step": 2089
+    },
+    {
+      "epoch": 1.6697302697302696,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9138,
+      "step": 2090
+    },
+    {
+      "epoch": 1.6705294705294707,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9243,
+      "step": 2091
+    },
+    {
+      "epoch": 1.6713286713286712,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0002,
+      "loss": 0.926,
+      "step": 2092
+    },
+    {
+      "epoch": 1.6721278721278723,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9198,
+      "step": 2093
+    },
+    {
+      "epoch": 1.6729270729270729,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0002,
+      "loss": 0.9263,
+      "step": 2094
+    },
+    {
+      "epoch": 1.6737262737262737,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0002,
+      "loss": 0.9124,
+      "step": 2095
+    },
+    {
+      "epoch": 1.6745254745254745,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0002,
+      "loss": 0.919,
+      "step": 2096
+    },
+    {
+      "epoch": 1.6753246753246753,
+      "grad_norm": 1.75,
+      "learning_rate": 0.0002,
+      "loss": 0.9461,
+      "step": 2097
+    },
+    {
+      "epoch": 1.6761238761238761,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.0002,
+      "loss": 0.9374,
+      "step": 2098
+    },
+    {
+      "epoch": 1.676923076923077,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.0002,
+      "loss": 0.9201,
+      "step": 2099
+    },
+    {
+      "epoch": 1.6777222777222778,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9268,
+      "step": 2100
+    },
+    {
+      "epoch": 1.6785214785214784,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9257,
+      "step": 2101
+    },
+    {
+      "epoch": 1.6793206793206794,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9213,
+      "step": 2102
+    },
+    {
+      "epoch": 1.68011988011988,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0002,
+      "loss": 0.927,
+      "step": 2103
+    },
+    {
+      "epoch": 1.680919080919081,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9173,
+      "step": 2104
+    },
+    {
+      "epoch": 1.6817182817182816,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.0002,
+      "loss": 0.9561,
+      "step": 2105
+    },
+    {
+      "epoch": 1.6825174825174827,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9239,
+      "step": 2106
+    },
+    {
+      "epoch": 1.6833166833166833,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.0002,
+      "loss": 0.9257,
+      "step": 2107
+    },
+    {
+      "epoch": 1.684115884115884,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9215,
+      "step": 2108
+    },
+    {
+      "epoch": 1.684915084915085,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9515,
+      "step": 2109
+    },
+    {
+      "epoch": 1.6857142857142857,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0002,
+      "loss": 0.9127,
+      "step": 2110
+    },
+    {
+      "epoch": 1.6865134865134865,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9163,
+      "step": 2111
+    },
+    {
+      "epoch": 1.6873126873126874,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0002,
+      "loss": 0.924,
+      "step": 2112
+    },
+    {
+      "epoch": 1.6881118881118882,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9318,
+      "step": 2113
+    },
+    {
+      "epoch": 1.6889110889110888,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0002,
+      "loss": 0.9206,
+      "step": 2114
+    },
+    {
+      "epoch": 1.6897102897102898,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0002,
+      "loss": 0.9255,
+      "step": 2115
+    },
+    {
+      "epoch": 1.6905094905094904,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0002,
+      "loss": 0.9744,
+      "step": 2116
+    },
+    {
+      "epoch": 1.6913086913086914,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.0002,
+      "loss": 0.9242,
+      "step": 2117
+    },
+    {
+      "epoch": 1.692107892107892,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0002,
+      "loss": 0.9187,
+      "step": 2118
+    },
+    {
+      "epoch": 1.692907092907093,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0002,
+      "loss": 0.9153,
+      "step": 2119
+    },
+    {
+      "epoch": 1.6937062937062937,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.0002,
+      "loss": 0.9145,
+      "step": 2120
+    },
+    {
+      "epoch": 1.6945054945054945,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.0002,
+      "loss": 0.9229,
+      "step": 2121
+    },
+    {
+      "epoch": 1.6953046953046953,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9323,
+      "step": 2122
+    },
+    {
+      "epoch": 1.6961038961038961,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9185,
+      "step": 2123
+    },
+    {
+      "epoch": 1.696903096903097,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.0002,
+      "loss": 0.9546,
+      "step": 2124
+    },
+    {
+      "epoch": 1.6977022977022977,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.0002,
+      "loss": 0.9146,
+      "step": 2125
+    },
+    {
+      "epoch": 1.6985014985014986,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0002,
+      "loss": 0.9225,
+      "step": 2126
+    },
+    {
+      "epoch": 1.6993006993006992,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0002,
+      "loss": 0.908,
+      "step": 2127
+    },
+    {
+      "epoch": 1.7000999000999002,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0002,
+      "loss": 0.9261,
+      "step": 2128
+    },
+    {
+      "epoch": 1.7008991008991008,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.0002,
+      "loss": 0.9604,
+      "step": 2129
+    },
+    {
+      "epoch": 1.7016983016983018,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0002,
+      "loss": 0.9175,
+      "step": 2130
+    },
+    {
+      "epoch": 1.7024975024975024,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.0002,
+      "loss": 0.9197,
+      "step": 2131
+    },
+    {
+      "epoch": 1.7032967032967035,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0002,
+      "loss": 0.9179,
+      "step": 2132
+    },
+    {
+      "epoch": 1.704095904095904,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9208,
+      "step": 2133
+    },
+    {
+      "epoch": 1.7048951048951049,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0002,
+      "loss": 0.92,
+      "step": 2134
+    },
+    {
+      "epoch": 1.7056943056943057,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0002,
+      "loss": 0.9144,
+      "step": 2135
+    },
+    {
+      "epoch": 1.7064935064935065,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0002,
+      "loss": 0.9213,
+      "step": 2136
+    },
+    {
+      "epoch": 1.7072927072927073,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0002,
+      "loss": 0.9213,
+      "step": 2137
+    },
+    {
+      "epoch": 1.7080919080919081,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0002,
+      "loss": 0.9139,
+      "step": 2138
+    },
+    {
+      "epoch": 1.708891108891109,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0002,
+      "loss": 0.9281,
+      "step": 2139
+    },
+    {
+      "epoch": 1.7096903096903096,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9219,
+      "step": 2140
+    },
+    {
+      "epoch": 1.7104895104895106,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.0002,
+      "loss": 0.922,
+      "step": 2141
+    },
+    {
+      "epoch": 1.7112887112887112,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.0002,
+      "loss": 0.9122,
+      "step": 2142
+    },
+    {
+      "epoch": 1.7120879120879122,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9133,
+      "step": 2143
+    },
+    {
+      "epoch": 1.7128871128871128,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.0002,
+      "loss": 0.9153,
+      "step": 2144
+    },
+    {
+      "epoch": 1.7136863136863136,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.0002,
+      "loss": 0.9225,
+      "step": 2145
+    },
+    {
+      "epoch": 1.7144855144855145,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0002,
+      "loss": 0.9251,
+      "step": 2146
+    },
+    {
+      "epoch": 1.7152847152847153,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9249,
+      "step": 2147
+    },
+    {
+      "epoch": 1.716083916083916,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0002,
+      "loss": 0.9147,
+      "step": 2148
+    },
+    {
+      "epoch": 1.716883116883117,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0002,
+      "loss": 0.9139,
+      "step": 2149
+    },
+    {
+      "epoch": 1.7176823176823177,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9182,
+      "step": 2150
+    },
+    {
+      "epoch": 1.7184815184815185,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9239,
+      "step": 2151
+    },
+    {
+      "epoch": 1.7192807192807193,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9208,
+      "step": 2152
+    },
+    {
+      "epoch": 1.72007992007992,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9119,
+      "step": 2153
+    },
+    {
+      "epoch": 1.720879120879121,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.0002,
+      "loss": 0.9161,
+      "step": 2154
+    },
+    {
+      "epoch": 1.7216783216783216,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0002,
+      "loss": 0.9232,
+      "step": 2155
+    },
+    {
+      "epoch": 1.7224775224775226,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0002,
+      "loss": 0.9202,
+      "step": 2156
+    },
+    {
+      "epoch": 1.7232767232767232,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9047,
+      "step": 2157
+    },
+    {
+      "epoch": 1.724075924075924,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0002,
+      "loss": 0.927,
+      "step": 2158
+    },
+    {
+      "epoch": 1.7248751248751248,
+      "grad_norm": 0.375,
+      "learning_rate": 0.0002,
+      "loss": 0.9273,
+      "step": 2159
+    },
+    {
+      "epoch": 1.7256743256743257,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9179,
+      "step": 2160
+    },
+    {
+      "epoch": 1.7264735264735265,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9294,
+      "step": 2161
+    },
+    {
+      "epoch": 1.7272727272727273,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0002,
+      "loss": 0.9111,
+      "step": 2162
+    },
+    {
+      "epoch": 1.728071928071928,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9077,
+      "step": 2163
+    },
+    {
+      "epoch": 1.728871128871129,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0002,
+      "loss": 0.9085,
+      "step": 2164
+    },
+    {
+      "epoch": 1.7296703296703297,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9154,
+      "step": 2165
+    },
+    {
+      "epoch": 1.7304695304695303,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9227,
+      "step": 2166
+    },
+    {
+      "epoch": 1.7312687312687314,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0002,
+      "loss": 0.923,
+      "step": 2167
+    },
+    {
+      "epoch": 1.732067932067932,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0002,
+      "loss": 0.9218,
+      "step": 2168
+    },
+    {
+      "epoch": 1.732867132867133,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0002,
+      "loss": 0.9177,
+      "step": 2169
+    },
+    {
+      "epoch": 1.7336663336663336,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9146,
+      "step": 2170
+    },
+    {
+      "epoch": 1.7344655344655344,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9132,
+      "step": 2171
+    },
+    {
+      "epoch": 1.7352647352647352,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9164,
+      "step": 2172
+    },
+    {
+      "epoch": 1.736063936063936,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0002,
+      "loss": 0.9087,
+      "step": 2173
+    },
+    {
+      "epoch": 1.7368631368631369,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0002,
+      "loss": 0.929,
+      "step": 2174
+    },
+    {
+      "epoch": 1.7376623376623377,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0002,
+      "loss": 0.9109,
+      "step": 2175
+    },
+    {
+      "epoch": 1.7384615384615385,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.0002,
+      "loss": 0.9208,
+      "step": 2176
+    },
+    {
+      "epoch": 1.739260739260739,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.0002,
+      "loss": 0.9193,
+      "step": 2177
+    },
+    {
+      "epoch": 1.7400599400599401,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9262,
+      "step": 2178
+    },
+    {
+      "epoch": 1.7408591408591407,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9248,
+      "step": 2179
+    },
+    {
+      "epoch": 1.7416583416583418,
+      "grad_norm": 0.86328125,
+      "learning_rate": 0.0002,
+      "loss": 0.9198,
+      "step": 2180
+    },
+    {
+      "epoch": 1.7424575424575424,
+      "grad_norm": 1.4765625,
+      "learning_rate": 0.0002,
+      "loss": 0.9124,
+      "step": 2181
+    },
+    {
+      "epoch": 1.7432567432567434,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.0002,
+      "loss": 0.9171,
+      "step": 2182
+    },
+    {
+      "epoch": 1.744055944055944,
+      "grad_norm": 0.84375,
+      "learning_rate": 0.0002,
+      "loss": 0.9242,
+      "step": 2183
+    },
+    {
+      "epoch": 1.7448551448551448,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9103,
+      "step": 2184
+    },
+    {
+      "epoch": 1.7456543456543456,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9221,
+      "step": 2185
+    },
+    {
+      "epoch": 1.7464535464535464,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.0002,
+      "loss": 0.913,
+      "step": 2186
+    },
+    {
+      "epoch": 1.7472527472527473,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.0002,
+      "loss": 0.9158,
+      "step": 2187
+    },
+    {
+      "epoch": 1.748051948051948,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0002,
+      "loss": 0.917,
+      "step": 2188
+    },
+    {
+      "epoch": 1.748851148851149,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0002,
+      "loss": 0.9123,
+      "step": 2189
+    },
+    {
+      "epoch": 1.7496503496503495,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9217,
+      "step": 2190
+    },
+    {
+      "epoch": 1.7504495504495505,
+      "grad_norm": 0.80078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9244,
+      "step": 2191
+    },
+    {
+      "epoch": 1.7512487512487511,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.0002,
+      "loss": 0.9114,
+      "step": 2192
+    },
+    {
+      "epoch": 1.7520479520479522,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0002,
+      "loss": 0.9177,
+      "step": 2193
+    },
+    {
+      "epoch": 1.7528471528471528,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.0002,
+      "loss": 0.9157,
+      "step": 2194
+    },
+    {
+      "epoch": 1.7536463536463538,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.0002,
+      "loss": 0.9155,
+      "step": 2195
+    },
+    {
+      "epoch": 1.7544455544455544,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.0002,
+      "loss": 0.9145,
+      "step": 2196
+    },
+    {
+      "epoch": 1.7552447552447552,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.0002,
+      "loss": 0.9221,
+      "step": 2197
+    },
+    {
+      "epoch": 1.756043956043956,
+      "grad_norm": 0.82421875,
+      "learning_rate": 0.0002,
+      "loss": 0.9047,
+      "step": 2198
+    },
+    {
+      "epoch": 1.7568431568431568,
+      "grad_norm": 0.82421875,
+      "learning_rate": 0.0002,
+      "loss": 0.9107,
+      "step": 2199
+    },
+    {
+      "epoch": 1.7576423576423577,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9247,
+      "step": 2200
+    },
+    {
+      "epoch": 1.7584415584415585,
+      "grad_norm": 0.91015625,
+      "learning_rate": 0.0002,
+      "loss": 0.9219,
+      "step": 2201
+    },
+    {
+      "epoch": 1.7592407592407593,
+      "grad_norm": 0.625,
+      "learning_rate": 0.0002,
+      "loss": 0.9231,
+      "step": 2202
+    },
+    {
+      "epoch": 1.7600399600399599,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9155,
+      "step": 2203
+    },
+    {
+      "epoch": 1.760839160839161,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0002,
+      "loss": 0.9032,
+      "step": 2204
+    },
+    {
+      "epoch": 1.7616383616383615,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.0002,
+      "loss": 0.9189,
+      "step": 2205
+    },
+    {
+      "epoch": 1.7624375624375626,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9189,
+      "step": 2206
+    },
+    {
+      "epoch": 1.7632367632367631,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0002,
+      "loss": 0.9169,
+      "step": 2207
+    },
+    {
+      "epoch": 1.7640359640359642,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0002,
+      "loss": 0.9115,
+      "step": 2208
+    },
+    {
+      "epoch": 1.7648351648351648,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9204,
+      "step": 2209
+    },
+    {
+      "epoch": 1.7656343656343656,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9158,
+      "step": 2210
+    },
+    {
+      "epoch": 1.7664335664335664,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0002,
+      "loss": 0.9124,
+      "step": 2211
+    },
+    {
+      "epoch": 1.7672327672327672,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9158,
+      "step": 2212
+    },
+    {
+      "epoch": 1.768031968031968,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9089,
+      "step": 2213
+    },
+    {
+      "epoch": 1.7688311688311689,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0002,
+      "loss": 0.9684,
+      "step": 2214
+    },
+    {
+      "epoch": 1.7696303696303697,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.0002,
+      "loss": 0.9149,
+      "step": 2215
+    },
+    {
+      "epoch": 1.7704295704295703,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9209,
+      "step": 2216
+    },
+    {
+      "epoch": 1.7712287712287713,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.0002,
+      "loss": 0.9176,
+      "step": 2217
+    },
+    {
+      "epoch": 1.772027972027972,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9216,
+      "step": 2218
+    },
+    {
+      "epoch": 1.772827172827173,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9183,
+      "step": 2219
+    },
+    {
+      "epoch": 1.7736263736263735,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0002,
+      "loss": 0.9137,
+      "step": 2220
+    },
+    {
+      "epoch": 1.7744255744255746,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9051,
+      "step": 2221
+    },
+    {
+      "epoch": 1.7752247752247752,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0002,
+      "loss": 0.9092,
+      "step": 2222
+    },
+    {
+      "epoch": 1.776023976023976,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0002,
+      "loss": 0.9277,
+      "step": 2223
+    },
+    {
+      "epoch": 1.7768231768231768,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9234,
+      "step": 2224
+    },
+    {
+      "epoch": 1.7776223776223776,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9004,
+      "step": 2225
+    },
+    {
+      "epoch": 1.7784215784215784,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0002,
+      "loss": 0.9054,
+      "step": 2226
+    },
+    {
+      "epoch": 1.7792207792207793,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9158,
+      "step": 2227
+    },
+    {
+      "epoch": 1.78001998001998,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0002,
+      "loss": 0.9209,
+      "step": 2228
+    },
+    {
+      "epoch": 1.7808191808191807,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9111,
+      "step": 2229
+    },
+    {
+      "epoch": 1.7816183816183817,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9183,
+      "step": 2230
+    },
+    {
+      "epoch": 1.7824175824175823,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0002,
+      "loss": 0.9152,
+      "step": 2231
+    },
+    {
+      "epoch": 1.7832167832167833,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0002,
+      "loss": 0.92,
+      "step": 2232
+    },
+    {
+      "epoch": 1.784015984015984,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9211,
+      "step": 2233
+    },
+    {
+      "epoch": 1.784815184815185,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0002,
+      "loss": 0.9286,
+      "step": 2234
+    },
+    {
+      "epoch": 1.7856143856143856,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0002,
+      "loss": 0.9184,
+      "step": 2235
+    },
+    {
+      "epoch": 1.7864135864135864,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9176,
+      "step": 2236
+    },
+    {
+      "epoch": 1.7872127872127872,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0002,
+      "loss": 0.9165,
+      "step": 2237
+    },
+    {
+      "epoch": 1.788011988011988,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0002,
+      "loss": 0.9108,
+      "step": 2238
+    },
+    {
+      "epoch": 1.7888111888111888,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9251,
+      "step": 2239
+    },
+    {
+      "epoch": 1.7896103896103897,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0002,
+      "loss": 0.9134,
+      "step": 2240
+    },
+    {
+      "epoch": 1.7904095904095905,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0002,
+      "loss": 0.9211,
+      "step": 2241
+    },
+    {
+      "epoch": 1.791208791208791,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9255,
+      "step": 2242
+    },
+    {
+      "epoch": 1.792007992007992,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0002,
+      "loss": 0.9673,
+      "step": 2243
+    },
+    {
+      "epoch": 1.7928071928071927,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0002,
+      "loss": 0.9183,
+      "step": 2244
+    },
+    {
+      "epoch": 1.7936063936063937,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9134,
+      "step": 2245
+    },
+    {
+      "epoch": 1.7944055944055943,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9187,
+      "step": 2246
+    },
+    {
+      "epoch": 1.7952047952047954,
+      "grad_norm": 1.578125,
+      "learning_rate": 0.0002,
+      "loss": 0.9369,
+      "step": 2247
+    },
+    {
+      "epoch": 1.796003996003996,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.0002,
+      "loss": 0.9484,
+      "step": 2248
+    },
+    {
+      "epoch": 1.7968031968031968,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0002,
+      "loss": 0.9103,
+      "step": 2249
+    },
+    {
+      "epoch": 1.7976023976023976,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.0002,
+      "loss": 0.9388,
+      "step": 2250
+    },
+    {
+      "epoch": 1.7984015984015984,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0002,
+      "loss": 0.9157,
+      "step": 2251
+    },
+    {
+      "epoch": 1.7992007992007992,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.0002,
+      "loss": 0.9099,
+      "step": 2252
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0002,
+      "loss": 0.9215,
+      "step": 2253
+    },
+    {
+      "epoch": 1.8007992007992009,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0002,
+      "loss": 0.9204,
+      "step": 2254
+    },
+    {
+      "epoch": 1.8015984015984015,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9137,
+      "step": 2255
+    },
+    {
+      "epoch": 1.8023976023976025,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0002,
+      "loss": 0.9107,
+      "step": 2256
+    },
+    {
+      "epoch": 1.803196803196803,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9294,
+      "step": 2257
+    },
+    {
+      "epoch": 1.8039960039960041,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9263,
+      "step": 2258
+    },
+    {
+      "epoch": 1.8047952047952047,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9265,
+      "step": 2259
+    },
+    {
+      "epoch": 1.8055944055944058,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0002,
+      "loss": 0.9151,
+      "step": 2260
+    },
+    {
+      "epoch": 1.8063936063936064,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0002,
+      "loss": 0.9185,
+      "step": 2261
+    },
+    {
+      "epoch": 1.8071928071928072,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0002,
+      "loss": 0.9125,
+      "step": 2262
+    },
+    {
+      "epoch": 1.807992007992008,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9171,
+      "step": 2263
+    },
+    {
+      "epoch": 1.8087912087912088,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0002,
+      "loss": 0.9181,
+      "step": 2264
+    },
+    {
+      "epoch": 1.8095904095904096,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9174,
+      "step": 2265
+    },
+    {
+      "epoch": 1.8103896103896104,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0002,
+      "loss": 0.9213,
+      "step": 2266
+    },
+    {
+      "epoch": 1.8111888111888113,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0002,
+      "loss": 0.9189,
+      "step": 2267
+    },
+    {
+      "epoch": 1.8119880119880118,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0002,
+      "loss": 0.9139,
+      "step": 2268
+    },
+    {
+      "epoch": 1.8127872127872129,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.0002,
+      "loss": 0.9163,
+      "step": 2269
+    },
+    {
+      "epoch": 1.8135864135864135,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9216,
+      "step": 2270
+    },
+    {
+      "epoch": 1.8143856143856145,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9161,
+      "step": 2271
+    },
+    {
+      "epoch": 1.8151848151848151,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0002,
+      "loss": 0.9116,
+      "step": 2272
+    },
+    {
+      "epoch": 1.815984015984016,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0002,
+      "loss": 0.9232,
+      "step": 2273
+    },
+    {
+      "epoch": 1.8167832167832167,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0002,
+      "loss": 0.9271,
+      "step": 2274
+    },
+    {
+      "epoch": 1.8175824175824176,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9257,
+      "step": 2275
+    },
+    {
+      "epoch": 1.8183816183816184,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9209,
+      "step": 2276
+    },
+    {
+      "epoch": 1.8191808191808192,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0002,
+      "loss": 0.9171,
+      "step": 2277
+    },
+    {
+      "epoch": 1.81998001998002,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0002,
+      "loss": 0.9169,
+      "step": 2278
+    },
+    {
+      "epoch": 1.8207792207792208,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0002,
+      "loss": 0.9194,
+      "step": 2279
+    },
+    {
+      "epoch": 1.8215784215784216,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0002,
+      "loss": 0.9256,
+      "step": 2280
+    },
+    {
+      "epoch": 1.8223776223776222,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.0002,
+      "loss": 0.9182,
+      "step": 2281
+    },
+    {
+      "epoch": 1.8231768231768233,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0002,
+      "loss": 0.9223,
+      "step": 2282
+    },
+    {
+      "epoch": 1.8239760239760239,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0002,
+      "loss": 0.9102,
+      "step": 2283
+    },
+    {
+      "epoch": 1.824775224775225,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9232,
+      "step": 2284
+    },
+    {
+      "epoch": 1.8255744255744255,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0002,
+      "loss": 0.9105,
+      "step": 2285
+    },
+    {
+      "epoch": 1.8263736263736263,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0002,
+      "loss": 0.9263,
+      "step": 2286
+    },
+    {
+      "epoch": 1.8271728271728271,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9256,
+      "step": 2287
+    },
+    {
+      "epoch": 1.827972027972028,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0002,
+      "loss": 0.9125,
+      "step": 2288
+    },
+    {
+      "epoch": 1.8287712287712288,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0002,
+      "loss": 0.9285,
+      "step": 2289
+    },
+    {
+      "epoch": 1.8295704295704296,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9099,
+      "step": 2290
+    },
+    {
+      "epoch": 1.8303696303696304,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0002,
+      "loss": 0.9233,
+      "step": 2291
+    },
+    {
+      "epoch": 1.8311688311688312,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0002,
+      "loss": 0.9229,
+      "step": 2292
+    },
+    {
+      "epoch": 1.831968031968032,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0002,
+      "loss": 0.9154,
+      "step": 2293
+    },
+    {
+      "epoch": 1.8327672327672326,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0002,
+      "loss": 0.9161,
+      "step": 2294
+    },
+    {
+      "epoch": 1.8335664335664337,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9129,
+      "step": 2295
+    },
+    {
+      "epoch": 1.8343656343656343,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0002,
+      "loss": 0.9183,
+      "step": 2296
+    },
+    {
+      "epoch": 1.8351648351648353,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0002,
+      "loss": 0.9291,
+      "step": 2297
+    },
+    {
+      "epoch": 1.835964035964036,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9147,
+      "step": 2298
+    },
+    {
+      "epoch": 1.8367632367632367,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.0002,
+      "loss": 0.9228,
+      "step": 2299
+    },
+    {
+      "epoch": 1.8375624375624375,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9208,
+      "step": 2300
+    },
+    {
+      "epoch": 1.8383616383616384,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.0002,
+      "loss": 0.916,
+      "step": 2301
+    },
+    {
+      "epoch": 1.8391608391608392,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.0002,
+      "loss": 0.9211,
+      "step": 2302
+    },
+    {
+      "epoch": 1.83996003996004,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.0002,
+      "loss": 0.9438,
+      "step": 2303
+    },
+    {
+      "epoch": 1.8407592407592408,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0002,
+      "loss": 0.9186,
+      "step": 2304
+    },
+    {
+      "epoch": 1.8415584415584414,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0002,
+      "loss": 0.9031,
+      "step": 2305
+    },
+    {
+      "epoch": 1.8423576423576424,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0002,
+      "loss": 0.9064,
+      "step": 2306
+    },
+    {
+      "epoch": 1.843156843156843,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9196,
+      "step": 2307
+    },
+    {
+      "epoch": 1.843956043956044,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0002,
+      "loss": 0.9178,
+      "step": 2308
+    },
+    {
+      "epoch": 1.8447552447552447,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9155,
+      "step": 2309
+    },
+    {
+      "epoch": 1.8455544455544457,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9167,
+      "step": 2310
+    },
+    {
+      "epoch": 1.8463536463536463,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0002,
+      "loss": 0.9254,
+      "step": 2311
+    },
+    {
+      "epoch": 1.847152847152847,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0002,
+      "loss": 0.9177,
+      "step": 2312
+    },
+    {
+      "epoch": 1.847952047952048,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.0002,
+      "loss": 0.9155,
+      "step": 2313
+    },
+    {
+      "epoch": 1.8487512487512487,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9262,
+      "step": 2314
+    },
+    {
+      "epoch": 1.8495504495504496,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9181,
+      "step": 2315
+    },
+    {
+      "epoch": 1.8503496503496504,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0002,
+      "loss": 0.914,
+      "step": 2316
+    },
+    {
+      "epoch": 1.8511488511488512,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9203,
+      "step": 2317
+    },
+    {
+      "epoch": 1.8519480519480518,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9246,
+      "step": 2318
+    },
+    {
+      "epoch": 1.8527472527472528,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.0002,
+      "loss": 0.906,
+      "step": 2319
+    },
+    {
+      "epoch": 1.8535464535464534,
+      "grad_norm": 0.86328125,
+      "learning_rate": 0.0002,
+      "loss": 0.9195,
+      "step": 2320
+    },
+    {
+      "epoch": 1.8543456543456545,
+      "grad_norm": 1.0,
+      "learning_rate": 0.0002,
+      "loss": 0.9177,
+      "step": 2321
+    },
+    {
+      "epoch": 1.855144855144855,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9228,
+      "step": 2322
+    },
+    {
+      "epoch": 1.855944055944056,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.0002,
+      "loss": 0.9055,
+      "step": 2323
+    },
+    {
+      "epoch": 1.8567432567432567,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0002,
+      "loss": 0.9074,
+      "step": 2324
+    },
+    {
+      "epoch": 1.8575424575424575,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.0002,
+      "loss": 0.9199,
+      "step": 2325
+    },
+    {
+      "epoch": 1.8583416583416583,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.0002,
+      "loss": 0.9168,
+      "step": 2326
+    },
+    {
+      "epoch": 1.8591408591408591,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.0002,
+      "loss": 0.9084,
+      "step": 2327
+    },
+    {
+      "epoch": 1.85994005994006,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9084,
+      "step": 2328
+    },
+    {
+      "epoch": 1.8607392607392608,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.0002,
+      "loss": 0.9214,
+      "step": 2329
+    },
+    {
+      "epoch": 1.8615384615384616,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9176,
+      "step": 2330
+    },
+    {
+      "epoch": 1.8623376623376622,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0002,
+      "loss": 0.9211,
+      "step": 2331
+    },
+    {
+      "epoch": 1.8631368631368632,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.0002,
+      "loss": 0.9171,
+      "step": 2332
+    },
+    {
+      "epoch": 1.8639360639360638,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.0002,
+      "loss": 0.9271,
+      "step": 2333
+    },
+    {
+      "epoch": 1.8647352647352649,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9219,
+      "step": 2334
+    },
+    {
+      "epoch": 1.8655344655344654,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9152,
+      "step": 2335
+    },
+    {
+      "epoch": 1.8663336663336665,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9042,
+      "step": 2336
+    },
+    {
+      "epoch": 1.867132867132867,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9204,
+      "step": 2337
+    },
+    {
+      "epoch": 1.867932067932068,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9074,
+      "step": 2338
+    },
+    {
+      "epoch": 1.8687312687312687,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.0002,
+      "loss": 0.9163,
+      "step": 2339
+    },
+    {
+      "epoch": 1.8695304695304695,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0002,
+      "loss": 0.9173,
+      "step": 2340
+    },
+    {
+      "epoch": 1.8703296703296703,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9113,
+      "step": 2341
+    },
+    {
+      "epoch": 1.8711288711288712,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.0002,
+      "loss": 0.9096,
+      "step": 2342
+    },
+    {
+      "epoch": 1.871928071928072,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0002,
+      "loss": 0.908,
+      "step": 2343
+    },
+    {
+      "epoch": 1.8727272727272726,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0002,
+      "loss": 0.9179,
+      "step": 2344
+    },
+    {
+      "epoch": 1.8735264735264736,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9287,
+      "step": 2345
+    },
+    {
+      "epoch": 1.8743256743256742,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0002,
+      "loss": 0.9157,
+      "step": 2346
+    },
+    {
+      "epoch": 1.8751248751248752,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9149,
+      "step": 2347
+    },
+    {
+      "epoch": 1.8759240759240758,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9091,
+      "step": 2348
+    },
+    {
+      "epoch": 1.8767232767232769,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0002,
+      "loss": 0.9179,
+      "step": 2349
+    },
+    {
+      "epoch": 1.8775224775224775,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0002,
+      "loss": 0.9188,
+      "step": 2350
+    },
+    {
+      "epoch": 1.8783216783216783,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9053,
+      "step": 2351
+    },
+    {
+      "epoch": 1.879120879120879,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0002,
+      "loss": 0.9338,
+      "step": 2352
+    },
+    {
+      "epoch": 1.87992007992008,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9225,
+      "step": 2353
+    },
+    {
+      "epoch": 1.8807192807192807,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0002,
+      "loss": 0.9142,
+      "step": 2354
+    },
+    {
+      "epoch": 1.8815184815184816,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.0002,
+      "loss": 0.9066,
+      "step": 2355
+    },
+    {
+      "epoch": 1.8823176823176824,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0002,
+      "loss": 0.9149,
+      "step": 2356
+    },
+    {
+      "epoch": 1.883116883116883,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9206,
+      "step": 2357
+    },
+    {
+      "epoch": 1.883916083916084,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9196,
+      "step": 2358
+    },
+    {
+      "epoch": 1.8847152847152846,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9222,
+      "step": 2359
+    },
+    {
+      "epoch": 1.8855144855144856,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.0002,
+      "loss": 0.912,
+      "step": 2360
+    },
+    {
+      "epoch": 1.8863136863136862,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0002,
+      "loss": 0.9252,
+      "step": 2361
+    },
+    {
+      "epoch": 1.8871128871128873,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9111,
+      "step": 2362
+    },
+    {
+      "epoch": 1.8879120879120879,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0002,
+      "loss": 0.9193,
+      "step": 2363
+    },
+    {
+      "epoch": 1.8887112887112887,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0002,
+      "loss": 0.9199,
+      "step": 2364
+    },
+    {
+      "epoch": 1.8895104895104895,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0002,
+      "loss": 0.9171,
+      "step": 2365
+    },
+    {
+      "epoch": 1.8903096903096903,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0002,
+      "loss": 0.9175,
+      "step": 2366
+    },
+    {
+      "epoch": 1.8911088911088911,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.0002,
+      "loss": 0.9113,
+      "step": 2367
+    },
+    {
+      "epoch": 1.891908091908092,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0002,
+      "loss": 0.9033,
+      "step": 2368
+    },
+    {
+      "epoch": 1.8927072927072928,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0002,
+      "loss": 0.9142,
+      "step": 2369
+    },
+    {
+      "epoch": 1.8935064935064934,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0002,
+      "loss": 0.916,
+      "step": 2370
+    },
+    {
+      "epoch": 1.8943056943056944,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0002,
+      "loss": 0.9111,
+      "step": 2371
+    },
+    {
+      "epoch": 1.895104895104895,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0002,
+      "loss": 0.9101,
+      "step": 2372
+    },
+    {
+      "epoch": 1.895904095904096,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9143,
+      "step": 2373
+    },
+    {
+      "epoch": 1.8967032967032966,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.0002,
+      "loss": 0.9194,
+      "step": 2374
+    },
+    {
+      "epoch": 1.8975024975024977,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9227,
+      "step": 2375
+    },
+    {
+      "epoch": 1.8983016983016983,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.0002,
+      "loss": 0.9157,
+      "step": 2376
+    },
+    {
+      "epoch": 1.899100899100899,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0002,
+      "loss": 0.9164,
+      "step": 2377
+    },
+    {
+      "epoch": 1.8999000999001,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9172,
+      "step": 2378
+    },
+    {
+      "epoch": 1.9006993006993007,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9113,
+      "step": 2379
+    },
+    {
+      "epoch": 1.9014985014985015,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0002,
+      "loss": 0.918,
+      "step": 2380
+    },
+    {
+      "epoch": 1.9022977022977023,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.0002,
+      "loss": 0.9089,
+      "step": 2381
+    },
+    {
+      "epoch": 1.9030969030969032,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9177,
+      "step": 2382
+    },
+    {
+      "epoch": 1.9038961038961038,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0002,
+      "loss": 0.9076,
+      "step": 2383
+    },
+    {
+      "epoch": 1.9046953046953048,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0002,
+      "loss": 0.9081,
+      "step": 2384
+    },
+    {
+      "epoch": 1.9054945054945054,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9138,
+      "step": 2385
+    },
+    {
+      "epoch": 1.9062937062937064,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9166,
+      "step": 2386
+    },
+    {
+      "epoch": 1.907092907092907,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.0002,
+      "loss": 0.9135,
+      "step": 2387
+    },
+    {
+      "epoch": 1.907892107892108,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0002,
+      "loss": 0.9176,
+      "step": 2388
+    },
+    {
+      "epoch": 1.9086913086913087,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0002,
+      "loss": 0.9208,
+      "step": 2389
+    },
+    {
+      "epoch": 1.9094905094905095,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0002,
+      "loss": 0.9184,
+      "step": 2390
+    },
+    {
+      "epoch": 1.9102897102897103,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.0002,
+      "loss": 0.9243,
+      "step": 2391
+    },
+    {
+      "epoch": 1.911088911088911,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0002,
+      "loss": 0.9204,
+      "step": 2392
+    },
+    {
+      "epoch": 1.911888111888112,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0002,
+      "loss": 0.9325,
+      "step": 2393
+    },
+    {
+      "epoch": 1.9126873126873127,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0002,
+      "loss": 0.9142,
+      "step": 2394
+    },
+    {
+      "epoch": 1.9134865134865136,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.0002,
+      "loss": 0.9116,
+      "step": 2395
+    },
+    {
+      "epoch": 1.9142857142857141,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9049,
+      "step": 2396
+    },
+    {
+      "epoch": 1.9150849150849152,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9124,
+      "step": 2397
+    },
+    {
+      "epoch": 1.9158841158841158,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0002,
+      "loss": 0.9145,
+      "step": 2398
+    },
+    {
+      "epoch": 1.9166833166833168,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0002,
+      "loss": 0.9229,
+      "step": 2399
+    },
+    {
+      "epoch": 1.9174825174825174,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9168,
+      "step": 2400
+    },
+    {
+      "epoch": 1.9182817182817182,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9183,
+      "step": 2401
+    },
+    {
+      "epoch": 1.919080919080919,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0002,
+      "loss": 0.9093,
+      "step": 2402
+    },
+    {
+      "epoch": 1.9198801198801199,
+      "grad_norm": 0.375,
+      "learning_rate": 0.0002,
+      "loss": 0.9113,
+      "step": 2403
+    },
+    {
+      "epoch": 1.9206793206793207,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0002,
+      "loss": 0.917,
+      "step": 2404
+    },
+    {
+      "epoch": 1.9214785214785215,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9155,
+      "step": 2405
+    },
+    {
+      "epoch": 1.9222777222777223,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0002,
+      "loss": 0.906,
+      "step": 2406
+    },
+    {
+      "epoch": 1.9230769230769231,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.0002,
+      "loss": 0.9196,
+      "step": 2407
+    },
+    {
+      "epoch": 1.923876123876124,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9253,
+      "step": 2408
+    },
+    {
+      "epoch": 1.9246753246753245,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9087,
+      "step": 2409
+    },
+    {
+      "epoch": 1.9254745254745256,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0002,
+      "loss": 0.9149,
+      "step": 2410
+    },
+    {
+      "epoch": 1.9262737262737262,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0002,
+      "loss": 0.9184,
+      "step": 2411
+    },
+    {
+      "epoch": 1.9270729270729272,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0002,
+      "loss": 0.9213,
+      "step": 2412
+    },
+    {
+      "epoch": 1.9278721278721278,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.0002,
+      "loss": 0.922,
+      "step": 2413
+    },
+    {
+      "epoch": 1.9286713286713286,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0002,
+      "loss": 0.9161,
+      "step": 2414
+    },
+    {
+      "epoch": 1.9294705294705294,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9291,
+      "step": 2415
+    },
+    {
+      "epoch": 1.9302697302697303,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0002,
+      "loss": 0.9139,
+      "step": 2416
+    },
+    {
+      "epoch": 1.931068931068931,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.0002,
+      "loss": 0.9146,
+      "step": 2417
+    },
+    {
+      "epoch": 1.9318681318681319,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9148,
+      "step": 2418
+    },
+    {
+      "epoch": 1.9326673326673327,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0002,
+      "loss": 0.9149,
+      "step": 2419
+    },
+    {
+      "epoch": 1.9334665334665335,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9112,
+      "step": 2420
+    },
+    {
+      "epoch": 1.9342657342657343,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9182,
+      "step": 2421
+    },
+    {
+      "epoch": 1.935064935064935,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0002,
+      "loss": 0.9151,
+      "step": 2422
+    },
+    {
+      "epoch": 1.935864135864136,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0002,
+      "loss": 0.9173,
+      "step": 2423
+    },
+    {
+      "epoch": 1.9366633366633366,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0002,
+      "loss": 0.9118,
+      "step": 2424
+    },
+    {
+      "epoch": 1.9374625374625376,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9149,
+      "step": 2425
+    },
+    {
+      "epoch": 1.9382617382617382,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.0002,
+      "loss": 0.9182,
+      "step": 2426
+    },
+    {
+      "epoch": 1.939060939060939,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9071,
+      "step": 2427
+    },
+    {
+      "epoch": 1.9398601398601398,
+      "grad_norm": 1.0,
+      "learning_rate": 0.0002,
+      "loss": 0.9355,
+      "step": 2428
+    },
+    {
+      "epoch": 1.9406593406593406,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0002,
+      "loss": 0.9073,
+      "step": 2429
+    },
+    {
+      "epoch": 1.9414585414585415,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0002,
+      "loss": 0.96,
+      "step": 2430
+    },
+    {
+      "epoch": 1.9422577422577423,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9152,
+      "step": 2431
+    },
+    {
+      "epoch": 1.943056943056943,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0002,
+      "loss": 0.918,
+      "step": 2432
+    },
+    {
+      "epoch": 1.9438561438561437,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9236,
+      "step": 2433
+    },
+    {
+      "epoch": 1.9446553446553447,
+      "grad_norm": 0.375,
+      "learning_rate": 0.0002,
+      "loss": 0.9135,
+      "step": 2434
+    },
+    {
+      "epoch": 1.9454545454545453,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0002,
+      "loss": 0.921,
+      "step": 2435
+    },
+    {
+      "epoch": 1.9462537462537464,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.0002,
+      "loss": 0.9146,
+      "step": 2436
+    },
+    {
+      "epoch": 1.947052947052947,
+      "grad_norm": 0.625,
+      "learning_rate": 0.0002,
+      "loss": 0.9241,
+      "step": 2437
+    },
+    {
+      "epoch": 1.947852147852148,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0002,
+      "loss": 0.914,
+      "step": 2438
+    },
+    {
+      "epoch": 1.9486513486513486,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9217,
+      "step": 2439
+    },
+    {
+      "epoch": 1.9494505494505494,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0002,
+      "loss": 0.9269,
+      "step": 2440
+    },
+    {
+      "epoch": 1.9502497502497502,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.0002,
+      "loss": 0.9107,
+      "step": 2441
+    },
+    {
+      "epoch": 1.951048951048951,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0002,
+      "loss": 0.9136,
+      "step": 2442
+    },
+    {
+      "epoch": 1.9518481518481519,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0002,
+      "loss": 0.9096,
+      "step": 2443
+    },
+    {
+      "epoch": 1.9526473526473527,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0002,
+      "loss": 0.9154,
+      "step": 2444
+    },
+    {
+      "epoch": 1.9534465534465535,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0002,
+      "loss": 0.9191,
+      "step": 2445
+    },
+    {
+      "epoch": 1.954245754245754,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.0002,
+      "loss": 0.9145,
+      "step": 2446
+    },
+    {
+      "epoch": 1.9550449550449551,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0002,
+      "loss": 0.909,
+      "step": 2447
+    },
+    {
+      "epoch": 1.9558441558441557,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.0002,
+      "loss": 0.9231,
+      "step": 2448
+    },
+    {
+      "epoch": 1.9566433566433568,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9048,
+      "step": 2449
+    },
+    {
+      "epoch": 1.9574425574425574,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0002,
+      "loss": 0.9283,
+      "step": 2450
+    },
+    {
+      "epoch": 1.9582417582417584,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0002,
+      "loss": 0.9119,
+      "step": 2451
+    },
+    {
+      "epoch": 1.959040959040959,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0002,
+      "loss": 0.9235,
+      "step": 2452
+    },
+    {
+      "epoch": 1.9598401598401598,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0002,
+      "loss": 0.9144,
+      "step": 2453
+    },
+    {
+      "epoch": 1.9606393606393606,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0002,
+      "loss": 0.9185,
+      "step": 2454
+    },
+    {
+      "epoch": 1.9614385614385614,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0002,
+      "loss": 0.9213,
+      "step": 2455
+    },
+    {
+      "epoch": 1.9622377622377623,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0002,
+      "loss": 0.9171,
+      "step": 2456
+    },
+    {
+      "epoch": 1.963036963036963,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0002,
+      "loss": 0.911,
+      "step": 2457
+    },
+    {
+      "epoch": 1.9638361638361639,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0002,
+      "loss": 0.9143,
+      "step": 2458
+    },
+    {
+      "epoch": 1.9646353646353645,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0002,
+      "loss": 0.9083,
+      "step": 2459
+    },
+    {
+      "epoch": 1.9654345654345655,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0002,
+      "loss": 0.9116,
+      "step": 2460
+    },
+    {
+      "epoch": 1.9662337662337661,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9087,
+      "step": 2461
+    },
+    {
+      "epoch": 1.9670329670329672,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0002,
+      "loss": 0.907,
+      "step": 2462
+    },
+    {
+      "epoch": 1.9678321678321677,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9088,
+      "step": 2463
+    },
+    {
+      "epoch": 1.9686313686313688,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9171,
+      "step": 2464
+    },
+    {
+      "epoch": 1.9694305694305694,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0002,
+      "loss": 0.9129,
+      "step": 2465
+    },
+    {
+      "epoch": 1.9702297702297702,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0002,
+      "loss": 0.9137,
+      "step": 2466
+    },
+    {
+      "epoch": 1.971028971028971,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.0002,
+      "loss": 0.9062,
+      "step": 2467
+    },
+    {
+      "epoch": 1.9718281718281718,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0002,
+      "loss": 0.9114,
+      "step": 2468
+    },
+    {
+      "epoch": 1.9726273726273726,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0002,
+      "loss": 0.918,
+      "step": 2469
+    },
+    {
+      "epoch": 1.9734265734265735,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0002,
+      "loss": 0.914,
+      "step": 2470
+    },
+    {
+      "epoch": 1.9742257742257743,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0002,
+      "loss": 0.927,
+      "step": 2471
+    },
+    {
+      "epoch": 1.9750249750249749,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9089,
+      "step": 2472
+    },
+    {
+      "epoch": 1.975824175824176,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0002,
+      "loss": 0.9116,
+      "step": 2473
+    },
+    {
+      "epoch": 1.9766233766233765,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9225,
+      "step": 2474
+    },
+    {
+      "epoch": 1.9774225774225775,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0002,
+      "loss": 0.9131,
+      "step": 2475
+    },
+    {
+      "epoch": 1.9782217782217781,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9065,
+      "step": 2476
+    },
+    {
+      "epoch": 1.9790209790209792,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9139,
+      "step": 2477
+    },
+    {
+      "epoch": 1.9798201798201798,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0002,
+      "loss": 0.9127,
+      "step": 2478
+    },
+    {
+      "epoch": 1.9806193806193806,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.0002,
+      "loss": 0.9058,
+      "step": 2479
+    },
+    {
+      "epoch": 1.9814185814185814,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.0002,
+      "loss": 0.9041,
+      "step": 2480
+    },
+    {
+      "epoch": 1.9822177822177822,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0002,
+      "loss": 0.915,
+      "step": 2481
+    },
+    {
+      "epoch": 1.983016983016983,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0002,
+      "loss": 0.9071,
+      "step": 2482
+    },
+    {
+      "epoch": 1.9838161838161839,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9132,
+      "step": 2483
+    },
+    {
+      "epoch": 1.9846153846153847,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.0002,
+      "loss": 0.9317,
+      "step": 2484
+    },
+    {
+      "epoch": 1.9854145854145853,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.0002,
+      "loss": 0.9105,
+      "step": 2485
+    },
+    {
+      "epoch": 1.9862137862137863,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0002,
+      "loss": 0.9183,
+      "step": 2486
+    },
+    {
+      "epoch": 1.987012987012987,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0002,
+      "loss": 0.9167,
+      "step": 2487
+    },
+    {
+      "epoch": 1.987812187812188,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.0002,
+      "loss": 0.9174,
+      "step": 2488
+    },
+    {
+      "epoch": 1.9886113886113885,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.0002,
+      "loss": 0.9136,
+      "step": 2489
+    },
+    {
+      "epoch": 1.9894105894105896,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.0002,
+      "loss": 0.915,
+      "step": 2490
+    },
+    {
+      "epoch": 1.9902097902097902,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9201,
+      "step": 2491
+    },
+    {
+      "epoch": 1.991008991008991,
+      "grad_norm": 0.375,
+      "learning_rate": 0.0002,
+      "loss": 0.9091,
+      "step": 2492
+    },
+    {
+      "epoch": 1.9918081918081918,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9112,
+      "step": 2493
+    },
+    {
+      "epoch": 1.9926073926073926,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.0002,
+      "loss": 0.9076,
+      "step": 2494
+    },
+    {
+      "epoch": 1.9934065934065934,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9117,
+      "step": 2495
+    },
+    {
+      "epoch": 1.9942057942057942,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0002,
+      "loss": 0.9102,
+      "step": 2496
+    },
+    {
+      "epoch": 1.995004995004995,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0002,
+      "loss": 0.9106,
+      "step": 2497
+    },
+    {
+      "epoch": 1.9958041958041957,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9085,
+      "step": 2498
+    },
+    {
+      "epoch": 1.9966033966033967,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0002,
+      "loss": 0.9089,
+      "step": 2499
+    },
+    {
+      "epoch": 1.9974025974025973,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0002,
+      "loss": 0.9245,
+      "step": 2500
+    },
+    {
+      "epoch": 1.9982017982017983,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9193,
+      "step": 2501
+    },
+    {
+      "epoch": 1.999000999000999,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.0002,
+      "loss": 0.927,
+      "step": 2502
+    },
+    {
+      "epoch": 1.9998001998002,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.0002,
+      "loss": 0.9108,
+      "step": 2503
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 0.2242,
+      "step": 2504
+    },
+    {
+      "epoch": 2.0007992007992006,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.0002,
+      "loss": 0.9162,
+      "step": 2505
+    },
+    {
+      "epoch": 2.0015984015984016,
+      "grad_norm": 0.890625,
+      "learning_rate": 0.0002,
+      "loss": 0.9171,
+      "step": 2506
+    },
+    {
+      "epoch": 2.0023976023976022,
+      "grad_norm": 1.4765625,
+      "learning_rate": 0.0002,
+      "loss": 0.9137,
+      "step": 2507
+    },
+    {
+      "epoch": 2.0031968031968033,
+      "grad_norm": 1.0,
+      "learning_rate": 0.0002,
+      "loss": 0.9228,
+      "step": 2508
+    },
+    {
+      "epoch": 2.003996003996004,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0002,
+      "loss": 0.916,
+      "step": 2509
+    },
+    {
+      "epoch": 2.004795204795205,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.0002,
+      "loss": 0.9169,
+      "step": 2510
+    },
+    {
+      "epoch": 2.0055944055944055,
+      "grad_norm": 0.97265625,
+      "learning_rate": 0.0002,
+      "loss": 0.9162,
+      "step": 2511
+    },
+    {
+      "epoch": 2.0063936063936065,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.0002,
+      "loss": 0.911,
+      "step": 2512
+    },
+    {
+      "epoch": 2.007192807192807,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0002,
+      "loss": 0.9013,
+      "step": 2513
+    },
+    {
+      "epoch": 2.007992007992008,
+      "grad_norm": 0.890625,
+      "learning_rate": 0.0002,
+      "loss": 0.9199,
+      "step": 2514
+    },
+    {
+      "epoch": 2.0087912087912088,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.0002,
+      "loss": 0.9166,
+      "step": 2515
+    },
+    {
+      "epoch": 2.00959040959041,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9068,
+      "step": 2516
+    },
+    {
+      "epoch": 2.0103896103896104,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0002,
+      "loss": 0.917,
+      "step": 2517
+    },
+    {
+      "epoch": 2.011188811188811,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0002,
+      "loss": 0.9166,
+      "step": 2518
+    },
+    {
+      "epoch": 2.011988011988012,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.0002,
+      "loss": 0.912,
+      "step": 2519
+    },
+    {
+      "epoch": 2.0127872127872126,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0002,
+      "loss": 0.915,
+      "step": 2520
+    },
+    {
+      "epoch": 2.0135864135864137,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9128,
+      "step": 2521
+    },
+    {
+      "epoch": 2.0143856143856143,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.0002,
+      "loss": 0.92,
+      "step": 2522
+    },
+    {
+      "epoch": 2.0151848151848153,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0002,
+      "loss": 0.9187,
+      "step": 2523
+    },
+    {
+      "epoch": 2.015984015984016,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.0002,
+      "loss": 0.9106,
+      "step": 2524
+    },
+    {
+      "epoch": 2.016783216783217,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0002,
+      "loss": 0.911,
+      "step": 2525
+    },
+    {
+      "epoch": 2.0175824175824175,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9118,
+      "step": 2526
+    },
+    {
+      "epoch": 2.0183816183816186,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.0002,
+      "loss": 0.9154,
+      "step": 2527
+    },
+    {
+      "epoch": 2.019180819180819,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9086,
+      "step": 2528
+    },
+    {
+      "epoch": 2.01998001998002,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9116,
+      "step": 2529
+    },
+    {
+      "epoch": 2.020779220779221,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9144,
+      "step": 2530
+    },
+    {
+      "epoch": 2.0215784215784214,
+      "grad_norm": 0.97265625,
+      "learning_rate": 0.0002,
+      "loss": 0.9415,
+      "step": 2531
+    },
+    {
+      "epoch": 2.0223776223776224,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0002,
+      "loss": 0.9127,
+      "step": 2532
+    },
+    {
+      "epoch": 2.023176823176823,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9086,
+      "step": 2533
+    },
+    {
+      "epoch": 2.023976023976024,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0002,
+      "loss": 0.9161,
+      "step": 2534
+    },
+    {
+      "epoch": 2.0247752247752246,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0002,
+      "loss": 0.9129,
+      "step": 2535
+    },
+    {
+      "epoch": 2.0255744255744257,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0002,
+      "loss": 0.9163,
+      "step": 2536
+    },
+    {
+      "epoch": 2.0263736263736263,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9087,
+      "step": 2537
+    },
+    {
+      "epoch": 2.0271728271728273,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0002,
+      "loss": 0.916,
+      "step": 2538
+    },
+    {
+      "epoch": 2.027972027972028,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.0002,
+      "loss": 0.9051,
+      "step": 2539
+    },
+    {
+      "epoch": 2.028771228771229,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0002,
+      "loss": 0.9164,
+      "step": 2540
+    },
+    {
+      "epoch": 2.0295704295704295,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0002,
+      "loss": 0.9171,
+      "step": 2541
+    },
+    {
+      "epoch": 2.0303696303696306,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9181,
+      "step": 2542
+    },
+    {
+      "epoch": 2.031168831168831,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9058,
+      "step": 2543
+    },
+    {
+      "epoch": 2.0319680319680318,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0002,
+      "loss": 0.9206,
+      "step": 2544
+    },
+    {
+      "epoch": 2.032767232767233,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9131,
+      "step": 2545
+    },
+    {
+      "epoch": 2.0335664335664334,
+      "grad_norm": 0.9609375,
+      "learning_rate": 0.0002,
+      "loss": 0.909,
+      "step": 2546
+    },
+    {
+      "epoch": 2.0343656343656344,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.0002,
+      "loss": 0.9118,
+      "step": 2547
+    },
+    {
+      "epoch": 2.035164835164835,
+      "grad_norm": 0.80859375,
+      "learning_rate": 0.0002,
+      "loss": 0.9119,
+      "step": 2548
+    },
+    {
+      "epoch": 2.035964035964036,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0002,
+      "loss": 0.9173,
+      "step": 2549
+    },
+    {
+      "epoch": 2.0367632367632367,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.0002,
+      "loss": 0.911,
+      "step": 2550
+    },
+    {
+      "epoch": 2.0375624375624377,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9193,
+      "step": 2551
+    },
+    {
+      "epoch": 2.0383616383616383,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9235,
+      "step": 2552
+    },
+    {
+      "epoch": 2.0391608391608393,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.0002,
+      "loss": 0.9088,
+      "step": 2553
+    },
+    {
+      "epoch": 2.03996003996004,
+      "grad_norm": 0.80859375,
+      "learning_rate": 0.0002,
+      "loss": 0.9073,
+      "step": 2554
+    },
+    {
+      "epoch": 2.040759240759241,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0002,
+      "loss": 0.925,
+      "step": 2555
+    },
+    {
+      "epoch": 2.0415584415584416,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9222,
+      "step": 2556
+    },
+    {
+      "epoch": 2.042357642357642,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0002,
+      "loss": 0.9229,
+      "step": 2557
+    },
+    {
+      "epoch": 2.043156843156843,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9086,
+      "step": 2558
+    },
+    {
+      "epoch": 2.043956043956044,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.0002,
+      "loss": 0.9281,
+      "step": 2559
+    },
+    {
+      "epoch": 2.044755244755245,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9139,
+      "step": 2560
+    },
+    {
+      "epoch": 2.0455544455544454,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.0002,
+      "loss": 0.9178,
+      "step": 2561
+    },
+    {
+      "epoch": 2.0463536463536465,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.0002,
+      "loss": 0.9324,
+      "step": 2562
+    },
+    {
+      "epoch": 2.047152847152847,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0002,
+      "loss": 0.91,
+      "step": 2563
+    },
+    {
+      "epoch": 2.047952047952048,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0002,
+      "loss": 0.916,
+      "step": 2564
+    },
+    {
+      "epoch": 2.0487512487512487,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9153,
+      "step": 2565
+    },
+    {
+      "epoch": 2.0495504495504497,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0002,
+      "loss": 0.9044,
+      "step": 2566
+    },
+    {
+      "epoch": 2.0503496503496503,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0002,
+      "loss": 0.9116,
+      "step": 2567
+    },
+    {
+      "epoch": 2.0511488511488514,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9104,
+      "step": 2568
+    },
+    {
+      "epoch": 2.051948051948052,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.0002,
+      "loss": 0.9215,
+      "step": 2569
+    },
+    {
+      "epoch": 2.0527472527472526,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9173,
+      "step": 2570
+    },
+    {
+      "epoch": 2.0535464535464536,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.0002,
+      "loss": 0.9154,
+      "step": 2571
+    },
+    {
+      "epoch": 2.054345654345654,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9054,
+      "step": 2572
+    },
+    {
+      "epoch": 2.0551448551448552,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9156,
+      "step": 2573
+    },
+    {
+      "epoch": 2.055944055944056,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0002,
+      "loss": 0.9091,
+      "step": 2574
+    },
+    {
+      "epoch": 2.056743256743257,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0002,
+      "loss": 0.918,
+      "step": 2575
+    },
+    {
+      "epoch": 2.0575424575424575,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0002,
+      "loss": 0.922,
+      "step": 2576
+    },
+    {
+      "epoch": 2.0583416583416585,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9001,
+      "step": 2577
+    },
+    {
+      "epoch": 2.059140859140859,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0002,
+      "loss": 0.9071,
+      "step": 2578
+    },
+    {
+      "epoch": 2.05994005994006,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0002,
+      "loss": 0.9163,
+      "step": 2579
+    },
+    {
+      "epoch": 2.0607392607392607,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9086,
+      "step": 2580
+    },
+    {
+      "epoch": 2.0615384615384613,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0002,
+      "loss": 0.9139,
+      "step": 2581
+    },
+    {
+      "epoch": 2.0623376623376624,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.0002,
+      "loss": 0.9257,
+      "step": 2582
+    },
+    {
+      "epoch": 2.063136863136863,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.0002,
+      "loss": 0.911,
+      "step": 2583
+    },
+    {
+      "epoch": 2.063936063936064,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9133,
+      "step": 2584
+    },
+    {
+      "epoch": 2.0647352647352646,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9157,
+      "step": 2585
+    },
+    {
+      "epoch": 2.0655344655344656,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.0002,
+      "loss": 0.9087,
+      "step": 2586
+    },
+    {
+      "epoch": 2.066333666333666,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0002,
+      "loss": 0.9143,
+      "step": 2587
+    },
+    {
+      "epoch": 2.0671328671328673,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.0002,
+      "loss": 0.9142,
+      "step": 2588
+    },
+    {
+      "epoch": 2.067932067932068,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0002,
+      "loss": 0.9135,
+      "step": 2589
+    },
+    {
+      "epoch": 2.068731268731269,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9017,
+      "step": 2590
+    },
+    {
+      "epoch": 2.0695304695304695,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0002,
+      "loss": 0.9127,
+      "step": 2591
+    },
+    {
+      "epoch": 2.0703296703296705,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9144,
+      "step": 2592
+    },
+    {
+      "epoch": 2.071128871128871,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9094,
+      "step": 2593
+    },
+    {
+      "epoch": 2.071928071928072,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0002,
+      "loss": 0.9206,
+      "step": 2594
+    },
+    {
+      "epoch": 2.0727272727272728,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0002,
+      "loss": 0.9097,
+      "step": 2595
+    },
+    {
+      "epoch": 2.0735264735264733,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0002,
+      "loss": 0.9169,
+      "step": 2596
+    },
+    {
+      "epoch": 2.0743256743256744,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.0002,
+      "loss": 0.9168,
+      "step": 2597
+    },
+    {
+      "epoch": 2.075124875124875,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0002,
+      "loss": 0.9196,
+      "step": 2598
+    },
+    {
+      "epoch": 2.075924075924076,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0002,
+      "loss": 0.9072,
+      "step": 2599
+    },
+    {
+      "epoch": 2.0767232767232766,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.0002,
+      "loss": 0.9058,
+      "step": 2600
+    },
+    {
+      "epoch": 2.0775224775224777,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9152,
+      "step": 2601
+    },
+    {
+      "epoch": 2.0783216783216782,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9124,
+      "step": 2602
+    },
+    {
+      "epoch": 2.0791208791208793,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9151,
+      "step": 2603
+    },
+    {
+      "epoch": 2.07992007992008,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.0002,
+      "loss": 0.9215,
+      "step": 2604
+    },
+    {
+      "epoch": 2.080719280719281,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.0002,
+      "loss": 0.9089,
+      "step": 2605
+    },
+    {
+      "epoch": 2.0815184815184815,
+      "grad_norm": 0.99609375,
+      "learning_rate": 0.0002,
+      "loss": 0.9126,
+      "step": 2606
+    },
+    {
+      "epoch": 2.082317682317682,
+      "grad_norm": 0.9609375,
+      "learning_rate": 0.0002,
+      "loss": 0.9101,
+      "step": 2607
+    },
+    {
+      "epoch": 2.083116883116883,
+      "grad_norm": 0.625,
+      "learning_rate": 0.0002,
+      "loss": 0.9052,
+      "step": 2608
+    },
+    {
+      "epoch": 2.0839160839160837,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0002,
+      "loss": 0.915,
+      "step": 2609
+    },
+    {
+      "epoch": 2.0847152847152848,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.0002,
+      "loss": 0.9135,
+      "step": 2610
+    },
+    {
+      "epoch": 2.0855144855144854,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0002,
+      "loss": 0.9051,
+      "step": 2611
+    },
+    {
+      "epoch": 2.0863136863136864,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.0002,
+      "loss": 0.9173,
+      "step": 2612
+    },
+    {
+      "epoch": 2.087112887112887,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.0002,
+      "loss": 0.906,
+      "step": 2613
+    },
+    {
+      "epoch": 2.087912087912088,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0002,
+      "loss": 0.9163,
+      "step": 2614
+    },
+    {
+      "epoch": 2.0887112887112886,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9117,
+      "step": 2615
+    },
+    {
+      "epoch": 2.0895104895104897,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0002,
+      "loss": 0.9136,
+      "step": 2616
+    },
+    {
+      "epoch": 2.0903096903096903,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9134,
+      "step": 2617
+    },
+    {
+      "epoch": 2.0911088911088913,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9104,
+      "step": 2618
+    },
+    {
+      "epoch": 2.091908091908092,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9051,
+      "step": 2619
+    },
+    {
+      "epoch": 2.0927072927072925,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0002,
+      "loss": 0.912,
+      "step": 2620
+    },
+    {
+      "epoch": 2.0935064935064935,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9143,
+      "step": 2621
+    },
+    {
+      "epoch": 2.094305694305694,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9175,
+      "step": 2622
+    },
+    {
+      "epoch": 2.095104895104895,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0002,
+      "loss": 0.9071,
+      "step": 2623
+    },
+    {
+      "epoch": 2.0959040959040958,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0002,
+      "loss": 0.9049,
+      "step": 2624
+    },
+    {
+      "epoch": 2.096703296703297,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0002,
+      "loss": 0.9112,
+      "step": 2625
+    },
+    {
+      "epoch": 2.0975024975024974,
+      "grad_norm": 0.890625,
+      "learning_rate": 0.0002,
+      "loss": 0.9425,
+      "step": 2626
+    },
+    {
+      "epoch": 2.0983016983016984,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0002,
+      "loss": 0.9184,
+      "step": 2627
+    },
+    {
+      "epoch": 2.099100899100899,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9133,
+      "step": 2628
+    },
+    {
+      "epoch": 2.0999000999001,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.0002,
+      "loss": 0.9163,
+      "step": 2629
+    },
+    {
+      "epoch": 2.1006993006993007,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0002,
+      "loss": 0.9135,
+      "step": 2630
+    },
+    {
+      "epoch": 2.1014985014985017,
+      "grad_norm": 0.375,
+      "learning_rate": 0.0002,
+      "loss": 0.9044,
+      "step": 2631
+    },
+    {
+      "epoch": 2.1022977022977023,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0002,
+      "loss": 0.9183,
+      "step": 2632
+    },
+    {
+      "epoch": 2.103096903096903,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0002,
+      "loss": 0.9102,
+      "step": 2633
+    },
+    {
+      "epoch": 2.103896103896104,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0002,
+      "loss": 0.9085,
+      "step": 2634
+    },
+    {
+      "epoch": 2.1046953046953045,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9049,
+      "step": 2635
+    },
+    {
+      "epoch": 2.1054945054945056,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9074,
+      "step": 2636
+    },
+    {
+      "epoch": 2.106293706293706,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9115,
+      "step": 2637
+    },
+    {
+      "epoch": 2.107092907092907,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0002,
+      "loss": 0.9116,
+      "step": 2638
+    },
+    {
+      "epoch": 2.107892107892108,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0002,
+      "loss": 0.9091,
+      "step": 2639
+    },
+    {
+      "epoch": 2.108691308691309,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0002,
+      "loss": 0.9087,
+      "step": 2640
+    },
+    {
+      "epoch": 2.1094905094905094,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0002,
+      "loss": 0.9138,
+      "step": 2641
+    },
+    {
+      "epoch": 2.1102897102897105,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.0002,
+      "loss": 0.9162,
+      "step": 2642
+    },
+    {
+      "epoch": 2.111088911088911,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9078,
+      "step": 2643
+    },
+    {
+      "epoch": 2.111888111888112,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9176,
+      "step": 2644
+    },
+    {
+      "epoch": 2.1126873126873127,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0002,
+      "loss": 0.9069,
+      "step": 2645
+    },
+    {
+      "epoch": 2.1134865134865133,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9139,
+      "step": 2646
+    },
+    {
+      "epoch": 2.1142857142857143,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9133,
+      "step": 2647
+    },
+    {
+      "epoch": 2.115084915084915,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0002,
+      "loss": 0.9049,
+      "step": 2648
+    },
+    {
+      "epoch": 2.115884115884116,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9094,
+      "step": 2649
+    },
+    {
+      "epoch": 2.1166833166833166,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0002,
+      "loss": 0.8976,
+      "step": 2650
+    },
+    {
+      "epoch": 2.1174825174825176,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.0002,
+      "loss": 0.9127,
+      "step": 2651
+    },
+    {
+      "epoch": 2.118281718281718,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0002,
+      "loss": 0.9119,
+      "step": 2652
+    },
+    {
+      "epoch": 2.1190809190809192,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9098,
+      "step": 2653
+    },
+    {
+      "epoch": 2.11988011988012,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0002,
+      "loss": 0.917,
+      "step": 2654
+    },
+    {
+      "epoch": 2.120679320679321,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.0002,
+      "loss": 0.9111,
+      "step": 2655
+    },
+    {
+      "epoch": 2.1214785214785215,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9166,
+      "step": 2656
+    },
+    {
+      "epoch": 2.1222777222777225,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.0002,
+      "loss": 0.9128,
+      "step": 2657
+    },
+    {
+      "epoch": 2.123076923076923,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9178,
+      "step": 2658
+    },
+    {
+      "epoch": 2.1238761238761237,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.0002,
+      "loss": 0.9148,
+      "step": 2659
+    },
+    {
+      "epoch": 2.1246753246753247,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9124,
+      "step": 2660
+    },
+    {
+      "epoch": 2.1254745254745253,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9133,
+      "step": 2661
+    },
+    {
+      "epoch": 2.1262737262737263,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0002,
+      "loss": 0.9131,
+      "step": 2662
+    },
+    {
+      "epoch": 2.127072927072927,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.0002,
+      "loss": 0.9115,
+      "step": 2663
+    },
+    {
+      "epoch": 2.127872127872128,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9055,
+      "step": 2664
+    },
+    {
+      "epoch": 2.1286713286713286,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0002,
+      "loss": 0.9083,
+      "step": 2665
+    },
+    {
+      "epoch": 2.1294705294705296,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9106,
+      "step": 2666
+    },
+    {
+      "epoch": 2.13026973026973,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.0002,
+      "loss": 0.9162,
+      "step": 2667
+    },
+    {
+      "epoch": 2.1310689310689312,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.0002,
+      "loss": 0.9131,
+      "step": 2668
+    },
+    {
+      "epoch": 2.131868131868132,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0002,
+      "loss": 0.9165,
+      "step": 2669
+    },
+    {
+      "epoch": 2.1326673326673324,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.0002,
+      "loss": 0.9121,
+      "step": 2670
+    },
+    {
+      "epoch": 2.1334665334665335,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.0002,
+      "loss": 0.9083,
+      "step": 2671
+    },
+    {
+      "epoch": 2.134265734265734,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9063,
+      "step": 2672
+    },
+    {
+      "epoch": 2.135064935064935,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0002,
+      "loss": 0.9168,
+      "step": 2673
+    },
+    {
+      "epoch": 2.1358641358641357,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0002,
+      "loss": 0.913,
+      "step": 2674
+    },
+    {
+      "epoch": 2.1366633366633367,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9068,
+      "step": 2675
+    },
+    {
+      "epoch": 2.1374625374625373,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.0002,
+      "loss": 0.9048,
+      "step": 2676
+    },
+    {
+      "epoch": 2.1382617382617384,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0002,
+      "loss": 0.9071,
+      "step": 2677
+    },
+    {
+      "epoch": 2.139060939060939,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0002,
+      "loss": 0.915,
+      "step": 2678
+    },
+    {
+      "epoch": 2.13986013986014,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.0002,
+      "loss": 0.9041,
+      "step": 2679
+    },
+    {
+      "epoch": 2.1406593406593406,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0002,
+      "loss": 0.9127,
+      "step": 2680
+    },
+    {
+      "epoch": 2.1414585414585416,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.0002,
+      "loss": 0.9079,
+      "step": 2681
+    },
+    {
+      "epoch": 2.1422577422577422,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.0002,
+      "loss": 0.917,
+      "step": 2682
+    },
+    {
+      "epoch": 2.1430569430569433,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9104,
+      "step": 2683
+    },
+    {
+      "epoch": 2.143856143856144,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9101,
+      "step": 2684
+    },
+    {
+      "epoch": 2.1446553446553445,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.0002,
+      "loss": 0.9028,
+      "step": 2685
+    },
+    {
+      "epoch": 2.1454545454545455,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9118,
+      "step": 2686
+    },
+    {
+      "epoch": 2.146253746253746,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.0002,
+      "loss": 0.9186,
+      "step": 2687
+    },
+    {
+      "epoch": 2.147052947052947,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.0002,
+      "loss": 0.9142,
+      "step": 2688
+    },
+    {
+      "epoch": 2.1478521478521477,
+      "grad_norm": 0.8828125,
+      "learning_rate": 0.0002,
+      "loss": 0.9143,
+      "step": 2689
+    },
+    {
+      "epoch": 2.1486513486513488,
+      "grad_norm": 0.92578125,
+      "learning_rate": 0.0002,
+      "loss": 0.9041,
+      "step": 2690
+    },
+    {
+      "epoch": 2.1494505494505494,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9113,
+      "step": 2691
+    },
+    {
+      "epoch": 2.1502497502497504,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0002,
+      "loss": 0.9097,
+      "step": 2692
+    },
+    {
+      "epoch": 2.151048951048951,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.0002,
+      "loss": 0.9214,
+      "step": 2693
+    },
+    {
+      "epoch": 2.151848151848152,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9128,
+      "step": 2694
+    },
+    {
+      "epoch": 2.1526473526473526,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.0002,
+      "loss": 0.9041,
+      "step": 2695
+    },
+    {
+      "epoch": 2.1534465534465532,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9099,
+      "step": 2696
+    },
+    {
+      "epoch": 2.1542457542457543,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0002,
+      "loss": 0.9182,
+      "step": 2697
+    },
+    {
+      "epoch": 2.155044955044955,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.0002,
+      "loss": 0.9095,
+      "step": 2698
+    },
+    {
+      "epoch": 2.155844155844156,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0002,
+      "loss": 0.9133,
+      "step": 2699
+    },
+    {
+      "epoch": 2.1566433566433565,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9079,
+      "step": 2700
+    },
+    {
+      "epoch": 2.1574425574425575,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.0002,
+      "loss": 0.9118,
+      "step": 2701
+    },
+    {
+      "epoch": 2.158241758241758,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9133,
+      "step": 2702
+    },
+    {
+      "epoch": 2.159040959040959,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0002,
+      "loss": 0.9094,
+      "step": 2703
+    },
+    {
+      "epoch": 2.1598401598401598,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0002,
+      "loss": 0.9102,
+      "step": 2704
+    },
+    {
+      "epoch": 2.160639360639361,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.0002,
+      "loss": 0.9093,
+      "step": 2705
+    },
+    {
+      "epoch": 2.1614385614385614,
+      "grad_norm": 0.8984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9231,
+      "step": 2706
+    },
+    {
+      "epoch": 2.1622377622377624,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0002,
+      "loss": 0.9063,
+      "step": 2707
+    },
+    {
+      "epoch": 2.163036963036963,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9161,
+      "step": 2708
+    },
+    {
+      "epoch": 2.163836163836164,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0002,
+      "loss": 0.91,
+      "step": 2709
+    },
+    {
+      "epoch": 2.1646353646353647,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0002,
+      "loss": 0.9134,
+      "step": 2710
+    },
+    {
+      "epoch": 2.1654345654345653,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9103,
+      "step": 2711
+    },
+    {
+      "epoch": 2.1662337662337663,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0002,
+      "loss": 0.9122,
+      "step": 2712
+    },
+    {
+      "epoch": 2.167032967032967,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0002,
+      "loss": 0.9098,
+      "step": 2713
+    },
+    {
+      "epoch": 2.167832167832168,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9112,
+      "step": 2714
+    },
+    {
+      "epoch": 2.1686313686313685,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0002,
+      "loss": 0.9087,
+      "step": 2715
+    },
+    {
+      "epoch": 2.1694305694305696,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0002,
+      "loss": 0.9138,
+      "step": 2716
+    },
+    {
+      "epoch": 2.17022977022977,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0002,
+      "loss": 0.9203,
+      "step": 2717
+    },
+    {
+      "epoch": 2.171028971028971,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0002,
+      "loss": 0.9222,
+      "step": 2718
+    },
+    {
+      "epoch": 2.171828171828172,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9036,
+      "step": 2719
+    },
+    {
+      "epoch": 2.172627372627373,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9063,
+      "step": 2720
+    },
+    {
+      "epoch": 2.1734265734265734,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0002,
+      "loss": 0.9166,
+      "step": 2721
+    },
+    {
+      "epoch": 2.174225774225774,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0002,
+      "loss": 0.9075,
+      "step": 2722
+    },
+    {
+      "epoch": 2.175024975024975,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9121,
+      "step": 2723
+    },
+    {
+      "epoch": 2.1758241758241756,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0002,
+      "loss": 0.9065,
+      "step": 2724
+    },
+    {
+      "epoch": 2.1766233766233767,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9093,
+      "step": 2725
+    },
+    {
+      "epoch": 2.1774225774225773,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0002,
+      "loss": 0.9126,
+      "step": 2726
+    },
+    {
+      "epoch": 2.1782217782217783,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0002,
+      "loss": 0.9056,
+      "step": 2727
+    },
+    {
+      "epoch": 2.179020979020979,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0002,
+      "loss": 0.903,
+      "step": 2728
+    },
+    {
+      "epoch": 2.17982017982018,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9077,
+      "step": 2729
+    },
+    {
+      "epoch": 2.1806193806193805,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0002,
+      "loss": 0.9086,
+      "step": 2730
+    },
+    {
+      "epoch": 2.1814185814185816,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0002,
+      "loss": 0.9066,
+      "step": 2731
+    },
+    {
+      "epoch": 2.182217782217782,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0002,
+      "loss": 0.9073,
+      "step": 2732
+    },
+    {
+      "epoch": 2.183016983016983,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9121,
+      "step": 2733
+    },
+    {
+      "epoch": 2.183816183816184,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9035,
+      "step": 2734
+    },
+    {
+      "epoch": 2.184615384615385,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0002,
+      "loss": 0.9075,
+      "step": 2735
+    },
+    {
+      "epoch": 2.1854145854145854,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9143,
+      "step": 2736
+    },
+    {
+      "epoch": 2.186213786213786,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9145,
+      "step": 2737
+    },
+    {
+      "epoch": 2.187012987012987,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0002,
+      "loss": 0.9117,
+      "step": 2738
+    },
+    {
+      "epoch": 2.1878121878121877,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0002,
+      "loss": 0.9118,
+      "step": 2739
+    },
+    {
+      "epoch": 2.1886113886113887,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9124,
+      "step": 2740
+    },
+    {
+      "epoch": 2.1894105894105893,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0002,
+      "loss": 0.915,
+      "step": 2741
+    },
+    {
+      "epoch": 2.1902097902097903,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0002,
+      "loss": 0.9144,
+      "step": 2742
+    },
+    {
+      "epoch": 2.191008991008991,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0002,
+      "loss": 0.9062,
+      "step": 2743
+    },
+    {
+      "epoch": 2.191808191808192,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0002,
+      "loss": 0.909,
+      "step": 2744
+    },
+    {
+      "epoch": 2.1926073926073926,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8992,
+      "step": 2745
+    },
+    {
+      "epoch": 2.1934065934065936,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9227,
+      "step": 2746
+    },
+    {
+      "epoch": 2.194205794205794,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0002,
+      "loss": 0.9073,
+      "step": 2747
+    },
+    {
+      "epoch": 2.195004995004995,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0002,
+      "loss": 0.9143,
+      "step": 2748
+    },
+    {
+      "epoch": 2.195804195804196,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0002,
+      "loss": 0.9063,
+      "step": 2749
+    },
+    {
+      "epoch": 2.1966033966033964,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9132,
+      "step": 2750
+    },
+    {
+      "epoch": 2.1974025974025975,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9156,
+      "step": 2751
+    },
+    {
+      "epoch": 2.198201798201798,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9122,
+      "step": 2752
+    },
+    {
+      "epoch": 2.199000999000999,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9175,
+      "step": 2753
+    },
+    {
+      "epoch": 2.1998001998001997,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0002,
+      "loss": 0.915,
+      "step": 2754
+    },
+    {
+      "epoch": 2.2005994005994007,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0002,
+      "loss": 0.9081,
+      "step": 2755
+    },
+    {
+      "epoch": 2.2013986013986013,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9155,
+      "step": 2756
+    },
+    {
+      "epoch": 2.2021978021978024,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0002,
+      "loss": 0.9068,
+      "step": 2757
+    },
+    {
+      "epoch": 2.202997002997003,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0002,
+      "loss": 0.9041,
+      "step": 2758
+    },
+    {
+      "epoch": 2.203796203796204,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9141,
+      "step": 2759
+    },
+    {
+      "epoch": 2.2045954045954046,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9106,
+      "step": 2760
+    },
+    {
+      "epoch": 2.205394605394605,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0002,
+      "loss": 0.915,
+      "step": 2761
+    },
+    {
+      "epoch": 2.2061938061938062,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0002,
+      "loss": 0.9083,
+      "step": 2762
+    },
+    {
+      "epoch": 2.206993006993007,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0002,
+      "loss": 0.9112,
+      "step": 2763
+    },
+    {
+      "epoch": 2.207792207792208,
+      "grad_norm": 1.75,
+      "learning_rate": 0.0002,
+      "loss": 0.9361,
+      "step": 2764
+    },
+    {
+      "epoch": 2.2085914085914085,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9118,
+      "step": 2765
+    },
+    {
+      "epoch": 2.2093906093906095,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0002,
+      "loss": 0.9091,
+      "step": 2766
+    },
+    {
+      "epoch": 2.21018981018981,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9118,
+      "step": 2767
+    },
+    {
+      "epoch": 2.210989010989011,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0002,
+      "loss": 0.9139,
+      "step": 2768
+    },
+    {
+      "epoch": 2.2117882117882117,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9159,
+      "step": 2769
+    },
+    {
+      "epoch": 2.2125874125874128,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9094,
+      "step": 2770
+    },
+    {
+      "epoch": 2.2133866133866134,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.0002,
+      "loss": 0.9026,
+      "step": 2771
+    },
+    {
+      "epoch": 2.2141858141858144,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0002,
+      "loss": 0.9027,
+      "step": 2772
+    },
+    {
+      "epoch": 2.214985014985015,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0002,
+      "loss": 0.9059,
+      "step": 2773
+    },
+    {
+      "epoch": 2.2157842157842156,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9118,
+      "step": 2774
+    },
+    {
+      "epoch": 2.2165834165834166,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.0002,
+      "loss": 0.9063,
+      "step": 2775
+    },
+    {
+      "epoch": 2.217382617382617,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0002,
+      "loss": 0.9079,
+      "step": 2776
+    },
+    {
+      "epoch": 2.2181818181818183,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9228,
+      "step": 2777
+    },
+    {
+      "epoch": 2.218981018981019,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0002,
+      "loss": 0.9208,
+      "step": 2778
+    },
+    {
+      "epoch": 2.21978021978022,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9028,
+      "step": 2779
+    },
+    {
+      "epoch": 2.2205794205794205,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0002,
+      "loss": 0.9071,
+      "step": 2780
+    },
+    {
+      "epoch": 2.2213786213786215,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0002,
+      "loss": 0.9101,
+      "step": 2781
+    },
+    {
+      "epoch": 2.222177822177822,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0002,
+      "loss": 0.9126,
+      "step": 2782
+    },
+    {
+      "epoch": 2.222977022977023,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0002,
+      "loss": 0.9031,
+      "step": 2783
+    },
+    {
+      "epoch": 2.2237762237762237,
+      "grad_norm": 0.375,
+      "learning_rate": 0.0002,
+      "loss": 0.914,
+      "step": 2784
+    },
+    {
+      "epoch": 2.2245754245754243,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0002,
+      "loss": 0.9136,
+      "step": 2785
+    },
+    {
+      "epoch": 2.2253746253746254,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0002,
+      "loss": 0.909,
+      "step": 2786
+    },
+    {
+      "epoch": 2.226173826173826,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9079,
+      "step": 2787
+    },
+    {
+      "epoch": 2.226973026973027,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0002,
+      "loss": 0.914,
+      "step": 2788
+    },
+    {
+      "epoch": 2.2277722277722276,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0002,
+      "loss": 0.9108,
+      "step": 2789
+    },
+    {
+      "epoch": 2.2285714285714286,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0002,
+      "loss": 0.9101,
+      "step": 2790
+    },
+    {
+      "epoch": 2.2293706293706292,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0002,
+      "loss": 0.9133,
+      "step": 2791
+    },
+    {
+      "epoch": 2.2301698301698303,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9104,
+      "step": 2792
+    },
+    {
+      "epoch": 2.230969030969031,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0002,
+      "loss": 0.909,
+      "step": 2793
+    },
+    {
+      "epoch": 2.231768231768232,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9126,
+      "step": 2794
+    },
+    {
+      "epoch": 2.2325674325674325,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0002,
+      "loss": 0.9038,
+      "step": 2795
+    },
+    {
+      "epoch": 2.2333666333666335,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0002,
+      "loss": 0.8946,
+      "step": 2796
+    },
+    {
+      "epoch": 2.234165834165834,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9052,
+      "step": 2797
+    },
+    {
+      "epoch": 2.234965034965035,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0002,
+      "loss": 0.915,
+      "step": 2798
+    },
+    {
+      "epoch": 2.2357642357642358,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.0002,
+      "loss": 0.9041,
+      "step": 2799
+    },
+    {
+      "epoch": 2.2365634365634364,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0002,
+      "loss": 0.8996,
+      "step": 2800
+    },
+    {
+      "epoch": 2.2373626373626374,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.0002,
+      "loss": 0.9094,
+      "step": 2801
+    },
+    {
+      "epoch": 2.238161838161838,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0002,
+      "loss": 0.9184,
+      "step": 2802
+    },
+    {
+      "epoch": 2.238961038961039,
+      "grad_norm": 0.375,
+      "learning_rate": 0.0002,
+      "loss": 0.9102,
+      "step": 2803
+    },
+    {
+      "epoch": 2.2397602397602396,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0002,
+      "loss": 0.8991,
+      "step": 2804
+    },
+    {
+      "epoch": 2.2405594405594407,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0002,
+      "loss": 0.9166,
+      "step": 2805
+    },
+    {
+      "epoch": 2.2413586413586413,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0002,
+      "loss": 0.9129,
+      "step": 2806
+    },
+    {
+      "epoch": 2.2421578421578423,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0002,
+      "loss": 0.9213,
+      "step": 2807
+    },
+    {
+      "epoch": 2.242957042957043,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0002,
+      "loss": 0.9009,
+      "step": 2808
+    },
+    {
+      "epoch": 2.243756243756244,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0002,
+      "loss": 0.905,
+      "step": 2809
+    },
+    {
+      "epoch": 2.2445554445554445,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.0002,
+      "loss": 0.9156,
+      "step": 2810
+    },
+    {
+      "epoch": 2.245354645354645,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0002,
+      "loss": 0.9026,
+      "step": 2811
+    },
+    {
+      "epoch": 2.246153846153846,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9024,
+      "step": 2812
+    },
+    {
+      "epoch": 2.2469530469530468,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0002,
+      "loss": 0.9194,
+      "step": 2813
+    },
+    {
+      "epoch": 2.247752247752248,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.0002,
+      "loss": 0.9064,
+      "step": 2814
+    },
+    {
+      "epoch": 2.2485514485514484,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0002,
+      "loss": 0.9064,
+      "step": 2815
+    },
+    {
+      "epoch": 2.2493506493506494,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.0002,
+      "loss": 0.9146,
+      "step": 2816
+    },
+    {
+      "epoch": 2.25014985014985,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0002,
+      "loss": 0.9078,
+      "step": 2817
+    },
+    {
+      "epoch": 2.250949050949051,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.0002,
+      "loss": 0.9088,
+      "step": 2818
+    },
+    {
+      "epoch": 2.2517482517482517,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0002,
+      "loss": 0.9127,
+      "step": 2819
+    },
+    {
+      "epoch": 2.2525474525474527,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8986,
+      "step": 2820
+    },
+    {
+      "epoch": 2.2533466533466533,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0002,
+      "loss": 0.9056,
+      "step": 2821
+    },
+    {
+      "epoch": 2.2541458541458543,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.0002,
+      "loss": 0.9298,
+      "step": 2822
+    },
+    {
+      "epoch": 2.254945054945055,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9065,
+      "step": 2823
+    },
+    {
+      "epoch": 2.255744255744256,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0002,
+      "loss": 0.9038,
+      "step": 2824
+    },
+    {
+      "epoch": 2.2565434565434566,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0002,
+      "loss": 0.9132,
+      "step": 2825
+    },
+    {
+      "epoch": 2.257342657342657,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.0002,
+      "loss": 0.9123,
+      "step": 2826
+    },
+    {
+      "epoch": 2.258141858141858,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0002,
+      "loss": 0.9256,
+      "step": 2827
+    },
+    {
+      "epoch": 2.258941058941059,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9201,
+      "step": 2828
+    },
+    {
+      "epoch": 2.25974025974026,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0002,
+      "loss": 0.9072,
+      "step": 2829
+    },
+    {
+      "epoch": 2.2605394605394604,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8982,
+      "step": 2830
+    },
+    {
+      "epoch": 2.2613386613386615,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.0002,
+      "loss": 0.9153,
+      "step": 2831
+    },
+    {
+      "epoch": 2.262137862137862,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.0002,
+      "loss": 0.9099,
+      "step": 2832
+    },
+    {
+      "epoch": 2.262937062937063,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.0002,
+      "loss": 0.9088,
+      "step": 2833
+    },
+    {
+      "epoch": 2.2637362637362637,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.0002,
+      "loss": 0.9014,
+      "step": 2834
+    },
+    {
+      "epoch": 2.2645354645354647,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0002,
+      "loss": 0.9119,
+      "step": 2835
+    },
+    {
+      "epoch": 2.2653346653346653,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9057,
+      "step": 2836
+    },
+    {
+      "epoch": 2.266133866133866,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0002,
+      "loss": 0.9116,
+      "step": 2837
+    },
+    {
+      "epoch": 2.266933066933067,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9111,
+      "step": 2838
+    },
+    {
+      "epoch": 2.2677322677322675,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9096,
+      "step": 2839
+    },
+    {
+      "epoch": 2.2685314685314686,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0002,
+      "loss": 0.9151,
+      "step": 2840
+    },
+    {
+      "epoch": 2.269330669330669,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0002,
+      "loss": 0.9177,
+      "step": 2841
+    },
+    {
+      "epoch": 2.27012987012987,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.0002,
+      "loss": 0.9074,
+      "step": 2842
+    },
+    {
+      "epoch": 2.270929070929071,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9029,
+      "step": 2843
+    },
+    {
+      "epoch": 2.271728271728272,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0002,
+      "loss": 0.917,
+      "step": 2844
+    },
+    {
+      "epoch": 2.2725274725274724,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.0002,
+      "loss": 0.9147,
+      "step": 2845
+    },
+    {
+      "epoch": 2.2733266733266735,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.0002,
+      "loss": 0.9071,
+      "step": 2846
+    },
+    {
+      "epoch": 2.274125874125874,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.0002,
+      "loss": 0.9021,
+      "step": 2847
+    },
+    {
+      "epoch": 2.274925074925075,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9083,
+      "step": 2848
+    },
+    {
+      "epoch": 2.2757242757242757,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.0002,
+      "loss": 0.915,
+      "step": 2849
+    },
+    {
+      "epoch": 2.2765234765234768,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9123,
+      "step": 2850
+    },
+    {
+      "epoch": 2.2773226773226773,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0002,
+      "loss": 0.9114,
+      "step": 2851
+    },
+    {
+      "epoch": 2.278121878121878,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9108,
+      "step": 2852
+    },
+    {
+      "epoch": 2.278921078921079,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9204,
+      "step": 2853
+    },
+    {
+      "epoch": 2.2797202797202796,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.0002,
+      "loss": 0.9111,
+      "step": 2854
+    },
+    {
+      "epoch": 2.2805194805194806,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.0002,
+      "loss": 0.9105,
+      "step": 2855
+    },
+    {
+      "epoch": 2.281318681318681,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0002,
+      "loss": 0.8926,
+      "step": 2856
+    },
+    {
+      "epoch": 2.2821178821178822,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9028,
+      "step": 2857
+    },
+    {
+      "epoch": 2.282917082917083,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.0002,
+      "loss": 0.9168,
+      "step": 2858
+    },
+    {
+      "epoch": 2.283716283716284,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0002,
+      "loss": 0.9175,
+      "step": 2859
+    },
+    {
+      "epoch": 2.2845154845154845,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9163,
+      "step": 2860
+    },
+    {
+      "epoch": 2.2853146853146855,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9048,
+      "step": 2861
+    },
+    {
+      "epoch": 2.286113886113886,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0002,
+      "loss": 0.9151,
+      "step": 2862
+    },
+    {
+      "epoch": 2.2869130869130867,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0002,
+      "loss": 0.9077,
+      "step": 2863
+    },
+    {
+      "epoch": 2.2877122877122877,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9154,
+      "step": 2864
+    },
+    {
+      "epoch": 2.2885114885114883,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0002,
+      "loss": 0.9119,
+      "step": 2865
+    },
+    {
+      "epoch": 2.2893106893106894,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9082,
+      "step": 2866
+    },
+    {
+      "epoch": 2.29010989010989,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.0002,
+      "loss": 0.9175,
+      "step": 2867
+    },
+    {
+      "epoch": 2.290909090909091,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9087,
+      "step": 2868
+    },
+    {
+      "epoch": 2.2917082917082916,
+      "grad_norm": 1.8515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9176,
+      "step": 2869
+    },
+    {
+      "epoch": 2.2925074925074926,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9025,
+      "step": 2870
+    },
+    {
+      "epoch": 2.2933066933066932,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0002,
+      "loss": 0.9359,
+      "step": 2871
+    },
+    {
+      "epoch": 2.2941058941058943,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.0002,
+      "loss": 0.9096,
+      "step": 2872
+    },
+    {
+      "epoch": 2.294905094905095,
+      "grad_norm": 1.4375,
+      "learning_rate": 0.0002,
+      "loss": 0.9116,
+      "step": 2873
+    },
+    {
+      "epoch": 2.2957042957042955,
+      "grad_norm": 0.85546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9128,
+      "step": 2874
+    },
+    {
+      "epoch": 2.2965034965034965,
+      "grad_norm": 0.86328125,
+      "learning_rate": 0.0002,
+      "loss": 0.9136,
+      "step": 2875
+    },
+    {
+      "epoch": 2.2973026973026975,
+      "grad_norm": 0.96875,
+      "learning_rate": 0.0002,
+      "loss": 0.9192,
+      "step": 2876
+    },
+    {
+      "epoch": 2.298101898101898,
+      "grad_norm": 0.94140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9156,
+      "step": 2877
+    },
+    {
+      "epoch": 2.2989010989010987,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.0002,
+      "loss": 0.911,
+      "step": 2878
+    },
+    {
+      "epoch": 2.2997002997002998,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0002,
+      "loss": 0.9158,
+      "step": 2879
+    },
+    {
+      "epoch": 2.3004995004995004,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.0002,
+      "loss": 0.9136,
+      "step": 2880
+    },
+    {
+      "epoch": 2.3012987012987014,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0002,
+      "loss": 0.9102,
+      "step": 2881
+    },
+    {
+      "epoch": 2.302097902097902,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9242,
+      "step": 2882
+    },
+    {
+      "epoch": 2.302897102897103,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0002,
+      "loss": 0.9083,
+      "step": 2883
+    },
+    {
+      "epoch": 2.3036963036963036,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0002,
+      "loss": 0.9139,
+      "step": 2884
+    },
+    {
+      "epoch": 2.3044955044955047,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9155,
+      "step": 2885
+    },
+    {
+      "epoch": 2.3052947052947053,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9102,
+      "step": 2886
+    },
+    {
+      "epoch": 2.3060939060939063,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0002,
+      "loss": 0.9123,
+      "step": 2887
+    },
+    {
+      "epoch": 2.306893106893107,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0002,
+      "loss": 0.9197,
+      "step": 2888
+    },
+    {
+      "epoch": 2.3076923076923075,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9117,
+      "step": 2889
+    },
+    {
+      "epoch": 2.3084915084915085,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0002,
+      "loss": 0.9081,
+      "step": 2890
+    },
+    {
+      "epoch": 2.309290709290709,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0002,
+      "loss": 0.904,
+      "step": 2891
+    },
+    {
+      "epoch": 2.31008991008991,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0002,
+      "loss": 0.9156,
+      "step": 2892
+    },
+    {
+      "epoch": 2.3108891108891108,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9193,
+      "step": 2893
+    },
+    {
+      "epoch": 2.311688311688312,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0002,
+      "loss": 0.918,
+      "step": 2894
+    },
+    {
+      "epoch": 2.3124875124875124,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9195,
+      "step": 2895
+    },
+    {
+      "epoch": 2.3132867132867134,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0002,
+      "loss": 0.912,
+      "step": 2896
+    },
+    {
+      "epoch": 2.314085914085914,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9145,
+      "step": 2897
+    },
+    {
+      "epoch": 2.314885114885115,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9005,
+      "step": 2898
+    },
+    {
+      "epoch": 2.3156843156843157,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0002,
+      "loss": 0.9129,
+      "step": 2899
+    },
+    {
+      "epoch": 2.3164835164835162,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0002,
+      "loss": 0.913,
+      "step": 2900
+    },
+    {
+      "epoch": 2.3172827172827173,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0002,
+      "loss": 0.9111,
+      "step": 2901
+    },
+    {
+      "epoch": 2.3180819180819183,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0002,
+      "loss": 0.9123,
+      "step": 2902
+    },
+    {
+      "epoch": 2.318881118881119,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0002,
+      "loss": 0.9247,
+      "step": 2903
+    },
+    {
+      "epoch": 2.3196803196803195,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9216,
+      "step": 2904
+    },
+    {
+      "epoch": 2.3204795204795206,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0002,
+      "loss": 0.9147,
+      "step": 2905
+    },
+    {
+      "epoch": 2.321278721278721,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0002,
+      "loss": 0.9089,
+      "step": 2906
+    },
+    {
+      "epoch": 2.322077922077922,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.0002,
+      "loss": 0.9097,
+      "step": 2907
+    },
+    {
+      "epoch": 2.322877122877123,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9079,
+      "step": 2908
+    },
+    {
+      "epoch": 2.323676323676324,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.0002,
+      "loss": 0.9137,
+      "step": 2909
+    },
+    {
+      "epoch": 2.3244755244755244,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0002,
+      "loss": 0.909,
+      "step": 2910
+    },
+    {
+      "epoch": 2.3252747252747255,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9141,
+      "step": 2911
+    },
+    {
+      "epoch": 2.326073926073926,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0002,
+      "loss": 0.9113,
+      "step": 2912
+    },
+    {
+      "epoch": 2.326873126873127,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0002,
+      "loss": 0.9131,
+      "step": 2913
+    },
+    {
+      "epoch": 2.3276723276723277,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9151,
+      "step": 2914
+    },
+    {
+      "epoch": 2.3284715284715283,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.0002,
+      "loss": 0.909,
+      "step": 2915
+    },
+    {
+      "epoch": 2.3292707292707293,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.0002,
+      "loss": 0.9059,
+      "step": 2916
+    },
+    {
+      "epoch": 2.33006993006993,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0002,
+      "loss": 0.9072,
+      "step": 2917
+    },
+    {
+      "epoch": 2.330869130869131,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9092,
+      "step": 2918
+    },
+    {
+      "epoch": 2.3316683316683315,
+      "grad_norm": 0.80859375,
+      "learning_rate": 0.0002,
+      "loss": 0.9078,
+      "step": 2919
+    },
+    {
+      "epoch": 2.3324675324675326,
+      "grad_norm": 0.890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8991,
+      "step": 2920
+    },
+    {
+      "epoch": 2.333266733266733,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9112,
+      "step": 2921
+    },
+    {
+      "epoch": 2.334065934065934,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0002,
+      "loss": 0.916,
+      "step": 2922
+    },
+    {
+      "epoch": 2.334865134865135,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0002,
+      "loss": 0.9019,
+      "step": 2923
+    },
+    {
+      "epoch": 2.335664335664336,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.0002,
+      "loss": 0.914,
+      "step": 2924
+    },
+    {
+      "epoch": 2.3364635364635364,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0002,
+      "loss": 0.9111,
+      "step": 2925
+    },
+    {
+      "epoch": 2.337262737262737,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0002,
+      "loss": 0.9137,
+      "step": 2926
+    },
+    {
+      "epoch": 2.338061938061938,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9148,
+      "step": 2927
+    },
+    {
+      "epoch": 2.338861138861139,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.0002,
+      "loss": 0.9134,
+      "step": 2928
+    },
+    {
+      "epoch": 2.3396603396603397,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0002,
+      "loss": 0.9051,
+      "step": 2929
+    },
+    {
+      "epoch": 2.3404595404595403,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8993,
+      "step": 2930
+    },
+    {
+      "epoch": 2.3412587412587413,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9103,
+      "step": 2931
+    },
+    {
+      "epoch": 2.342057942057942,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0002,
+      "loss": 0.9081,
+      "step": 2932
+    },
+    {
+      "epoch": 2.342857142857143,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9095,
+      "step": 2933
+    },
+    {
+      "epoch": 2.3436563436563436,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0002,
+      "loss": 0.9153,
+      "step": 2934
+    },
+    {
+      "epoch": 2.3444555444555446,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9145,
+      "step": 2935
+    },
+    {
+      "epoch": 2.345254745254745,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.0002,
+      "loss": 0.8996,
+      "step": 2936
+    },
+    {
+      "epoch": 2.3460539460539462,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9105,
+      "step": 2937
+    },
+    {
+      "epoch": 2.346853146853147,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0002,
+      "loss": 0.9248,
+      "step": 2938
+    },
+    {
+      "epoch": 2.347652347652348,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0002,
+      "loss": 0.9076,
+      "step": 2939
+    },
+    {
+      "epoch": 2.3484515484515485,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.0002,
+      "loss": 0.898,
+      "step": 2940
+    },
+    {
+      "epoch": 2.349250749250749,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.0002,
+      "loss": 0.9138,
+      "step": 2941
+    },
+    {
+      "epoch": 2.35004995004995,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.0002,
+      "loss": 0.9178,
+      "step": 2942
+    },
+    {
+      "epoch": 2.3508491508491507,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0002,
+      "loss": 0.9079,
+      "step": 2943
+    },
+    {
+      "epoch": 2.3516483516483517,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.0002,
+      "loss": 0.9118,
+      "step": 2944
+    },
+    {
+      "epoch": 2.3524475524475523,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0002,
+      "loss": 0.9156,
+      "step": 2945
+    },
+    {
+      "epoch": 2.3532467532467534,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0002,
+      "loss": 0.9025,
+      "step": 2946
+    },
+    {
+      "epoch": 2.354045954045954,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0002,
+      "loss": 0.9105,
+      "step": 2947
+    },
+    {
+      "epoch": 2.354845154845155,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0002,
+      "loss": 0.9093,
+      "step": 2948
+    },
+    {
+      "epoch": 2.3556443556443556,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9127,
+      "step": 2949
+    },
+    {
+      "epoch": 2.3564435564435566,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0002,
+      "loss": 0.9021,
+      "step": 2950
+    },
+    {
+      "epoch": 2.3572427572427572,
+      "grad_norm": 0.8359375,
+      "learning_rate": 0.0002,
+      "loss": 0.92,
+      "step": 2951
+    },
+    {
+      "epoch": 2.358041958041958,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0002,
+      "loss": 0.905,
+      "step": 2952
+    },
+    {
+      "epoch": 2.358841158841159,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0002,
+      "loss": 0.9075,
+      "step": 2953
+    },
+    {
+      "epoch": 2.3596403596403595,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9096,
+      "step": 2954
+    },
+    {
+      "epoch": 2.3604395604395605,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0002,
+      "loss": 0.9117,
+      "step": 2955
+    },
+    {
+      "epoch": 2.361238761238761,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0002,
+      "loss": 0.9077,
+      "step": 2956
+    },
+    {
+      "epoch": 2.362037962037962,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0002,
+      "loss": 0.9099,
+      "step": 2957
+    },
+    {
+      "epoch": 2.3628371628371627,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0002,
+      "loss": 0.9109,
+      "step": 2958
+    },
+    {
+      "epoch": 2.3636363636363638,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8972,
+      "step": 2959
+    },
+    {
+      "epoch": 2.3644355644355644,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9158,
+      "step": 2960
+    },
+    {
+      "epoch": 2.3652347652347654,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9204,
+      "step": 2961
+    },
+    {
+      "epoch": 2.366033966033966,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.0002,
+      "loss": 0.9098,
+      "step": 2962
+    },
+    {
+      "epoch": 2.366833166833167,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.0002,
+      "loss": 0.9078,
+      "step": 2963
+    },
+    {
+      "epoch": 2.3676323676323676,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.0002,
+      "loss": 0.9077,
+      "step": 2964
+    },
+    {
+      "epoch": 2.3684315684315687,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.0002,
+      "loss": 0.9073,
+      "step": 2965
+    },
+    {
+      "epoch": 2.3692307692307693,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9055,
+      "step": 2966
+    },
+    {
+      "epoch": 2.37002997002997,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0002,
+      "loss": 0.9116,
+      "step": 2967
+    },
+    {
+      "epoch": 2.370829170829171,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9094,
+      "step": 2968
+    },
+    {
+      "epoch": 2.3716283716283715,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0002,
+      "loss": 0.9053,
+      "step": 2969
+    },
+    {
+      "epoch": 2.3724275724275725,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0002,
+      "loss": 0.909,
+      "step": 2970
+    },
+    {
+      "epoch": 2.373226773226773,
+      "grad_norm": 0.375,
+      "learning_rate": 0.0002,
+      "loss": 0.9131,
+      "step": 2971
+    },
+    {
+      "epoch": 2.374025974025974,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9124,
+      "step": 2972
+    },
+    {
+      "epoch": 2.3748251748251747,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.0002,
+      "loss": 0.916,
+      "step": 2973
+    },
+    {
+      "epoch": 2.375624375624376,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.0002,
+      "loss": 0.9144,
+      "step": 2974
+    },
+    {
+      "epoch": 2.3764235764235764,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0002,
+      "loss": 0.9066,
+      "step": 2975
+    },
+    {
+      "epoch": 2.3772227772227774,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9067,
+      "step": 2976
+    },
+    {
+      "epoch": 2.378021978021978,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0002,
+      "loss": 0.9071,
+      "step": 2977
+    },
+    {
+      "epoch": 2.3788211788211786,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0002,
+      "loss": 0.9073,
+      "step": 2978
+    },
+    {
+      "epoch": 2.3796203796203796,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0002,
+      "loss": 0.9121,
+      "step": 2979
+    },
+    {
+      "epoch": 2.3804195804195802,
+      "grad_norm": 0.375,
+      "learning_rate": 0.0002,
+      "loss": 0.9093,
+      "step": 2980
+    },
+    {
+      "epoch": 2.3812187812187813,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0002,
+      "loss": 0.9052,
+      "step": 2981
+    },
+    {
+      "epoch": 2.382017982017982,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.0002,
+      "loss": 0.9091,
+      "step": 2982
+    },
+    {
+      "epoch": 2.382817182817183,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0002,
+      "loss": 0.914,
+      "step": 2983
+    },
+    {
+      "epoch": 2.3836163836163835,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0002,
+      "loss": 0.9107,
+      "step": 2984
+    },
+    {
+      "epoch": 2.3844155844155845,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.0002,
+      "loss": 0.906,
+      "step": 2985
+    },
+    {
+      "epoch": 2.385214785214785,
+      "grad_norm": 0.85546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8982,
+      "step": 2986
+    },
+    {
+      "epoch": 2.386013986013986,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0002,
+      "loss": 0.9095,
+      "step": 2987
+    },
+    {
+      "epoch": 2.3868131868131868,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0002,
+      "loss": 0.9088,
+      "step": 2988
+    },
+    {
+      "epoch": 2.3876123876123874,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.0002,
+      "loss": 0.9073,
+      "step": 2989
+    },
+    {
+      "epoch": 2.3884115884115884,
+      "grad_norm": 0.828125,
+      "learning_rate": 0.0002,
+      "loss": 0.9075,
+      "step": 2990
+    },
+    {
+      "epoch": 2.3892107892107894,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9086,
+      "step": 2991
+    },
+    {
+      "epoch": 2.39000999000999,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0002,
+      "loss": 0.9111,
+      "step": 2992
+    },
+    {
+      "epoch": 2.3908091908091906,
+      "grad_norm": 0.625,
+      "learning_rate": 0.0002,
+      "loss": 0.9132,
+      "step": 2993
+    },
+    {
+      "epoch": 2.3916083916083917,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.0002,
+      "loss": 0.905,
+      "step": 2994
+    },
+    {
+      "epoch": 2.3924075924075923,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9027,
+      "step": 2995
+    },
+    {
+      "epoch": 2.3932067932067933,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0002,
+      "loss": 0.9076,
+      "step": 2996
+    },
+    {
+      "epoch": 2.394005994005994,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0002,
+      "loss": 0.9086,
+      "step": 2997
+    },
+    {
+      "epoch": 2.394805194805195,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.0002,
+      "loss": 0.9171,
+      "step": 2998
+    },
+    {
+      "epoch": 2.3956043956043955,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9231,
+      "step": 2999
+    },
+    {
+      "epoch": 2.3964035964035966,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.0002,
+      "loss": 0.902,
+      "step": 3000
+    },
+    {
+      "epoch": 2.397202797202797,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0002,
+      "loss": 0.9161,
+      "step": 3001
+    },
+    {
+      "epoch": 2.398001998001998,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0002,
+      "loss": 0.9051,
+      "step": 3002
+    },
+    {
+      "epoch": 2.398801198801199,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9114,
+      "step": 3003
+    },
+    {
+      "epoch": 2.3996003996003994,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0002,
+      "loss": 0.914,
+      "step": 3004
+    },
+    {
+      "epoch": 2.4003996003996004,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0002,
+      "loss": 0.9069,
+      "step": 3005
+    },
+    {
+      "epoch": 2.401198801198801,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0002,
+      "loss": 0.8968,
+      "step": 3006
+    },
+    {
+      "epoch": 2.401998001998002,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.0002,
+      "loss": 0.9135,
+      "step": 3007
+    },
+    {
+      "epoch": 2.4027972027972027,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0002,
+      "loss": 0.9001,
+      "step": 3008
+    },
+    {
+      "epoch": 2.4035964035964037,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.0002,
+      "loss": 0.9092,
+      "step": 3009
+    },
+    {
+      "epoch": 2.4043956043956043,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9109,
+      "step": 3010
+    },
+    {
+      "epoch": 2.4051948051948053,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.0002,
+      "loss": 0.9168,
+      "step": 3011
+    },
+    {
+      "epoch": 2.405994005994006,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.0002,
+      "loss": 0.9099,
+      "step": 3012
+    },
+    {
+      "epoch": 2.406793206793207,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9016,
+      "step": 3013
+    },
+    {
+      "epoch": 2.4075924075924076,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.0002,
+      "loss": 0.9081,
+      "step": 3014
+    },
+    {
+      "epoch": 2.408391608391608,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9148,
+      "step": 3015
+    },
+    {
+      "epoch": 2.409190809190809,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0002,
+      "loss": 0.9177,
+      "step": 3016
+    },
+    {
+      "epoch": 2.4099900099900102,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0002,
+      "loss": 0.9071,
+      "step": 3017
+    },
+    {
+      "epoch": 2.410789210789211,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0002,
+      "loss": 0.8955,
+      "step": 3018
+    },
+    {
+      "epoch": 2.4115884115884114,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0002,
+      "loss": 0.905,
+      "step": 3019
+    },
+    {
+      "epoch": 2.4123876123876125,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.0002,
+      "loss": 0.9085,
+      "step": 3020
+    },
+    {
+      "epoch": 2.413186813186813,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.0002,
+      "loss": 0.9143,
+      "step": 3021
+    },
+    {
+      "epoch": 2.413986013986014,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.0002,
+      "loss": 0.9165,
+      "step": 3022
+    },
+    {
+      "epoch": 2.4147852147852147,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0002,
+      "loss": 0.9015,
+      "step": 3023
+    },
+    {
+      "epoch": 2.4155844155844157,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9015,
+      "step": 3024
+    },
+    {
+      "epoch": 2.4163836163836163,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0002,
+      "loss": 0.9176,
+      "step": 3025
+    },
+    {
+      "epoch": 2.4171828171828174,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0002,
+      "loss": 0.9137,
+      "step": 3026
+    },
+    {
+      "epoch": 2.417982017982018,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0002,
+      "loss": 0.9065,
+      "step": 3027
+    },
+    {
+      "epoch": 2.418781218781219,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0002,
+      "loss": 0.9157,
+      "step": 3028
+    },
+    {
+      "epoch": 2.4195804195804196,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9052,
+      "step": 3029
+    },
+    {
+      "epoch": 2.42037962037962,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.0002,
+      "loss": 0.9092,
+      "step": 3030
+    },
+    {
+      "epoch": 2.421178821178821,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0002,
+      "loss": 0.913,
+      "step": 3031
+    },
+    {
+      "epoch": 2.421978021978022,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0002,
+      "loss": 0.9112,
+      "step": 3032
+    },
+    {
+      "epoch": 2.422777222777223,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.0002,
+      "loss": 0.9188,
+      "step": 3033
+    },
+    {
+      "epoch": 2.4235764235764234,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.0002,
+      "loss": 0.9079,
+      "step": 3034
+    },
+    {
+      "epoch": 2.4243756243756245,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9018,
+      "step": 3035
+    },
+    {
+      "epoch": 2.425174825174825,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.0002,
+      "loss": 0.8996,
+      "step": 3036
+    },
+    {
+      "epoch": 2.425974025974026,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0002,
+      "loss": 0.9165,
+      "step": 3037
+    },
+    {
+      "epoch": 2.4267732267732267,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.0002,
+      "loss": 0.9013,
+      "step": 3038
+    },
+    {
+      "epoch": 2.4275724275724277,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.0002,
+      "loss": 0.9022,
+      "step": 3039
+    },
+    {
+      "epoch": 2.4283716283716283,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0002,
+      "loss": 0.9136,
+      "step": 3040
+    },
+    {
+      "epoch": 2.429170829170829,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0002,
+      "loss": 0.9158,
+      "step": 3041
+    },
+    {
+      "epoch": 2.42997002997003,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.0002,
+      "loss": 0.9136,
+      "step": 3042
+    },
+    {
+      "epoch": 2.430769230769231,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9111,
+      "step": 3043
+    },
+    {
+      "epoch": 2.4315684315684316,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9101,
+      "step": 3044
+    },
+    {
+      "epoch": 2.432367632367632,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0002,
+      "loss": 0.9128,
+      "step": 3045
+    },
+    {
+      "epoch": 2.4331668331668332,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9165,
+      "step": 3046
+    },
+    {
+      "epoch": 2.433966033966034,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0002,
+      "loss": 0.9005,
+      "step": 3047
+    },
+    {
+      "epoch": 2.434765234765235,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0002,
+      "loss": 0.9073,
+      "step": 3048
+    },
+    {
+      "epoch": 2.4355644355644355,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0002,
+      "loss": 0.9232,
+      "step": 3049
+    },
+    {
+      "epoch": 2.4363636363636365,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0002,
+      "loss": 0.9108,
+      "step": 3050
+    },
+    {
+      "epoch": 2.437162837162837,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0002,
+      "loss": 0.9137,
+      "step": 3051
+    },
+    {
+      "epoch": 2.437962037962038,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0002,
+      "loss": 0.9102,
+      "step": 3052
+    },
+    {
+      "epoch": 2.4387612387612387,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0002,
+      "loss": 0.9031,
+      "step": 3053
+    },
+    {
+      "epoch": 2.4395604395604398,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9023,
+      "step": 3054
+    },
+    {
+      "epoch": 2.4403596403596404,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.0002,
+      "loss": 0.9107,
+      "step": 3055
+    },
+    {
+      "epoch": 2.441158841158841,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0002,
+      "loss": 0.9053,
+      "step": 3056
+    },
+    {
+      "epoch": 2.441958041958042,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0002,
+      "loss": 0.9079,
+      "step": 3057
+    },
+    {
+      "epoch": 2.4427572427572426,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0002,
+      "loss": 0.9051,
+      "step": 3058
+    },
+    {
+      "epoch": 2.4435564435564436,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0002,
+      "loss": 0.9051,
+      "step": 3059
+    },
+    {
+      "epoch": 2.4443556443556442,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9134,
+      "step": 3060
+    },
+    {
+      "epoch": 2.4451548451548453,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9071,
+      "step": 3061
+    },
+    {
+      "epoch": 2.445954045954046,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9043,
+      "step": 3062
+    },
+    {
+      "epoch": 2.446753246753247,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.0002,
+      "loss": 0.9391,
+      "step": 3063
+    },
+    {
+      "epoch": 2.4475524475524475,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.0002,
+      "loss": 0.9067,
+      "step": 3064
+    },
+    {
+      "epoch": 2.4483516483516485,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9133,
+      "step": 3065
+    },
+    {
+      "epoch": 2.449150849150849,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.0002,
+      "loss": 0.9078,
+      "step": 3066
+    },
+    {
+      "epoch": 2.4499500499500497,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.0002,
+      "loss": 0.9066,
+      "step": 3067
+    },
+    {
+      "epoch": 2.4507492507492508,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0002,
+      "loss": 0.9041,
+      "step": 3068
+    },
+    {
+      "epoch": 2.4515484515484514,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0002,
+      "loss": 0.9026,
+      "step": 3069
+    },
+    {
+      "epoch": 2.4523476523476524,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0002,
+      "loss": 0.8975,
+      "step": 3070
+    },
+    {
+      "epoch": 2.453146853146853,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9042,
+      "step": 3071
+    },
+    {
+      "epoch": 2.453946053946054,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.0002,
+      "loss": 0.906,
+      "step": 3072
+    },
+    {
+      "epoch": 2.4547452547452546,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0002,
+      "loss": 0.9046,
+      "step": 3073
+    },
+    {
+      "epoch": 2.4555444555444557,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0002,
+      "loss": 0.9117,
+      "step": 3074
+    },
+    {
+      "epoch": 2.4563436563436563,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0002,
+      "loss": 0.9104,
+      "step": 3075
+    },
+    {
+      "epoch": 2.4571428571428573,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0002,
+      "loss": 0.9066,
+      "step": 3076
+    },
+    {
+      "epoch": 2.457942057942058,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0002,
+      "loss": 0.9156,
+      "step": 3077
+    },
+    {
+      "epoch": 2.458741258741259,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0002,
+      "loss": 0.9065,
+      "step": 3078
+    },
+    {
+      "epoch": 2.4595404595404595,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9032,
+      "step": 3079
+    },
+    {
+      "epoch": 2.4603396603396606,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9059,
+      "step": 3080
+    },
+    {
+      "epoch": 2.461138861138861,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0002,
+      "loss": 0.9131,
+      "step": 3081
+    },
+    {
+      "epoch": 2.4619380619380618,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0002,
+      "loss": 0.905,
+      "step": 3082
+    },
+    {
+      "epoch": 2.462737262737263,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9097,
+      "step": 3083
+    },
+    {
+      "epoch": 2.4635364635364634,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0002,
+      "loss": 0.9134,
+      "step": 3084
+    },
+    {
+      "epoch": 2.4643356643356644,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.0002,
+      "loss": 0.907,
+      "step": 3085
+    },
+    {
+      "epoch": 2.465134865134865,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9221,
+      "step": 3086
+    },
+    {
+      "epoch": 2.465934065934066,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.0002,
+      "loss": 0.907,
+      "step": 3087
+    },
+    {
+      "epoch": 2.4667332667332666,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.0002,
+      "loss": 0.9114,
+      "step": 3088
+    },
+    {
+      "epoch": 2.4675324675324677,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.0002,
+      "loss": 0.9092,
+      "step": 3089
+    },
+    {
+      "epoch": 2.4683316683316683,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0002,
+      "loss": 0.9132,
+      "step": 3090
+    },
+    {
+      "epoch": 2.4691308691308693,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0002,
+      "loss": 0.9102,
+      "step": 3091
+    },
+    {
+      "epoch": 2.46993006993007,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9107,
+      "step": 3092
+    },
+    {
+      "epoch": 2.4707292707292705,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0002,
+      "loss": 0.9031,
+      "step": 3093
+    },
+    {
+      "epoch": 2.4715284715284715,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0002,
+      "loss": 0.9103,
+      "step": 3094
+    },
+    {
+      "epoch": 2.472327672327672,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.0002,
+      "loss": 0.9042,
+      "step": 3095
+    },
+    {
+      "epoch": 2.473126873126873,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9125,
+      "step": 3096
+    },
+    {
+      "epoch": 2.4739260739260738,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0002,
+      "loss": 0.9072,
+      "step": 3097
+    },
+    {
+      "epoch": 2.474725274725275,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9026,
+      "step": 3098
+    },
+    {
+      "epoch": 2.4755244755244754,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0002,
+      "loss": 0.9018,
+      "step": 3099
+    },
+    {
+      "epoch": 2.4763236763236764,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9079,
+      "step": 3100
+    },
+    {
+      "epoch": 2.477122877122877,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0002,
+      "loss": 0.902,
+      "step": 3101
+    },
+    {
+      "epoch": 2.477922077922078,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9041,
+      "step": 3102
+    },
+    {
+      "epoch": 2.4787212787212787,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0002,
+      "loss": 0.9111,
+      "step": 3103
+    },
+    {
+      "epoch": 2.4795204795204797,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.0002,
+      "loss": 0.9038,
+      "step": 3104
+    },
+    {
+      "epoch": 2.4803196803196803,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.0002,
+      "loss": 0.9045,
+      "step": 3105
+    },
+    {
+      "epoch": 2.4811188811188813,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9096,
+      "step": 3106
+    },
+    {
+      "epoch": 2.481918081918082,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0002,
+      "loss": 0.9091,
+      "step": 3107
+    },
+    {
+      "epoch": 2.4827172827172825,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0002,
+      "loss": 0.9042,
+      "step": 3108
+    },
+    {
+      "epoch": 2.4835164835164836,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.0002,
+      "loss": 0.9109,
+      "step": 3109
+    },
+    {
+      "epoch": 2.484315684315684,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9212,
+      "step": 3110
+    },
+    {
+      "epoch": 2.485114885114885,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0002,
+      "loss": 0.9144,
+      "step": 3111
+    },
+    {
+      "epoch": 2.485914085914086,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9117,
+      "step": 3112
+    },
+    {
+      "epoch": 2.486713286713287,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9045,
+      "step": 3113
+    },
+    {
+      "epoch": 2.4875124875124874,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.0002,
+      "loss": 0.9079,
+      "step": 3114
+    },
+    {
+      "epoch": 2.4883116883116885,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0002,
+      "loss": 0.9117,
+      "step": 3115
+    },
+    {
+      "epoch": 2.489110889110889,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.0002,
+      "loss": 0.9137,
+      "step": 3116
+    },
+    {
+      "epoch": 2.48991008991009,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0002,
+      "loss": 0.9062,
+      "step": 3117
+    },
+    {
+      "epoch": 2.4907092907092907,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0002,
+      "loss": 0.9108,
+      "step": 3118
+    },
+    {
+      "epoch": 2.4915084915084913,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0002,
+      "loss": 0.8982,
+      "step": 3119
+    },
+    {
+      "epoch": 2.4923076923076923,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0002,
+      "loss": 0.9087,
+      "step": 3120
+    },
+    {
+      "epoch": 2.493106893106893,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0002,
+      "loss": 0.9018,
+      "step": 3121
+    },
+    {
+      "epoch": 2.493906093906094,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0002,
+      "loss": 0.9067,
+      "step": 3122
+    },
+    {
+      "epoch": 2.4947052947052946,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0002,
+      "loss": 0.8986,
+      "step": 3123
+    },
+    {
+      "epoch": 2.4955044955044956,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9024,
+      "step": 3124
+    },
+    {
+      "epoch": 2.496303696303696,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9054,
+      "step": 3125
+    },
+    {
+      "epoch": 2.4971028971028972,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0002,
+      "loss": 0.9069,
+      "step": 3126
+    },
+    {
+      "epoch": 2.497902097902098,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0002,
+      "loss": 0.8999,
+      "step": 3127
+    },
+    {
+      "epoch": 2.498701298701299,
+      "grad_norm": 0.375,
+      "learning_rate": 0.0002,
+      "loss": 0.9054,
+      "step": 3128
+    },
+    {
+      "epoch": 2.4995004995004995,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0002,
+      "loss": 0.8938,
+      "step": 3129
+    },
+    {
+      "epoch": 2.5002997002997,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.0002,
+      "loss": 0.9121,
+      "step": 3130
+    },
+    {
+      "epoch": 2.501098901098901,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9115,
+      "step": 3131
+    },
+    {
+      "epoch": 2.501898101898102,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0002,
+      "loss": 0.9084,
+      "step": 3132
+    },
+    {
+      "epoch": 2.5026973026973027,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0002,
+      "loss": 0.908,
+      "step": 3133
+    },
+    {
+      "epoch": 2.5034965034965033,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9091,
+      "step": 3134
+    },
+    {
+      "epoch": 2.5042957042957044,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0002,
+      "loss": 0.9073,
+      "step": 3135
+    },
+    {
+      "epoch": 2.505094905094905,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0002,
+      "loss": 0.9113,
+      "step": 3136
+    },
+    {
+      "epoch": 2.505894105894106,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.0002,
+      "loss": 0.9024,
+      "step": 3137
+    },
+    {
+      "epoch": 2.5066933066933066,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0002,
+      "loss": 0.9101,
+      "step": 3138
+    },
+    {
+      "epoch": 2.5074925074925076,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0002,
+      "loss": 0.909,
+      "step": 3139
+    },
+    {
+      "epoch": 2.508291708291708,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0002,
+      "loss": 0.9137,
+      "step": 3140
+    },
+    {
+      "epoch": 2.509090909090909,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9081,
+      "step": 3141
+    },
+    {
+      "epoch": 2.50989010989011,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.0002,
+      "loss": 0.9096,
+      "step": 3142
+    },
+    {
+      "epoch": 2.510689310689311,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.0002,
+      "loss": 0.9037,
+      "step": 3143
+    },
+    {
+      "epoch": 2.5114885114885115,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.0002,
+      "loss": 0.9068,
+      "step": 3144
+    },
+    {
+      "epoch": 2.512287712287712,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0002,
+      "loss": 0.9092,
+      "step": 3145
+    },
+    {
+      "epoch": 2.513086913086913,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0002,
+      "loss": 0.9023,
+      "step": 3146
+    },
+    {
+      "epoch": 2.513886113886114,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0002,
+      "loss": 0.9054,
+      "step": 3147
+    },
+    {
+      "epoch": 2.5146853146853148,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9001,
+      "step": 3148
+    },
+    {
+      "epoch": 2.5154845154845153,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.0002,
+      "loss": 0.9116,
+      "step": 3149
+    },
+    {
+      "epoch": 2.5162837162837164,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0002,
+      "loss": 0.9163,
+      "step": 3150
+    },
+    {
+      "epoch": 2.517082917082917,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0002,
+      "loss": 0.9094,
+      "step": 3151
+    },
+    {
+      "epoch": 2.517882117882118,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.0002,
+      "loss": 0.9104,
+      "step": 3152
+    },
+    {
+      "epoch": 2.5186813186813186,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.0002,
+      "loss": 0.9096,
+      "step": 3153
+    },
+    {
+      "epoch": 2.5194805194805197,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0002,
+      "loss": 0.902,
+      "step": 3154
+    },
+    {
+      "epoch": 2.5202797202797202,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0002,
+      "loss": 0.9045,
+      "step": 3155
+    },
+    {
+      "epoch": 2.521078921078921,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0002,
+      "loss": 0.9082,
+      "step": 3156
+    },
+    {
+      "epoch": 2.521878121878122,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9197,
+      "step": 3157
+    },
+    {
+      "epoch": 2.522677322677323,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.0002,
+      "loss": 0.9163,
+      "step": 3158
+    },
+    {
+      "epoch": 2.5234765234765235,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.0002,
+      "loss": 0.9103,
+      "step": 3159
+    },
+    {
+      "epoch": 2.524275724275724,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9108,
+      "step": 3160
+    },
+    {
+      "epoch": 2.525074925074925,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9078,
+      "step": 3161
+    },
+    {
+      "epoch": 2.5258741258741257,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0002,
+      "loss": 0.9439,
+      "step": 3162
+    },
+    {
+      "epoch": 2.526673326673327,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.0002,
+      "loss": 0.9032,
+      "step": 3163
+    },
+    {
+      "epoch": 2.5274725274725274,
+      "grad_norm": 3.171875,
+      "learning_rate": 0.0002,
+      "loss": 0.926,
+      "step": 3164
+    },
+    {
+      "epoch": 2.5282717282717284,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.0002,
+      "loss": 0.912,
+      "step": 3165
+    },
+    {
+      "epoch": 2.529070929070929,
+      "grad_norm": 3.15625,
+      "learning_rate": 0.0002,
+      "loss": 0.9224,
+      "step": 3166
+    },
+    {
+      "epoch": 2.5298701298701296,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0002,
+      "loss": 0.9057,
+      "step": 3167
+    },
+    {
+      "epoch": 2.5306693306693306,
+      "grad_norm": 1.6640625,
+      "learning_rate": 0.0002,
+      "loss": 0.9171,
+      "step": 3168
+    },
+    {
+      "epoch": 2.5314685314685317,
+      "grad_norm": 1.625,
+      "learning_rate": 0.0002,
+      "loss": 0.927,
+      "step": 3169
+    },
+    {
+      "epoch": 2.5322677322677323,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0002,
+      "loss": 0.9094,
+      "step": 3170
+    },
+    {
+      "epoch": 2.533066933066933,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.0002,
+      "loss": 0.9144,
+      "step": 3171
+    },
+    {
+      "epoch": 2.533866133866134,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.0002,
+      "loss": 0.9225,
+      "step": 3172
+    },
+    {
+      "epoch": 2.5346653346653345,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0002,
+      "loss": 0.9192,
+      "step": 3173
+    },
+    {
+      "epoch": 2.5354645354645355,
+      "grad_norm": 0.90625,
+      "learning_rate": 0.0002,
+      "loss": 0.9137,
+      "step": 3174
+    },
+    {
+      "epoch": 2.536263736263736,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0002,
+      "loss": 0.919,
+      "step": 3175
+    },
+    {
+      "epoch": 2.537062937062937,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.0002,
+      "loss": 0.9173,
+      "step": 3176
+    },
+    {
+      "epoch": 2.5378621378621378,
+      "grad_norm": 1.5859375,
+      "learning_rate": 0.0002,
+      "loss": 0.9199,
+      "step": 3177
+    },
+    {
+      "epoch": 2.538661338661339,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9299,
+      "step": 3178
+    },
+    {
+      "epoch": 2.5394605394605394,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9246,
+      "step": 3179
+    },
+    {
+      "epoch": 2.5402597402597404,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.0002,
+      "loss": 0.923,
+      "step": 3180
+    },
+    {
+      "epoch": 2.541058941058941,
+      "grad_norm": 1.53125,
+      "learning_rate": 0.0002,
+      "loss": 0.9199,
+      "step": 3181
+    },
+    {
+      "epoch": 2.5418581418581416,
+      "grad_norm": 0.86328125,
+      "learning_rate": 0.0002,
+      "loss": 0.9243,
+      "step": 3182
+    },
+    {
+      "epoch": 2.5426573426573427,
+      "grad_norm": 1.5,
+      "learning_rate": 0.0002,
+      "loss": 0.9191,
+      "step": 3183
+    },
+    {
+      "epoch": 2.5434565434565437,
+      "grad_norm": 0.9453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9037,
+      "step": 3184
+    },
+    {
+      "epoch": 2.5442557442557443,
+      "grad_norm": 1.6796875,
+      "learning_rate": 0.0002,
+      "loss": 0.9125,
+      "step": 3185
+    },
+    {
+      "epoch": 2.545054945054945,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.0002,
+      "loss": 0.9214,
+      "step": 3186
+    },
+    {
+      "epoch": 2.545854145854146,
+      "grad_norm": 1.5703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9175,
+      "step": 3187
+    },
+    {
+      "epoch": 2.5466533466533465,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9073,
+      "step": 3188
+    },
+    {
+      "epoch": 2.5474525474525476,
+      "grad_norm": 1.4609375,
+      "learning_rate": 0.0002,
+      "loss": 0.9131,
+      "step": 3189
+    },
+    {
+      "epoch": 2.548251748251748,
+      "grad_norm": 1.265625,
+      "learning_rate": 0.0002,
+      "loss": 0.917,
+      "step": 3190
+    },
+    {
+      "epoch": 2.549050949050949,
+      "grad_norm": 1.4296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9199,
+      "step": 3191
+    },
+    {
+      "epoch": 2.54985014985015,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.0002,
+      "loss": 0.8997,
+      "step": 3192
+    },
+    {
+      "epoch": 2.5506493506493504,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.0002,
+      "loss": 0.915,
+      "step": 3193
+    },
+    {
+      "epoch": 2.5514485514485514,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.0002,
+      "loss": 0.9161,
+      "step": 3194
+    },
+    {
+      "epoch": 2.5522477522477525,
+      "grad_norm": 1.6796875,
+      "learning_rate": 0.0002,
+      "loss": 0.9057,
+      "step": 3195
+    },
+    {
+      "epoch": 2.553046953046953,
+      "grad_norm": 1.671875,
+      "learning_rate": 0.0002,
+      "loss": 0.9144,
+      "step": 3196
+    },
+    {
+      "epoch": 2.5538461538461537,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.0002,
+      "loss": 0.911,
+      "step": 3197
+    },
+    {
+      "epoch": 2.5546453546453547,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.0002,
+      "loss": 0.9103,
+      "step": 3198
+    },
+    {
+      "epoch": 2.5554445554445553,
+      "grad_norm": 1.5859375,
+      "learning_rate": 0.0002,
+      "loss": 0.899,
+      "step": 3199
+    },
+    {
+      "epoch": 2.5562437562437563,
+      "grad_norm": 1.4375,
+      "learning_rate": 0.0002,
+      "loss": 0.9234,
+      "step": 3200
+    },
+    {
+      "epoch": 2.557042957042957,
+      "grad_norm": 1.5625,
+      "learning_rate": 0.0002,
+      "loss": 0.9082,
+      "step": 3201
+    },
+    {
+      "epoch": 2.557842157842158,
+      "grad_norm": 1.453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9098,
+      "step": 3202
+    },
+    {
+      "epoch": 2.5586413586413586,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.0002,
+      "loss": 0.909,
+      "step": 3203
+    },
+    {
+      "epoch": 2.5594405594405596,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.0002,
+      "loss": 0.9081,
+      "step": 3204
+    },
+    {
+      "epoch": 2.56023976023976,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.0002,
+      "loss": 0.9078,
+      "step": 3205
+    },
+    {
+      "epoch": 2.5610389610389612,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.0002,
+      "loss": 0.917,
+      "step": 3206
+    },
+    {
+      "epoch": 2.561838161838162,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.0002,
+      "loss": 0.9219,
+      "step": 3207
+    },
+    {
+      "epoch": 2.5626373626373624,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.0002,
+      "loss": 0.9172,
+      "step": 3208
+    },
+    {
+      "epoch": 2.5634365634365635,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.0002,
+      "loss": 0.9092,
+      "step": 3209
+    },
+    {
+      "epoch": 2.5642357642357645,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.0002,
+      "loss": 0.9185,
+      "step": 3210
+    },
+    {
+      "epoch": 2.565034965034965,
+      "grad_norm": 1.4921875,
+      "learning_rate": 0.0002,
+      "loss": 0.9098,
+      "step": 3211
+    },
+    {
+      "epoch": 2.5658341658341657,
+      "grad_norm": 1.265625,
+      "learning_rate": 0.0002,
+      "loss": 0.9067,
+      "step": 3212
+    },
+    {
+      "epoch": 2.5666333666333667,
+      "grad_norm": 1.734375,
+      "learning_rate": 0.0002,
+      "loss": 0.9161,
+      "step": 3213
+    },
+    {
+      "epoch": 2.5674325674325673,
+      "grad_norm": 1.453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9122,
+      "step": 3214
+    },
+    {
+      "epoch": 2.5682317682317684,
+      "grad_norm": 1.3984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9104,
+      "step": 3215
+    },
+    {
+      "epoch": 2.569030969030969,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.0002,
+      "loss": 0.9175,
+      "step": 3216
+    },
+    {
+      "epoch": 2.56983016983017,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9114,
+      "step": 3217
+    },
+    {
+      "epoch": 2.5706293706293706,
+      "grad_norm": 0.98828125,
+      "learning_rate": 0.0002,
+      "loss": 0.9242,
+      "step": 3218
+    },
+    {
+      "epoch": 2.571428571428571,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.0002,
+      "loss": 0.903,
+      "step": 3219
+    },
+    {
+      "epoch": 2.572227772227772,
+      "grad_norm": 0.94921875,
+      "learning_rate": 0.0002,
+      "loss": 0.9154,
+      "step": 3220
+    },
+    {
+      "epoch": 2.5730269730269733,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.0002,
+      "loss": 0.9097,
+      "step": 3221
+    },
+    {
+      "epoch": 2.573826173826174,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.0002,
+      "loss": 0.9129,
+      "step": 3222
+    },
+    {
+      "epoch": 2.5746253746253744,
+      "grad_norm": 1.765625,
+      "learning_rate": 0.0002,
+      "loss": 0.9102,
+      "step": 3223
+    },
+    {
+      "epoch": 2.5754245754245755,
+      "grad_norm": 1.5390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9071,
+      "step": 3224
+    },
+    {
+      "epoch": 2.576223776223776,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.0002,
+      "loss": 0.9085,
+      "step": 3225
+    },
+    {
+      "epoch": 2.577022977022977,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.0002,
+      "loss": 0.9039,
+      "step": 3226
+    },
+    {
+      "epoch": 2.5778221778221777,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.0002,
+      "loss": 0.9109,
+      "step": 3227
+    },
+    {
+      "epoch": 2.5786213786213787,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.0002,
+      "loss": 0.9112,
+      "step": 3228
+    },
+    {
+      "epoch": 2.5794205794205793,
+      "grad_norm": 1.46875,
+      "learning_rate": 0.0002,
+      "loss": 0.9158,
+      "step": 3229
+    },
+    {
+      "epoch": 2.5802197802197804,
+      "grad_norm": 1.3515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9156,
+      "step": 3230
+    },
+    {
+      "epoch": 2.581018981018981,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.0002,
+      "loss": 0.9044,
+      "step": 3231
+    },
+    {
+      "epoch": 2.581818181818182,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.0002,
+      "loss": 0.919,
+      "step": 3232
+    },
+    {
+      "epoch": 2.5826173826173826,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.0002,
+      "loss": 0.9144,
+      "step": 3233
+    },
+    {
+      "epoch": 2.583416583416583,
+      "grad_norm": 1.625,
+      "learning_rate": 0.0002,
+      "loss": 0.9297,
+      "step": 3234
+    },
+    {
+      "epoch": 2.5842157842157842,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0002,
+      "loss": 0.9101,
+      "step": 3235
+    },
+    {
+      "epoch": 2.5850149850149853,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.0002,
+      "loss": 0.9062,
+      "step": 3236
+    },
+    {
+      "epoch": 2.585814185814186,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9026,
+      "step": 3237
+    },
+    {
+      "epoch": 2.5866133866133865,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9085,
+      "step": 3238
+    },
+    {
+      "epoch": 2.5874125874125875,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0002,
+      "loss": 0.9043,
+      "step": 3239
+    },
+    {
+      "epoch": 2.588211788211788,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.0002,
+      "loss": 0.9053,
+      "step": 3240
+    },
+    {
+      "epoch": 2.589010989010989,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0002,
+      "loss": 0.9041,
+      "step": 3241
+    },
+    {
+      "epoch": 2.5898101898101897,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0002,
+      "loss": 0.9175,
+      "step": 3242
+    },
+    {
+      "epoch": 2.5906093906093908,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0002,
+      "loss": 0.9038,
+      "step": 3243
+    },
+    {
+      "epoch": 2.5914085914085914,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.0002,
+      "loss": 0.9085,
+      "step": 3244
+    },
+    {
+      "epoch": 2.592207792207792,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0002,
+      "loss": 0.9057,
+      "step": 3245
+    },
+    {
+      "epoch": 2.593006993006993,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0002,
+      "loss": 0.904,
+      "step": 3246
+    },
+    {
+      "epoch": 2.593806193806194,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9066,
+      "step": 3247
+    },
+    {
+      "epoch": 2.5946053946053946,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9144,
+      "step": 3248
+    },
+    {
+      "epoch": 2.5954045954045952,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9121,
+      "step": 3249
+    },
+    {
+      "epoch": 2.5962037962037963,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0002,
+      "loss": 0.9115,
+      "step": 3250
+    },
+    {
+      "epoch": 2.597002997002997,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.0002,
+      "loss": 0.9405,
+      "step": 3251
+    },
+    {
+      "epoch": 2.597802197802198,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9081,
+      "step": 3252
+    },
+    {
+      "epoch": 2.5986013986013985,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0002,
+      "loss": 0.9134,
+      "step": 3253
+    },
+    {
+      "epoch": 2.5994005994005995,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9119,
+      "step": 3254
+    },
+    {
+      "epoch": 2.6001998001998,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0002,
+      "loss": 0.9016,
+      "step": 3255
+    },
+    {
+      "epoch": 2.600999000999001,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0002,
+      "loss": 0.9128,
+      "step": 3256
+    },
+    {
+      "epoch": 2.6017982017982018,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0002,
+      "loss": 0.9074,
+      "step": 3257
+    },
+    {
+      "epoch": 2.602597402597403,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0002,
+      "loss": 0.9077,
+      "step": 3258
+    },
+    {
+      "epoch": 2.6033966033966034,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9071,
+      "step": 3259
+    },
+    {
+      "epoch": 2.604195804195804,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0002,
+      "loss": 0.9148,
+      "step": 3260
+    },
+    {
+      "epoch": 2.604995004995005,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0002,
+      "loss": 0.9159,
+      "step": 3261
+    },
+    {
+      "epoch": 2.605794205794206,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9107,
+      "step": 3262
+    },
+    {
+      "epoch": 2.6065934065934067,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0002,
+      "loss": 0.9233,
+      "step": 3263
+    },
+    {
+      "epoch": 2.6073926073926073,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0002,
+      "loss": 0.9012,
+      "step": 3264
+    },
+    {
+      "epoch": 2.6081918081918083,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0002,
+      "loss": 0.9068,
+      "step": 3265
+    },
+    {
+      "epoch": 2.608991008991009,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9042,
+      "step": 3266
+    },
+    {
+      "epoch": 2.60979020979021,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0002,
+      "loss": 0.9161,
+      "step": 3267
+    },
+    {
+      "epoch": 2.6105894105894105,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.0002,
+      "loss": 0.9039,
+      "step": 3268
+    },
+    {
+      "epoch": 2.6113886113886116,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0002,
+      "loss": 0.9029,
+      "step": 3269
+    },
+    {
+      "epoch": 2.612187812187812,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9076,
+      "step": 3270
+    },
+    {
+      "epoch": 2.6129870129870127,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9025,
+      "step": 3271
+    },
+    {
+      "epoch": 2.613786213786214,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.0002,
+      "loss": 0.9038,
+      "step": 3272
+    },
+    {
+      "epoch": 2.614585414585415,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.0002,
+      "loss": 0.9086,
+      "step": 3273
+    },
+    {
+      "epoch": 2.6153846153846154,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0002,
+      "loss": 0.912,
+      "step": 3274
+    },
+    {
+      "epoch": 2.616183816183816,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0002,
+      "loss": 0.9069,
+      "step": 3275
+    },
+    {
+      "epoch": 2.616983016983017,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9199,
+      "step": 3276
+    },
+    {
+      "epoch": 2.6177822177822176,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9082,
+      "step": 3277
+    },
+    {
+      "epoch": 2.6185814185814187,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9052,
+      "step": 3278
+    },
+    {
+      "epoch": 2.6193806193806193,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0002,
+      "loss": 0.9118,
+      "step": 3279
+    },
+    {
+      "epoch": 2.6201798201798203,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0002,
+      "loss": 0.9015,
+      "step": 3280
+    },
+    {
+      "epoch": 2.620979020979021,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0002,
+      "loss": 0.8982,
+      "step": 3281
+    },
+    {
+      "epoch": 2.6217782217782215,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9061,
+      "step": 3282
+    },
+    {
+      "epoch": 2.6225774225774225,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8993,
+      "step": 3283
+    },
+    {
+      "epoch": 2.6233766233766236,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0002,
+      "loss": 0.9013,
+      "step": 3284
+    },
+    {
+      "epoch": 2.624175824175824,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0002,
+      "loss": 0.8978,
+      "step": 3285
+    },
+    {
+      "epoch": 2.6249750249750248,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.0002,
+      "loss": 0.898,
+      "step": 3286
+    },
+    {
+      "epoch": 2.625774225774226,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0002,
+      "loss": 0.9036,
+      "step": 3287
+    },
+    {
+      "epoch": 2.626573426573427,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0002,
+      "loss": 0.9077,
+      "step": 3288
+    },
+    {
+      "epoch": 2.6273726273726274,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0002,
+      "loss": 0.903,
+      "step": 3289
+    },
+    {
+      "epoch": 2.628171828171828,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9045,
+      "step": 3290
+    },
+    {
+      "epoch": 2.628971028971029,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9066,
+      "step": 3291
+    },
+    {
+      "epoch": 2.6297702297702297,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0002,
+      "loss": 0.9074,
+      "step": 3292
+    },
+    {
+      "epoch": 2.6305694305694307,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9119,
+      "step": 3293
+    },
+    {
+      "epoch": 2.6313686313686313,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9108,
+      "step": 3294
+    },
+    {
+      "epoch": 2.6321678321678323,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.0002,
+      "loss": 0.9174,
+      "step": 3295
+    },
+    {
+      "epoch": 2.632967032967033,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.0002,
+      "loss": 0.906,
+      "step": 3296
+    },
+    {
+      "epoch": 2.6337662337662335,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.0002,
+      "loss": 0.91,
+      "step": 3297
+    },
+    {
+      "epoch": 2.6345654345654346,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.0002,
+      "loss": 0.9065,
+      "step": 3298
+    },
+    {
+      "epoch": 2.6353646353646356,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0002,
+      "loss": 0.9039,
+      "step": 3299
+    },
+    {
+      "epoch": 2.636163836163836,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8979,
+      "step": 3300
+    },
+    {
+      "epoch": 2.636963036963037,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0002,
+      "loss": 0.9037,
+      "step": 3301
+    },
+    {
+      "epoch": 2.637762237762238,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0002,
+      "loss": 0.9102,
+      "step": 3302
+    },
+    {
+      "epoch": 2.6385614385614384,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.0002,
+      "loss": 0.9027,
+      "step": 3303
+    },
+    {
+      "epoch": 2.6393606393606395,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9123,
+      "step": 3304
+    },
+    {
+      "epoch": 2.64015984015984,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9073,
+      "step": 3305
+    },
+    {
+      "epoch": 2.640959040959041,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.0002,
+      "loss": 0.9097,
+      "step": 3306
+    },
+    {
+      "epoch": 2.6417582417582417,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9141,
+      "step": 3307
+    },
+    {
+      "epoch": 2.6425574425574423,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.0002,
+      "loss": 0.9044,
+      "step": 3308
+    },
+    {
+      "epoch": 2.6433566433566433,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.0002,
+      "loss": 0.9025,
+      "step": 3309
+    },
+    {
+      "epoch": 2.6441558441558444,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9032,
+      "step": 3310
+    },
+    {
+      "epoch": 2.644955044955045,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0002,
+      "loss": 0.9068,
+      "step": 3311
+    },
+    {
+      "epoch": 2.6457542457542456,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.0002,
+      "loss": 0.9041,
+      "step": 3312
+    },
+    {
+      "epoch": 2.6465534465534466,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0002,
+      "loss": 0.9076,
+      "step": 3313
+    },
+    {
+      "epoch": 2.647352647352647,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0002,
+      "loss": 0.9076,
+      "step": 3314
+    },
+    {
+      "epoch": 2.6481518481518482,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0002,
+      "loss": 0.9005,
+      "step": 3315
+    },
+    {
+      "epoch": 2.648951048951049,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0002,
+      "loss": 0.9041,
+      "step": 3316
+    },
+    {
+      "epoch": 2.64975024975025,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0002,
+      "loss": 0.9031,
+      "step": 3317
+    },
+    {
+      "epoch": 2.6505494505494505,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9,
+      "step": 3318
+    },
+    {
+      "epoch": 2.6513486513486515,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0002,
+      "loss": 0.9043,
+      "step": 3319
+    },
+    {
+      "epoch": 2.652147852147852,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0002,
+      "loss": 0.9032,
+      "step": 3320
+    },
+    {
+      "epoch": 2.652947052947053,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0002,
+      "loss": 0.9006,
+      "step": 3321
+    },
+    {
+      "epoch": 2.6537462537462537,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0002,
+      "loss": 0.9109,
+      "step": 3322
+    },
+    {
+      "epoch": 2.6545454545454543,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0002,
+      "loss": 0.9077,
+      "step": 3323
+    },
+    {
+      "epoch": 2.6553446553446554,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9051,
+      "step": 3324
+    },
+    {
+      "epoch": 2.6561438561438564,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9146,
+      "step": 3325
+    },
+    {
+      "epoch": 2.656943056943057,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0002,
+      "loss": 0.9061,
+      "step": 3326
+    },
+    {
+      "epoch": 2.6577422577422576,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9174,
+      "step": 3327
+    },
+    {
+      "epoch": 2.6585414585414586,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9077,
+      "step": 3328
+    },
+    {
+      "epoch": 2.659340659340659,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9087,
+      "step": 3329
+    },
+    {
+      "epoch": 2.6601398601398603,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0002,
+      "loss": 0.9081,
+      "step": 3330
+    },
+    {
+      "epoch": 2.660939060939061,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.0002,
+      "loss": 0.917,
+      "step": 3331
+    },
+    {
+      "epoch": 2.661738261738262,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0002,
+      "loss": 0.903,
+      "step": 3332
+    },
+    {
+      "epoch": 2.6625374625374625,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9169,
+      "step": 3333
+    },
+    {
+      "epoch": 2.663336663336663,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0002,
+      "loss": 0.9182,
+      "step": 3334
+    },
+    {
+      "epoch": 2.664135864135864,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.0002,
+      "loss": 0.8975,
+      "step": 3335
+    },
+    {
+      "epoch": 2.664935064935065,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0002,
+      "loss": 0.9023,
+      "step": 3336
+    },
+    {
+      "epoch": 2.6657342657342658,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0002,
+      "loss": 0.9082,
+      "step": 3337
+    },
+    {
+      "epoch": 2.6665334665334663,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0002,
+      "loss": 0.908,
+      "step": 3338
+    },
+    {
+      "epoch": 2.6673326673326674,
+      "grad_norm": 0.8984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9265,
+      "step": 3339
+    },
+    {
+      "epoch": 2.668131868131868,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9038,
+      "step": 3340
+    },
+    {
+      "epoch": 2.668931068931069,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9046,
+      "step": 3341
+    },
+    {
+      "epoch": 2.6697302697302696,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0002,
+      "loss": 0.9078,
+      "step": 3342
+    },
+    {
+      "epoch": 2.6705294705294707,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9036,
+      "step": 3343
+    },
+    {
+      "epoch": 2.6713286713286712,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0002,
+      "loss": 0.9018,
+      "step": 3344
+    },
+    {
+      "epoch": 2.6721278721278723,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0002,
+      "loss": 0.9078,
+      "step": 3345
+    },
+    {
+      "epoch": 2.672927072927073,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9144,
+      "step": 3346
+    },
+    {
+      "epoch": 2.673726273726274,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0002,
+      "loss": 0.915,
+      "step": 3347
+    },
+    {
+      "epoch": 2.6745254745254745,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0002,
+      "loss": 0.9117,
+      "step": 3348
+    },
+    {
+      "epoch": 2.675324675324675,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9114,
+      "step": 3349
+    },
+    {
+      "epoch": 2.676123876123876,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.0002,
+      "loss": 0.9086,
+      "step": 3350
+    },
+    {
+      "epoch": 2.676923076923077,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9114,
+      "step": 3351
+    },
+    {
+      "epoch": 2.6777222777222778,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0002,
+      "loss": 0.911,
+      "step": 3352
+    },
+    {
+      "epoch": 2.6785214785214784,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9133,
+      "step": 3353
+    },
+    {
+      "epoch": 2.6793206793206794,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0002,
+      "loss": 0.9107,
+      "step": 3354
+    },
+    {
+      "epoch": 2.68011988011988,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9043,
+      "step": 3355
+    },
+    {
+      "epoch": 2.680919080919081,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0002,
+      "loss": 0.9088,
+      "step": 3356
+    },
+    {
+      "epoch": 2.6817182817182816,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0002,
+      "loss": 0.9069,
+      "step": 3357
+    },
+    {
+      "epoch": 2.6825174825174827,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9124,
+      "step": 3358
+    },
+    {
+      "epoch": 2.6833166833166833,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9163,
+      "step": 3359
+    },
+    {
+      "epoch": 2.684115884115884,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.0002,
+      "loss": 0.9076,
+      "step": 3360
+    },
+    {
+      "epoch": 2.684915084915085,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9266,
+      "step": 3361
+    },
+    {
+      "epoch": 2.685714285714286,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0002,
+      "loss": 0.9091,
+      "step": 3362
+    },
+    {
+      "epoch": 2.6865134865134865,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0002,
+      "loss": 0.8926,
+      "step": 3363
+    },
+    {
+      "epoch": 2.687312687312687,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9107,
+      "step": 3364
+    },
+    {
+      "epoch": 2.688111888111888,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0002,
+      "loss": 0.9177,
+      "step": 3365
+    },
+    {
+      "epoch": 2.6889110889110888,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9087,
+      "step": 3366
+    },
+    {
+      "epoch": 2.68971028971029,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0002,
+      "loss": 0.9047,
+      "step": 3367
+    },
+    {
+      "epoch": 2.6905094905094904,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9081,
+      "step": 3368
+    },
+    {
+      "epoch": 2.6913086913086914,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0002,
+      "loss": 0.9046,
+      "step": 3369
+    },
+    {
+      "epoch": 2.692107892107892,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0002,
+      "loss": 0.9098,
+      "step": 3370
+    },
+    {
+      "epoch": 2.692907092907093,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0002,
+      "loss": 0.9038,
+      "step": 3371
+    },
+    {
+      "epoch": 2.6937062937062937,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0002,
+      "loss": 0.9305,
+      "step": 3372
+    },
+    {
+      "epoch": 2.6945054945054947,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0002,
+      "loss": 0.9079,
+      "step": 3373
+    },
+    {
+      "epoch": 2.6953046953046953,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0002,
+      "loss": 0.8961,
+      "step": 3374
+    },
+    {
+      "epoch": 2.696103896103896,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0002,
+      "loss": 0.9004,
+      "step": 3375
+    },
+    {
+      "epoch": 2.696903096903097,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0002,
+      "loss": 0.9053,
+      "step": 3376
+    },
+    {
+      "epoch": 2.697702297702298,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9016,
+      "step": 3377
+    },
+    {
+      "epoch": 2.6985014985014986,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0002,
+      "loss": 0.9017,
+      "step": 3378
+    },
+    {
+      "epoch": 2.699300699300699,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0002,
+      "loss": 0.907,
+      "step": 3379
+    },
+    {
+      "epoch": 2.7000999000999,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0002,
+      "loss": 0.9,
+      "step": 3380
+    },
+    {
+      "epoch": 2.700899100899101,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0002,
+      "loss": 0.9146,
+      "step": 3381
+    },
+    {
+      "epoch": 2.701698301698302,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0002,
+      "loss": 0.9189,
+      "step": 3382
+    },
+    {
+      "epoch": 2.7024975024975024,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0002,
+      "loss": 0.9047,
+      "step": 3383
+    },
+    {
+      "epoch": 2.7032967032967035,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0002,
+      "loss": 0.9053,
+      "step": 3384
+    },
+    {
+      "epoch": 2.704095904095904,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0002,
+      "loss": 0.904,
+      "step": 3385
+    },
+    {
+      "epoch": 2.7048951048951047,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0002,
+      "loss": 0.9067,
+      "step": 3386
+    },
+    {
+      "epoch": 2.7056943056943057,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0002,
+      "loss": 0.9043,
+      "step": 3387
+    },
+    {
+      "epoch": 2.7064935064935067,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8959,
+      "step": 3388
+    },
+    {
+      "epoch": 2.7072927072927073,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0002,
+      "loss": 0.9065,
+      "step": 3389
+    },
+    {
+      "epoch": 2.708091908091908,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0002,
+      "loss": 0.9083,
+      "step": 3390
+    },
+    {
+      "epoch": 2.708891108891109,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0002,
+      "loss": 0.9056,
+      "step": 3391
+    },
+    {
+      "epoch": 2.7096903096903096,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0002,
+      "loss": 0.9002,
+      "step": 3392
+    },
+    {
+      "epoch": 2.7104895104895106,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0002,
+      "loss": 0.9084,
+      "step": 3393
+    },
+    {
+      "epoch": 2.711288711288711,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0002,
+      "loss": 0.9012,
+      "step": 3394
+    },
+    {
+      "epoch": 2.7120879120879122,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0002,
+      "loss": 0.8984,
+      "step": 3395
+    },
+    {
+      "epoch": 2.712887112887113,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0002,
+      "loss": 0.9126,
+      "step": 3396
+    },
+    {
+      "epoch": 2.7136863136863134,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0002,
+      "loss": 0.8956,
+      "step": 3397
+    },
+    {
+      "epoch": 2.7144855144855145,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.0002,
+      "loss": 0.9061,
+      "step": 3398
+    },
+    {
+      "epoch": 2.7152847152847155,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0002,
+      "loss": 0.903,
+      "step": 3399
+    },
+    {
+      "epoch": 2.716083916083916,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9156,
+      "step": 3400
+    },
+    {
+      "epoch": 2.7168831168831167,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0002,
+      "loss": 0.898,
+      "step": 3401
+    },
+    {
+      "epoch": 2.7176823176823177,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0002,
+      "loss": 0.9051,
+      "step": 3402
+    },
+    {
+      "epoch": 2.7184815184815188,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9125,
+      "step": 3403
+    },
+    {
+      "epoch": 2.7192807192807193,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9096,
+      "step": 3404
+    },
+    {
+      "epoch": 2.72007992007992,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0002,
+      "loss": 0.903,
+      "step": 3405
+    },
+    {
+      "epoch": 2.720879120879121,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0002,
+      "loss": 0.9,
+      "step": 3406
+    },
+    {
+      "epoch": 2.7216783216783216,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0002,
+      "loss": 0.9118,
+      "step": 3407
+    },
+    {
+      "epoch": 2.7224775224775226,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0002,
+      "loss": 0.9037,
+      "step": 3408
+    },
+    {
+      "epoch": 2.723276723276723,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0002,
+      "loss": 0.9019,
+      "step": 3409
+    },
+    {
+      "epoch": 2.7240759240759242,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0002,
+      "loss": 0.9125,
+      "step": 3410
+    },
+    {
+      "epoch": 2.724875124875125,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0002,
+      "loss": 0.908,
+      "step": 3411
+    },
+    {
+      "epoch": 2.7256743256743254,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0002,
+      "loss": 0.9035,
+      "step": 3412
+    },
+    {
+      "epoch": 2.7264735264735265,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0002,
+      "loss": 0.9152,
+      "step": 3413
+    },
+    {
+      "epoch": 2.7272727272727275,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9066,
+      "step": 3414
+    },
+    {
+      "epoch": 2.728071928071928,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.0002,
+      "loss": 0.8981,
+      "step": 3415
+    },
+    {
+      "epoch": 2.7288711288711287,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0002,
+      "loss": 0.9034,
+      "step": 3416
+    },
+    {
+      "epoch": 2.7296703296703297,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0002,
+      "loss": 0.9116,
+      "step": 3417
+    },
+    {
+      "epoch": 2.7304695304695303,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.0002,
+      "loss": 0.9054,
+      "step": 3418
+    },
+    {
+      "epoch": 2.7312687312687314,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8967,
+      "step": 3419
+    },
+    {
+      "epoch": 2.732067932067932,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9052,
+      "step": 3420
+    },
+    {
+      "epoch": 2.732867132867133,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0002,
+      "loss": 0.9091,
+      "step": 3421
+    },
+    {
+      "epoch": 2.7336663336663336,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0002,
+      "loss": 0.908,
+      "step": 3422
+    },
+    {
+      "epoch": 2.734465534465534,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0002,
+      "loss": 0.9014,
+      "step": 3423
+    },
+    {
+      "epoch": 2.7352647352647352,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.0002,
+      "loss": 0.9065,
+      "step": 3424
+    },
+    {
+      "epoch": 2.7360639360639363,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0002,
+      "loss": 0.8942,
+      "step": 3425
+    },
+    {
+      "epoch": 2.736863136863137,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9082,
+      "step": 3426
+    },
+    {
+      "epoch": 2.7376623376623375,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9015,
+      "step": 3427
+    },
+    {
+      "epoch": 2.7384615384615385,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0002,
+      "loss": 0.908,
+      "step": 3428
+    },
+    {
+      "epoch": 2.739260739260739,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9036,
+      "step": 3429
+    },
+    {
+      "epoch": 2.74005994005994,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0002,
+      "loss": 0.9054,
+      "step": 3430
+    },
+    {
+      "epoch": 2.7408591408591407,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0002,
+      "loss": 0.9028,
+      "step": 3431
+    },
+    {
+      "epoch": 2.7416583416583418,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9029,
+      "step": 3432
+    },
+    {
+      "epoch": 2.7424575424575424,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0002,
+      "loss": 0.9073,
+      "step": 3433
+    },
+    {
+      "epoch": 2.7432567432567434,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0002,
+      "loss": 0.9002,
+      "step": 3434
+    },
+    {
+      "epoch": 2.744055944055944,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0002,
+      "loss": 0.8937,
+      "step": 3435
+    },
+    {
+      "epoch": 2.744855144855145,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0002,
+      "loss": 0.9035,
+      "step": 3436
+    },
+    {
+      "epoch": 2.7456543456543456,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0002,
+      "loss": 0.9023,
+      "step": 3437
+    },
+    {
+      "epoch": 2.7464535464535462,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.0002,
+      "loss": 0.9061,
+      "step": 3438
+    },
+    {
+      "epoch": 2.7472527472527473,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0002,
+      "loss": 0.8982,
+      "step": 3439
+    },
+    {
+      "epoch": 2.7480519480519483,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0002,
+      "loss": 0.9115,
+      "step": 3440
+    },
+    {
+      "epoch": 2.748851148851149,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9108,
+      "step": 3441
+    },
+    {
+      "epoch": 2.7496503496503495,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9152,
+      "step": 3442
+    },
+    {
+      "epoch": 2.7504495504495505,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9063,
+      "step": 3443
+    },
+    {
+      "epoch": 2.751248751248751,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0002,
+      "loss": 0.9074,
+      "step": 3444
+    },
+    {
+      "epoch": 2.752047952047952,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0002,
+      "loss": 0.9117,
+      "step": 3445
+    },
+    {
+      "epoch": 2.7528471528471528,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0002,
+      "loss": 0.9097,
+      "step": 3446
+    },
+    {
+      "epoch": 2.753646353646354,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9057,
+      "step": 3447
+    },
+    {
+      "epoch": 2.7544455544455544,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0002,
+      "loss": 0.8985,
+      "step": 3448
+    },
+    {
+      "epoch": 2.755244755244755,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0002,
+      "loss": 0.9104,
+      "step": 3449
+    },
+    {
+      "epoch": 2.756043956043956,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0002,
+      "loss": 0.8983,
+      "step": 3450
+    },
+    {
+      "epoch": 2.756843156843157,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0002,
+      "loss": 0.904,
+      "step": 3451
+    },
+    {
+      "epoch": 2.7576423576423577,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.0002,
+      "loss": 0.9123,
+      "step": 3452
+    },
+    {
+      "epoch": 2.7584415584415583,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9073,
+      "step": 3453
+    },
+    {
+      "epoch": 2.7592407592407593,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.0002,
+      "loss": 0.9118,
+      "step": 3454
+    },
+    {
+      "epoch": 2.76003996003996,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0002,
+      "loss": 0.916,
+      "step": 3455
+    },
+    {
+      "epoch": 2.760839160839161,
+      "grad_norm": 0.375,
+      "learning_rate": 0.0002,
+      "loss": 0.9152,
+      "step": 3456
+    },
+    {
+      "epoch": 2.7616383616383615,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0002,
+      "loss": 0.8962,
+      "step": 3457
+    },
+    {
+      "epoch": 2.7624375624375626,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9091,
+      "step": 3458
+    },
+    {
+      "epoch": 2.763236763236763,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0002,
+      "loss": 0.9136,
+      "step": 3459
+    },
+    {
+      "epoch": 2.764035964035964,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.0002,
+      "loss": 0.9068,
+      "step": 3460
+    },
+    {
+      "epoch": 2.764835164835165,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9122,
+      "step": 3461
+    },
+    {
+      "epoch": 2.765634365634366,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9011,
+      "step": 3462
+    },
+    {
+      "epoch": 2.7664335664335664,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9006,
+      "step": 3463
+    },
+    {
+      "epoch": 2.767232767232767,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.0002,
+      "loss": 0.906,
+      "step": 3464
+    },
+    {
+      "epoch": 2.768031968031968,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9116,
+      "step": 3465
+    },
+    {
+      "epoch": 2.768831168831169,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.0002,
+      "loss": 0.9059,
+      "step": 3466
+    },
+    {
+      "epoch": 2.7696303696303697,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9128,
+      "step": 3467
+    },
+    {
+      "epoch": 2.7704295704295703,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9039,
+      "step": 3468
+    },
+    {
+      "epoch": 2.7712287712287713,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0002,
+      "loss": 0.9055,
+      "step": 3469
+    },
+    {
+      "epoch": 2.772027972027972,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0002,
+      "loss": 0.8985,
+      "step": 3470
+    },
+    {
+      "epoch": 2.772827172827173,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.0002,
+      "loss": 0.9035,
+      "step": 3471
+    },
+    {
+      "epoch": 2.7736263736263735,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0002,
+      "loss": 0.9197,
+      "step": 3472
+    },
+    {
+      "epoch": 2.7744255744255746,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0002,
+      "loss": 0.9045,
+      "step": 3473
+    },
+    {
+      "epoch": 2.775224775224775,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.0002,
+      "loss": 0.9036,
+      "step": 3474
+    },
+    {
+      "epoch": 2.7760239760239758,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0002,
+      "loss": 0.9025,
+      "step": 3475
+    },
+    {
+      "epoch": 2.776823176823177,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0002,
+      "loss": 0.8941,
+      "step": 3476
+    },
+    {
+      "epoch": 2.777622377622378,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9039,
+      "step": 3477
+    },
+    {
+      "epoch": 2.7784215784215784,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9005,
+      "step": 3478
+    },
+    {
+      "epoch": 2.779220779220779,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0002,
+      "loss": 0.907,
+      "step": 3479
+    },
+    {
+      "epoch": 2.78001998001998,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0002,
+      "loss": 0.9046,
+      "step": 3480
+    },
+    {
+      "epoch": 2.7808191808191807,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0002,
+      "loss": 0.9013,
+      "step": 3481
+    },
+    {
+      "epoch": 2.7816183816183817,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8991,
+      "step": 3482
+    },
+    {
+      "epoch": 2.7824175824175823,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0002,
+      "loss": 0.9183,
+      "step": 3483
+    },
+    {
+      "epoch": 2.7832167832167833,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8973,
+      "step": 3484
+    },
+    {
+      "epoch": 2.784015984015984,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0002,
+      "loss": 0.9025,
+      "step": 3485
+    },
+    {
+      "epoch": 2.784815184815185,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0002,
+      "loss": 0.9098,
+      "step": 3486
+    },
+    {
+      "epoch": 2.7856143856143856,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0002,
+      "loss": 0.9513,
+      "step": 3487
+    },
+    {
+      "epoch": 2.7864135864135866,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0002,
+      "loss": 0.8985,
+      "step": 3488
+    },
+    {
+      "epoch": 2.787212787212787,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8936,
+      "step": 3489
+    },
+    {
+      "epoch": 2.788011988011988,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9126,
+      "step": 3490
+    },
+    {
+      "epoch": 2.788811188811189,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0002,
+      "loss": 0.9105,
+      "step": 3491
+    },
+    {
+      "epoch": 2.78961038961039,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0002,
+      "loss": 0.904,
+      "step": 3492
+    },
+    {
+      "epoch": 2.7904095904095905,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9004,
+      "step": 3493
+    },
+    {
+      "epoch": 2.791208791208791,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0002,
+      "loss": 0.908,
+      "step": 3494
+    },
+    {
+      "epoch": 2.792007992007992,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0002,
+      "loss": 0.8967,
+      "step": 3495
+    },
+    {
+      "epoch": 2.7928071928071927,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9023,
+      "step": 3496
+    },
+    {
+      "epoch": 2.7936063936063937,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9005,
+      "step": 3497
+    },
+    {
+      "epoch": 2.7944055944055943,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9097,
+      "step": 3498
+    },
+    {
+      "epoch": 2.7952047952047954,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0002,
+      "loss": 0.9009,
+      "step": 3499
+    },
+    {
+      "epoch": 2.796003996003996,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0002,
+      "loss": 0.908,
+      "step": 3500
+    },
+    {
+      "epoch": 2.7968031968031966,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8971,
+      "step": 3501
+    },
+    {
+      "epoch": 2.7976023976023976,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.0002,
+      "loss": 0.9012,
+      "step": 3502
+    },
+    {
+      "epoch": 2.7984015984015986,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8933,
+      "step": 3503
+    },
+    {
+      "epoch": 2.7992007992007992,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0002,
+      "loss": 0.9113,
+      "step": 3504
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0002,
+      "loss": 0.9094,
+      "step": 3505
+    },
+    {
+      "epoch": 2.800799200799201,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8987,
+      "step": 3506
+    },
+    {
+      "epoch": 2.8015984015984015,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0002,
+      "loss": 0.9045,
+      "step": 3507
+    },
+    {
+      "epoch": 2.8023976023976025,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9056,
+      "step": 3508
+    },
+    {
+      "epoch": 2.803196803196803,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0002,
+      "loss": 0.9013,
+      "step": 3509
+    },
+    {
+      "epoch": 2.803996003996004,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0002,
+      "loss": 0.9035,
+      "step": 3510
+    },
+    {
+      "epoch": 2.8047952047952047,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0002,
+      "loss": 0.9097,
+      "step": 3511
+    },
+    {
+      "epoch": 2.8055944055944058,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0002,
+      "loss": 0.9052,
+      "step": 3512
+    },
+    {
+      "epoch": 2.8063936063936064,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0002,
+      "loss": 0.9078,
+      "step": 3513
+    },
+    {
+      "epoch": 2.8071928071928074,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.0002,
+      "loss": 0.9078,
+      "step": 3514
+    },
+    {
+      "epoch": 2.807992007992008,
+      "grad_norm": 3.515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9342,
+      "step": 3515
+    },
+    {
+      "epoch": 2.8087912087912086,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9,
+      "step": 3516
+    },
+    {
+      "epoch": 2.8095904095904096,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0002,
+      "loss": 0.899,
+      "step": 3517
+    },
+    {
+      "epoch": 2.8103896103896107,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.0002,
+      "loss": 0.9218,
+      "step": 3518
+    },
+    {
+      "epoch": 2.8111888111888113,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0002,
+      "loss": 0.9091,
+      "step": 3519
+    },
+    {
+      "epoch": 2.811988011988012,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9161,
+      "step": 3520
+    },
+    {
+      "epoch": 2.812787212787213,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9132,
+      "step": 3521
+    },
+    {
+      "epoch": 2.8135864135864135,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0002,
+      "loss": 0.9081,
+      "step": 3522
+    },
+    {
+      "epoch": 2.8143856143856145,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0002,
+      "loss": 0.9084,
+      "step": 3523
+    },
+    {
+      "epoch": 2.815184815184815,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0002,
+      "loss": 0.9126,
+      "step": 3524
+    },
+    {
+      "epoch": 2.815984015984016,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.0002,
+      "loss": 0.9102,
+      "step": 3525
+    },
+    {
+      "epoch": 2.8167832167832167,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0002,
+      "loss": 0.9122,
+      "step": 3526
+    },
+    {
+      "epoch": 2.8175824175824173,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9182,
+      "step": 3527
+    },
+    {
+      "epoch": 2.8183816183816184,
+      "grad_norm": 0.375,
+      "learning_rate": 0.0002,
+      "loss": 0.9144,
+      "step": 3528
+    },
+    {
+      "epoch": 2.8191808191808194,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0002,
+      "loss": 0.9048,
+      "step": 3529
+    },
+    {
+      "epoch": 2.81998001998002,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.0002,
+      "loss": 0.8958,
+      "step": 3530
+    },
+    {
+      "epoch": 2.8207792207792206,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.0002,
+      "loss": 0.8953,
+      "step": 3531
+    },
+    {
+      "epoch": 2.8215784215784216,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0002,
+      "loss": 0.9006,
+      "step": 3532
+    },
+    {
+      "epoch": 2.8223776223776222,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.0002,
+      "loss": 0.9138,
+      "step": 3533
+    },
+    {
+      "epoch": 2.8231768231768233,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9033,
+      "step": 3534
+    },
+    {
+      "epoch": 2.823976023976024,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.0002,
+      "loss": 0.8971,
+      "step": 3535
+    },
+    {
+      "epoch": 2.824775224775225,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9085,
+      "step": 3536
+    },
+    {
+      "epoch": 2.8255744255744255,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.0002,
+      "loss": 0.9351,
+      "step": 3537
+    },
+    {
+      "epoch": 2.826373626373626,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0002,
+      "loss": 0.9145,
+      "step": 3538
+    },
+    {
+      "epoch": 2.827172827172827,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0002,
+      "loss": 0.9123,
+      "step": 3539
+    },
+    {
+      "epoch": 2.827972027972028,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0002,
+      "loss": 0.9005,
+      "step": 3540
+    },
+    {
+      "epoch": 2.8287712287712288,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0002,
+      "loss": 0.899,
+      "step": 3541
+    },
+    {
+      "epoch": 2.8295704295704294,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0002,
+      "loss": 0.9079,
+      "step": 3542
+    },
+    {
+      "epoch": 2.8303696303696304,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0002,
+      "loss": 0.9113,
+      "step": 3543
+    },
+    {
+      "epoch": 2.8311688311688314,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9128,
+      "step": 3544
+    },
+    {
+      "epoch": 2.831968031968032,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0002,
+      "loss": 0.9098,
+      "step": 3545
+    },
+    {
+      "epoch": 2.8327672327672326,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0002,
+      "loss": 0.8988,
+      "step": 3546
+    },
+    {
+      "epoch": 2.8335664335664337,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9069,
+      "step": 3547
+    },
+    {
+      "epoch": 2.8343656343656343,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.0002,
+      "loss": 0.9085,
+      "step": 3548
+    },
+    {
+      "epoch": 2.8351648351648353,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9047,
+      "step": 3549
+    },
+    {
+      "epoch": 2.835964035964036,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9055,
+      "step": 3550
+    },
+    {
+      "epoch": 2.836763236763237,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0002,
+      "loss": 0.9026,
+      "step": 3551
+    },
+    {
+      "epoch": 2.8375624375624375,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0002,
+      "loss": 0.9104,
+      "step": 3552
+    },
+    {
+      "epoch": 2.838361638361638,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0002,
+      "loss": 0.9051,
+      "step": 3553
+    },
+    {
+      "epoch": 2.839160839160839,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9042,
+      "step": 3554
+    },
+    {
+      "epoch": 2.83996003996004,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0002,
+      "loss": 0.9052,
+      "step": 3555
+    },
+    {
+      "epoch": 2.840759240759241,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0002,
+      "loss": 0.9086,
+      "step": 3556
+    },
+    {
+      "epoch": 2.8415584415584414,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0002,
+      "loss": 0.9143,
+      "step": 3557
+    },
+    {
+      "epoch": 2.8423576423576424,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.0002,
+      "loss": 0.9062,
+      "step": 3558
+    },
+    {
+      "epoch": 2.843156843156843,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0002,
+      "loss": 0.9147,
+      "step": 3559
+    },
+    {
+      "epoch": 2.843956043956044,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8943,
+      "step": 3560
+    },
+    {
+      "epoch": 2.8447552447552447,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9128,
+      "step": 3561
+    },
+    {
+      "epoch": 2.8455544455544457,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9099,
+      "step": 3562
+    },
+    {
+      "epoch": 2.8463536463536463,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0002,
+      "loss": 0.9171,
+      "step": 3563
+    },
+    {
+      "epoch": 2.847152847152847,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9027,
+      "step": 3564
+    },
+    {
+      "epoch": 2.847952047952048,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0002,
+      "loss": 0.9094,
+      "step": 3565
+    },
+    {
+      "epoch": 2.848751248751249,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.0002,
+      "loss": 0.8953,
+      "step": 3566
+    },
+    {
+      "epoch": 2.8495504495504496,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.0002,
+      "loss": 0.9134,
+      "step": 3567
+    },
+    {
+      "epoch": 2.85034965034965,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.0002,
+      "loss": 0.905,
+      "step": 3568
+    },
+    {
+      "epoch": 2.851148851148851,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0002,
+      "loss": 0.9071,
+      "step": 3569
+    },
+    {
+      "epoch": 2.851948051948052,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9084,
+      "step": 3570
+    },
+    {
+      "epoch": 2.852747252747253,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0002,
+      "loss": 0.8968,
+      "step": 3571
+    },
+    {
+      "epoch": 2.8535464535464534,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9051,
+      "step": 3572
+    },
+    {
+      "epoch": 2.8543456543456545,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0002,
+      "loss": 0.8967,
+      "step": 3573
+    },
+    {
+      "epoch": 2.855144855144855,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0002,
+      "loss": 0.9069,
+      "step": 3574
+    },
+    {
+      "epoch": 2.855944055944056,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0002,
+      "loss": 0.9051,
+      "step": 3575
+    },
+    {
+      "epoch": 2.8567432567432567,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0002,
+      "loss": 0.9114,
+      "step": 3576
+    },
+    {
+      "epoch": 2.8575424575424577,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0002,
+      "loss": 0.9039,
+      "step": 3577
+    },
+    {
+      "epoch": 2.8583416583416583,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9102,
+      "step": 3578
+    },
+    {
+      "epoch": 2.859140859140859,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9058,
+      "step": 3579
+    },
+    {
+      "epoch": 2.85994005994006,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0002,
+      "loss": 0.9001,
+      "step": 3580
+    },
+    {
+      "epoch": 2.860739260739261,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0002,
+      "loss": 0.9001,
+      "step": 3581
+    },
+    {
+      "epoch": 2.8615384615384616,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9023,
+      "step": 3582
+    },
+    {
+      "epoch": 2.862337662337662,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0002,
+      "loss": 0.9034,
+      "step": 3583
+    },
+    {
+      "epoch": 2.863136863136863,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0002,
+      "loss": 0.9108,
+      "step": 3584
+    },
+    {
+      "epoch": 2.863936063936064,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0002,
+      "loss": 0.8997,
+      "step": 3585
+    },
+    {
+      "epoch": 2.864735264735265,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.0002,
+      "loss": 0.9019,
+      "step": 3586
+    },
+    {
+      "epoch": 2.8655344655344654,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0002,
+      "loss": 0.9044,
+      "step": 3587
+    },
+    {
+      "epoch": 2.8663336663336665,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0002,
+      "loss": 0.8962,
+      "step": 3588
+    },
+    {
+      "epoch": 2.867132867132867,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0002,
+      "loss": 0.9064,
+      "step": 3589
+    },
+    {
+      "epoch": 2.8679320679320677,
+      "grad_norm": 0.375,
+      "learning_rate": 0.0002,
+      "loss": 0.9116,
+      "step": 3590
+    },
+    {
+      "epoch": 2.8687312687312687,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9036,
+      "step": 3591
+    },
+    {
+      "epoch": 2.8695304695304698,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.0002,
+      "loss": 0.9096,
+      "step": 3592
+    },
+    {
+      "epoch": 2.8703296703296703,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0002,
+      "loss": 0.9101,
+      "step": 3593
+    },
+    {
+      "epoch": 2.871128871128871,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.0002,
+      "loss": 0.8998,
+      "step": 3594
+    },
+    {
+      "epoch": 2.871928071928072,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0002,
+      "loss": 0.9115,
+      "step": 3595
+    },
+    {
+      "epoch": 2.8727272727272726,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.0002,
+      "loss": 0.9052,
+      "step": 3596
+    },
+    {
+      "epoch": 2.8735264735264736,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0002,
+      "loss": 0.9039,
+      "step": 3597
+    },
+    {
+      "epoch": 2.874325674325674,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9061,
+      "step": 3598
+    },
+    {
+      "epoch": 2.8751248751248752,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0002,
+      "loss": 0.9006,
+      "step": 3599
+    },
+    {
+      "epoch": 2.875924075924076,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.0002,
+      "loss": 0.9033,
+      "step": 3600
+    },
+    {
+      "epoch": 2.876723276723277,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0002,
+      "loss": 0.8953,
+      "step": 3601
+    },
+    {
+      "epoch": 2.8775224775224775,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9106,
+      "step": 3602
+    },
+    {
+      "epoch": 2.8783216783216785,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0002,
+      "loss": 0.9129,
+      "step": 3603
+    },
+    {
+      "epoch": 2.879120879120879,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9147,
+      "step": 3604
+    },
+    {
+      "epoch": 2.8799200799200797,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0002,
+      "loss": 0.9194,
+      "step": 3605
+    },
+    {
+      "epoch": 2.8807192807192807,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9029,
+      "step": 3606
+    },
+    {
+      "epoch": 2.8815184815184818,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9017,
+      "step": 3607
+    },
+    {
+      "epoch": 2.8823176823176824,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0002,
+      "loss": 0.9148,
+      "step": 3608
+    },
+    {
+      "epoch": 2.883116883116883,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0002,
+      "loss": 0.9039,
+      "step": 3609
+    },
+    {
+      "epoch": 2.883916083916084,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.0002,
+      "loss": 0.8941,
+      "step": 3610
+    },
+    {
+      "epoch": 2.8847152847152846,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0002,
+      "loss": 0.9127,
+      "step": 3611
+    },
+    {
+      "epoch": 2.8855144855144856,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.0002,
+      "loss": 0.9068,
+      "step": 3612
+    },
+    {
+      "epoch": 2.8863136863136862,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.0002,
+      "loss": 0.906,
+      "step": 3613
+    },
+    {
+      "epoch": 2.8871128871128873,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9032,
+      "step": 3614
+    },
+    {
+      "epoch": 2.887912087912088,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9098,
+      "step": 3615
+    },
+    {
+      "epoch": 2.8887112887112885,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0002,
+      "loss": 0.9037,
+      "step": 3616
+    },
+    {
+      "epoch": 2.8895104895104895,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8994,
+      "step": 3617
+    },
+    {
+      "epoch": 2.8903096903096905,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9166,
+      "step": 3618
+    },
+    {
+      "epoch": 2.891108891108891,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.0002,
+      "loss": 0.9054,
+      "step": 3619
+    },
+    {
+      "epoch": 2.8919080919080917,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9058,
+      "step": 3620
+    },
+    {
+      "epoch": 2.8927072927072928,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0002,
+      "loss": 0.9011,
+      "step": 3621
+    },
+    {
+      "epoch": 2.8935064935064934,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9064,
+      "step": 3622
+    },
+    {
+      "epoch": 2.8943056943056944,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8991,
+      "step": 3623
+    },
+    {
+      "epoch": 2.895104895104895,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9142,
+      "step": 3624
+    },
+    {
+      "epoch": 2.895904095904096,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.0002,
+      "loss": 0.9064,
+      "step": 3625
+    },
+    {
+      "epoch": 2.8967032967032966,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9075,
+      "step": 3626
+    },
+    {
+      "epoch": 2.8975024975024977,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9026,
+      "step": 3627
+    },
+    {
+      "epoch": 2.8983016983016983,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0002,
+      "loss": 0.9046,
+      "step": 3628
+    },
+    {
+      "epoch": 2.8991008991008993,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.0002,
+      "loss": 0.8956,
+      "step": 3629
+    },
+    {
+      "epoch": 2.8999000999001,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8998,
+      "step": 3630
+    },
+    {
+      "epoch": 2.9006993006993005,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0002,
+      "loss": 0.9112,
+      "step": 3631
+    },
+    {
+      "epoch": 2.9014985014985015,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0002,
+      "loss": 0.9091,
+      "step": 3632
+    },
+    {
+      "epoch": 2.9022977022977026,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.0002,
+      "loss": 0.9045,
+      "step": 3633
+    },
+    {
+      "epoch": 2.903096903096903,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.0002,
+      "loss": 0.928,
+      "step": 3634
+    },
+    {
+      "epoch": 2.9038961038961038,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.0002,
+      "loss": 0.9047,
+      "step": 3635
+    },
+    {
+      "epoch": 2.904695304695305,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.0002,
+      "loss": 0.9082,
+      "step": 3636
+    },
+    {
+      "epoch": 2.9054945054945054,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9071,
+      "step": 3637
+    },
+    {
+      "epoch": 2.9062937062937064,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.0002,
+      "loss": 0.9048,
+      "step": 3638
+    },
+    {
+      "epoch": 2.907092907092907,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9016,
+      "step": 3639
+    },
+    {
+      "epoch": 2.907892107892108,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0002,
+      "loss": 0.8981,
+      "step": 3640
+    },
+    {
+      "epoch": 2.9086913086913087,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.0002,
+      "loss": 0.895,
+      "step": 3641
+    },
+    {
+      "epoch": 2.9094905094905092,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8975,
+      "step": 3642
+    },
+    {
+      "epoch": 2.9102897102897103,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.0002,
+      "loss": 0.9079,
+      "step": 3643
+    },
+    {
+      "epoch": 2.9110889110889113,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.0002,
+      "loss": 0.9133,
+      "step": 3644
+    },
+    {
+      "epoch": 2.911888111888112,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9141,
+      "step": 3645
+    },
+    {
+      "epoch": 2.9126873126873125,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.0002,
+      "loss": 0.9046,
+      "step": 3646
+    },
+    {
+      "epoch": 2.9134865134865136,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9059,
+      "step": 3647
+    },
+    {
+      "epoch": 2.914285714285714,
+      "grad_norm": 0.84765625,
+      "learning_rate": 0.0002,
+      "loss": 0.8975,
+      "step": 3648
+    },
+    {
+      "epoch": 2.915084915084915,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9023,
+      "step": 3649
+    },
+    {
+      "epoch": 2.915884115884116,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.0002,
+      "loss": 0.9074,
+      "step": 3650
+    },
+    {
+      "epoch": 2.916683316683317,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0002,
+      "loss": 0.9029,
+      "step": 3651
+    },
+    {
+      "epoch": 2.9174825174825174,
+      "grad_norm": 0.80859375,
+      "learning_rate": 0.0002,
+      "loss": 0.8952,
+      "step": 3652
+    },
+    {
+      "epoch": 2.918281718281718,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.0002,
+      "loss": 0.9092,
+      "step": 3653
+    },
+    {
+      "epoch": 2.919080919080919,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0002,
+      "loss": 0.9059,
+      "step": 3654
+    },
+    {
+      "epoch": 2.91988011988012,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0002,
+      "loss": 0.908,
+      "step": 3655
+    },
+    {
+      "epoch": 2.9206793206793207,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9093,
+      "step": 3656
+    },
+    {
+      "epoch": 2.9214785214785213,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.0002,
+      "loss": 0.9064,
+      "step": 3657
+    },
+    {
+      "epoch": 2.9222777222777223,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8907,
+      "step": 3658
+    },
+    {
+      "epoch": 2.9230769230769234,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.0002,
+      "loss": 0.9039,
+      "step": 3659
+    },
+    {
+      "epoch": 2.923876123876124,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0002,
+      "loss": 0.9013,
+      "step": 3660
+    },
+    {
+      "epoch": 2.9246753246753245,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0002,
+      "loss": 0.9093,
+      "step": 3661
+    },
+    {
+      "epoch": 2.9254745254745256,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9004,
+      "step": 3662
+    },
+    {
+      "epoch": 2.926273726273726,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9021,
+      "step": 3663
+    },
+    {
+      "epoch": 2.927072927072927,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9062,
+      "step": 3664
+    },
+    {
+      "epoch": 2.927872127872128,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.0002,
+      "loss": 0.9062,
+      "step": 3665
+    },
+    {
+      "epoch": 2.928671328671329,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0002,
+      "loss": 0.9024,
+      "step": 3666
+    },
+    {
+      "epoch": 2.9294705294705294,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9048,
+      "step": 3667
+    },
+    {
+      "epoch": 2.93026973026973,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9078,
+      "step": 3668
+    },
+    {
+      "epoch": 2.931068931068931,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.0002,
+      "loss": 0.9004,
+      "step": 3669
+    },
+    {
+      "epoch": 2.931868131868132,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0002,
+      "loss": 0.907,
+      "step": 3670
+    },
+    {
+      "epoch": 2.9326673326673327,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0002,
+      "loss": 0.8981,
+      "step": 3671
+    },
+    {
+      "epoch": 2.9334665334665333,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0002,
+      "loss": 0.902,
+      "step": 3672
+    },
+    {
+      "epoch": 2.9342657342657343,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9129,
+      "step": 3673
+    },
+    {
+      "epoch": 2.935064935064935,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0002,
+      "loss": 0.9062,
+      "step": 3674
+    },
+    {
+      "epoch": 2.935864135864136,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0002,
+      "loss": 0.909,
+      "step": 3675
+    },
+    {
+      "epoch": 2.9366633366633366,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9071,
+      "step": 3676
+    },
+    {
+      "epoch": 2.9374625374625376,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9117,
+      "step": 3677
+    },
+    {
+      "epoch": 2.938261738261738,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0002,
+      "loss": 0.9068,
+      "step": 3678
+    },
+    {
+      "epoch": 2.939060939060939,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9019,
+      "step": 3679
+    },
+    {
+      "epoch": 2.93986013986014,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0002,
+      "loss": 0.9143,
+      "step": 3680
+    },
+    {
+      "epoch": 2.940659340659341,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0002,
+      "loss": 0.9029,
+      "step": 3681
+    },
+    {
+      "epoch": 2.9414585414585415,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0002,
+      "loss": 0.9105,
+      "step": 3682
+    },
+    {
+      "epoch": 2.942257742257742,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.0002,
+      "loss": 0.9029,
+      "step": 3683
+    },
+    {
+      "epoch": 2.943056943056943,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0002,
+      "loss": 0.9052,
+      "step": 3684
+    },
+    {
+      "epoch": 2.9438561438561437,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9035,
+      "step": 3685
+    },
+    {
+      "epoch": 2.9446553446553447,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0002,
+      "loss": 0.9011,
+      "step": 3686
+    },
+    {
+      "epoch": 2.9454545454545453,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0002,
+      "loss": 0.9057,
+      "step": 3687
+    },
+    {
+      "epoch": 2.9462537462537464,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0002,
+      "loss": 0.904,
+      "step": 3688
+    },
+    {
+      "epoch": 2.947052947052947,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0002,
+      "loss": 0.8955,
+      "step": 3689
+    },
+    {
+      "epoch": 2.947852147852148,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0002,
+      "loss": 0.9096,
+      "step": 3690
+    },
+    {
+      "epoch": 2.9486513486513486,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0002,
+      "loss": 0.9107,
+      "step": 3691
+    },
+    {
+      "epoch": 2.9494505494505496,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9055,
+      "step": 3692
+    },
+    {
+      "epoch": 2.9502497502497502,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8941,
+      "step": 3693
+    },
+    {
+      "epoch": 2.951048951048951,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0002,
+      "loss": 0.8996,
+      "step": 3694
+    },
+    {
+      "epoch": 2.951848151848152,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.0002,
+      "loss": 0.8992,
+      "step": 3695
+    },
+    {
+      "epoch": 2.952647352647353,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9106,
+      "step": 3696
+    },
+    {
+      "epoch": 2.9534465534465535,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0002,
+      "loss": 0.9026,
+      "step": 3697
+    },
+    {
+      "epoch": 2.954245754245754,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0002,
+      "loss": 0.895,
+      "step": 3698
+    },
+    {
+      "epoch": 2.955044955044955,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0002,
+      "loss": 0.9052,
+      "step": 3699
+    },
+    {
+      "epoch": 2.9558441558441557,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9048,
+      "step": 3700
+    },
+    {
+      "epoch": 2.9566433566433568,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0002,
+      "loss": 0.9026,
+      "step": 3701
+    },
+    {
+      "epoch": 2.9574425574425574,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9125,
+      "step": 3702
+    },
+    {
+      "epoch": 2.9582417582417584,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9154,
+      "step": 3703
+    },
+    {
+      "epoch": 2.959040959040959,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0002,
+      "loss": 0.9004,
+      "step": 3704
+    },
+    {
+      "epoch": 2.9598401598401596,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9034,
+      "step": 3705
+    },
+    {
+      "epoch": 2.9606393606393606,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9047,
+      "step": 3706
+    },
+    {
+      "epoch": 2.9614385614385617,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9101,
+      "step": 3707
+    },
+    {
+      "epoch": 2.9622377622377623,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0002,
+      "loss": 0.9159,
+      "step": 3708
+    },
+    {
+      "epoch": 2.963036963036963,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0002,
+      "loss": 0.9079,
+      "step": 3709
+    },
+    {
+      "epoch": 2.963836163836164,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0002,
+      "loss": 0.9142,
+      "step": 3710
+    },
+    {
+      "epoch": 2.9646353646353645,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0002,
+      "loss": 0.9083,
+      "step": 3711
+    },
+    {
+      "epoch": 2.9654345654345655,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9094,
+      "step": 3712
+    },
+    {
+      "epoch": 2.966233766233766,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0002,
+      "loss": 0.9022,
+      "step": 3713
+    },
+    {
+      "epoch": 2.967032967032967,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0002,
+      "loss": 0.902,
+      "step": 3714
+    },
+    {
+      "epoch": 2.9678321678321677,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0002,
+      "loss": 0.9081,
+      "step": 3715
+    },
+    {
+      "epoch": 2.968631368631369,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9016,
+      "step": 3716
+    },
+    {
+      "epoch": 2.9694305694305694,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0002,
+      "loss": 0.9103,
+      "step": 3717
+    },
+    {
+      "epoch": 2.9702297702297704,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002,
+      "loss": 0.901,
+      "step": 3718
+    },
+    {
+      "epoch": 2.971028971028971,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9125,
+      "step": 3719
+    },
+    {
+      "epoch": 2.9718281718281716,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0002,
+      "loss": 0.8949,
+      "step": 3720
+    },
+    {
+      "epoch": 2.9726273726273726,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0002,
+      "loss": 0.902,
+      "step": 3721
+    },
+    {
+      "epoch": 2.9734265734265737,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0002,
+      "loss": 0.9096,
+      "step": 3722
+    },
+    {
+      "epoch": 2.9742257742257743,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0002,
+      "loss": 0.9038,
+      "step": 3723
+    },
+    {
+      "epoch": 2.975024975024975,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0002,
+      "loss": 0.9049,
+      "step": 3724
+    },
+    {
+      "epoch": 2.975824175824176,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0002,
+      "loss": 0.9097,
+      "step": 3725
+    },
+    {
+      "epoch": 2.9766233766233765,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9158,
+      "step": 3726
+    },
+    {
+      "epoch": 2.9774225774225775,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9035,
+      "step": 3727
+    },
+    {
+      "epoch": 2.978221778221778,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0002,
+      "loss": 0.9076,
+      "step": 3728
+    },
+    {
+      "epoch": 2.979020979020979,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0002,
+      "loss": 0.9062,
+      "step": 3729
+    },
+    {
+      "epoch": 2.9798201798201798,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0002,
+      "loss": 0.907,
+      "step": 3730
+    },
+    {
+      "epoch": 2.9806193806193804,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0002,
+      "loss": 0.9123,
+      "step": 3731
+    },
+    {
+      "epoch": 2.9814185814185814,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0002,
+      "loss": 0.9077,
+      "step": 3732
+    },
+    {
+      "epoch": 2.9822177822177824,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0002,
+      "loss": 0.9026,
+      "step": 3733
+    },
+    {
+      "epoch": 2.983016983016983,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0002,
+      "loss": 0.9054,
+      "step": 3734
+    },
+    {
+      "epoch": 2.9838161838161836,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0002,
+      "loss": 0.9063,
+      "step": 3735
+    },
+    {
+      "epoch": 2.9846153846153847,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9054,
+      "step": 3736
+    },
+    {
+      "epoch": 2.9854145854145853,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0002,
+      "loss": 0.9024,
+      "step": 3737
+    },
+    {
+      "epoch": 2.9862137862137863,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0002,
+      "loss": 0.8928,
+      "step": 3738
+    },
+    {
+      "epoch": 2.987012987012987,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8933,
+      "step": 3739
+    },
+    {
+      "epoch": 2.987812187812188,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9067,
+      "step": 3740
+    },
+    {
+      "epoch": 2.9886113886113885,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0002,
+      "loss": 0.8912,
+      "step": 3741
+    },
+    {
+      "epoch": 2.9894105894105896,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9047,
+      "step": 3742
+    },
+    {
+      "epoch": 2.99020979020979,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0002,
+      "loss": 0.9002,
+      "step": 3743
+    },
+    {
+      "epoch": 2.991008991008991,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.0002,
+      "loss": 0.8971,
+      "step": 3744
+    },
+    {
+      "epoch": 2.991808191808192,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0002,
+      "loss": 0.9013,
+      "step": 3745
+    },
+    {
+      "epoch": 2.9926073926073924,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9149,
+      "step": 3746
+    },
+    {
+      "epoch": 2.9934065934065934,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0002,
+      "loss": 0.9081,
+      "step": 3747
+    },
+    {
+      "epoch": 2.9942057942057945,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0002,
+      "loss": 0.9065,
+      "step": 3748
+    },
+    {
+      "epoch": 2.995004995004995,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0002,
+      "loss": 0.9086,
+      "step": 3749
+    },
+    {
+      "epoch": 2.9958041958041957,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0002,
+      "loss": 0.9117,
+      "step": 3750
+    },
+    {
+      "epoch": 2.9966033966033967,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0002,
+      "loss": 0.904,
+      "step": 3751
+    },
+    {
+      "epoch": 2.9974025974025973,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0002,
+      "loss": 0.9008,
+      "step": 3752
+    },
+    {
+      "epoch": 2.9982017982017983,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9092,
+      "step": 3753
+    },
+    {
+      "epoch": 2.999000999000999,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0002,
+      "loss": 0.9067,
+      "step": 3754
+    },
+    {
+      "epoch": 2.9998001998002,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0002,
+      "loss": 0.9064,
+      "step": 3755
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0002,
+      "loss": 0.2235,
+      "step": 3756
+    },
+    {
+      "epoch": 3.0007992007992006,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0002,
+      "loss": 0.9013,
+      "step": 3757
+    },
+    {
+      "epoch": 3.0015984015984016,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8952,
+      "step": 3758
+    },
+    {
+      "epoch": 3.0023976023976022,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8967,
+      "step": 3759
+    },
+    {
+      "epoch": 3.0031968031968033,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0002,
+      "loss": 0.9078,
+      "step": 3760
+    },
+    {
+      "epoch": 3.003996003996004,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8979,
+      "step": 3761
+    },
+    {
+      "epoch": 3.004795204795205,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0002,
+      "loss": 0.9125,
+      "step": 3762
+    },
+    {
+      "epoch": 3.0055944055944055,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9338,
+      "step": 3763
+    },
+    {
+      "epoch": 3.0063936063936065,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0002,
+      "loss": 0.913,
+      "step": 3764
+    },
+    {
+      "epoch": 3.007192807192807,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9082,
+      "step": 3765
+    },
+    {
+      "epoch": 3.007992007992008,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0002,
+      "loss": 0.9012,
+      "step": 3766
+    },
+    {
+      "epoch": 3.0087912087912088,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0002,
+      "loss": 0.9078,
+      "step": 3767
+    },
+    {
+      "epoch": 3.00959040959041,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0002,
+      "loss": 0.9063,
+      "step": 3768
+    },
+    {
+      "epoch": 3.0103896103896104,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0002,
+      "loss": 0.8962,
+      "step": 3769
+    },
+    {
+      "epoch": 3.011188811188811,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0002,
+      "loss": 0.9052,
+      "step": 3770
+    },
+    {
+      "epoch": 3.011988011988012,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0002,
+      "loss": 0.8964,
+      "step": 3771
+    },
+    {
+      "epoch": 3.0127872127872126,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0002,
+      "loss": 0.905,
+      "step": 3772
+    },
+    {
+      "epoch": 3.0135864135864137,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.0002,
+      "loss": 0.9063,
+      "step": 3773
+    },
+    {
+      "epoch": 3.0143856143856143,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0002,
+      "loss": 0.904,
+      "step": 3774
+    },
+    {
+      "epoch": 3.0151848151848153,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0002,
+      "loss": 0.9016,
+      "step": 3775
+    },
+    {
+      "epoch": 3.015984015984016,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0002,
+      "loss": 0.9103,
+      "step": 3776
+    },
+    {
+      "epoch": 3.016783216783217,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.0002,
+      "loss": 0.9057,
+      "step": 3777
+    },
+    {
+      "epoch": 3.0175824175824175,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0002,
+      "loss": 0.8959,
+      "step": 3778
+    },
+    {
+      "epoch": 3.0183816183816186,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0002,
+      "loss": 0.904,
+      "step": 3779
+    },
+    {
+      "epoch": 3.019180819180819,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.0002,
+      "loss": 0.8969,
+      "step": 3780
+    },
+    {
+      "epoch": 3.01998001998002,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0002,
+      "loss": 0.9029,
+      "step": 3781
+    },
+    {
+      "epoch": 3.020779220779221,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9073,
+      "step": 3782
+    },
+    {
+      "epoch": 3.0215784215784214,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.0002,
+      "loss": 0.8972,
+      "step": 3783
+    },
+    {
+      "epoch": 3.0223776223776224,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8982,
+      "step": 3784
+    },
+    {
+      "epoch": 3.023176823176823,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9119,
+      "step": 3785
+    },
+    {
+      "epoch": 3.023976023976024,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9002,
+      "step": 3786
+    },
+    {
+      "epoch": 3.0247752247752246,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.0002,
+      "loss": 0.9001,
+      "step": 3787
+    },
+    {
+      "epoch": 3.0255744255744257,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9025,
+      "step": 3788
+    },
+    {
+      "epoch": 3.0263736263736263,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0002,
+      "loss": 0.9034,
+      "step": 3789
+    },
+    {
+      "epoch": 3.0271728271728273,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0002,
+      "loss": 0.9021,
+      "step": 3790
+    },
+    {
+      "epoch": 3.027972027972028,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9096,
+      "step": 3791
+    },
+    {
+      "epoch": 3.028771228771229,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8997,
+      "step": 3792
+    },
+    {
+      "epoch": 3.0295704295704295,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9137,
+      "step": 3793
+    },
+    {
+      "epoch": 3.0303696303696306,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8992,
+      "step": 3794
+    },
+    {
+      "epoch": 3.031168831168831,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9024,
+      "step": 3795
+    },
+    {
+      "epoch": 3.0319680319680318,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9032,
+      "step": 3796
+    },
+    {
+      "epoch": 3.032767232767233,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0002,
+      "loss": 0.9047,
+      "step": 3797
+    },
+    {
+      "epoch": 3.0335664335664334,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0002,
+      "loss": 0.8962,
+      "step": 3798
+    },
+    {
+      "epoch": 3.0343656343656344,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0002,
+      "loss": 0.9093,
+      "step": 3799
+    },
+    {
+      "epoch": 3.035164835164835,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0002,
+      "loss": 0.9037,
+      "step": 3800
+    },
+    {
+      "epoch": 3.035964035964036,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9078,
+      "step": 3801
+    },
+    {
+      "epoch": 3.0367632367632367,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.0002,
+      "loss": 0.8978,
+      "step": 3802
+    },
+    {
+      "epoch": 3.0375624375624377,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9019,
+      "step": 3803
+    },
+    {
+      "epoch": 3.0383616383616383,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9087,
+      "step": 3804
+    },
+    {
+      "epoch": 3.0391608391608393,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9103,
+      "step": 3805
+    },
+    {
+      "epoch": 3.03996003996004,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9038,
+      "step": 3806
+    },
+    {
+      "epoch": 3.040759240759241,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0002,
+      "loss": 0.9022,
+      "step": 3807
+    },
+    {
+      "epoch": 3.0415584415584416,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0002,
+      "loss": 0.9098,
+      "step": 3808
+    },
+    {
+      "epoch": 3.042357642357642,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9072,
+      "step": 3809
+    },
+    {
+      "epoch": 3.043156843156843,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0002,
+      "loss": 0.9029,
+      "step": 3810
+    },
+    {
+      "epoch": 3.043956043956044,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9008,
+      "step": 3811
+    },
+    {
+      "epoch": 3.044755244755245,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0002,
+      "loss": 0.905,
+      "step": 3812
+    },
+    {
+      "epoch": 3.0455544455544454,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0002,
+      "loss": 0.9057,
+      "step": 3813
+    },
+    {
+      "epoch": 3.0463536463536465,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0002,
+      "loss": 0.9092,
+      "step": 3814
+    },
+    {
+      "epoch": 3.047152847152847,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0002,
+      "loss": 0.8959,
+      "step": 3815
+    },
+    {
+      "epoch": 3.047952047952048,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0002,
+      "loss": 0.9038,
+      "step": 3816
+    },
+    {
+      "epoch": 3.0487512487512487,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0002,
+      "loss": 0.8951,
+      "step": 3817
+    },
+    {
+      "epoch": 3.0495504495504497,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0002,
+      "loss": 0.899,
+      "step": 3818
+    },
+    {
+      "epoch": 3.0503496503496503,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0002,
+      "loss": 0.8981,
+      "step": 3819
+    },
+    {
+      "epoch": 3.0511488511488514,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0002,
+      "loss": 0.8954,
+      "step": 3820
+    },
+    {
+      "epoch": 3.051948051948052,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0002,
+      "loss": 0.9078,
+      "step": 3821
+    },
+    {
+      "epoch": 3.0527472527472526,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9124,
+      "step": 3822
+    },
+    {
+      "epoch": 3.0535464535464536,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9084,
+      "step": 3823
+    },
+    {
+      "epoch": 3.054345654345654,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0002,
+      "loss": 0.8965,
+      "step": 3824
+    },
+    {
+      "epoch": 3.0551448551448552,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0002,
+      "loss": 0.8995,
+      "step": 3825
+    },
+    {
+      "epoch": 3.055944055944056,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0002,
+      "loss": 0.9084,
+      "step": 3826
+    },
+    {
+      "epoch": 3.056743256743257,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0002,
+      "loss": 0.9105,
+      "step": 3827
+    },
+    {
+      "epoch": 3.0575424575424575,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0002,
+      "loss": 0.908,
+      "step": 3828
+    },
+    {
+      "epoch": 3.0583416583416585,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0002,
+      "loss": 0.8983,
+      "step": 3829
+    },
+    {
+      "epoch": 3.059140859140859,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9006,
+      "step": 3830
+    },
+    {
+      "epoch": 3.05994005994006,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9004,
+      "step": 3831
+    },
+    {
+      "epoch": 3.0607392607392607,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.0002,
+      "loss": 0.9051,
+      "step": 3832
+    },
+    {
+      "epoch": 3.0615384615384613,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0002,
+      "loss": 0.908,
+      "step": 3833
+    },
+    {
+      "epoch": 3.0623376623376624,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.0002,
+      "loss": 0.9118,
+      "step": 3834
+    },
+    {
+      "epoch": 3.063136863136863,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0002,
+      "loss": 0.8986,
+      "step": 3835
+    },
+    {
+      "epoch": 3.063936063936064,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0002,
+      "loss": 0.9059,
+      "step": 3836
+    },
+    {
+      "epoch": 3.0647352647352646,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0002,
+      "loss": 0.9118,
+      "step": 3837
+    },
+    {
+      "epoch": 3.0655344655344656,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.0002,
+      "loss": 0.9027,
+      "step": 3838
+    },
+    {
+      "epoch": 3.066333666333666,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8936,
+      "step": 3839
+    },
+    {
+      "epoch": 3.0671328671328673,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.0002,
+      "loss": 0.8998,
+      "step": 3840
+    },
+    {
+      "epoch": 3.067932067932068,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0002,
+      "loss": 0.9081,
+      "step": 3841
+    },
+    {
+      "epoch": 3.068731268731269,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9061,
+      "step": 3842
+    },
+    {
+      "epoch": 3.0695304695304695,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9033,
+      "step": 3843
+    },
+    {
+      "epoch": 3.0703296703296705,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0002,
+      "loss": 0.8976,
+      "step": 3844
+    },
+    {
+      "epoch": 3.071128871128871,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0002,
+      "loss": 0.9039,
+      "step": 3845
+    },
+    {
+      "epoch": 3.071928071928072,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8905,
+      "step": 3846
+    },
+    {
+      "epoch": 3.0727272727272728,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9098,
+      "step": 3847
+    },
+    {
+      "epoch": 3.0735264735264733,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.0002,
+      "loss": 0.9038,
+      "step": 3848
+    },
+    {
+      "epoch": 3.0743256743256744,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9039,
+      "step": 3849
+    },
+    {
+      "epoch": 3.075124875124875,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9121,
+      "step": 3850
+    },
+    {
+      "epoch": 3.075924075924076,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8984,
+      "step": 3851
+    },
+    {
+      "epoch": 3.0767232767232766,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9017,
+      "step": 3852
+    },
+    {
+      "epoch": 3.0775224775224777,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9036,
+      "step": 3853
+    },
+    {
+      "epoch": 3.0783216783216782,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9013,
+      "step": 3854
+    },
+    {
+      "epoch": 3.0791208791208793,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.0002,
+      "loss": 0.9051,
+      "step": 3855
+    },
+    {
+      "epoch": 3.07992007992008,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0002,
+      "loss": 0.9064,
+      "step": 3856
+    },
+    {
+      "epoch": 3.080719280719281,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0002,
+      "loss": 0.9047,
+      "step": 3857
+    },
+    {
+      "epoch": 3.0815184815184815,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.0002,
+      "loss": 0.9066,
+      "step": 3858
+    },
+    {
+      "epoch": 3.082317682317682,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0002,
+      "loss": 0.9025,
+      "step": 3859
+    },
+    {
+      "epoch": 3.083116883116883,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0002,
+      "loss": 0.9027,
+      "step": 3860
+    },
+    {
+      "epoch": 3.0839160839160837,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0002,
+      "loss": 0.906,
+      "step": 3861
+    },
+    {
+      "epoch": 3.0847152847152848,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9042,
+      "step": 3862
+    },
+    {
+      "epoch": 3.0855144855144854,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0002,
+      "loss": 0.9062,
+      "step": 3863
+    },
+    {
+      "epoch": 3.0863136863136864,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9104,
+      "step": 3864
+    },
+    {
+      "epoch": 3.087112887112887,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9097,
+      "step": 3865
+    },
+    {
+      "epoch": 3.087912087912088,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0002,
+      "loss": 0.8993,
+      "step": 3866
+    },
+    {
+      "epoch": 3.0887112887112886,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9022,
+      "step": 3867
+    },
+    {
+      "epoch": 3.0895104895104897,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0002,
+      "loss": 0.8994,
+      "step": 3868
+    },
+    {
+      "epoch": 3.0903096903096903,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0002,
+      "loss": 0.9021,
+      "step": 3869
+    },
+    {
+      "epoch": 3.0911088911088913,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0002,
+      "loss": 0.907,
+      "step": 3870
+    },
+    {
+      "epoch": 3.091908091908092,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0002,
+      "loss": 0.8985,
+      "step": 3871
+    },
+    {
+      "epoch": 3.0927072927072925,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0002,
+      "loss": 0.9033,
+      "step": 3872
+    },
+    {
+      "epoch": 3.0935064935064935,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0002,
+      "loss": 0.9078,
+      "step": 3873
+    },
+    {
+      "epoch": 3.094305694305694,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9138,
+      "step": 3874
+    },
+    {
+      "epoch": 3.095104895104895,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8889,
+      "step": 3875
+    },
+    {
+      "epoch": 3.0959040959040958,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0002,
+      "loss": 0.9123,
+      "step": 3876
+    },
+    {
+      "epoch": 3.096703296703297,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0002,
+      "loss": 0.9041,
+      "step": 3877
+    },
+    {
+      "epoch": 3.0975024975024974,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.0002,
+      "loss": 0.9024,
+      "step": 3878
+    },
+    {
+      "epoch": 3.0983016983016984,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0002,
+      "loss": 0.906,
+      "step": 3879
+    },
+    {
+      "epoch": 3.099100899100899,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.0002,
+      "loss": 0.8995,
+      "step": 3880
+    },
+    {
+      "epoch": 3.0999000999001,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0002,
+      "loss": 0.9098,
+      "step": 3881
+    },
+    {
+      "epoch": 3.1006993006993007,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0002,
+      "loss": 0.9003,
+      "step": 3882
+    },
+    {
+      "epoch": 3.1014985014985017,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0002,
+      "loss": 0.9053,
+      "step": 3883
+    },
+    {
+      "epoch": 3.1022977022977023,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0002,
+      "loss": 0.9004,
+      "step": 3884
+    },
+    {
+      "epoch": 3.103096903096903,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0002,
+      "loss": 0.897,
+      "step": 3885
+    },
+    {
+      "epoch": 3.103896103896104,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9046,
+      "step": 3886
+    },
+    {
+      "epoch": 3.1046953046953045,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8951,
+      "step": 3887
+    },
+    {
+      "epoch": 3.1054945054945056,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.0002,
+      "loss": 0.8985,
+      "step": 3888
+    },
+    {
+      "epoch": 3.106293706293706,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0002,
+      "loss": 0.9048,
+      "step": 3889
+    },
+    {
+      "epoch": 3.107092907092907,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9063,
+      "step": 3890
+    },
+    {
+      "epoch": 3.107892107892108,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0002,
+      "loss": 0.9038,
+      "step": 3891
+    },
+    {
+      "epoch": 3.108691308691309,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9069,
+      "step": 3892
+    },
+    {
+      "epoch": 3.1094905094905094,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9158,
+      "step": 3893
+    },
+    {
+      "epoch": 3.1102897102897105,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9141,
+      "step": 3894
+    },
+    {
+      "epoch": 3.111088911088911,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9052,
+      "step": 3895
+    },
+    {
+      "epoch": 3.111888111888112,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0002,
+      "loss": 0.8953,
+      "step": 3896
+    },
+    {
+      "epoch": 3.1126873126873127,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9022,
+      "step": 3897
+    },
+    {
+      "epoch": 3.1134865134865133,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0002,
+      "loss": 0.8997,
+      "step": 3898
+    },
+    {
+      "epoch": 3.1142857142857143,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0002,
+      "loss": 0.9065,
+      "step": 3899
+    },
+    {
+      "epoch": 3.115084915084915,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0002,
+      "loss": 0.897,
+      "step": 3900
+    },
+    {
+      "epoch": 3.115884115884116,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9084,
+      "step": 3901
+    },
+    {
+      "epoch": 3.1166833166833166,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0002,
+      "loss": 0.8978,
+      "step": 3902
+    },
+    {
+      "epoch": 3.1174825174825176,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0002,
+      "loss": 0.8943,
+      "step": 3903
+    },
+    {
+      "epoch": 3.118281718281718,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.0002,
+      "loss": 0.9041,
+      "step": 3904
+    },
+    {
+      "epoch": 3.1190809190809192,
+      "grad_norm": 0.625,
+      "learning_rate": 0.0002,
+      "loss": 0.8988,
+      "step": 3905
+    },
+    {
+      "epoch": 3.11988011988012,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0002,
+      "loss": 0.905,
+      "step": 3906
+    },
+    {
+      "epoch": 3.120679320679321,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9052,
+      "step": 3907
+    },
+    {
+      "epoch": 3.1214785214785215,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0002,
+      "loss": 0.9074,
+      "step": 3908
+    },
+    {
+      "epoch": 3.1222777222777225,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9119,
+      "step": 3909
+    },
+    {
+      "epoch": 3.123076923076923,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0002,
+      "loss": 0.9027,
+      "step": 3910
+    },
+    {
+      "epoch": 3.1238761238761237,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0002,
+      "loss": 0.9056,
+      "step": 3911
+    },
+    {
+      "epoch": 3.1246753246753247,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0002,
+      "loss": 0.8986,
+      "step": 3912
+    },
+    {
+      "epoch": 3.1254745254745253,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9026,
+      "step": 3913
+    },
+    {
+      "epoch": 3.1262737262737263,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0002,
+      "loss": 0.907,
+      "step": 3914
+    },
+    {
+      "epoch": 3.127072927072927,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8925,
+      "step": 3915
+    },
+    {
+      "epoch": 3.127872127872128,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.0002,
+      "loss": 0.9012,
+      "step": 3916
+    },
+    {
+      "epoch": 3.1286713286713286,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0002,
+      "loss": 0.9027,
+      "step": 3917
+    },
+    {
+      "epoch": 3.1294705294705296,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0002,
+      "loss": 0.8977,
+      "step": 3918
+    },
+    {
+      "epoch": 3.13026973026973,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0002,
+      "loss": 0.8962,
+      "step": 3919
+    },
+    {
+      "epoch": 3.1310689310689312,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0002,
+      "loss": 0.896,
+      "step": 3920
+    },
+    {
+      "epoch": 3.131868131868132,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9052,
+      "step": 3921
+    },
+    {
+      "epoch": 3.1326673326673324,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0002,
+      "loss": 0.9017,
+      "step": 3922
+    },
+    {
+      "epoch": 3.1334665334665335,
+      "grad_norm": 0.625,
+      "learning_rate": 0.0002,
+      "loss": 0.9057,
+      "step": 3923
+    },
+    {
+      "epoch": 3.134265734265734,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0002,
+      "loss": 0.9017,
+      "step": 3924
+    },
+    {
+      "epoch": 3.135064935064935,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.0002,
+      "loss": 0.9006,
+      "step": 3925
+    },
+    {
+      "epoch": 3.1358641358641357,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0002,
+      "loss": 0.9044,
+      "step": 3926
+    },
+    {
+      "epoch": 3.1366633366633367,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0002,
+      "loss": 0.9057,
+      "step": 3927
+    },
+    {
+      "epoch": 3.1374625374625373,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0002,
+      "loss": 0.9084,
+      "step": 3928
+    },
+    {
+      "epoch": 3.1382617382617384,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8954,
+      "step": 3929
+    },
+    {
+      "epoch": 3.139060939060939,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.0002,
+      "loss": 0.8974,
+      "step": 3930
+    },
+    {
+      "epoch": 3.13986013986014,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9042,
+      "step": 3931
+    },
+    {
+      "epoch": 3.1406593406593406,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.0002,
+      "loss": 0.9015,
+      "step": 3932
+    },
+    {
+      "epoch": 3.1414585414585416,
+      "grad_norm": 0.375,
+      "learning_rate": 0.0002,
+      "loss": 0.9018,
+      "step": 3933
+    },
+    {
+      "epoch": 3.1422577422577422,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9032,
+      "step": 3934
+    },
+    {
+      "epoch": 3.1430569430569433,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9094,
+      "step": 3935
+    },
+    {
+      "epoch": 3.143856143856144,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0002,
+      "loss": 0.9043,
+      "step": 3936
+    },
+    {
+      "epoch": 3.1446553446553445,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9082,
+      "step": 3937
+    },
+    {
+      "epoch": 3.1454545454545455,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0002,
+      "loss": 0.8952,
+      "step": 3938
+    },
+    {
+      "epoch": 3.146253746253746,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0002,
+      "loss": 0.8997,
+      "step": 3939
+    },
+    {
+      "epoch": 3.147052947052947,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0002,
+      "loss": 0.9047,
+      "step": 3940
+    },
+    {
+      "epoch": 3.1478521478521477,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9117,
+      "step": 3941
+    },
+    {
+      "epoch": 3.1486513486513488,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0002,
+      "loss": 0.904,
+      "step": 3942
+    },
+    {
+      "epoch": 3.1494505494505494,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9071,
+      "step": 3943
+    },
+    {
+      "epoch": 3.1502497502497504,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0002,
+      "loss": 0.9092,
+      "step": 3944
+    },
+    {
+      "epoch": 3.151048951048951,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9075,
+      "step": 3945
+    },
+    {
+      "epoch": 3.151848151848152,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9046,
+      "step": 3946
+    },
+    {
+      "epoch": 3.1526473526473526,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9037,
+      "step": 3947
+    },
+    {
+      "epoch": 3.1534465534465532,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0002,
+      "loss": 0.899,
+      "step": 3948
+    },
+    {
+      "epoch": 3.1542457542457543,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0002,
+      "loss": 0.899,
+      "step": 3949
+    },
+    {
+      "epoch": 3.155044955044955,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8966,
+      "step": 3950
+    },
+    {
+      "epoch": 3.155844155844156,
+      "grad_norm": 1.265625,
+      "learning_rate": 0.0002,
+      "loss": 0.9416,
+      "step": 3951
+    },
+    {
+      "epoch": 3.1566433566433565,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0002,
+      "loss": 0.9028,
+      "step": 3952
+    },
+    {
+      "epoch": 3.1574425574425575,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0002,
+      "loss": 0.9036,
+      "step": 3953
+    },
+    {
+      "epoch": 3.158241758241758,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.0002,
+      "loss": 0.9051,
+      "step": 3954
+    },
+    {
+      "epoch": 3.159040959040959,
+      "grad_norm": 1.453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9249,
+      "step": 3955
+    },
+    {
+      "epoch": 3.1598401598401598,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9126,
+      "step": 3956
+    },
+    {
+      "epoch": 3.160639360639361,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8988,
+      "step": 3957
+    },
+    {
+      "epoch": 3.1614385614385614,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9032,
+      "step": 3958
+    },
+    {
+      "epoch": 3.1622377622377624,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0002,
+      "loss": 0.8973,
+      "step": 3959
+    },
+    {
+      "epoch": 3.163036963036963,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9055,
+      "step": 3960
+    },
+    {
+      "epoch": 3.163836163836164,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0002,
+      "loss": 0.8992,
+      "step": 3961
+    },
+    {
+      "epoch": 3.1646353646353647,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0002,
+      "loss": 0.9075,
+      "step": 3962
+    },
+    {
+      "epoch": 3.1654345654345653,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0002,
+      "loss": 0.9101,
+      "step": 3963
+    },
+    {
+      "epoch": 3.1662337662337663,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8952,
+      "step": 3964
+    },
+    {
+      "epoch": 3.167032967032967,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9037,
+      "step": 3965
+    },
+    {
+      "epoch": 3.167832167832168,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0002,
+      "loss": 0.9018,
+      "step": 3966
+    },
+    {
+      "epoch": 3.1686313686313685,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8988,
+      "step": 3967
+    },
+    {
+      "epoch": 3.1694305694305696,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0002,
+      "loss": 0.9023,
+      "step": 3968
+    },
+    {
+      "epoch": 3.17022977022977,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.0002,
+      "loss": 0.9126,
+      "step": 3969
+    },
+    {
+      "epoch": 3.171028971028971,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8983,
+      "step": 3970
+    },
+    {
+      "epoch": 3.171828171828172,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0002,
+      "loss": 0.9094,
+      "step": 3971
+    },
+    {
+      "epoch": 3.172627372627373,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0002,
+      "loss": 0.904,
+      "step": 3972
+    },
+    {
+      "epoch": 3.1734265734265734,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0002,
+      "loss": 0.9044,
+      "step": 3973
+    },
+    {
+      "epoch": 3.174225774225774,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8991,
+      "step": 3974
+    },
+    {
+      "epoch": 3.175024975024975,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0002,
+      "loss": 0.9039,
+      "step": 3975
+    },
+    {
+      "epoch": 3.1758241758241756,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0002,
+      "loss": 0.9042,
+      "step": 3976
+    },
+    {
+      "epoch": 3.1766233766233767,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0002,
+      "loss": 0.9111,
+      "step": 3977
+    },
+    {
+      "epoch": 3.1774225774225773,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0002,
+      "loss": 0.9005,
+      "step": 3978
+    },
+    {
+      "epoch": 3.1782217782217783,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0002,
+      "loss": 0.8956,
+      "step": 3979
+    },
+    {
+      "epoch": 3.179020979020979,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0002,
+      "loss": 0.9036,
+      "step": 3980
+    },
+    {
+      "epoch": 3.17982017982018,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0002,
+      "loss": 0.9121,
+      "step": 3981
+    },
+    {
+      "epoch": 3.1806193806193805,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0002,
+      "loss": 0.9038,
+      "step": 3982
+    },
+    {
+      "epoch": 3.1814185814185816,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9076,
+      "step": 3983
+    },
+    {
+      "epoch": 3.182217782217782,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0002,
+      "loss": 0.9046,
+      "step": 3984
+    },
+    {
+      "epoch": 3.183016983016983,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0002,
+      "loss": 0.9032,
+      "step": 3985
+    },
+    {
+      "epoch": 3.183816183816184,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0002,
+      "loss": 0.9043,
+      "step": 3986
+    },
+    {
+      "epoch": 3.184615384615385,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9101,
+      "step": 3987
+    },
+    {
+      "epoch": 3.1854145854145854,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0002,
+      "loss": 0.9084,
+      "step": 3988
+    },
+    {
+      "epoch": 3.186213786213786,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8975,
+      "step": 3989
+    },
+    {
+      "epoch": 3.187012987012987,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0002,
+      "loss": 0.908,
+      "step": 3990
+    },
+    {
+      "epoch": 3.1878121878121877,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0002,
+      "loss": 0.8973,
+      "step": 3991
+    },
+    {
+      "epoch": 3.1886113886113887,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0002,
+      "loss": 0.9082,
+      "step": 3992
+    },
+    {
+      "epoch": 3.1894105894105893,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0002,
+      "loss": 0.9007,
+      "step": 3993
+    },
+    {
+      "epoch": 3.1902097902097903,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0002,
+      "loss": 0.9067,
+      "step": 3994
+    },
+    {
+      "epoch": 3.191008991008991,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8961,
+      "step": 3995
+    },
+    {
+      "epoch": 3.191808191808192,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0002,
+      "loss": 0.9046,
+      "step": 3996
+    },
+    {
+      "epoch": 3.1926073926073926,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0002,
+      "loss": 0.8946,
+      "step": 3997
+    },
+    {
+      "epoch": 3.1934065934065936,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0002,
+      "loss": 0.9032,
+      "step": 3998
+    },
+    {
+      "epoch": 3.194205794205794,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0002,
+      "loss": 0.9068,
+      "step": 3999
+    },
+    {
+      "epoch": 3.195004995004995,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8906,
+      "step": 4000
+    },
+    {
+      "epoch": 3.195804195804196,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9011,
+      "step": 4001
+    },
+    {
+      "epoch": 3.1966033966033964,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0002,
+      "loss": 0.9029,
+      "step": 4002
+    },
+    {
+      "epoch": 3.1974025974025975,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8908,
+      "step": 4003
+    },
+    {
+      "epoch": 3.198201798201798,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0002,
+      "loss": 0.9105,
+      "step": 4004
+    },
+    {
+      "epoch": 3.199000999000999,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0002,
+      "loss": 0.8945,
+      "step": 4005
+    },
+    {
+      "epoch": 3.1998001998001997,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9085,
+      "step": 4006
+    },
+    {
+      "epoch": 3.2005994005994007,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0002,
+      "loss": 0.913,
+      "step": 4007
+    },
+    {
+      "epoch": 3.2013986013986013,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0002,
+      "loss": 0.9065,
+      "step": 4008
+    },
+    {
+      "epoch": 3.2021978021978024,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9026,
+      "step": 4009
+    },
+    {
+      "epoch": 3.202997002997003,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8964,
+      "step": 4010
+    },
+    {
+      "epoch": 3.203796203796204,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0002,
+      "loss": 0.9025,
+      "step": 4011
+    },
+    {
+      "epoch": 3.2045954045954046,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0002,
+      "loss": 0.8905,
+      "step": 4012
+    },
+    {
+      "epoch": 3.205394605394605,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9133,
+      "step": 4013
+    },
+    {
+      "epoch": 3.2061938061938062,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0002,
+      "loss": 0.9042,
+      "step": 4014
+    },
+    {
+      "epoch": 3.206993006993007,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9021,
+      "step": 4015
+    },
+    {
+      "epoch": 3.207792207792208,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8952,
+      "step": 4016
+    },
+    {
+      "epoch": 3.2085914085914085,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9045,
+      "step": 4017
+    },
+    {
+      "epoch": 3.2093906093906095,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0002,
+      "loss": 0.9122,
+      "step": 4018
+    },
+    {
+      "epoch": 3.21018981018981,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0002,
+      "loss": 0.8965,
+      "step": 4019
+    },
+    {
+      "epoch": 3.210989010989011,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0002,
+      "loss": 0.9071,
+      "step": 4020
+    },
+    {
+      "epoch": 3.2117882117882117,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0002,
+      "loss": 0.9136,
+      "step": 4021
+    },
+    {
+      "epoch": 3.2125874125874128,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0002,
+      "loss": 0.902,
+      "step": 4022
+    },
+    {
+      "epoch": 3.2133866133866134,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9025,
+      "step": 4023
+    },
+    {
+      "epoch": 3.2141858141858144,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0002,
+      "loss": 0.9099,
+      "step": 4024
+    },
+    {
+      "epoch": 3.214985014985015,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0002,
+      "loss": 0.8927,
+      "step": 4025
+    },
+    {
+      "epoch": 3.2157842157842156,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8977,
+      "step": 4026
+    },
+    {
+      "epoch": 3.2165834165834166,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9045,
+      "step": 4027
+    },
+    {
+      "epoch": 3.217382617382617,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.0002,
+      "loss": 0.8936,
+      "step": 4028
+    },
+    {
+      "epoch": 3.2181818181818183,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8957,
+      "step": 4029
+    },
+    {
+      "epoch": 3.218981018981019,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0002,
+      "loss": 0.9075,
+      "step": 4030
+    },
+    {
+      "epoch": 3.21978021978022,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8973,
+      "step": 4031
+    },
+    {
+      "epoch": 3.2205794205794205,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0002,
+      "loss": 0.9104,
+      "step": 4032
+    },
+    {
+      "epoch": 3.2213786213786215,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9048,
+      "step": 4033
+    },
+    {
+      "epoch": 3.222177822177822,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9055,
+      "step": 4034
+    },
+    {
+      "epoch": 3.222977022977023,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0002,
+      "loss": 0.8913,
+      "step": 4035
+    },
+    {
+      "epoch": 3.2237762237762237,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0002,
+      "loss": 0.9128,
+      "step": 4036
+    },
+    {
+      "epoch": 3.2245754245754243,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8976,
+      "step": 4037
+    },
+    {
+      "epoch": 3.2253746253746254,
+      "grad_norm": 0.8203125,
+      "learning_rate": 0.0002,
+      "loss": 0.9145,
+      "step": 4038
+    },
+    {
+      "epoch": 3.226173826173826,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.0002,
+      "loss": 0.902,
+      "step": 4039
+    },
+    {
+      "epoch": 3.226973026973027,
+      "grad_norm": 1.40625,
+      "learning_rate": 0.0002,
+      "loss": 0.9043,
+      "step": 4040
+    },
+    {
+      "epoch": 3.2277722277722276,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0002,
+      "loss": 0.9116,
+      "step": 4041
+    },
+    {
+      "epoch": 3.2285714285714286,
+      "grad_norm": 1.4921875,
+      "learning_rate": 0.0002,
+      "loss": 0.9017,
+      "step": 4042
+    },
+    {
+      "epoch": 3.2293706293706292,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9082,
+      "step": 4043
+    },
+    {
+      "epoch": 3.2301698301698303,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.0002,
+      "loss": 0.9003,
+      "step": 4044
+    },
+    {
+      "epoch": 3.230969030969031,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0002,
+      "loss": 0.9021,
+      "step": 4045
+    },
+    {
+      "epoch": 3.231768231768232,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0002,
+      "loss": 0.901,
+      "step": 4046
+    },
+    {
+      "epoch": 3.2325674325674325,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9029,
+      "step": 4047
+    },
+    {
+      "epoch": 3.2333666333666335,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0002,
+      "loss": 0.9018,
+      "step": 4048
+    },
+    {
+      "epoch": 3.234165834165834,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.0002,
+      "loss": 0.9066,
+      "step": 4049
+    },
+    {
+      "epoch": 3.234965034965035,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.0002,
+      "loss": 0.8969,
+      "step": 4050
+    },
+    {
+      "epoch": 3.2357642357642358,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0002,
+      "loss": 0.9,
+      "step": 4051
+    },
+    {
+      "epoch": 3.2365634365634364,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9041,
+      "step": 4052
+    },
+    {
+      "epoch": 3.2373626373626374,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.0002,
+      "loss": 0.8934,
+      "step": 4053
+    },
+    {
+      "epoch": 3.238161838161838,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0002,
+      "loss": 0.9059,
+      "step": 4054
+    },
+    {
+      "epoch": 3.238961038961039,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0002,
+      "loss": 0.9076,
+      "step": 4055
+    },
+    {
+      "epoch": 3.2397602397602396,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0002,
+      "loss": 0.889,
+      "step": 4056
+    },
+    {
+      "epoch": 3.2405594405594407,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9062,
+      "step": 4057
+    },
+    {
+      "epoch": 3.2413586413586413,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0002,
+      "loss": 0.9029,
+      "step": 4058
+    },
+    {
+      "epoch": 3.2421578421578423,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8975,
+      "step": 4059
+    },
+    {
+      "epoch": 3.242957042957043,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9022,
+      "step": 4060
+    },
+    {
+      "epoch": 3.243756243756244,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9072,
+      "step": 4061
+    },
+    {
+      "epoch": 3.2445554445554445,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8921,
+      "step": 4062
+    },
+    {
+      "epoch": 3.245354645354645,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.0002,
+      "loss": 0.9035,
+      "step": 4063
+    },
+    {
+      "epoch": 3.246153846153846,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8922,
+      "step": 4064
+    },
+    {
+      "epoch": 3.2469530469530468,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0002,
+      "loss": 0.907,
+      "step": 4065
+    },
+    {
+      "epoch": 3.247752247752248,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8987,
+      "step": 4066
+    },
+    {
+      "epoch": 3.2485514485514484,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0002,
+      "loss": 0.9091,
+      "step": 4067
+    },
+    {
+      "epoch": 3.2493506493506494,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0002,
+      "loss": 0.902,
+      "step": 4068
+    },
+    {
+      "epoch": 3.25014985014985,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.0002,
+      "loss": 0.895,
+      "step": 4069
+    },
+    {
+      "epoch": 3.250949050949051,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9067,
+      "step": 4070
+    },
+    {
+      "epoch": 3.2517482517482517,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8947,
+      "step": 4071
+    },
+    {
+      "epoch": 3.2525474525474527,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0002,
+      "loss": 0.897,
+      "step": 4072
+    },
+    {
+      "epoch": 3.2533466533466533,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0002,
+      "loss": 0.9094,
+      "step": 4073
+    },
+    {
+      "epoch": 3.2541458541458543,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8943,
+      "step": 4074
+    },
+    {
+      "epoch": 3.254945054945055,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8994,
+      "step": 4075
+    },
+    {
+      "epoch": 3.255744255744256,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9023,
+      "step": 4076
+    },
+    {
+      "epoch": 3.2565434565434566,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0002,
+      "loss": 0.9004,
+      "step": 4077
+    },
+    {
+      "epoch": 3.257342657342657,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0002,
+      "loss": 0.8946,
+      "step": 4078
+    },
+    {
+      "epoch": 3.258141858141858,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0002,
+      "loss": 0.9038,
+      "step": 4079
+    },
+    {
+      "epoch": 3.258941058941059,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8968,
+      "step": 4080
+    },
+    {
+      "epoch": 3.25974025974026,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0002,
+      "loss": 0.9009,
+      "step": 4081
+    },
+    {
+      "epoch": 3.2605394605394604,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0002,
+      "loss": 0.8946,
+      "step": 4082
+    },
+    {
+      "epoch": 3.2613386613386615,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8989,
+      "step": 4083
+    },
+    {
+      "epoch": 3.262137862137862,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9037,
+      "step": 4084
+    },
+    {
+      "epoch": 3.262937062937063,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9086,
+      "step": 4085
+    },
+    {
+      "epoch": 3.2637362637362637,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0002,
+      "loss": 0.9059,
+      "step": 4086
+    },
+    {
+      "epoch": 3.2645354645354647,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0002,
+      "loss": 0.9023,
+      "step": 4087
+    },
+    {
+      "epoch": 3.2653346653346653,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0002,
+      "loss": 0.8967,
+      "step": 4088
+    },
+    {
+      "epoch": 3.266133866133866,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0002,
+      "loss": 0.905,
+      "step": 4089
+    },
+    {
+      "epoch": 3.266933066933067,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0002,
+      "loss": 0.9011,
+      "step": 4090
+    },
+    {
+      "epoch": 3.2677322677322675,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8959,
+      "step": 4091
+    },
+    {
+      "epoch": 3.2685314685314686,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0002,
+      "loss": 0.9042,
+      "step": 4092
+    },
+    {
+      "epoch": 3.269330669330669,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0002,
+      "loss": 0.9009,
+      "step": 4093
+    },
+    {
+      "epoch": 3.27012987012987,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0002,
+      "loss": 0.9036,
+      "step": 4094
+    },
+    {
+      "epoch": 3.270929070929071,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0002,
+      "loss": 0.908,
+      "step": 4095
+    },
+    {
+      "epoch": 3.271728271728272,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0002,
+      "loss": 0.9006,
+      "step": 4096
+    },
+    {
+      "epoch": 3.2725274725274724,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0002,
+      "loss": 0.9064,
+      "step": 4097
+    },
+    {
+      "epoch": 3.2733266733266735,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8951,
+      "step": 4098
+    },
+    {
+      "epoch": 3.274125874125874,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0002,
+      "loss": 0.9027,
+      "step": 4099
+    },
+    {
+      "epoch": 3.274925074925075,
+      "grad_norm": 1.5078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9053,
+      "step": 4100
+    },
+    {
+      "epoch": 3.2757242757242757,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.0002,
+      "loss": 0.9063,
+      "step": 4101
+    },
+    {
+      "epoch": 3.2765234765234768,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9052,
+      "step": 4102
+    },
+    {
+      "epoch": 3.2773226773226773,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9003,
+      "step": 4103
+    },
+    {
+      "epoch": 3.278121878121878,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0002,
+      "loss": 0.9068,
+      "step": 4104
+    },
+    {
+      "epoch": 3.278921078921079,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8981,
+      "step": 4105
+    },
+    {
+      "epoch": 3.2797202797202796,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0002,
+      "loss": 0.8924,
+      "step": 4106
+    },
+    {
+      "epoch": 3.2805194805194806,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0002,
+      "loss": 0.8903,
+      "step": 4107
+    },
+    {
+      "epoch": 3.281318681318681,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0002,
+      "loss": 0.901,
+      "step": 4108
+    },
+    {
+      "epoch": 3.2821178821178822,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0002,
+      "loss": 0.899,
+      "step": 4109
+    },
+    {
+      "epoch": 3.282917082917083,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0002,
+      "loss": 0.9117,
+      "step": 4110
+    },
+    {
+      "epoch": 3.283716283716284,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0002,
+      "loss": 0.9035,
+      "step": 4111
+    },
+    {
+      "epoch": 3.2845154845154845,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9026,
+      "step": 4112
+    },
+    {
+      "epoch": 3.2853146853146855,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8942,
+      "step": 4113
+    },
+    {
+      "epoch": 3.286113886113886,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0002,
+      "loss": 0.9003,
+      "step": 4114
+    },
+    {
+      "epoch": 3.2869130869130867,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0002,
+      "loss": 0.8982,
+      "step": 4115
+    },
+    {
+      "epoch": 3.2877122877122877,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0002,
+      "loss": 0.8963,
+      "step": 4116
+    },
+    {
+      "epoch": 3.2885114885114883,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0002,
+      "loss": 0.9013,
+      "step": 4117
+    },
+    {
+      "epoch": 3.2893106893106894,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0002,
+      "loss": 0.9055,
+      "step": 4118
+    },
+    {
+      "epoch": 3.29010989010989,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0002,
+      "loss": 0.9018,
+      "step": 4119
+    },
+    {
+      "epoch": 3.290909090909091,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9009,
+      "step": 4120
+    },
+    {
+      "epoch": 3.2917082917082916,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0002,
+      "loss": 0.892,
+      "step": 4121
+    },
+    {
+      "epoch": 3.2925074925074926,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0002,
+      "loss": 0.9044,
+      "step": 4122
+    },
+    {
+      "epoch": 3.2933066933066932,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.0002,
+      "loss": 0.9154,
+      "step": 4123
+    },
+    {
+      "epoch": 3.2941058941058943,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.0002,
+      "loss": 0.9081,
+      "step": 4124
+    },
+    {
+      "epoch": 3.294905094905095,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0002,
+      "loss": 0.9019,
+      "step": 4125
+    },
+    {
+      "epoch": 3.2957042957042955,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9038,
+      "step": 4126
+    },
+    {
+      "epoch": 3.2965034965034965,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.0002,
+      "loss": 0.9011,
+      "step": 4127
+    },
+    {
+      "epoch": 3.2973026973026975,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0002,
+      "loss": 0.8992,
+      "step": 4128
+    },
+    {
+      "epoch": 3.298101898101898,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.0002,
+      "loss": 0.9007,
+      "step": 4129
+    },
+    {
+      "epoch": 3.2989010989010987,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0002,
+      "loss": 0.9101,
+      "step": 4130
+    },
+    {
+      "epoch": 3.2997002997002998,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0002,
+      "loss": 0.8957,
+      "step": 4131
+    },
+    {
+      "epoch": 3.3004995004995004,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0002,
+      "loss": 0.9122,
+      "step": 4132
+    },
+    {
+      "epoch": 3.3012987012987014,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0002,
+      "loss": 0.8959,
+      "step": 4133
+    },
+    {
+      "epoch": 3.302097902097902,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0002,
+      "loss": 0.903,
+      "step": 4134
+    },
+    {
+      "epoch": 3.302897102897103,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0002,
+      "loss": 0.9017,
+      "step": 4135
+    },
+    {
+      "epoch": 3.3036963036963036,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0002,
+      "loss": 0.9017,
+      "step": 4136
+    },
+    {
+      "epoch": 3.3044955044955047,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0002,
+      "loss": 0.9108,
+      "step": 4137
+    },
+    {
+      "epoch": 3.3052947052947053,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8887,
+      "step": 4138
+    },
+    {
+      "epoch": 3.3060939060939063,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0002,
+      "loss": 0.949,
+      "step": 4139
+    },
+    {
+      "epoch": 3.306893106893107,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.0002,
+      "loss": 0.8993,
+      "step": 4140
+    },
+    {
+      "epoch": 3.3076923076923075,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9122,
+      "step": 4141
+    },
+    {
+      "epoch": 3.3084915084915085,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9007,
+      "step": 4142
+    },
+    {
+      "epoch": 3.309290709290709,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.0002,
+      "loss": 0.9059,
+      "step": 4143
+    },
+    {
+      "epoch": 3.31008991008991,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0002,
+      "loss": 0.8987,
+      "step": 4144
+    },
+    {
+      "epoch": 3.3108891108891108,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8996,
+      "step": 4145
+    },
+    {
+      "epoch": 3.311688311688312,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0002,
+      "loss": 0.899,
+      "step": 4146
+    },
+    {
+      "epoch": 3.3124875124875124,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0002,
+      "loss": 0.9009,
+      "step": 4147
+    },
+    {
+      "epoch": 3.3132867132867134,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.0002,
+      "loss": 0.8937,
+      "step": 4148
+    },
+    {
+      "epoch": 3.314085914085914,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9104,
+      "step": 4149
+    },
+    {
+      "epoch": 3.314885114885115,
+      "grad_norm": 0.9921875,
+      "learning_rate": 0.0002,
+      "loss": 0.9363,
+      "step": 4150
+    },
+    {
+      "epoch": 3.3156843156843157,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9042,
+      "step": 4151
+    },
+    {
+      "epoch": 3.3164835164835162,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9007,
+      "step": 4152
+    },
+    {
+      "epoch": 3.3172827172827173,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0002,
+      "loss": 0.9028,
+      "step": 4153
+    },
+    {
+      "epoch": 3.3180819180819183,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.0002,
+      "loss": 0.8994,
+      "step": 4154
+    },
+    {
+      "epoch": 3.318881118881119,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0002,
+      "loss": 0.9002,
+      "step": 4155
+    },
+    {
+      "epoch": 3.3196803196803195,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9088,
+      "step": 4156
+    },
+    {
+      "epoch": 3.3204795204795206,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8911,
+      "step": 4157
+    },
+    {
+      "epoch": 3.321278721278721,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9004,
+      "step": 4158
+    },
+    {
+      "epoch": 3.322077922077922,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0002,
+      "loss": 0.8945,
+      "step": 4159
+    },
+    {
+      "epoch": 3.322877122877123,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0002,
+      "loss": 0.8984,
+      "step": 4160
+    },
+    {
+      "epoch": 3.323676323676324,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8998,
+      "step": 4161
+    },
+    {
+      "epoch": 3.3244755244755244,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0002,
+      "loss": 0.9056,
+      "step": 4162
+    },
+    {
+      "epoch": 3.3252747252747255,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9031,
+      "step": 4163
+    },
+    {
+      "epoch": 3.326073926073926,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9,
+      "step": 4164
+    },
+    {
+      "epoch": 3.326873126873127,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9051,
+      "step": 4165
+    },
+    {
+      "epoch": 3.3276723276723277,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8965,
+      "step": 4166
+    },
+    {
+      "epoch": 3.3284715284715283,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8953,
+      "step": 4167
+    },
+    {
+      "epoch": 3.3292707292707293,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0002,
+      "loss": 0.8962,
+      "step": 4168
+    },
+    {
+      "epoch": 3.33006993006993,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0002,
+      "loss": 0.8971,
+      "step": 4169
+    },
+    {
+      "epoch": 3.330869130869131,
+      "grad_norm": 0.8515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9158,
+      "step": 4170
+    },
+    {
+      "epoch": 3.3316683316683315,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0002,
+      "loss": 0.9063,
+      "step": 4171
+    },
+    {
+      "epoch": 3.3324675324675326,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8934,
+      "step": 4172
+    },
+    {
+      "epoch": 3.333266733266733,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0002,
+      "loss": 0.8989,
+      "step": 4173
+    },
+    {
+      "epoch": 3.334065934065934,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0002,
+      "loss": 0.897,
+      "step": 4174
+    },
+    {
+      "epoch": 3.334865134865135,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0002,
+      "loss": 0.896,
+      "step": 4175
+    },
+    {
+      "epoch": 3.335664335664336,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0002,
+      "loss": 0.8956,
+      "step": 4176
+    },
+    {
+      "epoch": 3.3364635364635364,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9044,
+      "step": 4177
+    },
+    {
+      "epoch": 3.337262737262737,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0002,
+      "loss": 0.907,
+      "step": 4178
+    },
+    {
+      "epoch": 3.338061938061938,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0002,
+      "loss": 0.8909,
+      "step": 4179
+    },
+    {
+      "epoch": 3.338861138861139,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.0002,
+      "loss": 0.9008,
+      "step": 4180
+    },
+    {
+      "epoch": 3.3396603396603397,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0002,
+      "loss": 0.8999,
+      "step": 4181
+    },
+    {
+      "epoch": 3.3404595404595403,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9012,
+      "step": 4182
+    },
+    {
+      "epoch": 3.3412587412587413,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0002,
+      "loss": 0.8901,
+      "step": 4183
+    },
+    {
+      "epoch": 3.342057942057942,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.0002,
+      "loss": 0.8918,
+      "step": 4184
+    },
+    {
+      "epoch": 3.342857142857143,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.0002,
+      "loss": 0.898,
+      "step": 4185
+    },
+    {
+      "epoch": 3.3436563436563436,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0002,
+      "loss": 0.9088,
+      "step": 4186
+    },
+    {
+      "epoch": 3.3444555444555446,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9,
+      "step": 4187
+    },
+    {
+      "epoch": 3.345254745254745,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.0002,
+      "loss": 0.9034,
+      "step": 4188
+    },
+    {
+      "epoch": 3.3460539460539462,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9075,
+      "step": 4189
+    },
+    {
+      "epoch": 3.346853146853147,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8989,
+      "step": 4190
+    },
+    {
+      "epoch": 3.347652347652348,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0002,
+      "loss": 0.8993,
+      "step": 4191
+    },
+    {
+      "epoch": 3.3484515484515485,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.0002,
+      "loss": 0.8953,
+      "step": 4192
+    },
+    {
+      "epoch": 3.349250749250749,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0002,
+      "loss": 0.8936,
+      "step": 4193
+    },
+    {
+      "epoch": 3.35004995004995,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0002,
+      "loss": 0.9133,
+      "step": 4194
+    },
+    {
+      "epoch": 3.3508491508491507,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9004,
+      "step": 4195
+    },
+    {
+      "epoch": 3.3516483516483517,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9031,
+      "step": 4196
+    },
+    {
+      "epoch": 3.3524475524475523,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.0002,
+      "loss": 0.9076,
+      "step": 4197
+    },
+    {
+      "epoch": 3.3532467532467534,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9037,
+      "step": 4198
+    },
+    {
+      "epoch": 3.354045954045954,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8992,
+      "step": 4199
+    },
+    {
+      "epoch": 3.354845154845155,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8994,
+      "step": 4200
+    },
+    {
+      "epoch": 3.3556443556443556,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0002,
+      "loss": 0.9178,
+      "step": 4201
+    },
+    {
+      "epoch": 3.3564435564435566,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8987,
+      "step": 4202
+    },
+    {
+      "epoch": 3.3572427572427572,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.0002,
+      "loss": 0.897,
+      "step": 4203
+    },
+    {
+      "epoch": 3.358041958041958,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9028,
+      "step": 4204
+    },
+    {
+      "epoch": 3.358841158841159,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0002,
+      "loss": 0.9094,
+      "step": 4205
+    },
+    {
+      "epoch": 3.3596403596403595,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9052,
+      "step": 4206
+    },
+    {
+      "epoch": 3.3604395604395605,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0002,
+      "loss": 0.9018,
+      "step": 4207
+    },
+    {
+      "epoch": 3.361238761238761,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.0002,
+      "loss": 0.8898,
+      "step": 4208
+    },
+    {
+      "epoch": 3.362037962037962,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8996,
+      "step": 4209
+    },
+    {
+      "epoch": 3.3628371628371627,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0002,
+      "loss": 0.9042,
+      "step": 4210
+    },
+    {
+      "epoch": 3.3636363636363638,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.0002,
+      "loss": 0.9185,
+      "step": 4211
+    },
+    {
+      "epoch": 3.3644355644355644,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9034,
+      "step": 4212
+    },
+    {
+      "epoch": 3.3652347652347654,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0002,
+      "loss": 0.8998,
+      "step": 4213
+    },
+    {
+      "epoch": 3.366033966033966,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0002,
+      "loss": 0.893,
+      "step": 4214
+    },
+    {
+      "epoch": 3.366833166833167,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0002,
+      "loss": 0.8984,
+      "step": 4215
+    },
+    {
+      "epoch": 3.3676323676323676,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9025,
+      "step": 4216
+    },
+    {
+      "epoch": 3.3684315684315687,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0002,
+      "loss": 0.8999,
+      "step": 4217
+    },
+    {
+      "epoch": 3.3692307692307693,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0002,
+      "loss": 0.9014,
+      "step": 4218
+    },
+    {
+      "epoch": 3.37002997002997,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0002,
+      "loss": 0.9104,
+      "step": 4219
+    },
+    {
+      "epoch": 3.370829170829171,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0002,
+      "loss": 0.8984,
+      "step": 4220
+    },
+    {
+      "epoch": 3.3716283716283715,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8953,
+      "step": 4221
+    },
+    {
+      "epoch": 3.3724275724275725,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0002,
+      "loss": 0.8983,
+      "step": 4222
+    },
+    {
+      "epoch": 3.373226773226773,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0002,
+      "loss": 0.9031,
+      "step": 4223
+    },
+    {
+      "epoch": 3.374025974025974,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0002,
+      "loss": 0.8985,
+      "step": 4224
+    },
+    {
+      "epoch": 3.3748251748251747,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0002,
+      "loss": 0.8958,
+      "step": 4225
+    },
+    {
+      "epoch": 3.375624375624376,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0002,
+      "loss": 0.8986,
+      "step": 4226
+    },
+    {
+      "epoch": 3.3764235764235764,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.0002,
+      "loss": 0.901,
+      "step": 4227
+    },
+    {
+      "epoch": 3.3772227772227774,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8932,
+      "step": 4228
+    },
+    {
+      "epoch": 3.378021978021978,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0002,
+      "loss": 0.9008,
+      "step": 4229
+    },
+    {
+      "epoch": 3.3788211788211786,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0002,
+      "loss": 0.8969,
+      "step": 4230
+    },
+    {
+      "epoch": 3.3796203796203796,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0002,
+      "loss": 0.909,
+      "step": 4231
+    },
+    {
+      "epoch": 3.3804195804195802,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0002,
+      "loss": 0.9018,
+      "step": 4232
+    },
+    {
+      "epoch": 3.3812187812187813,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0002,
+      "loss": 0.909,
+      "step": 4233
+    },
+    {
+      "epoch": 3.382017982017982,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9003,
+      "step": 4234
+    },
+    {
+      "epoch": 3.382817182817183,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9083,
+      "step": 4235
+    },
+    {
+      "epoch": 3.3836163836163835,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8917,
+      "step": 4236
+    },
+    {
+      "epoch": 3.3844155844155845,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0002,
+      "loss": 0.8983,
+      "step": 4237
+    },
+    {
+      "epoch": 3.385214785214785,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8974,
+      "step": 4238
+    },
+    {
+      "epoch": 3.386013986013986,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0002,
+      "loss": 0.893,
+      "step": 4239
+    },
+    {
+      "epoch": 3.3868131868131868,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9015,
+      "step": 4240
+    },
+    {
+      "epoch": 3.3876123876123874,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0002,
+      "loss": 0.9075,
+      "step": 4241
+    },
+    {
+      "epoch": 3.3884115884115884,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.0002,
+      "loss": 0.8974,
+      "step": 4242
+    },
+    {
+      "epoch": 3.3892107892107894,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0002,
+      "loss": 0.9078,
+      "step": 4243
+    },
+    {
+      "epoch": 3.39000999000999,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9,
+      "step": 4244
+    },
+    {
+      "epoch": 3.3908091908091906,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8951,
+      "step": 4245
+    },
+    {
+      "epoch": 3.3916083916083917,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.0002,
+      "loss": 0.8972,
+      "step": 4246
+    },
+    {
+      "epoch": 3.3924075924075923,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9094,
+      "step": 4247
+    },
+    {
+      "epoch": 3.3932067932067933,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0002,
+      "loss": 0.9061,
+      "step": 4248
+    },
+    {
+      "epoch": 3.394005994005994,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0002,
+      "loss": 0.9072,
+      "step": 4249
+    },
+    {
+      "epoch": 3.394805194805195,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9054,
+      "step": 4250
+    },
+    {
+      "epoch": 3.3956043956043955,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0002,
+      "loss": 0.898,
+      "step": 4251
+    },
+    {
+      "epoch": 3.3964035964035966,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.0002,
+      "loss": 0.894,
+      "step": 4252
+    },
+    {
+      "epoch": 3.397202797202797,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0002,
+      "loss": 0.8924,
+      "step": 4253
+    },
+    {
+      "epoch": 3.398001998001998,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8937,
+      "step": 4254
+    },
+    {
+      "epoch": 3.398801198801199,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9014,
+      "step": 4255
+    },
+    {
+      "epoch": 3.3996003996003994,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0002,
+      "loss": 0.8969,
+      "step": 4256
+    },
+    {
+      "epoch": 3.4003996003996004,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0002,
+      "loss": 0.897,
+      "step": 4257
+    },
+    {
+      "epoch": 3.401198801198801,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9043,
+      "step": 4258
+    },
+    {
+      "epoch": 3.401998001998002,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0002,
+      "loss": 0.8977,
+      "step": 4259
+    },
+    {
+      "epoch": 3.4027972027972027,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8983,
+      "step": 4260
+    },
+    {
+      "epoch": 3.4035964035964037,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0002,
+      "loss": 0.894,
+      "step": 4261
+    },
+    {
+      "epoch": 3.4043956043956043,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9005,
+      "step": 4262
+    },
+    {
+      "epoch": 3.4051948051948053,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9031,
+      "step": 4263
+    },
+    {
+      "epoch": 3.405994005994006,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.0002,
+      "loss": 0.8918,
+      "step": 4264
+    },
+    {
+      "epoch": 3.406793206793207,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0002,
+      "loss": 0.9003,
+      "step": 4265
+    },
+    {
+      "epoch": 3.4075924075924076,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0002,
+      "loss": 0.8978,
+      "step": 4266
+    },
+    {
+      "epoch": 3.408391608391608,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9053,
+      "step": 4267
+    },
+    {
+      "epoch": 3.409190809190809,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.0002,
+      "loss": 0.8922,
+      "step": 4268
+    },
+    {
+      "epoch": 3.4099900099900102,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.0002,
+      "loss": 0.8951,
+      "step": 4269
+    },
+    {
+      "epoch": 3.410789210789211,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0002,
+      "loss": 0.8896,
+      "step": 4270
+    },
+    {
+      "epoch": 3.4115884115884114,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9131,
+      "step": 4271
+    },
+    {
+      "epoch": 3.4123876123876125,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8976,
+      "step": 4272
+    },
+    {
+      "epoch": 3.413186813186813,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0002,
+      "loss": 0.8987,
+      "step": 4273
+    },
+    {
+      "epoch": 3.413986013986014,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0002,
+      "loss": 0.8993,
+      "step": 4274
+    },
+    {
+      "epoch": 3.4147852147852147,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.0002,
+      "loss": 0.9045,
+      "step": 4275
+    },
+    {
+      "epoch": 3.4155844155844157,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0002,
+      "loss": 0.8953,
+      "step": 4276
+    },
+    {
+      "epoch": 3.4163836163836163,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0002,
+      "loss": 0.9024,
+      "step": 4277
+    },
+    {
+      "epoch": 3.4171828171828174,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0002,
+      "loss": 0.897,
+      "step": 4278
+    },
+    {
+      "epoch": 3.417982017982018,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8978,
+      "step": 4279
+    },
+    {
+      "epoch": 3.418781218781219,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9083,
+      "step": 4280
+    },
+    {
+      "epoch": 3.4195804195804196,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0002,
+      "loss": 0.9023,
+      "step": 4281
+    },
+    {
+      "epoch": 3.42037962037962,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0002,
+      "loss": 0.902,
+      "step": 4282
+    },
+    {
+      "epoch": 3.421178821178821,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0002,
+      "loss": 0.9029,
+      "step": 4283
+    },
+    {
+      "epoch": 3.421978021978022,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.0002,
+      "loss": 0.9005,
+      "step": 4284
+    },
+    {
+      "epoch": 3.422777222777223,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0002,
+      "loss": 0.9026,
+      "step": 4285
+    },
+    {
+      "epoch": 3.4235764235764234,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0002,
+      "loss": 0.8921,
+      "step": 4286
+    },
+    {
+      "epoch": 3.4243756243756245,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0002,
+      "loss": 0.9034,
+      "step": 4287
+    },
+    {
+      "epoch": 3.425174825174825,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9037,
+      "step": 4288
+    },
+    {
+      "epoch": 3.425974025974026,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8917,
+      "step": 4289
+    },
+    {
+      "epoch": 3.4267732267732267,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0002,
+      "loss": 0.8943,
+      "step": 4290
+    },
+    {
+      "epoch": 3.4275724275724277,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0002,
+      "loss": 0.9034,
+      "step": 4291
+    },
+    {
+      "epoch": 3.4283716283716283,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0002,
+      "loss": 0.8982,
+      "step": 4292
+    },
+    {
+      "epoch": 3.429170829170829,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.0002,
+      "loss": 0.8986,
+      "step": 4293
+    },
+    {
+      "epoch": 3.42997002997003,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0002,
+      "loss": 0.9024,
+      "step": 4294
+    },
+    {
+      "epoch": 3.430769230769231,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.0002,
+      "loss": 0.9045,
+      "step": 4295
+    },
+    {
+      "epoch": 3.4315684315684316,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8905,
+      "step": 4296
+    },
+    {
+      "epoch": 3.432367632367632,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.0002,
+      "loss": 0.8936,
+      "step": 4297
+    },
+    {
+      "epoch": 3.4331668331668332,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0002,
+      "loss": 0.9522,
+      "step": 4298
+    },
+    {
+      "epoch": 3.433966033966034,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8972,
+      "step": 4299
+    },
+    {
+      "epoch": 3.434765234765235,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9427,
+      "step": 4300
+    },
+    {
+      "epoch": 3.4355644355644355,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0002,
+      "loss": 0.9038,
+      "step": 4301
+    },
+    {
+      "epoch": 3.4363636363636365,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.0002,
+      "loss": 0.9032,
+      "step": 4302
+    },
+    {
+      "epoch": 3.437162837162837,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9058,
+      "step": 4303
+    },
+    {
+      "epoch": 3.437962037962038,
+      "grad_norm": 1.7265625,
+      "learning_rate": 0.0002,
+      "loss": 0.939,
+      "step": 4304
+    },
+    {
+      "epoch": 3.4387612387612387,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9048,
+      "step": 4305
+    },
+    {
+      "epoch": 3.4395604395604398,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8976,
+      "step": 4306
+    },
+    {
+      "epoch": 3.4403596403596404,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9099,
+      "step": 4307
+    },
+    {
+      "epoch": 3.441158841158841,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0002,
+      "loss": 0.8986,
+      "step": 4308
+    },
+    {
+      "epoch": 3.441958041958042,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8958,
+      "step": 4309
+    },
+    {
+      "epoch": 3.4427572427572426,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9028,
+      "step": 4310
+    },
+    {
+      "epoch": 3.4435564435564436,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.0002,
+      "loss": 0.9054,
+      "step": 4311
+    },
+    {
+      "epoch": 3.4443556443556442,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.0002,
+      "loss": 0.904,
+      "step": 4312
+    },
+    {
+      "epoch": 3.4451548451548453,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0002,
+      "loss": 0.8992,
+      "step": 4313
+    },
+    {
+      "epoch": 3.445954045954046,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.0002,
+      "loss": 0.8957,
+      "step": 4314
+    },
+    {
+      "epoch": 3.446753246753247,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0002,
+      "loss": 0.9081,
+      "step": 4315
+    },
+    {
+      "epoch": 3.4475524475524475,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8843,
+      "step": 4316
+    },
+    {
+      "epoch": 3.4483516483516485,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0002,
+      "loss": 0.9081,
+      "step": 4317
+    },
+    {
+      "epoch": 3.449150849150849,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9089,
+      "step": 4318
+    },
+    {
+      "epoch": 3.4499500499500497,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0002,
+      "loss": 0.9063,
+      "step": 4319
+    },
+    {
+      "epoch": 3.4507492507492508,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8946,
+      "step": 4320
+    },
+    {
+      "epoch": 3.4515484515484514,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0002,
+      "loss": 0.8956,
+      "step": 4321
+    },
+    {
+      "epoch": 3.4523476523476524,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0002,
+      "loss": 0.8964,
+      "step": 4322
+    },
+    {
+      "epoch": 3.453146853146853,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0002,
+      "loss": 0.8952,
+      "step": 4323
+    },
+    {
+      "epoch": 3.453946053946054,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8965,
+      "step": 4324
+    },
+    {
+      "epoch": 3.4547452547452546,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8973,
+      "step": 4325
+    },
+    {
+      "epoch": 3.4555444555444557,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8977,
+      "step": 4326
+    },
+    {
+      "epoch": 3.4563436563436563,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0002,
+      "loss": 0.9043,
+      "step": 4327
+    },
+    {
+      "epoch": 3.4571428571428573,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0002,
+      "loss": 0.8994,
+      "step": 4328
+    },
+    {
+      "epoch": 3.457942057942058,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0002,
+      "loss": 0.9051,
+      "step": 4329
+    },
+    {
+      "epoch": 3.458741258741259,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0002,
+      "loss": 0.904,
+      "step": 4330
+    },
+    {
+      "epoch": 3.4595404595404595,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8962,
+      "step": 4331
+    },
+    {
+      "epoch": 3.4603396603396606,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0002,
+      "loss": 0.8966,
+      "step": 4332
+    },
+    {
+      "epoch": 3.461138861138861,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0002,
+      "loss": 0.8989,
+      "step": 4333
+    },
+    {
+      "epoch": 3.4619380619380618,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9034,
+      "step": 4334
+    },
+    {
+      "epoch": 3.462737262737263,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0002,
+      "loss": 0.8997,
+      "step": 4335
+    },
+    {
+      "epoch": 3.4635364635364634,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0002,
+      "loss": 0.9084,
+      "step": 4336
+    },
+    {
+      "epoch": 3.4643356643356644,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0002,
+      "loss": 0.908,
+      "step": 4337
+    },
+    {
+      "epoch": 3.465134865134865,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0002,
+      "loss": 0.903,
+      "step": 4338
+    },
+    {
+      "epoch": 3.465934065934066,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8954,
+      "step": 4339
+    },
+    {
+      "epoch": 3.4667332667332666,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8988,
+      "step": 4340
+    },
+    {
+      "epoch": 3.4675324675324677,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0002,
+      "loss": 0.9001,
+      "step": 4341
+    },
+    {
+      "epoch": 3.4683316683316683,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8968,
+      "step": 4342
+    },
+    {
+      "epoch": 3.4691308691308693,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0002,
+      "loss": 0.894,
+      "step": 4343
+    },
+    {
+      "epoch": 3.46993006993007,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0002,
+      "loss": 0.8916,
+      "step": 4344
+    },
+    {
+      "epoch": 3.4707292707292705,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0002,
+      "loss": 0.9053,
+      "step": 4345
+    },
+    {
+      "epoch": 3.4715284715284715,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8962,
+      "step": 4346
+    },
+    {
+      "epoch": 3.472327672327672,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8952,
+      "step": 4347
+    },
+    {
+      "epoch": 3.473126873126873,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8988,
+      "step": 4348
+    },
+    {
+      "epoch": 3.4739260739260738,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0002,
+      "loss": 0.9011,
+      "step": 4349
+    },
+    {
+      "epoch": 3.474725274725275,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9014,
+      "step": 4350
+    },
+    {
+      "epoch": 3.4755244755244754,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9028,
+      "step": 4351
+    },
+    {
+      "epoch": 3.4763236763236764,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9016,
+      "step": 4352
+    },
+    {
+      "epoch": 3.477122877122877,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0002,
+      "loss": 0.9007,
+      "step": 4353
+    },
+    {
+      "epoch": 3.477922077922078,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8983,
+      "step": 4354
+    },
+    {
+      "epoch": 3.4787212787212787,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9043,
+      "step": 4355
+    },
+    {
+      "epoch": 3.4795204795204797,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0002,
+      "loss": 0.8935,
+      "step": 4356
+    },
+    {
+      "epoch": 3.4803196803196803,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.0002,
+      "loss": 0.9028,
+      "step": 4357
+    },
+    {
+      "epoch": 3.4811188811188813,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.0002,
+      "loss": 0.908,
+      "step": 4358
+    },
+    {
+      "epoch": 3.481918081918082,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.0002,
+      "loss": 0.9014,
+      "step": 4359
+    },
+    {
+      "epoch": 3.4827172827172825,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0002,
+      "loss": 0.9029,
+      "step": 4360
+    },
+    {
+      "epoch": 3.4835164835164836,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9076,
+      "step": 4361
+    },
+    {
+      "epoch": 3.484315684315684,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.0002,
+      "loss": 0.9,
+      "step": 4362
+    },
+    {
+      "epoch": 3.485114885114885,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.0002,
+      "loss": 0.8979,
+      "step": 4363
+    },
+    {
+      "epoch": 3.485914085914086,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8901,
+      "step": 4364
+    },
+    {
+      "epoch": 3.486713286713287,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0002,
+      "loss": 0.9074,
+      "step": 4365
+    },
+    {
+      "epoch": 3.4875124875124874,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0002,
+      "loss": 0.8988,
+      "step": 4366
+    },
+    {
+      "epoch": 3.4883116883116885,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0002,
+      "loss": 0.8975,
+      "step": 4367
+    },
+    {
+      "epoch": 3.489110889110889,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0002,
+      "loss": 0.9023,
+      "step": 4368
+    },
+    {
+      "epoch": 3.48991008991009,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0002,
+      "loss": 0.897,
+      "step": 4369
+    },
+    {
+      "epoch": 3.4907092907092907,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0002,
+      "loss": 0.8943,
+      "step": 4370
+    },
+    {
+      "epoch": 3.4915084915084913,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0002,
+      "loss": 0.9038,
+      "step": 4371
+    },
+    {
+      "epoch": 3.4923076923076923,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9004,
+      "step": 4372
+    },
+    {
+      "epoch": 3.493106893106893,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0002,
+      "loss": 0.9003,
+      "step": 4373
+    },
+    {
+      "epoch": 3.493906093906094,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0002,
+      "loss": 0.9115,
+      "step": 4374
+    },
+    {
+      "epoch": 3.4947052947052946,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0002,
+      "loss": 0.8969,
+      "step": 4375
+    },
+    {
+      "epoch": 3.4955044955044956,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8976,
+      "step": 4376
+    },
+    {
+      "epoch": 3.496303696303696,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8995,
+      "step": 4377
+    },
+    {
+      "epoch": 3.4971028971028972,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.0002,
+      "loss": 0.8997,
+      "step": 4378
+    },
+    {
+      "epoch": 3.497902097902098,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0002,
+      "loss": 0.8924,
+      "step": 4379
+    },
+    {
+      "epoch": 3.498701298701299,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.0002,
+      "loss": 0.9054,
+      "step": 4380
+    },
+    {
+      "epoch": 3.4995004995004995,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0002,
+      "loss": 0.8989,
+      "step": 4381
+    },
+    {
+      "epoch": 3.5002997002997,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0002,
+      "loss": 0.9024,
+      "step": 4382
+    },
+    {
+      "epoch": 3.501098901098901,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8968,
+      "step": 4383
+    },
+    {
+      "epoch": 3.501898101898102,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9136,
+      "step": 4384
+    },
+    {
+      "epoch": 3.5026973026973027,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9085,
+      "step": 4385
+    },
+    {
+      "epoch": 3.5034965034965033,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0002,
+      "loss": 0.9067,
+      "step": 4386
+    },
+    {
+      "epoch": 3.5042957042957044,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.0002,
+      "loss": 0.9103,
+      "step": 4387
+    },
+    {
+      "epoch": 3.505094905094905,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0002,
+      "loss": 0.9014,
+      "step": 4388
+    },
+    {
+      "epoch": 3.505894105894106,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0002,
+      "loss": 0.8932,
+      "step": 4389
+    },
+    {
+      "epoch": 3.5066933066933066,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9014,
+      "step": 4390
+    },
+    {
+      "epoch": 3.5074925074925076,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0002,
+      "loss": 0.8903,
+      "step": 4391
+    },
+    {
+      "epoch": 3.508291708291708,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.0002,
+      "loss": 0.9212,
+      "step": 4392
+    },
+    {
+      "epoch": 3.509090909090909,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9011,
+      "step": 4393
+    },
+    {
+      "epoch": 3.50989010989011,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9118,
+      "step": 4394
+    },
+    {
+      "epoch": 3.510689310689311,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9031,
+      "step": 4395
+    },
+    {
+      "epoch": 3.5114885114885115,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0002,
+      "loss": 0.8851,
+      "step": 4396
+    },
+    {
+      "epoch": 3.512287712287712,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0002,
+      "loss": 0.9003,
+      "step": 4397
+    },
+    {
+      "epoch": 3.513086913086913,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0002,
+      "loss": 0.8913,
+      "step": 4398
+    },
+    {
+      "epoch": 3.513886113886114,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0002,
+      "loss": 0.8931,
+      "step": 4399
+    },
+    {
+      "epoch": 3.5146853146853148,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0002,
+      "loss": 0.9009,
+      "step": 4400
+    },
+    {
+      "epoch": 3.5154845154845153,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0002,
+      "loss": 0.8971,
+      "step": 4401
+    },
+    {
+      "epoch": 3.5162837162837164,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8948,
+      "step": 4402
+    },
+    {
+      "epoch": 3.517082917082917,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9059,
+      "step": 4403
+    },
+    {
+      "epoch": 3.517882117882118,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9025,
+      "step": 4404
+    },
+    {
+      "epoch": 3.5186813186813186,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9019,
+      "step": 4405
+    },
+    {
+      "epoch": 3.5194805194805197,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0002,
+      "loss": 0.9032,
+      "step": 4406
+    },
+    {
+      "epoch": 3.5202797202797202,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0002,
+      "loss": 0.8955,
+      "step": 4407
+    },
+    {
+      "epoch": 3.521078921078921,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.0002,
+      "loss": 0.9026,
+      "step": 4408
+    },
+    {
+      "epoch": 3.521878121878122,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0002,
+      "loss": 0.9032,
+      "step": 4409
+    },
+    {
+      "epoch": 3.522677322677323,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0002,
+      "loss": 0.9048,
+      "step": 4410
+    },
+    {
+      "epoch": 3.5234765234765235,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0002,
+      "loss": 0.9028,
+      "step": 4411
+    },
+    {
+      "epoch": 3.524275724275724,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8967,
+      "step": 4412
+    },
+    {
+      "epoch": 3.525074925074925,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0002,
+      "loss": 0.9015,
+      "step": 4413
+    },
+    {
+      "epoch": 3.5258741258741257,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0002,
+      "loss": 0.9058,
+      "step": 4414
+    },
+    {
+      "epoch": 3.526673326673327,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0002,
+      "loss": 0.9013,
+      "step": 4415
+    },
+    {
+      "epoch": 3.5274725274725274,
+      "grad_norm": 0.375,
+      "learning_rate": 0.0002,
+      "loss": 0.8873,
+      "step": 4416
+    },
+    {
+      "epoch": 3.5282717282717284,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.0002,
+      "loss": 0.892,
+      "step": 4417
+    },
+    {
+      "epoch": 3.529070929070929,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8949,
+      "step": 4418
+    },
+    {
+      "epoch": 3.5298701298701296,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0002,
+      "loss": 0.9034,
+      "step": 4419
+    },
+    {
+      "epoch": 3.5306693306693306,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0002,
+      "loss": 0.9059,
+      "step": 4420
+    },
+    {
+      "epoch": 3.5314685314685317,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8981,
+      "step": 4421
+    },
+    {
+      "epoch": 3.5322677322677323,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0002,
+      "loss": 0.895,
+      "step": 4422
+    },
+    {
+      "epoch": 3.533066933066933,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8986,
+      "step": 4423
+    },
+    {
+      "epoch": 3.533866133866134,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8981,
+      "step": 4424
+    },
+    {
+      "epoch": 3.5346653346653345,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0002,
+      "loss": 0.9076,
+      "step": 4425
+    },
+    {
+      "epoch": 3.5354645354645355,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8984,
+      "step": 4426
+    },
+    {
+      "epoch": 3.536263736263736,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0002,
+      "loss": 0.8956,
+      "step": 4427
+    },
+    {
+      "epoch": 3.537062937062937,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8975,
+      "step": 4428
+    },
+    {
+      "epoch": 3.5378621378621378,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9046,
+      "step": 4429
+    },
+    {
+      "epoch": 3.538661338661339,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0002,
+      "loss": 0.899,
+      "step": 4430
+    },
+    {
+      "epoch": 3.5394605394605394,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0002,
+      "loss": 0.89,
+      "step": 4431
+    },
+    {
+      "epoch": 3.5402597402597404,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8965,
+      "step": 4432
+    },
+    {
+      "epoch": 3.541058941058941,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0002,
+      "loss": 0.9007,
+      "step": 4433
+    },
+    {
+      "epoch": 3.5418581418581416,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9063,
+      "step": 4434
+    },
+    {
+      "epoch": 3.5426573426573427,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0002,
+      "loss": 0.9016,
+      "step": 4435
+    },
+    {
+      "epoch": 3.5434565434565437,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0002,
+      "loss": 0.9044,
+      "step": 4436
+    },
+    {
+      "epoch": 3.5442557442557443,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0002,
+      "loss": 0.9052,
+      "step": 4437
+    },
+    {
+      "epoch": 3.545054945054945,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0002,
+      "loss": 0.9018,
+      "step": 4438
+    },
+    {
+      "epoch": 3.545854145854146,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8979,
+      "step": 4439
+    },
+    {
+      "epoch": 3.5466533466533465,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0002,
+      "loss": 0.9036,
+      "step": 4440
+    },
+    {
+      "epoch": 3.5474525474525476,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8968,
+      "step": 4441
+    },
+    {
+      "epoch": 3.548251748251748,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.0002,
+      "loss": 0.8964,
+      "step": 4442
+    },
+    {
+      "epoch": 3.549050949050949,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0002,
+      "loss": 0.9092,
+      "step": 4443
+    },
+    {
+      "epoch": 3.54985014985015,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8943,
+      "step": 4444
+    },
+    {
+      "epoch": 3.5506493506493504,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8997,
+      "step": 4445
+    },
+    {
+      "epoch": 3.5514485514485514,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.0002,
+      "loss": 0.901,
+      "step": 4446
+    },
+    {
+      "epoch": 3.5522477522477525,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0002,
+      "loss": 0.8934,
+      "step": 4447
+    },
+    {
+      "epoch": 3.553046953046953,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0002,
+      "loss": 0.9007,
+      "step": 4448
+    },
+    {
+      "epoch": 3.5538461538461537,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0002,
+      "loss": 0.9084,
+      "step": 4449
+    },
+    {
+      "epoch": 3.5546453546453547,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8961,
+      "step": 4450
+    },
+    {
+      "epoch": 3.5554445554445553,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9042,
+      "step": 4451
+    },
+    {
+      "epoch": 3.5562437562437563,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0002,
+      "loss": 0.8987,
+      "step": 4452
+    },
+    {
+      "epoch": 3.557042957042957,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0002,
+      "loss": 0.909,
+      "step": 4453
+    },
+    {
+      "epoch": 3.557842157842158,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0002,
+      "loss": 0.906,
+      "step": 4454
+    },
+    {
+      "epoch": 3.5586413586413586,
+      "grad_norm": 0.8828125,
+      "learning_rate": 0.0002,
+      "loss": 0.9071,
+      "step": 4455
+    },
+    {
+      "epoch": 3.5594405594405596,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0002,
+      "loss": 0.9016,
+      "step": 4456
+    },
+    {
+      "epoch": 3.56023976023976,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9004,
+      "step": 4457
+    },
+    {
+      "epoch": 3.5610389610389612,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.0002,
+      "loss": 0.896,
+      "step": 4458
+    },
+    {
+      "epoch": 3.561838161838162,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.0002,
+      "loss": 0.9049,
+      "step": 4459
+    },
+    {
+      "epoch": 3.5626373626373624,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8996,
+      "step": 4460
+    },
+    {
+      "epoch": 3.5634365634365635,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.0002,
+      "loss": 0.8997,
+      "step": 4461
+    },
+    {
+      "epoch": 3.5642357642357645,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8994,
+      "step": 4462
+    },
+    {
+      "epoch": 3.565034965034965,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0002,
+      "loss": 0.8984,
+      "step": 4463
+    },
+    {
+      "epoch": 3.5658341658341657,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9031,
+      "step": 4464
+    },
+    {
+      "epoch": 3.5666333666333667,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0002,
+      "loss": 0.8975,
+      "step": 4465
+    },
+    {
+      "epoch": 3.5674325674325673,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0002,
+      "loss": 0.8997,
+      "step": 4466
+    },
+    {
+      "epoch": 3.5682317682317684,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0002,
+      "loss": 0.9001,
+      "step": 4467
+    },
+    {
+      "epoch": 3.569030969030969,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0002,
+      "loss": 0.9067,
+      "step": 4468
+    },
+    {
+      "epoch": 3.56983016983017,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0002,
+      "loss": 0.8996,
+      "step": 4469
+    },
+    {
+      "epoch": 3.5706293706293706,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0002,
+      "loss": 0.8984,
+      "step": 4470
+    },
+    {
+      "epoch": 3.571428571428571,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9034,
+      "step": 4471
+    },
+    {
+      "epoch": 3.572227772227772,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0002,
+      "loss": 0.9012,
+      "step": 4472
+    },
+    {
+      "epoch": 3.5730269730269733,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8977,
+      "step": 4473
+    },
+    {
+      "epoch": 3.573826173826174,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.0002,
+      "loss": 0.8945,
+      "step": 4474
+    },
+    {
+      "epoch": 3.5746253746253744,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0002,
+      "loss": 0.9028,
+      "step": 4475
+    },
+    {
+      "epoch": 3.5754245754245755,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8971,
+      "step": 4476
+    },
+    {
+      "epoch": 3.576223776223776,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.0002,
+      "loss": 0.8954,
+      "step": 4477
+    },
+    {
+      "epoch": 3.577022977022977,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0002,
+      "loss": 0.9021,
+      "step": 4478
+    },
+    {
+      "epoch": 3.5778221778221777,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0002,
+      "loss": 0.9086,
+      "step": 4479
+    },
+    {
+      "epoch": 3.5786213786213787,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0002,
+      "loss": 0.8984,
+      "step": 4480
+    },
+    {
+      "epoch": 3.5794205794205793,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0002,
+      "loss": 0.9002,
+      "step": 4481
+    },
+    {
+      "epoch": 3.5802197802197804,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0002,
+      "loss": 0.9006,
+      "step": 4482
+    },
+    {
+      "epoch": 3.581018981018981,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8999,
+      "step": 4483
+    },
+    {
+      "epoch": 3.581818181818182,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0002,
+      "loss": 0.9074,
+      "step": 4484
+    },
+    {
+      "epoch": 3.5826173826173826,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8981,
+      "step": 4485
+    },
+    {
+      "epoch": 3.583416583416583,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0002,
+      "loss": 0.9013,
+      "step": 4486
+    },
+    {
+      "epoch": 3.5842157842157842,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.0002,
+      "loss": 0.9152,
+      "step": 4487
+    },
+    {
+      "epoch": 3.5850149850149853,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0002,
+      "loss": 0.9043,
+      "step": 4488
+    },
+    {
+      "epoch": 3.585814185814186,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0002,
+      "loss": 0.9065,
+      "step": 4489
+    },
+    {
+      "epoch": 3.5866133866133865,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9088,
+      "step": 4490
+    },
+    {
+      "epoch": 3.5874125874125875,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8955,
+      "step": 4491
+    },
+    {
+      "epoch": 3.588211788211788,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0002,
+      "loss": 0.894,
+      "step": 4492
+    },
+    {
+      "epoch": 3.589010989010989,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8858,
+      "step": 4493
+    },
+    {
+      "epoch": 3.5898101898101897,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0002,
+      "loss": 0.8947,
+      "step": 4494
+    },
+    {
+      "epoch": 3.5906093906093908,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0002,
+      "loss": 0.9083,
+      "step": 4495
+    },
+    {
+      "epoch": 3.5914085914085914,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9032,
+      "step": 4496
+    },
+    {
+      "epoch": 3.592207792207792,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0002,
+      "loss": 0.8988,
+      "step": 4497
+    },
+    {
+      "epoch": 3.593006993006993,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.0002,
+      "loss": 0.8989,
+      "step": 4498
+    },
+    {
+      "epoch": 3.593806193806194,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8957,
+      "step": 4499
+    },
+    {
+      "epoch": 3.5946053946053946,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0002,
+      "loss": 0.908,
+      "step": 4500
+    },
+    {
+      "epoch": 3.5954045954045952,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8982,
+      "step": 4501
+    },
+    {
+      "epoch": 3.5962037962037963,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0002,
+      "loss": 0.9049,
+      "step": 4502
+    },
+    {
+      "epoch": 3.597002997002997,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0002,
+      "loss": 0.8937,
+      "step": 4503
+    },
+    {
+      "epoch": 3.597802197802198,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.0002,
+      "loss": 0.9073,
+      "step": 4504
+    },
+    {
+      "epoch": 3.5986013986013985,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0002,
+      "loss": 0.8943,
+      "step": 4505
+    },
+    {
+      "epoch": 3.5994005994005995,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8917,
+      "step": 4506
+    },
+    {
+      "epoch": 3.6001998001998,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9021,
+      "step": 4507
+    },
+    {
+      "epoch": 3.600999000999001,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0002,
+      "loss": 0.9028,
+      "step": 4508
+    },
+    {
+      "epoch": 3.6017982017982018,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0002,
+      "loss": 0.9019,
+      "step": 4509
+    },
+    {
+      "epoch": 3.602597402597403,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0002,
+      "loss": 0.8946,
+      "step": 4510
+    },
+    {
+      "epoch": 3.6033966033966034,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0002,
+      "loss": 0.9033,
+      "step": 4511
+    },
+    {
+      "epoch": 3.604195804195804,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0002,
+      "loss": 0.9013,
+      "step": 4512
+    },
+    {
+      "epoch": 3.604995004995005,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9096,
+      "step": 4513
+    },
+    {
+      "epoch": 3.605794205794206,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9073,
+      "step": 4514
+    },
+    {
+      "epoch": 3.6065934065934067,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9006,
+      "step": 4515
+    },
+    {
+      "epoch": 3.6073926073926073,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9036,
+      "step": 4516
+    },
+    {
+      "epoch": 3.6081918081918083,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0002,
+      "loss": 0.9042,
+      "step": 4517
+    },
+    {
+      "epoch": 3.608991008991009,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.0002,
+      "loss": 0.9035,
+      "step": 4518
+    },
+    {
+      "epoch": 3.60979020979021,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0002,
+      "loss": 0.8941,
+      "step": 4519
+    },
+    {
+      "epoch": 3.6105894105894105,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0002,
+      "loss": 0.9044,
+      "step": 4520
+    },
+    {
+      "epoch": 3.6113886113886116,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0002,
+      "loss": 0.9056,
+      "step": 4521
+    },
+    {
+      "epoch": 3.612187812187812,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0002,
+      "loss": 0.8968,
+      "step": 4522
+    },
+    {
+      "epoch": 3.6129870129870127,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0002,
+      "loss": 0.8979,
+      "step": 4523
+    },
+    {
+      "epoch": 3.613786213786214,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.0002,
+      "loss": 0.8964,
+      "step": 4524
+    },
+    {
+      "epoch": 3.614585414585415,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9073,
+      "step": 4525
+    },
+    {
+      "epoch": 3.6153846153846154,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8997,
+      "step": 4526
+    },
+    {
+      "epoch": 3.616183816183816,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9029,
+      "step": 4527
+    },
+    {
+      "epoch": 3.616983016983017,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8968,
+      "step": 4528
+    },
+    {
+      "epoch": 3.6177822177822176,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8984,
+      "step": 4529
+    },
+    {
+      "epoch": 3.6185814185814187,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0002,
+      "loss": 0.8989,
+      "step": 4530
+    },
+    {
+      "epoch": 3.6193806193806193,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8967,
+      "step": 4531
+    },
+    {
+      "epoch": 3.6201798201798203,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0002,
+      "loss": 0.9081,
+      "step": 4532
+    },
+    {
+      "epoch": 3.620979020979021,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0002,
+      "loss": 0.907,
+      "step": 4533
+    },
+    {
+      "epoch": 3.6217782217782215,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0002,
+      "loss": 0.893,
+      "step": 4534
+    },
+    {
+      "epoch": 3.6225774225774225,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9006,
+      "step": 4535
+    },
+    {
+      "epoch": 3.6233766233766236,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9004,
+      "step": 4536
+    },
+    {
+      "epoch": 3.624175824175824,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0002,
+      "loss": 0.8963,
+      "step": 4537
+    },
+    {
+      "epoch": 3.6249750249750248,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.0002,
+      "loss": 0.8995,
+      "step": 4538
+    },
+    {
+      "epoch": 3.625774225774226,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0002,
+      "loss": 0.8948,
+      "step": 4539
+    },
+    {
+      "epoch": 3.626573426573427,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0002,
+      "loss": 0.9021,
+      "step": 4540
+    },
+    {
+      "epoch": 3.6273726273726274,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.0002,
+      "loss": 0.8989,
+      "step": 4541
+    },
+    {
+      "epoch": 3.628171828171828,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8948,
+      "step": 4542
+    },
+    {
+      "epoch": 3.628971028971029,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9044,
+      "step": 4543
+    },
+    {
+      "epoch": 3.6297702297702297,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0002,
+      "loss": 0.9012,
+      "step": 4544
+    },
+    {
+      "epoch": 3.6305694305694307,
+      "grad_norm": 0.92578125,
+      "learning_rate": 0.0002,
+      "loss": 0.9034,
+      "step": 4545
+    },
+    {
+      "epoch": 3.6313686313686313,
+      "grad_norm": 0.92578125,
+      "learning_rate": 0.0002,
+      "loss": 0.9159,
+      "step": 4546
+    },
+    {
+      "epoch": 3.6321678321678323,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8969,
+      "step": 4547
+    },
+    {
+      "epoch": 3.632967032967033,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0002,
+      "loss": 0.9066,
+      "step": 4548
+    },
+    {
+      "epoch": 3.6337662337662335,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9138,
+      "step": 4549
+    },
+    {
+      "epoch": 3.6345654345654346,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0002,
+      "loss": 0.8989,
+      "step": 4550
+    },
+    {
+      "epoch": 3.6353646353646356,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0002,
+      "loss": 0.9034,
+      "step": 4551
+    },
+    {
+      "epoch": 3.636163836163836,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.0002,
+      "loss": 0.897,
+      "step": 4552
+    },
+    {
+      "epoch": 3.636963036963037,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0002,
+      "loss": 0.8977,
+      "step": 4553
+    },
+    {
+      "epoch": 3.637762237762238,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.0002,
+      "loss": 0.8953,
+      "step": 4554
+    },
+    {
+      "epoch": 3.6385614385614384,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8932,
+      "step": 4555
+    },
+    {
+      "epoch": 3.6393606393606395,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9009,
+      "step": 4556
+    },
+    {
+      "epoch": 3.64015984015984,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8963,
+      "step": 4557
+    },
+    {
+      "epoch": 3.640959040959041,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9038,
+      "step": 4558
+    },
+    {
+      "epoch": 3.6417582417582417,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8996,
+      "step": 4559
+    },
+    {
+      "epoch": 3.6425574425574423,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9014,
+      "step": 4560
+    },
+    {
+      "epoch": 3.6433566433566433,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8905,
+      "step": 4561
+    },
+    {
+      "epoch": 3.6441558441558444,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.0002,
+      "loss": 0.8999,
+      "step": 4562
+    },
+    {
+      "epoch": 3.644955044955045,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0002,
+      "loss": 0.9028,
+      "step": 4563
+    },
+    {
+      "epoch": 3.6457542457542456,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0002,
+      "loss": 0.9021,
+      "step": 4564
+    },
+    {
+      "epoch": 3.6465534465534466,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0002,
+      "loss": 0.8975,
+      "step": 4565
+    },
+    {
+      "epoch": 3.647352647352647,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0002,
+      "loss": 0.8974,
+      "step": 4566
+    },
+    {
+      "epoch": 3.6481518481518482,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0002,
+      "loss": 0.8993,
+      "step": 4567
+    },
+    {
+      "epoch": 3.648951048951049,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8927,
+      "step": 4568
+    },
+    {
+      "epoch": 3.64975024975025,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9026,
+      "step": 4569
+    },
+    {
+      "epoch": 3.6505494505494505,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9027,
+      "step": 4570
+    },
+    {
+      "epoch": 3.6513486513486515,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0002,
+      "loss": 0.8928,
+      "step": 4571
+    },
+    {
+      "epoch": 3.652147852147852,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8808,
+      "step": 4572
+    },
+    {
+      "epoch": 3.652947052947053,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9047,
+      "step": 4573
+    },
+    {
+      "epoch": 3.6537462537462537,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0002,
+      "loss": 0.8981,
+      "step": 4574
+    },
+    {
+      "epoch": 3.6545454545454543,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8904,
+      "step": 4575
+    },
+    {
+      "epoch": 3.6553446553446554,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.0002,
+      "loss": 0.9043,
+      "step": 4576
+    },
+    {
+      "epoch": 3.6561438561438564,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.0002,
+      "loss": 0.8996,
+      "step": 4577
+    },
+    {
+      "epoch": 3.656943056943057,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0002,
+      "loss": 0.9426,
+      "step": 4578
+    },
+    {
+      "epoch": 3.6577422577422576,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9027,
+      "step": 4579
+    },
+    {
+      "epoch": 3.6585414585414586,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0002,
+      "loss": 0.901,
+      "step": 4580
+    },
+    {
+      "epoch": 3.659340659340659,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.0002,
+      "loss": 0.9007,
+      "step": 4581
+    },
+    {
+      "epoch": 3.6601398601398603,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.0002,
+      "loss": 0.8996,
+      "step": 4582
+    },
+    {
+      "epoch": 3.660939060939061,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9009,
+      "step": 4583
+    },
+    {
+      "epoch": 3.661738261738262,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0002,
+      "loss": 0.9082,
+      "step": 4584
+    },
+    {
+      "epoch": 3.6625374625374625,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8922,
+      "step": 4585
+    },
+    {
+      "epoch": 3.663336663336663,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0002,
+      "loss": 0.8926,
+      "step": 4586
+    },
+    {
+      "epoch": 3.664135864135864,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0002,
+      "loss": 0.8975,
+      "step": 4587
+    },
+    {
+      "epoch": 3.664935064935065,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0002,
+      "loss": 0.8982,
+      "step": 4588
+    },
+    {
+      "epoch": 3.6657342657342658,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8938,
+      "step": 4589
+    },
+    {
+      "epoch": 3.6665334665334663,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8974,
+      "step": 4590
+    },
+    {
+      "epoch": 3.6673326673326674,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0002,
+      "loss": 0.887,
+      "step": 4591
+    },
+    {
+      "epoch": 3.668131868131868,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0002,
+      "loss": 0.899,
+      "step": 4592
+    },
+    {
+      "epoch": 3.668931068931069,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0002,
+      "loss": 0.8967,
+      "step": 4593
+    },
+    {
+      "epoch": 3.6697302697302696,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8953,
+      "step": 4594
+    },
+    {
+      "epoch": 3.6705294705294707,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0002,
+      "loss": 0.8986,
+      "step": 4595
+    },
+    {
+      "epoch": 3.6713286713286712,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0002,
+      "loss": 0.9027,
+      "step": 4596
+    },
+    {
+      "epoch": 3.6721278721278723,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0002,
+      "loss": 0.8939,
+      "step": 4597
+    },
+    {
+      "epoch": 3.672927072927073,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0002,
+      "loss": 0.8956,
+      "step": 4598
+    },
+    {
+      "epoch": 3.673726273726274,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0002,
+      "loss": 0.8923,
+      "step": 4599
+    },
+    {
+      "epoch": 3.6745254745254745,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0002,
+      "loss": 0.9008,
+      "step": 4600
+    },
+    {
+      "epoch": 3.675324675324675,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0002,
+      "loss": 0.897,
+      "step": 4601
+    },
+    {
+      "epoch": 3.676123876123876,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8914,
+      "step": 4602
+    },
+    {
+      "epoch": 3.676923076923077,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0002,
+      "loss": 0.8822,
+      "step": 4603
+    },
+    {
+      "epoch": 3.6777222777222778,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9047,
+      "step": 4604
+    },
+    {
+      "epoch": 3.6785214785214784,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0002,
+      "loss": 0.8998,
+      "step": 4605
+    },
+    {
+      "epoch": 3.6793206793206794,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0002,
+      "loss": 0.9007,
+      "step": 4606
+    },
+    {
+      "epoch": 3.68011988011988,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0002,
+      "loss": 0.9049,
+      "step": 4607
+    },
+    {
+      "epoch": 3.680919080919081,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002,
+      "loss": 0.901,
+      "step": 4608
+    },
+    {
+      "epoch": 3.6817182817182816,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0002,
+      "loss": 0.9012,
+      "step": 4609
+    },
+    {
+      "epoch": 3.6825174825174827,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0002,
+      "loss": 0.8941,
+      "step": 4610
+    },
+    {
+      "epoch": 3.6833166833166833,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0002,
+      "loss": 0.8968,
+      "step": 4611
+    },
+    {
+      "epoch": 3.684115884115884,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8981,
+      "step": 4612
+    },
+    {
+      "epoch": 3.684915084915085,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0002,
+      "loss": 0.8943,
+      "step": 4613
+    },
+    {
+      "epoch": 3.685714285714286,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8962,
+      "step": 4614
+    },
+    {
+      "epoch": 3.6865134865134865,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0002,
+      "loss": 0.8988,
+      "step": 4615
+    },
+    {
+      "epoch": 3.687312687312687,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8982,
+      "step": 4616
+    },
+    {
+      "epoch": 3.688111888111888,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8958,
+      "step": 4617
+    },
+    {
+      "epoch": 3.6889110889110888,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0002,
+      "loss": 0.8979,
+      "step": 4618
+    },
+    {
+      "epoch": 3.68971028971029,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0002,
+      "loss": 0.9044,
+      "step": 4619
+    },
+    {
+      "epoch": 3.6905094905094904,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9055,
+      "step": 4620
+    },
+    {
+      "epoch": 3.6913086913086914,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9106,
+      "step": 4621
+    },
+    {
+      "epoch": 3.692107892107892,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.0002,
+      "loss": 0.8942,
+      "step": 4622
+    },
+    {
+      "epoch": 3.692907092907093,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8922,
+      "step": 4623
+    },
+    {
+      "epoch": 3.6937062937062937,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9004,
+      "step": 4624
+    },
+    {
+      "epoch": 3.6945054945054947,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0002,
+      "loss": 0.903,
+      "step": 4625
+    },
+    {
+      "epoch": 3.6953046953046953,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0002,
+      "loss": 0.8957,
+      "step": 4626
+    },
+    {
+      "epoch": 3.696103896103896,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0002,
+      "loss": 0.9117,
+      "step": 4627
+    },
+    {
+      "epoch": 3.696903096903097,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9036,
+      "step": 4628
+    },
+    {
+      "epoch": 3.697702297702298,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0002,
+      "loss": 0.8982,
+      "step": 4629
+    },
+    {
+      "epoch": 3.6985014985014986,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8984,
+      "step": 4630
+    },
+    {
+      "epoch": 3.699300699300699,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8946,
+      "step": 4631
+    },
+    {
+      "epoch": 3.7000999000999,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0002,
+      "loss": 0.8939,
+      "step": 4632
+    },
+    {
+      "epoch": 3.700899100899101,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0002,
+      "loss": 0.8866,
+      "step": 4633
+    },
+    {
+      "epoch": 3.701698301698302,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0002,
+      "loss": 0.9047,
+      "step": 4634
+    },
+    {
+      "epoch": 3.7024975024975024,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0002,
+      "loss": 0.9,
+      "step": 4635
+    },
+    {
+      "epoch": 3.7032967032967035,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0002,
+      "loss": 0.897,
+      "step": 4636
+    },
+    {
+      "epoch": 3.704095904095904,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8981,
+      "step": 4637
+    },
+    {
+      "epoch": 3.7048951048951047,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8953,
+      "step": 4638
+    },
+    {
+      "epoch": 3.7056943056943057,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8951,
+      "step": 4639
+    },
+    {
+      "epoch": 3.7064935064935067,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0002,
+      "loss": 0.898,
+      "step": 4640
+    },
+    {
+      "epoch": 3.7072927072927073,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0002,
+      "loss": 0.8985,
+      "step": 4641
+    },
+    {
+      "epoch": 3.708091908091908,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0002,
+      "loss": 0.8942,
+      "step": 4642
+    },
+    {
+      "epoch": 3.708891108891109,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8923,
+      "step": 4643
+    },
+    {
+      "epoch": 3.7096903096903096,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0002,
+      "loss": 0.8977,
+      "step": 4644
+    },
+    {
+      "epoch": 3.7104895104895106,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0002,
+      "loss": 0.9054,
+      "step": 4645
+    },
+    {
+      "epoch": 3.711288711288711,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8983,
+      "step": 4646
+    },
+    {
+      "epoch": 3.7120879120879122,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0002,
+      "loss": 0.9024,
+      "step": 4647
+    },
+    {
+      "epoch": 3.712887112887113,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0002,
+      "loss": 0.8936,
+      "step": 4648
+    },
+    {
+      "epoch": 3.7136863136863134,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8958,
+      "step": 4649
+    },
+    {
+      "epoch": 3.7144855144855145,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9081,
+      "step": 4650
+    },
+    {
+      "epoch": 3.7152847152847155,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9014,
+      "step": 4651
+    },
+    {
+      "epoch": 3.716083916083916,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0002,
+      "loss": 0.9034,
+      "step": 4652
+    },
+    {
+      "epoch": 3.7168831168831167,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9117,
+      "step": 4653
+    },
+    {
+      "epoch": 3.7176823176823177,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8974,
+      "step": 4654
+    },
+    {
+      "epoch": 3.7184815184815188,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0002,
+      "loss": 0.908,
+      "step": 4655
+    },
+    {
+      "epoch": 3.7192807192807193,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0002,
+      "loss": 0.8903,
+      "step": 4656
+    },
+    {
+      "epoch": 3.72007992007992,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0002,
+      "loss": 0.9004,
+      "step": 4657
+    },
+    {
+      "epoch": 3.720879120879121,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0002,
+      "loss": 0.8966,
+      "step": 4658
+    },
+    {
+      "epoch": 3.7216783216783216,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0002,
+      "loss": 0.9009,
+      "step": 4659
+    },
+    {
+      "epoch": 3.7224775224775226,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.0002,
+      "loss": 0.8956,
+      "step": 4660
+    },
+    {
+      "epoch": 3.723276723276723,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0002,
+      "loss": 0.9016,
+      "step": 4661
+    },
+    {
+      "epoch": 3.7240759240759242,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0002,
+      "loss": 0.9044,
+      "step": 4662
+    },
+    {
+      "epoch": 3.724875124875125,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9097,
+      "step": 4663
+    },
+    {
+      "epoch": 3.7256743256743254,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0002,
+      "loss": 0.9037,
+      "step": 4664
+    },
+    {
+      "epoch": 3.7264735264735265,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0002,
+      "loss": 0.8967,
+      "step": 4665
+    },
+    {
+      "epoch": 3.7272727272727275,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8995,
+      "step": 4666
+    },
+    {
+      "epoch": 3.728071928071928,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0002,
+      "loss": 0.883,
+      "step": 4667
+    },
+    {
+      "epoch": 3.7288711288711287,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8818,
+      "step": 4668
+    },
+    {
+      "epoch": 3.7296703296703297,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8958,
+      "step": 4669
+    },
+    {
+      "epoch": 3.7304695304695303,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8934,
+      "step": 4670
+    },
+    {
+      "epoch": 3.7312687312687314,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0002,
+      "loss": 0.8991,
+      "step": 4671
+    },
+    {
+      "epoch": 3.732067932067932,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.0002,
+      "loss": 0.8954,
+      "step": 4672
+    },
+    {
+      "epoch": 3.732867132867133,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.0002,
+      "loss": 0.937,
+      "step": 4673
+    },
+    {
+      "epoch": 3.7336663336663336,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0002,
+      "loss": 0.8998,
+      "step": 4674
+    },
+    {
+      "epoch": 3.734465534465534,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.0002,
+      "loss": 0.8992,
+      "step": 4675
+    },
+    {
+      "epoch": 3.7352647352647352,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0002,
+      "loss": 0.8948,
+      "step": 4676
+    },
+    {
+      "epoch": 3.7360639360639363,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9039,
+      "step": 4677
+    },
+    {
+      "epoch": 3.736863136863137,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8977,
+      "step": 4678
+    },
+    {
+      "epoch": 3.7376623376623375,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0002,
+      "loss": 0.8993,
+      "step": 4679
+    },
+    {
+      "epoch": 3.7384615384615385,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9403,
+      "step": 4680
+    },
+    {
+      "epoch": 3.739260739260739,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0002,
+      "loss": 0.9033,
+      "step": 4681
+    },
+    {
+      "epoch": 3.74005994005994,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9007,
+      "step": 4682
+    },
+    {
+      "epoch": 3.7408591408591407,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9032,
+      "step": 4683
+    },
+    {
+      "epoch": 3.7416583416583418,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0002,
+      "loss": 0.8908,
+      "step": 4684
+    },
+    {
+      "epoch": 3.7424575424575424,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0002,
+      "loss": 0.8987,
+      "step": 4685
+    },
+    {
+      "epoch": 3.7432567432567434,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8937,
+      "step": 4686
+    },
+    {
+      "epoch": 3.744055944055944,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0002,
+      "loss": 0.8907,
+      "step": 4687
+    },
+    {
+      "epoch": 3.744855144855145,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0002,
+      "loss": 0.911,
+      "step": 4688
+    },
+    {
+      "epoch": 3.7456543456543456,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0002,
+      "loss": 0.8874,
+      "step": 4689
+    },
+    {
+      "epoch": 3.7464535464535462,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0002,
+      "loss": 0.8968,
+      "step": 4690
+    },
+    {
+      "epoch": 3.7472527472527473,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0002,
+      "loss": 0.8946,
+      "step": 4691
+    },
+    {
+      "epoch": 3.7480519480519483,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0002,
+      "loss": 0.8962,
+      "step": 4692
+    },
+    {
+      "epoch": 3.748851148851149,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.0002,
+      "loss": 0.8908,
+      "step": 4693
+    },
+    {
+      "epoch": 3.7496503496503495,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9039,
+      "step": 4694
+    },
+    {
+      "epoch": 3.7504495504495505,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9013,
+      "step": 4695
+    },
+    {
+      "epoch": 3.751248751248751,
+      "grad_norm": 1.875,
+      "learning_rate": 0.0002,
+      "loss": 0.9421,
+      "step": 4696
+    },
+    {
+      "epoch": 3.752047952047952,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0002,
+      "loss": 0.9017,
+      "step": 4697
+    },
+    {
+      "epoch": 3.7528471528471528,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8979,
+      "step": 4698
+    },
+    {
+      "epoch": 3.753646353646354,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8874,
+      "step": 4699
+    },
+    {
+      "epoch": 3.7544455544455544,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0002,
+      "loss": 0.9036,
+      "step": 4700
+    },
+    {
+      "epoch": 3.755244755244755,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0002,
+      "loss": 0.8921,
+      "step": 4701
+    },
+    {
+      "epoch": 3.756043956043956,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0002,
+      "loss": 0.8931,
+      "step": 4702
+    },
+    {
+      "epoch": 3.756843156843157,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8967,
+      "step": 4703
+    },
+    {
+      "epoch": 3.7576423576423577,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8929,
+      "step": 4704
+    },
+    {
+      "epoch": 3.7584415584415583,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.0002,
+      "loss": 0.9157,
+      "step": 4705
+    },
+    {
+      "epoch": 3.7592407592407593,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0002,
+      "loss": 0.8894,
+      "step": 4706
+    },
+    {
+      "epoch": 3.76003996003996,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8979,
+      "step": 4707
+    },
+    {
+      "epoch": 3.760839160839161,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8932,
+      "step": 4708
+    },
+    {
+      "epoch": 3.7616383616383615,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0002,
+      "loss": 0.9048,
+      "step": 4709
+    },
+    {
+      "epoch": 3.7624375624375626,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0002,
+      "loss": 0.9003,
+      "step": 4710
+    },
+    {
+      "epoch": 3.763236763236763,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8955,
+      "step": 4711
+    },
+    {
+      "epoch": 3.764035964035964,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0002,
+      "loss": 0.894,
+      "step": 4712
+    },
+    {
+      "epoch": 3.764835164835165,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0002,
+      "loss": 0.8979,
+      "step": 4713
+    },
+    {
+      "epoch": 3.765634365634366,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8915,
+      "step": 4714
+    },
+    {
+      "epoch": 3.7664335664335664,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8997,
+      "step": 4715
+    },
+    {
+      "epoch": 3.767232767232767,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8968,
+      "step": 4716
+    },
+    {
+      "epoch": 3.768031968031968,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9061,
+      "step": 4717
+    },
+    {
+      "epoch": 3.768831168831169,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0002,
+      "loss": 0.9058,
+      "step": 4718
+    },
+    {
+      "epoch": 3.7696303696303697,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0002,
+      "loss": 0.8974,
+      "step": 4719
+    },
+    {
+      "epoch": 3.7704295704295703,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9022,
+      "step": 4720
+    },
+    {
+      "epoch": 3.7712287712287713,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0002,
+      "loss": 0.9076,
+      "step": 4721
+    },
+    {
+      "epoch": 3.772027972027972,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8887,
+      "step": 4722
+    },
+    {
+      "epoch": 3.772827172827173,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8825,
+      "step": 4723
+    },
+    {
+      "epoch": 3.7736263736263735,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0002,
+      "loss": 0.9013,
+      "step": 4724
+    },
+    {
+      "epoch": 3.7744255744255746,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.0002,
+      "loss": 0.9075,
+      "step": 4725
+    },
+    {
+      "epoch": 3.775224775224775,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8929,
+      "step": 4726
+    },
+    {
+      "epoch": 3.7760239760239758,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0002,
+      "loss": 0.8927,
+      "step": 4727
+    },
+    {
+      "epoch": 3.776823176823177,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8997,
+      "step": 4728
+    },
+    {
+      "epoch": 3.777622377622378,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9036,
+      "step": 4729
+    },
+    {
+      "epoch": 3.7784215784215784,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8939,
+      "step": 4730
+    },
+    {
+      "epoch": 3.779220779220779,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.0002,
+      "loss": 0.8938,
+      "step": 4731
+    },
+    {
+      "epoch": 3.78001998001998,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0002,
+      "loss": 0.895,
+      "step": 4732
+    },
+    {
+      "epoch": 3.7808191808191807,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.0002,
+      "loss": 0.9014,
+      "step": 4733
+    },
+    {
+      "epoch": 3.7816183816183817,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8936,
+      "step": 4734
+    },
+    {
+      "epoch": 3.7824175824175823,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0002,
+      "loss": 0.8941,
+      "step": 4735
+    },
+    {
+      "epoch": 3.7832167832167833,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.0002,
+      "loss": 0.8946,
+      "step": 4736
+    },
+    {
+      "epoch": 3.784015984015984,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0002,
+      "loss": 0.898,
+      "step": 4737
+    },
+    {
+      "epoch": 3.784815184815185,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0002,
+      "loss": 0.8984,
+      "step": 4738
+    },
+    {
+      "epoch": 3.7856143856143856,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0002,
+      "loss": 0.9022,
+      "step": 4739
+    },
+    {
+      "epoch": 3.7864135864135866,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0002,
+      "loss": 0.8992,
+      "step": 4740
+    },
+    {
+      "epoch": 3.787212787212787,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0002,
+      "loss": 0.905,
+      "step": 4741
+    },
+    {
+      "epoch": 3.788011988011988,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8929,
+      "step": 4742
+    },
+    {
+      "epoch": 3.788811188811189,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0002,
+      "loss": 0.8841,
+      "step": 4743
+    },
+    {
+      "epoch": 3.78961038961039,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0002,
+      "loss": 0.8976,
+      "step": 4744
+    },
+    {
+      "epoch": 3.7904095904095905,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0002,
+      "loss": 0.912,
+      "step": 4745
+    },
+    {
+      "epoch": 3.791208791208791,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0002,
+      "loss": 0.9083,
+      "step": 4746
+    },
+    {
+      "epoch": 3.792007992007992,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0002,
+      "loss": 0.906,
+      "step": 4747
+    },
+    {
+      "epoch": 3.7928071928071927,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8908,
+      "step": 4748
+    },
+    {
+      "epoch": 3.7936063936063937,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8988,
+      "step": 4749
+    },
+    {
+      "epoch": 3.7944055944055943,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0002,
+      "loss": 0.898,
+      "step": 4750
+    },
+    {
+      "epoch": 3.7952047952047954,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8934,
+      "step": 4751
+    },
+    {
+      "epoch": 3.796003996003996,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9043,
+      "step": 4752
+    },
+    {
+      "epoch": 3.7968031968031966,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8885,
+      "step": 4753
+    },
+    {
+      "epoch": 3.7976023976023976,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9057,
+      "step": 4754
+    },
+    {
+      "epoch": 3.7984015984015986,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0002,
+      "loss": 0.8903,
+      "step": 4755
+    },
+    {
+      "epoch": 3.7992007992007992,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.0002,
+      "loss": 0.9021,
+      "step": 4756
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0002,
+      "loss": 0.8871,
+      "step": 4757
+    },
+    {
+      "epoch": 3.800799200799201,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.0002,
+      "loss": 0.8962,
+      "step": 4758
+    },
+    {
+      "epoch": 3.8015984015984015,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.0002,
+      "loss": 0.8991,
+      "step": 4759
+    },
+    {
+      "epoch": 3.8023976023976025,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0002,
+      "loss": 0.8954,
+      "step": 4760
+    },
+    {
+      "epoch": 3.803196803196803,
+      "grad_norm": 0.92578125,
+      "learning_rate": 0.0002,
+      "loss": 0.9148,
+      "step": 4761
+    },
+    {
+      "epoch": 3.803996003996004,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0002,
+      "loss": 0.9006,
+      "step": 4762
+    },
+    {
+      "epoch": 3.8047952047952047,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9075,
+      "step": 4763
+    },
+    {
+      "epoch": 3.8055944055944058,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9386,
+      "step": 4764
+    },
+    {
+      "epoch": 3.8063936063936064,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8996,
+      "step": 4765
+    },
+    {
+      "epoch": 3.8071928071928074,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0002,
+      "loss": 0.9003,
+      "step": 4766
+    },
+    {
+      "epoch": 3.807992007992008,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9008,
+      "step": 4767
+    },
+    {
+      "epoch": 3.8087912087912086,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8931,
+      "step": 4768
+    },
+    {
+      "epoch": 3.8095904095904096,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.0002,
+      "loss": 0.9109,
+      "step": 4769
+    },
+    {
+      "epoch": 3.8103896103896107,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8962,
+      "step": 4770
+    },
+    {
+      "epoch": 3.8111888111888113,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0002,
+      "loss": 0.9006,
+      "step": 4771
+    },
+    {
+      "epoch": 3.811988011988012,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0002,
+      "loss": 0.8933,
+      "step": 4772
+    },
+    {
+      "epoch": 3.812787212787213,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0002,
+      "loss": 0.889,
+      "step": 4773
+    },
+    {
+      "epoch": 3.8135864135864135,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0002,
+      "loss": 0.902,
+      "step": 4774
+    },
+    {
+      "epoch": 3.8143856143856145,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0002,
+      "loss": 0.891,
+      "step": 4775
+    },
+    {
+      "epoch": 3.815184815184815,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0002,
+      "loss": 0.9028,
+      "step": 4776
+    },
+    {
+      "epoch": 3.815984015984016,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0002,
+      "loss": 0.8889,
+      "step": 4777
+    },
+    {
+      "epoch": 3.8167832167832167,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8986,
+      "step": 4778
+    },
+    {
+      "epoch": 3.8175824175824173,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9046,
+      "step": 4779
+    },
+    {
+      "epoch": 3.8183816183816184,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0002,
+      "loss": 0.8981,
+      "step": 4780
+    },
+    {
+      "epoch": 3.8191808191808194,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8948,
+      "step": 4781
+    },
+    {
+      "epoch": 3.81998001998002,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0002,
+      "loss": 0.9001,
+      "step": 4782
+    },
+    {
+      "epoch": 3.8207792207792206,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0002,
+      "loss": 0.8915,
+      "step": 4783
+    },
+    {
+      "epoch": 3.8215784215784216,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8965,
+      "step": 4784
+    },
+    {
+      "epoch": 3.8223776223776222,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0002,
+      "loss": 0.8968,
+      "step": 4785
+    },
+    {
+      "epoch": 3.8231768231768233,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0002,
+      "loss": 0.9027,
+      "step": 4786
+    },
+    {
+      "epoch": 3.823976023976024,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0002,
+      "loss": 0.8966,
+      "step": 4787
+    },
+    {
+      "epoch": 3.824775224775225,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0002,
+      "loss": 0.8924,
+      "step": 4788
+    },
+    {
+      "epoch": 3.8255744255744255,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0002,
+      "loss": 0.8945,
+      "step": 4789
+    },
+    {
+      "epoch": 3.826373626373626,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8948,
+      "step": 4790
+    },
+    {
+      "epoch": 3.827172827172827,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0002,
+      "loss": 0.9047,
+      "step": 4791
+    },
+    {
+      "epoch": 3.827972027972028,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9016,
+      "step": 4792
+    },
+    {
+      "epoch": 3.8287712287712288,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0002,
+      "loss": 0.8946,
+      "step": 4793
+    },
+    {
+      "epoch": 3.8295704295704294,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8948,
+      "step": 4794
+    },
+    {
+      "epoch": 3.8303696303696304,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0002,
+      "loss": 0.9002,
+      "step": 4795
+    },
+    {
+      "epoch": 3.8311688311688314,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9008,
+      "step": 4796
+    },
+    {
+      "epoch": 3.831968031968032,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8952,
+      "step": 4797
+    },
+    {
+      "epoch": 3.8327672327672326,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8985,
+      "step": 4798
+    },
+    {
+      "epoch": 3.8335664335664337,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8954,
+      "step": 4799
+    },
+    {
+      "epoch": 3.8343656343656343,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0002,
+      "loss": 0.8956,
+      "step": 4800
+    },
+    {
+      "epoch": 3.8351648351648353,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0002,
+      "loss": 0.8855,
+      "step": 4801
+    },
+    {
+      "epoch": 3.835964035964036,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9054,
+      "step": 4802
+    },
+    {
+      "epoch": 3.836763236763237,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0002,
+      "loss": 0.8963,
+      "step": 4803
+    },
+    {
+      "epoch": 3.8375624375624375,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0002,
+      "loss": 0.9022,
+      "step": 4804
+    },
+    {
+      "epoch": 3.838361638361638,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9016,
+      "step": 4805
+    },
+    {
+      "epoch": 3.839160839160839,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8978,
+      "step": 4806
+    },
+    {
+      "epoch": 3.83996003996004,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0002,
+      "loss": 0.9015,
+      "step": 4807
+    },
+    {
+      "epoch": 3.840759240759241,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8956,
+      "step": 4808
+    },
+    {
+      "epoch": 3.8415584415584414,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8959,
+      "step": 4809
+    },
+    {
+      "epoch": 3.8423576423576424,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9041,
+      "step": 4810
+    },
+    {
+      "epoch": 3.843156843156843,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0002,
+      "loss": 0.8958,
+      "step": 4811
+    },
+    {
+      "epoch": 3.843956043956044,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.0002,
+      "loss": 0.9017,
+      "step": 4812
+    },
+    {
+      "epoch": 3.8447552447552447,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0002,
+      "loss": 0.8928,
+      "step": 4813
+    },
+    {
+      "epoch": 3.8455544455544457,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9005,
+      "step": 4814
+    },
+    {
+      "epoch": 3.8463536463536463,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8916,
+      "step": 4815
+    },
+    {
+      "epoch": 3.847152847152847,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0002,
+      "loss": 0.8893,
+      "step": 4816
+    },
+    {
+      "epoch": 3.847952047952048,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.0002,
+      "loss": 0.8906,
+      "step": 4817
+    },
+    {
+      "epoch": 3.848751248751249,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0002,
+      "loss": 0.9022,
+      "step": 4818
+    },
+    {
+      "epoch": 3.8495504495504496,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.0002,
+      "loss": 0.9026,
+      "step": 4819
+    },
+    {
+      "epoch": 3.85034965034965,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.0002,
+      "loss": 0.8886,
+      "step": 4820
+    },
+    {
+      "epoch": 3.851148851148851,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0002,
+      "loss": 0.9036,
+      "step": 4821
+    },
+    {
+      "epoch": 3.851948051948052,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9,
+      "step": 4822
+    },
+    {
+      "epoch": 3.852747252747253,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.0002,
+      "loss": 0.8927,
+      "step": 4823
+    },
+    {
+      "epoch": 3.8535464535464534,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0002,
+      "loss": 0.9027,
+      "step": 4824
+    },
+    {
+      "epoch": 3.8543456543456545,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0002,
+      "loss": 0.9028,
+      "step": 4825
+    },
+    {
+      "epoch": 3.855144855144855,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0002,
+      "loss": 0.8956,
+      "step": 4826
+    },
+    {
+      "epoch": 3.855944055944056,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.0002,
+      "loss": 0.8892,
+      "step": 4827
+    },
+    {
+      "epoch": 3.8567432567432567,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0002,
+      "loss": 0.8966,
+      "step": 4828
+    },
+    {
+      "epoch": 3.8575424575424577,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.0002,
+      "loss": 0.9002,
+      "step": 4829
+    },
+    {
+      "epoch": 3.8583416583416583,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8854,
+      "step": 4830
+    },
+    {
+      "epoch": 3.859140859140859,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8951,
+      "step": 4831
+    },
+    {
+      "epoch": 3.85994005994006,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.0002,
+      "loss": 0.8988,
+      "step": 4832
+    },
+    {
+      "epoch": 3.860739260739261,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.0002,
+      "loss": 0.8929,
+      "step": 4833
+    },
+    {
+      "epoch": 3.8615384615384616,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.0002,
+      "loss": 0.9,
+      "step": 4834
+    },
+    {
+      "epoch": 3.862337662337662,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.0002,
+      "loss": 0.899,
+      "step": 4835
+    },
+    {
+      "epoch": 3.863136863136863,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.0002,
+      "loss": 0.8991,
+      "step": 4836
+    },
+    {
+      "epoch": 3.863936063936064,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8985,
+      "step": 4837
+    },
+    {
+      "epoch": 3.864735264735265,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8982,
+      "step": 4838
+    },
+    {
+      "epoch": 3.8655344655344654,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8994,
+      "step": 4839
+    },
+    {
+      "epoch": 3.8663336663336665,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0002,
+      "loss": 0.8918,
+      "step": 4840
+    },
+    {
+      "epoch": 3.867132867132867,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9033,
+      "step": 4841
+    },
+    {
+      "epoch": 3.8679320679320677,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9026,
+      "step": 4842
+    },
+    {
+      "epoch": 3.8687312687312687,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0002,
+      "loss": 0.9028,
+      "step": 4843
+    },
+    {
+      "epoch": 3.8695304695304698,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0002,
+      "loss": 0.9073,
+      "step": 4844
+    },
+    {
+      "epoch": 3.8703296703296703,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0002,
+      "loss": 0.8981,
+      "step": 4845
+    },
+    {
+      "epoch": 3.871128871128871,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0002,
+      "loss": 0.8969,
+      "step": 4846
+    },
+    {
+      "epoch": 3.871928071928072,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.0002,
+      "loss": 0.8958,
+      "step": 4847
+    },
+    {
+      "epoch": 3.8727272727272726,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0002,
+      "loss": 0.895,
+      "step": 4848
+    },
+    {
+      "epoch": 3.8735264735264736,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0002,
+      "loss": 0.8935,
+      "step": 4849
+    },
+    {
+      "epoch": 3.874325674325674,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0002,
+      "loss": 0.8988,
+      "step": 4850
+    },
+    {
+      "epoch": 3.8751248751248752,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0002,
+      "loss": 0.898,
+      "step": 4851
+    },
+    {
+      "epoch": 3.875924075924076,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0002,
+      "loss": 0.892,
+      "step": 4852
+    },
+    {
+      "epoch": 3.876723276723277,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8983,
+      "step": 4853
+    },
+    {
+      "epoch": 3.8775224775224775,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0002,
+      "loss": 0.8867,
+      "step": 4854
+    },
+    {
+      "epoch": 3.8783216783216785,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0002,
+      "loss": 0.8963,
+      "step": 4855
+    },
+    {
+      "epoch": 3.879120879120879,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0002,
+      "loss": 0.8932,
+      "step": 4856
+    },
+    {
+      "epoch": 3.8799200799200797,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0002,
+      "loss": 0.8988,
+      "step": 4857
+    },
+    {
+      "epoch": 3.8807192807192807,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9009,
+      "step": 4858
+    },
+    {
+      "epoch": 3.8815184815184818,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0002,
+      "loss": 0.889,
+      "step": 4859
+    },
+    {
+      "epoch": 3.8823176823176824,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0002,
+      "loss": 0.893,
+      "step": 4860
+    },
+    {
+      "epoch": 3.883116883116883,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.0002,
+      "loss": 0.8995,
+      "step": 4861
+    },
+    {
+      "epoch": 3.883916083916084,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0002,
+      "loss": 0.9059,
+      "step": 4862
+    },
+    {
+      "epoch": 3.8847152847152846,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0002,
+      "loss": 0.8938,
+      "step": 4863
+    },
+    {
+      "epoch": 3.8855144855144856,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0002,
+      "loss": 0.8948,
+      "step": 4864
+    },
+    {
+      "epoch": 3.8863136863136862,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.0002,
+      "loss": 0.9462,
+      "step": 4865
+    },
+    {
+      "epoch": 3.8871128871128873,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0002,
+      "loss": 0.8997,
+      "step": 4866
+    },
+    {
+      "epoch": 3.887912087912088,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0002,
+      "loss": 0.8918,
+      "step": 4867
+    },
+    {
+      "epoch": 3.8887112887112885,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0002,
+      "loss": 0.8999,
+      "step": 4868
+    },
+    {
+      "epoch": 3.8895104895104895,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0002,
+      "loss": 0.8996,
+      "step": 4869
+    },
+    {
+      "epoch": 3.8903096903096905,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0002,
+      "loss": 0.9047,
+      "step": 4870
+    },
+    {
+      "epoch": 3.891108891108891,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.0002,
+      "loss": 0.8942,
+      "step": 4871
+    },
+    {
+      "epoch": 3.8919080919080917,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0002,
+      "loss": 0.9,
+      "step": 4872
+    },
+    {
+      "epoch": 3.8927072927072928,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8963,
+      "step": 4873
+    },
+    {
+      "epoch": 3.8935064935064934,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0002,
+      "loss": 0.8873,
+      "step": 4874
+    },
+    {
+      "epoch": 3.8943056943056944,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0002,
+      "loss": 0.9032,
+      "step": 4875
+    },
+    {
+      "epoch": 3.895104895104895,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9062,
+      "step": 4876
+    },
+    {
+      "epoch": 3.895904095904096,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8937,
+      "step": 4877
+    },
+    {
+      "epoch": 3.8967032967032966,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.0002,
+      "loss": 0.9004,
+      "step": 4878
+    },
+    {
+      "epoch": 3.8975024975024977,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0002,
+      "loss": 0.8969,
+      "step": 4879
+    },
+    {
+      "epoch": 3.8983016983016983,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0002,
+      "loss": 0.8926,
+      "step": 4880
+    },
+    {
+      "epoch": 3.8991008991008993,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9045,
+      "step": 4881
+    },
+    {
+      "epoch": 3.8999000999001,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0002,
+      "loss": 0.907,
+      "step": 4882
+    },
+    {
+      "epoch": 3.9006993006993005,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0002,
+      "loss": 0.8972,
+      "step": 4883
+    },
+    {
+      "epoch": 3.9014985014985015,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.0002,
+      "loss": 0.898,
+      "step": 4884
+    },
+    {
+      "epoch": 3.9022977022977026,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.0002,
+      "loss": 0.9061,
+      "step": 4885
+    },
+    {
+      "epoch": 3.903096903096903,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0002,
+      "loss": 0.8864,
+      "step": 4886
+    },
+    {
+      "epoch": 3.9038961038961038,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0002,
+      "loss": 0.901,
+      "step": 4887
+    },
+    {
+      "epoch": 3.904695304695305,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0002,
+      "loss": 0.8985,
+      "step": 4888
+    },
+    {
+      "epoch": 3.9054945054945054,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8941,
+      "step": 4889
+    },
+    {
+      "epoch": 3.9062937062937064,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9,
+      "step": 4890
+    },
+    {
+      "epoch": 3.907092907092907,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8959,
+      "step": 4891
+    },
+    {
+      "epoch": 3.907892107892108,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.0002,
+      "loss": 0.8914,
+      "step": 4892
+    },
+    {
+      "epoch": 3.9086913086913087,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0002,
+      "loss": 0.8919,
+      "step": 4893
+    },
+    {
+      "epoch": 3.9094905094905092,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8921,
+      "step": 4894
+    },
+    {
+      "epoch": 3.9102897102897103,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0002,
+      "loss": 0.8934,
+      "step": 4895
+    },
+    {
+      "epoch": 3.9110889110889113,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8965,
+      "step": 4896
+    },
+    {
+      "epoch": 3.911888111888112,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8912,
+      "step": 4897
+    },
+    {
+      "epoch": 3.9126873126873125,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0002,
+      "loss": 0.8971,
+      "step": 4898
+    },
+    {
+      "epoch": 3.9134865134865136,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8898,
+      "step": 4899
+    },
+    {
+      "epoch": 3.914285714285714,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9022,
+      "step": 4900
+    },
+    {
+      "epoch": 3.915084915084915,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9048,
+      "step": 4901
+    },
+    {
+      "epoch": 3.915884115884116,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0002,
+      "loss": 0.8976,
+      "step": 4902
+    },
+    {
+      "epoch": 3.916683316683317,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9054,
+      "step": 4903
+    },
+    {
+      "epoch": 3.9174825174825174,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0002,
+      "loss": 0.9258,
+      "step": 4904
+    },
+    {
+      "epoch": 3.918281718281718,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.0002,
+      "loss": 0.89,
+      "step": 4905
+    },
+    {
+      "epoch": 3.919080919080919,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9103,
+      "step": 4906
+    },
+    {
+      "epoch": 3.91988011988012,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8974,
+      "step": 4907
+    },
+    {
+      "epoch": 3.9206793206793207,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0002,
+      "loss": 0.9005,
+      "step": 4908
+    },
+    {
+      "epoch": 3.9214785214785213,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.0002,
+      "loss": 0.8916,
+      "step": 4909
+    },
+    {
+      "epoch": 3.9222777222777223,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8945,
+      "step": 4910
+    },
+    {
+      "epoch": 3.9230769230769234,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9066,
+      "step": 4911
+    },
+    {
+      "epoch": 3.923876123876124,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0002,
+      "loss": 0.898,
+      "step": 4912
+    },
+    {
+      "epoch": 3.9246753246753245,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0002,
+      "loss": 0.894,
+      "step": 4913
+    },
+    {
+      "epoch": 3.9254745254745256,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0002,
+      "loss": 0.8977,
+      "step": 4914
+    },
+    {
+      "epoch": 3.926273726273726,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0002,
+      "loss": 0.9004,
+      "step": 4915
+    },
+    {
+      "epoch": 3.927072927072927,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0002,
+      "loss": 0.8981,
+      "step": 4916
+    },
+    {
+      "epoch": 3.927872127872128,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8922,
+      "step": 4917
+    },
+    {
+      "epoch": 3.928671328671329,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.0002,
+      "loss": 0.8947,
+      "step": 4918
+    },
+    {
+      "epoch": 3.9294705294705294,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8951,
+      "step": 4919
+    },
+    {
+      "epoch": 3.93026973026973,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0002,
+      "loss": 0.904,
+      "step": 4920
+    },
+    {
+      "epoch": 3.931068931068931,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0002,
+      "loss": 0.9005,
+      "step": 4921
+    },
+    {
+      "epoch": 3.931868131868132,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0002,
+      "loss": 0.9008,
+      "step": 4922
+    },
+    {
+      "epoch": 3.9326673326673327,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0002,
+      "loss": 0.893,
+      "step": 4923
+    },
+    {
+      "epoch": 3.9334665334665333,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.0002,
+      "loss": 0.8999,
+      "step": 4924
+    },
+    {
+      "epoch": 3.9342657342657343,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0002,
+      "loss": 0.895,
+      "step": 4925
+    },
+    {
+      "epoch": 3.935064935064935,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0002,
+      "loss": 0.895,
+      "step": 4926
+    },
+    {
+      "epoch": 3.935864135864136,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8936,
+      "step": 4927
+    },
+    {
+      "epoch": 3.9366633366633366,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8936,
+      "step": 4928
+    },
+    {
+      "epoch": 3.9374625374625376,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0002,
+      "loss": 0.9009,
+      "step": 4929
+    },
+    {
+      "epoch": 3.938261738261738,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8915,
+      "step": 4930
+    },
+    {
+      "epoch": 3.939060939060939,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0002,
+      "loss": 0.8968,
+      "step": 4931
+    },
+    {
+      "epoch": 3.93986013986014,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9016,
+      "step": 4932
+    },
+    {
+      "epoch": 3.940659340659341,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0002,
+      "loss": 0.897,
+      "step": 4933
+    },
+    {
+      "epoch": 3.9414585414585415,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8975,
+      "step": 4934
+    },
+    {
+      "epoch": 3.942257742257742,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9006,
+      "step": 4935
+    },
+    {
+      "epoch": 3.943056943056943,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0002,
+      "loss": 0.9004,
+      "step": 4936
+    },
+    {
+      "epoch": 3.9438561438561437,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0002,
+      "loss": 0.8966,
+      "step": 4937
+    },
+    {
+      "epoch": 3.9446553446553447,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8953,
+      "step": 4938
+    },
+    {
+      "epoch": 3.9454545454545453,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.0002,
+      "loss": 0.8913,
+      "step": 4939
+    },
+    {
+      "epoch": 3.9462537462537464,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.0002,
+      "loss": 0.904,
+      "step": 4940
+    },
+    {
+      "epoch": 3.947052947052947,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.0002,
+      "loss": 0.907,
+      "step": 4941
+    },
+    {
+      "epoch": 3.947852147852148,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0002,
+      "loss": 0.8977,
+      "step": 4942
+    },
+    {
+      "epoch": 3.9486513486513486,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0002,
+      "loss": 0.8944,
+      "step": 4943
+    },
+    {
+      "epoch": 3.9494505494505496,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0002,
+      "loss": 0.892,
+      "step": 4944
+    },
+    {
+      "epoch": 3.9502497502497502,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.0002,
+      "loss": 0.886,
+      "step": 4945
+    },
+    {
+      "epoch": 3.951048951048951,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0002,
+      "loss": 0.901,
+      "step": 4946
+    },
+    {
+      "epoch": 3.951848151848152,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0002,
+      "loss": 0.8953,
+      "step": 4947
+    },
+    {
+      "epoch": 3.952647352647353,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8878,
+      "step": 4948
+    },
+    {
+      "epoch": 3.9534465534465535,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0002,
+      "loss": 0.896,
+      "step": 4949
+    },
+    {
+      "epoch": 3.954245754245754,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0002,
+      "loss": 0.8885,
+      "step": 4950
+    },
+    {
+      "epoch": 3.955044955044955,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0002,
+      "loss": 0.8932,
+      "step": 4951
+    },
+    {
+      "epoch": 3.9558441558441557,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.0002,
+      "loss": 0.9065,
+      "step": 4952
+    },
+    {
+      "epoch": 3.9566433566433568,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0002,
+      "loss": 0.8975,
+      "step": 4953
+    },
+    {
+      "epoch": 3.9574425574425574,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0002,
+      "loss": 0.8916,
+      "step": 4954
+    },
+    {
+      "epoch": 3.9582417582417584,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0002,
+      "loss": 0.9046,
+      "step": 4955
+    },
+    {
+      "epoch": 3.959040959040959,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8924,
+      "step": 4956
+    },
+    {
+      "epoch": 3.9598401598401596,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8998,
+      "step": 4957
+    },
+    {
+      "epoch": 3.9606393606393606,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0002,
+      "loss": 0.8996,
+      "step": 4958
+    },
+    {
+      "epoch": 3.9614385614385617,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0002,
+      "loss": 0.8931,
+      "step": 4959
+    },
+    {
+      "epoch": 3.9622377622377623,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.0002,
+      "loss": 0.9001,
+      "step": 4960
+    },
+    {
+      "epoch": 3.963036963036963,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0002,
+      "loss": 0.8978,
+      "step": 4961
+    },
+    {
+      "epoch": 3.963836163836164,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8936,
+      "step": 4962
+    },
+    {
+      "epoch": 3.9646353646353645,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0002,
+      "loss": 0.8996,
+      "step": 4963
+    },
+    {
+      "epoch": 3.9654345654345655,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.0002,
+      "loss": 0.8937,
+      "step": 4964
+    },
+    {
+      "epoch": 3.966233766233766,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.0002,
+      "loss": 0.8897,
+      "step": 4965
+    },
+    {
+      "epoch": 3.967032967032967,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0002,
+      "loss": 0.8984,
+      "step": 4966
+    },
+    {
+      "epoch": 3.9678321678321677,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.0002,
+      "loss": 0.9035,
+      "step": 4967
+    },
+    {
+      "epoch": 3.968631368631369,
+      "grad_norm": 0.375,
+      "learning_rate": 0.0002,
+      "loss": 0.8944,
+      "step": 4968
+    },
+    {
+      "epoch": 3.9694305694305694,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0002,
+      "loss": 0.9146,
+      "step": 4969
+    },
+    {
+      "epoch": 3.9702297702297704,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8881,
+      "step": 4970
+    },
+    {
+      "epoch": 3.971028971028971,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8888,
+      "step": 4971
+    },
+    {
+      "epoch": 3.9718281718281716,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9053,
+      "step": 4972
+    },
+    {
+      "epoch": 3.9726273726273726,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0002,
+      "loss": 0.8957,
+      "step": 4973
+    },
+    {
+      "epoch": 3.9734265734265737,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0002,
+      "loss": 0.9014,
+      "step": 4974
+    },
+    {
+      "epoch": 3.9742257742257743,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0002,
+      "loss": 0.896,
+      "step": 4975
+    },
+    {
+      "epoch": 3.975024975024975,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.0002,
+      "loss": 0.9007,
+      "step": 4976
+    },
+    {
+      "epoch": 3.975824175824176,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.0002,
+      "loss": 0.901,
+      "step": 4977
+    },
+    {
+      "epoch": 3.9766233766233765,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0002,
+      "loss": 0.8946,
+      "step": 4978
+    },
+    {
+      "epoch": 3.9774225774225775,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9068,
+      "step": 4979
+    },
+    {
+      "epoch": 3.978221778221778,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0002,
+      "loss": 0.8982,
+      "step": 4980
+    },
+    {
+      "epoch": 3.979020979020979,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0002,
+      "loss": 0.9022,
+      "step": 4981
+    },
+    {
+      "epoch": 3.9798201798201798,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0002,
+      "loss": 0.8983,
+      "step": 4982
+    },
+    {
+      "epoch": 3.9806193806193804,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0002,
+      "loss": 0.9015,
+      "step": 4983
+    },
+    {
+      "epoch": 3.9814185814185814,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0002,
+      "loss": 0.893,
+      "step": 4984
+    },
+    {
+      "epoch": 3.9822177822177824,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8978,
+      "step": 4985
+    },
+    {
+      "epoch": 3.983016983016983,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0002,
+      "loss": 0.902,
+      "step": 4986
+    },
+    {
+      "epoch": 3.9838161838161836,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9028,
+      "step": 4987
+    },
+    {
+      "epoch": 3.9846153846153847,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0002,
+      "loss": 0.8969,
+      "step": 4988
+    },
+    {
+      "epoch": 3.9854145854145853,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0002,
+      "loss": 0.9025,
+      "step": 4989
+    },
+    {
+      "epoch": 3.9862137862137863,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0002,
+      "loss": 0.8956,
+      "step": 4990
+    },
+    {
+      "epoch": 3.987012987012987,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8971,
+      "step": 4991
+    },
+    {
+      "epoch": 3.987812187812188,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0002,
+      "loss": 0.8973,
+      "step": 4992
+    },
+    {
+      "epoch": 3.9886113886113885,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.0002,
+      "loss": 0.9031,
+      "step": 4993
+    },
+    {
+      "epoch": 3.9894105894105896,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0002,
+      "loss": 0.8985,
+      "step": 4994
+    },
+    {
+      "epoch": 3.99020979020979,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0002,
+      "loss": 0.892,
+      "step": 4995
+    },
+    {
+      "epoch": 3.991008991008991,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.0002,
+      "loss": 0.8957,
+      "step": 4996
+    },
+    {
+      "epoch": 3.991808191808192,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9105,
+      "step": 4997
+    },
+    {
+      "epoch": 3.9926073926073924,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.0002,
+      "loss": 0.8928,
+      "step": 4998
+    },
+    {
+      "epoch": 3.9934065934065934,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8897,
+      "step": 4999
+    },
+    {
+      "epoch": 3.9942057942057945,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0002,
+      "loss": 0.9037,
+      "step": 5000
+    },
+    {
+      "epoch": 3.995004995004995,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0002,
+      "loss": 0.89,
+      "step": 5001
+    },
+    {
+      "epoch": 3.9958041958041957,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0002,
+      "loss": 0.8954,
+      "step": 5002
+    },
+    {
+      "epoch": 3.9966033966033967,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9044,
+      "step": 5003
+    },
+    {
+      "epoch": 3.9974025974025973,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.0002,
+      "loss": 0.9,
+      "step": 5004
+    },
+    {
+      "epoch": 3.9982017982017983,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8937,
+      "step": 5005
+    },
+    {
+      "epoch": 3.999000999000999,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0002,
+      "loss": 0.8967,
+      "step": 5006
+    },
+    {
+      "epoch": 3.9998001998002,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.0002,
+      "loss": 0.902,
+      "step": 5007
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0002,
+      "loss": 0.2225,
+      "step": 5008
+    },
+    {
+      "epoch": 4.000799200799201,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8947,
+      "step": 5009
+    },
+    {
+      "epoch": 4.001598401598401,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.0002,
+      "loss": 0.895,
+      "step": 5010
+    },
+    {
+      "epoch": 4.002397602397602,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.0002,
+      "loss": 0.8947,
+      "step": 5011
+    },
+    {
+      "epoch": 4.003196803196803,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0002,
+      "loss": 0.9005,
+      "step": 5012
+    },
+    {
+      "epoch": 4.003996003996004,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0002,
+      "loss": 0.8964,
+      "step": 5013
+    },
+    {
+      "epoch": 4.0047952047952045,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.0002,
+      "loss": 0.8976,
+      "step": 5014
+    },
+    {
+      "epoch": 4.0055944055944055,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0002,
+      "loss": 0.8916,
+      "step": 5015
+    },
+    {
+      "epoch": 4.0063936063936065,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0002,
+      "loss": 0.8947,
+      "step": 5016
+    },
+    {
+      "epoch": 4.007192807192808,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.0002,
+      "loss": 0.898,
+      "step": 5017
+    },
+    {
+      "epoch": 4.007992007992008,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0002,
+      "loss": 0.9013,
+      "step": 5018
+    },
+    {
+      "epoch": 4.008791208791209,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0002,
+      "loss": 0.8964,
+      "step": 5019
+    },
+    {
+      "epoch": 4.00959040959041,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0002,
+      "loss": 0.8898,
+      "step": 5020
+    },
+    {
+      "epoch": 4.01038961038961,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0002,
+      "loss": 0.9087,
+      "step": 5021
+    },
+    {
+      "epoch": 4.011188811188811,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0002,
+      "loss": 0.9005,
+      "step": 5022
+    },
+    {
+      "epoch": 4.011988011988012,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8925,
+      "step": 5023
+    },
+    {
+      "epoch": 4.012787212787213,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0002,
+      "loss": 0.9006,
+      "step": 5024
+    },
+    {
+      "epoch": 4.013586413586413,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0002,
+      "loss": 0.907,
+      "step": 5025
+    },
+    {
+      "epoch": 4.014385614385614,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9004,
+      "step": 5026
+    },
+    {
+      "epoch": 4.015184815184815,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.0002,
+      "loss": 0.9052,
+      "step": 5027
+    },
+    {
+      "epoch": 4.015984015984016,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9008,
+      "step": 5028
+    },
+    {
+      "epoch": 4.0167832167832165,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0002,
+      "loss": 0.8939,
+      "step": 5029
+    },
+    {
+      "epoch": 4.0175824175824175,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8976,
+      "step": 5030
+    },
+    {
+      "epoch": 4.018381618381619,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0002,
+      "loss": 0.9039,
+      "step": 5031
+    },
+    {
+      "epoch": 4.01918081918082,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8995,
+      "step": 5032
+    },
+    {
+      "epoch": 4.01998001998002,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0002,
+      "loss": 0.8967,
+      "step": 5033
+    },
+    {
+      "epoch": 4.020779220779221,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8937,
+      "step": 5034
+    },
+    {
+      "epoch": 4.021578421578422,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8959,
+      "step": 5035
+    },
+    {
+      "epoch": 4.022377622377622,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0002,
+      "loss": 0.8917,
+      "step": 5036
+    },
+    {
+      "epoch": 4.023176823176823,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0002,
+      "loss": 0.8978,
+      "step": 5037
+    },
+    {
+      "epoch": 4.023976023976024,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8873,
+      "step": 5038
+    },
+    {
+      "epoch": 4.024775224775225,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0002,
+      "loss": 0.8938,
+      "step": 5039
+    },
+    {
+      "epoch": 4.025574425574425,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8945,
+      "step": 5040
+    },
+    {
+      "epoch": 4.026373626373626,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8964,
+      "step": 5041
+    },
+    {
+      "epoch": 4.027172827172827,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8994,
+      "step": 5042
+    },
+    {
+      "epoch": 4.027972027972028,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0002,
+      "loss": 0.8957,
+      "step": 5043
+    },
+    {
+      "epoch": 4.0287712287712285,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9055,
+      "step": 5044
+    },
+    {
+      "epoch": 4.0295704295704295,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0002,
+      "loss": 0.9018,
+      "step": 5045
+    },
+    {
+      "epoch": 4.030369630369631,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8939,
+      "step": 5046
+    },
+    {
+      "epoch": 4.031168831168831,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0002,
+      "loss": 0.8928,
+      "step": 5047
+    },
+    {
+      "epoch": 4.031968031968032,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9351,
+      "step": 5048
+    },
+    {
+      "epoch": 4.032767232767233,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9047,
+      "step": 5049
+    },
+    {
+      "epoch": 4.033566433566434,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9009,
+      "step": 5050
+    },
+    {
+      "epoch": 4.034365634365634,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8988,
+      "step": 5051
+    },
+    {
+      "epoch": 4.035164835164835,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0002,
+      "loss": 0.8978,
+      "step": 5052
+    },
+    {
+      "epoch": 4.035964035964036,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8933,
+      "step": 5053
+    },
+    {
+      "epoch": 4.036763236763237,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8979,
+      "step": 5054
+    },
+    {
+      "epoch": 4.037562437562437,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0002,
+      "loss": 0.8984,
+      "step": 5055
+    },
+    {
+      "epoch": 4.038361638361638,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0002,
+      "loss": 0.8868,
+      "step": 5056
+    },
+    {
+      "epoch": 4.039160839160839,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0002,
+      "loss": 0.8991,
+      "step": 5057
+    },
+    {
+      "epoch": 4.03996003996004,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0002,
+      "loss": 0.893,
+      "step": 5058
+    },
+    {
+      "epoch": 4.0407592407592405,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0002,
+      "loss": 0.891,
+      "step": 5059
+    },
+    {
+      "epoch": 4.041558441558442,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8976,
+      "step": 5060
+    },
+    {
+      "epoch": 4.042357642357643,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9014,
+      "step": 5061
+    },
+    {
+      "epoch": 4.043156843156843,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8898,
+      "step": 5062
+    },
+    {
+      "epoch": 4.043956043956044,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0002,
+      "loss": 0.8958,
+      "step": 5063
+    },
+    {
+      "epoch": 4.044755244755245,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.0002,
+      "loss": 0.8935,
+      "step": 5064
+    },
+    {
+      "epoch": 4.045554445554446,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0002,
+      "loss": 0.8937,
+      "step": 5065
+    },
+    {
+      "epoch": 4.046353646353646,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8947,
+      "step": 5066
+    },
+    {
+      "epoch": 4.047152847152847,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0002,
+      "loss": 0.9014,
+      "step": 5067
+    },
+    {
+      "epoch": 4.047952047952048,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.0002,
+      "loss": 0.9042,
+      "step": 5068
+    },
+    {
+      "epoch": 4.048751248751249,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0002,
+      "loss": 0.8965,
+      "step": 5069
+    },
+    {
+      "epoch": 4.049550449550449,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9029,
+      "step": 5070
+    },
+    {
+      "epoch": 4.05034965034965,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0002,
+      "loss": 0.8905,
+      "step": 5071
+    },
+    {
+      "epoch": 4.051148851148851,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.0002,
+      "loss": 0.8975,
+      "step": 5072
+    },
+    {
+      "epoch": 4.0519480519480515,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0002,
+      "loss": 0.899,
+      "step": 5073
+    },
+    {
+      "epoch": 4.052747252747253,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.0002,
+      "loss": 0.9141,
+      "step": 5074
+    },
+    {
+      "epoch": 4.053546453546454,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0002,
+      "loss": 0.8984,
+      "step": 5075
+    },
+    {
+      "epoch": 4.054345654345655,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0002,
+      "loss": 0.898,
+      "step": 5076
+    },
+    {
+      "epoch": 4.055144855144855,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0002,
+      "loss": 0.8959,
+      "step": 5077
+    },
+    {
+      "epoch": 4.055944055944056,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8939,
+      "step": 5078
+    },
+    {
+      "epoch": 4.056743256743257,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0002,
+      "loss": 0.8813,
+      "step": 5079
+    },
+    {
+      "epoch": 4.057542457542458,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0002,
+      "loss": 0.9022,
+      "step": 5080
+    },
+    {
+      "epoch": 4.058341658341658,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0002,
+      "loss": 0.8987,
+      "step": 5081
+    },
+    {
+      "epoch": 4.059140859140859,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8923,
+      "step": 5082
+    },
+    {
+      "epoch": 4.05994005994006,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8878,
+      "step": 5083
+    },
+    {
+      "epoch": 4.060739260739261,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0002,
+      "loss": 0.8989,
+      "step": 5084
+    },
+    {
+      "epoch": 4.061538461538461,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.0002,
+      "loss": 0.8894,
+      "step": 5085
+    },
+    {
+      "epoch": 4.062337662337662,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9012,
+      "step": 5086
+    },
+    {
+      "epoch": 4.063136863136863,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0002,
+      "loss": 0.8922,
+      "step": 5087
+    },
+    {
+      "epoch": 4.0639360639360635,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9036,
+      "step": 5088
+    },
+    {
+      "epoch": 4.064735264735265,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0002,
+      "loss": 0.9019,
+      "step": 5089
+    },
+    {
+      "epoch": 4.065534465534466,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8903,
+      "step": 5090
+    },
+    {
+      "epoch": 4.066333666333667,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0002,
+      "loss": 0.8945,
+      "step": 5091
+    },
+    {
+      "epoch": 4.067132867132867,
+      "grad_norm": 0.375,
+      "learning_rate": 0.0002,
+      "loss": 0.8937,
+      "step": 5092
+    },
+    {
+      "epoch": 4.067932067932068,
+      "grad_norm": 0.91796875,
+      "learning_rate": 0.0002,
+      "loss": 0.9019,
+      "step": 5093
+    },
+    {
+      "epoch": 4.068731268731269,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.0002,
+      "loss": 0.8953,
+      "step": 5094
+    },
+    {
+      "epoch": 4.06953046953047,
+      "grad_norm": 0.921875,
+      "learning_rate": 0.0002,
+      "loss": 0.8903,
+      "step": 5095
+    },
+    {
+      "epoch": 4.07032967032967,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0002,
+      "loss": 0.9081,
+      "step": 5096
+    },
+    {
+      "epoch": 4.071128871128871,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0002,
+      "loss": 0.9016,
+      "step": 5097
+    },
+    {
+      "epoch": 4.071928071928072,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0002,
+      "loss": 0.8982,
+      "step": 5098
+    },
+    {
+      "epoch": 4.072727272727272,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8911,
+      "step": 5099
+    },
+    {
+      "epoch": 4.073526473526473,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.0002,
+      "loss": 0.9017,
+      "step": 5100
+    },
+    {
+      "epoch": 4.074325674325674,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0002,
+      "loss": 0.8898,
+      "step": 5101
+    },
+    {
+      "epoch": 4.075124875124875,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0002,
+      "loss": 0.9001,
+      "step": 5102
+    },
+    {
+      "epoch": 4.075924075924076,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8906,
+      "step": 5103
+    },
+    {
+      "epoch": 4.076723276723277,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9082,
+      "step": 5104
+    },
+    {
+      "epoch": 4.077522477522478,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.0002,
+      "loss": 0.8982,
+      "step": 5105
+    },
+    {
+      "epoch": 4.078321678321679,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.0002,
+      "loss": 0.8952,
+      "step": 5106
+    },
+    {
+      "epoch": 4.079120879120879,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0002,
+      "loss": 0.8911,
+      "step": 5107
+    },
+    {
+      "epoch": 4.07992007992008,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0002,
+      "loss": 0.8888,
+      "step": 5108
+    },
+    {
+      "epoch": 4.080719280719281,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.0002,
+      "loss": 0.9018,
+      "step": 5109
+    },
+    {
+      "epoch": 4.081518481518482,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8906,
+      "step": 5110
+    },
+    {
+      "epoch": 4.082317682317682,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8939,
+      "step": 5111
+    },
+    {
+      "epoch": 4.083116883116883,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.0002,
+      "loss": 0.891,
+      "step": 5112
+    },
+    {
+      "epoch": 4.083916083916084,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0002,
+      "loss": 0.8957,
+      "step": 5113
+    },
+    {
+      "epoch": 4.084715284715284,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9036,
+      "step": 5114
+    },
+    {
+      "epoch": 4.085514485514485,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0002,
+      "loss": 0.9034,
+      "step": 5115
+    },
+    {
+      "epoch": 4.086313686313686,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0002,
+      "loss": 0.888,
+      "step": 5116
+    },
+    {
+      "epoch": 4.0871128871128874,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0002,
+      "loss": 0.892,
+      "step": 5117
+    },
+    {
+      "epoch": 4.087912087912088,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.0002,
+      "loss": 0.8961,
+      "step": 5118
+    },
+    {
+      "epoch": 4.088711288711289,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8959,
+      "step": 5119
+    },
+    {
+      "epoch": 4.08951048951049,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8922,
+      "step": 5120
+    },
+    {
+      "epoch": 4.090309690309691,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8991,
+      "step": 5121
+    },
+    {
+      "epoch": 4.091108891108891,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.0002,
+      "loss": 0.893,
+      "step": 5122
+    },
+    {
+      "epoch": 4.091908091908092,
+      "grad_norm": 0.98046875,
+      "learning_rate": 0.0002,
+      "loss": 0.9368,
+      "step": 5123
+    },
+    {
+      "epoch": 4.092707292707293,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0002,
+      "loss": 0.902,
+      "step": 5124
+    },
+    {
+      "epoch": 4.093506493506493,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0002,
+      "loss": 0.9007,
+      "step": 5125
+    },
+    {
+      "epoch": 4.094305694305694,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9024,
+      "step": 5126
+    },
+    {
+      "epoch": 4.095104895104895,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0002,
+      "loss": 0.8933,
+      "step": 5127
+    },
+    {
+      "epoch": 4.095904095904096,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.0002,
+      "loss": 0.8913,
+      "step": 5128
+    },
+    {
+      "epoch": 4.096703296703296,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8894,
+      "step": 5129
+    },
+    {
+      "epoch": 4.097502497502497,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0002,
+      "loss": 0.8919,
+      "step": 5130
+    },
+    {
+      "epoch": 4.098301698301698,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0002,
+      "loss": 0.89,
+      "step": 5131
+    },
+    {
+      "epoch": 4.0991008991008995,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9025,
+      "step": 5132
+    },
+    {
+      "epoch": 4.0999000999001,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8941,
+      "step": 5133
+    },
+    {
+      "epoch": 4.100699300699301,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0002,
+      "loss": 0.8921,
+      "step": 5134
+    },
+    {
+      "epoch": 4.101498501498502,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.0002,
+      "loss": 0.893,
+      "step": 5135
+    },
+    {
+      "epoch": 4.102297702297703,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8877,
+      "step": 5136
+    },
+    {
+      "epoch": 4.103096903096903,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.0002,
+      "loss": 0.8958,
+      "step": 5137
+    },
+    {
+      "epoch": 4.103896103896104,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0002,
+      "loss": 0.8967,
+      "step": 5138
+    },
+    {
+      "epoch": 4.104695304695305,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0002,
+      "loss": 0.897,
+      "step": 5139
+    },
+    {
+      "epoch": 4.105494505494505,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.0002,
+      "loss": 0.8917,
+      "step": 5140
+    },
+    {
+      "epoch": 4.106293706293706,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0002,
+      "loss": 0.8877,
+      "step": 5141
+    },
+    {
+      "epoch": 4.107092907092907,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.0002,
+      "loss": 0.897,
+      "step": 5142
+    },
+    {
+      "epoch": 4.107892107892108,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0002,
+      "loss": 0.9008,
+      "step": 5143
+    },
+    {
+      "epoch": 4.108691308691308,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9009,
+      "step": 5144
+    },
+    {
+      "epoch": 4.109490509490509,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.0002,
+      "loss": 0.8974,
+      "step": 5145
+    },
+    {
+      "epoch": 4.1102897102897105,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.0002,
+      "loss": 0.8976,
+      "step": 5146
+    },
+    {
+      "epoch": 4.1110889110889115,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8954,
+      "step": 5147
+    },
+    {
+      "epoch": 4.111888111888112,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0002,
+      "loss": 0.8922,
+      "step": 5148
+    },
+    {
+      "epoch": 4.112687312687313,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8959,
+      "step": 5149
+    },
+    {
+      "epoch": 4.113486513486514,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8892,
+      "step": 5150
+    },
+    {
+      "epoch": 4.114285714285714,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0002,
+      "loss": 0.901,
+      "step": 5151
+    },
+    {
+      "epoch": 4.115084915084915,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0002,
+      "loss": 0.8928,
+      "step": 5152
+    },
+    {
+      "epoch": 4.115884115884116,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8904,
+      "step": 5153
+    },
+    {
+      "epoch": 4.116683316683317,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8868,
+      "step": 5154
+    },
+    {
+      "epoch": 4.117482517482517,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0002,
+      "loss": 0.8893,
+      "step": 5155
+    },
+    {
+      "epoch": 4.118281718281718,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8987,
+      "step": 5156
+    },
+    {
+      "epoch": 4.119080919080919,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8942,
+      "step": 5157
+    },
+    {
+      "epoch": 4.11988011988012,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0002,
+      "loss": 0.902,
+      "step": 5158
+    },
+    {
+      "epoch": 4.12067932067932,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0002,
+      "loss": 0.8902,
+      "step": 5159
+    },
+    {
+      "epoch": 4.1214785214785215,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0002,
+      "loss": 0.8947,
+      "step": 5160
+    },
+    {
+      "epoch": 4.1222777222777225,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0002,
+      "loss": 0.901,
+      "step": 5161
+    },
+    {
+      "epoch": 4.123076923076923,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8984,
+      "step": 5162
+    },
+    {
+      "epoch": 4.123876123876124,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.0002,
+      "loss": 0.9009,
+      "step": 5163
+    },
+    {
+      "epoch": 4.124675324675325,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8868,
+      "step": 5164
+    },
+    {
+      "epoch": 4.125474525474526,
+      "grad_norm": 0.375,
+      "learning_rate": 0.0002,
+      "loss": 0.889,
+      "step": 5165
+    },
+    {
+      "epoch": 4.126273726273726,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9014,
+      "step": 5166
+    },
+    {
+      "epoch": 4.127072927072927,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8993,
+      "step": 5167
+    },
+    {
+      "epoch": 4.127872127872128,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.0002,
+      "loss": 0.8896,
+      "step": 5168
+    },
+    {
+      "epoch": 4.128671328671329,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9015,
+      "step": 5169
+    },
+    {
+      "epoch": 4.129470529470529,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0002,
+      "loss": 0.8967,
+      "step": 5170
+    },
+    {
+      "epoch": 4.13026973026973,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.0002,
+      "loss": 0.9013,
+      "step": 5171
+    },
+    {
+      "epoch": 4.131068931068931,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0002,
+      "loss": 0.9035,
+      "step": 5172
+    },
+    {
+      "epoch": 4.131868131868132,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0002,
+      "loss": 0.8918,
+      "step": 5173
+    },
+    {
+      "epoch": 4.132667332667332,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.0002,
+      "loss": 0.9135,
+      "step": 5174
+    },
+    {
+      "epoch": 4.1334665334665335,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0002,
+      "loss": 0.8924,
+      "step": 5175
+    },
+    {
+      "epoch": 4.1342657342657345,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0002,
+      "loss": 0.8977,
+      "step": 5176
+    },
+    {
+      "epoch": 4.135064935064935,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0002,
+      "loss": 0.8986,
+      "step": 5177
+    },
+    {
+      "epoch": 4.135864135864136,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0002,
+      "loss": 0.9082,
+      "step": 5178
+    },
+    {
+      "epoch": 4.136663336663337,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8965,
+      "step": 5179
+    },
+    {
+      "epoch": 4.137462537462538,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0002,
+      "loss": 0.9011,
+      "step": 5180
+    },
+    {
+      "epoch": 4.138261738261738,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0002,
+      "loss": 0.8822,
+      "step": 5181
+    },
+    {
+      "epoch": 4.139060939060939,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.0002,
+      "loss": 0.8922,
+      "step": 5182
+    },
+    {
+      "epoch": 4.13986013986014,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0002,
+      "loss": 0.8989,
+      "step": 5183
+    },
+    {
+      "epoch": 4.140659340659341,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8927,
+      "step": 5184
+    },
+    {
+      "epoch": 4.141458541458541,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0002,
+      "loss": 0.9007,
+      "step": 5185
+    },
+    {
+      "epoch": 4.142257742257742,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9024,
+      "step": 5186
+    },
+    {
+      "epoch": 4.143056943056943,
+      "grad_norm": 0.375,
+      "learning_rate": 0.0002,
+      "loss": 0.8953,
+      "step": 5187
+    },
+    {
+      "epoch": 4.143856143856144,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0002,
+      "loss": 0.9006,
+      "step": 5188
+    },
+    {
+      "epoch": 4.1446553446553445,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0002,
+      "loss": 0.8957,
+      "step": 5189
+    },
+    {
+      "epoch": 4.1454545454545455,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0002,
+      "loss": 0.904,
+      "step": 5190
+    },
+    {
+      "epoch": 4.1462537462537465,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8988,
+      "step": 5191
+    },
+    {
+      "epoch": 4.147052947052947,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8886,
+      "step": 5192
+    },
+    {
+      "epoch": 4.147852147852148,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8998,
+      "step": 5193
+    },
+    {
+      "epoch": 4.148651348651349,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0002,
+      "loss": 0.8954,
+      "step": 5194
+    },
+    {
+      "epoch": 4.14945054945055,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0002,
+      "loss": 0.8934,
+      "step": 5195
+    },
+    {
+      "epoch": 4.15024975024975,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8922,
+      "step": 5196
+    },
+    {
+      "epoch": 4.151048951048951,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0002,
+      "loss": 0.9025,
+      "step": 5197
+    },
+    {
+      "epoch": 4.151848151848152,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0002,
+      "loss": 0.899,
+      "step": 5198
+    },
+    {
+      "epoch": 4.152647352647353,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8862,
+      "step": 5199
+    },
+    {
+      "epoch": 4.153446553446553,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.0002,
+      "loss": 0.9015,
+      "step": 5200
+    },
+    {
+      "epoch": 4.154245754245754,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0002,
+      "loss": 0.8954,
+      "step": 5201
+    },
+    {
+      "epoch": 4.155044955044955,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0002,
+      "loss": 0.8982,
+      "step": 5202
+    },
+    {
+      "epoch": 4.1558441558441555,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0002,
+      "loss": 0.9037,
+      "step": 5203
+    },
+    {
+      "epoch": 4.1566433566433565,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8979,
+      "step": 5204
+    },
+    {
+      "epoch": 4.1574425574425575,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8949,
+      "step": 5205
+    },
+    {
+      "epoch": 4.158241758241759,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0002,
+      "loss": 0.8934,
+      "step": 5206
+    },
+    {
+      "epoch": 4.159040959040959,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0002,
+      "loss": 0.8921,
+      "step": 5207
+    },
+    {
+      "epoch": 4.15984015984016,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0002,
+      "loss": 0.894,
+      "step": 5208
+    },
+    {
+      "epoch": 4.160639360639361,
+      "grad_norm": 0.375,
+      "learning_rate": 0.0002,
+      "loss": 0.8963,
+      "step": 5209
+    },
+    {
+      "epoch": 4.161438561438562,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0002,
+      "loss": 0.8932,
+      "step": 5210
+    },
+    {
+      "epoch": 4.162237762237762,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8907,
+      "step": 5211
+    },
+    {
+      "epoch": 4.163036963036963,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0002,
+      "loss": 0.9026,
+      "step": 5212
+    },
+    {
+      "epoch": 4.163836163836164,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0002,
+      "loss": 0.8919,
+      "step": 5213
+    },
+    {
+      "epoch": 4.164635364635364,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8952,
+      "step": 5214
+    },
+    {
+      "epoch": 4.165434565434565,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8959,
+      "step": 5215
+    },
+    {
+      "epoch": 4.166233766233766,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0002,
+      "loss": 0.8854,
+      "step": 5216
+    },
+    {
+      "epoch": 4.167032967032967,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9034,
+      "step": 5217
+    },
+    {
+      "epoch": 4.1678321678321675,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.0002,
+      "loss": 0.9017,
+      "step": 5218
+    },
+    {
+      "epoch": 4.1686313686313685,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0002,
+      "loss": 0.895,
+      "step": 5219
+    },
+    {
+      "epoch": 4.1694305694305696,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.0002,
+      "loss": 0.8907,
+      "step": 5220
+    },
+    {
+      "epoch": 4.170229770229771,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8986,
+      "step": 5221
+    },
+    {
+      "epoch": 4.171028971028971,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0002,
+      "loss": 0.8981,
+      "step": 5222
+    },
+    {
+      "epoch": 4.171828171828172,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.0002,
+      "loss": 0.8928,
+      "step": 5223
+    },
+    {
+      "epoch": 4.172627372627373,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8928,
+      "step": 5224
+    },
+    {
+      "epoch": 4.173426573426573,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0002,
+      "loss": 0.9018,
+      "step": 5225
+    },
+    {
+      "epoch": 4.174225774225774,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8991,
+      "step": 5226
+    },
+    {
+      "epoch": 4.175024975024975,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0002,
+      "loss": 0.8937,
+      "step": 5227
+    },
+    {
+      "epoch": 4.175824175824176,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8875,
+      "step": 5228
+    },
+    {
+      "epoch": 4.176623376623376,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.0002,
+      "loss": 0.9052,
+      "step": 5229
+    },
+    {
+      "epoch": 4.177422577422577,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0002,
+      "loss": 0.896,
+      "step": 5230
+    },
+    {
+      "epoch": 4.178221778221778,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0002,
+      "loss": 0.8915,
+      "step": 5231
+    },
+    {
+      "epoch": 4.179020979020979,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0002,
+      "loss": 0.885,
+      "step": 5232
+    },
+    {
+      "epoch": 4.1798201798201795,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0002,
+      "loss": 0.9,
+      "step": 5233
+    },
+    {
+      "epoch": 4.1806193806193805,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0002,
+      "loss": 0.9037,
+      "step": 5234
+    },
+    {
+      "epoch": 4.181418581418582,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8929,
+      "step": 5235
+    },
+    {
+      "epoch": 4.182217782217783,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0002,
+      "loss": 0.8998,
+      "step": 5236
+    },
+    {
+      "epoch": 4.183016983016983,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8983,
+      "step": 5237
+    },
+    {
+      "epoch": 4.183816183816184,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8974,
+      "step": 5238
+    },
+    {
+      "epoch": 4.184615384615385,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0002,
+      "loss": 0.888,
+      "step": 5239
+    },
+    {
+      "epoch": 4.185414585414585,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8954,
+      "step": 5240
+    },
+    {
+      "epoch": 4.186213786213786,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0002,
+      "loss": 0.8901,
+      "step": 5241
+    },
+    {
+      "epoch": 4.187012987012987,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8956,
+      "step": 5242
+    },
+    {
+      "epoch": 4.187812187812188,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002,
+      "loss": 0.894,
+      "step": 5243
+    },
+    {
+      "epoch": 4.188611388611388,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8929,
+      "step": 5244
+    },
+    {
+      "epoch": 4.189410589410589,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0002,
+      "loss": 0.9054,
+      "step": 5245
+    },
+    {
+      "epoch": 4.19020979020979,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0002,
+      "loss": 0.8932,
+      "step": 5246
+    },
+    {
+      "epoch": 4.191008991008991,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9019,
+      "step": 5247
+    },
+    {
+      "epoch": 4.1918081918081915,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0002,
+      "loss": 0.8973,
+      "step": 5248
+    },
+    {
+      "epoch": 4.192607392607393,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8995,
+      "step": 5249
+    },
+    {
+      "epoch": 4.193406593406594,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8916,
+      "step": 5250
+    },
+    {
+      "epoch": 4.194205794205795,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9007,
+      "step": 5251
+    },
+    {
+      "epoch": 4.195004995004995,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0002,
+      "loss": 0.8898,
+      "step": 5252
+    },
+    {
+      "epoch": 4.195804195804196,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0002,
+      "loss": 0.9013,
+      "step": 5253
+    },
+    {
+      "epoch": 4.196603396603397,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0002,
+      "loss": 0.8869,
+      "step": 5254
+    },
+    {
+      "epoch": 4.197402597402597,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9055,
+      "step": 5255
+    },
+    {
+      "epoch": 4.198201798201798,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0002,
+      "loss": 0.9002,
+      "step": 5256
+    },
+    {
+      "epoch": 4.199000999000999,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0002,
+      "loss": 0.8884,
+      "step": 5257
+    },
+    {
+      "epoch": 4.1998001998002,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0002,
+      "loss": 0.894,
+      "step": 5258
+    },
+    {
+      "epoch": 4.2005994005994,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0002,
+      "loss": 0.8919,
+      "step": 5259
+    },
+    {
+      "epoch": 4.201398601398601,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8934,
+      "step": 5260
+    },
+    {
+      "epoch": 4.202197802197802,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9115,
+      "step": 5261
+    },
+    {
+      "epoch": 4.202997002997003,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8965,
+      "step": 5262
+    },
+    {
+      "epoch": 4.203796203796204,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.0002,
+      "loss": 0.8938,
+      "step": 5263
+    },
+    {
+      "epoch": 4.204595404595405,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8993,
+      "step": 5264
+    },
+    {
+      "epoch": 4.205394605394606,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0002,
+      "loss": 0.8976,
+      "step": 5265
+    },
+    {
+      "epoch": 4.206193806193806,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0002,
+      "loss": 0.8907,
+      "step": 5266
+    },
+    {
+      "epoch": 4.206993006993007,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8917,
+      "step": 5267
+    },
+    {
+      "epoch": 4.207792207792208,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8919,
+      "step": 5268
+    },
+    {
+      "epoch": 4.208591408591409,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8973,
+      "step": 5269
+    },
+    {
+      "epoch": 4.209390609390609,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.0002,
+      "loss": 0.8917,
+      "step": 5270
+    },
+    {
+      "epoch": 4.21018981018981,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0002,
+      "loss": 0.9007,
+      "step": 5271
+    },
+    {
+      "epoch": 4.210989010989011,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0002,
+      "loss": 0.8945,
+      "step": 5272
+    },
+    {
+      "epoch": 4.211788211788212,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0002,
+      "loss": 0.8975,
+      "step": 5273
+    },
+    {
+      "epoch": 4.212587412587412,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0002,
+      "loss": 0.8925,
+      "step": 5274
+    },
+    {
+      "epoch": 4.213386613386613,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0002,
+      "loss": 0.8964,
+      "step": 5275
+    },
+    {
+      "epoch": 4.214185814185814,
+      "grad_norm": 0.375,
+      "learning_rate": 0.0002,
+      "loss": 0.8966,
+      "step": 5276
+    },
+    {
+      "epoch": 4.2149850149850145,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0002,
+      "loss": 0.8907,
+      "step": 5277
+    },
+    {
+      "epoch": 4.215784215784216,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.0002,
+      "loss": 0.8934,
+      "step": 5278
+    },
+    {
+      "epoch": 4.216583416583417,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8927,
+      "step": 5279
+    },
+    {
+      "epoch": 4.217382617382618,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.0002,
+      "loss": 0.891,
+      "step": 5280
+    },
+    {
+      "epoch": 4.218181818181818,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0002,
+      "loss": 0.8878,
+      "step": 5281
+    },
+    {
+      "epoch": 4.218981018981019,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0002,
+      "loss": 0.9055,
+      "step": 5282
+    },
+    {
+      "epoch": 4.21978021978022,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0002,
+      "loss": 0.8994,
+      "step": 5283
+    },
+    {
+      "epoch": 4.220579420579421,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0002,
+      "loss": 0.889,
+      "step": 5284
+    },
+    {
+      "epoch": 4.221378621378621,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.0002,
+      "loss": 0.8944,
+      "step": 5285
+    },
+    {
+      "epoch": 4.222177822177822,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8967,
+      "step": 5286
+    },
+    {
+      "epoch": 4.222977022977023,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.0002,
+      "loss": 0.8957,
+      "step": 5287
+    },
+    {
+      "epoch": 4.223776223776224,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0002,
+      "loss": 0.9039,
+      "step": 5288
+    },
+    {
+      "epoch": 4.224575424575424,
+      "grad_norm": 0.375,
+      "learning_rate": 0.0002,
+      "loss": 0.9012,
+      "step": 5289
+    },
+    {
+      "epoch": 4.225374625374625,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8915,
+      "step": 5290
+    },
+    {
+      "epoch": 4.226173826173826,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9188,
+      "step": 5291
+    },
+    {
+      "epoch": 4.226973026973027,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0002,
+      "loss": 0.9001,
+      "step": 5292
+    },
+    {
+      "epoch": 4.227772227772228,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8945,
+      "step": 5293
+    },
+    {
+      "epoch": 4.228571428571429,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.0002,
+      "loss": 0.8971,
+      "step": 5294
+    },
+    {
+      "epoch": 4.22937062937063,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0002,
+      "loss": 0.8924,
+      "step": 5295
+    },
+    {
+      "epoch": 4.23016983016983,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8927,
+      "step": 5296
+    },
+    {
+      "epoch": 4.230969030969031,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9525,
+      "step": 5297
+    },
+    {
+      "epoch": 4.231768231768232,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0002,
+      "loss": 0.9055,
+      "step": 5298
+    },
+    {
+      "epoch": 4.232567432567433,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0002,
+      "loss": 0.8961,
+      "step": 5299
+    },
+    {
+      "epoch": 4.233366633366633,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8944,
+      "step": 5300
+    },
+    {
+      "epoch": 4.234165834165834,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8904,
+      "step": 5301
+    },
+    {
+      "epoch": 4.234965034965035,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0002,
+      "loss": 0.8984,
+      "step": 5302
+    },
+    {
+      "epoch": 4.235764235764236,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0002,
+      "loss": 0.9059,
+      "step": 5303
+    },
+    {
+      "epoch": 4.236563436563436,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.0002,
+      "loss": 0.8964,
+      "step": 5304
+    },
+    {
+      "epoch": 4.237362637362637,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9058,
+      "step": 5305
+    },
+    {
+      "epoch": 4.2381618381618384,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0002,
+      "loss": 0.8942,
+      "step": 5306
+    },
+    {
+      "epoch": 4.238961038961039,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9001,
+      "step": 5307
+    },
+    {
+      "epoch": 4.23976023976024,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0002,
+      "loss": 0.898,
+      "step": 5308
+    },
+    {
+      "epoch": 4.240559440559441,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8926,
+      "step": 5309
+    },
+    {
+      "epoch": 4.241358641358642,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0002,
+      "loss": 0.8978,
+      "step": 5310
+    },
+    {
+      "epoch": 4.242157842157842,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.0002,
+      "loss": 0.9001,
+      "step": 5311
+    },
+    {
+      "epoch": 4.242957042957043,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0002,
+      "loss": 0.8905,
+      "step": 5312
+    },
+    {
+      "epoch": 4.243756243756244,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0002,
+      "loss": 0.8885,
+      "step": 5313
+    },
+    {
+      "epoch": 4.244555444555445,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0002,
+      "loss": 0.904,
+      "step": 5314
+    },
+    {
+      "epoch": 4.245354645354645,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8933,
+      "step": 5315
+    },
+    {
+      "epoch": 4.246153846153846,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.0002,
+      "loss": 0.9056,
+      "step": 5316
+    },
+    {
+      "epoch": 4.246953046953047,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8998,
+      "step": 5317
+    },
+    {
+      "epoch": 4.247752247752247,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0002,
+      "loss": 0.8974,
+      "step": 5318
+    },
+    {
+      "epoch": 4.248551448551448,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0002,
+      "loss": 0.8816,
+      "step": 5319
+    },
+    {
+      "epoch": 4.249350649350649,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0002,
+      "loss": 0.8997,
+      "step": 5320
+    },
+    {
+      "epoch": 4.2501498501498505,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0002,
+      "loss": 0.8969,
+      "step": 5321
+    },
+    {
+      "epoch": 4.250949050949051,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8991,
+      "step": 5322
+    },
+    {
+      "epoch": 4.251748251748252,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.0002,
+      "loss": 0.8947,
+      "step": 5323
+    },
+    {
+      "epoch": 4.252547452547453,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0002,
+      "loss": 0.8939,
+      "step": 5324
+    },
+    {
+      "epoch": 4.253346653346654,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8903,
+      "step": 5325
+    },
+    {
+      "epoch": 4.254145854145854,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8991,
+      "step": 5326
+    },
+    {
+      "epoch": 4.254945054945055,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.0002,
+      "loss": 0.8839,
+      "step": 5327
+    },
+    {
+      "epoch": 4.255744255744256,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0002,
+      "loss": 0.9028,
+      "step": 5328
+    },
+    {
+      "epoch": 4.256543456543456,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0002,
+      "loss": 0.895,
+      "step": 5329
+    },
+    {
+      "epoch": 4.257342657342657,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0002,
+      "loss": 0.8899,
+      "step": 5330
+    },
+    {
+      "epoch": 4.258141858141858,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0002,
+      "loss": 0.9025,
+      "step": 5331
+    },
+    {
+      "epoch": 4.258941058941059,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.0002,
+      "loss": 0.8951,
+      "step": 5332
+    },
+    {
+      "epoch": 4.259740259740259,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0002,
+      "loss": 0.8977,
+      "step": 5333
+    },
+    {
+      "epoch": 4.26053946053946,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8998,
+      "step": 5334
+    },
+    {
+      "epoch": 4.2613386613386615,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0002,
+      "loss": 0.9025,
+      "step": 5335
+    },
+    {
+      "epoch": 4.2621378621378625,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0002,
+      "loss": 0.8991,
+      "step": 5336
+    },
+    {
+      "epoch": 4.262937062937063,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0002,
+      "loss": 0.8936,
+      "step": 5337
+    },
+    {
+      "epoch": 4.263736263736264,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0002,
+      "loss": 0.895,
+      "step": 5338
+    },
+    {
+      "epoch": 4.264535464535465,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9023,
+      "step": 5339
+    },
+    {
+      "epoch": 4.265334665334665,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0002,
+      "loss": 0.9018,
+      "step": 5340
+    },
+    {
+      "epoch": 4.266133866133866,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0002,
+      "loss": 0.9025,
+      "step": 5341
+    },
+    {
+      "epoch": 4.266933066933067,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.0002,
+      "loss": 0.8929,
+      "step": 5342
+    },
+    {
+      "epoch": 4.267732267732268,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0002,
+      "loss": 0.8958,
+      "step": 5343
+    },
+    {
+      "epoch": 4.268531468531468,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.0002,
+      "loss": 0.8942,
+      "step": 5344
+    },
+    {
+      "epoch": 4.269330669330669,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0002,
+      "loss": 0.9055,
+      "step": 5345
+    },
+    {
+      "epoch": 4.27012987012987,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.0002,
+      "loss": 0.9028,
+      "step": 5346
+    },
+    {
+      "epoch": 4.270929070929071,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9046,
+      "step": 5347
+    },
+    {
+      "epoch": 4.271728271728271,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.0002,
+      "loss": 0.893,
+      "step": 5348
+    },
+    {
+      "epoch": 4.2725274725274724,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0002,
+      "loss": 0.894,
+      "step": 5349
+    },
+    {
+      "epoch": 4.2733266733266735,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0002,
+      "loss": 0.8922,
+      "step": 5350
+    },
+    {
+      "epoch": 4.2741258741258745,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0002,
+      "loss": 0.8909,
+      "step": 5351
+    },
+    {
+      "epoch": 4.274925074925075,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0002,
+      "loss": 0.9052,
+      "step": 5352
+    },
+    {
+      "epoch": 4.275724275724276,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8997,
+      "step": 5353
+    },
+    {
+      "epoch": 4.276523476523477,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.0002,
+      "loss": 0.8926,
+      "step": 5354
+    },
+    {
+      "epoch": 4.277322677322678,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0002,
+      "loss": 0.8906,
+      "step": 5355
+    },
+    {
+      "epoch": 4.278121878121878,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.0002,
+      "loss": 0.9391,
+      "step": 5356
+    },
+    {
+      "epoch": 4.278921078921079,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0002,
+      "loss": 0.8923,
+      "step": 5357
+    },
+    {
+      "epoch": 4.27972027972028,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0002,
+      "loss": 0.8954,
+      "step": 5358
+    },
+    {
+      "epoch": 4.28051948051948,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0002,
+      "loss": 0.8997,
+      "step": 5359
+    },
+    {
+      "epoch": 4.281318681318681,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0002,
+      "loss": 0.8964,
+      "step": 5360
+    },
+    {
+      "epoch": 4.282117882117882,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8981,
+      "step": 5361
+    },
+    {
+      "epoch": 4.282917082917083,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0002,
+      "loss": 0.8959,
+      "step": 5362
+    },
+    {
+      "epoch": 4.283716283716283,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0002,
+      "loss": 0.8929,
+      "step": 5363
+    },
+    {
+      "epoch": 4.2845154845154845,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8937,
+      "step": 5364
+    },
+    {
+      "epoch": 4.2853146853146855,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0002,
+      "loss": 0.9021,
+      "step": 5365
+    },
+    {
+      "epoch": 4.2861138861138866,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8904,
+      "step": 5366
+    },
+    {
+      "epoch": 4.286913086913087,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8971,
+      "step": 5367
+    },
+    {
+      "epoch": 4.287712287712288,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8974,
+      "step": 5368
+    },
+    {
+      "epoch": 4.288511488511489,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.0002,
+      "loss": 0.8895,
+      "step": 5369
+    },
+    {
+      "epoch": 4.289310689310689,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0002,
+      "loss": 0.905,
+      "step": 5370
+    },
+    {
+      "epoch": 4.29010989010989,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0002,
+      "loss": 0.8971,
+      "step": 5371
+    },
+    {
+      "epoch": 4.290909090909091,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8939,
+      "step": 5372
+    },
+    {
+      "epoch": 4.291708291708292,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0002,
+      "loss": 0.8929,
+      "step": 5373
+    },
+    {
+      "epoch": 4.292507492507492,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8924,
+      "step": 5374
+    },
+    {
+      "epoch": 4.293306693306693,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0002,
+      "loss": 0.8927,
+      "step": 5375
+    },
+    {
+      "epoch": 4.294105894105894,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8907,
+      "step": 5376
+    },
+    {
+      "epoch": 4.294905094905095,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0002,
+      "loss": 0.8881,
+      "step": 5377
+    },
+    {
+      "epoch": 4.2957042957042955,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0002,
+      "loss": 0.9041,
+      "step": 5378
+    },
+    {
+      "epoch": 4.2965034965034965,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0002,
+      "loss": 0.9015,
+      "step": 5379
+    },
+    {
+      "epoch": 4.2973026973026975,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8905,
+      "step": 5380
+    },
+    {
+      "epoch": 4.298101898101898,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.0002,
+      "loss": 0.9346,
+      "step": 5381
+    },
+    {
+      "epoch": 4.298901098901099,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.0002,
+      "loss": 0.9056,
+      "step": 5382
+    },
+    {
+      "epoch": 4.2997002997003,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.0002,
+      "loss": 0.9055,
+      "step": 5383
+    },
+    {
+      "epoch": 4.300499500499501,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0002,
+      "loss": 0.8931,
+      "step": 5384
+    },
+    {
+      "epoch": 4.301298701298701,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.0002,
+      "loss": 0.8971,
+      "step": 5385
+    },
+    {
+      "epoch": 4.302097902097902,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.0002,
+      "loss": 0.9043,
+      "step": 5386
+    },
+    {
+      "epoch": 4.302897102897103,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.0002,
+      "loss": 0.8973,
+      "step": 5387
+    },
+    {
+      "epoch": 4.303696303696304,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8998,
+      "step": 5388
+    },
+    {
+      "epoch": 4.304495504495504,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0002,
+      "loss": 0.8891,
+      "step": 5389
+    },
+    {
+      "epoch": 4.305294705294705,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8979,
+      "step": 5390
+    },
+    {
+      "epoch": 4.306093906093906,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8941,
+      "step": 5391
+    },
+    {
+      "epoch": 4.3068931068931064,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8931,
+      "step": 5392
+    },
+    {
+      "epoch": 4.3076923076923075,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0002,
+      "loss": 0.8915,
+      "step": 5393
+    },
+    {
+      "epoch": 4.3084915084915085,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0002,
+      "loss": 0.8911,
+      "step": 5394
+    },
+    {
+      "epoch": 4.30929070929071,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.0002,
+      "loss": 0.888,
+      "step": 5395
+    },
+    {
+      "epoch": 4.31008991008991,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0002,
+      "loss": 0.8944,
+      "step": 5396
+    },
+    {
+      "epoch": 4.310889110889111,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8997,
+      "step": 5397
+    },
+    {
+      "epoch": 4.311688311688312,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8999,
+      "step": 5398
+    },
+    {
+      "epoch": 4.312487512487513,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0002,
+      "loss": 0.8964,
+      "step": 5399
+    },
+    {
+      "epoch": 4.313286713286713,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.0002,
+      "loss": 0.9003,
+      "step": 5400
+    },
+    {
+      "epoch": 4.314085914085914,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8987,
+      "step": 5401
+    },
+    {
+      "epoch": 4.314885114885115,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9005,
+      "step": 5402
+    },
+    {
+      "epoch": 4.315684315684316,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0002,
+      "loss": 0.8891,
+      "step": 5403
+    },
+    {
+      "epoch": 4.316483516483516,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0002,
+      "loss": 0.8898,
+      "step": 5404
+    },
+    {
+      "epoch": 4.317282717282717,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0002,
+      "loss": 0.8974,
+      "step": 5405
+    },
+    {
+      "epoch": 4.318081918081918,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0002,
+      "loss": 0.8962,
+      "step": 5406
+    },
+    {
+      "epoch": 4.3188811188811185,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8978,
+      "step": 5407
+    },
+    {
+      "epoch": 4.3196803196803195,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8965,
+      "step": 5408
+    },
+    {
+      "epoch": 4.3204795204795206,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0002,
+      "loss": 0.8992,
+      "step": 5409
+    },
+    {
+      "epoch": 4.321278721278722,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0002,
+      "loss": 0.9053,
+      "step": 5410
+    },
+    {
+      "epoch": 4.322077922077922,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.0002,
+      "loss": 0.8963,
+      "step": 5411
+    },
+    {
+      "epoch": 4.322877122877123,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8994,
+      "step": 5412
+    },
+    {
+      "epoch": 4.323676323676324,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0002,
+      "loss": 0.8915,
+      "step": 5413
+    },
+    {
+      "epoch": 4.324475524475525,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0002,
+      "loss": 0.8916,
+      "step": 5414
+    },
+    {
+      "epoch": 4.325274725274725,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.0002,
+      "loss": 0.8905,
+      "step": 5415
+    },
+    {
+      "epoch": 4.326073926073926,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0002,
+      "loss": 0.9037,
+      "step": 5416
+    },
+    {
+      "epoch": 4.326873126873127,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0002,
+      "loss": 0.8914,
+      "step": 5417
+    },
+    {
+      "epoch": 4.327672327672328,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.0002,
+      "loss": 0.9017,
+      "step": 5418
+    },
+    {
+      "epoch": 4.328471528471528,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0002,
+      "loss": 0.8842,
+      "step": 5419
+    },
+    {
+      "epoch": 4.329270729270729,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.0002,
+      "loss": 0.8879,
+      "step": 5420
+    },
+    {
+      "epoch": 4.33006993006993,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0002,
+      "loss": 0.9043,
+      "step": 5421
+    },
+    {
+      "epoch": 4.3308691308691305,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.0002,
+      "loss": 0.8858,
+      "step": 5422
+    },
+    {
+      "epoch": 4.3316683316683315,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0002,
+      "loss": 0.8933,
+      "step": 5423
+    },
+    {
+      "epoch": 4.332467532467533,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.0002,
+      "loss": 0.9046,
+      "step": 5424
+    },
+    {
+      "epoch": 4.333266733266734,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0002,
+      "loss": 0.9022,
+      "step": 5425
+    },
+    {
+      "epoch": 4.334065934065934,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9007,
+      "step": 5426
+    },
+    {
+      "epoch": 4.334865134865135,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0002,
+      "loss": 0.9059,
+      "step": 5427
+    },
+    {
+      "epoch": 4.335664335664336,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8967,
+      "step": 5428
+    },
+    {
+      "epoch": 4.336463536463537,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9055,
+      "step": 5429
+    },
+    {
+      "epoch": 4.337262737262737,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8965,
+      "step": 5430
+    },
+    {
+      "epoch": 4.338061938061938,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0002,
+      "loss": 0.8993,
+      "step": 5431
+    },
+    {
+      "epoch": 4.338861138861139,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8963,
+      "step": 5432
+    },
+    {
+      "epoch": 4.339660339660339,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8928,
+      "step": 5433
+    },
+    {
+      "epoch": 4.34045954045954,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0002,
+      "loss": 0.8929,
+      "step": 5434
+    },
+    {
+      "epoch": 4.341258741258741,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0002,
+      "loss": 0.8878,
+      "step": 5435
+    },
+    {
+      "epoch": 4.342057942057942,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8889,
+      "step": 5436
+    },
+    {
+      "epoch": 4.3428571428571425,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8932,
+      "step": 5437
+    },
+    {
+      "epoch": 4.343656343656344,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.0002,
+      "loss": 0.9021,
+      "step": 5438
+    },
+    {
+      "epoch": 4.344455544455545,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0002,
+      "loss": 0.8919,
+      "step": 5439
+    },
+    {
+      "epoch": 4.345254745254746,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.0002,
+      "loss": 0.9025,
+      "step": 5440
+    },
+    {
+      "epoch": 4.346053946053946,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.0002,
+      "loss": 0.9077,
+      "step": 5441
+    },
+    {
+      "epoch": 4.346853146853147,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8934,
+      "step": 5442
+    },
+    {
+      "epoch": 4.347652347652348,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8896,
+      "step": 5443
+    },
+    {
+      "epoch": 4.348451548451548,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0002,
+      "loss": 0.9001,
+      "step": 5444
+    },
+    {
+      "epoch": 4.349250749250749,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0002,
+      "loss": 0.8918,
+      "step": 5445
+    },
+    {
+      "epoch": 4.35004995004995,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0002,
+      "loss": 0.8926,
+      "step": 5446
+    },
+    {
+      "epoch": 4.350849150849151,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8982,
+      "step": 5447
+    },
+    {
+      "epoch": 4.351648351648351,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0002,
+      "loss": 0.8945,
+      "step": 5448
+    },
+    {
+      "epoch": 4.352447552447552,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9016,
+      "step": 5449
+    },
+    {
+      "epoch": 4.353246753246753,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0002,
+      "loss": 0.8879,
+      "step": 5450
+    },
+    {
+      "epoch": 4.354045954045954,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8912,
+      "step": 5451
+    },
+    {
+      "epoch": 4.3548451548451546,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0002,
+      "loss": 0.8985,
+      "step": 5452
+    },
+    {
+      "epoch": 4.355644355644356,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0002,
+      "loss": 0.8915,
+      "step": 5453
+    },
+    {
+      "epoch": 4.356443556443557,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8861,
+      "step": 5454
+    },
+    {
+      "epoch": 4.357242757242757,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0002,
+      "loss": 0.8874,
+      "step": 5455
+    },
+    {
+      "epoch": 4.358041958041958,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.0002,
+      "loss": 0.8923,
+      "step": 5456
+    },
+    {
+      "epoch": 4.358841158841159,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.0002,
+      "loss": 0.8975,
+      "step": 5457
+    },
+    {
+      "epoch": 4.35964035964036,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0002,
+      "loss": 0.891,
+      "step": 5458
+    },
+    {
+      "epoch": 4.36043956043956,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8923,
+      "step": 5459
+    },
+    {
+      "epoch": 4.361238761238761,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0002,
+      "loss": 0.9044,
+      "step": 5460
+    },
+    {
+      "epoch": 4.362037962037962,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8958,
+      "step": 5461
+    },
+    {
+      "epoch": 4.362837162837163,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8941,
+      "step": 5462
+    },
+    {
+      "epoch": 4.363636363636363,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.0002,
+      "loss": 0.8961,
+      "step": 5463
+    },
+    {
+      "epoch": 4.364435564435564,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.0002,
+      "loss": 0.8982,
+      "step": 5464
+    },
+    {
+      "epoch": 4.365234765234765,
+      "grad_norm": 0.95703125,
+      "learning_rate": 0.0002,
+      "loss": 0.8895,
+      "step": 5465
+    },
+    {
+      "epoch": 4.366033966033966,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.0002,
+      "loss": 0.8906,
+      "step": 5466
+    },
+    {
+      "epoch": 4.366833166833167,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.0002,
+      "loss": 0.8976,
+      "step": 5467
+    },
+    {
+      "epoch": 4.367632367632368,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.0002,
+      "loss": 0.905,
+      "step": 5468
+    },
+    {
+      "epoch": 4.368431568431569,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.0002,
+      "loss": 0.8896,
+      "step": 5469
+    },
+    {
+      "epoch": 4.36923076923077,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0002,
+      "loss": 0.8958,
+      "step": 5470
+    },
+    {
+      "epoch": 4.37002997002997,
+      "grad_norm": 0.98828125,
+      "learning_rate": 0.0002,
+      "loss": 0.8933,
+      "step": 5471
+    },
+    {
+      "epoch": 4.370829170829171,
+      "grad_norm": 0.90234375,
+      "learning_rate": 0.0002,
+      "loss": 0.892,
+      "step": 5472
+    },
+    {
+      "epoch": 4.371628371628372,
+      "grad_norm": 0.9765625,
+      "learning_rate": 0.0002,
+      "loss": 0.8943,
+      "step": 5473
+    },
+    {
+      "epoch": 4.372427572427572,
+      "grad_norm": 0.8984375,
+      "learning_rate": 0.0002,
+      "loss": 0.8979,
+      "step": 5474
+    },
+    {
+      "epoch": 4.373226773226773,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.0002,
+      "loss": 0.8856,
+      "step": 5475
+    },
+    {
+      "epoch": 4.374025974025974,
+      "grad_norm": 0.8125,
+      "learning_rate": 0.0002,
+      "loss": 0.9014,
+      "step": 5476
+    },
+    {
+      "epoch": 4.374825174825175,
+      "grad_norm": 0.83203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8926,
+      "step": 5477
+    },
+    {
+      "epoch": 4.375624375624375,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.0002,
+      "loss": 0.9043,
+      "step": 5478
+    },
+    {
+      "epoch": 4.376423576423576,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.0002,
+      "loss": 0.9053,
+      "step": 5479
+    },
+    {
+      "epoch": 4.377222777222777,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9059,
+      "step": 5480
+    },
+    {
+      "epoch": 4.3780219780219785,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0002,
+      "loss": 0.9036,
+      "step": 5481
+    },
+    {
+      "epoch": 4.378821178821179,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.0002,
+      "loss": 0.8977,
+      "step": 5482
+    },
+    {
+      "epoch": 4.37962037962038,
+      "grad_norm": 0.81640625,
+      "learning_rate": 0.0002,
+      "loss": 0.9047,
+      "step": 5483
+    },
+    {
+      "epoch": 4.380419580419581,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0002,
+      "loss": 0.9001,
+      "step": 5484
+    },
+    {
+      "epoch": 4.381218781218781,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.0002,
+      "loss": 0.8931,
+      "step": 5485
+    },
+    {
+      "epoch": 4.382017982017982,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.0002,
+      "loss": 0.8955,
+      "step": 5486
+    },
+    {
+      "epoch": 4.382817182817183,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.0002,
+      "loss": 0.8943,
+      "step": 5487
+    },
+    {
+      "epoch": 4.383616383616384,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0002,
+      "loss": 0.8921,
+      "step": 5488
+    },
+    {
+      "epoch": 4.384415584415584,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0002,
+      "loss": 0.9012,
+      "step": 5489
+    },
+    {
+      "epoch": 4.385214785214785,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0002,
+      "loss": 0.8975,
+      "step": 5490
+    },
+    {
+      "epoch": 4.386013986013986,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9026,
+      "step": 5491
+    },
+    {
+      "epoch": 4.386813186813187,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0002,
+      "loss": 0.8986,
+      "step": 5492
+    },
+    {
+      "epoch": 4.387612387612387,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0002,
+      "loss": 0.8995,
+      "step": 5493
+    },
+    {
+      "epoch": 4.388411588411588,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0002,
+      "loss": 0.8924,
+      "step": 5494
+    },
+    {
+      "epoch": 4.389210789210789,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8851,
+      "step": 5495
+    },
+    {
+      "epoch": 4.39000999000999,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0002,
+      "loss": 0.8943,
+      "step": 5496
+    },
+    {
+      "epoch": 4.390809190809191,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0002,
+      "loss": 0.8949,
+      "step": 5497
+    },
+    {
+      "epoch": 4.391608391608392,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8954,
+      "step": 5498
+    },
+    {
+      "epoch": 4.392407592407593,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8907,
+      "step": 5499
+    },
+    {
+      "epoch": 4.393206793206793,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0002,
+      "loss": 0.8904,
+      "step": 5500
+    },
+    {
+      "epoch": 4.394005994005994,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8983,
+      "step": 5501
+    },
+    {
+      "epoch": 4.394805194805195,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8892,
+      "step": 5502
+    },
+    {
+      "epoch": 4.395604395604396,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9023,
+      "step": 5503
+    },
+    {
+      "epoch": 4.396403596403596,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8911,
+      "step": 5504
+    },
+    {
+      "epoch": 4.397202797202797,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8912,
+      "step": 5505
+    },
+    {
+      "epoch": 4.398001998001998,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0002,
+      "loss": 0.9084,
+      "step": 5506
+    },
+    {
+      "epoch": 4.398801198801198,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8958,
+      "step": 5507
+    },
+    {
+      "epoch": 4.399600399600399,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8928,
+      "step": 5508
+    },
+    {
+      "epoch": 4.4003996003996,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0002,
+      "loss": 0.8914,
+      "step": 5509
+    },
+    {
+      "epoch": 4.4011988011988015,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9012,
+      "step": 5510
+    },
+    {
+      "epoch": 4.401998001998002,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0002,
+      "loss": 0.9012,
+      "step": 5511
+    },
+    {
+      "epoch": 4.402797202797203,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0002,
+      "loss": 0.8887,
+      "step": 5512
+    },
+    {
+      "epoch": 4.403596403596404,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0002,
+      "loss": 0.8903,
+      "step": 5513
+    },
+    {
+      "epoch": 4.404395604395605,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0002,
+      "loss": 0.893,
+      "step": 5514
+    },
+    {
+      "epoch": 4.405194805194805,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8901,
+      "step": 5515
+    },
+    {
+      "epoch": 4.405994005994006,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0002,
+      "loss": 0.8913,
+      "step": 5516
+    },
+    {
+      "epoch": 4.406793206793207,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8952,
+      "step": 5517
+    },
+    {
+      "epoch": 4.407592407592408,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0002,
+      "loss": 0.8908,
+      "step": 5518
+    },
+    {
+      "epoch": 4.408391608391608,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8964,
+      "step": 5519
+    },
+    {
+      "epoch": 4.409190809190809,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0002,
+      "loss": 0.8986,
+      "step": 5520
+    },
+    {
+      "epoch": 4.40999000999001,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0002,
+      "loss": 0.8988,
+      "step": 5521
+    },
+    {
+      "epoch": 4.41078921078921,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0002,
+      "loss": 0.8945,
+      "step": 5522
+    },
+    {
+      "epoch": 4.411588411588411,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0002,
+      "loss": 0.8966,
+      "step": 5523
+    },
+    {
+      "epoch": 4.4123876123876125,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8975,
+      "step": 5524
+    },
+    {
+      "epoch": 4.4131868131868135,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0002,
+      "loss": 0.894,
+      "step": 5525
+    },
+    {
+      "epoch": 4.413986013986014,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0002,
+      "loss": 0.89,
+      "step": 5526
+    },
+    {
+      "epoch": 4.414785214785215,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8912,
+      "step": 5527
+    },
+    {
+      "epoch": 4.415584415584416,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0002,
+      "loss": 0.8936,
+      "step": 5528
+    },
+    {
+      "epoch": 4.416383616383617,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0002,
+      "loss": 0.893,
+      "step": 5529
+    },
+    {
+      "epoch": 4.417182817182817,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8949,
+      "step": 5530
+    },
+    {
+      "epoch": 4.417982017982018,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0002,
+      "loss": 0.892,
+      "step": 5531
+    },
+    {
+      "epoch": 4.418781218781219,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0002,
+      "loss": 0.8974,
+      "step": 5532
+    },
+    {
+      "epoch": 4.41958041958042,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8886,
+      "step": 5533
+    },
+    {
+      "epoch": 4.42037962037962,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0002,
+      "loss": 0.9035,
+      "step": 5534
+    },
+    {
+      "epoch": 4.421178821178821,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8925,
+      "step": 5535
+    },
+    {
+      "epoch": 4.421978021978022,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0002,
+      "loss": 0.8918,
+      "step": 5536
+    },
+    {
+      "epoch": 4.422777222777222,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8906,
+      "step": 5537
+    },
+    {
+      "epoch": 4.4235764235764234,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8995,
+      "step": 5538
+    },
+    {
+      "epoch": 4.4243756243756245,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0002,
+      "loss": 0.9034,
+      "step": 5539
+    },
+    {
+      "epoch": 4.4251748251748255,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8918,
+      "step": 5540
+    },
+    {
+      "epoch": 4.425974025974026,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.0002,
+      "loss": 0.8878,
+      "step": 5541
+    },
+    {
+      "epoch": 4.426773226773227,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0002,
+      "loss": 0.8965,
+      "step": 5542
+    },
+    {
+      "epoch": 4.427572427572428,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0002,
+      "loss": 0.8901,
+      "step": 5543
+    },
+    {
+      "epoch": 4.428371628371629,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0002,
+      "loss": 0.8932,
+      "step": 5544
+    },
+    {
+      "epoch": 4.429170829170829,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0002,
+      "loss": 0.9064,
+      "step": 5545
+    },
+    {
+      "epoch": 4.42997002997003,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0002,
+      "loss": 0.8948,
+      "step": 5546
+    },
+    {
+      "epoch": 4.430769230769231,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0002,
+      "loss": 0.892,
+      "step": 5547
+    },
+    {
+      "epoch": 4.431568431568431,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0002,
+      "loss": 0.8874,
+      "step": 5548
+    },
+    {
+      "epoch": 4.432367632367632,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9006,
+      "step": 5549
+    },
+    {
+      "epoch": 4.433166833166833,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.0002,
+      "loss": 0.9011,
+      "step": 5550
+    },
+    {
+      "epoch": 4.433966033966034,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0002,
+      "loss": 0.903,
+      "step": 5551
+    },
+    {
+      "epoch": 4.434765234765234,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9081,
+      "step": 5552
+    },
+    {
+      "epoch": 4.4355644355644355,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8847,
+      "step": 5553
+    },
+    {
+      "epoch": 4.4363636363636365,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0002,
+      "loss": 0.8869,
+      "step": 5554
+    },
+    {
+      "epoch": 4.4371628371628375,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0002,
+      "loss": 0.8901,
+      "step": 5555
+    },
+    {
+      "epoch": 4.437962037962038,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8947,
+      "step": 5556
+    },
+    {
+      "epoch": 4.438761238761239,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8902,
+      "step": 5557
+    },
+    {
+      "epoch": 4.43956043956044,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0002,
+      "loss": 0.9021,
+      "step": 5558
+    },
+    {
+      "epoch": 4.44035964035964,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0002,
+      "loss": 0.8939,
+      "step": 5559
+    },
+    {
+      "epoch": 4.441158841158841,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.0002,
+      "loss": 0.8936,
+      "step": 5560
+    },
+    {
+      "epoch": 4.441958041958042,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8957,
+      "step": 5561
+    },
+    {
+      "epoch": 4.442757242757243,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8961,
+      "step": 5562
+    },
+    {
+      "epoch": 4.443556443556443,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0002,
+      "loss": 0.9021,
+      "step": 5563
+    },
+    {
+      "epoch": 4.444355644355644,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0002,
+      "loss": 0.8967,
+      "step": 5564
+    },
+    {
+      "epoch": 4.445154845154845,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0002,
+      "loss": 0.895,
+      "step": 5565
+    },
+    {
+      "epoch": 4.445954045954046,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8939,
+      "step": 5566
+    },
+    {
+      "epoch": 4.4467532467532465,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0002,
+      "loss": 0.8989,
+      "step": 5567
+    },
+    {
+      "epoch": 4.4475524475524475,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0002,
+      "loss": 0.8995,
+      "step": 5568
+    },
+    {
+      "epoch": 4.4483516483516485,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8976,
+      "step": 5569
+    },
+    {
+      "epoch": 4.449150849150849,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8968,
+      "step": 5570
+    },
+    {
+      "epoch": 4.44995004995005,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9069,
+      "step": 5571
+    },
+    {
+      "epoch": 4.450749250749251,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0002,
+      "loss": 0.9016,
+      "step": 5572
+    },
+    {
+      "epoch": 4.451548451548452,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0002,
+      "loss": 0.8989,
+      "step": 5573
+    },
+    {
+      "epoch": 4.452347652347652,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9028,
+      "step": 5574
+    },
+    {
+      "epoch": 4.453146853146853,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0002,
+      "loss": 0.8969,
+      "step": 5575
+    },
+    {
+      "epoch": 4.453946053946054,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9061,
+      "step": 5576
+    },
+    {
+      "epoch": 4.454745254745255,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8927,
+      "step": 5577
+    },
+    {
+      "epoch": 4.455544455544455,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8995,
+      "step": 5578
+    },
+    {
+      "epoch": 4.456343656343656,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0002,
+      "loss": 0.881,
+      "step": 5579
+    },
+    {
+      "epoch": 4.457142857142857,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0002,
+      "loss": 0.902,
+      "step": 5580
+    },
+    {
+      "epoch": 4.457942057942058,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8976,
+      "step": 5581
+    },
+    {
+      "epoch": 4.4587412587412585,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0002,
+      "loss": 0.8973,
+      "step": 5582
+    },
+    {
+      "epoch": 4.4595404595404595,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8942,
+      "step": 5583
+    },
+    {
+      "epoch": 4.460339660339661,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0002,
+      "loss": 0.9008,
+      "step": 5584
+    },
+    {
+      "epoch": 4.461138861138862,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0002,
+      "loss": 0.8924,
+      "step": 5585
+    },
+    {
+      "epoch": 4.461938061938062,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0002,
+      "loss": 0.8901,
+      "step": 5586
+    },
+    {
+      "epoch": 4.462737262737263,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0002,
+      "loss": 0.8922,
+      "step": 5587
+    },
+    {
+      "epoch": 4.463536463536464,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8871,
+      "step": 5588
+    },
+    {
+      "epoch": 4.464335664335664,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0002,
+      "loss": 0.8885,
+      "step": 5589
+    },
+    {
+      "epoch": 4.465134865134865,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8938,
+      "step": 5590
+    },
+    {
+      "epoch": 4.465934065934066,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0002,
+      "loss": 0.8944,
+      "step": 5591
+    },
+    {
+      "epoch": 4.466733266733267,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8931,
+      "step": 5592
+    },
+    {
+      "epoch": 4.467532467532467,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0002,
+      "loss": 0.8873,
+      "step": 5593
+    },
+    {
+      "epoch": 4.468331668331668,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0002,
+      "loss": 0.8995,
+      "step": 5594
+    },
+    {
+      "epoch": 4.469130869130869,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0002,
+      "loss": 0.881,
+      "step": 5595
+    },
+    {
+      "epoch": 4.46993006993007,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0002,
+      "loss": 0.8992,
+      "step": 5596
+    },
+    {
+      "epoch": 4.4707292707292705,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0002,
+      "loss": 0.8955,
+      "step": 5597
+    },
+    {
+      "epoch": 4.4715284715284715,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8895,
+      "step": 5598
+    },
+    {
+      "epoch": 4.472327672327673,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0002,
+      "loss": 0.8889,
+      "step": 5599
+    },
+    {
+      "epoch": 4.473126873126873,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0002,
+      "loss": 0.8893,
+      "step": 5600
+    },
+    {
+      "epoch": 4.473926073926074,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9012,
+      "step": 5601
+    },
+    {
+      "epoch": 4.474725274725275,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8917,
+      "step": 5602
+    },
+    {
+      "epoch": 4.475524475524476,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8945,
+      "step": 5603
+    },
+    {
+      "epoch": 4.476323676323676,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8931,
+      "step": 5604
+    },
+    {
+      "epoch": 4.477122877122877,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0002,
+      "loss": 0.889,
+      "step": 5605
+    },
+    {
+      "epoch": 4.477922077922078,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8932,
+      "step": 5606
+    },
+    {
+      "epoch": 4.478721278721279,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0002,
+      "loss": 0.893,
+      "step": 5607
+    },
+    {
+      "epoch": 4.479520479520479,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.0002,
+      "loss": 0.8967,
+      "step": 5608
+    },
+    {
+      "epoch": 4.48031968031968,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8885,
+      "step": 5609
+    },
+    {
+      "epoch": 4.481118881118881,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0002,
+      "loss": 0.8995,
+      "step": 5610
+    },
+    {
+      "epoch": 4.4819180819180815,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0002,
+      "loss": 0.9014,
+      "step": 5611
+    },
+    {
+      "epoch": 4.4827172827172825,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0002,
+      "loss": 0.9021,
+      "step": 5612
+    },
+    {
+      "epoch": 4.483516483516484,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0002,
+      "loss": 0.893,
+      "step": 5613
+    },
+    {
+      "epoch": 4.484315684315685,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8965,
+      "step": 5614
+    },
+    {
+      "epoch": 4.485114885114885,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0002,
+      "loss": 0.8954,
+      "step": 5615
+    },
+    {
+      "epoch": 4.485914085914086,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8942,
+      "step": 5616
+    },
+    {
+      "epoch": 4.486713286713287,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.0002,
+      "loss": 0.8931,
+      "step": 5617
+    },
+    {
+      "epoch": 4.487512487512488,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8959,
+      "step": 5618
+    },
+    {
+      "epoch": 4.488311688311688,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8864,
+      "step": 5619
+    },
+    {
+      "epoch": 4.489110889110889,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8952,
+      "step": 5620
+    },
+    {
+      "epoch": 4.48991008991009,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0002,
+      "loss": 0.8953,
+      "step": 5621
+    },
+    {
+      "epoch": 4.49070929070929,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9057,
+      "step": 5622
+    },
+    {
+      "epoch": 4.491508491508491,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8993,
+      "step": 5623
+    },
+    {
+      "epoch": 4.492307692307692,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0002,
+      "loss": 0.8922,
+      "step": 5624
+    },
+    {
+      "epoch": 4.493106893106893,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0002,
+      "loss": 0.895,
+      "step": 5625
+    },
+    {
+      "epoch": 4.4939060939060935,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8919,
+      "step": 5626
+    },
+    {
+      "epoch": 4.494705294705295,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0002,
+      "loss": 0.8941,
+      "step": 5627
+    },
+    {
+      "epoch": 4.495504495504496,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8962,
+      "step": 5628
+    },
+    {
+      "epoch": 4.496303696303697,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0002,
+      "loss": 0.8969,
+      "step": 5629
+    },
+    {
+      "epoch": 4.497102897102897,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9023,
+      "step": 5630
+    },
+    {
+      "epoch": 4.497902097902098,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.0002,
+      "loss": 0.8961,
+      "step": 5631
+    },
+    {
+      "epoch": 4.498701298701299,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0002,
+      "loss": 0.8976,
+      "step": 5632
+    },
+    {
+      "epoch": 4.4995004995005,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8903,
+      "step": 5633
+    },
+    {
+      "epoch": 4.5002997002997,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0002,
+      "loss": 0.8966,
+      "step": 5634
+    },
+    {
+      "epoch": 4.501098901098901,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8867,
+      "step": 5635
+    },
+    {
+      "epoch": 4.501898101898102,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8959,
+      "step": 5636
+    },
+    {
+      "epoch": 4.502697302697303,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9019,
+      "step": 5637
+    },
+    {
+      "epoch": 4.503496503496503,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8905,
+      "step": 5638
+    },
+    {
+      "epoch": 4.504295704295704,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.0002,
+      "loss": 0.9031,
+      "step": 5639
+    },
+    {
+      "epoch": 4.505094905094905,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0002,
+      "loss": 0.8927,
+      "step": 5640
+    },
+    {
+      "epoch": 4.5058941058941056,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8919,
+      "step": 5641
+    },
+    {
+      "epoch": 4.506693306693307,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.0002,
+      "loss": 0.8827,
+      "step": 5642
+    },
+    {
+      "epoch": 4.507492507492508,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0002,
+      "loss": 0.897,
+      "step": 5643
+    },
+    {
+      "epoch": 4.508291708291709,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8912,
+      "step": 5644
+    },
+    {
+      "epoch": 4.509090909090909,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.0002,
+      "loss": 0.8959,
+      "step": 5645
+    },
+    {
+      "epoch": 4.50989010989011,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8981,
+      "step": 5646
+    },
+    {
+      "epoch": 4.510689310689311,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8938,
+      "step": 5647
+    },
+    {
+      "epoch": 4.511488511488512,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8973,
+      "step": 5648
+    },
+    {
+      "epoch": 4.512287712287712,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0002,
+      "loss": 0.8939,
+      "step": 5649
+    },
+    {
+      "epoch": 4.513086913086913,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8979,
+      "step": 5650
+    },
+    {
+      "epoch": 4.513886113886114,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8921,
+      "step": 5651
+    },
+    {
+      "epoch": 4.514685314685314,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0002,
+      "loss": 0.9442,
+      "step": 5652
+    },
+    {
+      "epoch": 4.515484515484515,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0002,
+      "loss": 0.884,
+      "step": 5653
+    },
+    {
+      "epoch": 4.516283716283716,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8895,
+      "step": 5654
+    },
+    {
+      "epoch": 4.517082917082917,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8939,
+      "step": 5655
+    },
+    {
+      "epoch": 4.517882117882118,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0002,
+      "loss": 0.8915,
+      "step": 5656
+    },
+    {
+      "epoch": 4.518681318681319,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8908,
+      "step": 5657
+    },
+    {
+      "epoch": 4.51948051948052,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8948,
+      "step": 5658
+    },
+    {
+      "epoch": 4.520279720279721,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8924,
+      "step": 5659
+    },
+    {
+      "epoch": 4.521078921078921,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0002,
+      "loss": 0.8905,
+      "step": 5660
+    },
+    {
+      "epoch": 4.521878121878122,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8952,
+      "step": 5661
+    },
+    {
+      "epoch": 4.522677322677323,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0002,
+      "loss": 0.9031,
+      "step": 5662
+    },
+    {
+      "epoch": 4.523476523476523,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0002,
+      "loss": 0.8935,
+      "step": 5663
+    },
+    {
+      "epoch": 4.524275724275724,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8897,
+      "step": 5664
+    },
+    {
+      "epoch": 4.525074925074925,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8996,
+      "step": 5665
+    },
+    {
+      "epoch": 4.525874125874126,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0002,
+      "loss": 0.8997,
+      "step": 5666
+    },
+    {
+      "epoch": 4.526673326673326,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.0002,
+      "loss": 0.8892,
+      "step": 5667
+    },
+    {
+      "epoch": 4.527472527472527,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.0002,
+      "loss": 0.8984,
+      "step": 5668
+    },
+    {
+      "epoch": 4.528271728271728,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0002,
+      "loss": 0.9021,
+      "step": 5669
+    },
+    {
+      "epoch": 4.5290709290709295,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0002,
+      "loss": 0.9049,
+      "step": 5670
+    },
+    {
+      "epoch": 4.52987012987013,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8955,
+      "step": 5671
+    },
+    {
+      "epoch": 4.530669330669331,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8904,
+      "step": 5672
+    },
+    {
+      "epoch": 4.531468531468532,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.0002,
+      "loss": 0.8918,
+      "step": 5673
+    },
+    {
+      "epoch": 4.532267732267732,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0002,
+      "loss": 0.8905,
+      "step": 5674
+    },
+    {
+      "epoch": 4.533066933066933,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9024,
+      "step": 5675
+    },
+    {
+      "epoch": 4.533866133866134,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.0002,
+      "loss": 0.8911,
+      "step": 5676
+    },
+    {
+      "epoch": 4.534665334665335,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0002,
+      "loss": 0.8956,
+      "step": 5677
+    },
+    {
+      "epoch": 4.535464535464535,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8988,
+      "step": 5678
+    },
+    {
+      "epoch": 4.536263736263736,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8999,
+      "step": 5679
+    },
+    {
+      "epoch": 4.537062937062937,
+      "grad_norm": 0.375,
+      "learning_rate": 0.0002,
+      "loss": 0.9003,
+      "step": 5680
+    },
+    {
+      "epoch": 4.537862137862138,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0002,
+      "loss": 0.8962,
+      "step": 5681
+    },
+    {
+      "epoch": 4.538661338661338,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0002,
+      "loss": 0.8837,
+      "step": 5682
+    },
+    {
+      "epoch": 4.539460539460539,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0002,
+      "loss": 0.8933,
+      "step": 5683
+    },
+    {
+      "epoch": 4.54025974025974,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8911,
+      "step": 5684
+    },
+    {
+      "epoch": 4.541058941058941,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.0002,
+      "loss": 0.8939,
+      "step": 5685
+    },
+    {
+      "epoch": 4.541858141858142,
+      "grad_norm": 0.94140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9028,
+      "step": 5686
+    },
+    {
+      "epoch": 4.542657342657343,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9023,
+      "step": 5687
+    },
+    {
+      "epoch": 4.543456543456544,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8955,
+      "step": 5688
+    },
+    {
+      "epoch": 4.544255744255745,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0002,
+      "loss": 0.8867,
+      "step": 5689
+    },
+    {
+      "epoch": 4.545054945054945,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8958,
+      "step": 5690
+    },
+    {
+      "epoch": 4.545854145854146,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9029,
+      "step": 5691
+    },
+    {
+      "epoch": 4.546653346653347,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8887,
+      "step": 5692
+    },
+    {
+      "epoch": 4.547452547452547,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0002,
+      "loss": 0.8956,
+      "step": 5693
+    },
+    {
+      "epoch": 4.548251748251748,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.0002,
+      "loss": 0.895,
+      "step": 5694
+    },
+    {
+      "epoch": 4.549050949050949,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0002,
+      "loss": 0.8993,
+      "step": 5695
+    },
+    {
+      "epoch": 4.54985014985015,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8943,
+      "step": 5696
+    },
+    {
+      "epoch": 4.55064935064935,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0002,
+      "loss": 0.8932,
+      "step": 5697
+    },
+    {
+      "epoch": 4.551448551448551,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8935,
+      "step": 5698
+    },
+    {
+      "epoch": 4.5522477522477525,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0002,
+      "loss": 0.8921,
+      "step": 5699
+    },
+    {
+      "epoch": 4.5530469530469535,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0002,
+      "loss": 0.8901,
+      "step": 5700
+    },
+    {
+      "epoch": 4.553846153846154,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0002,
+      "loss": 0.8912,
+      "step": 5701
+    },
+    {
+      "epoch": 4.554645354645355,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0002,
+      "loss": 0.8928,
+      "step": 5702
+    },
+    {
+      "epoch": 4.555444555444556,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8858,
+      "step": 5703
+    },
+    {
+      "epoch": 4.556243756243756,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0002,
+      "loss": 0.8928,
+      "step": 5704
+    },
+    {
+      "epoch": 4.557042957042957,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8959,
+      "step": 5705
+    },
+    {
+      "epoch": 4.557842157842158,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8938,
+      "step": 5706
+    },
+    {
+      "epoch": 4.558641358641359,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0002,
+      "loss": 0.8987,
+      "step": 5707
+    },
+    {
+      "epoch": 4.559440559440559,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8976,
+      "step": 5708
+    },
+    {
+      "epoch": 4.56023976023976,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8902,
+      "step": 5709
+    },
+    {
+      "epoch": 4.561038961038961,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0002,
+      "loss": 0.9013,
+      "step": 5710
+    },
+    {
+      "epoch": 4.561838161838162,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8935,
+      "step": 5711
+    },
+    {
+      "epoch": 4.562637362637362,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0002,
+      "loss": 0.8898,
+      "step": 5712
+    },
+    {
+      "epoch": 4.5634365634365635,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8978,
+      "step": 5713
+    },
+    {
+      "epoch": 4.5642357642357645,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8942,
+      "step": 5714
+    },
+    {
+      "epoch": 4.565034965034965,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8972,
+      "step": 5715
+    },
+    {
+      "epoch": 4.565834165834166,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.0002,
+      "loss": 0.8976,
+      "step": 5716
+    },
+    {
+      "epoch": 4.566633366633367,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8939,
+      "step": 5717
+    },
+    {
+      "epoch": 4.567432567432568,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0002,
+      "loss": 0.8891,
+      "step": 5718
+    },
+    {
+      "epoch": 4.568231768231768,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8963,
+      "step": 5719
+    },
+    {
+      "epoch": 4.569030969030969,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8952,
+      "step": 5720
+    },
+    {
+      "epoch": 4.56983016983017,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0002,
+      "loss": 0.8871,
+      "step": 5721
+    },
+    {
+      "epoch": 4.570629370629371,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0002,
+      "loss": 0.8976,
+      "step": 5722
+    },
+    {
+      "epoch": 4.571428571428571,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.0002,
+      "loss": 0.901,
+      "step": 5723
+    },
+    {
+      "epoch": 4.572227772227772,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.0002,
+      "loss": 0.8994,
+      "step": 5724
+    },
+    {
+      "epoch": 4.573026973026973,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8978,
+      "step": 5725
+    },
+    {
+      "epoch": 4.573826173826173,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8901,
+      "step": 5726
+    },
+    {
+      "epoch": 4.574625374625374,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0002,
+      "loss": 0.8997,
+      "step": 5727
+    },
+    {
+      "epoch": 4.5754245754245755,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0002,
+      "loss": 0.898,
+      "step": 5728
+    },
+    {
+      "epoch": 4.5762237762237765,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.0002,
+      "loss": 0.8993,
+      "step": 5729
+    },
+    {
+      "epoch": 4.577022977022977,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0002,
+      "loss": 0.8999,
+      "step": 5730
+    },
+    {
+      "epoch": 4.577822177822178,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8999,
+      "step": 5731
+    },
+    {
+      "epoch": 4.578621378621379,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8899,
+      "step": 5732
+    },
+    {
+      "epoch": 4.57942057942058,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8941,
+      "step": 5733
+    },
+    {
+      "epoch": 4.58021978021978,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0002,
+      "loss": 0.8903,
+      "step": 5734
+    },
+    {
+      "epoch": 4.581018981018981,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8966,
+      "step": 5735
+    },
+    {
+      "epoch": 4.581818181818182,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0002,
+      "loss": 0.8983,
+      "step": 5736
+    },
+    {
+      "epoch": 4.582617382617382,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8951,
+      "step": 5737
+    },
+    {
+      "epoch": 4.583416583416583,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0002,
+      "loss": 0.8936,
+      "step": 5738
+    },
+    {
+      "epoch": 4.584215784215784,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8916,
+      "step": 5739
+    },
+    {
+      "epoch": 4.585014985014985,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.0002,
+      "loss": 0.8986,
+      "step": 5740
+    },
+    {
+      "epoch": 4.585814185814185,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0002,
+      "loss": 0.8987,
+      "step": 5741
+    },
+    {
+      "epoch": 4.5866133866133865,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0002,
+      "loss": 0.8921,
+      "step": 5742
+    },
+    {
+      "epoch": 4.5874125874125875,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8878,
+      "step": 5743
+    },
+    {
+      "epoch": 4.5882117882117885,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0002,
+      "loss": 0.9014,
+      "step": 5744
+    },
+    {
+      "epoch": 4.589010989010989,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9048,
+      "step": 5745
+    },
+    {
+      "epoch": 4.58981018981019,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0002,
+      "loss": 0.8975,
+      "step": 5746
+    },
+    {
+      "epoch": 4.590609390609391,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8951,
+      "step": 5747
+    },
+    {
+      "epoch": 4.591408591408591,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0002,
+      "loss": 0.8904,
+      "step": 5748
+    },
+    {
+      "epoch": 4.592207792207792,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0002,
+      "loss": 0.8917,
+      "step": 5749
+    },
+    {
+      "epoch": 4.593006993006993,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.0002,
+      "loss": 0.8881,
+      "step": 5750
+    },
+    {
+      "epoch": 4.593806193806194,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8875,
+      "step": 5751
+    },
+    {
+      "epoch": 4.594605394605395,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8861,
+      "step": 5752
+    },
+    {
+      "epoch": 4.595404595404595,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8863,
+      "step": 5753
+    },
+    {
+      "epoch": 4.596203796203796,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0002,
+      "loss": 0.8883,
+      "step": 5754
+    },
+    {
+      "epoch": 4.597002997002997,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0002,
+      "loss": 0.8913,
+      "step": 5755
+    },
+    {
+      "epoch": 4.5978021978021975,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0002,
+      "loss": 0.896,
+      "step": 5756
+    },
+    {
+      "epoch": 4.5986013986013985,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8968,
+      "step": 5757
+    },
+    {
+      "epoch": 4.5994005994005995,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8957,
+      "step": 5758
+    },
+    {
+      "epoch": 4.600199800199801,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0002,
+      "loss": 0.898,
+      "step": 5759
+    },
+    {
+      "epoch": 4.600999000999001,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8996,
+      "step": 5760
+    },
+    {
+      "epoch": 4.601798201798202,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8941,
+      "step": 5761
+    },
+    {
+      "epoch": 4.602597402597403,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0002,
+      "loss": 0.8971,
+      "step": 5762
+    },
+    {
+      "epoch": 4.603396603396604,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8867,
+      "step": 5763
+    },
+    {
+      "epoch": 4.604195804195804,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0002,
+      "loss": 0.8914,
+      "step": 5764
+    },
+    {
+      "epoch": 4.604995004995005,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8889,
+      "step": 5765
+    },
+    {
+      "epoch": 4.605794205794206,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8902,
+      "step": 5766
+    },
+    {
+      "epoch": 4.606593406593406,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8912,
+      "step": 5767
+    },
+    {
+      "epoch": 4.607392607392607,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9051,
+      "step": 5768
+    },
+    {
+      "epoch": 4.608191808191808,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.0002,
+      "loss": 0.8891,
+      "step": 5769
+    },
+    {
+      "epoch": 4.608991008991009,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0002,
+      "loss": 0.8947,
+      "step": 5770
+    },
+    {
+      "epoch": 4.6097902097902095,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8905,
+      "step": 5771
+    },
+    {
+      "epoch": 4.6105894105894105,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9019,
+      "step": 5772
+    },
+    {
+      "epoch": 4.611388611388612,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8829,
+      "step": 5773
+    },
+    {
+      "epoch": 4.612187812187813,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0002,
+      "loss": 0.8945,
+      "step": 5774
+    },
+    {
+      "epoch": 4.612987012987013,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8902,
+      "step": 5775
+    },
+    {
+      "epoch": 4.613786213786214,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8907,
+      "step": 5776
+    },
+    {
+      "epoch": 4.614585414585415,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.0002,
+      "loss": 0.8978,
+      "step": 5777
+    },
+    {
+      "epoch": 4.615384615384615,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8972,
+      "step": 5778
+    },
+    {
+      "epoch": 4.616183816183816,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8922,
+      "step": 5779
+    },
+    {
+      "epoch": 4.616983016983017,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0002,
+      "loss": 0.886,
+      "step": 5780
+    },
+    {
+      "epoch": 4.617782217782218,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8966,
+      "step": 5781
+    },
+    {
+      "epoch": 4.618581418581418,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0002,
+      "loss": 0.8898,
+      "step": 5782
+    },
+    {
+      "epoch": 4.619380619380619,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0002,
+      "loss": 0.8913,
+      "step": 5783
+    },
+    {
+      "epoch": 4.62017982017982,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8941,
+      "step": 5784
+    },
+    {
+      "epoch": 4.620979020979021,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0002,
+      "loss": 0.8932,
+      "step": 5785
+    },
+    {
+      "epoch": 4.6217782217782215,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0002,
+      "loss": 0.9013,
+      "step": 5786
+    },
+    {
+      "epoch": 4.6225774225774225,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0002,
+      "loss": 0.8891,
+      "step": 5787
+    },
+    {
+      "epoch": 4.623376623376624,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8966,
+      "step": 5788
+    },
+    {
+      "epoch": 4.624175824175824,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8948,
+      "step": 5789
+    },
+    {
+      "epoch": 4.624975024975025,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8934,
+      "step": 5790
+    },
+    {
+      "epoch": 4.625774225774226,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8951,
+      "step": 5791
+    },
+    {
+      "epoch": 4.626573426573427,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0002,
+      "loss": 0.8958,
+      "step": 5792
+    },
+    {
+      "epoch": 4.627372627372627,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8963,
+      "step": 5793
+    },
+    {
+      "epoch": 4.628171828171828,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0002,
+      "loss": 0.9019,
+      "step": 5794
+    },
+    {
+      "epoch": 4.628971028971029,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.0002,
+      "loss": 0.8857,
+      "step": 5795
+    },
+    {
+      "epoch": 4.62977022977023,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0002,
+      "loss": 0.8947,
+      "step": 5796
+    },
+    {
+      "epoch": 4.63056943056943,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0002,
+      "loss": 0.893,
+      "step": 5797
+    },
+    {
+      "epoch": 4.631368631368631,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0002,
+      "loss": 0.8997,
+      "step": 5798
+    },
+    {
+      "epoch": 4.632167832167832,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0002,
+      "loss": 0.8877,
+      "step": 5799
+    },
+    {
+      "epoch": 4.6329670329670325,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8869,
+      "step": 5800
+    },
+    {
+      "epoch": 4.6337662337662335,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8929,
+      "step": 5801
+    },
+    {
+      "epoch": 4.634565434565435,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0002,
+      "loss": 0.8863,
+      "step": 5802
+    },
+    {
+      "epoch": 4.635364635364636,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8886,
+      "step": 5803
+    },
+    {
+      "epoch": 4.636163836163837,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8929,
+      "step": 5804
+    },
+    {
+      "epoch": 4.636963036963037,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8942,
+      "step": 5805
+    },
+    {
+      "epoch": 4.637762237762238,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0002,
+      "loss": 0.8968,
+      "step": 5806
+    },
+    {
+      "epoch": 4.638561438561439,
+      "grad_norm": 0.9609375,
+      "learning_rate": 0.0002,
+      "loss": 0.898,
+      "step": 5807
+    },
+    {
+      "epoch": 4.639360639360639,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0002,
+      "loss": 0.891,
+      "step": 5808
+    },
+    {
+      "epoch": 4.64015984015984,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0002,
+      "loss": 0.9015,
+      "step": 5809
+    },
+    {
+      "epoch": 4.640959040959041,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0002,
+      "loss": 0.8932,
+      "step": 5810
+    },
+    {
+      "epoch": 4.641758241758242,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8948,
+      "step": 5811
+    },
+    {
+      "epoch": 4.642557442557442,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8866,
+      "step": 5812
+    },
+    {
+      "epoch": 4.643356643356643,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0002,
+      "loss": 0.8924,
+      "step": 5813
+    },
+    {
+      "epoch": 4.644155844155844,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0002,
+      "loss": 0.895,
+      "step": 5814
+    },
+    {
+      "epoch": 4.644955044955045,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8914,
+      "step": 5815
+    },
+    {
+      "epoch": 4.645754245754246,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8909,
+      "step": 5816
+    },
+    {
+      "epoch": 4.646553446553447,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8935,
+      "step": 5817
+    },
+    {
+      "epoch": 4.647352647352648,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0002,
+      "loss": 0.9022,
+      "step": 5818
+    },
+    {
+      "epoch": 4.648151848151848,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8933,
+      "step": 5819
+    },
+    {
+      "epoch": 4.648951048951049,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0002,
+      "loss": 0.893,
+      "step": 5820
+    },
+    {
+      "epoch": 4.64975024975025,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8992,
+      "step": 5821
+    },
+    {
+      "epoch": 4.650549450549451,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8935,
+      "step": 5822
+    },
+    {
+      "epoch": 4.651348651348651,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0002,
+      "loss": 0.887,
+      "step": 5823
+    },
+    {
+      "epoch": 4.652147852147852,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0002,
+      "loss": 0.8815,
+      "step": 5824
+    },
+    {
+      "epoch": 4.652947052947053,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0002,
+      "loss": 0.8913,
+      "step": 5825
+    },
+    {
+      "epoch": 4.653746253746254,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8999,
+      "step": 5826
+    },
+    {
+      "epoch": 4.654545454545454,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0002,
+      "loss": 0.893,
+      "step": 5827
+    },
+    {
+      "epoch": 4.655344655344655,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0002,
+      "loss": 0.8954,
+      "step": 5828
+    },
+    {
+      "epoch": 4.656143856143856,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8885,
+      "step": 5829
+    },
+    {
+      "epoch": 4.6569430569430565,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0002,
+      "loss": 0.8971,
+      "step": 5830
+    },
+    {
+      "epoch": 4.657742257742258,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0002,
+      "loss": 0.896,
+      "step": 5831
+    },
+    {
+      "epoch": 4.658541458541459,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8954,
+      "step": 5832
+    },
+    {
+      "epoch": 4.65934065934066,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0002,
+      "loss": 0.8961,
+      "step": 5833
+    },
+    {
+      "epoch": 4.66013986013986,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8963,
+      "step": 5834
+    },
+    {
+      "epoch": 4.660939060939061,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8892,
+      "step": 5835
+    },
+    {
+      "epoch": 4.661738261738262,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9,
+      "step": 5836
+    },
+    {
+      "epoch": 4.662537462537463,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.0002,
+      "loss": 0.9316,
+      "step": 5837
+    },
+    {
+      "epoch": 4.663336663336663,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0002,
+      "loss": 0.9043,
+      "step": 5838
+    },
+    {
+      "epoch": 4.664135864135864,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0002,
+      "loss": 0.8898,
+      "step": 5839
+    },
+    {
+      "epoch": 4.664935064935065,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8827,
+      "step": 5840
+    },
+    {
+      "epoch": 4.665734265734265,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0002,
+      "loss": 0.8878,
+      "step": 5841
+    },
+    {
+      "epoch": 4.666533466533466,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0002,
+      "loss": 0.8905,
+      "step": 5842
+    },
+    {
+      "epoch": 4.667332667332667,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0002,
+      "loss": 0.9015,
+      "step": 5843
+    },
+    {
+      "epoch": 4.668131868131868,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0002,
+      "loss": 0.8892,
+      "step": 5844
+    },
+    {
+      "epoch": 4.668931068931069,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0002,
+      "loss": 0.8868,
+      "step": 5845
+    },
+    {
+      "epoch": 4.66973026973027,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8946,
+      "step": 5846
+    },
+    {
+      "epoch": 4.670529470529471,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8928,
+      "step": 5847
+    },
+    {
+      "epoch": 4.671328671328672,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0002,
+      "loss": 0.8924,
+      "step": 5848
+    },
+    {
+      "epoch": 4.672127872127872,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0002,
+      "loss": 0.8934,
+      "step": 5849
+    },
+    {
+      "epoch": 4.672927072927073,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.0002,
+      "loss": 0.8927,
+      "step": 5850
+    },
+    {
+      "epoch": 4.673726273726274,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0002,
+      "loss": 0.8985,
+      "step": 5851
+    },
+    {
+      "epoch": 4.674525474525474,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.0002,
+      "loss": 0.9008,
+      "step": 5852
+    },
+    {
+      "epoch": 4.675324675324675,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0002,
+      "loss": 0.8994,
+      "step": 5853
+    },
+    {
+      "epoch": 4.676123876123876,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0002,
+      "loss": 0.9028,
+      "step": 5854
+    },
+    {
+      "epoch": 4.676923076923077,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0002,
+      "loss": 0.9037,
+      "step": 5855
+    },
+    {
+      "epoch": 4.677722277722278,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8893,
+      "step": 5856
+    },
+    {
+      "epoch": 4.678521478521478,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8991,
+      "step": 5857
+    },
+    {
+      "epoch": 4.679320679320679,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0002,
+      "loss": 0.8927,
+      "step": 5858
+    },
+    {
+      "epoch": 4.6801198801198804,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.0002,
+      "loss": 0.8966,
+      "step": 5859
+    },
+    {
+      "epoch": 4.680919080919081,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8844,
+      "step": 5860
+    },
+    {
+      "epoch": 4.681718281718282,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8939,
+      "step": 5861
+    },
+    {
+      "epoch": 4.682517482517483,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8988,
+      "step": 5862
+    },
+    {
+      "epoch": 4.683316683316683,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8963,
+      "step": 5863
+    },
+    {
+      "epoch": 4.684115884115884,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8993,
+      "step": 5864
+    },
+    {
+      "epoch": 4.684915084915085,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0002,
+      "loss": 0.8878,
+      "step": 5865
+    },
+    {
+      "epoch": 4.685714285714286,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0002,
+      "loss": 0.9004,
+      "step": 5866
+    },
+    {
+      "epoch": 4.686513486513487,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8963,
+      "step": 5867
+    },
+    {
+      "epoch": 4.687312687312687,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.0002,
+      "loss": 0.8996,
+      "step": 5868
+    },
+    {
+      "epoch": 4.688111888111888,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0002,
+      "loss": 0.8952,
+      "step": 5869
+    },
+    {
+      "epoch": 4.688911088911089,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0002,
+      "loss": 0.8934,
+      "step": 5870
+    },
+    {
+      "epoch": 4.689710289710289,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9046,
+      "step": 5871
+    },
+    {
+      "epoch": 4.69050949050949,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.0002,
+      "loss": 0.9028,
+      "step": 5872
+    },
+    {
+      "epoch": 4.691308691308691,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8886,
+      "step": 5873
+    },
+    {
+      "epoch": 4.6921078921078925,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0002,
+      "loss": 0.8914,
+      "step": 5874
+    },
+    {
+      "epoch": 4.692907092907093,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.0002,
+      "loss": 0.8954,
+      "step": 5875
+    },
+    {
+      "epoch": 4.693706293706294,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0002,
+      "loss": 0.8899,
+      "step": 5876
+    },
+    {
+      "epoch": 4.694505494505495,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0002,
+      "loss": 0.9024,
+      "step": 5877
+    },
+    {
+      "epoch": 4.695304695304696,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.0002,
+      "loss": 0.8936,
+      "step": 5878
+    },
+    {
+      "epoch": 4.696103896103896,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8947,
+      "step": 5879
+    },
+    {
+      "epoch": 4.696903096903097,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.0002,
+      "loss": 0.9137,
+      "step": 5880
+    },
+    {
+      "epoch": 4.697702297702298,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0002,
+      "loss": 0.8841,
+      "step": 5881
+    },
+    {
+      "epoch": 4.698501498501498,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.0002,
+      "loss": 0.8864,
+      "step": 5882
+    },
+    {
+      "epoch": 4.699300699300699,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0002,
+      "loss": 0.8858,
+      "step": 5883
+    },
+    {
+      "epoch": 4.7000999000999,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0002,
+      "loss": 0.8908,
+      "step": 5884
+    },
+    {
+      "epoch": 4.700899100899101,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0002,
+      "loss": 0.89,
+      "step": 5885
+    },
+    {
+      "epoch": 4.701698301698301,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9009,
+      "step": 5886
+    },
+    {
+      "epoch": 4.702497502497502,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8983,
+      "step": 5887
+    },
+    {
+      "epoch": 4.7032967032967035,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0002,
+      "loss": 0.8918,
+      "step": 5888
+    },
+    {
+      "epoch": 4.7040959040959045,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.0002,
+      "loss": 0.8901,
+      "step": 5889
+    },
+    {
+      "epoch": 4.704895104895105,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8883,
+      "step": 5890
+    },
+    {
+      "epoch": 4.705694305694306,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.0002,
+      "loss": 0.9046,
+      "step": 5891
+    },
+    {
+      "epoch": 4.706493506493507,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8928,
+      "step": 5892
+    },
+    {
+      "epoch": 4.707292707292707,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.0002,
+      "loss": 0.909,
+      "step": 5893
+    },
+    {
+      "epoch": 4.708091908091908,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9089,
+      "step": 5894
+    },
+    {
+      "epoch": 4.708891108891109,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.0002,
+      "loss": 0.8985,
+      "step": 5895
+    },
+    {
+      "epoch": 4.70969030969031,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8965,
+      "step": 5896
+    },
+    {
+      "epoch": 4.71048951048951,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8877,
+      "step": 5897
+    },
+    {
+      "epoch": 4.711288711288711,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0002,
+      "loss": 0.8986,
+      "step": 5898
+    },
+    {
+      "epoch": 4.712087912087912,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0002,
+      "loss": 0.9062,
+      "step": 5899
+    },
+    {
+      "epoch": 4.712887112887113,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8885,
+      "step": 5900
+    },
+    {
+      "epoch": 4.713686313686313,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.0002,
+      "loss": 0.9077,
+      "step": 5901
+    },
+    {
+      "epoch": 4.7144855144855145,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.0002,
+      "loss": 0.8868,
+      "step": 5902
+    },
+    {
+      "epoch": 4.7152847152847155,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8929,
+      "step": 5903
+    },
+    {
+      "epoch": 4.716083916083916,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.0002,
+      "loss": 0.8894,
+      "step": 5904
+    },
+    {
+      "epoch": 4.716883116883117,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.0002,
+      "loss": 0.8951,
+      "step": 5905
+    },
+    {
+      "epoch": 4.717682317682318,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0002,
+      "loss": 0.8925,
+      "step": 5906
+    },
+    {
+      "epoch": 4.718481518481519,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.0002,
+      "loss": 0.8908,
+      "step": 5907
+    },
+    {
+      "epoch": 4.719280719280719,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0002,
+      "loss": 0.9021,
+      "step": 5908
+    },
+    {
+      "epoch": 4.72007992007992,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.0002,
+      "loss": 0.9029,
+      "step": 5909
+    },
+    {
+      "epoch": 4.720879120879121,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.0002,
+      "loss": 0.881,
+      "step": 5910
+    },
+    {
+      "epoch": 4.721678321678322,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8869,
+      "step": 5911
+    },
+    {
+      "epoch": 4.722477522477522,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.0002,
+      "loss": 0.889,
+      "step": 5912
+    },
+    {
+      "epoch": 4.723276723276723,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0002,
+      "loss": 0.8857,
+      "step": 5913
+    },
+    {
+      "epoch": 4.724075924075924,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0002,
+      "loss": 0.8995,
+      "step": 5914
+    },
+    {
+      "epoch": 4.724875124875124,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8943,
+      "step": 5915
+    },
+    {
+      "epoch": 4.725674325674325,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8971,
+      "step": 5916
+    },
+    {
+      "epoch": 4.7264735264735265,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0002,
+      "loss": 0.8896,
+      "step": 5917
+    },
+    {
+      "epoch": 4.7272727272727275,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8987,
+      "step": 5918
+    },
+    {
+      "epoch": 4.7280719280719286,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0002,
+      "loss": 0.8877,
+      "step": 5919
+    },
+    {
+      "epoch": 4.728871128871129,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0002,
+      "loss": 0.9006,
+      "step": 5920
+    },
+    {
+      "epoch": 4.72967032967033,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.0002,
+      "loss": 0.8958,
+      "step": 5921
+    },
+    {
+      "epoch": 4.730469530469531,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.0002,
+      "loss": 0.8894,
+      "step": 5922
+    },
+    {
+      "epoch": 4.731268731268731,
+      "grad_norm": 0.375,
+      "learning_rate": 0.0002,
+      "loss": 0.8895,
+      "step": 5923
+    },
+    {
+      "epoch": 4.732067932067932,
+      "grad_norm": 0.375,
+      "learning_rate": 0.0002,
+      "loss": 0.8872,
+      "step": 5924
+    },
+    {
+      "epoch": 4.732867132867133,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0002,
+      "loss": 0.8879,
+      "step": 5925
+    },
+    {
+      "epoch": 4.733666333666334,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.0002,
+      "loss": 0.9023,
+      "step": 5926
+    },
+    {
+      "epoch": 4.734465534465534,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8939,
+      "step": 5927
+    },
+    {
+      "epoch": 4.735264735264735,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0002,
+      "loss": 0.8919,
+      "step": 5928
+    },
+    {
+      "epoch": 4.736063936063936,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0002,
+      "loss": 0.8931,
+      "step": 5929
+    },
+    {
+      "epoch": 4.736863136863137,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0002,
+      "loss": 0.9028,
+      "step": 5930
+    },
+    {
+      "epoch": 4.7376623376623375,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8945,
+      "step": 5931
+    },
+    {
+      "epoch": 4.7384615384615385,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8902,
+      "step": 5932
+    },
+    {
+      "epoch": 4.7392607392607395,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0002,
+      "loss": 0.8956,
+      "step": 5933
+    },
+    {
+      "epoch": 4.74005994005994,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0002,
+      "loss": 0.8945,
+      "step": 5934
+    },
+    {
+      "epoch": 4.740859140859141,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0002,
+      "loss": 0.8849,
+      "step": 5935
+    },
+    {
+      "epoch": 4.741658341658342,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0002,
+      "loss": 0.8864,
+      "step": 5936
+    },
+    {
+      "epoch": 4.742457542457543,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0002,
+      "loss": 0.8968,
+      "step": 5937
+    },
+    {
+      "epoch": 4.743256743256743,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8938,
+      "step": 5938
+    },
+    {
+      "epoch": 4.744055944055944,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8916,
+      "step": 5939
+    },
+    {
+      "epoch": 4.744855144855145,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0002,
+      "loss": 0.8983,
+      "step": 5940
+    },
+    {
+      "epoch": 4.745654345654346,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0002,
+      "loss": 0.894,
+      "step": 5941
+    },
+    {
+      "epoch": 4.746453546453546,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8914,
+      "step": 5942
+    },
+    {
+      "epoch": 4.747252747252747,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0002,
+      "loss": 0.8952,
+      "step": 5943
+    },
+    {
+      "epoch": 4.748051948051948,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0002,
+      "loss": 0.8959,
+      "step": 5944
+    },
+    {
+      "epoch": 4.7488511488511485,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0002,
+      "loss": 0.895,
+      "step": 5945
+    },
+    {
+      "epoch": 4.7496503496503495,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0002,
+      "loss": 0.8954,
+      "step": 5946
+    },
+    {
+      "epoch": 4.7504495504495505,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8903,
+      "step": 5947
+    },
+    {
+      "epoch": 4.751248751248752,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0002,
+      "loss": 0.9011,
+      "step": 5948
+    },
+    {
+      "epoch": 4.752047952047952,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0002,
+      "loss": 0.8992,
+      "step": 5949
+    },
+    {
+      "epoch": 4.752847152847153,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8919,
+      "step": 5950
+    },
+    {
+      "epoch": 4.753646353646354,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8905,
+      "step": 5951
+    },
+    {
+      "epoch": 4.754445554445555,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0002,
+      "loss": 0.8993,
+      "step": 5952
+    },
+    {
+      "epoch": 4.755244755244755,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8893,
+      "step": 5953
+    },
+    {
+      "epoch": 4.756043956043956,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0002,
+      "loss": 0.9,
+      "step": 5954
+    },
+    {
+      "epoch": 4.756843156843157,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0002,
+      "loss": 0.8891,
+      "step": 5955
+    },
+    {
+      "epoch": 4.757642357642357,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8987,
+      "step": 5956
+    },
+    {
+      "epoch": 4.758441558441558,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8958,
+      "step": 5957
+    },
+    {
+      "epoch": 4.759240759240759,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8977,
+      "step": 5958
+    },
+    {
+      "epoch": 4.76003996003996,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.0002,
+      "loss": 0.9167,
+      "step": 5959
+    },
+    {
+      "epoch": 4.7608391608391605,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0002,
+      "loss": 0.8995,
+      "step": 5960
+    },
+    {
+      "epoch": 4.7616383616383615,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0002,
+      "loss": 0.9032,
+      "step": 5961
+    },
+    {
+      "epoch": 4.7624375624375626,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8976,
+      "step": 5962
+    },
+    {
+      "epoch": 4.763236763236764,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8928,
+      "step": 5963
+    },
+    {
+      "epoch": 4.764035964035964,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0002,
+      "loss": 0.8939,
+      "step": 5964
+    },
+    {
+      "epoch": 4.764835164835165,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0002,
+      "loss": 0.9039,
+      "step": 5965
+    },
+    {
+      "epoch": 4.765634365634366,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0002,
+      "loss": 0.8939,
+      "step": 5966
+    },
+    {
+      "epoch": 4.766433566433566,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0002,
+      "loss": 0.9,
+      "step": 5967
+    },
+    {
+      "epoch": 4.767232767232767,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0002,
+      "loss": 0.9023,
+      "step": 5968
+    },
+    {
+      "epoch": 4.768031968031968,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0002,
+      "loss": 0.8901,
+      "step": 5969
+    },
+    {
+      "epoch": 4.768831168831169,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8962,
+      "step": 5970
+    },
+    {
+      "epoch": 4.76963036963037,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8946,
+      "step": 5971
+    },
+    {
+      "epoch": 4.77042957042957,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002,
+      "loss": 0.887,
+      "step": 5972
+    },
+    {
+      "epoch": 4.771228771228771,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0002,
+      "loss": 0.8962,
+      "step": 5973
+    },
+    {
+      "epoch": 4.772027972027972,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8986,
+      "step": 5974
+    },
+    {
+      "epoch": 4.7728271728271725,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0002,
+      "loss": 0.8984,
+      "step": 5975
+    },
+    {
+      "epoch": 4.7736263736263735,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0002,
+      "loss": 0.8974,
+      "step": 5976
+    },
+    {
+      "epoch": 4.774425574425575,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8922,
+      "step": 5977
+    },
+    {
+      "epoch": 4.775224775224775,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.0002,
+      "loss": 0.8967,
+      "step": 5978
+    },
+    {
+      "epoch": 4.776023976023976,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0002,
+      "loss": 0.8922,
+      "step": 5979
+    },
+    {
+      "epoch": 4.776823176823177,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8989,
+      "step": 5980
+    },
+    {
+      "epoch": 4.777622377622378,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8972,
+      "step": 5981
+    },
+    {
+      "epoch": 4.778421578421579,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0002,
+      "loss": 0.8993,
+      "step": 5982
+    },
+    {
+      "epoch": 4.779220779220779,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0002,
+      "loss": 0.8902,
+      "step": 5983
+    },
+    {
+      "epoch": 4.78001998001998,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0002,
+      "loss": 0.9021,
+      "step": 5984
+    },
+    {
+      "epoch": 4.780819180819181,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0002,
+      "loss": 0.8893,
+      "step": 5985
+    },
+    {
+      "epoch": 4.781618381618381,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0002,
+      "loss": 0.8861,
+      "step": 5986
+    },
+    {
+      "epoch": 4.782417582417582,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8971,
+      "step": 5987
+    },
+    {
+      "epoch": 4.783216783216783,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002,
+      "loss": 0.89,
+      "step": 5988
+    },
+    {
+      "epoch": 4.784015984015984,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8884,
+      "step": 5989
+    },
+    {
+      "epoch": 4.7848151848151845,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8945,
+      "step": 5990
+    },
+    {
+      "epoch": 4.785614385614386,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0002,
+      "loss": 0.8901,
+      "step": 5991
+    },
+    {
+      "epoch": 4.786413586413587,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8977,
+      "step": 5992
+    },
+    {
+      "epoch": 4.787212787212788,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8918,
+      "step": 5993
+    },
+    {
+      "epoch": 4.788011988011988,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0002,
+      "loss": 0.8933,
+      "step": 5994
+    },
+    {
+      "epoch": 4.788811188811189,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9167,
+      "step": 5995
+    },
+    {
+      "epoch": 4.78961038961039,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0002,
+      "loss": 0.8928,
+      "step": 5996
+    },
+    {
+      "epoch": 4.79040959040959,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.0002,
+      "loss": 0.9225,
+      "step": 5997
+    },
+    {
+      "epoch": 4.791208791208791,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8906,
+      "step": 5998
+    },
+    {
+      "epoch": 4.792007992007992,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0002,
+      "loss": 0.8958,
+      "step": 5999
+    },
+    {
+      "epoch": 4.792807192807193,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8965,
+      "step": 6000
+    },
+    {
+      "epoch": 4.793606393606393,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0002,
+      "loss": 0.8985,
+      "step": 6001
+    },
+    {
+      "epoch": 4.794405594405594,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8843,
+      "step": 6002
+    },
+    {
+      "epoch": 4.795204795204795,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.0002,
+      "loss": 0.8867,
+      "step": 6003
+    },
+    {
+      "epoch": 4.796003996003996,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0002,
+      "loss": 0.8967,
+      "step": 6004
+    },
+    {
+      "epoch": 4.796803196803197,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.0002,
+      "loss": 0.8904,
+      "step": 6005
+    },
+    {
+      "epoch": 4.797602397602398,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0002,
+      "loss": 0.8903,
+      "step": 6006
+    },
+    {
+      "epoch": 4.798401598401599,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.0002,
+      "loss": 0.8902,
+      "step": 6007
+    },
+    {
+      "epoch": 4.799200799200799,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8911,
+      "step": 6008
+    },
+    {
+      "epoch": 4.8,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8891,
+      "step": 6009
+    },
+    {
+      "epoch": 4.800799200799201,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0002,
+      "loss": 0.8955,
+      "step": 6010
+    },
+    {
+      "epoch": 4.801598401598402,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.0002,
+      "loss": 0.8952,
+      "step": 6011
+    },
+    {
+      "epoch": 4.802397602397602,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8915,
+      "step": 6012
+    },
+    {
+      "epoch": 4.803196803196803,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0002,
+      "loss": 0.889,
+      "step": 6013
+    },
+    {
+      "epoch": 4.803996003996004,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.0002,
+      "loss": 0.8926,
+      "step": 6014
+    },
+    {
+      "epoch": 4.804795204795205,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8937,
+      "step": 6015
+    },
+    {
+      "epoch": 4.805594405594405,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.0002,
+      "loss": 0.9026,
+      "step": 6016
+    },
+    {
+      "epoch": 4.806393606393606,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9025,
+      "step": 6017
+    },
+    {
+      "epoch": 4.807192807192807,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8927,
+      "step": 6018
+    },
+    {
+      "epoch": 4.8079920079920075,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8945,
+      "step": 6019
+    },
+    {
+      "epoch": 4.808791208791209,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.0002,
+      "loss": 0.89,
+      "step": 6020
+    },
+    {
+      "epoch": 4.80959040959041,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0002,
+      "loss": 0.9004,
+      "step": 6021
+    },
+    {
+      "epoch": 4.810389610389611,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.0002,
+      "loss": 0.89,
+      "step": 6022
+    },
+    {
+      "epoch": 4.811188811188811,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9011,
+      "step": 6023
+    },
+    {
+      "epoch": 4.811988011988012,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0002,
+      "loss": 0.8918,
+      "step": 6024
+    },
+    {
+      "epoch": 4.812787212787213,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.0002,
+      "loss": 0.9017,
+      "step": 6025
+    },
+    {
+      "epoch": 4.813586413586414,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8981,
+      "step": 6026
+    },
+    {
+      "epoch": 4.814385614385614,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0002,
+      "loss": 0.897,
+      "step": 6027
+    },
+    {
+      "epoch": 4.815184815184815,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0002,
+      "loss": 0.8985,
+      "step": 6028
+    },
+    {
+      "epoch": 4.815984015984016,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8996,
+      "step": 6029
+    },
+    {
+      "epoch": 4.816783216783216,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8971,
+      "step": 6030
+    },
+    {
+      "epoch": 4.817582417582417,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8925,
+      "step": 6031
+    },
+    {
+      "epoch": 4.818381618381618,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8959,
+      "step": 6032
+    },
+    {
+      "epoch": 4.819180819180819,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.0002,
+      "loss": 0.9035,
+      "step": 6033
+    },
+    {
+      "epoch": 4.8199800199800205,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0002,
+      "loss": 0.8915,
+      "step": 6034
+    },
+    {
+      "epoch": 4.820779220779221,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9002,
+      "step": 6035
+    },
+    {
+      "epoch": 4.821578421578422,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0002,
+      "loss": 0.896,
+      "step": 6036
+    },
+    {
+      "epoch": 4.822377622377623,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0002,
+      "loss": 0.8935,
+      "step": 6037
+    },
+    {
+      "epoch": 4.823176823176823,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8924,
+      "step": 6038
+    },
+    {
+      "epoch": 4.823976023976024,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0002,
+      "loss": 0.9031,
+      "step": 6039
+    },
+    {
+      "epoch": 4.824775224775225,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0002,
+      "loss": 0.898,
+      "step": 6040
+    },
+    {
+      "epoch": 4.825574425574426,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8895,
+      "step": 6041
+    },
+    {
+      "epoch": 4.826373626373626,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0002,
+      "loss": 0.899,
+      "step": 6042
+    },
+    {
+      "epoch": 4.827172827172827,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.0002,
+      "loss": 0.8912,
+      "step": 6043
+    },
+    {
+      "epoch": 4.827972027972028,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0002,
+      "loss": 0.8934,
+      "step": 6044
+    },
+    {
+      "epoch": 4.828771228771229,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8961,
+      "step": 6045
+    },
+    {
+      "epoch": 4.829570429570429,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0002,
+      "loss": 0.8883,
+      "step": 6046
+    },
+    {
+      "epoch": 4.83036963036963,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0002,
+      "loss": 0.8945,
+      "step": 6047
+    },
+    {
+      "epoch": 4.8311688311688314,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0002,
+      "loss": 0.8843,
+      "step": 6048
+    },
+    {
+      "epoch": 4.831968031968032,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0002,
+      "loss": 0.8911,
+      "step": 6049
+    },
+    {
+      "epoch": 4.832767232767233,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0002,
+      "loss": 0.9058,
+      "step": 6050
+    },
+    {
+      "epoch": 4.833566433566434,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.0002,
+      "loss": 0.8946,
+      "step": 6051
+    },
+    {
+      "epoch": 4.834365634365635,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.0002,
+      "loss": 0.8931,
+      "step": 6052
+    },
+    {
+      "epoch": 4.835164835164835,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0002,
+      "loss": 0.9008,
+      "step": 6053
+    },
+    {
+      "epoch": 4.835964035964036,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.0002,
+      "loss": 0.8988,
+      "step": 6054
+    },
+    {
+      "epoch": 4.836763236763237,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0002,
+      "loss": 0.9,
+      "step": 6055
+    },
+    {
+      "epoch": 4.837562437562438,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0002,
+      "loss": 0.8902,
+      "step": 6056
+    },
+    {
+      "epoch": 4.838361638361638,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0002,
+      "loss": 0.8906,
+      "step": 6057
+    },
+    {
+      "epoch": 4.839160839160839,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0002,
+      "loss": 0.8887,
+      "step": 6058
+    },
+    {
+      "epoch": 4.83996003996004,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0002,
+      "loss": 0.8952,
+      "step": 6059
+    },
+    {
+      "epoch": 4.84075924075924,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0002,
+      "loss": 0.8904,
+      "step": 6060
+    },
+    {
+      "epoch": 4.841558441558441,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8858,
+      "step": 6061
+    },
+    {
+      "epoch": 4.842357642357642,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8943,
+      "step": 6062
+    },
+    {
+      "epoch": 4.8431568431568435,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8902,
+      "step": 6063
+    },
+    {
+      "epoch": 4.843956043956044,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8947,
+      "step": 6064
+    },
+    {
+      "epoch": 4.844755244755245,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8948,
+      "step": 6065
+    },
+    {
+      "epoch": 4.845554445554446,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8988,
+      "step": 6066
+    },
+    {
+      "epoch": 4.846353646353647,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0002,
+      "loss": 0.8918,
+      "step": 6067
+    },
+    {
+      "epoch": 4.847152847152847,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8987,
+      "step": 6068
+    },
+    {
+      "epoch": 4.847952047952048,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8914,
+      "step": 6069
+    },
+    {
+      "epoch": 4.848751248751249,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0002,
+      "loss": 0.8961,
+      "step": 6070
+    },
+    {
+      "epoch": 4.849550449550449,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8892,
+      "step": 6071
+    },
+    {
+      "epoch": 4.85034965034965,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.0002,
+      "loss": 0.9021,
+      "step": 6072
+    },
+    {
+      "epoch": 4.851148851148851,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8887,
+      "step": 6073
+    },
+    {
+      "epoch": 4.851948051948052,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0002,
+      "loss": 0.8926,
+      "step": 6074
+    },
+    {
+      "epoch": 4.852747252747252,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8962,
+      "step": 6075
+    },
+    {
+      "epoch": 4.853546453546453,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8971,
+      "step": 6076
+    },
+    {
+      "epoch": 4.8543456543456545,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0002,
+      "loss": 0.8943,
+      "step": 6077
+    },
+    {
+      "epoch": 4.8551448551448555,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9006,
+      "step": 6078
+    },
+    {
+      "epoch": 4.855944055944056,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0002,
+      "loss": 0.896,
+      "step": 6079
+    },
+    {
+      "epoch": 4.856743256743257,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8921,
+      "step": 6080
+    },
+    {
+      "epoch": 4.857542457542458,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8999,
+      "step": 6081
+    },
+    {
+      "epoch": 4.858341658341658,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0002,
+      "loss": 0.8944,
+      "step": 6082
+    },
+    {
+      "epoch": 4.859140859140859,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.0002,
+      "loss": 0.8949,
+      "step": 6083
+    },
+    {
+      "epoch": 4.85994005994006,
+      "grad_norm": 0.375,
+      "learning_rate": 0.0002,
+      "loss": 0.8884,
+      "step": 6084
+    },
+    {
+      "epoch": 4.860739260739261,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.0002,
+      "loss": 0.8903,
+      "step": 6085
+    },
+    {
+      "epoch": 4.861538461538462,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0002,
+      "loss": 0.8936,
+      "step": 6086
+    },
+    {
+      "epoch": 4.862337662337662,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8909,
+      "step": 6087
+    },
+    {
+      "epoch": 4.863136863136863,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0002,
+      "loss": 0.8925,
+      "step": 6088
+    },
+    {
+      "epoch": 4.863936063936064,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8812,
+      "step": 6089
+    },
+    {
+      "epoch": 4.864735264735264,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0002,
+      "loss": 0.9001,
+      "step": 6090
+    },
+    {
+      "epoch": 4.8655344655344654,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0002,
+      "loss": 0.888,
+      "step": 6091
+    },
+    {
+      "epoch": 4.8663336663336665,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8989,
+      "step": 6092
+    },
+    {
+      "epoch": 4.867132867132867,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0002,
+      "loss": 0.8846,
+      "step": 6093
+    },
+    {
+      "epoch": 4.867932067932068,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.0002,
+      "loss": 0.9245,
+      "step": 6094
+    },
+    {
+      "epoch": 4.868731268731269,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.0002,
+      "loss": 0.8914,
+      "step": 6095
+    },
+    {
+      "epoch": 4.86953046953047,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0002,
+      "loss": 0.8977,
+      "step": 6096
+    },
+    {
+      "epoch": 4.870329670329671,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0002,
+      "loss": 0.892,
+      "step": 6097
+    },
+    {
+      "epoch": 4.871128871128871,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002,
+      "loss": 0.898,
+      "step": 6098
+    },
+    {
+      "epoch": 4.871928071928072,
+      "grad_norm": 0.375,
+      "learning_rate": 0.0002,
+      "loss": 0.8931,
+      "step": 6099
+    },
+    {
+      "epoch": 4.872727272727273,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.0002,
+      "loss": 0.9008,
+      "step": 6100
+    },
+    {
+      "epoch": 4.873526473526473,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0002,
+      "loss": 0.9027,
+      "step": 6101
+    },
+    {
+      "epoch": 4.874325674325674,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.0002,
+      "loss": 0.9041,
+      "step": 6102
+    },
+    {
+      "epoch": 4.875124875124875,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8979,
+      "step": 6103
+    },
+    {
+      "epoch": 4.875924075924076,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0002,
+      "loss": 0.8932,
+      "step": 6104
+    },
+    {
+      "epoch": 4.876723276723276,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8975,
+      "step": 6105
+    },
+    {
+      "epoch": 4.8775224775224775,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0002,
+      "loss": 0.8912,
+      "step": 6106
+    },
+    {
+      "epoch": 4.8783216783216785,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.0002,
+      "loss": 0.8964,
+      "step": 6107
+    },
+    {
+      "epoch": 4.8791208791208796,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0002,
+      "loss": 0.8954,
+      "step": 6108
+    },
+    {
+      "epoch": 4.87992007992008,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9117,
+      "step": 6109
+    },
+    {
+      "epoch": 4.880719280719281,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8872,
+      "step": 6110
+    },
+    {
+      "epoch": 4.881518481518482,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8987,
+      "step": 6111
+    },
+    {
+      "epoch": 4.882317682317682,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.0002,
+      "loss": 0.8969,
+      "step": 6112
+    },
+    {
+      "epoch": 4.883116883116883,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8833,
+      "step": 6113
+    },
+    {
+      "epoch": 4.883916083916084,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8921,
+      "step": 6114
+    },
+    {
+      "epoch": 4.884715284715285,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8871,
+      "step": 6115
+    },
+    {
+      "epoch": 4.885514485514485,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0002,
+      "loss": 0.8898,
+      "step": 6116
+    },
+    {
+      "epoch": 4.886313686313686,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8921,
+      "step": 6117
+    },
+    {
+      "epoch": 4.887112887112887,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.0002,
+      "loss": 0.8947,
+      "step": 6118
+    },
+    {
+      "epoch": 4.887912087912088,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8948,
+      "step": 6119
+    },
+    {
+      "epoch": 4.8887112887112885,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0002,
+      "loss": 0.8932,
+      "step": 6120
+    },
+    {
+      "epoch": 4.8895104895104895,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0002,
+      "loss": 0.8916,
+      "step": 6121
+    },
+    {
+      "epoch": 4.8903096903096905,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8913,
+      "step": 6122
+    },
+    {
+      "epoch": 4.891108891108891,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8904,
+      "step": 6123
+    },
+    {
+      "epoch": 4.891908091908092,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0002,
+      "loss": 0.8976,
+      "step": 6124
+    },
+    {
+      "epoch": 4.892707292707293,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0002,
+      "loss": 0.8988,
+      "step": 6125
+    },
+    {
+      "epoch": 4.893506493506494,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8945,
+      "step": 6126
+    },
+    {
+      "epoch": 4.894305694305694,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8892,
+      "step": 6127
+    },
+    {
+      "epoch": 4.895104895104895,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0002,
+      "loss": 0.8951,
+      "step": 6128
+    },
+    {
+      "epoch": 4.895904095904096,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0002,
+      "loss": 0.8939,
+      "step": 6129
+    },
+    {
+      "epoch": 4.896703296703297,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0002,
+      "loss": 0.8954,
+      "step": 6130
+    },
+    {
+      "epoch": 4.897502497502497,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.0002,
+      "loss": 0.892,
+      "step": 6131
+    },
+    {
+      "epoch": 4.898301698301698,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0002,
+      "loss": 0.89,
+      "step": 6132
+    },
+    {
+      "epoch": 4.899100899100899,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0002,
+      "loss": 0.8978,
+      "step": 6133
+    },
+    {
+      "epoch": 4.8999000999000994,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8922,
+      "step": 6134
+    },
+    {
+      "epoch": 4.9006993006993005,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0002,
+      "loss": 0.8939,
+      "step": 6135
+    },
+    {
+      "epoch": 4.9014985014985015,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8924,
+      "step": 6136
+    },
+    {
+      "epoch": 4.902297702297703,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0002,
+      "loss": 0.8969,
+      "step": 6137
+    },
+    {
+      "epoch": 4.903096903096903,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9022,
+      "step": 6138
+    },
+    {
+      "epoch": 4.903896103896104,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8977,
+      "step": 6139
+    },
+    {
+      "epoch": 4.904695304695305,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0002,
+      "loss": 0.8819,
+      "step": 6140
+    },
+    {
+      "epoch": 4.905494505494506,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.0002,
+      "loss": 0.8963,
+      "step": 6141
+    },
+    {
+      "epoch": 4.906293706293706,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0002,
+      "loss": 0.9024,
+      "step": 6142
+    },
+    {
+      "epoch": 4.907092907092907,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.0002,
+      "loss": 0.8885,
+      "step": 6143
+    },
+    {
+      "epoch": 4.907892107892108,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0002,
+      "loss": 0.8953,
+      "step": 6144
+    },
+    {
+      "epoch": 4.908691308691308,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8838,
+      "step": 6145
+    },
+    {
+      "epoch": 4.909490509490509,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8912,
+      "step": 6146
+    },
+    {
+      "epoch": 4.91028971028971,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0002,
+      "loss": 0.8904,
+      "step": 6147
+    },
+    {
+      "epoch": 4.911088911088911,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0002,
+      "loss": 0.8882,
+      "step": 6148
+    },
+    {
+      "epoch": 4.911888111888112,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0002,
+      "loss": 0.8884,
+      "step": 6149
+    },
+    {
+      "epoch": 4.9126873126873125,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.0002,
+      "loss": 0.8955,
+      "step": 6150
+    },
+    {
+      "epoch": 4.9134865134865136,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0002,
+      "loss": 0.8879,
+      "step": 6151
+    },
+    {
+      "epoch": 4.914285714285715,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0002,
+      "loss": 0.8956,
+      "step": 6152
+    },
+    {
+      "epoch": 4.915084915084915,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8855,
+      "step": 6153
+    },
+    {
+      "epoch": 4.915884115884116,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8958,
+      "step": 6154
+    },
+    {
+      "epoch": 4.916683316683317,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0002,
+      "loss": 0.894,
+      "step": 6155
+    },
+    {
+      "epoch": 4.917482517482518,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0002,
+      "loss": 0.8911,
+      "step": 6156
+    },
+    {
+      "epoch": 4.918281718281718,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8936,
+      "step": 6157
+    },
+    {
+      "epoch": 4.919080919080919,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0002,
+      "loss": 0.8919,
+      "step": 6158
+    },
+    {
+      "epoch": 4.91988011988012,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8967,
+      "step": 6159
+    },
+    {
+      "epoch": 4.920679320679321,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0002,
+      "loss": 0.9022,
+      "step": 6160
+    },
+    {
+      "epoch": 4.921478521478521,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0002,
+      "loss": 0.8943,
+      "step": 6161
+    },
+    {
+      "epoch": 4.922277722277722,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8984,
+      "step": 6162
+    },
+    {
+      "epoch": 4.923076923076923,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0002,
+      "loss": 0.894,
+      "step": 6163
+    },
+    {
+      "epoch": 4.9238761238761235,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8931,
+      "step": 6164
+    },
+    {
+      "epoch": 4.9246753246753245,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0002,
+      "loss": 0.8912,
+      "step": 6165
+    },
+    {
+      "epoch": 4.925474525474526,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.0002,
+      "loss": 0.8982,
+      "step": 6166
+    },
+    {
+      "epoch": 4.926273726273727,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0002,
+      "loss": 0.8854,
+      "step": 6167
+    },
+    {
+      "epoch": 4.927072927072927,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0002,
+      "loss": 0.887,
+      "step": 6168
+    },
+    {
+      "epoch": 4.927872127872128,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0002,
+      "loss": 0.8957,
+      "step": 6169
+    },
+    {
+      "epoch": 4.928671328671329,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.0002,
+      "loss": 0.8918,
+      "step": 6170
+    },
+    {
+      "epoch": 4.92947052947053,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0002,
+      "loss": 0.8963,
+      "step": 6171
+    },
+    {
+      "epoch": 4.93026973026973,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.0002,
+      "loss": 0.9001,
+      "step": 6172
+    },
+    {
+      "epoch": 4.931068931068931,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0002,
+      "loss": 0.8977,
+      "step": 6173
+    },
+    {
+      "epoch": 4.931868131868132,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.0002,
+      "loss": 0.8858,
+      "step": 6174
+    },
+    {
+      "epoch": 4.932667332667332,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9074,
+      "step": 6175
+    },
+    {
+      "epoch": 4.933466533466533,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.0002,
+      "loss": 0.8934,
+      "step": 6176
+    },
+    {
+      "epoch": 4.934265734265734,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0002,
+      "loss": 0.8882,
+      "step": 6177
+    },
+    {
+      "epoch": 4.935064935064935,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0002,
+      "loss": 0.8975,
+      "step": 6178
+    },
+    {
+      "epoch": 4.9358641358641355,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8876,
+      "step": 6179
+    },
+    {
+      "epoch": 4.936663336663337,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0002,
+      "loss": 0.8964,
+      "step": 6180
+    },
+    {
+      "epoch": 4.937462537462538,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8913,
+      "step": 6181
+    },
+    {
+      "epoch": 4.938261738261739,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0002,
+      "loss": 0.895,
+      "step": 6182
+    },
+    {
+      "epoch": 4.939060939060939,
+      "grad_norm": 0.93359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9087,
+      "step": 6183
+    },
+    {
+      "epoch": 4.93986013986014,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0002,
+      "loss": 0.9011,
+      "step": 6184
+    },
+    {
+      "epoch": 4.940659340659341,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.0002,
+      "loss": 0.9,
+      "step": 6185
+    },
+    {
+      "epoch": 4.941458541458541,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8889,
+      "step": 6186
+    },
+    {
+      "epoch": 4.942257742257742,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0002,
+      "loss": 0.8971,
+      "step": 6187
+    },
+    {
+      "epoch": 4.943056943056943,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0002,
+      "loss": 0.8909,
+      "step": 6188
+    },
+    {
+      "epoch": 4.943856143856144,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0002,
+      "loss": 0.8994,
+      "step": 6189
+    },
+    {
+      "epoch": 4.944655344655344,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0002,
+      "loss": 0.8823,
+      "step": 6190
+    },
+    {
+      "epoch": 4.945454545454545,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8868,
+      "step": 6191
+    },
+    {
+      "epoch": 4.946253746253746,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8905,
+      "step": 6192
+    },
+    {
+      "epoch": 4.947052947052947,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0002,
+      "loss": 0.8964,
+      "step": 6193
+    },
+    {
+      "epoch": 4.9478521478521476,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0002,
+      "loss": 0.8998,
+      "step": 6194
+    },
+    {
+      "epoch": 4.948651348651349,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0002,
+      "loss": 0.8969,
+      "step": 6195
+    },
+    {
+      "epoch": 4.94945054945055,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0002,
+      "loss": 0.8963,
+      "step": 6196
+    },
+    {
+      "epoch": 4.95024975024975,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0002,
+      "loss": 0.8837,
+      "step": 6197
+    },
+    {
+      "epoch": 4.951048951048951,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0002,
+      "loss": 0.8925,
+      "step": 6198
+    },
+    {
+      "epoch": 4.951848151848152,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0002,
+      "loss": 0.8994,
+      "step": 6199
+    },
+    {
+      "epoch": 4.952647352647353,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8863,
+      "step": 6200
+    },
+    {
+      "epoch": 4.953446553446554,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9018,
+      "step": 6201
+    },
+    {
+      "epoch": 4.954245754245754,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8965,
+      "step": 6202
+    },
+    {
+      "epoch": 4.955044955044955,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0002,
+      "loss": 0.8995,
+      "step": 6203
+    },
+    {
+      "epoch": 4.955844155844156,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8862,
+      "step": 6204
+    },
+    {
+      "epoch": 4.956643356643356,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.0002,
+      "loss": 0.8925,
+      "step": 6205
+    },
+    {
+      "epoch": 4.957442557442557,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.0002,
+      "loss": 0.8864,
+      "step": 6206
+    },
+    {
+      "epoch": 4.958241758241758,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.0002,
+      "loss": 0.8942,
+      "step": 6207
+    },
+    {
+      "epoch": 4.959040959040959,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.0002,
+      "loss": 0.8928,
+      "step": 6208
+    },
+    {
+      "epoch": 4.95984015984016,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.0002,
+      "loss": 0.8925,
+      "step": 6209
+    },
+    {
+      "epoch": 4.960639360639361,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.0002,
+      "loss": 0.8919,
+      "step": 6210
+    },
+    {
+      "epoch": 4.961438561438562,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0002,
+      "loss": 0.8959,
+      "step": 6211
+    },
+    {
+      "epoch": 4.962237762237763,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8961,
+      "step": 6212
+    },
+    {
+      "epoch": 4.963036963036963,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0002,
+      "loss": 0.888,
+      "step": 6213
+    },
+    {
+      "epoch": 4.963836163836164,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8856,
+      "step": 6214
+    },
+    {
+      "epoch": 4.964635364635365,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.0002,
+      "loss": 0.9048,
+      "step": 6215
+    },
+    {
+      "epoch": 4.965434565434565,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8895,
+      "step": 6216
+    },
+    {
+      "epoch": 4.966233766233766,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0002,
+      "loss": 0.9104,
+      "step": 6217
+    },
+    {
+      "epoch": 4.967032967032967,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.0002,
+      "loss": 0.9109,
+      "step": 6218
+    },
+    {
+      "epoch": 4.967832167832168,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0002,
+      "loss": 0.892,
+      "step": 6219
+    },
+    {
+      "epoch": 4.968631368631368,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8989,
+      "step": 6220
+    },
+    {
+      "epoch": 4.969430569430569,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8925,
+      "step": 6221
+    },
+    {
+      "epoch": 4.97022977022977,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8953,
+      "step": 6222
+    },
+    {
+      "epoch": 4.9710289710289715,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0002,
+      "loss": 0.8947,
+      "step": 6223
+    },
+    {
+      "epoch": 4.971828171828172,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8908,
+      "step": 6224
+    },
+    {
+      "epoch": 4.972627372627373,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8952,
+      "step": 6225
+    },
+    {
+      "epoch": 4.973426573426574,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0002,
+      "loss": 0.8868,
+      "step": 6226
+    },
+    {
+      "epoch": 4.974225774225774,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0002,
+      "loss": 0.8879,
+      "step": 6227
+    },
+    {
+      "epoch": 4.975024975024975,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8914,
+      "step": 6228
+    },
+    {
+      "epoch": 4.975824175824176,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8934,
+      "step": 6229
+    },
+    {
+      "epoch": 4.976623376623377,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0002,
+      "loss": 0.8996,
+      "step": 6230
+    },
+    {
+      "epoch": 4.977422577422577,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8936,
+      "step": 6231
+    },
+    {
+      "epoch": 4.978221778221778,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9069,
+      "step": 6232
+    },
+    {
+      "epoch": 4.979020979020979,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0002,
+      "loss": 0.8905,
+      "step": 6233
+    },
+    {
+      "epoch": 4.97982017982018,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8912,
+      "step": 6234
+    },
+    {
+      "epoch": 4.98061938061938,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0002,
+      "loss": 0.887,
+      "step": 6235
+    },
+    {
+      "epoch": 4.981418581418581,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0002,
+      "loss": 0.8914,
+      "step": 6236
+    },
+    {
+      "epoch": 4.982217782217782,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8902,
+      "step": 6237
+    },
+    {
+      "epoch": 4.983016983016983,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0002,
+      "loss": 0.8911,
+      "step": 6238
+    },
+    {
+      "epoch": 4.983816183816184,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8892,
+      "step": 6239
+    },
+    {
+      "epoch": 4.984615384615385,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0002,
+      "loss": 0.8966,
+      "step": 6240
+    },
+    {
+      "epoch": 4.985414585414586,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8974,
+      "step": 6241
+    },
+    {
+      "epoch": 4.986213786213786,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8925,
+      "step": 6242
+    },
+    {
+      "epoch": 4.987012987012987,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0002,
+      "loss": 0.8835,
+      "step": 6243
+    },
+    {
+      "epoch": 4.987812187812188,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0002,
+      "loss": 0.89,
+      "step": 6244
+    },
+    {
+      "epoch": 4.988611388611389,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0002,
+      "loss": 0.8859,
+      "step": 6245
+    },
+    {
+      "epoch": 4.989410589410589,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8985,
+      "step": 6246
+    },
+    {
+      "epoch": 4.99020979020979,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0002,
+      "loss": 0.8869,
+      "step": 6247
+    },
+    {
+      "epoch": 4.991008991008991,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8918,
+      "step": 6248
+    },
+    {
+      "epoch": 4.991808191808191,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0002,
+      "loss": 0.888,
+      "step": 6249
+    },
+    {
+      "epoch": 4.992607392607392,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0002,
+      "loss": 0.9022,
+      "step": 6250
+    },
+    {
+      "epoch": 4.993406593406593,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8993,
+      "step": 6251
+    },
+    {
+      "epoch": 4.9942057942057945,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0002,
+      "loss": 0.8905,
+      "step": 6252
+    },
+    {
+      "epoch": 4.995004995004995,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8904,
+      "step": 6253
+    },
+    {
+      "epoch": 4.995804195804196,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8918,
+      "step": 6254
+    },
+    {
+      "epoch": 4.996603396603397,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8886,
+      "step": 6255
+    },
+    {
+      "epoch": 4.997402597402598,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8895,
+      "step": 6256
+    },
+    {
+      "epoch": 4.998201798201798,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8925,
+      "step": 6257
+    },
+    {
+      "epoch": 4.999000999000999,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0002,
+      "loss": 0.8826,
+      "step": 6258
+    },
+    {
+      "epoch": 4.9998001998002,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8968,
+      "step": 6259
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0002,
+      "loss": 0.226,
+      "step": 6260
+    },
+    {
+      "epoch": 5.000799200799201,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8885,
+      "step": 6261
+    },
+    {
+      "epoch": 5.001598401598401,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8909,
+      "step": 6262
+    },
+    {
+      "epoch": 5.002397602397602,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0002,
+      "loss": 0.8836,
+      "step": 6263
+    },
+    {
+      "epoch": 5.003196803196803,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0002,
+      "loss": 0.8913,
+      "step": 6264
+    },
+    {
+      "epoch": 5.003996003996004,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8923,
+      "step": 6265
+    },
+    {
+      "epoch": 5.0047952047952045,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0002,
+      "loss": 0.8938,
+      "step": 6266
+    },
+    {
+      "epoch": 5.0055944055944055,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.0002,
+      "loss": 0.8874,
+      "step": 6267
+    },
+    {
+      "epoch": 5.0063936063936065,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0002,
+      "loss": 0.8968,
+      "step": 6268
+    },
+    {
+      "epoch": 5.007192807192808,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0002,
+      "loss": 0.8924,
+      "step": 6269
+    },
+    {
+      "epoch": 5.007992007992008,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.0002,
+      "loss": 0.9003,
+      "step": 6270
+    },
+    {
+      "epoch": 5.008791208791209,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8925,
+      "step": 6271
+    },
+    {
+      "epoch": 5.00959040959041,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0002,
+      "loss": 0.892,
+      "step": 6272
+    },
+    {
+      "epoch": 5.01038961038961,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0002,
+      "loss": 0.8958,
+      "step": 6273
+    },
+    {
+      "epoch": 5.011188811188811,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8908,
+      "step": 6274
+    },
+    {
+      "epoch": 5.011988011988012,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8957,
+      "step": 6275
+    },
+    {
+      "epoch": 5.012787212787213,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0002,
+      "loss": 0.896,
+      "step": 6276
+    },
+    {
+      "epoch": 5.013586413586413,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8937,
+      "step": 6277
+    },
+    {
+      "epoch": 5.014385614385614,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8887,
+      "step": 6278
+    },
+    {
+      "epoch": 5.015184815184815,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8865,
+      "step": 6279
+    },
+    {
+      "epoch": 5.015984015984016,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0002,
+      "loss": 0.9018,
+      "step": 6280
+    },
+    {
+      "epoch": 5.0167832167832165,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8913,
+      "step": 6281
+    },
+    {
+      "epoch": 5.0175824175824175,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8882,
+      "step": 6282
+    },
+    {
+      "epoch": 5.018381618381619,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0002,
+      "loss": 0.8791,
+      "step": 6283
+    },
+    {
+      "epoch": 5.01918081918082,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8896,
+      "step": 6284
+    },
+    {
+      "epoch": 5.01998001998002,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.0002,
+      "loss": 0.9016,
+      "step": 6285
+    },
+    {
+      "epoch": 5.020779220779221,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0002,
+      "loss": 0.8925,
+      "step": 6286
+    },
+    {
+      "epoch": 5.021578421578422,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0002,
+      "loss": 0.8907,
+      "step": 6287
+    },
+    {
+      "epoch": 5.022377622377622,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0002,
+      "loss": 0.8916,
+      "step": 6288
+    },
+    {
+      "epoch": 5.023176823176823,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0002,
+      "loss": 0.9001,
+      "step": 6289
+    },
+    {
+      "epoch": 5.023976023976024,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.0002,
+      "loss": 0.9083,
+      "step": 6290
+    },
+    {
+      "epoch": 5.024775224775225,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8911,
+      "step": 6291
+    },
+    {
+      "epoch": 5.025574425574425,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0002,
+      "loss": 0.9003,
+      "step": 6292
+    },
+    {
+      "epoch": 5.026373626373626,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0002,
+      "loss": 0.8971,
+      "step": 6293
+    },
+    {
+      "epoch": 5.027172827172827,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0002,
+      "loss": 0.884,
+      "step": 6294
+    },
+    {
+      "epoch": 5.027972027972028,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8835,
+      "step": 6295
+    },
+    {
+      "epoch": 5.0287712287712285,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0002,
+      "loss": 0.8975,
+      "step": 6296
+    },
+    {
+      "epoch": 5.0295704295704295,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8973,
+      "step": 6297
+    },
+    {
+      "epoch": 5.030369630369631,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0002,
+      "loss": 0.8922,
+      "step": 6298
+    },
+    {
+      "epoch": 5.031168831168831,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8971,
+      "step": 6299
+    },
+    {
+      "epoch": 5.031968031968032,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0002,
+      "loss": 0.893,
+      "step": 6300
+    },
+    {
+      "epoch": 5.032767232767233,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0002,
+      "loss": 0.89,
+      "step": 6301
+    },
+    {
+      "epoch": 5.033566433566434,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8803,
+      "step": 6302
+    },
+    {
+      "epoch": 5.034365634365634,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0002,
+      "loss": 0.8945,
+      "step": 6303
+    },
+    {
+      "epoch": 5.035164835164835,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.0002,
+      "loss": 0.8943,
+      "step": 6304
+    },
+    {
+      "epoch": 5.035964035964036,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8955,
+      "step": 6305
+    },
+    {
+      "epoch": 5.036763236763237,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.0002,
+      "loss": 0.8957,
+      "step": 6306
+    },
+    {
+      "epoch": 5.037562437562437,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0002,
+      "loss": 0.889,
+      "step": 6307
+    },
+    {
+      "epoch": 5.038361638361638,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0002,
+      "loss": 0.892,
+      "step": 6308
+    },
+    {
+      "epoch": 5.039160839160839,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0002,
+      "loss": 0.8879,
+      "step": 6309
+    },
+    {
+      "epoch": 5.03996003996004,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9027,
+      "step": 6310
+    },
+    {
+      "epoch": 5.0407592407592405,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0002,
+      "loss": 0.8902,
+      "step": 6311
+    },
+    {
+      "epoch": 5.041558441558442,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0002,
+      "loss": 0.942,
+      "step": 6312
+    },
+    {
+      "epoch": 5.042357642357643,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8919,
+      "step": 6313
+    },
+    {
+      "epoch": 5.043156843156843,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0002,
+      "loss": 0.8897,
+      "step": 6314
+    },
+    {
+      "epoch": 5.043956043956044,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0002,
+      "loss": 0.8961,
+      "step": 6315
+    },
+    {
+      "epoch": 5.044755244755245,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0002,
+      "loss": 0.8961,
+      "step": 6316
+    },
+    {
+      "epoch": 5.045554445554446,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8906,
+      "step": 6317
+    },
+    {
+      "epoch": 5.046353646353646,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8941,
+      "step": 6318
+    },
+    {
+      "epoch": 5.047152847152847,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0002,
+      "loss": 0.8954,
+      "step": 6319
+    },
+    {
+      "epoch": 5.047952047952048,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8886,
+      "step": 6320
+    },
+    {
+      "epoch": 5.048751248751249,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8882,
+      "step": 6321
+    },
+    {
+      "epoch": 5.049550449550449,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0002,
+      "loss": 0.8925,
+      "step": 6322
+    },
+    {
+      "epoch": 5.05034965034965,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.0002,
+      "loss": 0.8887,
+      "step": 6323
+    },
+    {
+      "epoch": 5.051148851148851,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0002,
+      "loss": 0.8936,
+      "step": 6324
+    },
+    {
+      "epoch": 5.0519480519480515,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0002,
+      "loss": 0.9008,
+      "step": 6325
+    },
+    {
+      "epoch": 5.052747252747253,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.0002,
+      "loss": 0.8926,
+      "step": 6326
+    },
+    {
+      "epoch": 5.053546453546454,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0002,
+      "loss": 0.891,
+      "step": 6327
+    },
+    {
+      "epoch": 5.054345654345655,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0002,
+      "loss": 0.8845,
+      "step": 6328
+    },
+    {
+      "epoch": 5.055144855144855,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0002,
+      "loss": 0.8976,
+      "step": 6329
+    },
+    {
+      "epoch": 5.055944055944056,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8908,
+      "step": 6330
+    },
+    {
+      "epoch": 5.056743256743257,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8884,
+      "step": 6331
+    },
+    {
+      "epoch": 5.057542457542458,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0002,
+      "loss": 0.892,
+      "step": 6332
+    },
+    {
+      "epoch": 5.058341658341658,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8982,
+      "step": 6333
+    },
+    {
+      "epoch": 5.059140859140859,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.0002,
+      "loss": 0.883,
+      "step": 6334
+    },
+    {
+      "epoch": 5.05994005994006,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8908,
+      "step": 6335
+    },
+    {
+      "epoch": 5.060739260739261,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8905,
+      "step": 6336
+    },
+    {
+      "epoch": 5.061538461538461,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8949,
+      "step": 6337
+    },
+    {
+      "epoch": 5.062337662337662,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8904,
+      "step": 6338
+    },
+    {
+      "epoch": 5.063136863136863,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0002,
+      "loss": 0.8992,
+      "step": 6339
+    },
+    {
+      "epoch": 5.0639360639360635,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8941,
+      "step": 6340
+    },
+    {
+      "epoch": 5.064735264735265,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8895,
+      "step": 6341
+    },
+    {
+      "epoch": 5.065534465534466,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8913,
+      "step": 6342
+    },
+    {
+      "epoch": 5.066333666333667,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8956,
+      "step": 6343
+    },
+    {
+      "epoch": 5.067132867132867,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8855,
+      "step": 6344
+    },
+    {
+      "epoch": 5.067932067932068,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0002,
+      "loss": 0.882,
+      "step": 6345
+    },
+    {
+      "epoch": 5.068731268731269,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8876,
+      "step": 6346
+    },
+    {
+      "epoch": 5.06953046953047,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8932,
+      "step": 6347
+    },
+    {
+      "epoch": 5.07032967032967,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0002,
+      "loss": 0.8884,
+      "step": 6348
+    },
+    {
+      "epoch": 5.071128871128871,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8923,
+      "step": 6349
+    },
+    {
+      "epoch": 5.071928071928072,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0002,
+      "loss": 0.9025,
+      "step": 6350
+    },
+    {
+      "epoch": 5.072727272727272,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8781,
+      "step": 6351
+    },
+    {
+      "epoch": 5.073526473526473,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8911,
+      "step": 6352
+    },
+    {
+      "epoch": 5.074325674325674,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8996,
+      "step": 6353
+    },
+    {
+      "epoch": 5.075124875124875,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9008,
+      "step": 6354
+    },
+    {
+      "epoch": 5.075924075924076,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8917,
+      "step": 6355
+    },
+    {
+      "epoch": 5.076723276723277,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8881,
+      "step": 6356
+    },
+    {
+      "epoch": 5.077522477522478,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0002,
+      "loss": 0.8877,
+      "step": 6357
+    },
+    {
+      "epoch": 5.078321678321679,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0002,
+      "loss": 0.9048,
+      "step": 6358
+    },
+    {
+      "epoch": 5.079120879120879,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8919,
+      "step": 6359
+    },
+    {
+      "epoch": 5.07992007992008,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8976,
+      "step": 6360
+    },
+    {
+      "epoch": 5.080719280719281,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0002,
+      "loss": 0.9007,
+      "step": 6361
+    },
+    {
+      "epoch": 5.081518481518482,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8821,
+      "step": 6362
+    },
+    {
+      "epoch": 5.082317682317682,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0002,
+      "loss": 0.8844,
+      "step": 6363
+    },
+    {
+      "epoch": 5.083116883116883,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0002,
+      "loss": 0.8975,
+      "step": 6364
+    },
+    {
+      "epoch": 5.083916083916084,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0002,
+      "loss": 0.8872,
+      "step": 6365
+    },
+    {
+      "epoch": 5.084715284715284,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0002,
+      "loss": 0.8795,
+      "step": 6366
+    },
+    {
+      "epoch": 5.085514485514485,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0002,
+      "loss": 0.9015,
+      "step": 6367
+    },
+    {
+      "epoch": 5.086313686313686,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0002,
+      "loss": 0.9022,
+      "step": 6368
+    },
+    {
+      "epoch": 5.0871128871128874,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8881,
+      "step": 6369
+    },
+    {
+      "epoch": 5.087912087912088,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0002,
+      "loss": 0.8751,
+      "step": 6370
+    },
+    {
+      "epoch": 5.088711288711289,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8817,
+      "step": 6371
+    },
+    {
+      "epoch": 5.08951048951049,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9068,
+      "step": 6372
+    },
+    {
+      "epoch": 5.090309690309691,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8923,
+      "step": 6373
+    },
+    {
+      "epoch": 5.091108891108891,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0002,
+      "loss": 0.9012,
+      "step": 6374
+    },
+    {
+      "epoch": 5.091908091908092,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0002,
+      "loss": 0.8997,
+      "step": 6375
+    },
+    {
+      "epoch": 5.092707292707293,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0002,
+      "loss": 0.8908,
+      "step": 6376
+    },
+    {
+      "epoch": 5.093506493506493,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0002,
+      "loss": 0.9136,
+      "step": 6377
+    },
+    {
+      "epoch": 5.094305694305694,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8951,
+      "step": 6378
+    },
+    {
+      "epoch": 5.095104895104895,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8835,
+      "step": 6379
+    },
+    {
+      "epoch": 5.095904095904096,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0002,
+      "loss": 0.8842,
+      "step": 6380
+    },
+    {
+      "epoch": 5.096703296703296,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0002,
+      "loss": 0.8935,
+      "step": 6381
+    },
+    {
+      "epoch": 5.097502497502497,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8905,
+      "step": 6382
+    },
+    {
+      "epoch": 5.098301698301698,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8948,
+      "step": 6383
+    },
+    {
+      "epoch": 5.0991008991008995,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.0002,
+      "loss": 0.8933,
+      "step": 6384
+    },
+    {
+      "epoch": 5.0999000999001,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.0002,
+      "loss": 0.899,
+      "step": 6385
+    },
+    {
+      "epoch": 5.100699300699301,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0002,
+      "loss": 0.8825,
+      "step": 6386
+    },
+    {
+      "epoch": 5.101498501498502,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8902,
+      "step": 6387
+    },
+    {
+      "epoch": 5.102297702297703,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.0002,
+      "loss": 0.9002,
+      "step": 6388
+    },
+    {
+      "epoch": 5.103096903096903,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.0002,
+      "loss": 0.8855,
+      "step": 6389
+    },
+    {
+      "epoch": 5.103896103896104,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0002,
+      "loss": 0.884,
+      "step": 6390
+    },
+    {
+      "epoch": 5.104695304695305,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0002,
+      "loss": 0.9012,
+      "step": 6391
+    },
+    {
+      "epoch": 5.105494505494505,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.0002,
+      "loss": 0.8953,
+      "step": 6392
+    },
+    {
+      "epoch": 5.106293706293706,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8953,
+      "step": 6393
+    },
+    {
+      "epoch": 5.107092907092907,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8951,
+      "step": 6394
+    },
+    {
+      "epoch": 5.107892107892108,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8913,
+      "step": 6395
+    },
+    {
+      "epoch": 5.108691308691308,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8833,
+      "step": 6396
+    },
+    {
+      "epoch": 5.109490509490509,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0002,
+      "loss": 0.8934,
+      "step": 6397
+    },
+    {
+      "epoch": 5.1102897102897105,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.0002,
+      "loss": 0.896,
+      "step": 6398
+    },
+    {
+      "epoch": 5.1110889110889115,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.0002,
+      "loss": 0.894,
+      "step": 6399
+    },
+    {
+      "epoch": 5.111888111888112,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.0002,
+      "loss": 0.8906,
+      "step": 6400
+    },
+    {
+      "epoch": 5.112687312687313,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8947,
+      "step": 6401
+    },
+    {
+      "epoch": 5.113486513486514,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8933,
+      "step": 6402
+    },
+    {
+      "epoch": 5.114285714285714,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9029,
+      "step": 6403
+    },
+    {
+      "epoch": 5.115084915084915,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0002,
+      "loss": 0.8876,
+      "step": 6404
+    },
+    {
+      "epoch": 5.115884115884116,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0002,
+      "loss": 0.8987,
+      "step": 6405
+    },
+    {
+      "epoch": 5.116683316683317,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8899,
+      "step": 6406
+    },
+    {
+      "epoch": 5.117482517482517,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0002,
+      "loss": 0.8935,
+      "step": 6407
+    },
+    {
+      "epoch": 5.118281718281718,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.0002,
+      "loss": 0.885,
+      "step": 6408
+    },
+    {
+      "epoch": 5.119080919080919,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.0002,
+      "loss": 0.9019,
+      "step": 6409
+    },
+    {
+      "epoch": 5.11988011988012,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8887,
+      "step": 6410
+    },
+    {
+      "epoch": 5.12067932067932,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0002,
+      "loss": 0.8978,
+      "step": 6411
+    },
+    {
+      "epoch": 5.1214785214785215,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8955,
+      "step": 6412
+    },
+    {
+      "epoch": 5.1222777222777225,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0002,
+      "loss": 0.8834,
+      "step": 6413
+    },
+    {
+      "epoch": 5.123076923076923,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0002,
+      "loss": 0.8984,
+      "step": 6414
+    },
+    {
+      "epoch": 5.123876123876124,
+      "grad_norm": 0.375,
+      "learning_rate": 0.0002,
+      "loss": 0.8961,
+      "step": 6415
+    },
+    {
+      "epoch": 5.124675324675325,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8894,
+      "step": 6416
+    },
+    {
+      "epoch": 5.125474525474526,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0002,
+      "loss": 0.8982,
+      "step": 6417
+    },
+    {
+      "epoch": 5.126273726273726,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0002,
+      "loss": 0.8848,
+      "step": 6418
+    },
+    {
+      "epoch": 5.127072927072927,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0002,
+      "loss": 0.8892,
+      "step": 6419
+    },
+    {
+      "epoch": 5.127872127872128,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0002,
+      "loss": 0.8999,
+      "step": 6420
+    },
+    {
+      "epoch": 5.128671328671329,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0002,
+      "loss": 0.8868,
+      "step": 6421
+    },
+    {
+      "epoch": 5.129470529470529,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0002,
+      "loss": 0.889,
+      "step": 6422
+    },
+    {
+      "epoch": 5.13026973026973,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0002,
+      "loss": 0.8949,
+      "step": 6423
+    },
+    {
+      "epoch": 5.131068931068931,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.0002,
+      "loss": 0.8956,
+      "step": 6424
+    },
+    {
+      "epoch": 5.131868131868132,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0002,
+      "loss": 0.8902,
+      "step": 6425
+    },
+    {
+      "epoch": 5.132667332667332,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0002,
+      "loss": 0.8952,
+      "step": 6426
+    },
+    {
+      "epoch": 5.1334665334665335,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0002,
+      "loss": 0.8984,
+      "step": 6427
+    },
+    {
+      "epoch": 5.1342657342657345,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8892,
+      "step": 6428
+    },
+    {
+      "epoch": 5.135064935064935,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0002,
+      "loss": 0.8922,
+      "step": 6429
+    },
+    {
+      "epoch": 5.135864135864136,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8875,
+      "step": 6430
+    },
+    {
+      "epoch": 5.136663336663337,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0002,
+      "loss": 0.9015,
+      "step": 6431
+    },
+    {
+      "epoch": 5.137462537462538,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0002,
+      "loss": 0.891,
+      "step": 6432
+    },
+    {
+      "epoch": 5.138261738261738,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0002,
+      "loss": 0.9004,
+      "step": 6433
+    },
+    {
+      "epoch": 5.139060939060939,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8952,
+      "step": 6434
+    },
+    {
+      "epoch": 5.13986013986014,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0002,
+      "loss": 0.9033,
+      "step": 6435
+    },
+    {
+      "epoch": 5.140659340659341,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.0002,
+      "loss": 0.8996,
+      "step": 6436
+    },
+    {
+      "epoch": 5.141458541458541,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0002,
+      "loss": 0.8952,
+      "step": 6437
+    },
+    {
+      "epoch": 5.142257742257742,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0002,
+      "loss": 0.8901,
+      "step": 6438
+    },
+    {
+      "epoch": 5.143056943056943,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8892,
+      "step": 6439
+    },
+    {
+      "epoch": 5.143856143856144,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0002,
+      "loss": 0.8888,
+      "step": 6440
+    },
+    {
+      "epoch": 5.1446553446553445,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0002,
+      "loss": 0.8838,
+      "step": 6441
+    },
+    {
+      "epoch": 5.1454545454545455,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0002,
+      "loss": 0.8876,
+      "step": 6442
+    },
+    {
+      "epoch": 5.1462537462537465,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0002,
+      "loss": 0.8931,
+      "step": 6443
+    },
+    {
+      "epoch": 5.147052947052947,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0002,
+      "loss": 0.8835,
+      "step": 6444
+    },
+    {
+      "epoch": 5.147852147852148,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8902,
+      "step": 6445
+    },
+    {
+      "epoch": 5.148651348651349,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0002,
+      "loss": 0.9037,
+      "step": 6446
+    },
+    {
+      "epoch": 5.14945054945055,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0002,
+      "loss": 0.8871,
+      "step": 6447
+    },
+    {
+      "epoch": 5.15024975024975,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0002,
+      "loss": 0.9018,
+      "step": 6448
+    },
+    {
+      "epoch": 5.151048951048951,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0002,
+      "loss": 0.8878,
+      "step": 6449
+    },
+    {
+      "epoch": 5.151848151848152,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8949,
+      "step": 6450
+    },
+    {
+      "epoch": 5.152647352647353,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0002,
+      "loss": 0.8931,
+      "step": 6451
+    },
+    {
+      "epoch": 5.153446553446553,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.0002,
+      "loss": 0.8948,
+      "step": 6452
+    },
+    {
+      "epoch": 5.154245754245754,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0002,
+      "loss": 0.8931,
+      "step": 6453
+    },
+    {
+      "epoch": 5.155044955044955,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8931,
+      "step": 6454
+    },
+    {
+      "epoch": 5.1558441558441555,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8859,
+      "step": 6455
+    },
+    {
+      "epoch": 5.1566433566433565,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8879,
+      "step": 6456
+    },
+    {
+      "epoch": 5.1574425574425575,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0002,
+      "loss": 0.8968,
+      "step": 6457
+    },
+    {
+      "epoch": 5.158241758241759,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8941,
+      "step": 6458
+    },
+    {
+      "epoch": 5.159040959040959,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8938,
+      "step": 6459
+    },
+    {
+      "epoch": 5.15984015984016,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8898,
+      "step": 6460
+    },
+    {
+      "epoch": 5.160639360639361,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0002,
+      "loss": 0.9006,
+      "step": 6461
+    },
+    {
+      "epoch": 5.161438561438562,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8931,
+      "step": 6462
+    },
+    {
+      "epoch": 5.162237762237762,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0002,
+      "loss": 0.8891,
+      "step": 6463
+    },
+    {
+      "epoch": 5.163036963036963,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.0002,
+      "loss": 0.9062,
+      "step": 6464
+    },
+    {
+      "epoch": 5.163836163836164,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0002,
+      "loss": 0.8965,
+      "step": 6465
+    },
+    {
+      "epoch": 5.164635364635364,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0002,
+      "loss": 0.8956,
+      "step": 6466
+    },
+    {
+      "epoch": 5.165434565434565,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8893,
+      "step": 6467
+    },
+    {
+      "epoch": 5.166233766233766,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.0002,
+      "loss": 0.8941,
+      "step": 6468
+    },
+    {
+      "epoch": 5.167032967032967,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8908,
+      "step": 6469
+    },
+    {
+      "epoch": 5.1678321678321675,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0002,
+      "loss": 0.891,
+      "step": 6470
+    },
+    {
+      "epoch": 5.1686313686313685,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8887,
+      "step": 6471
+    },
+    {
+      "epoch": 5.1694305694305696,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8916,
+      "step": 6472
+    },
+    {
+      "epoch": 5.170229770229771,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8972,
+      "step": 6473
+    },
+    {
+      "epoch": 5.171028971028971,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0002,
+      "loss": 0.8788,
+      "step": 6474
+    },
+    {
+      "epoch": 5.171828171828172,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8919,
+      "step": 6475
+    },
+    {
+      "epoch": 5.172627372627373,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.0002,
+      "loss": 0.889,
+      "step": 6476
+    },
+    {
+      "epoch": 5.173426573426573,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0002,
+      "loss": 0.8895,
+      "step": 6477
+    },
+    {
+      "epoch": 5.174225774225774,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0002,
+      "loss": 0.8897,
+      "step": 6478
+    },
+    {
+      "epoch": 5.175024975024975,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8988,
+      "step": 6479
+    },
+    {
+      "epoch": 5.175824175824176,
+      "grad_norm": 1.546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9192,
+      "step": 6480
+    },
+    {
+      "epoch": 5.176623376623376,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8997,
+      "step": 6481
+    },
+    {
+      "epoch": 5.177422577422577,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8884,
+      "step": 6482
+    },
+    {
+      "epoch": 5.178221778221778,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0002,
+      "loss": 0.9013,
+      "step": 6483
+    },
+    {
+      "epoch": 5.179020979020979,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0002,
+      "loss": 0.9083,
+      "step": 6484
+    },
+    {
+      "epoch": 5.1798201798201795,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8913,
+      "step": 6485
+    },
+    {
+      "epoch": 5.1806193806193805,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0002,
+      "loss": 0.8918,
+      "step": 6486
+    },
+    {
+      "epoch": 5.181418581418582,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8863,
+      "step": 6487
+    },
+    {
+      "epoch": 5.182217782217783,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0002,
+      "loss": 0.8942,
+      "step": 6488
+    },
+    {
+      "epoch": 5.183016983016983,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8922,
+      "step": 6489
+    },
+    {
+      "epoch": 5.183816183816184,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0002,
+      "loss": 0.8923,
+      "step": 6490
+    },
+    {
+      "epoch": 5.184615384615385,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8941,
+      "step": 6491
+    },
+    {
+      "epoch": 5.185414585414585,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0002,
+      "loss": 0.8994,
+      "step": 6492
+    },
+    {
+      "epoch": 5.186213786213786,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8874,
+      "step": 6493
+    },
+    {
+      "epoch": 5.187012987012987,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8937,
+      "step": 6494
+    },
+    {
+      "epoch": 5.187812187812188,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8956,
+      "step": 6495
+    },
+    {
+      "epoch": 5.188611388611388,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8892,
+      "step": 6496
+    },
+    {
+      "epoch": 5.189410589410589,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0002,
+      "loss": 0.9017,
+      "step": 6497
+    },
+    {
+      "epoch": 5.19020979020979,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8858,
+      "step": 6498
+    },
+    {
+      "epoch": 5.191008991008991,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0002,
+      "loss": 0.8925,
+      "step": 6499
+    },
+    {
+      "epoch": 5.1918081918081915,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.0002,
+      "loss": 0.8996,
+      "step": 6500
+    },
+    {
+      "epoch": 5.192607392607393,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0002,
+      "loss": 0.8944,
+      "step": 6501
+    },
+    {
+      "epoch": 5.193406593406594,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0002,
+      "loss": 0.8878,
+      "step": 6502
+    },
+    {
+      "epoch": 5.194205794205795,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.0002,
+      "loss": 0.89,
+      "step": 6503
+    },
+    {
+      "epoch": 5.195004995004995,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8891,
+      "step": 6504
+    },
+    {
+      "epoch": 5.195804195804196,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0002,
+      "loss": 0.892,
+      "step": 6505
+    },
+    {
+      "epoch": 5.196603396603397,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0002,
+      "loss": 0.9006,
+      "step": 6506
+    },
+    {
+      "epoch": 5.197402597402597,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0002,
+      "loss": 0.8925,
+      "step": 6507
+    },
+    {
+      "epoch": 5.198201798201798,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8992,
+      "step": 6508
+    },
+    {
+      "epoch": 5.199000999000999,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8994,
+      "step": 6509
+    },
+    {
+      "epoch": 5.1998001998002,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0002,
+      "loss": 0.8897,
+      "step": 6510
+    },
+    {
+      "epoch": 5.2005994005994,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.0002,
+      "loss": 0.8882,
+      "step": 6511
+    },
+    {
+      "epoch": 5.201398601398601,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0002,
+      "loss": 0.8852,
+      "step": 6512
+    },
+    {
+      "epoch": 5.202197802197802,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0002,
+      "loss": 0.8852,
+      "step": 6513
+    },
+    {
+      "epoch": 5.202997002997003,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0002,
+      "loss": 0.8874,
+      "step": 6514
+    },
+    {
+      "epoch": 5.203796203796204,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0002,
+      "loss": 0.895,
+      "step": 6515
+    },
+    {
+      "epoch": 5.204595404595405,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.0002,
+      "loss": 0.8864,
+      "step": 6516
+    },
+    {
+      "epoch": 5.205394605394606,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0002,
+      "loss": 0.8911,
+      "step": 6517
+    },
+    {
+      "epoch": 5.206193806193806,
+      "grad_norm": 1.5,
+      "learning_rate": 0.0002,
+      "loss": 0.9111,
+      "step": 6518
+    },
+    {
+      "epoch": 5.206993006993007,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8841,
+      "step": 6519
+    },
+    {
+      "epoch": 5.207792207792208,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.0002,
+      "loss": 0.8918,
+      "step": 6520
+    },
+    {
+      "epoch": 5.208591408591409,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0002,
+      "loss": 0.8826,
+      "step": 6521
+    },
+    {
+      "epoch": 5.209390609390609,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.0002,
+      "loss": 0.8789,
+      "step": 6522
+    },
+    {
+      "epoch": 5.21018981018981,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0002,
+      "loss": 0.8871,
+      "step": 6523
+    },
+    {
+      "epoch": 5.210989010989011,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8988,
+      "step": 6524
+    },
+    {
+      "epoch": 5.211788211788212,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8847,
+      "step": 6525
+    },
+    {
+      "epoch": 5.212587412587412,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8888,
+      "step": 6526
+    },
+    {
+      "epoch": 5.213386613386613,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8966,
+      "step": 6527
+    },
+    {
+      "epoch": 5.214185814185814,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0002,
+      "loss": 0.8937,
+      "step": 6528
+    },
+    {
+      "epoch": 5.2149850149850145,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8972,
+      "step": 6529
+    },
+    {
+      "epoch": 5.215784215784216,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.0002,
+      "loss": 0.8936,
+      "step": 6530
+    },
+    {
+      "epoch": 5.216583416583417,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0002,
+      "loss": 0.8848,
+      "step": 6531
+    },
+    {
+      "epoch": 5.217382617382618,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.0002,
+      "loss": 0.9027,
+      "step": 6532
+    },
+    {
+      "epoch": 5.218181818181818,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0002,
+      "loss": 0.8831,
+      "step": 6533
+    },
+    {
+      "epoch": 5.218981018981019,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0002,
+      "loss": 0.8923,
+      "step": 6534
+    },
+    {
+      "epoch": 5.21978021978022,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.0002,
+      "loss": 0.8889,
+      "step": 6535
+    },
+    {
+      "epoch": 5.220579420579421,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0002,
+      "loss": 0.8904,
+      "step": 6536
+    },
+    {
+      "epoch": 5.221378621378621,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0002,
+      "loss": 0.8915,
+      "step": 6537
+    },
+    {
+      "epoch": 5.222177822177822,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8894,
+      "step": 6538
+    },
+    {
+      "epoch": 5.222977022977023,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0002,
+      "loss": 0.8804,
+      "step": 6539
+    },
+    {
+      "epoch": 5.223776223776224,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.0002,
+      "loss": 0.8873,
+      "step": 6540
+    },
+    {
+      "epoch": 5.224575424575424,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0002,
+      "loss": 0.8958,
+      "step": 6541
+    },
+    {
+      "epoch": 5.225374625374625,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8878,
+      "step": 6542
+    },
+    {
+      "epoch": 5.226173826173826,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8931,
+      "step": 6543
+    },
+    {
+      "epoch": 5.226973026973027,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8862,
+      "step": 6544
+    },
+    {
+      "epoch": 5.227772227772228,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0002,
+      "loss": 0.8958,
+      "step": 6545
+    },
+    {
+      "epoch": 5.228571428571429,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0002,
+      "loss": 0.8988,
+      "step": 6546
+    },
+    {
+      "epoch": 5.22937062937063,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8989,
+      "step": 6547
+    },
+    {
+      "epoch": 5.23016983016983,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0002,
+      "loss": 0.8983,
+      "step": 6548
+    },
+    {
+      "epoch": 5.230969030969031,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8962,
+      "step": 6549
+    },
+    {
+      "epoch": 5.231768231768232,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8931,
+      "step": 6550
+    },
+    {
+      "epoch": 5.232567432567433,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0002,
+      "loss": 0.8946,
+      "step": 6551
+    },
+    {
+      "epoch": 5.233366633366633,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8937,
+      "step": 6552
+    },
+    {
+      "epoch": 5.234165834165834,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8917,
+      "step": 6553
+    },
+    {
+      "epoch": 5.234965034965035,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0002,
+      "loss": 0.891,
+      "step": 6554
+    },
+    {
+      "epoch": 5.235764235764236,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8976,
+      "step": 6555
+    },
+    {
+      "epoch": 5.236563436563436,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0002,
+      "loss": 0.8955,
+      "step": 6556
+    },
+    {
+      "epoch": 5.237362637362637,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8879,
+      "step": 6557
+    },
+    {
+      "epoch": 5.2381618381618384,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0002,
+      "loss": 0.893,
+      "step": 6558
+    },
+    {
+      "epoch": 5.238961038961039,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0002,
+      "loss": 0.8892,
+      "step": 6559
+    },
+    {
+      "epoch": 5.23976023976024,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8897,
+      "step": 6560
+    },
+    {
+      "epoch": 5.240559440559441,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8875,
+      "step": 6561
+    },
+    {
+      "epoch": 5.241358641358642,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8958,
+      "step": 6562
+    },
+    {
+      "epoch": 5.242157842157842,
+      "grad_norm": 1.515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9095,
+      "step": 6563
+    },
+    {
+      "epoch": 5.242957042957043,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0002,
+      "loss": 0.8935,
+      "step": 6564
+    },
+    {
+      "epoch": 5.243756243756244,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0002,
+      "loss": 0.8952,
+      "step": 6565
+    },
+    {
+      "epoch": 5.244555444555445,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8975,
+      "step": 6566
+    },
+    {
+      "epoch": 5.245354645354645,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0002,
+      "loss": 0.89,
+      "step": 6567
+    },
+    {
+      "epoch": 5.246153846153846,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8883,
+      "step": 6568
+    },
+    {
+      "epoch": 5.246953046953047,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8854,
+      "step": 6569
+    },
+    {
+      "epoch": 5.247752247752247,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0002,
+      "loss": 0.8934,
+      "step": 6570
+    },
+    {
+      "epoch": 5.248551448551448,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8973,
+      "step": 6571
+    },
+    {
+      "epoch": 5.249350649350649,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8865,
+      "step": 6572
+    },
+    {
+      "epoch": 5.2501498501498505,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8918,
+      "step": 6573
+    },
+    {
+      "epoch": 5.250949050949051,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0002,
+      "loss": 0.889,
+      "step": 6574
+    },
+    {
+      "epoch": 5.251748251748252,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0002,
+      "loss": 0.8949,
+      "step": 6575
+    },
+    {
+      "epoch": 5.252547452547453,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8832,
+      "step": 6576
+    },
+    {
+      "epoch": 5.253346653346654,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0002,
+      "loss": 0.889,
+      "step": 6577
+    },
+    {
+      "epoch": 5.254145854145854,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0002,
+      "loss": 0.8919,
+      "step": 6578
+    },
+    {
+      "epoch": 5.254945054945055,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8868,
+      "step": 6579
+    },
+    {
+      "epoch": 5.255744255744256,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0002,
+      "loss": 0.8773,
+      "step": 6580
+    },
+    {
+      "epoch": 5.256543456543456,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0002,
+      "loss": 0.8865,
+      "step": 6581
+    },
+    {
+      "epoch": 5.257342657342657,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8948,
+      "step": 6582
+    },
+    {
+      "epoch": 5.258141858141858,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0002,
+      "loss": 0.8988,
+      "step": 6583
+    },
+    {
+      "epoch": 5.258941058941059,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8922,
+      "step": 6584
+    },
+    {
+      "epoch": 5.259740259740259,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0002,
+      "loss": 0.8944,
+      "step": 6585
+    },
+    {
+      "epoch": 5.26053946053946,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8788,
+      "step": 6586
+    },
+    {
+      "epoch": 5.2613386613386615,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8959,
+      "step": 6587
+    },
+    {
+      "epoch": 5.2621378621378625,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0002,
+      "loss": 0.8884,
+      "step": 6588
+    },
+    {
+      "epoch": 5.262937062937063,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8863,
+      "step": 6589
+    },
+    {
+      "epoch": 5.263736263736264,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8943,
+      "step": 6590
+    },
+    {
+      "epoch": 5.264535464535465,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8927,
+      "step": 6591
+    },
+    {
+      "epoch": 5.265334665334665,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8977,
+      "step": 6592
+    },
+    {
+      "epoch": 5.266133866133866,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8849,
+      "step": 6593
+    },
+    {
+      "epoch": 5.266933066933067,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0002,
+      "loss": 0.8878,
+      "step": 6594
+    },
+    {
+      "epoch": 5.267732267732268,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8878,
+      "step": 6595
+    },
+    {
+      "epoch": 5.268531468531468,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8893,
+      "step": 6596
+    },
+    {
+      "epoch": 5.269330669330669,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0002,
+      "loss": 0.8884,
+      "step": 6597
+    },
+    {
+      "epoch": 5.27012987012987,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0002,
+      "loss": 0.89,
+      "step": 6598
+    },
+    {
+      "epoch": 5.270929070929071,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8975,
+      "step": 6599
+    },
+    {
+      "epoch": 5.271728271728271,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0002,
+      "loss": 0.8743,
+      "step": 6600
+    },
+    {
+      "epoch": 5.2725274725274724,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8928,
+      "step": 6601
+    },
+    {
+      "epoch": 5.2733266733266735,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8939,
+      "step": 6602
+    },
+    {
+      "epoch": 5.2741258741258745,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0002,
+      "loss": 0.8968,
+      "step": 6603
+    },
+    {
+      "epoch": 5.274925074925075,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0002,
+      "loss": 0.8949,
+      "step": 6604
+    },
+    {
+      "epoch": 5.275724275724276,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0002,
+      "loss": 0.8903,
+      "step": 6605
+    },
+    {
+      "epoch": 5.276523476523477,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0002,
+      "loss": 0.8992,
+      "step": 6606
+    },
+    {
+      "epoch": 5.277322677322678,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8882,
+      "step": 6607
+    },
+    {
+      "epoch": 5.278121878121878,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.0002,
+      "loss": 0.892,
+      "step": 6608
+    },
+    {
+      "epoch": 5.278921078921079,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8995,
+      "step": 6609
+    },
+    {
+      "epoch": 5.27972027972028,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.0002,
+      "loss": 0.8967,
+      "step": 6610
+    },
+    {
+      "epoch": 5.28051948051948,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0002,
+      "loss": 0.8948,
+      "step": 6611
+    },
+    {
+      "epoch": 5.281318681318681,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.0002,
+      "loss": 0.8972,
+      "step": 6612
+    },
+    {
+      "epoch": 5.282117882117882,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0002,
+      "loss": 0.8946,
+      "step": 6613
+    },
+    {
+      "epoch": 5.282917082917083,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8835,
+      "step": 6614
+    },
+    {
+      "epoch": 5.283716283716283,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8843,
+      "step": 6615
+    },
+    {
+      "epoch": 5.2845154845154845,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0002,
+      "loss": 0.8906,
+      "step": 6616
+    },
+    {
+      "epoch": 5.2853146853146855,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0002,
+      "loss": 0.9004,
+      "step": 6617
+    },
+    {
+      "epoch": 5.2861138861138866,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.0002,
+      "loss": 0.8876,
+      "step": 6618
+    },
+    {
+      "epoch": 5.286913086913087,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0002,
+      "loss": 0.884,
+      "step": 6619
+    },
+    {
+      "epoch": 5.287712287712288,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8924,
+      "step": 6620
+    },
+    {
+      "epoch": 5.288511488511489,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.0002,
+      "loss": 0.8955,
+      "step": 6621
+    },
+    {
+      "epoch": 5.289310689310689,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.0002,
+      "loss": 0.8994,
+      "step": 6622
+    },
+    {
+      "epoch": 5.29010989010989,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0002,
+      "loss": 0.8954,
+      "step": 6623
+    },
+    {
+      "epoch": 5.290909090909091,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0002,
+      "loss": 0.8923,
+      "step": 6624
+    },
+    {
+      "epoch": 5.291708291708292,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0002,
+      "loss": 0.895,
+      "step": 6625
+    },
+    {
+      "epoch": 5.292507492507492,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8998,
+      "step": 6626
+    },
+    {
+      "epoch": 5.293306693306693,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8888,
+      "step": 6627
+    },
+    {
+      "epoch": 5.294105894105894,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0002,
+      "loss": 0.8933,
+      "step": 6628
+    },
+    {
+      "epoch": 5.294905094905095,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0002,
+      "loss": 0.8942,
+      "step": 6629
+    },
+    {
+      "epoch": 5.2957042957042955,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0002,
+      "loss": 0.9001,
+      "step": 6630
+    },
+    {
+      "epoch": 5.2965034965034965,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0002,
+      "loss": 0.8833,
+      "step": 6631
+    },
+    {
+      "epoch": 5.2973026973026975,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.0002,
+      "loss": 0.8903,
+      "step": 6632
+    },
+    {
+      "epoch": 5.298101898101898,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.0002,
+      "loss": 0.8916,
+      "step": 6633
+    },
+    {
+      "epoch": 5.298901098901099,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.0002,
+      "loss": 0.8841,
+      "step": 6634
+    },
+    {
+      "epoch": 5.2997002997003,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8818,
+      "step": 6635
+    },
+    {
+      "epoch": 5.300499500499501,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0002,
+      "loss": 0.8892,
+      "step": 6636
+    },
+    {
+      "epoch": 5.301298701298701,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8853,
+      "step": 6637
+    },
+    {
+      "epoch": 5.302097902097902,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0002,
+      "loss": 0.8811,
+      "step": 6638
+    },
+    {
+      "epoch": 5.302897102897103,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0002,
+      "loss": 0.8978,
+      "step": 6639
+    },
+    {
+      "epoch": 5.303696303696304,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0002,
+      "loss": 0.8955,
+      "step": 6640
+    },
+    {
+      "epoch": 5.304495504495504,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.0002,
+      "loss": 0.8834,
+      "step": 6641
+    },
+    {
+      "epoch": 5.305294705294705,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8907,
+      "step": 6642
+    },
+    {
+      "epoch": 5.306093906093906,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8937,
+      "step": 6643
+    },
+    {
+      "epoch": 5.3068931068931064,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0002,
+      "loss": 0.8915,
+      "step": 6644
+    },
+    {
+      "epoch": 5.3076923076923075,
+      "grad_norm": 1.671875,
+      "learning_rate": 0.0002,
+      "loss": 0.8992,
+      "step": 6645
+    },
+    {
+      "epoch": 5.3084915084915085,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8916,
+      "step": 6646
+    },
+    {
+      "epoch": 5.30929070929071,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0002,
+      "loss": 0.8897,
+      "step": 6647
+    },
+    {
+      "epoch": 5.31008991008991,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.0002,
+      "loss": 0.8881,
+      "step": 6648
+    },
+    {
+      "epoch": 5.310889110889111,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8965,
+      "step": 6649
+    },
+    {
+      "epoch": 5.311688311688312,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.0002,
+      "loss": 0.8894,
+      "step": 6650
+    },
+    {
+      "epoch": 5.312487512487513,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.0002,
+      "loss": 0.8928,
+      "step": 6651
+    },
+    {
+      "epoch": 5.313286713286713,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0002,
+      "loss": 0.8913,
+      "step": 6652
+    },
+    {
+      "epoch": 5.314085914085914,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0002,
+      "loss": 0.8898,
+      "step": 6653
+    },
+    {
+      "epoch": 5.314885114885115,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.0002,
+      "loss": 0.898,
+      "step": 6654
+    },
+    {
+      "epoch": 5.315684315684316,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8916,
+      "step": 6655
+    },
+    {
+      "epoch": 5.316483516483516,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.0002,
+      "loss": 0.8787,
+      "step": 6656
+    },
+    {
+      "epoch": 5.317282717282717,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9021,
+      "step": 6657
+    },
+    {
+      "epoch": 5.318081918081918,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0002,
+      "loss": 0.8867,
+      "step": 6658
+    },
+    {
+      "epoch": 5.3188811188811185,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8991,
+      "step": 6659
+    },
+    {
+      "epoch": 5.3196803196803195,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8926,
+      "step": 6660
+    },
+    {
+      "epoch": 5.3204795204795206,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0002,
+      "loss": 0.884,
+      "step": 6661
+    },
+    {
+      "epoch": 5.321278721278722,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.0002,
+      "loss": 0.8903,
+      "step": 6662
+    },
+    {
+      "epoch": 5.322077922077922,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8956,
+      "step": 6663
+    },
+    {
+      "epoch": 5.322877122877123,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.0002,
+      "loss": 0.8938,
+      "step": 6664
+    },
+    {
+      "epoch": 5.323676323676324,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0002,
+      "loss": 0.8936,
+      "step": 6665
+    },
+    {
+      "epoch": 5.324475524475525,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0002,
+      "loss": 0.8966,
+      "step": 6666
+    },
+    {
+      "epoch": 5.325274725274725,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8935,
+      "step": 6667
+    },
+    {
+      "epoch": 5.326073926073926,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0002,
+      "loss": 0.8852,
+      "step": 6668
+    },
+    {
+      "epoch": 5.326873126873127,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0002,
+      "loss": 0.9009,
+      "step": 6669
+    },
+    {
+      "epoch": 5.327672327672328,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8911,
+      "step": 6670
+    },
+    {
+      "epoch": 5.328471528471528,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0002,
+      "loss": 0.9043,
+      "step": 6671
+    },
+    {
+      "epoch": 5.329270729270729,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9011,
+      "step": 6672
+    },
+    {
+      "epoch": 5.33006993006993,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.0002,
+      "loss": 0.895,
+      "step": 6673
+    },
+    {
+      "epoch": 5.3308691308691305,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0002,
+      "loss": 0.8982,
+      "step": 6674
+    },
+    {
+      "epoch": 5.3316683316683315,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0002,
+      "loss": 0.8965,
+      "step": 6675
+    },
+    {
+      "epoch": 5.332467532467533,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0002,
+      "loss": 0.8936,
+      "step": 6676
+    },
+    {
+      "epoch": 5.333266733266734,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8931,
+      "step": 6677
+    },
+    {
+      "epoch": 5.334065934065934,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0002,
+      "loss": 0.8851,
+      "step": 6678
+    },
+    {
+      "epoch": 5.334865134865135,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.0002,
+      "loss": 0.9014,
+      "step": 6679
+    },
+    {
+      "epoch": 5.335664335664336,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8862,
+      "step": 6680
+    },
+    {
+      "epoch": 5.336463536463537,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0002,
+      "loss": 0.898,
+      "step": 6681
+    },
+    {
+      "epoch": 5.337262737262737,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0002,
+      "loss": 0.889,
+      "step": 6682
+    },
+    {
+      "epoch": 5.338061938061938,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0002,
+      "loss": 0.8931,
+      "step": 6683
+    },
+    {
+      "epoch": 5.338861138861139,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8895,
+      "step": 6684
+    },
+    {
+      "epoch": 5.339660339660339,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0002,
+      "loss": 0.8924,
+      "step": 6685
+    },
+    {
+      "epoch": 5.34045954045954,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8925,
+      "step": 6686
+    },
+    {
+      "epoch": 5.341258741258741,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.0002,
+      "loss": 0.898,
+      "step": 6687
+    },
+    {
+      "epoch": 5.342057942057942,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8926,
+      "step": 6688
+    },
+    {
+      "epoch": 5.3428571428571425,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0002,
+      "loss": 0.8971,
+      "step": 6689
+    },
+    {
+      "epoch": 5.343656343656344,
+      "grad_norm": 0.375,
+      "learning_rate": 0.0002,
+      "loss": 0.8845,
+      "step": 6690
+    },
+    {
+      "epoch": 5.344455544455545,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8873,
+      "step": 6691
+    },
+    {
+      "epoch": 5.345254745254746,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8924,
+      "step": 6692
+    },
+    {
+      "epoch": 5.346053946053946,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.0002,
+      "loss": 0.8981,
+      "step": 6693
+    },
+    {
+      "epoch": 5.346853146853147,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8955,
+      "step": 6694
+    },
+    {
+      "epoch": 5.347652347652348,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0002,
+      "loss": 0.8911,
+      "step": 6695
+    },
+    {
+      "epoch": 5.348451548451548,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8919,
+      "step": 6696
+    },
+    {
+      "epoch": 5.349250749250749,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8869,
+      "step": 6697
+    },
+    {
+      "epoch": 5.35004995004995,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.0002,
+      "loss": 0.8928,
+      "step": 6698
+    },
+    {
+      "epoch": 5.350849150849151,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0002,
+      "loss": 0.8915,
+      "step": 6699
+    },
+    {
+      "epoch": 5.351648351648351,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0002,
+      "loss": 0.899,
+      "step": 6700
+    },
+    {
+      "epoch": 5.352447552447552,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.0002,
+      "loss": 0.8892,
+      "step": 6701
+    },
+    {
+      "epoch": 5.353246753246753,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0002,
+      "loss": 0.8881,
+      "step": 6702
+    },
+    {
+      "epoch": 5.354045954045954,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8898,
+      "step": 6703
+    },
+    {
+      "epoch": 5.3548451548451546,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.0002,
+      "loss": 0.9075,
+      "step": 6704
+    },
+    {
+      "epoch": 5.355644355644356,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.0002,
+      "loss": 0.8964,
+      "step": 6705
+    },
+    {
+      "epoch": 5.356443556443557,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0002,
+      "loss": 0.9011,
+      "step": 6706
+    },
+    {
+      "epoch": 5.357242757242757,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0002,
+      "loss": 0.8882,
+      "step": 6707
+    },
+    {
+      "epoch": 5.358041958041958,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0002,
+      "loss": 0.894,
+      "step": 6708
+    },
+    {
+      "epoch": 5.358841158841159,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8906,
+      "step": 6709
+    },
+    {
+      "epoch": 5.35964035964036,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0002,
+      "loss": 0.8881,
+      "step": 6710
+    },
+    {
+      "epoch": 5.36043956043956,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8852,
+      "step": 6711
+    },
+    {
+      "epoch": 5.361238761238761,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0002,
+      "loss": 0.8903,
+      "step": 6712
+    },
+    {
+      "epoch": 5.362037962037962,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0002,
+      "loss": 0.8909,
+      "step": 6713
+    },
+    {
+      "epoch": 5.362837162837163,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8945,
+      "step": 6714
+    },
+    {
+      "epoch": 5.363636363636363,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0002,
+      "loss": 0.8855,
+      "step": 6715
+    },
+    {
+      "epoch": 5.364435564435564,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8902,
+      "step": 6716
+    },
+    {
+      "epoch": 5.365234765234765,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0002,
+      "loss": 0.8922,
+      "step": 6717
+    },
+    {
+      "epoch": 5.366033966033966,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0002,
+      "loss": 0.893,
+      "step": 6718
+    },
+    {
+      "epoch": 5.366833166833167,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8884,
+      "step": 6719
+    },
+    {
+      "epoch": 5.367632367632368,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8872,
+      "step": 6720
+    },
+    {
+      "epoch": 5.368431568431569,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0002,
+      "loss": 0.8979,
+      "step": 6721
+    },
+    {
+      "epoch": 5.36923076923077,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8851,
+      "step": 6722
+    },
+    {
+      "epoch": 5.37002997002997,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0002,
+      "loss": 0.9013,
+      "step": 6723
+    },
+    {
+      "epoch": 5.370829170829171,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0002,
+      "loss": 0.8894,
+      "step": 6724
+    },
+    {
+      "epoch": 5.371628371628372,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0002,
+      "loss": 0.8891,
+      "step": 6725
+    },
+    {
+      "epoch": 5.372427572427572,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0002,
+      "loss": 0.8944,
+      "step": 6726
+    },
+    {
+      "epoch": 5.373226773226773,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.0002,
+      "loss": 0.9317,
+      "step": 6727
+    },
+    {
+      "epoch": 5.374025974025974,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0002,
+      "loss": 0.8907,
+      "step": 6728
+    },
+    {
+      "epoch": 5.374825174825175,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8968,
+      "step": 6729
+    },
+    {
+      "epoch": 5.375624375624375,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8961,
+      "step": 6730
+    },
+    {
+      "epoch": 5.376423576423576,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8882,
+      "step": 6731
+    },
+    {
+      "epoch": 5.377222777222777,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8958,
+      "step": 6732
+    },
+    {
+      "epoch": 5.3780219780219785,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8878,
+      "step": 6733
+    },
+    {
+      "epoch": 5.378821178821179,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8869,
+      "step": 6734
+    },
+    {
+      "epoch": 5.37962037962038,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8884,
+      "step": 6735
+    },
+    {
+      "epoch": 5.380419580419581,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0002,
+      "loss": 0.902,
+      "step": 6736
+    },
+    {
+      "epoch": 5.381218781218781,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8983,
+      "step": 6737
+    },
+    {
+      "epoch": 5.382017982017982,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0002,
+      "loss": 0.8956,
+      "step": 6738
+    },
+    {
+      "epoch": 5.382817182817183,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8985,
+      "step": 6739
+    },
+    {
+      "epoch": 5.383616383616384,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8918,
+      "step": 6740
+    },
+    {
+      "epoch": 5.384415584415584,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8927,
+      "step": 6741
+    },
+    {
+      "epoch": 5.385214785214785,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.0002,
+      "loss": 0.9122,
+      "step": 6742
+    },
+    {
+      "epoch": 5.386013986013986,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0002,
+      "loss": 0.8915,
+      "step": 6743
+    },
+    {
+      "epoch": 5.386813186813187,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9029,
+      "step": 6744
+    },
+    {
+      "epoch": 5.387612387612387,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8899,
+      "step": 6745
+    },
+    {
+      "epoch": 5.388411588411588,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0002,
+      "loss": 0.8885,
+      "step": 6746
+    },
+    {
+      "epoch": 5.389210789210789,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0002,
+      "loss": 0.8947,
+      "step": 6747
+    },
+    {
+      "epoch": 5.39000999000999,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0002,
+      "loss": 0.8932,
+      "step": 6748
+    },
+    {
+      "epoch": 5.390809190809191,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8903,
+      "step": 6749
+    },
+    {
+      "epoch": 5.391608391608392,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8929,
+      "step": 6750
+    },
+    {
+      "epoch": 5.392407592407593,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0002,
+      "loss": 0.9156,
+      "step": 6751
+    },
+    {
+      "epoch": 5.393206793206793,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8943,
+      "step": 6752
+    },
+    {
+      "epoch": 5.394005994005994,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8872,
+      "step": 6753
+    },
+    {
+      "epoch": 5.394805194805195,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8829,
+      "step": 6754
+    },
+    {
+      "epoch": 5.395604395604396,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.0002,
+      "loss": 0.8922,
+      "step": 6755
+    },
+    {
+      "epoch": 5.396403596403596,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8943,
+      "step": 6756
+    },
+    {
+      "epoch": 5.397202797202797,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0002,
+      "loss": 0.8908,
+      "step": 6757
+    },
+    {
+      "epoch": 5.398001998001998,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8893,
+      "step": 6758
+    },
+    {
+      "epoch": 5.398801198801198,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0002,
+      "loss": 0.8943,
+      "step": 6759
+    },
+    {
+      "epoch": 5.399600399600399,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0002,
+      "loss": 0.8933,
+      "step": 6760
+    },
+    {
+      "epoch": 5.4003996003996,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0002,
+      "loss": 0.8864,
+      "step": 6761
+    },
+    {
+      "epoch": 5.4011988011988015,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8914,
+      "step": 6762
+    },
+    {
+      "epoch": 5.401998001998002,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.0002,
+      "loss": 0.8995,
+      "step": 6763
+    },
+    {
+      "epoch": 5.402797202797203,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0002,
+      "loss": 0.9005,
+      "step": 6764
+    },
+    {
+      "epoch": 5.403596403596404,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0002,
+      "loss": 0.8857,
+      "step": 6765
+    },
+    {
+      "epoch": 5.404395604395605,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8926,
+      "step": 6766
+    },
+    {
+      "epoch": 5.405194805194805,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8977,
+      "step": 6767
+    },
+    {
+      "epoch": 5.405994005994006,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8762,
+      "step": 6768
+    },
+    {
+      "epoch": 5.406793206793207,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0002,
+      "loss": 0.8932,
+      "step": 6769
+    },
+    {
+      "epoch": 5.407592407592408,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0002,
+      "loss": 0.89,
+      "step": 6770
+    },
+    {
+      "epoch": 5.408391608391608,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8849,
+      "step": 6771
+    },
+    {
+      "epoch": 5.409190809190809,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8958,
+      "step": 6772
+    },
+    {
+      "epoch": 5.40999000999001,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8937,
+      "step": 6773
+    },
+    {
+      "epoch": 5.41078921078921,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0002,
+      "loss": 0.9025,
+      "step": 6774
+    },
+    {
+      "epoch": 5.411588411588411,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0002,
+      "loss": 0.8903,
+      "step": 6775
+    },
+    {
+      "epoch": 5.4123876123876125,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.0002,
+      "loss": 0.8853,
+      "step": 6776
+    },
+    {
+      "epoch": 5.4131868131868135,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8976,
+      "step": 6777
+    },
+    {
+      "epoch": 5.413986013986014,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8862,
+      "step": 6778
+    },
+    {
+      "epoch": 5.414785214785215,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0002,
+      "loss": 0.8987,
+      "step": 6779
+    },
+    {
+      "epoch": 5.415584415584416,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0002,
+      "loss": 0.8959,
+      "step": 6780
+    },
+    {
+      "epoch": 5.416383616383617,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8994,
+      "step": 6781
+    },
+    {
+      "epoch": 5.417182817182817,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8877,
+      "step": 6782
+    },
+    {
+      "epoch": 5.417982017982018,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0002,
+      "loss": 0.8837,
+      "step": 6783
+    },
+    {
+      "epoch": 5.418781218781219,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0002,
+      "loss": 0.8913,
+      "step": 6784
+    },
+    {
+      "epoch": 5.41958041958042,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0002,
+      "loss": 0.891,
+      "step": 6785
+    },
+    {
+      "epoch": 5.42037962037962,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0002,
+      "loss": 0.8858,
+      "step": 6786
+    },
+    {
+      "epoch": 5.421178821178821,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0002,
+      "loss": 0.8913,
+      "step": 6787
+    },
+    {
+      "epoch": 5.421978021978022,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0002,
+      "loss": 0.886,
+      "step": 6788
+    },
+    {
+      "epoch": 5.422777222777222,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0002,
+      "loss": 0.889,
+      "step": 6789
+    },
+    {
+      "epoch": 5.4235764235764234,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0002,
+      "loss": 0.885,
+      "step": 6790
+    },
+    {
+      "epoch": 5.4243756243756245,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0002,
+      "loss": 0.9004,
+      "step": 6791
+    },
+    {
+      "epoch": 5.4251748251748255,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0002,
+      "loss": 0.8856,
+      "step": 6792
+    },
+    {
+      "epoch": 5.425974025974026,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8926,
+      "step": 6793
+    },
+    {
+      "epoch": 5.426773226773227,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8949,
+      "step": 6794
+    },
+    {
+      "epoch": 5.427572427572428,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8855,
+      "step": 6795
+    },
+    {
+      "epoch": 5.428371628371629,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8831,
+      "step": 6796
+    },
+    {
+      "epoch": 5.429170829170829,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8989,
+      "step": 6797
+    },
+    {
+      "epoch": 5.42997002997003,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.0002,
+      "loss": 0.8901,
+      "step": 6798
+    },
+    {
+      "epoch": 5.430769230769231,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8979,
+      "step": 6799
+    },
+    {
+      "epoch": 5.431568431568431,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.0002,
+      "loss": 0.8897,
+      "step": 6800
+    },
+    {
+      "epoch": 5.432367632367632,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8912,
+      "step": 6801
+    },
+    {
+      "epoch": 5.433166833166833,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0002,
+      "loss": 0.8885,
+      "step": 6802
+    },
+    {
+      "epoch": 5.433966033966034,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8939,
+      "step": 6803
+    },
+    {
+      "epoch": 5.434765234765234,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0002,
+      "loss": 0.8871,
+      "step": 6804
+    },
+    {
+      "epoch": 5.4355644355644355,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0002,
+      "loss": 0.8856,
+      "step": 6805
+    },
+    {
+      "epoch": 5.4363636363636365,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8972,
+      "step": 6806
+    },
+    {
+      "epoch": 5.4371628371628375,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0002,
+      "loss": 0.8913,
+      "step": 6807
+    },
+    {
+      "epoch": 5.437962037962038,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8938,
+      "step": 6808
+    },
+    {
+      "epoch": 5.438761238761239,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8881,
+      "step": 6809
+    },
+    {
+      "epoch": 5.43956043956044,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0002,
+      "loss": 0.8927,
+      "step": 6810
+    },
+    {
+      "epoch": 5.44035964035964,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8976,
+      "step": 6811
+    },
+    {
+      "epoch": 5.441158841158841,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0002,
+      "loss": 0.9021,
+      "step": 6812
+    },
+    {
+      "epoch": 5.441958041958042,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8886,
+      "step": 6813
+    },
+    {
+      "epoch": 5.442757242757243,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8891,
+      "step": 6814
+    },
+    {
+      "epoch": 5.443556443556443,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8868,
+      "step": 6815
+    },
+    {
+      "epoch": 5.444355644355644,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8914,
+      "step": 6816
+    },
+    {
+      "epoch": 5.445154845154845,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0002,
+      "loss": 0.8811,
+      "step": 6817
+    },
+    {
+      "epoch": 5.445954045954046,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8949,
+      "step": 6818
+    },
+    {
+      "epoch": 5.4467532467532465,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8921,
+      "step": 6819
+    },
+    {
+      "epoch": 5.4475524475524475,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8956,
+      "step": 6820
+    },
+    {
+      "epoch": 5.4483516483516485,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0002,
+      "loss": 0.8955,
+      "step": 6821
+    },
+    {
+      "epoch": 5.449150849150849,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0002,
+      "loss": 0.8878,
+      "step": 6822
+    },
+    {
+      "epoch": 5.44995004995005,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0002,
+      "loss": 0.8918,
+      "step": 6823
+    },
+    {
+      "epoch": 5.450749250749251,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0002,
+      "loss": 0.8902,
+      "step": 6824
+    },
+    {
+      "epoch": 5.451548451548452,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0002,
+      "loss": 0.9031,
+      "step": 6825
+    },
+    {
+      "epoch": 5.452347652347652,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0002,
+      "loss": 0.8918,
+      "step": 6826
+    },
+    {
+      "epoch": 5.453146853146853,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8903,
+      "step": 6827
+    },
+    {
+      "epoch": 5.453946053946054,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0002,
+      "loss": 0.8887,
+      "step": 6828
+    },
+    {
+      "epoch": 5.454745254745255,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9052,
+      "step": 6829
+    },
+    {
+      "epoch": 5.455544455544455,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8991,
+      "step": 6830
+    },
+    {
+      "epoch": 5.456343656343656,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8832,
+      "step": 6831
+    },
+    {
+      "epoch": 5.457142857142857,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8803,
+      "step": 6832
+    },
+    {
+      "epoch": 5.457942057942058,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0002,
+      "loss": 0.8907,
+      "step": 6833
+    },
+    {
+      "epoch": 5.4587412587412585,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8955,
+      "step": 6834
+    },
+    {
+      "epoch": 5.4595404595404595,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8861,
+      "step": 6835
+    },
+    {
+      "epoch": 5.460339660339661,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9324,
+      "step": 6836
+    },
+    {
+      "epoch": 5.461138861138862,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8798,
+      "step": 6837
+    },
+    {
+      "epoch": 5.461938061938062,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0002,
+      "loss": 0.8901,
+      "step": 6838
+    },
+    {
+      "epoch": 5.462737262737263,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8927,
+      "step": 6839
+    },
+    {
+      "epoch": 5.463536463536464,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8958,
+      "step": 6840
+    },
+    {
+      "epoch": 5.464335664335664,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8928,
+      "step": 6841
+    },
+    {
+      "epoch": 5.465134865134865,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8898,
+      "step": 6842
+    },
+    {
+      "epoch": 5.465934065934066,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0002,
+      "loss": 0.8969,
+      "step": 6843
+    },
+    {
+      "epoch": 5.466733266733267,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8903,
+      "step": 6844
+    },
+    {
+      "epoch": 5.467532467532467,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.0002,
+      "loss": 0.8993,
+      "step": 6845
+    },
+    {
+      "epoch": 5.468331668331668,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0002,
+      "loss": 0.8964,
+      "step": 6846
+    },
+    {
+      "epoch": 5.469130869130869,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8928,
+      "step": 6847
+    },
+    {
+      "epoch": 5.46993006993007,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8897,
+      "step": 6848
+    },
+    {
+      "epoch": 5.4707292707292705,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8929,
+      "step": 6849
+    },
+    {
+      "epoch": 5.4715284715284715,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0002,
+      "loss": 0.8953,
+      "step": 6850
+    },
+    {
+      "epoch": 5.472327672327673,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8956,
+      "step": 6851
+    },
+    {
+      "epoch": 5.473126873126873,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8982,
+      "step": 6852
+    },
+    {
+      "epoch": 5.473926073926074,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8908,
+      "step": 6853
+    },
+    {
+      "epoch": 5.474725274725275,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.0002,
+      "loss": 0.892,
+      "step": 6854
+    },
+    {
+      "epoch": 5.475524475524476,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8888,
+      "step": 6855
+    },
+    {
+      "epoch": 5.476323676323676,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8918,
+      "step": 6856
+    },
+    {
+      "epoch": 5.477122877122877,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8929,
+      "step": 6857
+    },
+    {
+      "epoch": 5.477922077922078,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8909,
+      "step": 6858
+    },
+    {
+      "epoch": 5.478721278721279,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8958,
+      "step": 6859
+    },
+    {
+      "epoch": 5.479520479520479,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0002,
+      "loss": 0.9042,
+      "step": 6860
+    },
+    {
+      "epoch": 5.48031968031968,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0002,
+      "loss": 0.8967,
+      "step": 6861
+    },
+    {
+      "epoch": 5.481118881118881,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8948,
+      "step": 6862
+    },
+    {
+      "epoch": 5.4819180819180815,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8932,
+      "step": 6863
+    },
+    {
+      "epoch": 5.4827172827172825,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0002,
+      "loss": 0.8935,
+      "step": 6864
+    },
+    {
+      "epoch": 5.483516483516484,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0002,
+      "loss": 0.8944,
+      "step": 6865
+    },
+    {
+      "epoch": 5.484315684315685,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8958,
+      "step": 6866
+    },
+    {
+      "epoch": 5.485114885114885,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8881,
+      "step": 6867
+    },
+    {
+      "epoch": 5.485914085914086,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0002,
+      "loss": 0.8823,
+      "step": 6868
+    },
+    {
+      "epoch": 5.486713286713287,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8961,
+      "step": 6869
+    },
+    {
+      "epoch": 5.487512487512488,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8934,
+      "step": 6870
+    },
+    {
+      "epoch": 5.488311688311688,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8913,
+      "step": 6871
+    },
+    {
+      "epoch": 5.489110889110889,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8861,
+      "step": 6872
+    },
+    {
+      "epoch": 5.48991008991009,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8934,
+      "step": 6873
+    },
+    {
+      "epoch": 5.49070929070929,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.0002,
+      "loss": 0.8953,
+      "step": 6874
+    },
+    {
+      "epoch": 5.491508491508491,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0002,
+      "loss": 0.8866,
+      "step": 6875
+    },
+    {
+      "epoch": 5.492307692307692,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8965,
+      "step": 6876
+    },
+    {
+      "epoch": 5.493106893106893,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0002,
+      "loss": 0.8867,
+      "step": 6877
+    },
+    {
+      "epoch": 5.4939060939060935,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.0002,
+      "loss": 0.8907,
+      "step": 6878
+    },
+    {
+      "epoch": 5.494705294705295,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8984,
+      "step": 6879
+    },
+    {
+      "epoch": 5.495504495504496,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.0002,
+      "loss": 0.9021,
+      "step": 6880
+    },
+    {
+      "epoch": 5.496303696303697,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0002,
+      "loss": 0.8929,
+      "step": 6881
+    },
+    {
+      "epoch": 5.497102897102897,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0002,
+      "loss": 0.8929,
+      "step": 6882
+    },
+    {
+      "epoch": 5.497902097902098,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8986,
+      "step": 6883
+    },
+    {
+      "epoch": 5.498701298701299,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0002,
+      "loss": 0.8844,
+      "step": 6884
+    },
+    {
+      "epoch": 5.4995004995005,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.0002,
+      "loss": 0.8861,
+      "step": 6885
+    },
+    {
+      "epoch": 5.5002997002997,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0002,
+      "loss": 0.8807,
+      "step": 6886
+    },
+    {
+      "epoch": 5.501098901098901,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8958,
+      "step": 6887
+    },
+    {
+      "epoch": 5.501898101898102,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8919,
+      "step": 6888
+    },
+    {
+      "epoch": 5.502697302697303,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8834,
+      "step": 6889
+    },
+    {
+      "epoch": 5.503496503496503,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8895,
+      "step": 6890
+    },
+    {
+      "epoch": 5.504295704295704,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8928,
+      "step": 6891
+    },
+    {
+      "epoch": 5.505094905094905,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0002,
+      "loss": 0.8928,
+      "step": 6892
+    },
+    {
+      "epoch": 5.5058941058941056,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0002,
+      "loss": 0.8862,
+      "step": 6893
+    },
+    {
+      "epoch": 5.506693306693307,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8969,
+      "step": 6894
+    },
+    {
+      "epoch": 5.507492507492508,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0002,
+      "loss": 0.8861,
+      "step": 6895
+    },
+    {
+      "epoch": 5.508291708291709,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8964,
+      "step": 6896
+    },
+    {
+      "epoch": 5.509090909090909,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0002,
+      "loss": 0.8907,
+      "step": 6897
+    },
+    {
+      "epoch": 5.50989010989011,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0002,
+      "loss": 0.8913,
+      "step": 6898
+    },
+    {
+      "epoch": 5.510689310689311,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0002,
+      "loss": 0.894,
+      "step": 6899
+    },
+    {
+      "epoch": 5.511488511488512,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8884,
+      "step": 6900
+    },
+    {
+      "epoch": 5.512287712287712,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8926,
+      "step": 6901
+    },
+    {
+      "epoch": 5.513086913086913,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8935,
+      "step": 6902
+    },
+    {
+      "epoch": 5.513886113886114,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0002,
+      "loss": 0.8854,
+      "step": 6903
+    },
+    {
+      "epoch": 5.514685314685314,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0002,
+      "loss": 0.8904,
+      "step": 6904
+    },
+    {
+      "epoch": 5.515484515484515,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8967,
+      "step": 6905
+    },
+    {
+      "epoch": 5.516283716283716,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0002,
+      "loss": 0.8847,
+      "step": 6906
+    },
+    {
+      "epoch": 5.517082917082917,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8808,
+      "step": 6907
+    },
+    {
+      "epoch": 5.517882117882118,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8885,
+      "step": 6908
+    },
+    {
+      "epoch": 5.518681318681319,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0002,
+      "loss": 0.887,
+      "step": 6909
+    },
+    {
+      "epoch": 5.51948051948052,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0002,
+      "loss": 0.8913,
+      "step": 6910
+    },
+    {
+      "epoch": 5.520279720279721,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0002,
+      "loss": 0.8923,
+      "step": 6911
+    },
+    {
+      "epoch": 5.521078921078921,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8946,
+      "step": 6912
+    },
+    {
+      "epoch": 5.521878121878122,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.0002,
+      "loss": 0.8938,
+      "step": 6913
+    },
+    {
+      "epoch": 5.522677322677323,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0002,
+      "loss": 0.8968,
+      "step": 6914
+    },
+    {
+      "epoch": 5.523476523476523,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8932,
+      "step": 6915
+    },
+    {
+      "epoch": 5.524275724275724,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8892,
+      "step": 6916
+    },
+    {
+      "epoch": 5.525074925074925,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0002,
+      "loss": 0.8814,
+      "step": 6917
+    },
+    {
+      "epoch": 5.525874125874126,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.0002,
+      "loss": 0.8858,
+      "step": 6918
+    },
+    {
+      "epoch": 5.526673326673326,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0002,
+      "loss": 0.8942,
+      "step": 6919
+    },
+    {
+      "epoch": 5.527472527472527,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0002,
+      "loss": 0.8871,
+      "step": 6920
+    },
+    {
+      "epoch": 5.528271728271728,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0002,
+      "loss": 0.8831,
+      "step": 6921
+    },
+    {
+      "epoch": 5.5290709290709295,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8926,
+      "step": 6922
+    },
+    {
+      "epoch": 5.52987012987013,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8939,
+      "step": 6923
+    },
+    {
+      "epoch": 5.530669330669331,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8896,
+      "step": 6924
+    },
+    {
+      "epoch": 5.531468531468532,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8943,
+      "step": 6925
+    },
+    {
+      "epoch": 5.532267732267732,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8885,
+      "step": 6926
+    },
+    {
+      "epoch": 5.533066933066933,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0002,
+      "loss": 0.903,
+      "step": 6927
+    },
+    {
+      "epoch": 5.533866133866134,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0002,
+      "loss": 0.8951,
+      "step": 6928
+    },
+    {
+      "epoch": 5.534665334665335,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8808,
+      "step": 6929
+    },
+    {
+      "epoch": 5.535464535464535,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8898,
+      "step": 6930
+    },
+    {
+      "epoch": 5.536263736263736,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8984,
+      "step": 6931
+    },
+    {
+      "epoch": 5.537062937062937,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0002,
+      "loss": 0.8926,
+      "step": 6932
+    },
+    {
+      "epoch": 5.537862137862138,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0002,
+      "loss": 0.893,
+      "step": 6933
+    },
+    {
+      "epoch": 5.538661338661338,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0002,
+      "loss": 0.8982,
+      "step": 6934
+    },
+    {
+      "epoch": 5.539460539460539,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8949,
+      "step": 6935
+    },
+    {
+      "epoch": 5.54025974025974,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0002,
+      "loss": 0.8901,
+      "step": 6936
+    },
+    {
+      "epoch": 5.541058941058941,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8932,
+      "step": 6937
+    },
+    {
+      "epoch": 5.541858141858142,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.0002,
+      "loss": 0.8812,
+      "step": 6938
+    },
+    {
+      "epoch": 5.542657342657343,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0002,
+      "loss": 0.881,
+      "step": 6939
+    },
+    {
+      "epoch": 5.543456543456544,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8903,
+      "step": 6940
+    },
+    {
+      "epoch": 5.544255744255745,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0002,
+      "loss": 0.895,
+      "step": 6941
+    },
+    {
+      "epoch": 5.545054945054945,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8986,
+      "step": 6942
+    },
+    {
+      "epoch": 5.545854145854146,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8872,
+      "step": 6943
+    },
+    {
+      "epoch": 5.546653346653347,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0002,
+      "loss": 0.8931,
+      "step": 6944
+    },
+    {
+      "epoch": 5.547452547452547,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0002,
+      "loss": 0.8817,
+      "step": 6945
+    },
+    {
+      "epoch": 5.548251748251748,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.0002,
+      "loss": 0.8876,
+      "step": 6946
+    },
+    {
+      "epoch": 5.549050949050949,
+      "grad_norm": 1.5703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9223,
+      "step": 6947
+    },
+    {
+      "epoch": 5.54985014985015,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8851,
+      "step": 6948
+    },
+    {
+      "epoch": 5.55064935064935,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8884,
+      "step": 6949
+    },
+    {
+      "epoch": 5.551448551448551,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0002,
+      "loss": 0.891,
+      "step": 6950
+    },
+    {
+      "epoch": 5.5522477522477525,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9045,
+      "step": 6951
+    },
+    {
+      "epoch": 5.5530469530469535,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8925,
+      "step": 6952
+    },
+    {
+      "epoch": 5.553846153846154,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0002,
+      "loss": 0.8941,
+      "step": 6953
+    },
+    {
+      "epoch": 5.554645354645355,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0002,
+      "loss": 0.8879,
+      "step": 6954
+    },
+    {
+      "epoch": 5.555444555444556,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8875,
+      "step": 6955
+    },
+    {
+      "epoch": 5.556243756243756,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8981,
+      "step": 6956
+    },
+    {
+      "epoch": 5.557042957042957,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8954,
+      "step": 6957
+    },
+    {
+      "epoch": 5.557842157842158,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8887,
+      "step": 6958
+    },
+    {
+      "epoch": 5.558641358641359,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8837,
+      "step": 6959
+    },
+    {
+      "epoch": 5.559440559440559,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0002,
+      "loss": 0.8912,
+      "step": 6960
+    },
+    {
+      "epoch": 5.56023976023976,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8935,
+      "step": 6961
+    },
+    {
+      "epoch": 5.561038961038961,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0002,
+      "loss": 0.892,
+      "step": 6962
+    },
+    {
+      "epoch": 5.561838161838162,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0002,
+      "loss": 0.8963,
+      "step": 6963
+    },
+    {
+      "epoch": 5.562637362637362,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8961,
+      "step": 6964
+    },
+    {
+      "epoch": 5.5634365634365635,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0002,
+      "loss": 0.894,
+      "step": 6965
+    },
+    {
+      "epoch": 5.5642357642357645,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8981,
+      "step": 6966
+    },
+    {
+      "epoch": 5.565034965034965,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0002,
+      "loss": 0.8934,
+      "step": 6967
+    },
+    {
+      "epoch": 5.565834165834166,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.0002,
+      "loss": 0.8939,
+      "step": 6968
+    },
+    {
+      "epoch": 5.566633366633367,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0002,
+      "loss": 0.8883,
+      "step": 6969
+    },
+    {
+      "epoch": 5.567432567432568,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.0002,
+      "loss": 0.8914,
+      "step": 6970
+    },
+    {
+      "epoch": 5.568231768231768,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8903,
+      "step": 6971
+    },
+    {
+      "epoch": 5.569030969030969,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0002,
+      "loss": 0.8895,
+      "step": 6972
+    },
+    {
+      "epoch": 5.56983016983017,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0002,
+      "loss": 0.8945,
+      "step": 6973
+    },
+    {
+      "epoch": 5.570629370629371,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0002,
+      "loss": 0.892,
+      "step": 6974
+    },
+    {
+      "epoch": 5.571428571428571,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8912,
+      "step": 6975
+    },
+    {
+      "epoch": 5.572227772227772,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8892,
+      "step": 6976
+    },
+    {
+      "epoch": 5.573026973026973,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.0002,
+      "loss": 0.8851,
+      "step": 6977
+    },
+    {
+      "epoch": 5.573826173826173,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0002,
+      "loss": 0.8913,
+      "step": 6978
+    },
+    {
+      "epoch": 5.574625374625374,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8943,
+      "step": 6979
+    },
+    {
+      "epoch": 5.5754245754245755,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0002,
+      "loss": 0.8969,
+      "step": 6980
+    },
+    {
+      "epoch": 5.5762237762237765,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8892,
+      "step": 6981
+    },
+    {
+      "epoch": 5.577022977022977,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8951,
+      "step": 6982
+    },
+    {
+      "epoch": 5.577822177822178,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0002,
+      "loss": 0.8878,
+      "step": 6983
+    },
+    {
+      "epoch": 5.578621378621379,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0002,
+      "loss": 0.894,
+      "step": 6984
+    },
+    {
+      "epoch": 5.57942057942058,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0002,
+      "loss": 0.8857,
+      "step": 6985
+    },
+    {
+      "epoch": 5.58021978021978,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0002,
+      "loss": 0.8957,
+      "step": 6986
+    },
+    {
+      "epoch": 5.581018981018981,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9047,
+      "step": 6987
+    },
+    {
+      "epoch": 5.581818181818182,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0002,
+      "loss": 0.8898,
+      "step": 6988
+    },
+    {
+      "epoch": 5.582617382617382,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.0002,
+      "loss": 0.8936,
+      "step": 6989
+    },
+    {
+      "epoch": 5.583416583416583,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9002,
+      "step": 6990
+    },
+    {
+      "epoch": 5.584215784215784,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.0002,
+      "loss": 0.8869,
+      "step": 6991
+    },
+    {
+      "epoch": 5.585014985014985,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0002,
+      "loss": 0.8905,
+      "step": 6992
+    },
+    {
+      "epoch": 5.585814185814185,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8927,
+      "step": 6993
+    },
+    {
+      "epoch": 5.5866133866133865,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0002,
+      "loss": 0.8863,
+      "step": 6994
+    },
+    {
+      "epoch": 5.5874125874125875,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0002,
+      "loss": 0.8918,
+      "step": 6995
+    },
+    {
+      "epoch": 5.5882117882117885,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8924,
+      "step": 6996
+    },
+    {
+      "epoch": 5.589010989010989,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.0002,
+      "loss": 0.8881,
+      "step": 6997
+    },
+    {
+      "epoch": 5.58981018981019,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.0002,
+      "loss": 0.8944,
+      "step": 6998
+    },
+    {
+      "epoch": 5.590609390609391,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0002,
+      "loss": 0.8897,
+      "step": 6999
+    },
+    {
+      "epoch": 5.591408591408591,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.0002,
+      "loss": 0.888,
+      "step": 7000
+    },
+    {
+      "epoch": 5.592207792207792,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0002,
+      "loss": 0.8949,
+      "step": 7001
+    },
+    {
+      "epoch": 5.593006993006993,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0002,
+      "loss": 0.8837,
+      "step": 7002
+    },
+    {
+      "epoch": 5.593806193806194,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8928,
+      "step": 7003
+    },
+    {
+      "epoch": 5.594605394605395,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.0002,
+      "loss": 0.8944,
+      "step": 7004
+    },
+    {
+      "epoch": 5.595404595404595,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8886,
+      "step": 7005
+    },
+    {
+      "epoch": 5.596203796203796,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0002,
+      "loss": 0.9025,
+      "step": 7006
+    },
+    {
+      "epoch": 5.597002997002997,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0002,
+      "loss": 0.8913,
+      "step": 7007
+    },
+    {
+      "epoch": 5.5978021978021975,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0002,
+      "loss": 0.8847,
+      "step": 7008
+    },
+    {
+      "epoch": 5.5986013986013985,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0002,
+      "loss": 0.8869,
+      "step": 7009
+    },
+    {
+      "epoch": 5.5994005994005995,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0002,
+      "loss": 0.887,
+      "step": 7010
+    },
+    {
+      "epoch": 5.600199800199801,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8988,
+      "step": 7011
+    },
+    {
+      "epoch": 5.600999000999001,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8929,
+      "step": 7012
+    },
+    {
+      "epoch": 5.601798201798202,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8969,
+      "step": 7013
+    },
+    {
+      "epoch": 5.602597402597403,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8927,
+      "step": 7014
+    },
+    {
+      "epoch": 5.603396603396604,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8916,
+      "step": 7015
+    },
+    {
+      "epoch": 5.604195804195804,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0002,
+      "loss": 0.883,
+      "step": 7016
+    },
+    {
+      "epoch": 5.604995004995005,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8841,
+      "step": 7017
+    },
+    {
+      "epoch": 5.605794205794206,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0002,
+      "loss": 0.9004,
+      "step": 7018
+    },
+    {
+      "epoch": 5.606593406593406,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0002,
+      "loss": 0.8918,
+      "step": 7019
+    },
+    {
+      "epoch": 5.607392607392607,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0002,
+      "loss": 0.8798,
+      "step": 7020
+    },
+    {
+      "epoch": 5.608191808191808,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0002,
+      "loss": 0.8873,
+      "step": 7021
+    },
+    {
+      "epoch": 5.608991008991009,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0002,
+      "loss": 0.8908,
+      "step": 7022
+    },
+    {
+      "epoch": 5.6097902097902095,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8904,
+      "step": 7023
+    },
+    {
+      "epoch": 5.6105894105894105,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0002,
+      "loss": 0.8949,
+      "step": 7024
+    },
+    {
+      "epoch": 5.611388611388612,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0002,
+      "loss": 0.8972,
+      "step": 7025
+    },
+    {
+      "epoch": 5.612187812187813,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0002,
+      "loss": 0.8926,
+      "step": 7026
+    },
+    {
+      "epoch": 5.612987012987013,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9009,
+      "step": 7027
+    },
+    {
+      "epoch": 5.613786213786214,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.0002,
+      "loss": 0.8925,
+      "step": 7028
+    },
+    {
+      "epoch": 5.614585414585415,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0002,
+      "loss": 0.891,
+      "step": 7029
+    },
+    {
+      "epoch": 5.615384615384615,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8904,
+      "step": 7030
+    },
+    {
+      "epoch": 5.616183816183816,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0002,
+      "loss": 0.8894,
+      "step": 7031
+    },
+    {
+      "epoch": 5.616983016983017,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8921,
+      "step": 7032
+    },
+    {
+      "epoch": 5.617782217782218,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8947,
+      "step": 7033
+    },
+    {
+      "epoch": 5.618581418581418,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.0002,
+      "loss": 0.8906,
+      "step": 7034
+    },
+    {
+      "epoch": 5.619380619380619,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.0002,
+      "loss": 0.8949,
+      "step": 7035
+    },
+    {
+      "epoch": 5.62017982017982,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0002,
+      "loss": 0.8903,
+      "step": 7036
+    },
+    {
+      "epoch": 5.620979020979021,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.0002,
+      "loss": 0.892,
+      "step": 7037
+    },
+    {
+      "epoch": 5.6217782217782215,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.0002,
+      "loss": 0.8984,
+      "step": 7038
+    },
+    {
+      "epoch": 5.6225774225774225,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.0002,
+      "loss": 0.8867,
+      "step": 7039
+    },
+    {
+      "epoch": 5.623376623376624,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0002,
+      "loss": 0.8858,
+      "step": 7040
+    },
+    {
+      "epoch": 5.624175824175824,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.0002,
+      "loss": 0.8899,
+      "step": 7041
+    },
+    {
+      "epoch": 5.624975024975025,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0002,
+      "loss": 0.8981,
+      "step": 7042
+    },
+    {
+      "epoch": 5.625774225774226,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0002,
+      "loss": 0.8823,
+      "step": 7043
+    },
+    {
+      "epoch": 5.626573426573427,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8874,
+      "step": 7044
+    },
+    {
+      "epoch": 5.627372627372627,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8963,
+      "step": 7045
+    },
+    {
+      "epoch": 5.628171828171828,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0002,
+      "loss": 0.893,
+      "step": 7046
+    },
+    {
+      "epoch": 5.628971028971029,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.0002,
+      "loss": 0.8858,
+      "step": 7047
+    },
+    {
+      "epoch": 5.62977022977023,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8868,
+      "step": 7048
+    },
+    {
+      "epoch": 5.63056943056943,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0002,
+      "loss": 0.8881,
+      "step": 7049
+    },
+    {
+      "epoch": 5.631368631368631,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0002,
+      "loss": 0.8845,
+      "step": 7050
+    },
+    {
+      "epoch": 5.632167832167832,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0002,
+      "loss": 0.8922,
+      "step": 7051
+    },
+    {
+      "epoch": 5.6329670329670325,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.0002,
+      "loss": 0.8951,
+      "step": 7052
+    },
+    {
+      "epoch": 5.6337662337662335,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0002,
+      "loss": 0.883,
+      "step": 7053
+    },
+    {
+      "epoch": 5.634565434565435,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.0002,
+      "loss": 0.8898,
+      "step": 7054
+    },
+    {
+      "epoch": 5.635364635364636,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0002,
+      "loss": 0.8931,
+      "step": 7055
+    },
+    {
+      "epoch": 5.636163836163837,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.0002,
+      "loss": 0.8868,
+      "step": 7056
+    },
+    {
+      "epoch": 5.636963036963037,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0002,
+      "loss": 0.8891,
+      "step": 7057
+    },
+    {
+      "epoch": 5.637762237762238,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.0002,
+      "loss": 0.8973,
+      "step": 7058
+    },
+    {
+      "epoch": 5.638561438561439,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8903,
+      "step": 7059
+    },
+    {
+      "epoch": 5.639360639360639,
+      "grad_norm": 0.80859375,
+      "learning_rate": 0.0002,
+      "loss": 0.8974,
+      "step": 7060
+    },
+    {
+      "epoch": 5.64015984015984,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.0002,
+      "loss": 0.8931,
+      "step": 7061
+    },
+    {
+      "epoch": 5.640959040959041,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0002,
+      "loss": 0.8836,
+      "step": 7062
+    },
+    {
+      "epoch": 5.641758241758242,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8954,
+      "step": 7063
+    },
+    {
+      "epoch": 5.642557442557442,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0002,
+      "loss": 0.8972,
+      "step": 7064
+    },
+    {
+      "epoch": 5.643356643356643,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0002,
+      "loss": 0.8873,
+      "step": 7065
+    },
+    {
+      "epoch": 5.644155844155844,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8866,
+      "step": 7066
+    },
+    {
+      "epoch": 5.644955044955045,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0002,
+      "loss": 0.8896,
+      "step": 7067
+    },
+    {
+      "epoch": 5.645754245754246,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8885,
+      "step": 7068
+    },
+    {
+      "epoch": 5.646553446553447,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0002,
+      "loss": 0.8963,
+      "step": 7069
+    },
+    {
+      "epoch": 5.647352647352648,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0002,
+      "loss": 0.8914,
+      "step": 7070
+    },
+    {
+      "epoch": 5.648151848151848,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8983,
+      "step": 7071
+    },
+    {
+      "epoch": 5.648951048951049,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0002,
+      "loss": 0.8898,
+      "step": 7072
+    },
+    {
+      "epoch": 5.64975024975025,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0002,
+      "loss": 0.8904,
+      "step": 7073
+    },
+    {
+      "epoch": 5.650549450549451,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0002,
+      "loss": 0.8871,
+      "step": 7074
+    },
+    {
+      "epoch": 5.651348651348651,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8948,
+      "step": 7075
+    },
+    {
+      "epoch": 5.652147852147852,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.0002,
+      "loss": 0.8913,
+      "step": 7076
+    },
+    {
+      "epoch": 5.652947052947053,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0002,
+      "loss": 0.8908,
+      "step": 7077
+    },
+    {
+      "epoch": 5.653746253746254,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0002,
+      "loss": 0.8928,
+      "step": 7078
+    },
+    {
+      "epoch": 5.654545454545454,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8878,
+      "step": 7079
+    },
+    {
+      "epoch": 5.655344655344655,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.0002,
+      "loss": 0.8877,
+      "step": 7080
+    },
+    {
+      "epoch": 5.656143856143856,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8931,
+      "step": 7081
+    },
+    {
+      "epoch": 5.6569430569430565,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8938,
+      "step": 7082
+    },
+    {
+      "epoch": 5.657742257742258,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8837,
+      "step": 7083
+    },
+    {
+      "epoch": 5.658541458541459,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8907,
+      "step": 7084
+    },
+    {
+      "epoch": 5.65934065934066,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0002,
+      "loss": 0.888,
+      "step": 7085
+    },
+    {
+      "epoch": 5.66013986013986,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.0002,
+      "loss": 0.888,
+      "step": 7086
+    },
+    {
+      "epoch": 5.660939060939061,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9013,
+      "step": 7087
+    },
+    {
+      "epoch": 5.661738261738262,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.0002,
+      "loss": 0.889,
+      "step": 7088
+    },
+    {
+      "epoch": 5.662537462537463,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8889,
+      "step": 7089
+    },
+    {
+      "epoch": 5.663336663336663,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.0002,
+      "loss": 0.8845,
+      "step": 7090
+    },
+    {
+      "epoch": 5.664135864135864,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8965,
+      "step": 7091
+    },
+    {
+      "epoch": 5.664935064935065,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8866,
+      "step": 7092
+    },
+    {
+      "epoch": 5.665734265734265,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0002,
+      "loss": 0.887,
+      "step": 7093
+    },
+    {
+      "epoch": 5.666533466533466,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0002,
+      "loss": 0.8844,
+      "step": 7094
+    },
+    {
+      "epoch": 5.667332667332667,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.0002,
+      "loss": 0.8936,
+      "step": 7095
+    },
+    {
+      "epoch": 5.668131868131868,
+      "grad_norm": 0.375,
+      "learning_rate": 0.0002,
+      "loss": 0.8899,
+      "step": 7096
+    },
+    {
+      "epoch": 5.668931068931069,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0002,
+      "loss": 0.8864,
+      "step": 7097
+    },
+    {
+      "epoch": 5.66973026973027,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0002,
+      "loss": 0.8837,
+      "step": 7098
+    },
+    {
+      "epoch": 5.670529470529471,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0002,
+      "loss": 0.8948,
+      "step": 7099
+    },
+    {
+      "epoch": 5.671328671328672,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0002,
+      "loss": 0.8941,
+      "step": 7100
+    },
+    {
+      "epoch": 5.672127872127872,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8868,
+      "step": 7101
+    },
+    {
+      "epoch": 5.672927072927073,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.0002,
+      "loss": 0.8894,
+      "step": 7102
+    },
+    {
+      "epoch": 5.673726273726274,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.0002,
+      "loss": 0.8981,
+      "step": 7103
+    },
+    {
+      "epoch": 5.674525474525474,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8902,
+      "step": 7104
+    },
+    {
+      "epoch": 5.675324675324675,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0002,
+      "loss": 0.8825,
+      "step": 7105
+    },
+    {
+      "epoch": 5.676123876123876,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8903,
+      "step": 7106
+    },
+    {
+      "epoch": 5.676923076923077,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.0002,
+      "loss": 0.886,
+      "step": 7107
+    },
+    {
+      "epoch": 5.677722277722278,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8933,
+      "step": 7108
+    },
+    {
+      "epoch": 5.678521478521478,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0002,
+      "loss": 0.8888,
+      "step": 7109
+    },
+    {
+      "epoch": 5.679320679320679,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0002,
+      "loss": 0.8912,
+      "step": 7110
+    },
+    {
+      "epoch": 5.6801198801198804,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8838,
+      "step": 7111
+    },
+    {
+      "epoch": 5.680919080919081,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8841,
+      "step": 7112
+    },
+    {
+      "epoch": 5.681718281718282,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8972,
+      "step": 7113
+    },
+    {
+      "epoch": 5.682517482517483,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0002,
+      "loss": 0.891,
+      "step": 7114
+    },
+    {
+      "epoch": 5.683316683316683,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0002,
+      "loss": 0.8936,
+      "step": 7115
+    },
+    {
+      "epoch": 5.684115884115884,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8913,
+      "step": 7116
+    },
+    {
+      "epoch": 5.684915084915085,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8872,
+      "step": 7117
+    },
+    {
+      "epoch": 5.685714285714286,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8906,
+      "step": 7118
+    },
+    {
+      "epoch": 5.686513486513487,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8908,
+      "step": 7119
+    },
+    {
+      "epoch": 5.687312687312687,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8825,
+      "step": 7120
+    },
+    {
+      "epoch": 5.688111888111888,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8949,
+      "step": 7121
+    },
+    {
+      "epoch": 5.688911088911089,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8899,
+      "step": 7122
+    },
+    {
+      "epoch": 5.689710289710289,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0002,
+      "loss": 0.8876,
+      "step": 7123
+    },
+    {
+      "epoch": 5.69050949050949,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8786,
+      "step": 7124
+    },
+    {
+      "epoch": 5.691308691308691,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.0002,
+      "loss": 0.8963,
+      "step": 7125
+    },
+    {
+      "epoch": 5.6921078921078925,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0002,
+      "loss": 0.8945,
+      "step": 7126
+    },
+    {
+      "epoch": 5.692907092907093,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8899,
+      "step": 7127
+    },
+    {
+      "epoch": 5.693706293706294,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0002,
+      "loss": 0.8803,
+      "step": 7128
+    },
+    {
+      "epoch": 5.694505494505495,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.0002,
+      "loss": 0.8937,
+      "step": 7129
+    },
+    {
+      "epoch": 5.695304695304696,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0002,
+      "loss": 0.8921,
+      "step": 7130
+    },
+    {
+      "epoch": 5.696103896103896,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8843,
+      "step": 7131
+    },
+    {
+      "epoch": 5.696903096903097,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8871,
+      "step": 7132
+    },
+    {
+      "epoch": 5.697702297702298,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8871,
+      "step": 7133
+    },
+    {
+      "epoch": 5.698501498501498,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0002,
+      "loss": 0.8944,
+      "step": 7134
+    },
+    {
+      "epoch": 5.699300699300699,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8912,
+      "step": 7135
+    },
+    {
+      "epoch": 5.7000999000999,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8787,
+      "step": 7136
+    },
+    {
+      "epoch": 5.700899100899101,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.0002,
+      "loss": 0.885,
+      "step": 7137
+    },
+    {
+      "epoch": 5.701698301698301,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8824,
+      "step": 7138
+    },
+    {
+      "epoch": 5.702497502497502,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0002,
+      "loss": 0.8961,
+      "step": 7139
+    },
+    {
+      "epoch": 5.7032967032967035,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8913,
+      "step": 7140
+    },
+    {
+      "epoch": 5.7040959040959045,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0002,
+      "loss": 0.8875,
+      "step": 7141
+    },
+    {
+      "epoch": 5.704895104895105,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.0002,
+      "loss": 0.8982,
+      "step": 7142
+    },
+    {
+      "epoch": 5.705694305694306,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8933,
+      "step": 7143
+    },
+    {
+      "epoch": 5.706493506493507,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.0002,
+      "loss": 0.89,
+      "step": 7144
+    },
+    {
+      "epoch": 5.707292707292707,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0002,
+      "loss": 0.8871,
+      "step": 7145
+    },
+    {
+      "epoch": 5.708091908091908,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8892,
+      "step": 7146
+    },
+    {
+      "epoch": 5.708891108891109,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8902,
+      "step": 7147
+    },
+    {
+      "epoch": 5.70969030969031,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8945,
+      "step": 7148
+    },
+    {
+      "epoch": 5.71048951048951,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0002,
+      "loss": 0.8899,
+      "step": 7149
+    },
+    {
+      "epoch": 5.711288711288711,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8853,
+      "step": 7150
+    },
+    {
+      "epoch": 5.712087912087912,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0002,
+      "loss": 0.895,
+      "step": 7151
+    },
+    {
+      "epoch": 5.712887112887113,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0002,
+      "loss": 0.8833,
+      "step": 7152
+    },
+    {
+      "epoch": 5.713686313686313,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0002,
+      "loss": 0.8856,
+      "step": 7153
+    },
+    {
+      "epoch": 5.7144855144855145,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0002,
+      "loss": 0.8773,
+      "step": 7154
+    },
+    {
+      "epoch": 5.7152847152847155,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8803,
+      "step": 7155
+    },
+    {
+      "epoch": 5.716083916083916,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8986,
+      "step": 7156
+    },
+    {
+      "epoch": 5.716883116883117,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0002,
+      "loss": 0.8963,
+      "step": 7157
+    },
+    {
+      "epoch": 5.717682317682318,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8824,
+      "step": 7158
+    },
+    {
+      "epoch": 5.718481518481519,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.0002,
+      "loss": 0.8987,
+      "step": 7159
+    },
+    {
+      "epoch": 5.719280719280719,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0002,
+      "loss": 0.8973,
+      "step": 7160
+    },
+    {
+      "epoch": 5.72007992007992,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0002,
+      "loss": 0.8799,
+      "step": 7161
+    },
+    {
+      "epoch": 5.720879120879121,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0002,
+      "loss": 0.8803,
+      "step": 7162
+    },
+    {
+      "epoch": 5.721678321678322,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8909,
+      "step": 7163
+    },
+    {
+      "epoch": 5.722477522477522,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0002,
+      "loss": 0.8975,
+      "step": 7164
+    },
+    {
+      "epoch": 5.723276723276723,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0002,
+      "loss": 0.881,
+      "step": 7165
+    },
+    {
+      "epoch": 5.724075924075924,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8865,
+      "step": 7166
+    },
+    {
+      "epoch": 5.724875124875124,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8809,
+      "step": 7167
+    },
+    {
+      "epoch": 5.725674325674325,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8836,
+      "step": 7168
+    },
+    {
+      "epoch": 5.7264735264735265,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.0002,
+      "loss": 0.8811,
+      "step": 7169
+    },
+    {
+      "epoch": 5.7272727272727275,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8919,
+      "step": 7170
+    },
+    {
+      "epoch": 5.7280719280719286,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8881,
+      "step": 7171
+    },
+    {
+      "epoch": 5.728871128871129,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8887,
+      "step": 7172
+    },
+    {
+      "epoch": 5.72967032967033,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0002,
+      "loss": 0.8908,
+      "step": 7173
+    },
+    {
+      "epoch": 5.730469530469531,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0002,
+      "loss": 0.8949,
+      "step": 7174
+    },
+    {
+      "epoch": 5.731268731268731,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8898,
+      "step": 7175
+    },
+    {
+      "epoch": 5.732067932067932,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0002,
+      "loss": 0.8872,
+      "step": 7176
+    },
+    {
+      "epoch": 5.732867132867133,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0002,
+      "loss": 0.8918,
+      "step": 7177
+    },
+    {
+      "epoch": 5.733666333666334,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.0002,
+      "loss": 0.8889,
+      "step": 7178
+    },
+    {
+      "epoch": 5.734465534465534,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0002,
+      "loss": 0.8863,
+      "step": 7179
+    },
+    {
+      "epoch": 5.735264735264735,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8896,
+      "step": 7180
+    },
+    {
+      "epoch": 5.736063936063936,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8919,
+      "step": 7181
+    },
+    {
+      "epoch": 5.736863136863137,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8958,
+      "step": 7182
+    },
+    {
+      "epoch": 5.7376623376623375,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8933,
+      "step": 7183
+    },
+    {
+      "epoch": 5.7384615384615385,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8937,
+      "step": 7184
+    },
+    {
+      "epoch": 5.7392607392607395,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8903,
+      "step": 7185
+    },
+    {
+      "epoch": 5.74005994005994,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8935,
+      "step": 7186
+    },
+    {
+      "epoch": 5.740859140859141,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8876,
+      "step": 7187
+    },
+    {
+      "epoch": 5.741658341658342,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8949,
+      "step": 7188
+    },
+    {
+      "epoch": 5.742457542457543,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0002,
+      "loss": 0.8957,
+      "step": 7189
+    },
+    {
+      "epoch": 5.743256743256743,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.0002,
+      "loss": 0.8957,
+      "step": 7190
+    },
+    {
+      "epoch": 5.744055944055944,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0002,
+      "loss": 0.881,
+      "step": 7191
+    },
+    {
+      "epoch": 5.744855144855145,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0002,
+      "loss": 0.8899,
+      "step": 7192
+    },
+    {
+      "epoch": 5.745654345654346,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0002,
+      "loss": 0.8935,
+      "step": 7193
+    },
+    {
+      "epoch": 5.746453546453546,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0002,
+      "loss": 0.8787,
+      "step": 7194
+    },
+    {
+      "epoch": 5.747252747252747,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0002,
+      "loss": 0.893,
+      "step": 7195
+    },
+    {
+      "epoch": 5.748051948051948,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0002,
+      "loss": 0.892,
+      "step": 7196
+    },
+    {
+      "epoch": 5.7488511488511485,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0002,
+      "loss": 0.8911,
+      "step": 7197
+    },
+    {
+      "epoch": 5.7496503496503495,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8859,
+      "step": 7198
+    },
+    {
+      "epoch": 5.7504495504495505,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0002,
+      "loss": 0.8845,
+      "step": 7199
+    },
+    {
+      "epoch": 5.751248751248752,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8948,
+      "step": 7200
+    },
+    {
+      "epoch": 5.752047952047952,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8885,
+      "step": 7201
+    },
+    {
+      "epoch": 5.752847152847153,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0002,
+      "loss": 0.8866,
+      "step": 7202
+    },
+    {
+      "epoch": 5.753646353646354,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0002,
+      "loss": 0.8923,
+      "step": 7203
+    },
+    {
+      "epoch": 5.754445554445555,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0002,
+      "loss": 0.8913,
+      "step": 7204
+    },
+    {
+      "epoch": 5.755244755244755,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8998,
+      "step": 7205
+    },
+    {
+      "epoch": 5.756043956043956,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8877,
+      "step": 7206
+    },
+    {
+      "epoch": 5.756843156843157,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0002,
+      "loss": 0.8927,
+      "step": 7207
+    },
+    {
+      "epoch": 5.757642357642357,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8968,
+      "step": 7208
+    },
+    {
+      "epoch": 5.758441558441558,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8919,
+      "step": 7209
+    },
+    {
+      "epoch": 5.759240759240759,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0002,
+      "loss": 0.8947,
+      "step": 7210
+    },
+    {
+      "epoch": 5.76003996003996,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0002,
+      "loss": 0.8901,
+      "step": 7211
+    },
+    {
+      "epoch": 5.7608391608391605,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0002,
+      "loss": 0.8899,
+      "step": 7212
+    },
+    {
+      "epoch": 5.7616383616383615,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0002,
+      "loss": 0.8922,
+      "step": 7213
+    },
+    {
+      "epoch": 5.7624375624375626,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8824,
+      "step": 7214
+    },
+    {
+      "epoch": 5.763236763236764,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0002,
+      "loss": 0.8903,
+      "step": 7215
+    },
+    {
+      "epoch": 5.764035964035964,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8986,
+      "step": 7216
+    },
+    {
+      "epoch": 5.764835164835165,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0002,
+      "loss": 0.8985,
+      "step": 7217
+    },
+    {
+      "epoch": 5.765634365634366,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8837,
+      "step": 7218
+    },
+    {
+      "epoch": 5.766433566433566,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8992,
+      "step": 7219
+    },
+    {
+      "epoch": 5.767232767232767,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0002,
+      "loss": 0.8892,
+      "step": 7220
+    },
+    {
+      "epoch": 5.768031968031968,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8978,
+      "step": 7221
+    },
+    {
+      "epoch": 5.768831168831169,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8892,
+      "step": 7222
+    },
+    {
+      "epoch": 5.76963036963037,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0002,
+      "loss": 0.8945,
+      "step": 7223
+    },
+    {
+      "epoch": 5.77042957042957,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8825,
+      "step": 7224
+    },
+    {
+      "epoch": 5.771228771228771,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8835,
+      "step": 7225
+    },
+    {
+      "epoch": 5.772027972027972,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8894,
+      "step": 7226
+    },
+    {
+      "epoch": 5.7728271728271725,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0002,
+      "loss": 0.886,
+      "step": 7227
+    },
+    {
+      "epoch": 5.7736263736263735,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0002,
+      "loss": 0.89,
+      "step": 7228
+    },
+    {
+      "epoch": 5.774425574425575,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.0002,
+      "loss": 0.8901,
+      "step": 7229
+    },
+    {
+      "epoch": 5.775224775224775,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8862,
+      "step": 7230
+    },
+    {
+      "epoch": 5.776023976023976,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0002,
+      "loss": 0.8875,
+      "step": 7231
+    },
+    {
+      "epoch": 5.776823176823177,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8866,
+      "step": 7232
+    },
+    {
+      "epoch": 5.777622377622378,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8921,
+      "step": 7233
+    },
+    {
+      "epoch": 5.778421578421579,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0002,
+      "loss": 0.8878,
+      "step": 7234
+    },
+    {
+      "epoch": 5.779220779220779,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8839,
+      "step": 7235
+    },
+    {
+      "epoch": 5.78001998001998,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.0002,
+      "loss": 0.886,
+      "step": 7236
+    },
+    {
+      "epoch": 5.780819180819181,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8895,
+      "step": 7237
+    },
+    {
+      "epoch": 5.781618381618381,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8907,
+      "step": 7238
+    },
+    {
+      "epoch": 5.782417582417582,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0002,
+      "loss": 0.889,
+      "step": 7239
+    },
+    {
+      "epoch": 5.783216783216783,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0002,
+      "loss": 0.8849,
+      "step": 7240
+    },
+    {
+      "epoch": 5.784015984015984,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0002,
+      "loss": 0.8881,
+      "step": 7241
+    },
+    {
+      "epoch": 5.7848151848151845,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0002,
+      "loss": 0.8846,
+      "step": 7242
+    },
+    {
+      "epoch": 5.785614385614386,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0002,
+      "loss": 0.8817,
+      "step": 7243
+    },
+    {
+      "epoch": 5.786413586413587,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0002,
+      "loss": 0.8829,
+      "step": 7244
+    },
+    {
+      "epoch": 5.787212787212788,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0002,
+      "loss": 0.8949,
+      "step": 7245
+    },
+    {
+      "epoch": 5.788011988011988,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.0002,
+      "loss": 0.8923,
+      "step": 7246
+    },
+    {
+      "epoch": 5.788811188811189,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8863,
+      "step": 7247
+    },
+    {
+      "epoch": 5.78961038961039,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.0002,
+      "loss": 0.887,
+      "step": 7248
+    },
+    {
+      "epoch": 5.79040959040959,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0002,
+      "loss": 0.8852,
+      "step": 7249
+    },
+    {
+      "epoch": 5.791208791208791,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8934,
+      "step": 7250
+    },
+    {
+      "epoch": 5.792007992007992,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.0002,
+      "loss": 0.8976,
+      "step": 7251
+    },
+    {
+      "epoch": 5.792807192807193,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8896,
+      "step": 7252
+    },
+    {
+      "epoch": 5.793606393606393,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8832,
+      "step": 7253
+    },
+    {
+      "epoch": 5.794405594405594,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0002,
+      "loss": 0.8863,
+      "step": 7254
+    },
+    {
+      "epoch": 5.795204795204795,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8811,
+      "step": 7255
+    },
+    {
+      "epoch": 5.796003996003996,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8921,
+      "step": 7256
+    },
+    {
+      "epoch": 5.796803196803197,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0002,
+      "loss": 0.8775,
+      "step": 7257
+    },
+    {
+      "epoch": 5.797602397602398,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8969,
+      "step": 7258
+    },
+    {
+      "epoch": 5.798401598401599,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0002,
+      "loss": 0.8986,
+      "step": 7259
+    },
+    {
+      "epoch": 5.799200799200799,
+      "grad_norm": 0.9609375,
+      "learning_rate": 0.0002,
+      "loss": 0.9165,
+      "step": 7260
+    },
+    {
+      "epoch": 5.8,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8964,
+      "step": 7261
+    },
+    {
+      "epoch": 5.800799200799201,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8957,
+      "step": 7262
+    },
+    {
+      "epoch": 5.801598401598402,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8844,
+      "step": 7263
+    },
+    {
+      "epoch": 5.802397602397602,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8846,
+      "step": 7264
+    },
+    {
+      "epoch": 5.803196803196803,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0002,
+      "loss": 0.8857,
+      "step": 7265
+    },
+    {
+      "epoch": 5.803996003996004,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8867,
+      "step": 7266
+    },
+    {
+      "epoch": 5.804795204795205,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8811,
+      "step": 7267
+    },
+    {
+      "epoch": 5.805594405594405,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8832,
+      "step": 7268
+    },
+    {
+      "epoch": 5.806393606393606,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0002,
+      "loss": 0.8897,
+      "step": 7269
+    },
+    {
+      "epoch": 5.807192807192807,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0002,
+      "loss": 0.8934,
+      "step": 7270
+    },
+    {
+      "epoch": 5.8079920079920075,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8978,
+      "step": 7271
+    },
+    {
+      "epoch": 5.808791208791209,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8875,
+      "step": 7272
+    },
+    {
+      "epoch": 5.80959040959041,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0002,
+      "loss": 0.8872,
+      "step": 7273
+    },
+    {
+      "epoch": 5.810389610389611,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8916,
+      "step": 7274
+    },
+    {
+      "epoch": 5.811188811188811,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8892,
+      "step": 7275
+    },
+    {
+      "epoch": 5.811988011988012,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0002,
+      "loss": 0.8916,
+      "step": 7276
+    },
+    {
+      "epoch": 5.812787212787213,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0002,
+      "loss": 0.8966,
+      "step": 7277
+    },
+    {
+      "epoch": 5.813586413586414,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0002,
+      "loss": 0.893,
+      "step": 7278
+    },
+    {
+      "epoch": 5.814385614385614,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8951,
+      "step": 7279
+    },
+    {
+      "epoch": 5.815184815184815,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8897,
+      "step": 7280
+    },
+    {
+      "epoch": 5.815984015984016,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8956,
+      "step": 7281
+    },
+    {
+      "epoch": 5.816783216783216,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0002,
+      "loss": 0.8857,
+      "step": 7282
+    },
+    {
+      "epoch": 5.817582417582417,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8909,
+      "step": 7283
+    },
+    {
+      "epoch": 5.818381618381618,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8832,
+      "step": 7284
+    },
+    {
+      "epoch": 5.819180819180819,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8854,
+      "step": 7285
+    },
+    {
+      "epoch": 5.8199800199800205,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0002,
+      "loss": 0.8867,
+      "step": 7286
+    },
+    {
+      "epoch": 5.820779220779221,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.0002,
+      "loss": 0.8858,
+      "step": 7287
+    },
+    {
+      "epoch": 5.821578421578422,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.0002,
+      "loss": 0.8887,
+      "step": 7288
+    },
+    {
+      "epoch": 5.822377622377623,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0002,
+      "loss": 0.8848,
+      "step": 7289
+    },
+    {
+      "epoch": 5.823176823176823,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.0002,
+      "loss": 0.8858,
+      "step": 7290
+    },
+    {
+      "epoch": 5.823976023976024,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0002,
+      "loss": 0.8917,
+      "step": 7291
+    },
+    {
+      "epoch": 5.824775224775225,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0002,
+      "loss": 0.8986,
+      "step": 7292
+    },
+    {
+      "epoch": 5.825574425574426,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8927,
+      "step": 7293
+    },
+    {
+      "epoch": 5.826373626373626,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.0002,
+      "loss": 0.8948,
+      "step": 7294
+    },
+    {
+      "epoch": 5.827172827172827,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.0002,
+      "loss": 0.8899,
+      "step": 7295
+    },
+    {
+      "epoch": 5.827972027972028,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8925,
+      "step": 7296
+    },
+    {
+      "epoch": 5.828771228771229,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8799,
+      "step": 7297
+    },
+    {
+      "epoch": 5.829570429570429,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8938,
+      "step": 7298
+    },
+    {
+      "epoch": 5.83036963036963,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0002,
+      "loss": 0.8944,
+      "step": 7299
+    },
+    {
+      "epoch": 5.8311688311688314,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.0002,
+      "loss": 0.8931,
+      "step": 7300
+    },
+    {
+      "epoch": 5.831968031968032,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8855,
+      "step": 7301
+    },
+    {
+      "epoch": 5.832767232767233,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8865,
+      "step": 7302
+    },
+    {
+      "epoch": 5.833566433566434,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8863,
+      "step": 7303
+    },
+    {
+      "epoch": 5.834365634365635,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0002,
+      "loss": 0.893,
+      "step": 7304
+    },
+    {
+      "epoch": 5.835164835164835,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0002,
+      "loss": 0.8904,
+      "step": 7305
+    },
+    {
+      "epoch": 5.835964035964036,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8967,
+      "step": 7306
+    },
+    {
+      "epoch": 5.836763236763237,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8887,
+      "step": 7307
+    },
+    {
+      "epoch": 5.837562437562438,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8909,
+      "step": 7308
+    },
+    {
+      "epoch": 5.838361638361638,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.0002,
+      "loss": 0.8913,
+      "step": 7309
+    },
+    {
+      "epoch": 5.839160839160839,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.0002,
+      "loss": 0.8884,
+      "step": 7310
+    },
+    {
+      "epoch": 5.83996003996004,
+      "grad_norm": 1.5703125,
+      "learning_rate": 0.0002,
+      "loss": 0.9149,
+      "step": 7311
+    },
+    {
+      "epoch": 5.84075924075924,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0002,
+      "loss": 0.892,
+      "step": 7312
+    },
+    {
+      "epoch": 5.841558441558441,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0002,
+      "loss": 0.8894,
+      "step": 7313
+    },
+    {
+      "epoch": 5.842357642357642,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0002,
+      "loss": 0.8832,
+      "step": 7314
+    },
+    {
+      "epoch": 5.8431568431568435,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8822,
+      "step": 7315
+    },
+    {
+      "epoch": 5.843956043956044,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0002,
+      "loss": 0.8886,
+      "step": 7316
+    },
+    {
+      "epoch": 5.844755244755245,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8901,
+      "step": 7317
+    },
+    {
+      "epoch": 5.845554445554446,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.0002,
+      "loss": 0.8878,
+      "step": 7318
+    },
+    {
+      "epoch": 5.846353646353647,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.0002,
+      "loss": 0.8872,
+      "step": 7319
+    },
+    {
+      "epoch": 5.847152847152847,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0002,
+      "loss": 0.891,
+      "step": 7320
+    },
+    {
+      "epoch": 5.847952047952048,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8925,
+      "step": 7321
+    },
+    {
+      "epoch": 5.848751248751249,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8962,
+      "step": 7322
+    },
+    {
+      "epoch": 5.849550449550449,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0002,
+      "loss": 0.8909,
+      "step": 7323
+    },
+    {
+      "epoch": 5.85034965034965,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8955,
+      "step": 7324
+    },
+    {
+      "epoch": 5.851148851148851,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0002,
+      "loss": 0.8889,
+      "step": 7325
+    },
+    {
+      "epoch": 5.851948051948052,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8837,
+      "step": 7326
+    },
+    {
+      "epoch": 5.852747252747252,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.0002,
+      "loss": 0.8791,
+      "step": 7327
+    },
+    {
+      "epoch": 5.853546453546453,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.0002,
+      "loss": 0.8946,
+      "step": 7328
+    },
+    {
+      "epoch": 5.8543456543456545,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.0002,
+      "loss": 0.8885,
+      "step": 7329
+    },
+    {
+      "epoch": 5.8551448551448555,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.0002,
+      "loss": 0.888,
+      "step": 7330
+    },
+    {
+      "epoch": 5.855944055944056,
+      "grad_norm": 0.375,
+      "learning_rate": 0.0002,
+      "loss": 0.8945,
+      "step": 7331
+    },
+    {
+      "epoch": 5.856743256743257,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.0002,
+      "loss": 0.8903,
+      "step": 7332
+    },
+    {
+      "epoch": 5.857542457542458,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.0002,
+      "loss": 0.8826,
+      "step": 7333
+    },
+    {
+      "epoch": 5.858341658341658,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.0002,
+      "loss": 0.8901,
+      "step": 7334
+    },
+    {
+      "epoch": 5.859140859140859,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8958,
+      "step": 7335
+    },
+    {
+      "epoch": 5.85994005994006,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.0002,
+      "loss": 0.8936,
+      "step": 7336
+    },
+    {
+      "epoch": 5.860739260739261,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.0002,
+      "loss": 0.895,
+      "step": 7337
+    },
+    {
+      "epoch": 5.861538461538462,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.0002,
+      "loss": 0.8864,
+      "step": 7338
+    },
+    {
+      "epoch": 5.862337662337662,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0002,
+      "loss": 0.8868,
+      "step": 7339
+    },
+    {
+      "epoch": 5.863136863136863,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8893,
+      "step": 7340
+    },
+    {
+      "epoch": 5.863936063936064,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9012,
+      "step": 7341
+    },
+    {
+      "epoch": 5.864735264735264,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.0002,
+      "loss": 0.8935,
+      "step": 7342
+    },
+    {
+      "epoch": 5.8655344655344654,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8851,
+      "step": 7343
+    },
+    {
+      "epoch": 5.8663336663336665,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.0002,
+      "loss": 0.8901,
+      "step": 7344
+    },
+    {
+      "epoch": 5.867132867132867,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.0002,
+      "loss": 0.8887,
+      "step": 7345
+    },
+    {
+      "epoch": 5.867932067932068,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0002,
+      "loss": 0.8834,
+      "step": 7346
+    },
+    {
+      "epoch": 5.868731268731269,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0002,
+      "loss": 0.8872,
+      "step": 7347
+    },
+    {
+      "epoch": 5.86953046953047,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0002,
+      "loss": 0.8937,
+      "step": 7348
+    },
+    {
+      "epoch": 5.870329670329671,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0002,
+      "loss": 0.8843,
+      "step": 7349
+    },
+    {
+      "epoch": 5.871128871128871,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8933,
+      "step": 7350
+    },
+    {
+      "epoch": 5.871928071928072,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0002,
+      "loss": 0.8879,
+      "step": 7351
+    },
+    {
+      "epoch": 5.872727272727273,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0002,
+      "loss": 0.8893,
+      "step": 7352
+    },
+    {
+      "epoch": 5.873526473526473,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8934,
+      "step": 7353
+    },
+    {
+      "epoch": 5.874325674325674,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9038,
+      "step": 7354
+    },
+    {
+      "epoch": 5.875124875124875,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0002,
+      "loss": 0.895,
+      "step": 7355
+    },
+    {
+      "epoch": 5.875924075924076,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0002,
+      "loss": 0.9153,
+      "step": 7356
+    },
+    {
+      "epoch": 5.876723276723276,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0002,
+      "loss": 0.8931,
+      "step": 7357
+    },
+    {
+      "epoch": 5.8775224775224775,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0002,
+      "loss": 0.887,
+      "step": 7358
+    },
+    {
+      "epoch": 5.8783216783216785,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0002,
+      "loss": 0.8888,
+      "step": 7359
+    },
+    {
+      "epoch": 5.8791208791208796,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0002,
+      "loss": 0.899,
+      "step": 7360
+    },
+    {
+      "epoch": 5.87992007992008,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8911,
+      "step": 7361
+    },
+    {
+      "epoch": 5.880719280719281,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8868,
+      "step": 7362
+    },
+    {
+      "epoch": 5.881518481518482,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8958,
+      "step": 7363
+    },
+    {
+      "epoch": 5.882317682317682,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8894,
+      "step": 7364
+    },
+    {
+      "epoch": 5.883116883116883,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0002,
+      "loss": 0.879,
+      "step": 7365
+    },
+    {
+      "epoch": 5.883916083916084,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8859,
+      "step": 7366
+    },
+    {
+      "epoch": 5.884715284715285,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0002,
+      "loss": 0.8953,
+      "step": 7367
+    },
+    {
+      "epoch": 5.885514485514485,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0002,
+      "loss": 0.8945,
+      "step": 7368
+    },
+    {
+      "epoch": 5.886313686313686,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0002,
+      "loss": 0.8907,
+      "step": 7369
+    },
+    {
+      "epoch": 5.887112887112887,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0002,
+      "loss": 0.8854,
+      "step": 7370
+    },
+    {
+      "epoch": 5.887912087912088,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0002,
+      "loss": 0.8922,
+      "step": 7371
+    },
+    {
+      "epoch": 5.8887112887112885,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.0002,
+      "loss": 0.8943,
+      "step": 7372
+    },
+    {
+      "epoch": 5.8895104895104895,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8902,
+      "step": 7373
+    },
+    {
+      "epoch": 5.8903096903096905,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8889,
+      "step": 7374
+    },
+    {
+      "epoch": 5.891108891108891,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8936,
+      "step": 7375
+    },
+    {
+      "epoch": 5.891908091908092,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0002,
+      "loss": 0.8915,
+      "step": 7376
+    },
+    {
+      "epoch": 5.892707292707293,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0002,
+      "loss": 0.8831,
+      "step": 7377
+    },
+    {
+      "epoch": 5.893506493506494,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0002,
+      "loss": 0.893,
+      "step": 7378
+    },
+    {
+      "epoch": 5.894305694305694,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8856,
+      "step": 7379
+    },
+    {
+      "epoch": 5.895104895104895,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8872,
+      "step": 7380
+    },
+    {
+      "epoch": 5.895904095904096,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.0002,
+      "loss": 0.8847,
+      "step": 7381
+    },
+    {
+      "epoch": 5.896703296703297,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0002,
+      "loss": 0.899,
+      "step": 7382
+    },
+    {
+      "epoch": 5.897502497502497,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.0002,
+      "loss": 0.8854,
+      "step": 7383
+    },
+    {
+      "epoch": 5.898301698301698,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0002,
+      "loss": 0.8835,
+      "step": 7384
+    },
+    {
+      "epoch": 5.899100899100899,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0002,
+      "loss": 0.9088,
+      "step": 7385
+    },
+    {
+      "epoch": 5.8999000999000994,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0002,
+      "loss": 0.8832,
+      "step": 7386
+    },
+    {
+      "epoch": 5.9006993006993005,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.0002,
+      "loss": 0.8845,
+      "step": 7387
+    },
+    {
+      "epoch": 5.9014985014985015,
+      "grad_norm": 0.98828125,
+      "learning_rate": 0.0002,
+      "loss": 0.9055,
+      "step": 7388
+    },
+    {
+      "epoch": 5.902297702297703,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8915,
+      "step": 7389
+    },
+    {
+      "epoch": 5.903096903096903,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.0002,
+      "loss": 0.8877,
+      "step": 7390
+    },
+    {
+      "epoch": 5.903896103896104,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.0002,
+      "loss": 0.9017,
+      "step": 7391
+    },
+    {
+      "epoch": 5.904695304695305,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.0002,
+      "loss": 0.8901,
+      "step": 7392
+    },
+    {
+      "epoch": 5.905494505494506,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0002,
+      "loss": 0.888,
+      "step": 7393
+    },
+    {
+      "epoch": 5.906293706293706,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.0002,
+      "loss": 0.8946,
+      "step": 7394
+    },
+    {
+      "epoch": 5.907092907092907,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8948,
+      "step": 7395
+    },
+    {
+      "epoch": 5.907892107892108,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.0002,
+      "loss": 0.9026,
+      "step": 7396
+    },
+    {
+      "epoch": 5.908691308691308,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0002,
+      "loss": 0.8947,
+      "step": 7397
+    },
+    {
+      "epoch": 5.909490509490509,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.0002,
+      "loss": 0.8866,
+      "step": 7398
+    },
+    {
+      "epoch": 5.91028971028971,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0002,
+      "loss": 0.8984,
+      "step": 7399
+    },
+    {
+      "epoch": 5.911088911088911,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.0002,
+      "loss": 0.8944,
+      "step": 7400
+    },
+    {
+      "epoch": 5.911888111888112,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0002,
+      "loss": 0.899,
+      "step": 7401
+    },
+    {
+      "epoch": 5.9126873126873125,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.0002,
+      "loss": 0.8976,
+      "step": 7402
+    },
+    {
+      "epoch": 5.9134865134865136,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0002,
+      "loss": 0.8845,
+      "step": 7403
+    },
+    {
+      "epoch": 5.914285714285715,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0002,
+      "loss": 0.8921,
+      "step": 7404
+    },
+    {
+      "epoch": 5.915084915084915,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0002,
+      "loss": 0.9,
+      "step": 7405
+    },
+    {
+      "epoch": 5.915884115884116,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8975,
+      "step": 7406
+    },
+    {
+      "epoch": 5.916683316683317,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.0002,
+      "loss": 0.8912,
+      "step": 7407
+    },
+    {
+      "epoch": 5.917482517482518,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.0002,
+      "loss": 0.891,
+      "step": 7408
+    },
+    {
+      "epoch": 5.918281718281718,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8952,
+      "step": 7409
+    },
+    {
+      "epoch": 5.919080919080919,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0002,
+      "loss": 0.8932,
+      "step": 7410
+    },
+    {
+      "epoch": 5.91988011988012,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8975,
+      "step": 7411
+    },
+    {
+      "epoch": 5.920679320679321,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8859,
+      "step": 7412
+    },
+    {
+      "epoch": 5.921478521478521,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8971,
+      "step": 7413
+    },
+    {
+      "epoch": 5.922277722277722,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0002,
+      "loss": 0.8963,
+      "step": 7414
+    },
+    {
+      "epoch": 5.923076923076923,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0002,
+      "loss": 0.8985,
+      "step": 7415
+    },
+    {
+      "epoch": 5.9238761238761235,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8894,
+      "step": 7416
+    },
+    {
+      "epoch": 5.9246753246753245,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8903,
+      "step": 7417
+    },
+    {
+      "epoch": 5.925474525474526,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8852,
+      "step": 7418
+    },
+    {
+      "epoch": 5.926273726273727,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0002,
+      "loss": 0.886,
+      "step": 7419
+    },
+    {
+      "epoch": 5.927072927072927,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8874,
+      "step": 7420
+    },
+    {
+      "epoch": 5.927872127872128,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0002,
+      "loss": 0.8897,
+      "step": 7421
+    },
+    {
+      "epoch": 5.928671328671329,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.0002,
+      "loss": 0.8859,
+      "step": 7422
+    },
+    {
+      "epoch": 5.92947052947053,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0002,
+      "loss": 0.8812,
+      "step": 7423
+    },
+    {
+      "epoch": 5.93026973026973,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8895,
+      "step": 7424
+    },
+    {
+      "epoch": 5.931068931068931,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8789,
+      "step": 7425
+    },
+    {
+      "epoch": 5.931868131868132,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0002,
+      "loss": 0.9047,
+      "step": 7426
+    },
+    {
+      "epoch": 5.932667332667332,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0002,
+      "loss": 0.8863,
+      "step": 7427
+    },
+    {
+      "epoch": 5.933466533466533,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0002,
+      "loss": 0.8864,
+      "step": 7428
+    },
+    {
+      "epoch": 5.934265734265734,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0002,
+      "loss": 0.8885,
+      "step": 7429
+    },
+    {
+      "epoch": 5.935064935064935,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0002,
+      "loss": 0.8876,
+      "step": 7430
+    },
+    {
+      "epoch": 5.9358641358641355,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0002,
+      "loss": 0.8829,
+      "step": 7431
+    },
+    {
+      "epoch": 5.936663336663337,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0002,
+      "loss": 0.8954,
+      "step": 7432
+    },
+    {
+      "epoch": 5.937462537462538,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0002,
+      "loss": 0.9063,
+      "step": 7433
+    },
+    {
+      "epoch": 5.938261738261739,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0002,
+      "loss": 0.8804,
+      "step": 7434
+    },
+    {
+      "epoch": 5.939060939060939,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0002,
+      "loss": 0.8899,
+      "step": 7435
+    },
+    {
+      "epoch": 5.93986013986014,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8794,
+      "step": 7436
+    },
+    {
+      "epoch": 5.940659340659341,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.0002,
+      "loss": 0.8953,
+      "step": 7437
+    },
+    {
+      "epoch": 5.941458541458541,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8894,
+      "step": 7438
+    },
+    {
+      "epoch": 5.942257742257742,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.0002,
+      "loss": 0.8916,
+      "step": 7439
+    },
+    {
+      "epoch": 5.943056943056943,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.0002,
+      "loss": 0.8965,
+      "step": 7440
+    },
+    {
+      "epoch": 5.943856143856144,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0002,
+      "loss": 0.8829,
+      "step": 7441
+    },
+    {
+      "epoch": 5.944655344655344,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8949,
+      "step": 7442
+    },
+    {
+      "epoch": 5.945454545454545,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0002,
+      "loss": 0.8916,
+      "step": 7443
+    },
+    {
+      "epoch": 5.946253746253746,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.0002,
+      "loss": 0.8908,
+      "step": 7444
+    },
+    {
+      "epoch": 5.947052947052947,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9055,
+      "step": 7445
+    },
+    {
+      "epoch": 5.9478521478521476,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8794,
+      "step": 7446
+    },
+    {
+      "epoch": 5.948651348651349,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.0002,
+      "loss": 0.8881,
+      "step": 7447
+    },
+    {
+      "epoch": 5.94945054945055,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8989,
+      "step": 7448
+    },
+    {
+      "epoch": 5.95024975024975,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.0002,
+      "loss": 0.8869,
+      "step": 7449
+    },
+    {
+      "epoch": 5.951048951048951,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0002,
+      "loss": 0.8851,
+      "step": 7450
+    },
+    {
+      "epoch": 5.951848151848152,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0002,
+      "loss": 0.8916,
+      "step": 7451
+    },
+    {
+      "epoch": 5.952647352647353,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0002,
+      "loss": 0.8855,
+      "step": 7452
+    },
+    {
+      "epoch": 5.953446553446554,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.0002,
+      "loss": 0.8858,
+      "step": 7453
+    },
+    {
+      "epoch": 5.954245754245754,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0002,
+      "loss": 0.8912,
+      "step": 7454
+    },
+    {
+      "epoch": 5.955044955044955,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.0002,
+      "loss": 0.9086,
+      "step": 7455
+    },
+    {
+      "epoch": 5.955844155844156,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0002,
+      "loss": 0.8901,
+      "step": 7456
+    },
+    {
+      "epoch": 5.956643356643356,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0002,
+      "loss": 0.8887,
+      "step": 7457
+    },
+    {
+      "epoch": 5.957442557442557,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0002,
+      "loss": 0.8885,
+      "step": 7458
+    },
+    {
+      "epoch": 5.958241758241758,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0002,
+      "loss": 0.8903,
+      "step": 7459
+    },
+    {
+      "epoch": 5.959040959040959,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0002,
+      "loss": 0.8893,
+      "step": 7460
+    },
+    {
+      "epoch": 5.95984015984016,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8894,
+      "step": 7461
+    },
+    {
+      "epoch": 5.960639360639361,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8863,
+      "step": 7462
+    },
+    {
+      "epoch": 5.961438561438562,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8818,
+      "step": 7463
+    },
+    {
+      "epoch": 5.962237762237763,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0002,
+      "loss": 0.8948,
+      "step": 7464
+    },
+    {
+      "epoch": 5.963036963036963,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8915,
+      "step": 7465
+    },
+    {
+      "epoch": 5.963836163836164,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8864,
+      "step": 7466
+    },
+    {
+      "epoch": 5.964635364635365,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0002,
+      "loss": 0.8867,
+      "step": 7467
+    },
+    {
+      "epoch": 5.965434565434565,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0002,
+      "loss": 0.886,
+      "step": 7468
+    },
+    {
+      "epoch": 5.966233766233766,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0002,
+      "loss": 0.8816,
+      "step": 7469
+    },
+    {
+      "epoch": 5.967032967032967,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8848,
+      "step": 7470
+    },
+    {
+      "epoch": 5.967832167832168,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.0002,
+      "loss": 0.879,
+      "step": 7471
+    },
+    {
+      "epoch": 5.968631368631368,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8868,
+      "step": 7472
+    },
+    {
+      "epoch": 5.969430569430569,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8909,
+      "step": 7473
+    },
+    {
+      "epoch": 5.97022977022977,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0002,
+      "loss": 0.888,
+      "step": 7474
+    },
+    {
+      "epoch": 5.9710289710289715,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8988,
+      "step": 7475
+    },
+    {
+      "epoch": 5.971828171828172,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.0002,
+      "loss": 0.8906,
+      "step": 7476
+    },
+    {
+      "epoch": 5.972627372627373,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8876,
+      "step": 7477
+    },
+    {
+      "epoch": 5.973426573426574,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8896,
+      "step": 7478
+    },
+    {
+      "epoch": 5.974225774225774,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0002,
+      "loss": 0.8971,
+      "step": 7479
+    },
+    {
+      "epoch": 5.975024975024975,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0002,
+      "loss": 0.8847,
+      "step": 7480
+    },
+    {
+      "epoch": 5.975824175824176,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0002,
+      "loss": 0.891,
+      "step": 7481
+    },
+    {
+      "epoch": 5.976623376623377,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0002,
+      "loss": 0.8847,
+      "step": 7482
+    },
+    {
+      "epoch": 5.977422577422577,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8866,
+      "step": 7483
+    },
+    {
+      "epoch": 5.978221778221778,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8818,
+      "step": 7484
+    },
+    {
+      "epoch": 5.979020979020979,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0002,
+      "loss": 0.894,
+      "step": 7485
+    },
+    {
+      "epoch": 5.97982017982018,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0002,
+      "loss": 0.8874,
+      "step": 7486
+    },
+    {
+      "epoch": 5.98061938061938,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8986,
+      "step": 7487
+    },
+    {
+      "epoch": 5.981418581418581,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8869,
+      "step": 7488
+    },
+    {
+      "epoch": 5.982217782217782,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0002,
+      "loss": 0.8828,
+      "step": 7489
+    },
+    {
+      "epoch": 5.983016983016983,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.0002,
+      "loss": 0.8907,
+      "step": 7490
+    },
+    {
+      "epoch": 5.983816183816184,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0002,
+      "loss": 0.884,
+      "step": 7491
+    },
+    {
+      "epoch": 5.984615384615385,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8849,
+      "step": 7492
+    },
+    {
+      "epoch": 5.985414585414586,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0002,
+      "loss": 0.8948,
+      "step": 7493
+    },
+    {
+      "epoch": 5.986213786213786,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0002,
+      "loss": 0.8936,
+      "step": 7494
+    },
+    {
+      "epoch": 5.987012987012987,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8898,
+      "step": 7495
+    },
+    {
+      "epoch": 5.987812187812188,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0002,
+      "loss": 0.8845,
+      "step": 7496
+    },
+    {
+      "epoch": 5.988611388611389,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.0002,
+      "loss": 0.8893,
+      "step": 7497
+    },
+    {
+      "epoch": 5.989410589410589,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0002,
+      "loss": 0.8882,
+      "step": 7498
+    },
+    {
+      "epoch": 5.99020979020979,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8912,
+      "step": 7499
+    },
+    {
+      "epoch": 5.991008991008991,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0002,
+      "loss": 0.8942,
+      "step": 7500
+    },
+    {
+      "epoch": 5.991808191808191,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0002,
+      "loss": 0.893,
+      "step": 7501
+    },
+    {
+      "epoch": 5.992607392607392,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0002,
+      "loss": 0.886,
+      "step": 7502
+    },
+    {
+      "epoch": 5.993406593406593,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.0002,
+      "loss": 0.8961,
+      "step": 7503
+    },
+    {
+      "epoch": 5.9942057942057945,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8913,
+      "step": 7504
+    },
+    {
+      "epoch": 5.995004995004995,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.0002,
+      "loss": 0.8827,
+      "step": 7505
+    },
+    {
+      "epoch": 5.995804195804196,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.0002,
+      "loss": 0.8933,
+      "step": 7506
+    },
+    {
+      "epoch": 5.996603396603397,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0002,
+      "loss": 0.8786,
+      "step": 7507
+    },
+    {
+      "epoch": 5.997402597402598,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0002,
+      "loss": 0.8963,
+      "step": 7508
+    },
+    {
+      "epoch": 5.998201798201798,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0002,
+      "loss": 0.8834,
+      "step": 7509
+    },
+    {
+      "epoch": 5.999000999000999,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.0002,
+      "loss": 0.8844,
+      "step": 7510
+    },
+    {
+      "epoch": 5.9998001998002,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.0002,
+      "loss": 0.8969,
+      "step": 7511
+    },
+    {
+      "epoch": 6.0,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0002,
+      "loss": 0.2212,
+      "step": 7512
+    },
+    {
+      "epoch": 6.000799200799201,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8876,
+      "step": 7513
+    },
+    {
+      "epoch": 6.001598401598401,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0002,
+      "loss": 0.8784,
+      "step": 7514
+    },
+    {
+      "epoch": 6.002397602397602,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8956,
+      "step": 7515
+    },
+    {
+      "epoch": 6.003196803196803,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002,
+      "loss": 0.895,
+      "step": 7516
+    },
+    {
+      "epoch": 6.003996003996004,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0002,
+      "loss": 0.8909,
+      "step": 7517
+    },
+    {
+      "epoch": 6.0047952047952045,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8921,
+      "step": 7518
+    },
+    {
+      "epoch": 6.0055944055944055,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0002,
+      "loss": 0.8828,
+      "step": 7519
+    },
+    {
+      "epoch": 6.0063936063936065,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0002,
+      "loss": 0.9074,
+      "step": 7520
+    },
+    {
+      "epoch": 6.007192807192808,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0002,
+      "loss": 0.9376,
+      "step": 7521
+    },
+    {
+      "epoch": 6.007992007992008,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8857,
+      "step": 7522
+    },
+    {
+      "epoch": 6.008791208791209,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0002,
+      "loss": 0.8828,
+      "step": 7523
+    },
+    {
+      "epoch": 6.00959040959041,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8881,
+      "step": 7524
+    },
+    {
+      "epoch": 6.01038961038961,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8861,
+      "step": 7525
+    },
+    {
+      "epoch": 6.011188811188811,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8852,
+      "step": 7526
+    },
+    {
+      "epoch": 6.011988011988012,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8949,
+      "step": 7527
+    },
+    {
+      "epoch": 6.012787212787213,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8985,
+      "step": 7528
+    },
+    {
+      "epoch": 6.013586413586413,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0002,
+      "loss": 0.8901,
+      "step": 7529
+    },
+    {
+      "epoch": 6.014385614385614,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0002,
+      "loss": 0.8969,
+      "step": 7530
+    },
+    {
+      "epoch": 6.015184815184815,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8809,
+      "step": 7531
+    },
+    {
+      "epoch": 6.015984015984016,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0002,
+      "loss": 0.8965,
+      "step": 7532
+    },
+    {
+      "epoch": 6.0167832167832165,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8897,
+      "step": 7533
+    },
+    {
+      "epoch": 6.0175824175824175,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0002,
+      "loss": 0.895,
+      "step": 7534
+    },
+    {
+      "epoch": 6.018381618381619,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0002,
+      "loss": 0.8809,
+      "step": 7535
+    },
+    {
+      "epoch": 6.01918081918082,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0002,
+      "loss": 0.8905,
+      "step": 7536
+    },
+    {
+      "epoch": 6.01998001998002,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8868,
+      "step": 7537
+    },
+    {
+      "epoch": 6.020779220779221,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8778,
+      "step": 7538
+    },
+    {
+      "epoch": 6.021578421578422,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0002,
+      "loss": 0.883,
+      "step": 7539
+    },
+    {
+      "epoch": 6.022377622377622,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0002,
+      "loss": 0.8795,
+      "step": 7540
+    },
+    {
+      "epoch": 6.023176823176823,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0002,
+      "loss": 0.8932,
+      "step": 7541
+    },
+    {
+      "epoch": 6.023976023976024,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0002,
+      "loss": 0.9306,
+      "step": 7542
+    },
+    {
+      "epoch": 6.024775224775225,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0002,
+      "loss": 0.889,
+      "step": 7543
+    },
+    {
+      "epoch": 6.025574425574425,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8992,
+      "step": 7544
+    },
+    {
+      "epoch": 6.026373626373626,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0002,
+      "loss": 0.8872,
+      "step": 7545
+    },
+    {
+      "epoch": 6.027172827172827,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0002,
+      "loss": 0.8894,
+      "step": 7546
+    },
+    {
+      "epoch": 6.027972027972028,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0002,
+      "loss": 0.8942,
+      "step": 7547
+    },
+    {
+      "epoch": 6.0287712287712285,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8919,
+      "step": 7548
+    },
+    {
+      "epoch": 6.0295704295704295,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0002,
+      "loss": 0.8878,
+      "step": 7549
+    },
+    {
+      "epoch": 6.030369630369631,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8823,
+      "step": 7550
+    },
+    {
+      "epoch": 6.031168831168831,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8876,
+      "step": 7551
+    },
+    {
+      "epoch": 6.031968031968032,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8944,
+      "step": 7552
+    },
+    {
+      "epoch": 6.032767232767233,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0002,
+      "loss": 0.8885,
+      "step": 7553
+    },
+    {
+      "epoch": 6.033566433566434,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8818,
+      "step": 7554
+    },
+    {
+      "epoch": 6.034365634365634,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8849,
+      "step": 7555
+    },
+    {
+      "epoch": 6.035164835164835,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8942,
+      "step": 7556
+    },
+    {
+      "epoch": 6.035964035964036,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.0002,
+      "loss": 0.898,
+      "step": 7557
+    },
+    {
+      "epoch": 6.036763236763237,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8904,
+      "step": 7558
+    },
+    {
+      "epoch": 6.037562437562437,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.0002,
+      "loss": 0.8876,
+      "step": 7559
+    },
+    {
+      "epoch": 6.038361638361638,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8834,
+      "step": 7560
+    },
+    {
+      "epoch": 6.039160839160839,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8855,
+      "step": 7561
+    },
+    {
+      "epoch": 6.03996003996004,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0002,
+      "loss": 0.8862,
+      "step": 7562
+    },
+    {
+      "epoch": 6.0407592407592405,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.0002,
+      "loss": 0.8882,
+      "step": 7563
+    },
+    {
+      "epoch": 6.041558441558442,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8827,
+      "step": 7564
+    },
+    {
+      "epoch": 6.042357642357643,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0002,
+      "loss": 0.8862,
+      "step": 7565
+    },
+    {
+      "epoch": 6.043156843156843,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0002,
+      "loss": 0.8848,
+      "step": 7566
+    },
+    {
+      "epoch": 6.043956043956044,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0002,
+      "loss": 0.8896,
+      "step": 7567
+    },
+    {
+      "epoch": 6.044755244755245,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0002,
+      "loss": 0.8886,
+      "step": 7568
+    },
+    {
+      "epoch": 6.045554445554446,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8898,
+      "step": 7569
+    },
+    {
+      "epoch": 6.046353646353646,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8904,
+      "step": 7570
+    },
+    {
+      "epoch": 6.047152847152847,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8922,
+      "step": 7571
+    },
+    {
+      "epoch": 6.047952047952048,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0002,
+      "loss": 0.8885,
+      "step": 7572
+    },
+    {
+      "epoch": 6.048751248751249,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8937,
+      "step": 7573
+    },
+    {
+      "epoch": 6.049550449550449,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8896,
+      "step": 7574
+    },
+    {
+      "epoch": 6.05034965034965,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.0002,
+      "loss": 0.8861,
+      "step": 7575
+    },
+    {
+      "epoch": 6.051148851148851,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.0002,
+      "loss": 0.8872,
+      "step": 7576
+    },
+    {
+      "epoch": 6.0519480519480515,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0002,
+      "loss": 0.8932,
+      "step": 7577
+    },
+    {
+      "epoch": 6.052747252747253,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0002,
+      "loss": 0.8908,
+      "step": 7578
+    },
+    {
+      "epoch": 6.053546453546454,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0002,
+      "loss": 0.8892,
+      "step": 7579
+    },
+    {
+      "epoch": 6.054345654345655,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0002,
+      "loss": 0.8886,
+      "step": 7580
+    },
+    {
+      "epoch": 6.055144855144855,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0002,
+      "loss": 0.888,
+      "step": 7581
+    },
+    {
+      "epoch": 6.055944055944056,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8888,
+      "step": 7582
+    },
+    {
+      "epoch": 6.056743256743257,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0002,
+      "loss": 0.891,
+      "step": 7583
+    },
+    {
+      "epoch": 6.057542457542458,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0002,
+      "loss": 0.8787,
+      "step": 7584
+    },
+    {
+      "epoch": 6.058341658341658,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0002,
+      "loss": 0.8898,
+      "step": 7585
+    },
+    {
+      "epoch": 6.059140859140859,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8946,
+      "step": 7586
+    },
+    {
+      "epoch": 6.05994005994006,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8906,
+      "step": 7587
+    },
+    {
+      "epoch": 6.060739260739261,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0002,
+      "loss": 0.8865,
+      "step": 7588
+    },
+    {
+      "epoch": 6.061538461538461,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0002,
+      "loss": 0.8855,
+      "step": 7589
+    },
+    {
+      "epoch": 6.062337662337662,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8848,
+      "step": 7590
+    },
+    {
+      "epoch": 6.063136863136863,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8971,
+      "step": 7591
+    },
+    {
+      "epoch": 6.0639360639360635,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0002,
+      "loss": 0.8912,
+      "step": 7592
+    },
+    {
+      "epoch": 6.064735264735265,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0002,
+      "loss": 0.8962,
+      "step": 7593
+    },
+    {
+      "epoch": 6.065534465534466,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8962,
+      "step": 7594
+    },
+    {
+      "epoch": 6.066333666333667,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0002,
+      "loss": 0.8866,
+      "step": 7595
+    },
+    {
+      "epoch": 6.067132867132867,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0002,
+      "loss": 0.8936,
+      "step": 7596
+    },
+    {
+      "epoch": 6.067932067932068,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0002,
+      "loss": 0.8914,
+      "step": 7597
+    },
+    {
+      "epoch": 6.068731268731269,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0002,
+      "loss": 0.8748,
+      "step": 7598
+    },
+    {
+      "epoch": 6.06953046953047,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0002,
+      "loss": 0.8917,
+      "step": 7599
+    },
+    {
+      "epoch": 6.07032967032967,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8932,
+      "step": 7600
+    },
+    {
+      "epoch": 6.071128871128871,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0002,
+      "loss": 0.8929,
+      "step": 7601
+    },
+    {
+      "epoch": 6.071928071928072,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0002,
+      "loss": 0.8884,
+      "step": 7602
+    },
+    {
+      "epoch": 6.072727272727272,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0002,
+      "loss": 0.8896,
+      "step": 7603
+    },
+    {
+      "epoch": 6.073526473526473,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.0002,
+      "loss": 0.8787,
+      "step": 7604
+    },
+    {
+      "epoch": 6.074325674325674,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0002,
+      "loss": 0.89,
+      "step": 7605
+    },
+    {
+      "epoch": 6.075124875124875,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0002,
+      "loss": 0.8782,
+      "step": 7606
+    },
+    {
+      "epoch": 6.075924075924076,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8919,
+      "step": 7607
+    },
+    {
+      "epoch": 6.076723276723277,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8909,
+      "step": 7608
+    },
+    {
+      "epoch": 6.077522477522478,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8844,
+      "step": 7609
+    },
+    {
+      "epoch": 6.078321678321679,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8884,
+      "step": 7610
+    },
+    {
+      "epoch": 6.079120879120879,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0002,
+      "loss": 0.8829,
+      "step": 7611
+    },
+    {
+      "epoch": 6.07992007992008,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0002,
+      "loss": 0.8937,
+      "step": 7612
+    },
+    {
+      "epoch": 6.080719280719281,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8812,
+      "step": 7613
+    },
+    {
+      "epoch": 6.081518481518482,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0002,
+      "loss": 0.89,
+      "step": 7614
+    },
+    {
+      "epoch": 6.082317682317682,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8917,
+      "step": 7615
+    },
+    {
+      "epoch": 6.083116883116883,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8885,
+      "step": 7616
+    },
+    {
+      "epoch": 6.083916083916084,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8979,
+      "step": 7617
+    },
+    {
+      "epoch": 6.084715284715284,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.0002,
+      "loss": 0.8878,
+      "step": 7618
+    },
+    {
+      "epoch": 6.085514485514485,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0002,
+      "loss": 0.8872,
+      "step": 7619
+    },
+    {
+      "epoch": 6.086313686313686,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0002,
+      "loss": 0.8873,
+      "step": 7620
+    },
+    {
+      "epoch": 6.0871128871128874,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8827,
+      "step": 7621
+    },
+    {
+      "epoch": 6.087912087912088,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0002,
+      "loss": 0.892,
+      "step": 7622
+    },
+    {
+      "epoch": 6.088711288711289,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8863,
+      "step": 7623
+    },
+    {
+      "epoch": 6.08951048951049,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8982,
+      "step": 7624
+    },
+    {
+      "epoch": 6.090309690309691,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0002,
+      "loss": 0.8869,
+      "step": 7625
+    },
+    {
+      "epoch": 6.091108891108891,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0002,
+      "loss": 0.8926,
+      "step": 7626
+    },
+    {
+      "epoch": 6.091908091908092,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8954,
+      "step": 7627
+    },
+    {
+      "epoch": 6.092707292707293,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8857,
+      "step": 7628
+    },
+    {
+      "epoch": 6.093506493506493,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.0002,
+      "loss": 0.8864,
+      "step": 7629
+    },
+    {
+      "epoch": 6.094305694305694,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0002,
+      "loss": 0.892,
+      "step": 7630
+    },
+    {
+      "epoch": 6.095104895104895,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9001,
+      "step": 7631
+    },
+    {
+      "epoch": 6.095904095904096,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8868,
+      "step": 7632
+    },
+    {
+      "epoch": 6.096703296703296,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0002,
+      "loss": 0.8853,
+      "step": 7633
+    },
+    {
+      "epoch": 6.097502497502497,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0002,
+      "loss": 0.8918,
+      "step": 7634
+    },
+    {
+      "epoch": 6.098301698301698,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8898,
+      "step": 7635
+    },
+    {
+      "epoch": 6.0991008991008995,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0002,
+      "loss": 0.8878,
+      "step": 7636
+    },
+    {
+      "epoch": 6.0999000999001,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8811,
+      "step": 7637
+    },
+    {
+      "epoch": 6.100699300699301,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8822,
+      "step": 7638
+    },
+    {
+      "epoch": 6.101498501498502,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.0002,
+      "loss": 0.8917,
+      "step": 7639
+    },
+    {
+      "epoch": 6.102297702297703,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0002,
+      "loss": 0.89,
+      "step": 7640
+    },
+    {
+      "epoch": 6.103096903096903,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.0002,
+      "loss": 0.8858,
+      "step": 7641
+    },
+    {
+      "epoch": 6.103896103896104,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8992,
+      "step": 7642
+    },
+    {
+      "epoch": 6.104695304695305,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8895,
+      "step": 7643
+    },
+    {
+      "epoch": 6.105494505494505,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8915,
+      "step": 7644
+    },
+    {
+      "epoch": 6.106293706293706,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0002,
+      "loss": 0.89,
+      "step": 7645
+    },
+    {
+      "epoch": 6.107092907092907,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0002,
+      "loss": 0.8943,
+      "step": 7646
+    },
+    {
+      "epoch": 6.107892107892108,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.0002,
+      "loss": 0.8975,
+      "step": 7647
+    },
+    {
+      "epoch": 6.108691308691308,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0002,
+      "loss": 0.8904,
+      "step": 7648
+    },
+    {
+      "epoch": 6.109490509490509,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8915,
+      "step": 7649
+    },
+    {
+      "epoch": 6.1102897102897105,
+      "grad_norm": 0.375,
+      "learning_rate": 0.0002,
+      "loss": 0.9039,
+      "step": 7650
+    },
+    {
+      "epoch": 6.1110889110889115,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8926,
+      "step": 7651
+    },
+    {
+      "epoch": 6.111888111888112,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8918,
+      "step": 7652
+    },
+    {
+      "epoch": 6.112687312687313,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8873,
+      "step": 7653
+    },
+    {
+      "epoch": 6.113486513486514,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8915,
+      "step": 7654
+    },
+    {
+      "epoch": 6.114285714285714,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8981,
+      "step": 7655
+    },
+    {
+      "epoch": 6.115084915084915,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.0002,
+      "loss": 0.91,
+      "step": 7656
+    },
+    {
+      "epoch": 6.115884115884116,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8887,
+      "step": 7657
+    },
+    {
+      "epoch": 6.116683316683317,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0002,
+      "loss": 0.8916,
+      "step": 7658
+    },
+    {
+      "epoch": 6.117482517482517,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.0002,
+      "loss": 0.8976,
+      "step": 7659
+    },
+    {
+      "epoch": 6.118281718281718,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0002,
+      "loss": 0.894,
+      "step": 7660
+    },
+    {
+      "epoch": 6.119080919080919,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0002,
+      "loss": 0.8941,
+      "step": 7661
+    },
+    {
+      "epoch": 6.11988011988012,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8819,
+      "step": 7662
+    },
+    {
+      "epoch": 6.12067932067932,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8904,
+      "step": 7663
+    },
+    {
+      "epoch": 6.1214785214785215,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8903,
+      "step": 7664
+    },
+    {
+      "epoch": 6.1222777222777225,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8832,
+      "step": 7665
+    },
+    {
+      "epoch": 6.123076923076923,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0002,
+      "loss": 0.8905,
+      "step": 7666
+    },
+    {
+      "epoch": 6.123876123876124,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8905,
+      "step": 7667
+    },
+    {
+      "epoch": 6.124675324675325,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8897,
+      "step": 7668
+    },
+    {
+      "epoch": 6.125474525474526,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8873,
+      "step": 7669
+    },
+    {
+      "epoch": 6.126273726273726,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.0002,
+      "loss": 0.8898,
+      "step": 7670
+    },
+    {
+      "epoch": 6.127072927072927,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8847,
+      "step": 7671
+    },
+    {
+      "epoch": 6.127872127872128,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8913,
+      "step": 7672
+    },
+    {
+      "epoch": 6.128671328671329,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0002,
+      "loss": 0.8897,
+      "step": 7673
+    },
+    {
+      "epoch": 6.129470529470529,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8963,
+      "step": 7674
+    },
+    {
+      "epoch": 6.13026973026973,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8851,
+      "step": 7675
+    },
+    {
+      "epoch": 6.131068931068931,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.0002,
+      "loss": 0.9035,
+      "step": 7676
+    },
+    {
+      "epoch": 6.131868131868132,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8937,
+      "step": 7677
+    },
+    {
+      "epoch": 6.132667332667332,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8845,
+      "step": 7678
+    },
+    {
+      "epoch": 6.1334665334665335,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8941,
+      "step": 7679
+    },
+    {
+      "epoch": 6.1342657342657345,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0002,
+      "loss": 0.8879,
+      "step": 7680
+    },
+    {
+      "epoch": 6.135064935064935,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0002,
+      "loss": 0.8981,
+      "step": 7681
+    },
+    {
+      "epoch": 6.135864135864136,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0002,
+      "loss": 0.892,
+      "step": 7682
+    },
+    {
+      "epoch": 6.136663336663337,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8917,
+      "step": 7683
+    },
+    {
+      "epoch": 6.137462537462538,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0002,
+      "loss": 0.8986,
+      "step": 7684
+    },
+    {
+      "epoch": 6.138261738261738,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0002,
+      "loss": 0.8892,
+      "step": 7685
+    },
+    {
+      "epoch": 6.139060939060939,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8858,
+      "step": 7686
+    },
+    {
+      "epoch": 6.13986013986014,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0002,
+      "loss": 0.8901,
+      "step": 7687
+    },
+    {
+      "epoch": 6.140659340659341,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0002,
+      "loss": 0.895,
+      "step": 7688
+    },
+    {
+      "epoch": 6.141458541458541,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0002,
+      "loss": 0.8892,
+      "step": 7689
+    },
+    {
+      "epoch": 6.142257742257742,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0002,
+      "loss": 0.8808,
+      "step": 7690
+    },
+    {
+      "epoch": 6.143056943056943,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0002,
+      "loss": 0.8989,
+      "step": 7691
+    },
+    {
+      "epoch": 6.143856143856144,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0002,
+      "loss": 0.8877,
+      "step": 7692
+    },
+    {
+      "epoch": 6.1446553446553445,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.0002,
+      "loss": 0.8881,
+      "step": 7693
+    },
+    {
+      "epoch": 6.1454545454545455,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8805,
+      "step": 7694
+    },
+    {
+      "epoch": 6.1462537462537465,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8903,
+      "step": 7695
+    },
+    {
+      "epoch": 6.147052947052947,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0002,
+      "loss": 0.8894,
+      "step": 7696
+    },
+    {
+      "epoch": 6.147852147852148,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0002,
+      "loss": 0.9,
+      "step": 7697
+    },
+    {
+      "epoch": 6.148651348651349,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0002,
+      "loss": 0.8888,
+      "step": 7698
+    },
+    {
+      "epoch": 6.14945054945055,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0002,
+      "loss": 0.9015,
+      "step": 7699
+    },
+    {
+      "epoch": 6.15024975024975,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8921,
+      "step": 7700
+    },
+    {
+      "epoch": 6.151048951048951,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8847,
+      "step": 7701
+    },
+    {
+      "epoch": 6.151848151848152,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8971,
+      "step": 7702
+    },
+    {
+      "epoch": 6.152647352647353,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8862,
+      "step": 7703
+    },
+    {
+      "epoch": 6.153446553446553,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.0002,
+      "loss": 0.89,
+      "step": 7704
+    },
+    {
+      "epoch": 6.154245754245754,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8859,
+      "step": 7705
+    },
+    {
+      "epoch": 6.155044955044955,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8799,
+      "step": 7706
+    },
+    {
+      "epoch": 6.1558441558441555,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0002,
+      "loss": 0.8932,
+      "step": 7707
+    },
+    {
+      "epoch": 6.1566433566433565,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8861,
+      "step": 7708
+    },
+    {
+      "epoch": 6.1574425574425575,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8851,
+      "step": 7709
+    },
+    {
+      "epoch": 6.158241758241759,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0002,
+      "loss": 0.8873,
+      "step": 7710
+    },
+    {
+      "epoch": 6.159040959040959,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8877,
+      "step": 7711
+    },
+    {
+      "epoch": 6.15984015984016,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8839,
+      "step": 7712
+    },
+    {
+      "epoch": 6.160639360639361,
+      "grad_norm": 0.375,
+      "learning_rate": 0.0002,
+      "loss": 0.88,
+      "step": 7713
+    },
+    {
+      "epoch": 6.161438561438562,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002,
+      "loss": 0.894,
+      "step": 7714
+    },
+    {
+      "epoch": 6.162237762237762,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0002,
+      "loss": 0.887,
+      "step": 7715
+    },
+    {
+      "epoch": 6.163036963036963,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.0002,
+      "loss": 0.8848,
+      "step": 7716
+    },
+    {
+      "epoch": 6.163836163836164,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8956,
+      "step": 7717
+    },
+    {
+      "epoch": 6.164635364635364,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8789,
+      "step": 7718
+    },
+    {
+      "epoch": 6.165434565434565,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0002,
+      "loss": 0.8916,
+      "step": 7719
+    },
+    {
+      "epoch": 6.166233766233766,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8858,
+      "step": 7720
+    },
+    {
+      "epoch": 6.167032967032967,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0002,
+      "loss": 0.8896,
+      "step": 7721
+    },
+    {
+      "epoch": 6.1678321678321675,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8876,
+      "step": 7722
+    },
+    {
+      "epoch": 6.1686313686313685,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0002,
+      "loss": 0.8859,
+      "step": 7723
+    },
+    {
+      "epoch": 6.1694305694305696,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0002,
+      "loss": 0.889,
+      "step": 7724
+    },
+    {
+      "epoch": 6.170229770229771,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8855,
+      "step": 7725
+    },
+    {
+      "epoch": 6.171028971028971,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0002,
+      "loss": 0.886,
+      "step": 7726
+    },
+    {
+      "epoch": 6.171828171828172,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8944,
+      "step": 7727
+    },
+    {
+      "epoch": 6.172627372627373,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8917,
+      "step": 7728
+    },
+    {
+      "epoch": 6.173426573426573,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8859,
+      "step": 7729
+    },
+    {
+      "epoch": 6.174225774225774,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8873,
+      "step": 7730
+    },
+    {
+      "epoch": 6.175024975024975,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0002,
+      "loss": 0.8931,
+      "step": 7731
+    },
+    {
+      "epoch": 6.175824175824176,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0002,
+      "loss": 0.8844,
+      "step": 7732
+    },
+    {
+      "epoch": 6.176623376623376,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0002,
+      "loss": 0.8882,
+      "step": 7733
+    },
+    {
+      "epoch": 6.177422577422577,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0002,
+      "loss": 0.9,
+      "step": 7734
+    },
+    {
+      "epoch": 6.178221778221778,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0002,
+      "loss": 0.8948,
+      "step": 7735
+    },
+    {
+      "epoch": 6.179020979020979,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0002,
+      "loss": 0.8885,
+      "step": 7736
+    },
+    {
+      "epoch": 6.1798201798201795,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0002,
+      "loss": 0.8937,
+      "step": 7737
+    },
+    {
+      "epoch": 6.1806193806193805,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.0002,
+      "loss": 0.8852,
+      "step": 7738
+    },
+    {
+      "epoch": 6.181418581418582,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0002,
+      "loss": 0.8891,
+      "step": 7739
+    },
+    {
+      "epoch": 6.182217782217783,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0002,
+      "loss": 0.8935,
+      "step": 7740
+    },
+    {
+      "epoch": 6.183016983016983,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0002,
+      "loss": 0.8951,
+      "step": 7741
+    },
+    {
+      "epoch": 6.183816183816184,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0002,
+      "loss": 0.8874,
+      "step": 7742
+    },
+    {
+      "epoch": 6.184615384615385,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8857,
+      "step": 7743
+    },
+    {
+      "epoch": 6.185414585414585,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8786,
+      "step": 7744
+    },
+    {
+      "epoch": 6.186213786213786,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8835,
+      "step": 7745
+    },
+    {
+      "epoch": 6.187012987012987,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8935,
+      "step": 7746
+    },
+    {
+      "epoch": 6.187812187812188,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8967,
+      "step": 7747
+    },
+    {
+      "epoch": 6.188611388611388,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8892,
+      "step": 7748
+    },
+    {
+      "epoch": 6.189410589410589,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8894,
+      "step": 7749
+    },
+    {
+      "epoch": 6.19020979020979,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.0002,
+      "loss": 0.8831,
+      "step": 7750
+    },
+    {
+      "epoch": 6.191008991008991,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.0002,
+      "loss": 0.9078,
+      "step": 7751
+    },
+    {
+      "epoch": 6.1918081918081915,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8784,
+      "step": 7752
+    },
+    {
+      "epoch": 6.192607392607393,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8882,
+      "step": 7753
+    },
+    {
+      "epoch": 6.193406593406594,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8908,
+      "step": 7754
+    },
+    {
+      "epoch": 6.194205794205795,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8907,
+      "step": 7755
+    },
+    {
+      "epoch": 6.195004995004995,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0002,
+      "loss": 0.8839,
+      "step": 7756
+    },
+    {
+      "epoch": 6.195804195804196,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0002,
+      "loss": 0.89,
+      "step": 7757
+    },
+    {
+      "epoch": 6.196603396603397,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0002,
+      "loss": 0.884,
+      "step": 7758
+    },
+    {
+      "epoch": 6.197402597402597,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0002,
+      "loss": 0.8818,
+      "step": 7759
+    },
+    {
+      "epoch": 6.198201798201798,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8965,
+      "step": 7760
+    },
+    {
+      "epoch": 6.199000999000999,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8915,
+      "step": 7761
+    },
+    {
+      "epoch": 6.1998001998002,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0002,
+      "loss": 0.8978,
+      "step": 7762
+    },
+    {
+      "epoch": 6.2005994005994,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8919,
+      "step": 7763
+    },
+    {
+      "epoch": 6.201398601398601,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0002,
+      "loss": 0.8843,
+      "step": 7764
+    },
+    {
+      "epoch": 6.202197802197802,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0002,
+      "loss": 0.881,
+      "step": 7765
+    },
+    {
+      "epoch": 6.202997002997003,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8731,
+      "step": 7766
+    },
+    {
+      "epoch": 6.203796203796204,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8912,
+      "step": 7767
+    },
+    {
+      "epoch": 6.204595404595405,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0002,
+      "loss": 0.8854,
+      "step": 7768
+    },
+    {
+      "epoch": 6.205394605394606,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0002,
+      "loss": 0.8917,
+      "step": 7769
+    },
+    {
+      "epoch": 6.206193806193806,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0002,
+      "loss": 0.8864,
+      "step": 7770
+    },
+    {
+      "epoch": 6.206993006993007,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0002,
+      "loss": 0.8864,
+      "step": 7771
+    },
+    {
+      "epoch": 6.207792207792208,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8927,
+      "step": 7772
+    },
+    {
+      "epoch": 6.208591408591409,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8878,
+      "step": 7773
+    },
+    {
+      "epoch": 6.209390609390609,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8852,
+      "step": 7774
+    },
+    {
+      "epoch": 6.21018981018981,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0002,
+      "loss": 0.8888,
+      "step": 7775
+    },
+    {
+      "epoch": 6.210989010989011,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0002,
+      "loss": 0.8903,
+      "step": 7776
+    },
+    {
+      "epoch": 6.211788211788212,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8849,
+      "step": 7777
+    },
+    {
+      "epoch": 6.212587412587412,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0002,
+      "loss": 0.8957,
+      "step": 7778
+    },
+    {
+      "epoch": 6.213386613386613,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0002,
+      "loss": 0.8979,
+      "step": 7779
+    },
+    {
+      "epoch": 6.214185814185814,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8923,
+      "step": 7780
+    },
+    {
+      "epoch": 6.2149850149850145,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0002,
+      "loss": 0.8823,
+      "step": 7781
+    },
+    {
+      "epoch": 6.215784215784216,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0002,
+      "loss": 0.8778,
+      "step": 7782
+    },
+    {
+      "epoch": 6.216583416583417,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0002,
+      "loss": 0.8891,
+      "step": 7783
+    },
+    {
+      "epoch": 6.217382617382618,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.0002,
+      "loss": 0.8847,
+      "step": 7784
+    },
+    {
+      "epoch": 6.218181818181818,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0002,
+      "loss": 0.8923,
+      "step": 7785
+    },
+    {
+      "epoch": 6.218981018981019,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0002,
+      "loss": 0.8977,
+      "step": 7786
+    },
+    {
+      "epoch": 6.21978021978022,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.0002,
+      "loss": 0.8858,
+      "step": 7787
+    },
+    {
+      "epoch": 6.220579420579421,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8914,
+      "step": 7788
+    },
+    {
+      "epoch": 6.221378621378621,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.0002,
+      "loss": 0.8954,
+      "step": 7789
+    },
+    {
+      "epoch": 6.222177822177822,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0002,
+      "loss": 0.8832,
+      "step": 7790
+    },
+    {
+      "epoch": 6.222977022977023,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0002,
+      "loss": 0.8779,
+      "step": 7791
+    },
+    {
+      "epoch": 6.223776223776224,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0002,
+      "loss": 0.8879,
+      "step": 7792
+    },
+    {
+      "epoch": 6.224575424575424,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0002,
+      "loss": 0.8842,
+      "step": 7793
+    },
+    {
+      "epoch": 6.225374625374625,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0002,
+      "loss": 0.8818,
+      "step": 7794
+    },
+    {
+      "epoch": 6.226173826173826,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0002,
+      "loss": 0.8804,
+      "step": 7795
+    },
+    {
+      "epoch": 6.226973026973027,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0002,
+      "loss": 0.8837,
+      "step": 7796
+    },
+    {
+      "epoch": 6.227772227772228,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0002,
+      "loss": 0.886,
+      "step": 7797
+    },
+    {
+      "epoch": 6.228571428571429,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8898,
+      "step": 7798
+    },
+    {
+      "epoch": 6.22937062937063,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8894,
+      "step": 7799
+    },
+    {
+      "epoch": 6.23016983016983,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8845,
+      "step": 7800
+    },
+    {
+      "epoch": 6.230969030969031,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0002,
+      "loss": 0.8868,
+      "step": 7801
+    },
+    {
+      "epoch": 6.231768231768232,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0002,
+      "loss": 0.8886,
+      "step": 7802
+    },
+    {
+      "epoch": 6.232567432567433,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8921,
+      "step": 7803
+    },
+    {
+      "epoch": 6.233366633366633,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0002,
+      "loss": 0.882,
+      "step": 7804
+    },
+    {
+      "epoch": 6.234165834165834,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0002,
+      "loss": 0.8783,
+      "step": 7805
+    },
+    {
+      "epoch": 6.234965034965035,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0002,
+      "loss": 0.8915,
+      "step": 7806
+    },
+    {
+      "epoch": 6.235764235764236,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0002,
+      "loss": 0.8897,
+      "step": 7807
+    },
+    {
+      "epoch": 6.236563436563436,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8808,
+      "step": 7808
+    },
+    {
+      "epoch": 6.237362637362637,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8844,
+      "step": 7809
+    },
+    {
+      "epoch": 6.2381618381618384,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8782,
+      "step": 7810
+    },
+    {
+      "epoch": 6.238961038961039,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0002,
+      "loss": 0.934,
+      "step": 7811
+    },
+    {
+      "epoch": 6.23976023976024,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0002,
+      "loss": 0.8891,
+      "step": 7812
+    },
+    {
+      "epoch": 6.240559440559441,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8869,
+      "step": 7813
+    },
+    {
+      "epoch": 6.241358641358642,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0002,
+      "loss": 0.8901,
+      "step": 7814
+    },
+    {
+      "epoch": 6.242157842157842,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0002,
+      "loss": 0.8878,
+      "step": 7815
+    },
+    {
+      "epoch": 6.242957042957043,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8849,
+      "step": 7816
+    },
+    {
+      "epoch": 6.243756243756244,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8936,
+      "step": 7817
+    },
+    {
+      "epoch": 6.244555444555445,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0002,
+      "loss": 0.8873,
+      "step": 7818
+    },
+    {
+      "epoch": 6.245354645354645,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8878,
+      "step": 7819
+    },
+    {
+      "epoch": 6.246153846153846,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8846,
+      "step": 7820
+    },
+    {
+      "epoch": 6.246953046953047,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8929,
+      "step": 7821
+    },
+    {
+      "epoch": 6.247752247752247,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8905,
+      "step": 7822
+    },
+    {
+      "epoch": 6.248551448551448,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0002,
+      "loss": 0.8958,
+      "step": 7823
+    },
+    {
+      "epoch": 6.249350649350649,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0002,
+      "loss": 0.8846,
+      "step": 7824
+    },
+    {
+      "epoch": 6.2501498501498505,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8927,
+      "step": 7825
+    },
+    {
+      "epoch": 6.250949050949051,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8874,
+      "step": 7826
+    },
+    {
+      "epoch": 6.251748251748252,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0002,
+      "loss": 0.888,
+      "step": 7827
+    },
+    {
+      "epoch": 6.252547452547453,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.0002,
+      "loss": 0.884,
+      "step": 7828
+    },
+    {
+      "epoch": 6.253346653346654,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0002,
+      "loss": 0.8971,
+      "step": 7829
+    },
+    {
+      "epoch": 6.254145854145854,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0002,
+      "loss": 0.881,
+      "step": 7830
+    },
+    {
+      "epoch": 6.254945054945055,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8842,
+      "step": 7831
+    },
+    {
+      "epoch": 6.255744255744256,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0002,
+      "loss": 0.888,
+      "step": 7832
+    },
+    {
+      "epoch": 6.256543456543456,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0002,
+      "loss": 0.8948,
+      "step": 7833
+    },
+    {
+      "epoch": 6.257342657342657,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0002,
+      "loss": 0.8817,
+      "step": 7834
+    },
+    {
+      "epoch": 6.258141858141858,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8861,
+      "step": 7835
+    },
+    {
+      "epoch": 6.258941058941059,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0002,
+      "loss": 0.8842,
+      "step": 7836
+    },
+    {
+      "epoch": 6.259740259740259,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8899,
+      "step": 7837
+    },
+    {
+      "epoch": 6.26053946053946,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0002,
+      "loss": 0.8863,
+      "step": 7838
+    },
+    {
+      "epoch": 6.2613386613386615,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0002,
+      "loss": 0.8942,
+      "step": 7839
+    },
+    {
+      "epoch": 6.2621378621378625,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8908,
+      "step": 7840
+    },
+    {
+      "epoch": 6.262937062937063,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8929,
+      "step": 7841
+    },
+    {
+      "epoch": 6.263736263736264,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0002,
+      "loss": 0.8934,
+      "step": 7842
+    },
+    {
+      "epoch": 6.264535464535465,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8909,
+      "step": 7843
+    },
+    {
+      "epoch": 6.265334665334665,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0002,
+      "loss": 0.8868,
+      "step": 7844
+    },
+    {
+      "epoch": 6.266133866133866,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0002,
+      "loss": 0.8932,
+      "step": 7845
+    },
+    {
+      "epoch": 6.266933066933067,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8835,
+      "step": 7846
+    },
+    {
+      "epoch": 6.267732267732268,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8835,
+      "step": 7847
+    },
+    {
+      "epoch": 6.268531468531468,
+      "grad_norm": 1.4140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9016,
+      "step": 7848
+    },
+    {
+      "epoch": 6.269330669330669,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8895,
+      "step": 7849
+    },
+    {
+      "epoch": 6.27012987012987,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0002,
+      "loss": 0.8923,
+      "step": 7850
+    },
+    {
+      "epoch": 6.270929070929071,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8877,
+      "step": 7851
+    },
+    {
+      "epoch": 6.271728271728271,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0002,
+      "loss": 0.884,
+      "step": 7852
+    },
+    {
+      "epoch": 6.2725274725274724,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8849,
+      "step": 7853
+    },
+    {
+      "epoch": 6.2733266733266735,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8868,
+      "step": 7854
+    },
+    {
+      "epoch": 6.2741258741258745,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0002,
+      "loss": 0.8952,
+      "step": 7855
+    },
+    {
+      "epoch": 6.274925074925075,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8912,
+      "step": 7856
+    },
+    {
+      "epoch": 6.275724275724276,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0002,
+      "loss": 0.899,
+      "step": 7857
+    },
+    {
+      "epoch": 6.276523476523477,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8861,
+      "step": 7858
+    },
+    {
+      "epoch": 6.277322677322678,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0002,
+      "loss": 0.8952,
+      "step": 7859
+    },
+    {
+      "epoch": 6.278121878121878,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0002,
+      "loss": 0.8824,
+      "step": 7860
+    },
+    {
+      "epoch": 6.278921078921079,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.0002,
+      "loss": 0.8861,
+      "step": 7861
+    },
+    {
+      "epoch": 6.27972027972028,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8933,
+      "step": 7862
+    },
+    {
+      "epoch": 6.28051948051948,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8832,
+      "step": 7863
+    },
+    {
+      "epoch": 6.281318681318681,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8832,
+      "step": 7864
+    },
+    {
+      "epoch": 6.282117882117882,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.0002,
+      "loss": 0.9002,
+      "step": 7865
+    },
+    {
+      "epoch": 6.282917082917083,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8996,
+      "step": 7866
+    },
+    {
+      "epoch": 6.283716283716283,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0002,
+      "loss": 0.894,
+      "step": 7867
+    },
+    {
+      "epoch": 6.2845154845154845,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0002,
+      "loss": 0.8939,
+      "step": 7868
+    },
+    {
+      "epoch": 6.2853146853146855,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8915,
+      "step": 7869
+    },
+    {
+      "epoch": 6.2861138861138866,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0002,
+      "loss": 0.8833,
+      "step": 7870
+    },
+    {
+      "epoch": 6.286913086913087,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8954,
+      "step": 7871
+    },
+    {
+      "epoch": 6.287712287712288,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9026,
+      "step": 7872
+    },
+    {
+      "epoch": 6.288511488511489,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8894,
+      "step": 7873
+    },
+    {
+      "epoch": 6.289310689310689,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8873,
+      "step": 7874
+    },
+    {
+      "epoch": 6.29010989010989,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8966,
+      "step": 7875
+    },
+    {
+      "epoch": 6.290909090909091,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8958,
+      "step": 7876
+    },
+    {
+      "epoch": 6.291708291708292,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8928,
+      "step": 7877
+    },
+    {
+      "epoch": 6.292507492507492,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0002,
+      "loss": 0.8874,
+      "step": 7878
+    },
+    {
+      "epoch": 6.293306693306693,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0002,
+      "loss": 0.8835,
+      "step": 7879
+    },
+    {
+      "epoch": 6.294105894105894,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8932,
+      "step": 7880
+    },
+    {
+      "epoch": 6.294905094905095,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0002,
+      "loss": 0.8885,
+      "step": 7881
+    },
+    {
+      "epoch": 6.2957042957042955,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.0002,
+      "loss": 0.8841,
+      "step": 7882
+    },
+    {
+      "epoch": 6.2965034965034965,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8887,
+      "step": 7883
+    },
+    {
+      "epoch": 6.2973026973026975,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0002,
+      "loss": 0.8943,
+      "step": 7884
+    },
+    {
+      "epoch": 6.298101898101898,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0002,
+      "loss": 0.8887,
+      "step": 7885
+    },
+    {
+      "epoch": 6.298901098901099,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0002,
+      "loss": 0.882,
+      "step": 7886
+    },
+    {
+      "epoch": 6.2997002997003,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8972,
+      "step": 7887
+    },
+    {
+      "epoch": 6.300499500499501,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.0002,
+      "loss": 0.893,
+      "step": 7888
+    },
+    {
+      "epoch": 6.301298701298701,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0002,
+      "loss": 0.8873,
+      "step": 7889
+    },
+    {
+      "epoch": 6.302097902097902,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0002,
+      "loss": 0.8927,
+      "step": 7890
+    },
+    {
+      "epoch": 6.302897102897103,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.0002,
+      "loss": 0.8963,
+      "step": 7891
+    },
+    {
+      "epoch": 6.303696303696304,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8826,
+      "step": 7892
+    },
+    {
+      "epoch": 6.304495504495504,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0002,
+      "loss": 0.8911,
+      "step": 7893
+    },
+    {
+      "epoch": 6.305294705294705,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8902,
+      "step": 7894
+    },
+    {
+      "epoch": 6.306093906093906,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8893,
+      "step": 7895
+    },
+    {
+      "epoch": 6.3068931068931064,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0002,
+      "loss": 0.8773,
+      "step": 7896
+    },
+    {
+      "epoch": 6.3076923076923075,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0002,
+      "loss": 0.8944,
+      "step": 7897
+    },
+    {
+      "epoch": 6.3084915084915085,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8916,
+      "step": 7898
+    },
+    {
+      "epoch": 6.30929070929071,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0002,
+      "loss": 0.8892,
+      "step": 7899
+    },
+    {
+      "epoch": 6.31008991008991,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0002,
+      "loss": 0.8932,
+      "step": 7900
+    },
+    {
+      "epoch": 6.310889110889111,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0002,
+      "loss": 0.8813,
+      "step": 7901
+    },
+    {
+      "epoch": 6.311688311688312,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0002,
+      "loss": 0.8902,
+      "step": 7902
+    },
+    {
+      "epoch": 6.312487512487513,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8957,
+      "step": 7903
+    },
+    {
+      "epoch": 6.313286713286713,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0002,
+      "loss": 0.89,
+      "step": 7904
+    },
+    {
+      "epoch": 6.314085914085914,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8796,
+      "step": 7905
+    },
+    {
+      "epoch": 6.314885114885115,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8871,
+      "step": 7906
+    },
+    {
+      "epoch": 6.315684315684316,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8885,
+      "step": 7907
+    },
+    {
+      "epoch": 6.316483516483516,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8935,
+      "step": 7908
+    },
+    {
+      "epoch": 6.317282717282717,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.0002,
+      "loss": 0.8776,
+      "step": 7909
+    },
+    {
+      "epoch": 6.318081918081918,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8885,
+      "step": 7910
+    },
+    {
+      "epoch": 6.3188811188811185,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0002,
+      "loss": 0.8753,
+      "step": 7911
+    },
+    {
+      "epoch": 6.3196803196803195,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0002,
+      "loss": 0.8931,
+      "step": 7912
+    },
+    {
+      "epoch": 6.3204795204795206,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8875,
+      "step": 7913
+    },
+    {
+      "epoch": 6.321278721278722,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8853,
+      "step": 7914
+    },
+    {
+      "epoch": 6.322077922077922,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0002,
+      "loss": 0.8834,
+      "step": 7915
+    },
+    {
+      "epoch": 6.322877122877123,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0002,
+      "loss": 0.8936,
+      "step": 7916
+    },
+    {
+      "epoch": 6.323676323676324,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8845,
+      "step": 7917
+    },
+    {
+      "epoch": 6.324475524475525,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0002,
+      "loss": 0.8858,
+      "step": 7918
+    },
+    {
+      "epoch": 6.325274725274725,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8834,
+      "step": 7919
+    },
+    {
+      "epoch": 6.326073926073926,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0002,
+      "loss": 0.8877,
+      "step": 7920
+    },
+    {
+      "epoch": 6.326873126873127,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0002,
+      "loss": 0.8907,
+      "step": 7921
+    },
+    {
+      "epoch": 6.327672327672328,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0002,
+      "loss": 0.8874,
+      "step": 7922
+    },
+    {
+      "epoch": 6.328471528471528,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0002,
+      "loss": 0.8921,
+      "step": 7923
+    },
+    {
+      "epoch": 6.329270729270729,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8878,
+      "step": 7924
+    },
+    {
+      "epoch": 6.33006993006993,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0002,
+      "loss": 0.8898,
+      "step": 7925
+    },
+    {
+      "epoch": 6.3308691308691305,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0002,
+      "loss": 0.886,
+      "step": 7926
+    },
+    {
+      "epoch": 6.3316683316683315,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0002,
+      "loss": 0.8848,
+      "step": 7927
+    },
+    {
+      "epoch": 6.332467532467533,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0002,
+      "loss": 0.887,
+      "step": 7928
+    },
+    {
+      "epoch": 6.333266733266734,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8925,
+      "step": 7929
+    },
+    {
+      "epoch": 6.334065934065934,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.0002,
+      "loss": 0.8865,
+      "step": 7930
+    },
+    {
+      "epoch": 6.334865134865135,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.0002,
+      "loss": 0.8947,
+      "step": 7931
+    },
+    {
+      "epoch": 6.335664335664336,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8863,
+      "step": 7932
+    },
+    {
+      "epoch": 6.336463536463537,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0002,
+      "loss": 0.8865,
+      "step": 7933
+    },
+    {
+      "epoch": 6.337262737262737,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8812,
+      "step": 7934
+    },
+    {
+      "epoch": 6.338061938061938,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8864,
+      "step": 7935
+    },
+    {
+      "epoch": 6.338861138861139,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0002,
+      "loss": 0.8949,
+      "step": 7936
+    },
+    {
+      "epoch": 6.339660339660339,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8875,
+      "step": 7937
+    },
+    {
+      "epoch": 6.34045954045954,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8866,
+      "step": 7938
+    },
+    {
+      "epoch": 6.341258741258741,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8905,
+      "step": 7939
+    },
+    {
+      "epoch": 6.342057942057942,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.0002,
+      "loss": 0.8903,
+      "step": 7940
+    },
+    {
+      "epoch": 6.3428571428571425,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0002,
+      "loss": 0.8899,
+      "step": 7941
+    },
+    {
+      "epoch": 6.343656343656344,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8902,
+      "step": 7942
+    },
+    {
+      "epoch": 6.344455544455545,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8842,
+      "step": 7943
+    },
+    {
+      "epoch": 6.345254745254746,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.0002,
+      "loss": 0.8757,
+      "step": 7944
+    },
+    {
+      "epoch": 6.346053946053946,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0002,
+      "loss": 0.8877,
+      "step": 7945
+    },
+    {
+      "epoch": 6.346853146853147,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.0002,
+      "loss": 0.893,
+      "step": 7946
+    },
+    {
+      "epoch": 6.347652347652348,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8857,
+      "step": 7947
+    },
+    {
+      "epoch": 6.348451548451548,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0002,
+      "loss": 0.885,
+      "step": 7948
+    },
+    {
+      "epoch": 6.349250749250749,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8804,
+      "step": 7949
+    },
+    {
+      "epoch": 6.35004995004995,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0002,
+      "loss": 0.8919,
+      "step": 7950
+    },
+    {
+      "epoch": 6.350849150849151,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0002,
+      "loss": 0.8919,
+      "step": 7951
+    },
+    {
+      "epoch": 6.351648351648351,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8912,
+      "step": 7952
+    },
+    {
+      "epoch": 6.352447552447552,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0002,
+      "loss": 0.8916,
+      "step": 7953
+    },
+    {
+      "epoch": 6.353246753246753,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0002,
+      "loss": 0.8961,
+      "step": 7954
+    },
+    {
+      "epoch": 6.354045954045954,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0002,
+      "loss": 0.8849,
+      "step": 7955
+    },
+    {
+      "epoch": 6.3548451548451546,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8877,
+      "step": 7956
+    },
+    {
+      "epoch": 6.355644355644356,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0002,
+      "loss": 0.9009,
+      "step": 7957
+    },
+    {
+      "epoch": 6.356443556443557,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0002,
+      "loss": 0.887,
+      "step": 7958
+    },
+    {
+      "epoch": 6.357242757242757,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0002,
+      "loss": 0.8962,
+      "step": 7959
+    },
+    {
+      "epoch": 6.358041958041958,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.0002,
+      "loss": 0.887,
+      "step": 7960
+    },
+    {
+      "epoch": 6.358841158841159,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8929,
+      "step": 7961
+    },
+    {
+      "epoch": 6.35964035964036,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8962,
+      "step": 7962
+    },
+    {
+      "epoch": 6.36043956043956,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8903,
+      "step": 7963
+    },
+    {
+      "epoch": 6.361238761238761,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8876,
+      "step": 7964
+    },
+    {
+      "epoch": 6.362037962037962,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0002,
+      "loss": 0.889,
+      "step": 7965
+    },
+    {
+      "epoch": 6.362837162837163,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8852,
+      "step": 7966
+    },
+    {
+      "epoch": 6.363636363636363,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0002,
+      "loss": 0.8875,
+      "step": 7967
+    },
+    {
+      "epoch": 6.364435564435564,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0002,
+      "loss": 0.8922,
+      "step": 7968
+    },
+    {
+      "epoch": 6.365234765234765,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8935,
+      "step": 7969
+    },
+    {
+      "epoch": 6.366033966033966,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0002,
+      "loss": 0.8854,
+      "step": 7970
+    },
+    {
+      "epoch": 6.366833166833167,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0002,
+      "loss": 0.8784,
+      "step": 7971
+    },
+    {
+      "epoch": 6.367632367632368,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0002,
+      "loss": 0.8857,
+      "step": 7972
+    },
+    {
+      "epoch": 6.368431568431569,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0002,
+      "loss": 0.8829,
+      "step": 7973
+    },
+    {
+      "epoch": 6.36923076923077,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8868,
+      "step": 7974
+    },
+    {
+      "epoch": 6.37002997002997,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0002,
+      "loss": 0.891,
+      "step": 7975
+    },
+    {
+      "epoch": 6.370829170829171,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8941,
+      "step": 7976
+    },
+    {
+      "epoch": 6.371628371628372,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8862,
+      "step": 7977
+    },
+    {
+      "epoch": 6.372427572427572,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8847,
+      "step": 7978
+    },
+    {
+      "epoch": 6.373226773226773,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0002,
+      "loss": 0.8873,
+      "step": 7979
+    },
+    {
+      "epoch": 6.374025974025974,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0002,
+      "loss": 0.899,
+      "step": 7980
+    },
+    {
+      "epoch": 6.374825174825175,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0002,
+      "loss": 0.8865,
+      "step": 7981
+    },
+    {
+      "epoch": 6.375624375624375,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0002,
+      "loss": 0.8856,
+      "step": 7982
+    },
+    {
+      "epoch": 6.376423576423576,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0002,
+      "loss": 0.8914,
+      "step": 7983
+    },
+    {
+      "epoch": 6.377222777222777,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0002,
+      "loss": 0.8988,
+      "step": 7984
+    },
+    {
+      "epoch": 6.3780219780219785,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8866,
+      "step": 7985
+    },
+    {
+      "epoch": 6.378821178821179,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0002,
+      "loss": 0.8853,
+      "step": 7986
+    },
+    {
+      "epoch": 6.37962037962038,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8927,
+      "step": 7987
+    },
+    {
+      "epoch": 6.380419580419581,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8942,
+      "step": 7988
+    },
+    {
+      "epoch": 6.381218781218781,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0002,
+      "loss": 0.8934,
+      "step": 7989
+    },
+    {
+      "epoch": 6.382017982017982,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0002,
+      "loss": 0.8809,
+      "step": 7990
+    },
+    {
+      "epoch": 6.382817182817183,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8893,
+      "step": 7991
+    },
+    {
+      "epoch": 6.383616383616384,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0002,
+      "loss": 0.8868,
+      "step": 7992
+    },
+    {
+      "epoch": 6.384415584415584,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8918,
+      "step": 7993
+    },
+    {
+      "epoch": 6.385214785214785,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8893,
+      "step": 7994
+    },
+    {
+      "epoch": 6.386013986013986,
+      "grad_norm": 2.828125,
+      "learning_rate": 0.0002,
+      "loss": 0.9121,
+      "step": 7995
+    },
+    {
+      "epoch": 6.386813186813187,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8819,
+      "step": 7996
+    },
+    {
+      "epoch": 6.387612387612387,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0002,
+      "loss": 0.8837,
+      "step": 7997
+    },
+    {
+      "epoch": 6.388411588411588,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8868,
+      "step": 7998
+    },
+    {
+      "epoch": 6.389210789210789,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0002,
+      "loss": 0.8841,
+      "step": 7999
+    },
+    {
+      "epoch": 6.39000999000999,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.0002,
+      "loss": 0.8831,
+      "step": 8000
+    },
+    {
+      "epoch": 6.390809190809191,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8836,
+      "step": 8001
+    },
+    {
+      "epoch": 6.391608391608392,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0002,
+      "loss": 0.8846,
+      "step": 8002
+    },
+    {
+      "epoch": 6.392407592407593,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.0002,
+      "loss": 0.8946,
+      "step": 8003
+    },
+    {
+      "epoch": 6.393206793206793,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.0002,
+      "loss": 0.886,
+      "step": 8004
+    },
+    {
+      "epoch": 6.394005994005994,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0002,
+      "loss": 0.8842,
+      "step": 8005
+    },
+    {
+      "epoch": 6.394805194805195,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0002,
+      "loss": 0.8902,
+      "step": 8006
+    },
+    {
+      "epoch": 6.395604395604396,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8967,
+      "step": 8007
+    },
+    {
+      "epoch": 6.396403596403596,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0002,
+      "loss": 0.892,
+      "step": 8008
+    },
+    {
+      "epoch": 6.397202797202797,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0002,
+      "loss": 0.8912,
+      "step": 8009
+    },
+    {
+      "epoch": 6.398001998001998,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8857,
+      "step": 8010
+    },
+    {
+      "epoch": 6.398801198801198,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0002,
+      "loss": 0.8754,
+      "step": 8011
+    },
+    {
+      "epoch": 6.399600399600399,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8899,
+      "step": 8012
+    },
+    {
+      "epoch": 6.4003996003996,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0002,
+      "loss": 0.8898,
+      "step": 8013
+    },
+    {
+      "epoch": 6.4011988011988015,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0002,
+      "loss": 0.8896,
+      "step": 8014
+    },
+    {
+      "epoch": 6.401998001998002,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.0002,
+      "loss": 0.8933,
+      "step": 8015
+    },
+    {
+      "epoch": 6.402797202797203,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.0002,
+      "loss": 0.8874,
+      "step": 8016
+    },
+    {
+      "epoch": 6.403596403596404,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8884,
+      "step": 8017
+    },
+    {
+      "epoch": 6.404395604395605,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0002,
+      "loss": 0.8942,
+      "step": 8018
+    },
+    {
+      "epoch": 6.405194805194805,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8889,
+      "step": 8019
+    },
+    {
+      "epoch": 6.405994005994006,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8875,
+      "step": 8020
+    },
+    {
+      "epoch": 6.406793206793207,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0002,
+      "loss": 0.8979,
+      "step": 8021
+    },
+    {
+      "epoch": 6.407592407592408,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8941,
+      "step": 8022
+    },
+    {
+      "epoch": 6.408391608391608,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8845,
+      "step": 8023
+    },
+    {
+      "epoch": 6.409190809190809,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8935,
+      "step": 8024
+    },
+    {
+      "epoch": 6.40999000999001,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0002,
+      "loss": 0.8874,
+      "step": 8025
+    },
+    {
+      "epoch": 6.41078921078921,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0002,
+      "loss": 0.8902,
+      "step": 8026
+    },
+    {
+      "epoch": 6.411588411588411,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0002,
+      "loss": 0.8951,
+      "step": 8027
+    },
+    {
+      "epoch": 6.4123876123876125,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0002,
+      "loss": 0.8839,
+      "step": 8028
+    },
+    {
+      "epoch": 6.4131868131868135,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0002,
+      "loss": 0.8895,
+      "step": 8029
+    },
+    {
+      "epoch": 6.413986013986014,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0002,
+      "loss": 0.8936,
+      "step": 8030
+    },
+    {
+      "epoch": 6.414785214785215,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8808,
+      "step": 8031
+    },
+    {
+      "epoch": 6.415584415584416,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0002,
+      "loss": 0.8883,
+      "step": 8032
+    },
+    {
+      "epoch": 6.416383616383617,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0002,
+      "loss": 0.8784,
+      "step": 8033
+    },
+    {
+      "epoch": 6.417182817182817,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0002,
+      "loss": 0.891,
+      "step": 8034
+    },
+    {
+      "epoch": 6.417982017982018,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0002,
+      "loss": 0.8963,
+      "step": 8035
+    },
+    {
+      "epoch": 6.418781218781219,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8909,
+      "step": 8036
+    },
+    {
+      "epoch": 6.41958041958042,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8903,
+      "step": 8037
+    },
+    {
+      "epoch": 6.42037962037962,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0002,
+      "loss": 0.8827,
+      "step": 8038
+    },
+    {
+      "epoch": 6.421178821178821,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0002,
+      "loss": 0.8869,
+      "step": 8039
+    },
+    {
+      "epoch": 6.421978021978022,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0002,
+      "loss": 0.8933,
+      "step": 8040
+    },
+    {
+      "epoch": 6.422777222777222,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.0002,
+      "loss": 0.88,
+      "step": 8041
+    },
+    {
+      "epoch": 6.4235764235764234,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8849,
+      "step": 8042
+    },
+    {
+      "epoch": 6.4243756243756245,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0002,
+      "loss": 0.8935,
+      "step": 8043
+    },
+    {
+      "epoch": 6.4251748251748255,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0002,
+      "loss": 0.8834,
+      "step": 8044
+    },
+    {
+      "epoch": 6.425974025974026,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8961,
+      "step": 8045
+    },
+    {
+      "epoch": 6.426773226773227,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8888,
+      "step": 8046
+    },
+    {
+      "epoch": 6.427572427572428,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.0002,
+      "loss": 0.8921,
+      "step": 8047
+    },
+    {
+      "epoch": 6.428371628371629,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8916,
+      "step": 8048
+    },
+    {
+      "epoch": 6.429170829170829,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.0002,
+      "loss": 0.895,
+      "step": 8049
+    },
+    {
+      "epoch": 6.42997002997003,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0002,
+      "loss": 0.8839,
+      "step": 8050
+    },
+    {
+      "epoch": 6.430769230769231,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.0002,
+      "loss": 0.8951,
+      "step": 8051
+    },
+    {
+      "epoch": 6.431568431568431,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8923,
+      "step": 8052
+    },
+    {
+      "epoch": 6.432367632367632,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.0002,
+      "loss": 0.8959,
+      "step": 8053
+    },
+    {
+      "epoch": 6.433166833166833,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0002,
+      "loss": 0.8941,
+      "step": 8054
+    },
+    {
+      "epoch": 6.433966033966034,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0002,
+      "loss": 0.8881,
+      "step": 8055
+    },
+    {
+      "epoch": 6.434765234765234,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0002,
+      "loss": 0.8856,
+      "step": 8056
+    },
+    {
+      "epoch": 6.4355644355644355,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8878,
+      "step": 8057
+    },
+    {
+      "epoch": 6.4363636363636365,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0002,
+      "loss": 0.8755,
+      "step": 8058
+    },
+    {
+      "epoch": 6.4371628371628375,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8875,
+      "step": 8059
+    },
+    {
+      "epoch": 6.437962037962038,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0002,
+      "loss": 0.888,
+      "step": 8060
+    },
+    {
+      "epoch": 6.438761238761239,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0002,
+      "loss": 0.9003,
+      "step": 8061
+    },
+    {
+      "epoch": 6.43956043956044,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0002,
+      "loss": 0.8831,
+      "step": 8062
+    },
+    {
+      "epoch": 6.44035964035964,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0002,
+      "loss": 0.8918,
+      "step": 8063
+    },
+    {
+      "epoch": 6.441158841158841,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0002,
+      "loss": 0.8889,
+      "step": 8064
+    },
+    {
+      "epoch": 6.441958041958042,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0002,
+      "loss": 0.8856,
+      "step": 8065
+    },
+    {
+      "epoch": 6.442757242757243,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0002,
+      "loss": 0.8846,
+      "step": 8066
+    },
+    {
+      "epoch": 6.443556443556443,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8861,
+      "step": 8067
+    },
+    {
+      "epoch": 6.444355644355644,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0002,
+      "loss": 0.8924,
+      "step": 8068
+    },
+    {
+      "epoch": 6.445154845154845,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0002,
+      "loss": 0.8949,
+      "step": 8069
+    },
+    {
+      "epoch": 6.445954045954046,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8874,
+      "step": 8070
+    },
+    {
+      "epoch": 6.4467532467532465,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8955,
+      "step": 8071
+    },
+    {
+      "epoch": 6.4475524475524475,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8894,
+      "step": 8072
+    },
+    {
+      "epoch": 6.4483516483516485,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0002,
+      "loss": 0.8861,
+      "step": 8073
+    },
+    {
+      "epoch": 6.449150849150849,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8962,
+      "step": 8074
+    },
+    {
+      "epoch": 6.44995004995005,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0002,
+      "loss": 0.8824,
+      "step": 8075
+    },
+    {
+      "epoch": 6.450749250749251,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8854,
+      "step": 8076
+    },
+    {
+      "epoch": 6.451548451548452,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8871,
+      "step": 8077
+    },
+    {
+      "epoch": 6.452347652347652,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8924,
+      "step": 8078
+    },
+    {
+      "epoch": 6.453146853146853,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0002,
+      "loss": 0.8919,
+      "step": 8079
+    },
+    {
+      "epoch": 6.453946053946054,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0002,
+      "loss": 0.8885,
+      "step": 8080
+    },
+    {
+      "epoch": 6.454745254745255,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0002,
+      "loss": 0.8823,
+      "step": 8081
+    },
+    {
+      "epoch": 6.455544455544455,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0002,
+      "loss": 0.8855,
+      "step": 8082
+    },
+    {
+      "epoch": 6.456343656343656,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8985,
+      "step": 8083
+    },
+    {
+      "epoch": 6.457142857142857,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8815,
+      "step": 8084
+    },
+    {
+      "epoch": 6.457942057942058,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0002,
+      "loss": 0.8901,
+      "step": 8085
+    },
+    {
+      "epoch": 6.4587412587412585,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8943,
+      "step": 8086
+    },
+    {
+      "epoch": 6.4595404595404595,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0002,
+      "loss": 0.8947,
+      "step": 8087
+    },
+    {
+      "epoch": 6.460339660339661,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8989,
+      "step": 8088
+    },
+    {
+      "epoch": 6.461138861138862,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8979,
+      "step": 8089
+    },
+    {
+      "epoch": 6.461938061938062,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.0002,
+      "loss": 0.8773,
+      "step": 8090
+    },
+    {
+      "epoch": 6.462737262737263,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0002,
+      "loss": 0.8887,
+      "step": 8091
+    },
+    {
+      "epoch": 6.463536463536464,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.0002,
+      "loss": 0.8796,
+      "step": 8092
+    },
+    {
+      "epoch": 6.464335664335664,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0002,
+      "loss": 0.8898,
+      "step": 8093
+    },
+    {
+      "epoch": 6.465134865134865,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8856,
+      "step": 8094
+    },
+    {
+      "epoch": 6.465934065934066,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0002,
+      "loss": 0.89,
+      "step": 8095
+    },
+    {
+      "epoch": 6.466733266733267,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8984,
+      "step": 8096
+    },
+    {
+      "epoch": 6.467532467532467,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8819,
+      "step": 8097
+    },
+    {
+      "epoch": 6.468331668331668,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8896,
+      "step": 8098
+    },
+    {
+      "epoch": 6.469130869130869,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0002,
+      "loss": 0.892,
+      "step": 8099
+    },
+    {
+      "epoch": 6.46993006993007,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.0002,
+      "loss": 0.886,
+      "step": 8100
+    },
+    {
+      "epoch": 6.4707292707292705,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0002,
+      "loss": 0.8924,
+      "step": 8101
+    },
+    {
+      "epoch": 6.4715284715284715,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.0002,
+      "loss": 0.8916,
+      "step": 8102
+    },
+    {
+      "epoch": 6.472327672327673,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8883,
+      "step": 8103
+    },
+    {
+      "epoch": 6.473126873126873,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8833,
+      "step": 8104
+    },
+    {
+      "epoch": 6.473926073926074,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0002,
+      "loss": 0.8954,
+      "step": 8105
+    },
+    {
+      "epoch": 6.474725274725275,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0002,
+      "loss": 0.8758,
+      "step": 8106
+    },
+    {
+      "epoch": 6.475524475524476,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0002,
+      "loss": 0.8889,
+      "step": 8107
+    },
+    {
+      "epoch": 6.476323676323676,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.0002,
+      "loss": 0.8954,
+      "step": 8108
+    },
+    {
+      "epoch": 6.477122877122877,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8796,
+      "step": 8109
+    },
+    {
+      "epoch": 6.477922077922078,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0002,
+      "loss": 0.897,
+      "step": 8110
+    },
+    {
+      "epoch": 6.478721278721279,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0002,
+      "loss": 0.8853,
+      "step": 8111
+    },
+    {
+      "epoch": 6.479520479520479,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0002,
+      "loss": 0.8861,
+      "step": 8112
+    },
+    {
+      "epoch": 6.48031968031968,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8867,
+      "step": 8113
+    },
+    {
+      "epoch": 6.481118881118881,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0002,
+      "loss": 0.892,
+      "step": 8114
+    },
+    {
+      "epoch": 6.4819180819180815,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.0002,
+      "loss": 0.8938,
+      "step": 8115
+    },
+    {
+      "epoch": 6.4827172827172825,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0002,
+      "loss": 0.8836,
+      "step": 8116
+    },
+    {
+      "epoch": 6.483516483516484,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.0002,
+      "loss": 0.8925,
+      "step": 8117
+    },
+    {
+      "epoch": 6.484315684315685,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8776,
+      "step": 8118
+    },
+    {
+      "epoch": 6.485114885114885,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8933,
+      "step": 8119
+    },
+    {
+      "epoch": 6.485914085914086,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0002,
+      "loss": 0.8765,
+      "step": 8120
+    },
+    {
+      "epoch": 6.486713286713287,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.0002,
+      "loss": 0.891,
+      "step": 8121
+    },
+    {
+      "epoch": 6.487512487512488,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.0002,
+      "loss": 0.8879,
+      "step": 8122
+    },
+    {
+      "epoch": 6.488311688311688,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8864,
+      "step": 8123
+    },
+    {
+      "epoch": 6.489110889110889,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.0002,
+      "loss": 0.8907,
+      "step": 8124
+    },
+    {
+      "epoch": 6.48991008991009,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0002,
+      "loss": 0.8796,
+      "step": 8125
+    },
+    {
+      "epoch": 6.49070929070929,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0002,
+      "loss": 0.8896,
+      "step": 8126
+    },
+    {
+      "epoch": 6.491508491508491,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.0002,
+      "loss": 0.8885,
+      "step": 8127
+    },
+    {
+      "epoch": 6.492307692307692,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0002,
+      "loss": 0.8869,
+      "step": 8128
+    },
+    {
+      "epoch": 6.493106893106893,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.0002,
+      "loss": 0.8994,
+      "step": 8129
+    },
+    {
+      "epoch": 6.4939060939060935,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0002,
+      "loss": 0.8886,
+      "step": 8130
+    },
+    {
+      "epoch": 6.494705294705295,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.0002,
+      "loss": 0.886,
+      "step": 8131
+    },
+    {
+      "epoch": 6.495504495504496,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0002,
+      "loss": 0.8889,
+      "step": 8132
+    },
+    {
+      "epoch": 6.496303696303697,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.0002,
+      "loss": 0.8946,
+      "step": 8133
+    },
+    {
+      "epoch": 6.497102897102897,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0002,
+      "loss": 0.8795,
+      "step": 8134
+    },
+    {
+      "epoch": 6.497902097902098,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8865,
+      "step": 8135
+    },
+    {
+      "epoch": 6.498701298701299,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0002,
+      "loss": 0.89,
+      "step": 8136
+    },
+    {
+      "epoch": 6.4995004995005,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8984,
+      "step": 8137
+    },
+    {
+      "epoch": 6.5002997002997,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8995,
+      "step": 8138
+    },
+    {
+      "epoch": 6.501098901098901,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8939,
+      "step": 8139
+    },
+    {
+      "epoch": 6.501898101898102,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8908,
+      "step": 8140
+    },
+    {
+      "epoch": 6.502697302697303,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0002,
+      "loss": 0.8897,
+      "step": 8141
+    },
+    {
+      "epoch": 6.503496503496503,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8786,
+      "step": 8142
+    },
+    {
+      "epoch": 6.504295704295704,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8857,
+      "step": 8143
+    },
+    {
+      "epoch": 6.505094905094905,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0002,
+      "loss": 0.8839,
+      "step": 8144
+    },
+    {
+      "epoch": 6.5058941058941056,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8948,
+      "step": 8145
+    },
+    {
+      "epoch": 6.506693306693307,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8872,
+      "step": 8146
+    },
+    {
+      "epoch": 6.507492507492508,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0002,
+      "loss": 0.8866,
+      "step": 8147
+    },
+    {
+      "epoch": 6.508291708291709,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0002,
+      "loss": 0.8889,
+      "step": 8148
+    },
+    {
+      "epoch": 6.509090909090909,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8829,
+      "step": 8149
+    },
+    {
+      "epoch": 6.50989010989011,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0002,
+      "loss": 0.8812,
+      "step": 8150
+    },
+    {
+      "epoch": 6.510689310689311,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8842,
+      "step": 8151
+    },
+    {
+      "epoch": 6.511488511488512,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0002,
+      "loss": 0.8876,
+      "step": 8152
+    },
+    {
+      "epoch": 6.512287712287712,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8911,
+      "step": 8153
+    },
+    {
+      "epoch": 6.513086913086913,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8953,
+      "step": 8154
+    },
+    {
+      "epoch": 6.513886113886114,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8897,
+      "step": 8155
+    },
+    {
+      "epoch": 6.514685314685314,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8948,
+      "step": 8156
+    },
+    {
+      "epoch": 6.515484515484515,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0002,
+      "loss": 0.89,
+      "step": 8157
+    },
+    {
+      "epoch": 6.516283716283716,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0002,
+      "loss": 0.8928,
+      "step": 8158
+    },
+    {
+      "epoch": 6.517082917082917,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.0002,
+      "loss": 0.8886,
+      "step": 8159
+    },
+    {
+      "epoch": 6.517882117882118,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8848,
+      "step": 8160
+    },
+    {
+      "epoch": 6.518681318681319,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0002,
+      "loss": 0.8809,
+      "step": 8161
+    },
+    {
+      "epoch": 6.51948051948052,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0002,
+      "loss": 0.892,
+      "step": 8162
+    },
+    {
+      "epoch": 6.520279720279721,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8897,
+      "step": 8163
+    },
+    {
+      "epoch": 6.521078921078921,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0002,
+      "loss": 0.8832,
+      "step": 8164
+    },
+    {
+      "epoch": 6.521878121878122,
+      "grad_norm": 2.9375,
+      "learning_rate": 0.0002,
+      "loss": 0.9355,
+      "step": 8165
+    },
+    {
+      "epoch": 6.522677322677323,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0002,
+      "loss": 0.8936,
+      "step": 8166
+    },
+    {
+      "epoch": 6.523476523476523,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8849,
+      "step": 8167
+    },
+    {
+      "epoch": 6.524275724275724,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8906,
+      "step": 8168
+    },
+    {
+      "epoch": 6.525074925074925,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0002,
+      "loss": 0.8823,
+      "step": 8169
+    },
+    {
+      "epoch": 6.525874125874126,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0002,
+      "loss": 0.9016,
+      "step": 8170
+    },
+    {
+      "epoch": 6.526673326673326,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8829,
+      "step": 8171
+    },
+    {
+      "epoch": 6.527472527472527,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0002,
+      "loss": 0.8793,
+      "step": 8172
+    },
+    {
+      "epoch": 6.528271728271728,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8882,
+      "step": 8173
+    },
+    {
+      "epoch": 6.5290709290709295,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0002,
+      "loss": 0.8864,
+      "step": 8174
+    },
+    {
+      "epoch": 6.52987012987013,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8868,
+      "step": 8175
+    },
+    {
+      "epoch": 6.530669330669331,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8991,
+      "step": 8176
+    },
+    {
+      "epoch": 6.531468531468532,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.0002,
+      "loss": 0.8767,
+      "step": 8177
+    },
+    {
+      "epoch": 6.532267732267732,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0002,
+      "loss": 0.8939,
+      "step": 8178
+    },
+    {
+      "epoch": 6.533066933066933,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.0002,
+      "loss": 0.8892,
+      "step": 8179
+    },
+    {
+      "epoch": 6.533866133866134,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8887,
+      "step": 8180
+    },
+    {
+      "epoch": 6.534665334665335,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.0002,
+      "loss": 0.8883,
+      "step": 8181
+    },
+    {
+      "epoch": 6.535464535464535,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002,
+      "loss": 0.877,
+      "step": 8182
+    },
+    {
+      "epoch": 6.536263736263736,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.0002,
+      "loss": 0.8848,
+      "step": 8183
+    },
+    {
+      "epoch": 6.537062937062937,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8922,
+      "step": 8184
+    },
+    {
+      "epoch": 6.537862137862138,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8896,
+      "step": 8185
+    },
+    {
+      "epoch": 6.538661338661338,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0002,
+      "loss": 0.9014,
+      "step": 8186
+    },
+    {
+      "epoch": 6.539460539460539,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0002,
+      "loss": 0.8771,
+      "step": 8187
+    },
+    {
+      "epoch": 6.54025974025974,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8782,
+      "step": 8188
+    },
+    {
+      "epoch": 6.541058941058941,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0002,
+      "loss": 0.8927,
+      "step": 8189
+    },
+    {
+      "epoch": 6.541858141858142,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8892,
+      "step": 8190
+    },
+    {
+      "epoch": 6.542657342657343,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8827,
+      "step": 8191
+    },
+    {
+      "epoch": 6.543456543456544,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8827,
+      "step": 8192
+    },
+    {
+      "epoch": 6.544255744255745,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0002,
+      "loss": 0.8879,
+      "step": 8193
+    },
+    {
+      "epoch": 6.545054945054945,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0002,
+      "loss": 0.8986,
+      "step": 8194
+    },
+    {
+      "epoch": 6.545854145854146,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0002,
+      "loss": 0.884,
+      "step": 8195
+    },
+    {
+      "epoch": 6.546653346653347,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8915,
+      "step": 8196
+    },
+    {
+      "epoch": 6.547452547452547,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0002,
+      "loss": 0.8845,
+      "step": 8197
+    },
+    {
+      "epoch": 6.548251748251748,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0002,
+      "loss": 0.8911,
+      "step": 8198
+    },
+    {
+      "epoch": 6.549050949050949,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0002,
+      "loss": 0.8884,
+      "step": 8199
+    },
+    {
+      "epoch": 6.54985014985015,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8879,
+      "step": 8200
+    },
+    {
+      "epoch": 6.55064935064935,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8904,
+      "step": 8201
+    },
+    {
+      "epoch": 6.551448551448551,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0002,
+      "loss": 0.9039,
+      "step": 8202
+    },
+    {
+      "epoch": 6.5522477522477525,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8884,
+      "step": 8203
+    },
+    {
+      "epoch": 6.5530469530469535,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.0002,
+      "loss": 0.8824,
+      "step": 8204
+    },
+    {
+      "epoch": 6.553846153846154,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8964,
+      "step": 8205
+    },
+    {
+      "epoch": 6.554645354645355,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.0002,
+      "loss": 0.8881,
+      "step": 8206
+    },
+    {
+      "epoch": 6.555444555444556,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0002,
+      "loss": 0.8812,
+      "step": 8207
+    },
+    {
+      "epoch": 6.556243756243756,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8857,
+      "step": 8208
+    },
+    {
+      "epoch": 6.557042957042957,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8902,
+      "step": 8209
+    },
+    {
+      "epoch": 6.557842157842158,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0002,
+      "loss": 0.8853,
+      "step": 8210
+    },
+    {
+      "epoch": 6.558641358641359,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8854,
+      "step": 8211
+    },
+    {
+      "epoch": 6.559440559440559,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.0002,
+      "loss": 0.8813,
+      "step": 8212
+    },
+    {
+      "epoch": 6.56023976023976,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.0002,
+      "loss": 0.8928,
+      "step": 8213
+    },
+    {
+      "epoch": 6.561038961038961,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8882,
+      "step": 8214
+    },
+    {
+      "epoch": 6.561838161838162,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0002,
+      "loss": 0.8844,
+      "step": 8215
+    },
+    {
+      "epoch": 6.562637362637362,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8807,
+      "step": 8216
+    },
+    {
+      "epoch": 6.5634365634365635,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8907,
+      "step": 8217
+    },
+    {
+      "epoch": 6.5642357642357645,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.0002,
+      "loss": 0.8865,
+      "step": 8218
+    },
+    {
+      "epoch": 6.565034965034965,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.0002,
+      "loss": 0.8816,
+      "step": 8219
+    },
+    {
+      "epoch": 6.565834165834166,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8898,
+      "step": 8220
+    },
+    {
+      "epoch": 6.566633366633367,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8845,
+      "step": 8221
+    },
+    {
+      "epoch": 6.567432567432568,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0002,
+      "loss": 0.8864,
+      "step": 8222
+    },
+    {
+      "epoch": 6.568231768231768,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.0002,
+      "loss": 0.8797,
+      "step": 8223
+    },
+    {
+      "epoch": 6.569030969030969,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0002,
+      "loss": 0.8886,
+      "step": 8224
+    },
+    {
+      "epoch": 6.56983016983017,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8839,
+      "step": 8225
+    },
+    {
+      "epoch": 6.570629370629371,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0002,
+      "loss": 0.891,
+      "step": 8226
+    },
+    {
+      "epoch": 6.571428571428571,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8945,
+      "step": 8227
+    },
+    {
+      "epoch": 6.572227772227772,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8866,
+      "step": 8228
+    },
+    {
+      "epoch": 6.573026973026973,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8912,
+      "step": 8229
+    },
+    {
+      "epoch": 6.573826173826173,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0002,
+      "loss": 0.8803,
+      "step": 8230
+    },
+    {
+      "epoch": 6.574625374625374,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0002,
+      "loss": 0.8879,
+      "step": 8231
+    },
+    {
+      "epoch": 6.5754245754245755,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8838,
+      "step": 8232
+    },
+    {
+      "epoch": 6.5762237762237765,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8868,
+      "step": 8233
+    },
+    {
+      "epoch": 6.577022977022977,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0002,
+      "loss": 0.8896,
+      "step": 8234
+    },
+    {
+      "epoch": 6.577822177822178,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0002,
+      "loss": 0.8862,
+      "step": 8235
+    },
+    {
+      "epoch": 6.578621378621379,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.0002,
+      "loss": 0.8864,
+      "step": 8236
+    },
+    {
+      "epoch": 6.57942057942058,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0002,
+      "loss": 0.8888,
+      "step": 8237
+    },
+    {
+      "epoch": 6.58021978021978,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0002,
+      "loss": 0.892,
+      "step": 8238
+    },
+    {
+      "epoch": 6.581018981018981,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8904,
+      "step": 8239
+    },
+    {
+      "epoch": 6.581818181818182,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0002,
+      "loss": 0.8857,
+      "step": 8240
+    },
+    {
+      "epoch": 6.582617382617382,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0002,
+      "loss": 0.8834,
+      "step": 8241
+    },
+    {
+      "epoch": 6.583416583416583,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8901,
+      "step": 8242
+    },
+    {
+      "epoch": 6.584215784215784,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8839,
+      "step": 8243
+    },
+    {
+      "epoch": 6.585014985014985,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0002,
+      "loss": 0.8854,
+      "step": 8244
+    },
+    {
+      "epoch": 6.585814185814185,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0002,
+      "loss": 0.888,
+      "step": 8245
+    },
+    {
+      "epoch": 6.5866133866133865,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0002,
+      "loss": 0.885,
+      "step": 8246
+    },
+    {
+      "epoch": 6.5874125874125875,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0002,
+      "loss": 0.8888,
+      "step": 8247
+    },
+    {
+      "epoch": 6.5882117882117885,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0002,
+      "loss": 0.8832,
+      "step": 8248
+    },
+    {
+      "epoch": 6.589010989010989,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8913,
+      "step": 8249
+    },
+    {
+      "epoch": 6.58981018981019,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8882,
+      "step": 8250
+    },
+    {
+      "epoch": 6.590609390609391,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8844,
+      "step": 8251
+    },
+    {
+      "epoch": 6.591408591408591,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0002,
+      "loss": 0.8936,
+      "step": 8252
+    },
+    {
+      "epoch": 6.592207792207792,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0002,
+      "loss": 0.8914,
+      "step": 8253
+    },
+    {
+      "epoch": 6.593006993006993,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0002,
+      "loss": 0.886,
+      "step": 8254
+    },
+    {
+      "epoch": 6.593806193806194,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8875,
+      "step": 8255
+    },
+    {
+      "epoch": 6.594605394605395,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0002,
+      "loss": 0.8945,
+      "step": 8256
+    },
+    {
+      "epoch": 6.595404595404595,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8903,
+      "step": 8257
+    },
+    {
+      "epoch": 6.596203796203796,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8905,
+      "step": 8258
+    },
+    {
+      "epoch": 6.597002997002997,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0002,
+      "loss": 0.8916,
+      "step": 8259
+    },
+    {
+      "epoch": 6.5978021978021975,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0002,
+      "loss": 0.8904,
+      "step": 8260
+    },
+    {
+      "epoch": 6.5986013986013985,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8923,
+      "step": 8261
+    },
+    {
+      "epoch": 6.5994005994005995,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0002,
+      "loss": 0.8978,
+      "step": 8262
+    },
+    {
+      "epoch": 6.600199800199801,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8797,
+      "step": 8263
+    },
+    {
+      "epoch": 6.600999000999001,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.0002,
+      "loss": 0.889,
+      "step": 8264
+    },
+    {
+      "epoch": 6.601798201798202,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8827,
+      "step": 8265
+    },
+    {
+      "epoch": 6.602597402597403,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8844,
+      "step": 8266
+    },
+    {
+      "epoch": 6.603396603396604,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8812,
+      "step": 8267
+    },
+    {
+      "epoch": 6.604195804195804,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8873,
+      "step": 8268
+    },
+    {
+      "epoch": 6.604995004995005,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8833,
+      "step": 8269
+    },
+    {
+      "epoch": 6.605794205794206,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8941,
+      "step": 8270
+    },
+    {
+      "epoch": 6.606593406593406,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8917,
+      "step": 8271
+    },
+    {
+      "epoch": 6.607392607392607,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8859,
+      "step": 8272
+    },
+    {
+      "epoch": 6.608191808191808,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8785,
+      "step": 8273
+    },
+    {
+      "epoch": 6.608991008991009,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8837,
+      "step": 8274
+    },
+    {
+      "epoch": 6.6097902097902095,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8789,
+      "step": 8275
+    },
+    {
+      "epoch": 6.6105894105894105,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0002,
+      "loss": 0.8882,
+      "step": 8276
+    },
+    {
+      "epoch": 6.611388611388612,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.0002,
+      "loss": 0.895,
+      "step": 8277
+    },
+    {
+      "epoch": 6.612187812187813,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0002,
+      "loss": 0.8908,
+      "step": 8278
+    },
+    {
+      "epoch": 6.612987012987013,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8912,
+      "step": 8279
+    },
+    {
+      "epoch": 6.613786213786214,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0002,
+      "loss": 0.9016,
+      "step": 8280
+    },
+    {
+      "epoch": 6.614585414585415,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8891,
+      "step": 8281
+    },
+    {
+      "epoch": 6.615384615384615,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0002,
+      "loss": 0.8956,
+      "step": 8282
+    },
+    {
+      "epoch": 6.616183816183816,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.0002,
+      "loss": 0.8916,
+      "step": 8283
+    },
+    {
+      "epoch": 6.616983016983017,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0002,
+      "loss": 0.8844,
+      "step": 8284
+    },
+    {
+      "epoch": 6.617782217782218,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8996,
+      "step": 8285
+    },
+    {
+      "epoch": 6.618581418581418,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0002,
+      "loss": 0.876,
+      "step": 8286
+    },
+    {
+      "epoch": 6.619380619380619,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8861,
+      "step": 8287
+    },
+    {
+      "epoch": 6.62017982017982,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8891,
+      "step": 8288
+    },
+    {
+      "epoch": 6.620979020979021,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0002,
+      "loss": 0.8878,
+      "step": 8289
+    },
+    {
+      "epoch": 6.6217782217782215,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0002,
+      "loss": 0.888,
+      "step": 8290
+    },
+    {
+      "epoch": 6.6225774225774225,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0002,
+      "loss": 0.8959,
+      "step": 8291
+    },
+    {
+      "epoch": 6.623376623376624,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0002,
+      "loss": 0.8803,
+      "step": 8292
+    },
+    {
+      "epoch": 6.624175824175824,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0002,
+      "loss": 0.8863,
+      "step": 8293
+    },
+    {
+      "epoch": 6.624975024975025,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8882,
+      "step": 8294
+    },
+    {
+      "epoch": 6.625774225774226,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0002,
+      "loss": 0.8818,
+      "step": 8295
+    },
+    {
+      "epoch": 6.626573426573427,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0002,
+      "loss": 0.881,
+      "step": 8296
+    },
+    {
+      "epoch": 6.627372627372627,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8907,
+      "step": 8297
+    },
+    {
+      "epoch": 6.628171828171828,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0002,
+      "loss": 0.8917,
+      "step": 8298
+    },
+    {
+      "epoch": 6.628971028971029,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8936,
+      "step": 8299
+    },
+    {
+      "epoch": 6.62977022977023,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8881,
+      "step": 8300
+    },
+    {
+      "epoch": 6.63056943056943,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0002,
+      "loss": 0.8743,
+      "step": 8301
+    },
+    {
+      "epoch": 6.631368631368631,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0002,
+      "loss": 0.8943,
+      "step": 8302
+    },
+    {
+      "epoch": 6.632167832167832,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0002,
+      "loss": 0.8927,
+      "step": 8303
+    },
+    {
+      "epoch": 6.6329670329670325,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0002,
+      "loss": 0.8907,
+      "step": 8304
+    },
+    {
+      "epoch": 6.6337662337662335,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8853,
+      "step": 8305
+    },
+    {
+      "epoch": 6.634565434565435,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0002,
+      "loss": 0.89,
+      "step": 8306
+    },
+    {
+      "epoch": 6.635364635364636,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8817,
+      "step": 8307
+    },
+    {
+      "epoch": 6.636163836163837,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8941,
+      "step": 8308
+    },
+    {
+      "epoch": 6.636963036963037,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.0002,
+      "loss": 0.8976,
+      "step": 8309
+    },
+    {
+      "epoch": 6.637762237762238,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8868,
+      "step": 8310
+    },
+    {
+      "epoch": 6.638561438561439,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8932,
+      "step": 8311
+    },
+    {
+      "epoch": 6.639360639360639,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0002,
+      "loss": 0.8895,
+      "step": 8312
+    },
+    {
+      "epoch": 6.64015984015984,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8886,
+      "step": 8313
+    },
+    {
+      "epoch": 6.640959040959041,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0002,
+      "loss": 0.883,
+      "step": 8314
+    },
+    {
+      "epoch": 6.641758241758242,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0002,
+      "loss": 0.8892,
+      "step": 8315
+    },
+    {
+      "epoch": 6.642557442557442,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0002,
+      "loss": 0.8842,
+      "step": 8316
+    },
+    {
+      "epoch": 6.643356643356643,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8839,
+      "step": 8317
+    },
+    {
+      "epoch": 6.644155844155844,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0002,
+      "loss": 0.878,
+      "step": 8318
+    },
+    {
+      "epoch": 6.644955044955045,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0002,
+      "loss": 0.888,
+      "step": 8319
+    },
+    {
+      "epoch": 6.645754245754246,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0002,
+      "loss": 0.8848,
+      "step": 8320
+    },
+    {
+      "epoch": 6.646553446553447,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8941,
+      "step": 8321
+    },
+    {
+      "epoch": 6.647352647352648,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0002,
+      "loss": 0.8916,
+      "step": 8322
+    },
+    {
+      "epoch": 6.648151848151848,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8882,
+      "step": 8323
+    },
+    {
+      "epoch": 6.648951048951049,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0002,
+      "loss": 0.8851,
+      "step": 8324
+    },
+    {
+      "epoch": 6.64975024975025,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8904,
+      "step": 8325
+    },
+    {
+      "epoch": 6.650549450549451,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0002,
+      "loss": 0.8896,
+      "step": 8326
+    },
+    {
+      "epoch": 6.651348651348651,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0002,
+      "loss": 0.8925,
+      "step": 8327
+    },
+    {
+      "epoch": 6.652147852147852,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0002,
+      "loss": 0.8952,
+      "step": 8328
+    },
+    {
+      "epoch": 6.652947052947053,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8826,
+      "step": 8329
+    },
+    {
+      "epoch": 6.653746253746254,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8908,
+      "step": 8330
+    },
+    {
+      "epoch": 6.654545454545454,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002,
+      "loss": 0.882,
+      "step": 8331
+    },
+    {
+      "epoch": 6.655344655344655,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0002,
+      "loss": 0.8915,
+      "step": 8332
+    },
+    {
+      "epoch": 6.656143856143856,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0002,
+      "loss": 0.8856,
+      "step": 8333
+    },
+    {
+      "epoch": 6.6569430569430565,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0002,
+      "loss": 0.8906,
+      "step": 8334
+    },
+    {
+      "epoch": 6.657742257742258,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8944,
+      "step": 8335
+    },
+    {
+      "epoch": 6.658541458541459,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8843,
+      "step": 8336
+    },
+    {
+      "epoch": 6.65934065934066,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0002,
+      "loss": 0.8877,
+      "step": 8337
+    },
+    {
+      "epoch": 6.66013986013986,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0002,
+      "loss": 0.8927,
+      "step": 8338
+    },
+    {
+      "epoch": 6.660939060939061,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0002,
+      "loss": 0.88,
+      "step": 8339
+    },
+    {
+      "epoch": 6.661738261738262,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8881,
+      "step": 8340
+    },
+    {
+      "epoch": 6.662537462537463,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0002,
+      "loss": 0.8853,
+      "step": 8341
+    },
+    {
+      "epoch": 6.663336663336663,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8858,
+      "step": 8342
+    },
+    {
+      "epoch": 6.664135864135864,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0002,
+      "loss": 0.8843,
+      "step": 8343
+    },
+    {
+      "epoch": 6.664935064935065,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8886,
+      "step": 8344
+    },
+    {
+      "epoch": 6.665734265734265,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0002,
+      "loss": 0.8872,
+      "step": 8345
+    },
+    {
+      "epoch": 6.666533466533466,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8859,
+      "step": 8346
+    },
+    {
+      "epoch": 6.667332667332667,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8786,
+      "step": 8347
+    },
+    {
+      "epoch": 6.668131868131868,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0002,
+      "loss": 0.8891,
+      "step": 8348
+    },
+    {
+      "epoch": 6.668931068931069,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0002,
+      "loss": 0.8924,
+      "step": 8349
+    },
+    {
+      "epoch": 6.66973026973027,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8803,
+      "step": 8350
+    },
+    {
+      "epoch": 6.670529470529471,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.0002,
+      "loss": 0.8833,
+      "step": 8351
+    },
+    {
+      "epoch": 6.671328671328672,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8869,
+      "step": 8352
+    },
+    {
+      "epoch": 6.672127872127872,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0002,
+      "loss": 0.89,
+      "step": 8353
+    },
+    {
+      "epoch": 6.672927072927073,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8853,
+      "step": 8354
+    },
+    {
+      "epoch": 6.673726273726274,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0002,
+      "loss": 0.8812,
+      "step": 8355
+    },
+    {
+      "epoch": 6.674525474525474,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0002,
+      "loss": 0.8768,
+      "step": 8356
+    },
+    {
+      "epoch": 6.675324675324675,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8823,
+      "step": 8357
+    },
+    {
+      "epoch": 6.676123876123876,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8982,
+      "step": 8358
+    },
+    {
+      "epoch": 6.676923076923077,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0002,
+      "loss": 0.8864,
+      "step": 8359
+    },
+    {
+      "epoch": 6.677722277722278,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.0002,
+      "loss": 0.8872,
+      "step": 8360
+    },
+    {
+      "epoch": 6.678521478521478,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0002,
+      "loss": 0.892,
+      "step": 8361
+    },
+    {
+      "epoch": 6.679320679320679,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0002,
+      "loss": 0.8955,
+      "step": 8362
+    },
+    {
+      "epoch": 6.6801198801198804,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.0002,
+      "loss": 0.8921,
+      "step": 8363
+    },
+    {
+      "epoch": 6.680919080919081,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0002,
+      "loss": 0.8834,
+      "step": 8364
+    },
+    {
+      "epoch": 6.681718281718282,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.0002,
+      "loss": 0.8826,
+      "step": 8365
+    },
+    {
+      "epoch": 6.682517482517483,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0002,
+      "loss": 0.8844,
+      "step": 8366
+    },
+    {
+      "epoch": 6.683316683316683,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8774,
+      "step": 8367
+    },
+    {
+      "epoch": 6.684115884115884,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0002,
+      "loss": 0.8922,
+      "step": 8368
+    },
+    {
+      "epoch": 6.684915084915085,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0002,
+      "loss": 0.881,
+      "step": 8369
+    },
+    {
+      "epoch": 6.685714285714286,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8936,
+      "step": 8370
+    },
+    {
+      "epoch": 6.686513486513487,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8944,
+      "step": 8371
+    },
+    {
+      "epoch": 6.687312687312687,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.0002,
+      "loss": 0.8957,
+      "step": 8372
+    },
+    {
+      "epoch": 6.688111888111888,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0002,
+      "loss": 0.8902,
+      "step": 8373
+    },
+    {
+      "epoch": 6.688911088911089,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0002,
+      "loss": 0.8807,
+      "step": 8374
+    },
+    {
+      "epoch": 6.689710289710289,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0002,
+      "loss": 0.8869,
+      "step": 8375
+    },
+    {
+      "epoch": 6.69050949050949,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0002,
+      "loss": 0.8871,
+      "step": 8376
+    },
+    {
+      "epoch": 6.691308691308691,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8834,
+      "step": 8377
+    },
+    {
+      "epoch": 6.6921078921078925,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0002,
+      "loss": 0.8799,
+      "step": 8378
+    },
+    {
+      "epoch": 6.692907092907093,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0002,
+      "loss": 0.8764,
+      "step": 8379
+    },
+    {
+      "epoch": 6.693706293706294,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0002,
+      "loss": 0.8892,
+      "step": 8380
+    },
+    {
+      "epoch": 6.694505494505495,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8906,
+      "step": 8381
+    },
+    {
+      "epoch": 6.695304695304696,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0002,
+      "loss": 0.8895,
+      "step": 8382
+    },
+    {
+      "epoch": 6.696103896103896,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8862,
+      "step": 8383
+    },
+    {
+      "epoch": 6.696903096903097,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8902,
+      "step": 8384
+    },
+    {
+      "epoch": 6.697702297702298,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8887,
+      "step": 8385
+    },
+    {
+      "epoch": 6.698501498501498,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8819,
+      "step": 8386
+    },
+    {
+      "epoch": 6.699300699300699,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0002,
+      "loss": 0.8829,
+      "step": 8387
+    },
+    {
+      "epoch": 6.7000999000999,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0002,
+      "loss": 0.8831,
+      "step": 8388
+    },
+    {
+      "epoch": 6.700899100899101,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0002,
+      "loss": 0.9428,
+      "step": 8389
+    },
+    {
+      "epoch": 6.701698301698301,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0002,
+      "loss": 0.8845,
+      "step": 8390
+    },
+    {
+      "epoch": 6.702497502497502,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8903,
+      "step": 8391
+    },
+    {
+      "epoch": 6.7032967032967035,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0002,
+      "loss": 0.8948,
+      "step": 8392
+    },
+    {
+      "epoch": 6.7040959040959045,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0002,
+      "loss": 0.8801,
+      "step": 8393
+    },
+    {
+      "epoch": 6.704895104895105,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8824,
+      "step": 8394
+    },
+    {
+      "epoch": 6.705694305694306,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8811,
+      "step": 8395
+    },
+    {
+      "epoch": 6.706493506493507,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0002,
+      "loss": 0.8867,
+      "step": 8396
+    },
+    {
+      "epoch": 6.707292707292707,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8853,
+      "step": 8397
+    },
+    {
+      "epoch": 6.708091908091908,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0002,
+      "loss": 0.8945,
+      "step": 8398
+    },
+    {
+      "epoch": 6.708891108891109,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8902,
+      "step": 8399
+    },
+    {
+      "epoch": 6.70969030969031,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8796,
+      "step": 8400
+    },
+    {
+      "epoch": 6.71048951048951,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.0002,
+      "loss": 0.8875,
+      "step": 8401
+    },
+    {
+      "epoch": 6.711288711288711,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0002,
+      "loss": 0.8952,
+      "step": 8402
+    },
+    {
+      "epoch": 6.712087912087912,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0002,
+      "loss": 0.8938,
+      "step": 8403
+    },
+    {
+      "epoch": 6.712887112887113,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8759,
+      "step": 8404
+    },
+    {
+      "epoch": 6.713686313686313,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0002,
+      "loss": 0.8874,
+      "step": 8405
+    },
+    {
+      "epoch": 6.7144855144855145,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8884,
+      "step": 8406
+    },
+    {
+      "epoch": 6.7152847152847155,
+      "grad_norm": 0.80859375,
+      "learning_rate": 0.0002,
+      "loss": 0.9429,
+      "step": 8407
+    },
+    {
+      "epoch": 6.716083916083916,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8731,
+      "step": 8408
+    },
+    {
+      "epoch": 6.716883116883117,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0002,
+      "loss": 0.8929,
+      "step": 8409
+    },
+    {
+      "epoch": 6.717682317682318,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8899,
+      "step": 8410
+    },
+    {
+      "epoch": 6.718481518481519,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0002,
+      "loss": 0.8899,
+      "step": 8411
+    },
+    {
+      "epoch": 6.719280719280719,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0002,
+      "loss": 0.89,
+      "step": 8412
+    },
+    {
+      "epoch": 6.72007992007992,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0002,
+      "loss": 0.8825,
+      "step": 8413
+    },
+    {
+      "epoch": 6.720879120879121,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.0002,
+      "loss": 0.8828,
+      "step": 8414
+    },
+    {
+      "epoch": 6.721678321678322,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0002,
+      "loss": 0.901,
+      "step": 8415
+    },
+    {
+      "epoch": 6.722477522477522,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0002,
+      "loss": 0.8801,
+      "step": 8416
+    },
+    {
+      "epoch": 6.723276723276723,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0002,
+      "loss": 0.891,
+      "step": 8417
+    },
+    {
+      "epoch": 6.724075924075924,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0002,
+      "loss": 0.8815,
+      "step": 8418
+    },
+    {
+      "epoch": 6.724875124875124,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0002,
+      "loss": 0.882,
+      "step": 8419
+    },
+    {
+      "epoch": 6.725674325674325,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0002,
+      "loss": 0.8938,
+      "step": 8420
+    },
+    {
+      "epoch": 6.7264735264735265,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.0002,
+      "loss": 0.8871,
+      "step": 8421
+    },
+    {
+      "epoch": 6.7272727272727275,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8912,
+      "step": 8422
+    },
+    {
+      "epoch": 6.7280719280719286,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8927,
+      "step": 8423
+    },
+    {
+      "epoch": 6.728871128871129,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0002,
+      "loss": 0.879,
+      "step": 8424
+    },
+    {
+      "epoch": 6.72967032967033,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0002,
+      "loss": 0.8804,
+      "step": 8425
+    },
+    {
+      "epoch": 6.730469530469531,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8823,
+      "step": 8426
+    },
+    {
+      "epoch": 6.731268731268731,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0002,
+      "loss": 0.8868,
+      "step": 8427
+    },
+    {
+      "epoch": 6.732067932067932,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8839,
+      "step": 8428
+    },
+    {
+      "epoch": 6.732867132867133,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.0002,
+      "loss": 0.8853,
+      "step": 8429
+    },
+    {
+      "epoch": 6.733666333666334,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.0002,
+      "loss": 0.8874,
+      "step": 8430
+    },
+    {
+      "epoch": 6.734465534465534,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0002,
+      "loss": 0.8831,
+      "step": 8431
+    },
+    {
+      "epoch": 6.735264735264735,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.0002,
+      "loss": 0.8851,
+      "step": 8432
+    },
+    {
+      "epoch": 6.736063936063936,
+      "grad_norm": 0.375,
+      "learning_rate": 0.0002,
+      "loss": 0.886,
+      "step": 8433
+    },
+    {
+      "epoch": 6.736863136863137,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0002,
+      "loss": 0.8925,
+      "step": 8434
+    },
+    {
+      "epoch": 6.7376623376623375,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0002,
+      "loss": 0.899,
+      "step": 8435
+    },
+    {
+      "epoch": 6.7384615384615385,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8918,
+      "step": 8436
+    },
+    {
+      "epoch": 6.7392607392607395,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0002,
+      "loss": 0.8912,
+      "step": 8437
+    },
+    {
+      "epoch": 6.74005994005994,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8767,
+      "step": 8438
+    },
+    {
+      "epoch": 6.740859140859141,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8839,
+      "step": 8439
+    },
+    {
+      "epoch": 6.741658341658342,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0002,
+      "loss": 0.8919,
+      "step": 8440
+    },
+    {
+      "epoch": 6.742457542457543,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0002,
+      "loss": 0.884,
+      "step": 8441
+    },
+    {
+      "epoch": 6.743256743256743,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8863,
+      "step": 8442
+    },
+    {
+      "epoch": 6.744055944055944,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0002,
+      "loss": 0.8881,
+      "step": 8443
+    },
+    {
+      "epoch": 6.744855144855145,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8843,
+      "step": 8444
+    },
+    {
+      "epoch": 6.745654345654346,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8916,
+      "step": 8445
+    },
+    {
+      "epoch": 6.746453546453546,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8976,
+      "step": 8446
+    },
+    {
+      "epoch": 6.747252747252747,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.0002,
+      "loss": 0.8909,
+      "step": 8447
+    },
+    {
+      "epoch": 6.748051948051948,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0002,
+      "loss": 0.8822,
+      "step": 8448
+    },
+    {
+      "epoch": 6.7488511488511485,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0002,
+      "loss": 0.8845,
+      "step": 8449
+    },
+    {
+      "epoch": 6.7496503496503495,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8898,
+      "step": 8450
+    },
+    {
+      "epoch": 6.7504495504495505,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.0002,
+      "loss": 0.8796,
+      "step": 8451
+    },
+    {
+      "epoch": 6.751248751248752,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0002,
+      "loss": 0.8768,
+      "step": 8452
+    },
+    {
+      "epoch": 6.752047952047952,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8761,
+      "step": 8453
+    },
+    {
+      "epoch": 6.752847152847153,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0002,
+      "loss": 0.8819,
+      "step": 8454
+    },
+    {
+      "epoch": 6.753646353646354,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0002,
+      "loss": 0.8893,
+      "step": 8455
+    },
+    {
+      "epoch": 6.754445554445555,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0002,
+      "loss": 0.888,
+      "step": 8456
+    },
+    {
+      "epoch": 6.755244755244755,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0002,
+      "loss": 0.8795,
+      "step": 8457
+    },
+    {
+      "epoch": 6.756043956043956,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0002,
+      "loss": 0.8811,
+      "step": 8458
+    },
+    {
+      "epoch": 6.756843156843157,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.0002,
+      "loss": 0.8806,
+      "step": 8459
+    },
+    {
+      "epoch": 6.757642357642357,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0002,
+      "loss": 0.8919,
+      "step": 8460
+    },
+    {
+      "epoch": 6.758441558441558,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.0002,
+      "loss": 0.8809,
+      "step": 8461
+    },
+    {
+      "epoch": 6.759240759240759,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0002,
+      "loss": 0.8942,
+      "step": 8462
+    },
+    {
+      "epoch": 6.76003996003996,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.0002,
+      "loss": 0.8889,
+      "step": 8463
+    },
+    {
+      "epoch": 6.7608391608391605,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0002,
+      "loss": 0.8813,
+      "step": 8464
+    },
+    {
+      "epoch": 6.7616383616383615,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0002,
+      "loss": 0.8828,
+      "step": 8465
+    },
+    {
+      "epoch": 6.7624375624375626,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0002,
+      "loss": 0.8905,
+      "step": 8466
+    },
+    {
+      "epoch": 6.763236763236764,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.0002,
+      "loss": 0.896,
+      "step": 8467
+    },
+    {
+      "epoch": 6.764035964035964,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0002,
+      "loss": 0.8793,
+      "step": 8468
+    },
+    {
+      "epoch": 6.764835164835165,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.0002,
+      "loss": 0.8978,
+      "step": 8469
+    },
+    {
+      "epoch": 6.765634365634366,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0002,
+      "loss": 0.8826,
+      "step": 8470
+    },
+    {
+      "epoch": 6.766433566433566,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0002,
+      "loss": 0.8901,
+      "step": 8471
+    },
+    {
+      "epoch": 6.767232767232767,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.0002,
+      "loss": 0.8788,
+      "step": 8472
+    },
+    {
+      "epoch": 6.768031968031968,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0002,
+      "loss": 0.8842,
+      "step": 8473
+    },
+    {
+      "epoch": 6.768831168831169,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8917,
+      "step": 8474
+    },
+    {
+      "epoch": 6.76963036963037,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0002,
+      "loss": 0.8882,
+      "step": 8475
+    },
+    {
+      "epoch": 6.77042957042957,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0002,
+      "loss": 0.8909,
+      "step": 8476
+    },
+    {
+      "epoch": 6.771228771228771,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8884,
+      "step": 8477
+    },
+    {
+      "epoch": 6.772027972027972,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8935,
+      "step": 8478
+    },
+    {
+      "epoch": 6.7728271728271725,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0002,
+      "loss": 0.8915,
+      "step": 8479
+    },
+    {
+      "epoch": 6.7736263736263735,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.0002,
+      "loss": 0.8826,
+      "step": 8480
+    },
+    {
+      "epoch": 6.774425574425575,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8888,
+      "step": 8481
+    },
+    {
+      "epoch": 6.775224775224775,
+      "grad_norm": 0.8671875,
+      "learning_rate": 0.0002,
+      "loss": 0.9084,
+      "step": 8482
+    },
+    {
+      "epoch": 6.776023976023976,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8779,
+      "step": 8483
+    },
+    {
+      "epoch": 6.776823176823177,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0002,
+      "loss": 0.8828,
+      "step": 8484
+    },
+    {
+      "epoch": 6.777622377622378,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8906,
+      "step": 8485
+    },
+    {
+      "epoch": 6.778421578421579,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8845,
+      "step": 8486
+    },
+    {
+      "epoch": 6.779220779220779,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8869,
+      "step": 8487
+    },
+    {
+      "epoch": 6.78001998001998,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0002,
+      "loss": 0.891,
+      "step": 8488
+    },
+    {
+      "epoch": 6.780819180819181,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0002,
+      "loss": 0.8921,
+      "step": 8489
+    },
+    {
+      "epoch": 6.781618381618381,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8823,
+      "step": 8490
+    },
+    {
+      "epoch": 6.782417582417582,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8882,
+      "step": 8491
+    },
+    {
+      "epoch": 6.783216783216783,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0002,
+      "loss": 0.886,
+      "step": 8492
+    },
+    {
+      "epoch": 6.784015984015984,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0002,
+      "loss": 0.8835,
+      "step": 8493
+    },
+    {
+      "epoch": 6.7848151848151845,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0002,
+      "loss": 0.8855,
+      "step": 8494
+    },
+    {
+      "epoch": 6.785614385614386,
+      "grad_norm": 2.890625,
+      "learning_rate": 0.0002,
+      "loss": 0.9272,
+      "step": 8495
+    },
+    {
+      "epoch": 6.786413586413587,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8826,
+      "step": 8496
+    },
+    {
+      "epoch": 6.787212787212788,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0002,
+      "loss": 0.8776,
+      "step": 8497
+    },
+    {
+      "epoch": 6.788011988011988,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0002,
+      "loss": 0.8908,
+      "step": 8498
+    },
+    {
+      "epoch": 6.788811188811189,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0002,
+      "loss": 0.8964,
+      "step": 8499
+    },
+    {
+      "epoch": 6.78961038961039,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0002,
+      "loss": 0.8913,
+      "step": 8500
+    },
+    {
+      "epoch": 6.79040959040959,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.0002,
+      "loss": 0.8923,
+      "step": 8501
+    },
+    {
+      "epoch": 6.791208791208791,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0002,
+      "loss": 0.886,
+      "step": 8502
+    },
+    {
+      "epoch": 6.792007992007992,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.0002,
+      "loss": 0.8798,
+      "step": 8503
+    },
+    {
+      "epoch": 6.792807192807193,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.0002,
+      "loss": 0.8884,
+      "step": 8504
+    },
+    {
+      "epoch": 6.793606393606393,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.0002,
+      "loss": 0.8843,
+      "step": 8505
+    },
+    {
+      "epoch": 6.794405594405594,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0002,
+      "loss": 0.8842,
+      "step": 8506
+    },
+    {
+      "epoch": 6.795204795204795,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.0002,
+      "loss": 0.885,
+      "step": 8507
+    },
+    {
+      "epoch": 6.796003996003996,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0002,
+      "loss": 0.8867,
+      "step": 8508
+    },
+    {
+      "epoch": 6.796803196803197,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8833,
+      "step": 8509
+    },
+    {
+      "epoch": 6.797602397602398,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8949,
+      "step": 8510
+    },
+    {
+      "epoch": 6.798401598401599,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8842,
+      "step": 8511
+    },
+    {
+      "epoch": 6.799200799200799,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8969,
+      "step": 8512
+    },
+    {
+      "epoch": 6.8,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8869,
+      "step": 8513
+    },
+    {
+      "epoch": 6.800799200799201,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.0002,
+      "loss": 0.8946,
+      "step": 8514
+    },
+    {
+      "epoch": 6.801598401598402,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.0002,
+      "loss": 0.8839,
+      "step": 8515
+    },
+    {
+      "epoch": 6.802397602397602,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8841,
+      "step": 8516
+    },
+    {
+      "epoch": 6.803196803196803,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8843,
+      "step": 8517
+    },
+    {
+      "epoch": 6.803996003996004,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8867,
+      "step": 8518
+    },
+    {
+      "epoch": 6.804795204795205,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8879,
+      "step": 8519
+    },
+    {
+      "epoch": 6.805594405594405,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0002,
+      "loss": 0.8861,
+      "step": 8520
+    },
+    {
+      "epoch": 6.806393606393606,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8866,
+      "step": 8521
+    },
+    {
+      "epoch": 6.807192807192807,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0002,
+      "loss": 0.885,
+      "step": 8522
+    },
+    {
+      "epoch": 6.8079920079920075,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8897,
+      "step": 8523
+    },
+    {
+      "epoch": 6.808791208791209,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0002,
+      "loss": 0.8896,
+      "step": 8524
+    },
+    {
+      "epoch": 6.80959040959041,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8821,
+      "step": 8525
+    },
+    {
+      "epoch": 6.810389610389611,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8888,
+      "step": 8526
+    },
+    {
+      "epoch": 6.811188811188811,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0002,
+      "loss": 0.8901,
+      "step": 8527
+    },
+    {
+      "epoch": 6.811988011988012,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.0002,
+      "loss": 0.887,
+      "step": 8528
+    },
+    {
+      "epoch": 6.812787212787213,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8912,
+      "step": 8529
+    },
+    {
+      "epoch": 6.813586413586414,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0002,
+      "loss": 0.8833,
+      "step": 8530
+    },
+    {
+      "epoch": 6.814385614385614,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8844,
+      "step": 8531
+    },
+    {
+      "epoch": 6.815184815184815,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8822,
+      "step": 8532
+    },
+    {
+      "epoch": 6.815984015984016,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0002,
+      "loss": 0.8778,
+      "step": 8533
+    },
+    {
+      "epoch": 6.816783216783216,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8887,
+      "step": 8534
+    },
+    {
+      "epoch": 6.817582417582417,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8861,
+      "step": 8535
+    },
+    {
+      "epoch": 6.818381618381618,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0002,
+      "loss": 0.8842,
+      "step": 8536
+    },
+    {
+      "epoch": 6.819180819180819,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8875,
+      "step": 8537
+    },
+    {
+      "epoch": 6.8199800199800205,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0002,
+      "loss": 0.891,
+      "step": 8538
+    },
+    {
+      "epoch": 6.820779220779221,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8902,
+      "step": 8539
+    },
+    {
+      "epoch": 6.821578421578422,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8903,
+      "step": 8540
+    },
+    {
+      "epoch": 6.822377622377623,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8917,
+      "step": 8541
+    },
+    {
+      "epoch": 6.823176823176823,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0002,
+      "loss": 0.8873,
+      "step": 8542
+    },
+    {
+      "epoch": 6.823976023976024,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.0002,
+      "loss": 0.891,
+      "step": 8543
+    },
+    {
+      "epoch": 6.824775224775225,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0002,
+      "loss": 0.8897,
+      "step": 8544
+    },
+    {
+      "epoch": 6.825574425574426,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8892,
+      "step": 8545
+    },
+    {
+      "epoch": 6.826373626373626,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8865,
+      "step": 8546
+    },
+    {
+      "epoch": 6.827172827172827,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8866,
+      "step": 8547
+    },
+    {
+      "epoch": 6.827972027972028,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8817,
+      "step": 8548
+    },
+    {
+      "epoch": 6.828771228771229,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0002,
+      "loss": 0.8961,
+      "step": 8549
+    },
+    {
+      "epoch": 6.829570429570429,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9002,
+      "step": 8550
+    },
+    {
+      "epoch": 6.83036963036963,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0002,
+      "loss": 0.8824,
+      "step": 8551
+    },
+    {
+      "epoch": 6.8311688311688314,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8927,
+      "step": 8552
+    },
+    {
+      "epoch": 6.831968031968032,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0002,
+      "loss": 0.8852,
+      "step": 8553
+    },
+    {
+      "epoch": 6.832767232767233,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0002,
+      "loss": 0.8739,
+      "step": 8554
+    },
+    {
+      "epoch": 6.833566433566434,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0002,
+      "loss": 0.895,
+      "step": 8555
+    },
+    {
+      "epoch": 6.834365634365635,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8886,
+      "step": 8556
+    },
+    {
+      "epoch": 6.835164835164835,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8964,
+      "step": 8557
+    },
+    {
+      "epoch": 6.835964035964036,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8793,
+      "step": 8558
+    },
+    {
+      "epoch": 6.836763236763237,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8882,
+      "step": 8559
+    },
+    {
+      "epoch": 6.837562437562438,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0002,
+      "loss": 0.8904,
+      "step": 8560
+    },
+    {
+      "epoch": 6.838361638361638,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8909,
+      "step": 8561
+    },
+    {
+      "epoch": 6.839160839160839,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8879,
+      "step": 8562
+    },
+    {
+      "epoch": 6.83996003996004,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0002,
+      "loss": 0.8897,
+      "step": 8563
+    },
+    {
+      "epoch": 6.84075924075924,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0002,
+      "loss": 0.8817,
+      "step": 8564
+    },
+    {
+      "epoch": 6.841558441558441,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0002,
+      "loss": 0.8919,
+      "step": 8565
+    },
+    {
+      "epoch": 6.842357642357642,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8854,
+      "step": 8566
+    },
+    {
+      "epoch": 6.8431568431568435,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0002,
+      "loss": 0.8873,
+      "step": 8567
+    },
+    {
+      "epoch": 6.843956043956044,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8797,
+      "step": 8568
+    },
+    {
+      "epoch": 6.844755244755245,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0002,
+      "loss": 0.886,
+      "step": 8569
+    },
+    {
+      "epoch": 6.845554445554446,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0002,
+      "loss": 0.8824,
+      "step": 8570
+    },
+    {
+      "epoch": 6.846353646353647,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0002,
+      "loss": 0.894,
+      "step": 8571
+    },
+    {
+      "epoch": 6.847152847152847,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8847,
+      "step": 8572
+    },
+    {
+      "epoch": 6.847952047952048,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8857,
+      "step": 8573
+    },
+    {
+      "epoch": 6.848751248751249,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0002,
+      "loss": 0.8902,
+      "step": 8574
+    },
+    {
+      "epoch": 6.849550449550449,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0002,
+      "loss": 0.88,
+      "step": 8575
+    },
+    {
+      "epoch": 6.85034965034965,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0002,
+      "loss": 0.8863,
+      "step": 8576
+    },
+    {
+      "epoch": 6.851148851148851,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8856,
+      "step": 8577
+    },
+    {
+      "epoch": 6.851948051948052,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0002,
+      "loss": 0.8834,
+      "step": 8578
+    },
+    {
+      "epoch": 6.852747252747252,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0002,
+      "loss": 0.8853,
+      "step": 8579
+    },
+    {
+      "epoch": 6.853546453546453,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8879,
+      "step": 8580
+    },
+    {
+      "epoch": 6.8543456543456545,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0002,
+      "loss": 0.8913,
+      "step": 8581
+    },
+    {
+      "epoch": 6.8551448551448555,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.0002,
+      "loss": 0.8822,
+      "step": 8582
+    },
+    {
+      "epoch": 6.855944055944056,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0002,
+      "loss": 0.8896,
+      "step": 8583
+    },
+    {
+      "epoch": 6.856743256743257,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0002,
+      "loss": 0.8967,
+      "step": 8584
+    },
+    {
+      "epoch": 6.857542457542458,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.0002,
+      "loss": 0.8962,
+      "step": 8585
+    },
+    {
+      "epoch": 6.858341658341658,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.0002,
+      "loss": 0.8904,
+      "step": 8586
+    },
+    {
+      "epoch": 6.859140859140859,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0002,
+      "loss": 0.8919,
+      "step": 8587
+    },
+    {
+      "epoch": 6.85994005994006,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0002,
+      "loss": 0.8884,
+      "step": 8588
+    },
+    {
+      "epoch": 6.860739260739261,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8887,
+      "step": 8589
+    },
+    {
+      "epoch": 6.861538461538462,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8891,
+      "step": 8590
+    },
+    {
+      "epoch": 6.862337662337662,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8846,
+      "step": 8591
+    },
+    {
+      "epoch": 6.863136863136863,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0002,
+      "loss": 0.8882,
+      "step": 8592
+    },
+    {
+      "epoch": 6.863936063936064,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0002,
+      "loss": 0.8823,
+      "step": 8593
+    },
+    {
+      "epoch": 6.864735264735264,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0002,
+      "loss": 0.8833,
+      "step": 8594
+    },
+    {
+      "epoch": 6.8655344655344654,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8847,
+      "step": 8595
+    },
+    {
+      "epoch": 6.8663336663336665,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0002,
+      "loss": 0.8948,
+      "step": 8596
+    },
+    {
+      "epoch": 6.867132867132867,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0002,
+      "loss": 0.8848,
+      "step": 8597
+    },
+    {
+      "epoch": 6.867932067932068,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8831,
+      "step": 8598
+    },
+    {
+      "epoch": 6.868731268731269,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8905,
+      "step": 8599
+    },
+    {
+      "epoch": 6.86953046953047,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0002,
+      "loss": 0.8895,
+      "step": 8600
+    },
+    {
+      "epoch": 6.870329670329671,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8828,
+      "step": 8601
+    },
+    {
+      "epoch": 6.871128871128871,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0002,
+      "loss": 0.8797,
+      "step": 8602
+    },
+    {
+      "epoch": 6.871928071928072,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0002,
+      "loss": 0.8834,
+      "step": 8603
+    },
+    {
+      "epoch": 6.872727272727273,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8911,
+      "step": 8604
+    },
+    {
+      "epoch": 6.873526473526473,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8938,
+      "step": 8605
+    },
+    {
+      "epoch": 6.874325674325674,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0002,
+      "loss": 0.8908,
+      "step": 8606
+    },
+    {
+      "epoch": 6.875124875124875,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8875,
+      "step": 8607
+    },
+    {
+      "epoch": 6.875924075924076,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8884,
+      "step": 8608
+    },
+    {
+      "epoch": 6.876723276723276,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8801,
+      "step": 8609
+    },
+    {
+      "epoch": 6.8775224775224775,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8971,
+      "step": 8610
+    },
+    {
+      "epoch": 6.8783216783216785,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.0002,
+      "loss": 0.8947,
+      "step": 8611
+    },
+    {
+      "epoch": 6.8791208791208796,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8821,
+      "step": 8612
+    },
+    {
+      "epoch": 6.87992007992008,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0002,
+      "loss": 0.8868,
+      "step": 8613
+    },
+    {
+      "epoch": 6.880719280719281,
+      "grad_norm": 0.375,
+      "learning_rate": 0.0002,
+      "loss": 0.8852,
+      "step": 8614
+    },
+    {
+      "epoch": 6.881518481518482,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8913,
+      "step": 8615
+    },
+    {
+      "epoch": 6.882317682317682,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.0002,
+      "loss": 0.8911,
+      "step": 8616
+    },
+    {
+      "epoch": 6.883116883116883,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0002,
+      "loss": 0.8831,
+      "step": 8617
+    },
+    {
+      "epoch": 6.883916083916084,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8815,
+      "step": 8618
+    },
+    {
+      "epoch": 6.884715284715285,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8936,
+      "step": 8619
+    },
+    {
+      "epoch": 6.885514485514485,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0002,
+      "loss": 0.886,
+      "step": 8620
+    },
+    {
+      "epoch": 6.886313686313686,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8822,
+      "step": 8621
+    },
+    {
+      "epoch": 6.887112887112887,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0002,
+      "loss": 0.886,
+      "step": 8622
+    },
+    {
+      "epoch": 6.887912087912088,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8951,
+      "step": 8623
+    },
+    {
+      "epoch": 6.8887112887112885,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8844,
+      "step": 8624
+    },
+    {
+      "epoch": 6.8895104895104895,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0002,
+      "loss": 0.8827,
+      "step": 8625
+    },
+    {
+      "epoch": 6.8903096903096905,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0002,
+      "loss": 0.889,
+      "step": 8626
+    },
+    {
+      "epoch": 6.891108891108891,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8819,
+      "step": 8627
+    },
+    {
+      "epoch": 6.891908091908092,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.0002,
+      "loss": 0.888,
+      "step": 8628
+    },
+    {
+      "epoch": 6.892707292707293,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8819,
+      "step": 8629
+    },
+    {
+      "epoch": 6.893506493506494,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0002,
+      "loss": 0.891,
+      "step": 8630
+    },
+    {
+      "epoch": 6.894305694305694,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8865,
+      "step": 8631
+    },
+    {
+      "epoch": 6.895104895104895,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0002,
+      "loss": 0.8855,
+      "step": 8632
+    },
+    {
+      "epoch": 6.895904095904096,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8849,
+      "step": 8633
+    },
+    {
+      "epoch": 6.896703296703297,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8898,
+      "step": 8634
+    },
+    {
+      "epoch": 6.897502497502497,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0002,
+      "loss": 0.8869,
+      "step": 8635
+    },
+    {
+      "epoch": 6.898301698301698,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8808,
+      "step": 8636
+    },
+    {
+      "epoch": 6.899100899100899,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0002,
+      "loss": 0.8896,
+      "step": 8637
+    },
+    {
+      "epoch": 6.8999000999000994,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8801,
+      "step": 8638
+    },
+    {
+      "epoch": 6.9006993006993005,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8952,
+      "step": 8639
+    },
+    {
+      "epoch": 6.9014985014985015,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0002,
+      "loss": 0.8867,
+      "step": 8640
+    },
+    {
+      "epoch": 6.902297702297703,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0002,
+      "loss": 0.8767,
+      "step": 8641
+    },
+    {
+      "epoch": 6.903096903096903,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0002,
+      "loss": 0.8943,
+      "step": 8642
+    },
+    {
+      "epoch": 6.903896103896104,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8817,
+      "step": 8643
+    },
+    {
+      "epoch": 6.904695304695305,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8882,
+      "step": 8644
+    },
+    {
+      "epoch": 6.905494505494506,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8846,
+      "step": 8645
+    },
+    {
+      "epoch": 6.906293706293706,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8908,
+      "step": 8646
+    },
+    {
+      "epoch": 6.907092907092907,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8865,
+      "step": 8647
+    },
+    {
+      "epoch": 6.907892107892108,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0002,
+      "loss": 0.8827,
+      "step": 8648
+    },
+    {
+      "epoch": 6.908691308691308,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0002,
+      "loss": 0.8906,
+      "step": 8649
+    },
+    {
+      "epoch": 6.909490509490509,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8867,
+      "step": 8650
+    },
+    {
+      "epoch": 6.91028971028971,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8905,
+      "step": 8651
+    },
+    {
+      "epoch": 6.911088911088911,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8825,
+      "step": 8652
+    },
+    {
+      "epoch": 6.911888111888112,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8799,
+      "step": 8653
+    },
+    {
+      "epoch": 6.9126873126873125,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8802,
+      "step": 8654
+    },
+    {
+      "epoch": 6.9134865134865136,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0002,
+      "loss": 0.8903,
+      "step": 8655
+    },
+    {
+      "epoch": 6.914285714285715,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8878,
+      "step": 8656
+    },
+    {
+      "epoch": 6.915084915084915,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8759,
+      "step": 8657
+    },
+    {
+      "epoch": 6.915884115884116,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0002,
+      "loss": 0.891,
+      "step": 8658
+    },
+    {
+      "epoch": 6.916683316683317,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.0002,
+      "loss": 0.886,
+      "step": 8659
+    },
+    {
+      "epoch": 6.917482517482518,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0002,
+      "loss": 0.8857,
+      "step": 8660
+    },
+    {
+      "epoch": 6.918281718281718,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.0002,
+      "loss": 0.895,
+      "step": 8661
+    },
+    {
+      "epoch": 6.919080919080919,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0002,
+      "loss": 0.8832,
+      "step": 8662
+    },
+    {
+      "epoch": 6.91988011988012,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0002,
+      "loss": 0.9028,
+      "step": 8663
+    },
+    {
+      "epoch": 6.920679320679321,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0002,
+      "loss": 0.8845,
+      "step": 8664
+    },
+    {
+      "epoch": 6.921478521478521,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0002,
+      "loss": 0.9007,
+      "step": 8665
+    },
+    {
+      "epoch": 6.922277722277722,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0002,
+      "loss": 0.8909,
+      "step": 8666
+    },
+    {
+      "epoch": 6.923076923076923,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0002,
+      "loss": 0.8897,
+      "step": 8667
+    },
+    {
+      "epoch": 6.9238761238761235,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0002,
+      "loss": 0.8737,
+      "step": 8668
+    },
+    {
+      "epoch": 6.9246753246753245,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.0002,
+      "loss": 0.8848,
+      "step": 8669
+    },
+    {
+      "epoch": 6.925474525474526,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0002,
+      "loss": 0.8896,
+      "step": 8670
+    },
+    {
+      "epoch": 6.926273726273727,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8842,
+      "step": 8671
+    },
+    {
+      "epoch": 6.927072927072927,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.0002,
+      "loss": 0.8833,
+      "step": 8672
+    },
+    {
+      "epoch": 6.927872127872128,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8967,
+      "step": 8673
+    },
+    {
+      "epoch": 6.928671328671329,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0002,
+      "loss": 0.8758,
+      "step": 8674
+    },
+    {
+      "epoch": 6.92947052947053,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8868,
+      "step": 8675
+    },
+    {
+      "epoch": 6.93026973026973,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0002,
+      "loss": 0.8951,
+      "step": 8676
+    },
+    {
+      "epoch": 6.931068931068931,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0002,
+      "loss": 0.8907,
+      "step": 8677
+    },
+    {
+      "epoch": 6.931868131868132,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.0002,
+      "loss": 0.8831,
+      "step": 8678
+    },
+    {
+      "epoch": 6.932667332667332,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0002,
+      "loss": 0.886,
+      "step": 8679
+    },
+    {
+      "epoch": 6.933466533466533,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0002,
+      "loss": 0.8931,
+      "step": 8680
+    },
+    {
+      "epoch": 6.934265734265734,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8891,
+      "step": 8681
+    },
+    {
+      "epoch": 6.935064935064935,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0002,
+      "loss": 0.8978,
+      "step": 8682
+    },
+    {
+      "epoch": 6.9358641358641355,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0002,
+      "loss": 0.8899,
+      "step": 8683
+    },
+    {
+      "epoch": 6.936663336663337,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0002,
+      "loss": 0.8932,
+      "step": 8684
+    },
+    {
+      "epoch": 6.937462537462538,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0002,
+      "loss": 0.8918,
+      "step": 8685
+    },
+    {
+      "epoch": 6.938261738261739,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8798,
+      "step": 8686
+    },
+    {
+      "epoch": 6.939060939060939,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8868,
+      "step": 8687
+    },
+    {
+      "epoch": 6.93986013986014,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0002,
+      "loss": 0.8814,
+      "step": 8688
+    },
+    {
+      "epoch": 6.940659340659341,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8886,
+      "step": 8689
+    },
+    {
+      "epoch": 6.941458541458541,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8871,
+      "step": 8690
+    },
+    {
+      "epoch": 6.942257742257742,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0002,
+      "loss": 0.8913,
+      "step": 8691
+    },
+    {
+      "epoch": 6.943056943056943,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0002,
+      "loss": 0.8959,
+      "step": 8692
+    },
+    {
+      "epoch": 6.943856143856144,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8929,
+      "step": 8693
+    },
+    {
+      "epoch": 6.944655344655344,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8765,
+      "step": 8694
+    },
+    {
+      "epoch": 6.945454545454545,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0002,
+      "loss": 0.8851,
+      "step": 8695
+    },
+    {
+      "epoch": 6.946253746253746,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0002,
+      "loss": 0.892,
+      "step": 8696
+    },
+    {
+      "epoch": 6.947052947052947,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8838,
+      "step": 8697
+    },
+    {
+      "epoch": 6.9478521478521476,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0002,
+      "loss": 0.8937,
+      "step": 8698
+    },
+    {
+      "epoch": 6.948651348651349,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8819,
+      "step": 8699
+    },
+    {
+      "epoch": 6.94945054945055,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8832,
+      "step": 8700
+    },
+    {
+      "epoch": 6.95024975024975,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8804,
+      "step": 8701
+    },
+    {
+      "epoch": 6.951048951048951,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8889,
+      "step": 8702
+    },
+    {
+      "epoch": 6.951848151848152,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0002,
+      "loss": 0.8918,
+      "step": 8703
+    },
+    {
+      "epoch": 6.952647352647353,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8949,
+      "step": 8704
+    },
+    {
+      "epoch": 6.953446553446554,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0002,
+      "loss": 0.875,
+      "step": 8705
+    },
+    {
+      "epoch": 6.954245754245754,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8868,
+      "step": 8706
+    },
+    {
+      "epoch": 6.955044955044955,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0002,
+      "loss": 0.8875,
+      "step": 8707
+    },
+    {
+      "epoch": 6.955844155844156,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8939,
+      "step": 8708
+    },
+    {
+      "epoch": 6.956643356643356,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0002,
+      "loss": 0.883,
+      "step": 8709
+    },
+    {
+      "epoch": 6.957442557442557,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8896,
+      "step": 8710
+    },
+    {
+      "epoch": 6.958241758241758,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0002,
+      "loss": 0.8918,
+      "step": 8711
+    },
+    {
+      "epoch": 6.959040959040959,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8874,
+      "step": 8712
+    },
+    {
+      "epoch": 6.95984015984016,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8853,
+      "step": 8713
+    },
+    {
+      "epoch": 6.960639360639361,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0002,
+      "loss": 0.8892,
+      "step": 8714
+    },
+    {
+      "epoch": 6.961438561438562,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8879,
+      "step": 8715
+    },
+    {
+      "epoch": 6.962237762237763,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0002,
+      "loss": 0.8823,
+      "step": 8716
+    },
+    {
+      "epoch": 6.963036963036963,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8939,
+      "step": 8717
+    },
+    {
+      "epoch": 6.963836163836164,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8829,
+      "step": 8718
+    },
+    {
+      "epoch": 6.964635364635365,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.0002,
+      "loss": 0.8839,
+      "step": 8719
+    },
+    {
+      "epoch": 6.965434565434565,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8937,
+      "step": 8720
+    },
+    {
+      "epoch": 6.966233766233766,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0002,
+      "loss": 0.8872,
+      "step": 8721
+    },
+    {
+      "epoch": 6.967032967032967,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8851,
+      "step": 8722
+    },
+    {
+      "epoch": 6.967832167832168,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0002,
+      "loss": 0.882,
+      "step": 8723
+    },
+    {
+      "epoch": 6.968631368631368,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0002,
+      "loss": 0.8821,
+      "step": 8724
+    },
+    {
+      "epoch": 6.969430569430569,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0002,
+      "loss": 0.882,
+      "step": 8725
+    },
+    {
+      "epoch": 6.97022977022977,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.0002,
+      "loss": 0.893,
+      "step": 8726
+    },
+    {
+      "epoch": 6.9710289710289715,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8905,
+      "step": 8727
+    },
+    {
+      "epoch": 6.971828171828172,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0002,
+      "loss": 0.8827,
+      "step": 8728
+    },
+    {
+      "epoch": 6.972627372627373,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8859,
+      "step": 8729
+    },
+    {
+      "epoch": 6.973426573426574,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0002,
+      "loss": 0.8907,
+      "step": 8730
+    },
+    {
+      "epoch": 6.974225774225774,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8867,
+      "step": 8731
+    },
+    {
+      "epoch": 6.975024975024975,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8837,
+      "step": 8732
+    },
+    {
+      "epoch": 6.975824175824176,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8774,
+      "step": 8733
+    },
+    {
+      "epoch": 6.976623376623377,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8925,
+      "step": 8734
+    },
+    {
+      "epoch": 6.977422577422577,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8915,
+      "step": 8735
+    },
+    {
+      "epoch": 6.978221778221778,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8867,
+      "step": 8736
+    },
+    {
+      "epoch": 6.979020979020979,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0002,
+      "loss": 0.8886,
+      "step": 8737
+    },
+    {
+      "epoch": 6.97982017982018,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8819,
+      "step": 8738
+    },
+    {
+      "epoch": 6.98061938061938,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0002,
+      "loss": 0.8936,
+      "step": 8739
+    },
+    {
+      "epoch": 6.981418581418581,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8836,
+      "step": 8740
+    },
+    {
+      "epoch": 6.982217782217782,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8829,
+      "step": 8741
+    },
+    {
+      "epoch": 6.983016983016983,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8882,
+      "step": 8742
+    },
+    {
+      "epoch": 6.983816183816184,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0002,
+      "loss": 0.8873,
+      "step": 8743
+    },
+    {
+      "epoch": 6.984615384615385,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0002,
+      "loss": 0.8996,
+      "step": 8744
+    },
+    {
+      "epoch": 6.985414585414586,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8853,
+      "step": 8745
+    },
+    {
+      "epoch": 6.986213786213786,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8872,
+      "step": 8746
+    },
+    {
+      "epoch": 6.987012987012987,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8843,
+      "step": 8747
+    },
+    {
+      "epoch": 6.987812187812188,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8859,
+      "step": 8748
+    },
+    {
+      "epoch": 6.988611388611389,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8824,
+      "step": 8749
+    },
+    {
+      "epoch": 6.989410589410589,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0002,
+      "loss": 0.8859,
+      "step": 8750
+    },
+    {
+      "epoch": 6.99020979020979,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8856,
+      "step": 8751
+    },
+    {
+      "epoch": 6.991008991008991,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8896,
+      "step": 8752
+    },
+    {
+      "epoch": 6.991808191808191,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0002,
+      "loss": 0.886,
+      "step": 8753
+    },
+    {
+      "epoch": 6.992607392607392,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8931,
+      "step": 8754
+    },
+    {
+      "epoch": 6.993406593406593,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0002,
+      "loss": 0.8883,
+      "step": 8755
+    },
+    {
+      "epoch": 6.9942057942057945,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8698,
+      "step": 8756
+    },
+    {
+      "epoch": 6.995004995004995,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8882,
+      "step": 8757
+    },
+    {
+      "epoch": 6.995804195804196,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8783,
+      "step": 8758
+    },
+    {
+      "epoch": 6.996603396603397,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0002,
+      "loss": 0.8868,
+      "step": 8759
+    },
+    {
+      "epoch": 6.997402597402598,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0002,
+      "loss": 0.8867,
+      "step": 8760
+    },
+    {
+      "epoch": 6.998201798201798,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8908,
+      "step": 8761
+    },
+    {
+      "epoch": 6.999000999000999,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8962,
+      "step": 8762
+    },
+    {
+      "epoch": 6.9998001998002,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0002,
+      "loss": 0.887,
+      "step": 8763
+    },
+    {
+      "epoch": 7.0,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0002,
+      "loss": 0.22,
+      "step": 8764
+    },
+    {
+      "epoch": 7.000799200799201,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0002,
+      "loss": 0.9319,
+      "step": 8765
+    },
+    {
+      "epoch": 7.001598401598401,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8889,
+      "step": 8766
+    },
+    {
+      "epoch": 7.002397602397602,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8969,
+      "step": 8767
+    },
+    {
+      "epoch": 7.003196803196803,
+      "grad_norm": 1.5078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8984,
+      "step": 8768
+    },
+    {
+      "epoch": 7.003996003996004,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0002,
+      "loss": 0.8816,
+      "step": 8769
+    },
+    {
+      "epoch": 7.0047952047952045,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8892,
+      "step": 8770
+    },
+    {
+      "epoch": 7.0055944055944055,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8886,
+      "step": 8771
+    },
+    {
+      "epoch": 7.0063936063936065,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.0002,
+      "loss": 0.8915,
+      "step": 8772
+    },
+    {
+      "epoch": 7.007192807192808,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0002,
+      "loss": 0.8949,
+      "step": 8773
+    },
+    {
+      "epoch": 7.007992007992008,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8836,
+      "step": 8774
+    },
+    {
+      "epoch": 7.008791208791209,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8904,
+      "step": 8775
+    },
+    {
+      "epoch": 7.00959040959041,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8855,
+      "step": 8776
+    },
+    {
+      "epoch": 7.01038961038961,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0002,
+      "loss": 0.8838,
+      "step": 8777
+    },
+    {
+      "epoch": 7.011188811188811,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8963,
+      "step": 8778
+    },
+    {
+      "epoch": 7.011988011988012,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0002,
+      "loss": 0.884,
+      "step": 8779
+    },
+    {
+      "epoch": 7.012787212787213,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0002,
+      "loss": 0.8924,
+      "step": 8780
+    },
+    {
+      "epoch": 7.013586413586413,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8805,
+      "step": 8781
+    },
+    {
+      "epoch": 7.014385614385614,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0002,
+      "loss": 0.8801,
+      "step": 8782
+    },
+    {
+      "epoch": 7.015184815184815,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0002,
+      "loss": 0.8809,
+      "step": 8783
+    },
+    {
+      "epoch": 7.015984015984016,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0002,
+      "loss": 0.8947,
+      "step": 8784
+    },
+    {
+      "epoch": 7.0167832167832165,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8954,
+      "step": 8785
+    },
+    {
+      "epoch": 7.0175824175824175,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0002,
+      "loss": 0.889,
+      "step": 8786
+    },
+    {
+      "epoch": 7.018381618381619,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8946,
+      "step": 8787
+    },
+    {
+      "epoch": 7.01918081918082,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0002,
+      "loss": 0.8809,
+      "step": 8788
+    },
+    {
+      "epoch": 7.01998001998002,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8891,
+      "step": 8789
+    },
+    {
+      "epoch": 7.020779220779221,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.0002,
+      "loss": 0.908,
+      "step": 8790
+    },
+    {
+      "epoch": 7.021578421578422,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0002,
+      "loss": 0.8894,
+      "step": 8791
+    },
+    {
+      "epoch": 7.022377622377622,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8822,
+      "step": 8792
+    },
+    {
+      "epoch": 7.023176823176823,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.0002,
+      "loss": 0.8854,
+      "step": 8793
+    },
+    {
+      "epoch": 7.023976023976024,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0002,
+      "loss": 0.8836,
+      "step": 8794
+    },
+    {
+      "epoch": 7.024775224775225,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8881,
+      "step": 8795
+    },
+    {
+      "epoch": 7.025574425574425,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8857,
+      "step": 8796
+    },
+    {
+      "epoch": 7.026373626373626,
+      "grad_norm": 0.375,
+      "learning_rate": 0.0002,
+      "loss": 0.876,
+      "step": 8797
+    },
+    {
+      "epoch": 7.027172827172827,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8844,
+      "step": 8798
+    },
+    {
+      "epoch": 7.027972027972028,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.0002,
+      "loss": 0.892,
+      "step": 8799
+    },
+    {
+      "epoch": 7.0287712287712285,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0002,
+      "loss": 0.8886,
+      "step": 8800
+    },
+    {
+      "epoch": 7.0295704295704295,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0002,
+      "loss": 0.888,
+      "step": 8801
+    },
+    {
+      "epoch": 7.030369630369631,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.0002,
+      "loss": 0.882,
+      "step": 8802
+    },
+    {
+      "epoch": 7.031168831168831,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0002,
+      "loss": 0.8946,
+      "step": 8803
+    },
+    {
+      "epoch": 7.031968031968032,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8795,
+      "step": 8804
+    },
+    {
+      "epoch": 7.032767232767233,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0002,
+      "loss": 0.885,
+      "step": 8805
+    },
+    {
+      "epoch": 7.033566433566434,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8899,
+      "step": 8806
+    },
+    {
+      "epoch": 7.034365634365634,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8913,
+      "step": 8807
+    },
+    {
+      "epoch": 7.035164835164835,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0002,
+      "loss": 0.8792,
+      "step": 8808
+    },
+    {
+      "epoch": 7.035964035964036,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8942,
+      "step": 8809
+    },
+    {
+      "epoch": 7.036763236763237,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8874,
+      "step": 8810
+    },
+    {
+      "epoch": 7.037562437562437,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0002,
+      "loss": 0.8866,
+      "step": 8811
+    },
+    {
+      "epoch": 7.038361638361638,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8918,
+      "step": 8812
+    },
+    {
+      "epoch": 7.039160839160839,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8886,
+      "step": 8813
+    },
+    {
+      "epoch": 7.03996003996004,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0002,
+      "loss": 0.8756,
+      "step": 8814
+    },
+    {
+      "epoch": 7.0407592407592405,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8812,
+      "step": 8815
+    },
+    {
+      "epoch": 7.041558441558442,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8808,
+      "step": 8816
+    },
+    {
+      "epoch": 7.042357642357643,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0002,
+      "loss": 0.8937,
+      "step": 8817
+    },
+    {
+      "epoch": 7.043156843156843,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.0002,
+      "loss": 0.8788,
+      "step": 8818
+    },
+    {
+      "epoch": 7.043956043956044,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0002,
+      "loss": 0.883,
+      "step": 8819
+    },
+    {
+      "epoch": 7.044755244755245,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8906,
+      "step": 8820
+    },
+    {
+      "epoch": 7.045554445554446,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0002,
+      "loss": 0.8851,
+      "step": 8821
+    },
+    {
+      "epoch": 7.046353646353646,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0002,
+      "loss": 0.8719,
+      "step": 8822
+    },
+    {
+      "epoch": 7.047152847152847,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0002,
+      "loss": 0.8837,
+      "step": 8823
+    },
+    {
+      "epoch": 7.047952047952048,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8869,
+      "step": 8824
+    },
+    {
+      "epoch": 7.048751248751249,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0002,
+      "loss": 0.8879,
+      "step": 8825
+    },
+    {
+      "epoch": 7.049550449550449,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0002,
+      "loss": 0.888,
+      "step": 8826
+    },
+    {
+      "epoch": 7.05034965034965,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0002,
+      "loss": 0.8883,
+      "step": 8827
+    },
+    {
+      "epoch": 7.051148851148851,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8844,
+      "step": 8828
+    },
+    {
+      "epoch": 7.0519480519480515,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8771,
+      "step": 8829
+    },
+    {
+      "epoch": 7.052747252747253,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0002,
+      "loss": 0.8811,
+      "step": 8830
+    },
+    {
+      "epoch": 7.053546453546454,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8941,
+      "step": 8831
+    },
+    {
+      "epoch": 7.054345654345655,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8848,
+      "step": 8832
+    },
+    {
+      "epoch": 7.055144855144855,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0002,
+      "loss": 0.8846,
+      "step": 8833
+    },
+    {
+      "epoch": 7.055944055944056,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0002,
+      "loss": 0.8822,
+      "step": 8834
+    },
+    {
+      "epoch": 7.056743256743257,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.0002,
+      "loss": 0.8855,
+      "step": 8835
+    },
+    {
+      "epoch": 7.057542457542458,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0002,
+      "loss": 0.886,
+      "step": 8836
+    },
+    {
+      "epoch": 7.058341658341658,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0002,
+      "loss": 0.8896,
+      "step": 8837
+    },
+    {
+      "epoch": 7.059140859140859,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0002,
+      "loss": 0.8873,
+      "step": 8838
+    },
+    {
+      "epoch": 7.05994005994006,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8869,
+      "step": 8839
+    },
+    {
+      "epoch": 7.060739260739261,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0002,
+      "loss": 0.8817,
+      "step": 8840
+    },
+    {
+      "epoch": 7.061538461538461,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0002,
+      "loss": 0.8792,
+      "step": 8841
+    },
+    {
+      "epoch": 7.062337662337662,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0002,
+      "loss": 0.8822,
+      "step": 8842
+    },
+    {
+      "epoch": 7.063136863136863,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8866,
+      "step": 8843
+    },
+    {
+      "epoch": 7.0639360639360635,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8795,
+      "step": 8844
+    },
+    {
+      "epoch": 7.064735264735265,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0002,
+      "loss": 0.8899,
+      "step": 8845
+    },
+    {
+      "epoch": 7.065534465534466,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0002,
+      "loss": 0.879,
+      "step": 8846
+    },
+    {
+      "epoch": 7.066333666333667,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0002,
+      "loss": 0.8874,
+      "step": 8847
+    },
+    {
+      "epoch": 7.067132867132867,
+      "grad_norm": 3.140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9294,
+      "step": 8848
+    },
+    {
+      "epoch": 7.067932067932068,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8847,
+      "step": 8849
+    },
+    {
+      "epoch": 7.068731268731269,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8826,
+      "step": 8850
+    },
+    {
+      "epoch": 7.06953046953047,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0002,
+      "loss": 0.8854,
+      "step": 8851
+    },
+    {
+      "epoch": 7.07032967032967,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8868,
+      "step": 8852
+    },
+    {
+      "epoch": 7.071128871128871,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8916,
+      "step": 8853
+    },
+    {
+      "epoch": 7.071928071928072,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8847,
+      "step": 8854
+    },
+    {
+      "epoch": 7.072727272727272,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0002,
+      "loss": 0.872,
+      "step": 8855
+    },
+    {
+      "epoch": 7.073526473526473,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0002,
+      "loss": 0.8929,
+      "step": 8856
+    },
+    {
+      "epoch": 7.074325674325674,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8827,
+      "step": 8857
+    },
+    {
+      "epoch": 7.075124875124875,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0002,
+      "loss": 0.8791,
+      "step": 8858
+    },
+    {
+      "epoch": 7.075924075924076,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0002,
+      "loss": 0.877,
+      "step": 8859
+    },
+    {
+      "epoch": 7.076723276723277,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0002,
+      "loss": 0.8854,
+      "step": 8860
+    },
+    {
+      "epoch": 7.077522477522478,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.0002,
+      "loss": 0.89,
+      "step": 8861
+    },
+    {
+      "epoch": 7.078321678321679,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8949,
+      "step": 8862
+    },
+    {
+      "epoch": 7.079120879120879,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0002,
+      "loss": 0.8859,
+      "step": 8863
+    },
+    {
+      "epoch": 7.07992007992008,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8881,
+      "step": 8864
+    },
+    {
+      "epoch": 7.080719280719281,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8939,
+      "step": 8865
+    },
+    {
+      "epoch": 7.081518481518482,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0002,
+      "loss": 0.8905,
+      "step": 8866
+    },
+    {
+      "epoch": 7.082317682317682,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0002,
+      "loss": 0.886,
+      "step": 8867
+    },
+    {
+      "epoch": 7.083116883116883,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.0002,
+      "loss": 0.8919,
+      "step": 8868
+    },
+    {
+      "epoch": 7.083916083916084,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.0002,
+      "loss": 0.8808,
+      "step": 8869
+    },
+    {
+      "epoch": 7.084715284715284,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8817,
+      "step": 8870
+    },
+    {
+      "epoch": 7.085514485514485,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8845,
+      "step": 8871
+    },
+    {
+      "epoch": 7.086313686313686,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.0002,
+      "loss": 0.8904,
+      "step": 8872
+    },
+    {
+      "epoch": 7.0871128871128874,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.0002,
+      "loss": 0.8843,
+      "step": 8873
+    },
+    {
+      "epoch": 7.087912087912088,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.0002,
+      "loss": 0.8852,
+      "step": 8874
+    },
+    {
+      "epoch": 7.088711288711289,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.0002,
+      "loss": 0.8883,
+      "step": 8875
+    },
+    {
+      "epoch": 7.08951048951049,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8786,
+      "step": 8876
+    },
+    {
+      "epoch": 7.090309690309691,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.0002,
+      "loss": 0.8849,
+      "step": 8877
+    },
+    {
+      "epoch": 7.091108891108891,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.0002,
+      "loss": 0.8957,
+      "step": 8878
+    },
+    {
+      "epoch": 7.091908091908092,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8838,
+      "step": 8879
+    },
+    {
+      "epoch": 7.092707292707293,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.0002,
+      "loss": 0.882,
+      "step": 8880
+    },
+    {
+      "epoch": 7.093506493506493,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8878,
+      "step": 8881
+    },
+    {
+      "epoch": 7.094305694305694,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.0002,
+      "loss": 0.8895,
+      "step": 8882
+    },
+    {
+      "epoch": 7.095104895104895,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0002,
+      "loss": 0.8863,
+      "step": 8883
+    },
+    {
+      "epoch": 7.095904095904096,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8936,
+      "step": 8884
+    },
+    {
+      "epoch": 7.096703296703296,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8852,
+      "step": 8885
+    },
+    {
+      "epoch": 7.097502497502497,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.0002,
+      "loss": 0.8932,
+      "step": 8886
+    },
+    {
+      "epoch": 7.098301698301698,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8924,
+      "step": 8887
+    },
+    {
+      "epoch": 7.0991008991008995,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.0002,
+      "loss": 0.8897,
+      "step": 8888
+    },
+    {
+      "epoch": 7.0999000999001,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8895,
+      "step": 8889
+    },
+    {
+      "epoch": 7.100699300699301,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0002,
+      "loss": 0.8909,
+      "step": 8890
+    },
+    {
+      "epoch": 7.101498501498502,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0002,
+      "loss": 0.888,
+      "step": 8891
+    },
+    {
+      "epoch": 7.102297702297703,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0002,
+      "loss": 0.888,
+      "step": 8892
+    },
+    {
+      "epoch": 7.103096903096903,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8827,
+      "step": 8893
+    },
+    {
+      "epoch": 7.103896103896104,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0002,
+      "loss": 0.8853,
+      "step": 8894
+    },
+    {
+      "epoch": 7.104695304695305,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0002,
+      "loss": 0.8826,
+      "step": 8895
+    },
+    {
+      "epoch": 7.105494505494505,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8877,
+      "step": 8896
+    },
+    {
+      "epoch": 7.106293706293706,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0002,
+      "loss": 0.8846,
+      "step": 8897
+    },
+    {
+      "epoch": 7.107092907092907,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0002,
+      "loss": 0.887,
+      "step": 8898
+    },
+    {
+      "epoch": 7.107892107892108,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0002,
+      "loss": 0.8803,
+      "step": 8899
+    },
+    {
+      "epoch": 7.108691308691308,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8873,
+      "step": 8900
+    },
+    {
+      "epoch": 7.109490509490509,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8868,
+      "step": 8901
+    },
+    {
+      "epoch": 7.1102897102897105,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8871,
+      "step": 8902
+    },
+    {
+      "epoch": 7.1110889110889115,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0002,
+      "loss": 0.8866,
+      "step": 8903
+    },
+    {
+      "epoch": 7.111888111888112,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.0002,
+      "loss": 0.8877,
+      "step": 8904
+    },
+    {
+      "epoch": 7.112687312687313,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8842,
+      "step": 8905
+    },
+    {
+      "epoch": 7.113486513486514,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0002,
+      "loss": 0.8855,
+      "step": 8906
+    },
+    {
+      "epoch": 7.114285714285714,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0002,
+      "loss": 0.889,
+      "step": 8907
+    },
+    {
+      "epoch": 7.115084915084915,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8852,
+      "step": 8908
+    },
+    {
+      "epoch": 7.115884115884116,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0002,
+      "loss": 0.8907,
+      "step": 8909
+    },
+    {
+      "epoch": 7.116683316683317,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8885,
+      "step": 8910
+    },
+    {
+      "epoch": 7.117482517482517,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8865,
+      "step": 8911
+    },
+    {
+      "epoch": 7.118281718281718,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8849,
+      "step": 8912
+    },
+    {
+      "epoch": 7.119080919080919,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0002,
+      "loss": 0.9008,
+      "step": 8913
+    },
+    {
+      "epoch": 7.11988011988012,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8813,
+      "step": 8914
+    },
+    {
+      "epoch": 7.12067932067932,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0002,
+      "loss": 0.8834,
+      "step": 8915
+    },
+    {
+      "epoch": 7.1214785214785215,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8872,
+      "step": 8916
+    },
+    {
+      "epoch": 7.1222777222777225,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8894,
+      "step": 8917
+    },
+    {
+      "epoch": 7.123076923076923,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.0002,
+      "loss": 0.8882,
+      "step": 8918
+    },
+    {
+      "epoch": 7.123876123876124,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0002,
+      "loss": 0.8872,
+      "step": 8919
+    },
+    {
+      "epoch": 7.124675324675325,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.0002,
+      "loss": 0.8878,
+      "step": 8920
+    },
+    {
+      "epoch": 7.125474525474526,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8856,
+      "step": 8921
+    },
+    {
+      "epoch": 7.126273726273726,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.0002,
+      "loss": 0.8851,
+      "step": 8922
+    },
+    {
+      "epoch": 7.127072927072927,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.0002,
+      "loss": 0.8916,
+      "step": 8923
+    },
+    {
+      "epoch": 7.127872127872128,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0002,
+      "loss": 0.8928,
+      "step": 8924
+    },
+    {
+      "epoch": 7.128671328671329,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.0002,
+      "loss": 0.8882,
+      "step": 8925
+    },
+    {
+      "epoch": 7.129470529470529,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0002,
+      "loss": 0.8887,
+      "step": 8926
+    },
+    {
+      "epoch": 7.13026973026973,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8926,
+      "step": 8927
+    },
+    {
+      "epoch": 7.131068931068931,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0002,
+      "loss": 0.8899,
+      "step": 8928
+    },
+    {
+      "epoch": 7.131868131868132,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8832,
+      "step": 8929
+    },
+    {
+      "epoch": 7.132667332667332,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0002,
+      "loss": 0.8874,
+      "step": 8930
+    },
+    {
+      "epoch": 7.1334665334665335,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8887,
+      "step": 8931
+    },
+    {
+      "epoch": 7.1342657342657345,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0002,
+      "loss": 0.8878,
+      "step": 8932
+    },
+    {
+      "epoch": 7.135064935064935,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0002,
+      "loss": 0.8852,
+      "step": 8933
+    },
+    {
+      "epoch": 7.135864135864136,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0002,
+      "loss": 0.8841,
+      "step": 8934
+    },
+    {
+      "epoch": 7.136663336663337,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8867,
+      "step": 8935
+    },
+    {
+      "epoch": 7.137462537462538,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8883,
+      "step": 8936
+    },
+    {
+      "epoch": 7.138261738261738,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0002,
+      "loss": 0.8849,
+      "step": 8937
+    },
+    {
+      "epoch": 7.139060939060939,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0002,
+      "loss": 0.8927,
+      "step": 8938
+    },
+    {
+      "epoch": 7.13986013986014,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0002,
+      "loss": 0.8836,
+      "step": 8939
+    },
+    {
+      "epoch": 7.140659340659341,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0002,
+      "loss": 0.8737,
+      "step": 8940
+    },
+    {
+      "epoch": 7.141458541458541,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8902,
+      "step": 8941
+    },
+    {
+      "epoch": 7.142257742257742,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8846,
+      "step": 8942
+    },
+    {
+      "epoch": 7.143056943056943,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8885,
+      "step": 8943
+    },
+    {
+      "epoch": 7.143856143856144,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0002,
+      "loss": 0.8751,
+      "step": 8944
+    },
+    {
+      "epoch": 7.1446553446553445,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0002,
+      "loss": 0.885,
+      "step": 8945
+    },
+    {
+      "epoch": 7.1454545454545455,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0002,
+      "loss": 0.8846,
+      "step": 8946
+    },
+    {
+      "epoch": 7.1462537462537465,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0002,
+      "loss": 0.884,
+      "step": 8947
+    },
+    {
+      "epoch": 7.147052947052947,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0002,
+      "loss": 0.8883,
+      "step": 8948
+    },
+    {
+      "epoch": 7.147852147852148,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8818,
+      "step": 8949
+    },
+    {
+      "epoch": 7.148651348651349,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0002,
+      "loss": 0.8793,
+      "step": 8950
+    },
+    {
+      "epoch": 7.14945054945055,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.0002,
+      "loss": 0.8906,
+      "step": 8951
+    },
+    {
+      "epoch": 7.15024975024975,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0002,
+      "loss": 0.8824,
+      "step": 8952
+    },
+    {
+      "epoch": 7.151048951048951,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.0002,
+      "loss": 0.8836,
+      "step": 8953
+    },
+    {
+      "epoch": 7.151848151848152,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8946,
+      "step": 8954
+    },
+    {
+      "epoch": 7.152647352647353,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8959,
+      "step": 8955
+    },
+    {
+      "epoch": 7.153446553446553,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8874,
+      "step": 8956
+    },
+    {
+      "epoch": 7.154245754245754,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.0002,
+      "loss": 0.8828,
+      "step": 8957
+    },
+    {
+      "epoch": 7.155044955044955,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8853,
+      "step": 8958
+    },
+    {
+      "epoch": 7.1558441558441555,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.0002,
+      "loss": 0.8842,
+      "step": 8959
+    },
+    {
+      "epoch": 7.1566433566433565,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8959,
+      "step": 8960
+    },
+    {
+      "epoch": 7.1574425574425575,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.0002,
+      "loss": 0.8933,
+      "step": 8961
+    },
+    {
+      "epoch": 7.158241758241759,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8841,
+      "step": 8962
+    },
+    {
+      "epoch": 7.159040959040959,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0002,
+      "loss": 0.8872,
+      "step": 8963
+    },
+    {
+      "epoch": 7.15984015984016,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0002,
+      "loss": 0.8849,
+      "step": 8964
+    },
+    {
+      "epoch": 7.160639360639361,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0002,
+      "loss": 0.8887,
+      "step": 8965
+    },
+    {
+      "epoch": 7.161438561438562,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0002,
+      "loss": 0.8818,
+      "step": 8966
+    },
+    {
+      "epoch": 7.162237762237762,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0002,
+      "loss": 0.892,
+      "step": 8967
+    },
+    {
+      "epoch": 7.163036963036963,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0002,
+      "loss": 0.8866,
+      "step": 8968
+    },
+    {
+      "epoch": 7.163836163836164,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0002,
+      "loss": 0.876,
+      "step": 8969
+    },
+    {
+      "epoch": 7.164635364635364,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0002,
+      "loss": 0.8919,
+      "step": 8970
+    },
+    {
+      "epoch": 7.165434565434565,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8837,
+      "step": 8971
+    },
+    {
+      "epoch": 7.166233766233766,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0002,
+      "loss": 0.891,
+      "step": 8972
+    },
+    {
+      "epoch": 7.167032967032967,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0002,
+      "loss": 0.8851,
+      "step": 8973
+    },
+    {
+      "epoch": 7.1678321678321675,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0002,
+      "loss": 0.8796,
+      "step": 8974
+    },
+    {
+      "epoch": 7.1686313686313685,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0002,
+      "loss": 0.8831,
+      "step": 8975
+    },
+    {
+      "epoch": 7.1694305694305696,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8908,
+      "step": 8976
+    },
+    {
+      "epoch": 7.170229770229771,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8915,
+      "step": 8977
+    },
+    {
+      "epoch": 7.171028971028971,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0002,
+      "loss": 0.8953,
+      "step": 8978
+    },
+    {
+      "epoch": 7.171828171828172,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8826,
+      "step": 8979
+    },
+    {
+      "epoch": 7.172627372627373,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.0002,
+      "loss": 0.8797,
+      "step": 8980
+    },
+    {
+      "epoch": 7.173426573426573,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8802,
+      "step": 8981
+    },
+    {
+      "epoch": 7.174225774225774,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0002,
+      "loss": 0.8857,
+      "step": 8982
+    },
+    {
+      "epoch": 7.175024975024975,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0002,
+      "loss": 0.8908,
+      "step": 8983
+    },
+    {
+      "epoch": 7.175824175824176,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.0002,
+      "loss": 0.8914,
+      "step": 8984
+    },
+    {
+      "epoch": 7.176623376623376,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0002,
+      "loss": 0.8881,
+      "step": 8985
+    },
+    {
+      "epoch": 7.177422577422577,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.0002,
+      "loss": 0.8829,
+      "step": 8986
+    },
+    {
+      "epoch": 7.178221778221778,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0002,
+      "loss": 0.8859,
+      "step": 8987
+    },
+    {
+      "epoch": 7.179020979020979,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8816,
+      "step": 8988
+    },
+    {
+      "epoch": 7.1798201798201795,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8781,
+      "step": 8989
+    },
+    {
+      "epoch": 7.1806193806193805,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8959,
+      "step": 8990
+    },
+    {
+      "epoch": 7.181418581418582,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.0002,
+      "loss": 0.8908,
+      "step": 8991
+    },
+    {
+      "epoch": 7.182217782217783,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.0002,
+      "loss": 0.8753,
+      "step": 8992
+    },
+    {
+      "epoch": 7.183016983016983,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8835,
+      "step": 8993
+    },
+    {
+      "epoch": 7.183816183816184,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8771,
+      "step": 8994
+    },
+    {
+      "epoch": 7.184615384615385,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.0002,
+      "loss": 0.8874,
+      "step": 8995
+    },
+    {
+      "epoch": 7.185414585414585,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0002,
+      "loss": 0.886,
+      "step": 8996
+    },
+    {
+      "epoch": 7.186213786213786,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8881,
+      "step": 8997
+    },
+    {
+      "epoch": 7.187012987012987,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0002,
+      "loss": 0.8933,
+      "step": 8998
+    },
+    {
+      "epoch": 7.187812187812188,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.0002,
+      "loss": 0.8892,
+      "step": 8999
+    },
+    {
+      "epoch": 7.188611388611388,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0002,
+      "loss": 0.896,
+      "step": 9000
+    },
+    {
+      "epoch": 7.189410589410589,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.0002,
+      "loss": 0.8858,
+      "step": 9001
+    },
+    {
+      "epoch": 7.19020979020979,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0002,
+      "loss": 0.8915,
+      "step": 9002
+    },
+    {
+      "epoch": 7.191008991008991,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8855,
+      "step": 9003
+    },
+    {
+      "epoch": 7.1918081918081915,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0002,
+      "loss": 0.8837,
+      "step": 9004
+    },
+    {
+      "epoch": 7.192607392607393,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8866,
+      "step": 9005
+    },
+    {
+      "epoch": 7.193406593406594,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8877,
+      "step": 9006
+    },
+    {
+      "epoch": 7.194205794205795,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0002,
+      "loss": 0.886,
+      "step": 9007
+    },
+    {
+      "epoch": 7.195004995004995,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0002,
+      "loss": 0.8826,
+      "step": 9008
+    },
+    {
+      "epoch": 7.195804195804196,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0002,
+      "loss": 0.8818,
+      "step": 9009
+    },
+    {
+      "epoch": 7.196603396603397,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8841,
+      "step": 9010
+    },
+    {
+      "epoch": 7.197402597402597,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8926,
+      "step": 9011
+    },
+    {
+      "epoch": 7.198201798201798,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0002,
+      "loss": 0.8962,
+      "step": 9012
+    },
+    {
+      "epoch": 7.199000999000999,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0002,
+      "loss": 0.889,
+      "step": 9013
+    },
+    {
+      "epoch": 7.1998001998002,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0002,
+      "loss": 0.8928,
+      "step": 9014
+    },
+    {
+      "epoch": 7.2005994005994,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8838,
+      "step": 9015
+    },
+    {
+      "epoch": 7.201398601398601,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0002,
+      "loss": 0.8931,
+      "step": 9016
+    },
+    {
+      "epoch": 7.202197802197802,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0002,
+      "loss": 0.8867,
+      "step": 9017
+    },
+    {
+      "epoch": 7.202997002997003,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0002,
+      "loss": 0.8872,
+      "step": 9018
+    },
+    {
+      "epoch": 7.203796203796204,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0002,
+      "loss": 0.8869,
+      "step": 9019
+    },
+    {
+      "epoch": 7.204595404595405,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8861,
+      "step": 9020
+    },
+    {
+      "epoch": 7.205394605394606,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0002,
+      "loss": 0.8893,
+      "step": 9021
+    },
+    {
+      "epoch": 7.206193806193806,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8873,
+      "step": 9022
+    },
+    {
+      "epoch": 7.206993006993007,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0002,
+      "loss": 0.887,
+      "step": 9023
+    },
+    {
+      "epoch": 7.207792207792208,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8791,
+      "step": 9024
+    },
+    {
+      "epoch": 7.208591408591409,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.0002,
+      "loss": 0.8837,
+      "step": 9025
+    },
+    {
+      "epoch": 7.209390609390609,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0002,
+      "loss": 0.8858,
+      "step": 9026
+    },
+    {
+      "epoch": 7.21018981018981,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.0002,
+      "loss": 0.8872,
+      "step": 9027
+    },
+    {
+      "epoch": 7.210989010989011,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0002,
+      "loss": 0.9425,
+      "step": 9028
+    },
+    {
+      "epoch": 7.211788211788212,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.0002,
+      "loss": 0.8894,
+      "step": 9029
+    },
+    {
+      "epoch": 7.212587412587412,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8881,
+      "step": 9030
+    },
+    {
+      "epoch": 7.213386613386613,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8933,
+      "step": 9031
+    },
+    {
+      "epoch": 7.214185814185814,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.0002,
+      "loss": 0.8912,
+      "step": 9032
+    },
+    {
+      "epoch": 7.2149850149850145,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.0002,
+      "loss": 0.8841,
+      "step": 9033
+    },
+    {
+      "epoch": 7.215784215784216,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0002,
+      "loss": 0.886,
+      "step": 9034
+    },
+    {
+      "epoch": 7.216583416583417,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8938,
+      "step": 9035
+    },
+    {
+      "epoch": 7.217382617382618,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8786,
+      "step": 9036
+    },
+    {
+      "epoch": 7.218181818181818,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.0002,
+      "loss": 0.8838,
+      "step": 9037
+    },
+    {
+      "epoch": 7.218981018981019,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0002,
+      "loss": 0.8928,
+      "step": 9038
+    },
+    {
+      "epoch": 7.21978021978022,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8855,
+      "step": 9039
+    },
+    {
+      "epoch": 7.220579420579421,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8833,
+      "step": 9040
+    },
+    {
+      "epoch": 7.221378621378621,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8851,
+      "step": 9041
+    },
+    {
+      "epoch": 7.222177822177822,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0002,
+      "loss": 0.8818,
+      "step": 9042
+    },
+    {
+      "epoch": 7.222977022977023,
+      "grad_norm": 0.625,
+      "learning_rate": 0.0002,
+      "loss": 0.8851,
+      "step": 9043
+    },
+    {
+      "epoch": 7.223776223776224,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8798,
+      "step": 9044
+    },
+    {
+      "epoch": 7.224575424575424,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.0002,
+      "loss": 0.8903,
+      "step": 9045
+    },
+    {
+      "epoch": 7.225374625374625,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8865,
+      "step": 9046
+    },
+    {
+      "epoch": 7.226173826173826,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8752,
+      "step": 9047
+    },
+    {
+      "epoch": 7.226973026973027,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.0002,
+      "loss": 0.8887,
+      "step": 9048
+    },
+    {
+      "epoch": 7.227772227772228,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0002,
+      "loss": 0.882,
+      "step": 9049
+    },
+    {
+      "epoch": 7.228571428571429,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0002,
+      "loss": 0.8743,
+      "step": 9050
+    },
+    {
+      "epoch": 7.22937062937063,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0002,
+      "loss": 0.8798,
+      "step": 9051
+    },
+    {
+      "epoch": 7.23016983016983,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8724,
+      "step": 9052
+    },
+    {
+      "epoch": 7.230969030969031,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0002,
+      "loss": 0.8824,
+      "step": 9053
+    },
+    {
+      "epoch": 7.231768231768232,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8828,
+      "step": 9054
+    },
+    {
+      "epoch": 7.232567432567433,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.0002,
+      "loss": 0.8882,
+      "step": 9055
+    },
+    {
+      "epoch": 7.233366633366633,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0002,
+      "loss": 0.883,
+      "step": 9056
+    },
+    {
+      "epoch": 7.234165834165834,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.0002,
+      "loss": 0.8873,
+      "step": 9057
+    },
+    {
+      "epoch": 7.234965034965035,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0002,
+      "loss": 0.8875,
+      "step": 9058
+    },
+    {
+      "epoch": 7.235764235764236,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.0002,
+      "loss": 0.885,
+      "step": 9059
+    },
+    {
+      "epoch": 7.236563436563436,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0002,
+      "loss": 0.8823,
+      "step": 9060
+    },
+    {
+      "epoch": 7.237362637362637,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.0002,
+      "loss": 0.8826,
+      "step": 9061
+    },
+    {
+      "epoch": 7.2381618381618384,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0002,
+      "loss": 0.8956,
+      "step": 9062
+    },
+    {
+      "epoch": 7.238961038961039,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.0002,
+      "loss": 0.8842,
+      "step": 9063
+    },
+    {
+      "epoch": 7.23976023976024,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0002,
+      "loss": 0.8783,
+      "step": 9064
+    },
+    {
+      "epoch": 7.240559440559441,
+      "grad_norm": 0.81640625,
+      "learning_rate": 0.0002,
+      "loss": 0.9383,
+      "step": 9065
+    },
+    {
+      "epoch": 7.241358641358642,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0002,
+      "loss": 0.8843,
+      "step": 9066
+    },
+    {
+      "epoch": 7.242157842157842,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.0002,
+      "loss": 0.8893,
+      "step": 9067
+    },
+    {
+      "epoch": 7.242957042957043,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8859,
+      "step": 9068
+    },
+    {
+      "epoch": 7.243756243756244,
+      "grad_norm": 0.83203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8982,
+      "step": 9069
+    },
+    {
+      "epoch": 7.244555444555445,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8898,
+      "step": 9070
+    },
+    {
+      "epoch": 7.245354645354645,
+      "grad_norm": 0.80859375,
+      "learning_rate": 0.0002,
+      "loss": 0.8963,
+      "step": 9071
+    },
+    {
+      "epoch": 7.246153846153846,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0002,
+      "loss": 0.8891,
+      "step": 9072
+    },
+    {
+      "epoch": 7.246953046953047,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.0002,
+      "loss": 0.8915,
+      "step": 9073
+    },
+    {
+      "epoch": 7.247752247752247,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8833,
+      "step": 9074
+    },
+    {
+      "epoch": 7.248551448551448,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8832,
+      "step": 9075
+    },
+    {
+      "epoch": 7.249350649350649,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0002,
+      "loss": 0.8936,
+      "step": 9076
+    },
+    {
+      "epoch": 7.2501498501498505,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8912,
+      "step": 9077
+    },
+    {
+      "epoch": 7.250949050949051,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8849,
+      "step": 9078
+    },
+    {
+      "epoch": 7.251748251748252,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8859,
+      "step": 9079
+    },
+    {
+      "epoch": 7.252547452547453,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0002,
+      "loss": 0.8819,
+      "step": 9080
+    },
+    {
+      "epoch": 7.253346653346654,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8842,
+      "step": 9081
+    },
+    {
+      "epoch": 7.254145854145854,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0002,
+      "loss": 0.8838,
+      "step": 9082
+    },
+    {
+      "epoch": 7.254945054945055,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8837,
+      "step": 9083
+    },
+    {
+      "epoch": 7.255744255744256,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0002,
+      "loss": 0.894,
+      "step": 9084
+    },
+    {
+      "epoch": 7.256543456543456,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8835,
+      "step": 9085
+    },
+    {
+      "epoch": 7.257342657342657,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8818,
+      "step": 9086
+    },
+    {
+      "epoch": 7.258141858141858,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.0002,
+      "loss": 0.8831,
+      "step": 9087
+    },
+    {
+      "epoch": 7.258941058941059,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0002,
+      "loss": 0.8757,
+      "step": 9088
+    },
+    {
+      "epoch": 7.259740259740259,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0002,
+      "loss": 0.8839,
+      "step": 9089
+    },
+    {
+      "epoch": 7.26053946053946,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8822,
+      "step": 9090
+    },
+    {
+      "epoch": 7.2613386613386615,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.0002,
+      "loss": 0.897,
+      "step": 9091
+    },
+    {
+      "epoch": 7.2621378621378625,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8788,
+      "step": 9092
+    },
+    {
+      "epoch": 7.262937062937063,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.0002,
+      "loss": 0.9082,
+      "step": 9093
+    },
+    {
+      "epoch": 7.263736263736264,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8917,
+      "step": 9094
+    },
+    {
+      "epoch": 7.264535464535465,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0002,
+      "loss": 0.8867,
+      "step": 9095
+    },
+    {
+      "epoch": 7.265334665334665,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0002,
+      "loss": 0.8884,
+      "step": 9096
+    },
+    {
+      "epoch": 7.266133866133866,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0002,
+      "loss": 0.8933,
+      "step": 9097
+    },
+    {
+      "epoch": 7.266933066933067,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8951,
+      "step": 9098
+    },
+    {
+      "epoch": 7.267732267732268,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8885,
+      "step": 9099
+    },
+    {
+      "epoch": 7.268531468531468,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8856,
+      "step": 9100
+    },
+    {
+      "epoch": 7.269330669330669,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8858,
+      "step": 9101
+    },
+    {
+      "epoch": 7.27012987012987,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8827,
+      "step": 9102
+    },
+    {
+      "epoch": 7.270929070929071,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0002,
+      "loss": 0.8776,
+      "step": 9103
+    },
+    {
+      "epoch": 7.271728271728271,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8943,
+      "step": 9104
+    },
+    {
+      "epoch": 7.2725274725274724,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0002,
+      "loss": 0.8913,
+      "step": 9105
+    },
+    {
+      "epoch": 7.2733266733266735,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0002,
+      "loss": 0.8795,
+      "step": 9106
+    },
+    {
+      "epoch": 7.2741258741258745,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.0002,
+      "loss": 0.8763,
+      "step": 9107
+    },
+    {
+      "epoch": 7.274925074925075,
+      "grad_norm": 0.375,
+      "learning_rate": 0.0002,
+      "loss": 0.8868,
+      "step": 9108
+    },
+    {
+      "epoch": 7.275724275724276,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8897,
+      "step": 9109
+    },
+    {
+      "epoch": 7.276523476523477,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0002,
+      "loss": 0.8748,
+      "step": 9110
+    },
+    {
+      "epoch": 7.277322677322678,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8912,
+      "step": 9111
+    },
+    {
+      "epoch": 7.278121878121878,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0002,
+      "loss": 0.8767,
+      "step": 9112
+    },
+    {
+      "epoch": 7.278921078921079,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0002,
+      "loss": 0.8799,
+      "step": 9113
+    },
+    {
+      "epoch": 7.27972027972028,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.0002,
+      "loss": 0.8916,
+      "step": 9114
+    },
+    {
+      "epoch": 7.28051948051948,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0002,
+      "loss": 0.897,
+      "step": 9115
+    },
+    {
+      "epoch": 7.281318681318681,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.0002,
+      "loss": 0.881,
+      "step": 9116
+    },
+    {
+      "epoch": 7.282117882117882,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0002,
+      "loss": 0.8831,
+      "step": 9117
+    },
+    {
+      "epoch": 7.282917082917083,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8915,
+      "step": 9118
+    },
+    {
+      "epoch": 7.283716283716283,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8916,
+      "step": 9119
+    },
+    {
+      "epoch": 7.2845154845154845,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.0002,
+      "loss": 0.8796,
+      "step": 9120
+    },
+    {
+      "epoch": 7.2853146853146855,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0002,
+      "loss": 0.8903,
+      "step": 9121
+    },
+    {
+      "epoch": 7.2861138861138866,
+      "grad_norm": 0.375,
+      "learning_rate": 0.0002,
+      "loss": 0.8792,
+      "step": 9122
+    },
+    {
+      "epoch": 7.286913086913087,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.0002,
+      "loss": 0.8838,
+      "step": 9123
+    },
+    {
+      "epoch": 7.287712287712288,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0002,
+      "loss": 0.8879,
+      "step": 9124
+    },
+    {
+      "epoch": 7.288511488511489,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8815,
+      "step": 9125
+    },
+    {
+      "epoch": 7.289310689310689,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0002,
+      "loss": 0.8833,
+      "step": 9126
+    },
+    {
+      "epoch": 7.29010989010989,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0002,
+      "loss": 0.8765,
+      "step": 9127
+    },
+    {
+      "epoch": 7.290909090909091,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0002,
+      "loss": 0.886,
+      "step": 9128
+    },
+    {
+      "epoch": 7.291708291708292,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0002,
+      "loss": 0.8939,
+      "step": 9129
+    },
+    {
+      "epoch": 7.292507492507492,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0002,
+      "loss": 0.8825,
+      "step": 9130
+    },
+    {
+      "epoch": 7.293306693306693,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8923,
+      "step": 9131
+    },
+    {
+      "epoch": 7.294105894105894,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8803,
+      "step": 9132
+    },
+    {
+      "epoch": 7.294905094905095,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0002,
+      "loss": 0.883,
+      "step": 9133
+    },
+    {
+      "epoch": 7.2957042957042955,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.0002,
+      "loss": 0.9017,
+      "step": 9134
+    },
+    {
+      "epoch": 7.2965034965034965,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8909,
+      "step": 9135
+    },
+    {
+      "epoch": 7.2973026973026975,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8828,
+      "step": 9136
+    },
+    {
+      "epoch": 7.298101898101898,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0002,
+      "loss": 0.8873,
+      "step": 9137
+    },
+    {
+      "epoch": 7.298901098901099,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0002,
+      "loss": 0.8787,
+      "step": 9138
+    },
+    {
+      "epoch": 7.2997002997003,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.0002,
+      "loss": 0.8897,
+      "step": 9139
+    },
+    {
+      "epoch": 7.300499500499501,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8928,
+      "step": 9140
+    },
+    {
+      "epoch": 7.301298701298701,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0002,
+      "loss": 0.9,
+      "step": 9141
+    },
+    {
+      "epoch": 7.302097902097902,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8816,
+      "step": 9142
+    },
+    {
+      "epoch": 7.302897102897103,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.0002,
+      "loss": 0.8821,
+      "step": 9143
+    },
+    {
+      "epoch": 7.303696303696304,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8854,
+      "step": 9144
+    },
+    {
+      "epoch": 7.304495504495504,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0002,
+      "loss": 0.8869,
+      "step": 9145
+    },
+    {
+      "epoch": 7.305294705294705,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0002,
+      "loss": 0.8983,
+      "step": 9146
+    },
+    {
+      "epoch": 7.306093906093906,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8895,
+      "step": 9147
+    },
+    {
+      "epoch": 7.3068931068931064,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8853,
+      "step": 9148
+    },
+    {
+      "epoch": 7.3076923076923075,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0002,
+      "loss": 0.8885,
+      "step": 9149
+    },
+    {
+      "epoch": 7.3084915084915085,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8859,
+      "step": 9150
+    },
+    {
+      "epoch": 7.30929070929071,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0002,
+      "loss": 0.893,
+      "step": 9151
+    },
+    {
+      "epoch": 7.31008991008991,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8805,
+      "step": 9152
+    },
+    {
+      "epoch": 7.310889110889111,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0002,
+      "loss": 0.887,
+      "step": 9153
+    },
+    {
+      "epoch": 7.311688311688312,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8887,
+      "step": 9154
+    },
+    {
+      "epoch": 7.312487512487513,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8904,
+      "step": 9155
+    },
+    {
+      "epoch": 7.313286713286713,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0002,
+      "loss": 0.882,
+      "step": 9156
+    },
+    {
+      "epoch": 7.314085914085914,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8959,
+      "step": 9157
+    },
+    {
+      "epoch": 7.314885114885115,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0002,
+      "loss": 0.897,
+      "step": 9158
+    },
+    {
+      "epoch": 7.315684315684316,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8852,
+      "step": 9159
+    },
+    {
+      "epoch": 7.316483516483516,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9007,
+      "step": 9160
+    },
+    {
+      "epoch": 7.317282717282717,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0002,
+      "loss": 0.8775,
+      "step": 9161
+    },
+    {
+      "epoch": 7.318081918081918,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8955,
+      "step": 9162
+    },
+    {
+      "epoch": 7.3188811188811185,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0002,
+      "loss": 0.8845,
+      "step": 9163
+    },
+    {
+      "epoch": 7.3196803196803195,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8845,
+      "step": 9164
+    },
+    {
+      "epoch": 7.3204795204795206,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.0002,
+      "loss": 0.8886,
+      "step": 9165
+    },
+    {
+      "epoch": 7.321278721278722,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8826,
+      "step": 9166
+    },
+    {
+      "epoch": 7.322077922077922,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8816,
+      "step": 9167
+    },
+    {
+      "epoch": 7.322877122877123,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8832,
+      "step": 9168
+    },
+    {
+      "epoch": 7.323676323676324,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0002,
+      "loss": 0.8839,
+      "step": 9169
+    },
+    {
+      "epoch": 7.324475524475525,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8891,
+      "step": 9170
+    },
+    {
+      "epoch": 7.325274725274725,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0002,
+      "loss": 0.8818,
+      "step": 9171
+    },
+    {
+      "epoch": 7.326073926073926,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0002,
+      "loss": 0.8869,
+      "step": 9172
+    },
+    {
+      "epoch": 7.326873126873127,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0002,
+      "loss": 0.8885,
+      "step": 9173
+    },
+    {
+      "epoch": 7.327672327672328,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0002,
+      "loss": 0.888,
+      "step": 9174
+    },
+    {
+      "epoch": 7.328471528471528,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0002,
+      "loss": 0.8844,
+      "step": 9175
+    },
+    {
+      "epoch": 7.329270729270729,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0002,
+      "loss": 0.893,
+      "step": 9176
+    },
+    {
+      "epoch": 7.33006993006993,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002,
+      "loss": 0.8921,
+      "step": 9177
+    },
+    {
+      "epoch": 7.3308691308691305,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8759,
+      "step": 9178
+    },
+    {
+      "epoch": 7.3316683316683315,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8917,
+      "step": 9179
+    },
+    {
+      "epoch": 7.332467532467533,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0002,
+      "loss": 0.8942,
+      "step": 9180
+    },
+    {
+      "epoch": 7.333266733266734,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.0002,
+      "loss": 0.8795,
+      "step": 9181
+    },
+    {
+      "epoch": 7.334065934065934,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8875,
+      "step": 9182
+    },
+    {
+      "epoch": 7.334865134865135,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0002,
+      "loss": 0.8861,
+      "step": 9183
+    },
+    {
+      "epoch": 7.335664335664336,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.0002,
+      "loss": 0.8765,
+      "step": 9184
+    },
+    {
+      "epoch": 7.336463536463537,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0002,
+      "loss": 0.8891,
+      "step": 9185
+    },
+    {
+      "epoch": 7.337262737262737,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.0002,
+      "loss": 0.889,
+      "step": 9186
+    },
+    {
+      "epoch": 7.338061938061938,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8833,
+      "step": 9187
+    },
+    {
+      "epoch": 7.338861138861139,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0002,
+      "loss": 0.8881,
+      "step": 9188
+    },
+    {
+      "epoch": 7.339660339660339,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.0002,
+      "loss": 0.8835,
+      "step": 9189
+    },
+    {
+      "epoch": 7.34045954045954,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8851,
+      "step": 9190
+    },
+    {
+      "epoch": 7.341258741258741,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0002,
+      "loss": 0.8779,
+      "step": 9191
+    },
+    {
+      "epoch": 7.342057942057942,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0002,
+      "loss": 0.8851,
+      "step": 9192
+    },
+    {
+      "epoch": 7.3428571428571425,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8831,
+      "step": 9193
+    },
+    {
+      "epoch": 7.343656343656344,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.0002,
+      "loss": 0.8816,
+      "step": 9194
+    },
+    {
+      "epoch": 7.344455544455545,
+      "grad_norm": 2.734375,
+      "learning_rate": 0.0002,
+      "loss": 0.9191,
+      "step": 9195
+    },
+    {
+      "epoch": 7.345254745254746,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.0002,
+      "loss": 0.8858,
+      "step": 9196
+    },
+    {
+      "epoch": 7.346053946053946,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8871,
+      "step": 9197
+    },
+    {
+      "epoch": 7.346853146853147,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0002,
+      "loss": 0.8962,
+      "step": 9198
+    },
+    {
+      "epoch": 7.347652347652348,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0002,
+      "loss": 0.882,
+      "step": 9199
+    },
+    {
+      "epoch": 7.348451548451548,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0002,
+      "loss": 0.8742,
+      "step": 9200
+    },
+    {
+      "epoch": 7.349250749250749,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0002,
+      "loss": 0.8926,
+      "step": 9201
+    },
+    {
+      "epoch": 7.35004995004995,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0002,
+      "loss": 0.8851,
+      "step": 9202
+    },
+    {
+      "epoch": 7.350849150849151,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0002,
+      "loss": 0.8849,
+      "step": 9203
+    },
+    {
+      "epoch": 7.351648351648351,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0002,
+      "loss": 0.8868,
+      "step": 9204
+    },
+    {
+      "epoch": 7.352447552447552,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8928,
+      "step": 9205
+    },
+    {
+      "epoch": 7.353246753246753,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0002,
+      "loss": 0.8851,
+      "step": 9206
+    },
+    {
+      "epoch": 7.354045954045954,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0002,
+      "loss": 0.8805,
+      "step": 9207
+    },
+    {
+      "epoch": 7.3548451548451546,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0002,
+      "loss": 0.8775,
+      "step": 9208
+    },
+    {
+      "epoch": 7.355644355644356,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8851,
+      "step": 9209
+    },
+    {
+      "epoch": 7.356443556443557,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0002,
+      "loss": 0.8899,
+      "step": 9210
+    },
+    {
+      "epoch": 7.357242757242757,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.0002,
+      "loss": 0.8984,
+      "step": 9211
+    },
+    {
+      "epoch": 7.358041958041958,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.0002,
+      "loss": 0.8832,
+      "step": 9212
+    },
+    {
+      "epoch": 7.358841158841159,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0002,
+      "loss": 0.8923,
+      "step": 9213
+    },
+    {
+      "epoch": 7.35964035964036,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0002,
+      "loss": 0.8856,
+      "step": 9214
+    },
+    {
+      "epoch": 7.36043956043956,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0002,
+      "loss": 0.8794,
+      "step": 9215
+    },
+    {
+      "epoch": 7.361238761238761,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0002,
+      "loss": 0.8943,
+      "step": 9216
+    },
+    {
+      "epoch": 7.362037962037962,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0002,
+      "loss": 0.8805,
+      "step": 9217
+    },
+    {
+      "epoch": 7.362837162837163,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0002,
+      "loss": 0.8931,
+      "step": 9218
+    },
+    {
+      "epoch": 7.363636363636363,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8918,
+      "step": 9219
+    },
+    {
+      "epoch": 7.364435564435564,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.0002,
+      "loss": 0.8981,
+      "step": 9220
+    },
+    {
+      "epoch": 7.365234765234765,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0002,
+      "loss": 0.8896,
+      "step": 9221
+    },
+    {
+      "epoch": 7.366033966033966,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0002,
+      "loss": 0.88,
+      "step": 9222
+    },
+    {
+      "epoch": 7.366833166833167,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.0002,
+      "loss": 0.8836,
+      "step": 9223
+    },
+    {
+      "epoch": 7.367632367632368,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8922,
+      "step": 9224
+    },
+    {
+      "epoch": 7.368431568431569,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0002,
+      "loss": 0.8848,
+      "step": 9225
+    },
+    {
+      "epoch": 7.36923076923077,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0002,
+      "loss": 0.8801,
+      "step": 9226
+    },
+    {
+      "epoch": 7.37002997002997,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0002,
+      "loss": 0.8906,
+      "step": 9227
+    },
+    {
+      "epoch": 7.370829170829171,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0002,
+      "loss": 0.8804,
+      "step": 9228
+    },
+    {
+      "epoch": 7.371628371628372,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.0002,
+      "loss": 0.875,
+      "step": 9229
+    },
+    {
+      "epoch": 7.372427572427572,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0002,
+      "loss": 0.8832,
+      "step": 9230
+    },
+    {
+      "epoch": 7.373226773226773,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0002,
+      "loss": 0.8874,
+      "step": 9231
+    },
+    {
+      "epoch": 7.374025974025974,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0002,
+      "loss": 0.8887,
+      "step": 9232
+    },
+    {
+      "epoch": 7.374825174825175,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0002,
+      "loss": 0.8817,
+      "step": 9233
+    },
+    {
+      "epoch": 7.375624375624375,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8882,
+      "step": 9234
+    },
+    {
+      "epoch": 7.376423576423576,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0002,
+      "loss": 0.8941,
+      "step": 9235
+    },
+    {
+      "epoch": 7.377222777222777,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0002,
+      "loss": 0.8804,
+      "step": 9236
+    },
+    {
+      "epoch": 7.3780219780219785,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0002,
+      "loss": 0.8921,
+      "step": 9237
+    },
+    {
+      "epoch": 7.378821178821179,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.0002,
+      "loss": 0.8946,
+      "step": 9238
+    },
+    {
+      "epoch": 7.37962037962038,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0002,
+      "loss": 0.8898,
+      "step": 9239
+    },
+    {
+      "epoch": 7.380419580419581,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.0002,
+      "loss": 0.8859,
+      "step": 9240
+    },
+    {
+      "epoch": 7.381218781218781,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0002,
+      "loss": 0.8824,
+      "step": 9241
+    },
+    {
+      "epoch": 7.382017982017982,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8891,
+      "step": 9242
+    },
+    {
+      "epoch": 7.382817182817183,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.0002,
+      "loss": 0.8907,
+      "step": 9243
+    },
+    {
+      "epoch": 7.383616383616384,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0002,
+      "loss": 0.8804,
+      "step": 9244
+    },
+    {
+      "epoch": 7.384415584415584,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0002,
+      "loss": 0.8856,
+      "step": 9245
+    },
+    {
+      "epoch": 7.385214785214785,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8947,
+      "step": 9246
+    },
+    {
+      "epoch": 7.386013986013986,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0002,
+      "loss": 0.8984,
+      "step": 9247
+    },
+    {
+      "epoch": 7.386813186813187,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0002,
+      "loss": 0.8868,
+      "step": 9248
+    },
+    {
+      "epoch": 7.387612387612387,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.0002,
+      "loss": 0.8922,
+      "step": 9249
+    },
+    {
+      "epoch": 7.388411588411588,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0002,
+      "loss": 0.8809,
+      "step": 9250
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 750600,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 600,
+  "save_steps": 250,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 4.0602973151585894e+18,
+  "train_batch_size": 64,
+  "trial_name": null,
+  "trial_params": null
+}